Preliminary Regulatory Impact Analysis

Reference 26 - FDA - PRIA - Docket No. FDA-2014-N-0053 - September 2020.pdf

Recordkeeping and Records Access Requirements for Food Facilities

Preliminary Regulatory Impact Analysis

OMB: 0910-0560

⚠️ Notice: This form may be outdated. More recent filings and information on OMB 0910-0560 can be found here:

Document [pdf]

Download: pdf | pdf

DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration

Requirements for Additional Traceability Records
for Certain Foods

Docket No. FDA-2014-N-0053

Preliminary Regulatory Impact Analysis
Initial Regulatory Flexibility Analysis
Unfunded Mandates Reform Act Analysis

Economics Staff
Office of Economics and Analysis
Office of Policy, Legislation, and International Affairs
Office of the Commissioner

Table of Contents
I. Introduction and Summary

A. Introduction

B. Summary of Costs and Benefits

C. Terminology

II. Preliminary Economic Analysis of Impacts

A. Background

B. Market Failure Potentially Relevant to Federal Regulatory Action

C. Purpose of the Proposed Rule

D. Baseline Conditions

1. Bioterrorism Act of 2002 and the 2004 BT Final Rule Recordkeeping
Requirements

2. Current Industry Practices

3. Coverage of the Rule

E. Benefits of the Proposed Rule

1. Potential Benefits from the Proposed Rule

2. Public Health Benefits from Averted Illnesses

3. Benefits from Avoiding Overly Broad Recalls Quantified for Selected FTL
Foods
60
4. Other Benefits
F. Costs of the Proposed Rule

68
70

1. Main Assumptions of Cost Analysis

2. Costs of Reading and Understanding the Rule

3. Costs of Capital Investment

4. Costs of Training in New Traceability Practices

5. Recordkeeping Requirements

6. Non-Quantified Costs

111

7. Summary of Costs

113

G. Distributional Effects

115

H. International Effects

115

I. Uncertainty and Sensitivity Analysis

119

J. Analysis of Regulatory Alternatives to the Proposed Rule

121

III. Initial Small Entity Analysis

125
2

A. Description and Number of Affected Small Entities

126

B. Description of the Potential Impacts of the Rule on Small Entities

131

C. Alternatives to Minimize the Burden on Small Entities

135

IV. Co-proposed Option 1

135

A. Summary

135

B. Benefits

137

C. Costs

137

D. International Effects

140

E. Regulatory Alternatives

140

F. Initial Small Entity Analysis

142

V. References

144

VI. Appendices

148

A. Food Traceability List (FTL)

148

B. Methodology Used to Estimate the Number of Illnesses

150

C. Outbreak Case Studies Used in Estimation of Public Health Benefits

154

I. Introduction and Summary
A. Introduction
We have examined the impacts of the proposed rule under Executive Order 12866,
Executive Order 13563, Executive Order 13771, the Regulatory Flexibility Act (5 U.S.C. 601612), and the Unfunded Mandates Reform Act of 1995 (Pub. L. 104-4). Executive Orders 12866
and 13563 direct us to assess all costs and benefits of available regulatory alternatives and, when
regulation is necessary, to select regulatory approaches that maximize net benefits (including
potential economic, environmental, public health and safety, and other advantages; distributive
impacts; and equity). Executive Order 13771 requires that the costs associated with significant
new regulations “shall, to the extent permitted by law, be offset by the elimination of existing
costs associated with at least two prior regulations.” This proposed rule is an economically
significant regulatory action as defined by Executive Order 12866.
The Regulatory Flexibility Act requires us to analyze regulatory options that would
minimize any significant impact of a rule on small entities. Because some small firms may incur
annualized costs that exceed one percent of their annual revenue, we find that the proposed rule
will have a significant economic impact on a substantial number of small entities.
The Unfunded Mandates Reform Act of 1995 (section 202(a)) requires us to prepare a
written statement, which includes an assessment of anticipated costs and benefits, before
proposing “any rule that includes any Federal mandate that may result in the expenditure by
State, local, and tribal governments, in the aggregate, or by the private sector, of $100,000,000 or
more (adjusted annually for inflation) in any one year.” The current threshold after adjustment
for inflation is $156 million, using the most current (2019) Implicit Price Deflator for the Gross

Domestic Product. This proposed rule would result in an expenditure in any year that meets or
exceeds this amount.
B. Summary of Costs and Benefits
This proposed rule, if finalized, would allow FDA and industry to more rapidly and
effectively trace food products that cause illnesses back through the food supply system to the
source and forward to determine recipients of the contaminated product. This rule would only
apply to foods we have designated for inclusion on the Food Traceability List. 1 By allowing
faster identification of contaminated foods and increasing rates of successful tracing
completions, the proposed rule may result in public health benefits if foodborne illnesses directly
related to those outbreaks are averted. This may also lead to more efficient use of FDA and
industry resources needed for outbreak investigations by potentially resulting in more precise
recalls and avoidance of overly broad market withdrawals and advisories for listed foods.
Benefits from this rule could be generated if the following two conditions hold: (1) a
foodborne outbreak occurs and (2) the traceability records required by this proposed rule help
FDA to quickly and accurately locate a commercially distributed violative product and ensure it
is removed from the market. The primary public health benefits of this rule are the value from
the reduction of the foodborne illnesses or deaths because records required by the proposed rule
are likely to reduce the time that a violative or contaminated food product is distributed in the
market.
Other non-health related benefits of this rule, if realized, would be from avoiding costs
associated with conducting overly broad recalls and market withdrawals that affect products that

1 The list of applicable foods may be updated by publication of a notice in the Federal Register following
consideration of comments on proposed changes. See Appendix A for the list as of this writing.

otherwise would not need to be withdrawn or recalled. Although recalls of rightly implicated
foods come with necessary costs, overly broad recalls that involve loosely related or unrelated
products can make overall recalls unnecessarily costly. The costs of a broad recall or market
withdrawal include lost revenues from unimplicated products, plus expenses associated with
notifying retailers and consumers, collection, shipping, disposal, inventory and legal costs. 2
There are no benefits from removing unimplicated products from the market. It is possible, but
not certain, that both of these categories of benefits separately or jointly could be experienced to
the extent quantified in this regulatory impact analysis. On the other hand, it is also possible, but
not certain, that a given instance of baseline contamination would lead to a very broad recall
(that could be narrowed by the proposed rule) or to illnesses (that could be avoided due to the
proposed rule) — but not both.
Additional benefits may include increased food supply system efficiencies, such as
improvements in supply chain management and inventory control; more expedient initiation and
completion of recalls; avoidance of costs due to unnecessary preventive actions by consumers;
and other food supply system efficiencies due to a standardized approach to traceability,
including an increase in transparency and trust and potential deterrence of fraud.
This proposed rule, if finalized, would impose compliance costs on covered entities by
increasing the number of records that are required for food products on the Food Traceability
List. Entities that manufacture, process, pack, or hold listed foods would incur costs to establish
and maintain traceability records. Some firms may also incur initial capital investment and

For example, in an undifferentiated product recall, a single firm’s investment in traceability may be ineffective
when competitors and partners have not instituted a traceability system. This is problematic because, for example, in
the event of an undifferentiated leafy greens outbreak, issuing a broad recall could be unavoidable, at least until the
implicated product is identified and removed from the market. In situations where the recalled products are insured,
targeted recalls will help prevent unnecessary recall of insured products which may have long term consequence to
retailers from increases in their insurance rates due to imprecise recalls.

training costs in systems that would enable them to establish, maintain, sort, and make available
upon our request their traceability records. Moreover, firms would incur one-time costs of
reading and understanding the rule. The information flows brought about by the proposed rule
may prompt new protective actions — for example, in farming, manufacturing or cooking
processes — that themselves would have costs. These potential costs have not been quantified
but their occurrence is likely to be correlated with the realization of health and longevity benefits
of this rule.
Tables 1a and 1b summarize the costs and the benefits of the proposed rule. Table 1a
shows RIA section IV’s estimates of the rule’s cost if proposed Option 1 of the co-proposal
regarding retail food establishments with 10 or fewer full-time equivalent employees (full
exemption from the proposed rule) were selected. At a seven percent discount rate, ten-year
annualized costs would range from approximately $34 million to $2.4 billion per year in 2018
dollars, with a primary estimate of $411 million per year. At a three percent discount rate,
annualized costs would range from approximately $33 million to $2.4 billion per year, with a
primary estimate of $400 million per year.
Table 1b shows estimates from RIA sections II.F and II.H of the rule’s cost under
proposed Option 2 of the co-proposal, which would exempt retail food establishments with 10 or
fewer fill-time equivalent employees from the requirement to provide FDA, under certain
circumstances, with an electronic sortable spreadsheet containing requested tracing information.
At a seven percent discount rate, annualized costs under Option 2 would range from
approximately $43 million to $3.2 billion per year in 2018 dollars, with a primary estimate of
$535 million per year. At a three percent discount rate, annualized costs would range from

approximately $42 million to $3.1 billion per year, with a primary estimate of $513 million per
year.
In RIA sections IV.B and II.E.2, we estimate public health benefits using several case
studies of outbreak tracebacks for four pathogens associated with illnesses caused by foods on
the Food Traceability List. These benefits have a tendency toward underestimation of the total
public health benefits because these four pathogens do not represent the total burden of all FTLassociated illnesses. 3 However, adjustments made for undiagnosed and unattributed illnesses
may have the opposite tendency of overstating both FTL-associated illnesses and benefits. We
calculate these monetized benefits from illnesses based on an estimated 84 percent reduction of
traceback time resulting from the requirements of this rule. Under Option 1 of the co-proposal,
for an estimated 84 percent traceback time improvement, the annualized monetized benefits
range from $33 million to $1.4 billion with a primary estimate of $567 million, discounted at
seven percent over ten years. At a three percent discount rate over ten years, the annualized
monetized benefits range from $33 million to $1.4 billion with a primary estimate of $580
million.
Under Option 2 of the co-proposal, for an estimated 84 percent traceback improvement,
the annualized monetized benefits range from $36 million to $1.5 billion with a primary estimate
of $626 million, discounted at a seven percent over ten years, and from $37 million to $1.5
billion with a primary estimate of $640 million, discounted at three percent over ten years. 4
Using examples from three recalls, RIA section II.E.3 presents estimates that additional (non-

3 We cannot scale up to 100% because our estimates of the percentage of illnesses potentially avoided with
improved traceability depend on data specific to each pathogen. We describe our methods in detail in section II.E.2
Public Health Benefits from Averted Illnesses. In short, these four pathogens may account for roughly 95% of the
total dollar value of the illnesses for which traceability might be an effective preventive measure.
4 These estimates reflect a wide uncertainty range because they were calculated using information from a small
(possibly under-representative) number of outbreaks.

health) benefits of avoiding overly broad recalls could range from $1.7 billion to $5.6 billion per
year at seven percent discount rate and from $1.7 billion to $5.8 billion using three percent
discount rate. As noted earlier, it is possible that both of these categories of benefits could be
experienced to the extent quantified in this regulatory impact analysis, either separately or
jointly. Therefore, Table 1a and Table 1b avoid a definitive statement that they should be
summed.
Costs are lower in Option 1, relative to Option 2, because fewer retail food
establishments (RFEs) would need to comply with the proposed rule. However, if RFEs with 10
or fewer full-time equivalent employees are exempt from Subpart S requirements, the timeliness,
precision, and accuracy of traceability efforts can be impacted and non-quantified benefits, such
as enhancement of our ability to narrow the number of lots in a recall and the ability of RFEs
with 10 or fewer full-time equivalent employees to have the data necessary to quickly identify
and remove contaminated products from shelves, will be lessened in comparison to Option 2
(Ref. [1]). Requiring recordkeeping by RFEs of all sizes allows for more consistent, organized,
and specific information that covers the entire supply chain.
Table 1a. Summary of Benefits, Costs and Distributional Effects of Proposed Rule (Option
1, in Millions of Dollars)
Category
Annualized
Monetized
$millions/year

Benefits

Primary
Estimate

Low
Estimate

High
Estimate

$567
$580

$33
$33

$1,355
$1,385

Year
Dollars
2018
2018

Units
Discount
Rate
7%
3%

Period
Covered
10 years
10 years

Notes
Monetized benefits from
an estimated 84%
improvement in
traceback time for four
pathogens. Additional
benefits of avoiding
overly broad recalls
could range from $1.7
billion to $5.6 billion
(7%, 10 years) and $1.7
billion to $5.8 billion
(3%, 10 years).

Category
Annualized
Quantified
Qualitative

Annualized
Monetized
$millions/year

Costs

Transfers

Effects

Primary
Estimate

Low
Estimate

High
Estimate

Year
Dollars

Additional potential benefits include
increased food supply system efficiencies;
more expedient initiation and completion of
recalls; avoidance of costs due to
unnecessary preventive actions; and other
efficiencies from a standardized approach to
traceability. However, if retail food
establishments with 10 or fewer full-time
equivalent employees are exempt from
Subpart S requirements, the timeliness,
precision, and accuracy of traceability
efforts can be impacted and qualitative
benefits such as the ability to narrow the
number of lots in a recall and the ability for
RFEs with 10 or fewer full-time equivalent
employees to have the data necessary to
quickly identify and remove contaminated
products from shelves will be lessened in
comparison to Option 2. 5
$411
$34
$2,425
2018
$400

$33

$2,352

2018

Units
Discount
Rate

Period
Covered

Notes

10 years

A portion of foreign
costs could be passed on
to domestic consumers.
We estimate that up to
$259 million in
annualized costs (7%,
10 years) to foreign
facilities could be
passed on to domestic
consumers.

Annualized
Quantified
Qualitative
Federal
Annualized
Monetized
$millions/year
From/ To
From:
To:
Other
Annualized
Monetized
$millions/year
From/To
From:
To:
State, Local or Tribal Government: No significant effect.
Small Business: Potential impact on some small entities that are currently not keeping traceability records
described by the proposed rule.
Wages: N/A
Growth: N/A

5 (Ref. [1]).

Table 1b. Summary of Benefits, Costs and Distributional Effects of Proposed Rule (Option
2, in Millions of Dollars)
Category
Annualized
Monetized
$millions/year

Primary
Estimate

Low
Estimate

High
Estimate

$626
$640

$36
$37

$1,497
$1,531

Year
Dollars
2018
2018

Units
Discount
Rate
7%
3%

Period
Covered
10 years
10 years

10 years

Benefits

Annualized
Quantified
Qualitative

Annualized
Monetized
$millions/year

$42

$3,063

2018

Costs

Transfers

Annualized
Quantified
Qualitative
Federal
Annualized
Monetized
$millions/year
From/ To

From:

Notes
Monetized benefits
from an estimated
84% improvement
in traceback time
for four pathogens.
Additional benefits
of avoiding overly
broad recalls could
range from $1.7
billion to $5.6
billion (7%, 10
years) and $1.7
billion to $5.8
billion (3%, 10
years).

A portion of foreign
costs could be
passed on to
domestic
consumers. We
estimate that up to
$259 million in
annualized costs
(7%, 10 years) to
foreign facilities
could be passed on
to domestic
consumers.

To:

Category

Effects

Primary
Estimate

Low
Estimate

High
Estimate

Year
Dollars

Units
Discount
Rate

Period
Covered

Notes

Other
Annualized
Monetized
$millions/year
From/To
From:
To:
State, Local or Tribal Government: No significant effect.
Small Business: Potential impact on small entities that are currently not keeping traceability records
described by the proposed rule.
Wages: N/A
Growth: N/A

In accordance with Executive Order 13771, in Tables 2a and 2b we estimate present and
annualized values of costs and cost savings of the proposed rule over an infinite time horizon.
This proposed rule is expected to be a regulatory action under Executive Order 13771.

Table 2a. EO 13771 Summary Table (Option 1, in Millions 2016 Dollars, Over an Infinite
Time Horizon)
Item
Primary Estimate Lower Estimate
Upper Estimate
(7%)
(7%)
(7%)
Present Value of Costs
Present Value of Cost Savings
Present Value of Net Costs
Annualized Costs
Annualized Cost Savings
Annualized Net Costs

$5,105
$$5,105
$357
$$357

$438
$$438
$31
$$31

$29,659
$$29,659
$2,076
$$2,076

Table 2b. EO 13771 Summary Table (Option 2, in Millions 2016 Dollars, Over an Infinite
Time Horizon)
Item
Primary Estimate Lower Estimate
Upper Estimate
(7%)
(7%)
(7%)
Present Value of Costs
Present Value of Cost Savings
Present Value of Net Costs
Annualized Costs
Annualized Cost Savings
Annualized Net Costs

$6,288
$$6,288
$440
$$440

$532
$$532
$37
$$37

$36,867
$$36,867
$2,581
$$2,581
12

We have also considered an alternative way of describing costs and benefits. Given
uncertainties in the data underlying our costs and benefits estimates, Tables 3a and 3b explore
the possibility that baseline costs of recalls are more fully internalized by market actors.
Column (a) of Tables 3a and 3b explores the possibility that market actors do not already
account for the costs of foodborne illnesses associated with listed foods (e.g. public health
benefits of products with better traceability are not captured in product price) and/or the costs of
overly broad recalls (e.g. firms do not invest enough in traceability because they do not expect
other firms to also invest). Primary estimates (and relatively large portions of the uncertainty
ranges) indicate that benefits of the rule would be greater than the rule’s cost. Column (b) of
Tables 3a and 3b considers scenarios where market actors already fully account for the costs of
overly broad recalls. Then recall-associated benefits would not be greater than the cost of the
rule. This means they have already invested in traceability to the point where further investment
would cost more than the benefit they would expect to receive. Then the total benefits of the
rule, including health benefits, may or may not be greater than the rule’s cost.

Table 3a. Summary of Benefits and Costs of Proposed Rule (Option 1), As a Function of
Assumptions Regarding Baseline Cost Internalization*

RIA Section IV.B

(a)
Neither adverse health effects nor
recall-associated costs fully
internalized in market transactions
for FTL foods

(b)

Health Benefits: $567M
(range: $33M to $1.4B)

Recall-associated costs, but not adverse health
effects, fully internalized in market
transactions for FTL foods

and/or

Recall-Associated Benefits:
$1.7B to $5.6B
RIA Section II.E.3

Recall-Associated Benefits: $1.7B
to $5.6B

Direct Compliance Costs >
$1.7B to $5.6B
Protective Action Costs (potential): not
quantified
or
Recall-Associated Benefits < Costs

Direct Compliance Costs (if foreign
passed through to U.S. supply chain
& consumers): $670M
(range: $52M to $4B)
RIA Sections
IV.C and IV.D

Direct Compliance Costs (if foreign
not passed through to U.S. supply
chain & consumers): $411M
(range: $34M to $2.4B)
Protective Action Costs (potential):
not quantified

Direct Compliance Costs (if foreign passed
through to U.S. supply chain & consumers):
$670M
(range: $52M to $4B)
Direct Compliance Costs (if foreign not
passed through to U.S. supply chain &
consumers): $411M
(range: $34M to $2.4B)
Protective Action Costs (potential): not
quantified

* Primary estimates presented in this table are calculated with a 7 percent discount rate; primary estimates
discounted at 3 percent differ only slightly. All estimates are expressed in 2018 dollars and annualized over 10
years. Abbreviations: M=million, B=billion.

Table 3b. Summary of Benefits and Costs of Proposed Rule (Option 2), As a Function of
Assumptions Regarding Baseline Cost Internalization*

RIA Section II.E.2

(a)
Neither adverse health effects nor
recall-associated costs fully
internalized in market transactions
for FTL foods

(b)

Health Benefits: $626M
(range: $36M to $1.5B)

Recall-associated costs, but not adverse health
effects, fully internalized in market
transactions for FTL foods

and/or

Recall-Associated Benefits:
$1.7B to $5.6B
RIA Section II.E.3

Recall-Associated Benefits: $1.7B
to $5.6B

Direct Compliance Costs >
$1.7B to $5.6B
Protective Action Costs (potential): not
quantified
or
Recall-Associated Benefits < Costs

Direct Compliance Costs (if foreign
passed through to U.S. supply chain
& consumers): $794M
(range: $61M to $4.8B)
RIA Sections
II.F and II.H

Direct Compliance Costs (if foreign
not passed through to U.S. supply
chain & consumers): $535M
(range: $43M to $3.2B)
Protective Action Costs (potential):
not quantified

Direct Compliance Costs (if foreign passed
through to U.S. supply chain & consumers):
$794M
(range: $61M to $4.8B)
Direct Compliance Costs (if foreign not
passed through to U.S. supply chain &
consumers): $535M
(range: $43M to $3.2B)
Protective Action Costs (potential): not
quantified

We request comment on our estimates of costs and benefits of this rule and on the extent
to which costs may already be internalized by covered entities.

C. Terminology
In Table 4, we describe the key terms we use in this document. We note that these
definitions only apply to this document.
Table 4. Key Terms in the Regulatory Impact Analysis
Term
Description
BT Act

Bioterrorism Act of 2002. We use Subpart J (of 21 CFR part 1)
and BT Act interchangeably.
15

BT rule

CDC
CTE
FD&C Act
FSMA
FTL
FTL foods
FTE
Lot, batch

KDE
NAICS
Persons, entities
UPC
USDA
We, our, us, FDA,
the Agency

Establishment and Maintenance of Records Under the Public
Health Security and Bioterrorism Preparedness and Response
Act of 2002 Act final rule (2004)
The Centers for Disease Control and Prevention
Critical tracking event
Federal Food, Drug, and Cosmetic Act
FDA Food Safety Modernization Act
Food Traceability List
Foods on the FTL
Full-time-equivalent employee
Food produced during a time period at a single physical location
and identified by a specific code. A lot is sometimes referred to
as a batch.
Key data element
North American Industry Classification System
We use these terms interchangeably to refer to businesses
covered by the proposed rule
Universal Product Code
The U.S. Department of Agriculture
We use these terms to refer to the U.S. Food and Drug
Administration.

II. Preliminary Economic Analysis of Impacts 6
A. Background
Current recordkeeping requirements that stem from the Bioterrorism Act (BT Act) of
2002 require firms to know and record the immediate previous source of their food products and
the immediate subsequent recipient (commonly referred to as one-up, one-back recordkeeping).
Since these requirements took effect, FDA has encountered significant limitations in the
available food tracing-related information upon which government agencies and industry rely for
rapid and effective tracing of food products in the event of an outbreak investigation. These
limitations arise from gaps in recordkeeping requirements, including: no requirement to collect

6 Sections II and III of this document discuss our analysis of policy Option 2. Please see section IV for our analysis
of Option 1.

specific key data elements for the type and quantity of food and ingredient source, a requirement
to maintain a record of the lot code or other unique identifier only if it exists, no requirement to
link incoming and outgoing product within a firm and from one point in the food supply chain to
the next, and address requirements that do not distinguish between corporate headquarters and
the physical location where the food was produced.
Inadequate traceability information and the challenge of having many point-of-service
firms (retail and foodservice) excluded from Subpart J requirements has hampered recalls of
potentially contaminated foods. In 2015, for example, an outbreak of Shiga toxin-producing
Escherichia coli (E. coli) O26 (STEC O26) resulted in 55 illnesses in 11 states, leading to 21
hospitalizations (Ref. [2]). Though an investigation conducted by the CDC, FDA, and the
USDA’s Food Safety and Inspection Service linked a specific restaurant chain to the outbreak as
early as October of 2015, investigators could not identify a particular ingredient or food item as
the likely source of contamination. The lack of information in the records maintained by the
restaurant caused an inability for regulatory officials to use traceability to narrow the ingredients
to further investigate which ingredients came from common sources. Additionally, available
consumer data could not identify a particular ingredient for tracing.
Inadequate traceability can also necessitate broad recalls that inadvertently affect noncontaminated product. In 2015, for example, FDA identified 36 farms as potentially having
produced leafy greens for a leafy greens mix linked to an E. coli outbreak. Without being able to
identify specific lots and growers of contaminated product, it was not possible to narrow
investigative efforts to the source of the outbreak which would have allowed the Agency to
narrow the scope of the recall (Ref. [3]).

On January 4, 2011, the FDA Food Safety Modernization Act (FSMA) (Public Law 111353) was signed into law. Section 204(d)(1) of FSMA requires FDA to establish recordkeeping
requirements for facilities that manufacture, process, pack, or hold foods that we designate as
high-risk foods. These recordkeeping requirements will be additional to the traceability
recordkeeping requirements in 21 CFR Part 1, Subpart J (the Subpart J requirements), which
were promulgated in accordance with the BT Act of 2002. Section 204(d)(2) of FSMA requires
the Agency to designate the foods for which these additional recordkeeping requirements are
appropriate and necessary to protect the public health, and to publish the list of such foods (the
FTL) on our web site when we finalize this proposed rule.
B. Market Failure Potentially Relevant to Federal Regulatory Action
Several types of market failure may impact current traceability efforts, creating a need for
regulatory action. First, firms currently utilize various traceability methods, creating
interoperability challenges, and costs of coordination can be prohibitively high. Second, the
return on each firm’s additional investment in traceability depends on the level and type of
investment made by other firms, potentially causing a disincentive for firms to invest. As a
result, the risk of FTL-associated foodborne illnesses is likely not fully priced into FTL products.
While current traceability systems may in part originate from requirements of the BT Act,
economic incentives, such as improved supply-side management and safety and quality control,
may have motivated some producers to develop traceability systems of varying sophistication
and comprehensiveness. Food producers in the U.S. use a variety of systems to trace the
movement of food in the supply system. Tracing systems vary by the type and amount of
information they collect and record, the record medium (e.g., paper vs. electronic), and the extent
of the supply chain covered (e.g., the immediate previous and next steps vs. the entire chain from
18

farm to retailer). In some instances, owners of large supply chains (e.g., major retailers, major
restaurant operators, brokers of different size that represent farms, food processors with many
ingredient suppliers, importers of seafood from many vessels) compete on supply chain
efficiency and consumer transparency, which requires traceability as a component of that
strategy. However, to maintain competitive advantage, most supply chain owners require their
suppliers share traceability data through private portals. This leads to a proliferation of different
portals and data standards, which reduces the potential for interoperability. A universal standard
for traceability would enable suppliers to insist their customers (and portals) accept, at minimum,
a standard list of CTEs and KDEs, which would lower the cost for suppliers among other
benefits.
The effectiveness of a tracing system depends on the accuracy, quality, uniformity and
extent of collected information. Firms generally have private incentives to avoid the deliberate or
accidental contamination of food linked to their products or facilities. Nevertheless, those
incentives may not be enough for all firms to provide the socially optimal amount of information
about their entire production and distribution network. Because firms’ revenues may not capture
all of the benefits that accrue to the public from improved food traceability, firms may collect
and supply less information than would be socially optimal for adequate protection of public
health. 7
FDA has experienced the significant limitations in the available tracing-related
information on which government agencies and industry currently rely to conduct tracing
operations. Industry often does not fully understand what data the FDA need to effectively

The socially optimal level of traceability considers all private costs and benefits (those faced by firms) as well as
public costs and benefits (those faced by everyone other than firms). In other words, the socially optimal level of
traceability maximizes the aggregate welfare of society, which includes firms and non-firms (e.g., consumers).

investigate foodborne illness outbreaks. Further, while standard production and distribution
records carry a lot of useful information, they do not necessarily capture the complete set of
information, in any standard format, that FDA would need to efficiently investigate a
contamination of unknown origin. The result is that many of the systems and approaches that
firms currently use for voluntary traceability are not interoperable, which results in potentially
avoidable costs for all entities in the food supply chain. This failure of interoperability also slows
outbreak investigations, sickens more consumers, and reduces trust in the U.S. food supply.
Although some supply chain owners have rapidly adopted traceability technology,
recordkeeping practices lack uniformity across supply chains. Different supply chain entities
such as growers, shippers, distributors, retailers and restaurants lack incentives to standardize
recordkeeping in the form of common key data elements. Without uniform recordkeeping
standards, competing software developers promote mutually exclusive, proprietary frameworks,
whose incompatibility increases traceability costs. The current lack of system interoperability
impedes collaboration in identifying sources or recipients of potentially contaminated food.
While the effectiveness of each traceability system increases with the number of participants
throughout the supply chain, the lack of standardization in recordkeeping among systems causes
duplication of efforts. In addition, high transaction costs and costs of coordination in setting up a
complete farm to retail national tracing system may even disincentivize some firms from
investing in traceability systems, particularly those firms that are not vertically integrated. This
proposed rule would standardize the key data elements and critical tracking events, would
significantly reduce the private coordination and transaction costs of setting up a complete
tracking system, and would enable FDA and other entities involved in a tracing investigation to
accelerate and enhance the acquisition of robust product tracing information.

Underscoring the need for standardized data elements, food trade associations,
technology providers, consumer advocacy groups, standards bodies, multi-unit restaurant
operators, retailers, distributors and food producers have asked FDA to describe the types of data
we need, and the format in which we prefer to receive such data, during an outbreak
investigation. 8 This information would enable companies and solution providers to develop
systems and procedures to efficiently collect that data, so it can be shared with the FDA when
needed. Ultimately this might lower traceability costs for most members of the food supply
chain because it would encourage the development of interoperable traceability systems.
From public comments 9 received as part of FDA’s New Era for Smarter Food Safety
Public Meeting held on October 21, 2019, one large food industry trade association representing
food companies from around the world commented that one of the most foundational and
significant actions FDA could take is identifying the key data elements that should be
communicated throughout the global supply chain. Similar comments echoed the need to
establish a common set of key data elements and to have clarification from FDA on the key data
elements needed to provide rapid identification.
In addition to standardized key data elements and critical tracking events, the
effectiveness of a tracing system depends on the extent to which firms throughout the supply
chain participate. Unfortunately, even a small number of breaks in tracing information through
the supply chain can prevent the FDA and others from being able to trace contaminated products
to their source. Full supply chain traceability requires policy intervention as some firms do not

8 https://www.regulations.gov/searchResults?rpp=25&po=0&s=FDA-2019-N-4187&fp=true&ns=true
9 This section references public comments from the New Era for Smarter Food Safety Public Meeting - Docket ID:
FDA-2019-N-4187(https://www.regulations.gov/searchResults?rpp=25&po=0&s=FDA-2019-N4187&fp=true&ns=true), including comments by the Grocery Manufacturers Association, the US Apple
Association, National Fisheries Institute, United Fresh, Produce Manufacturers Association, among others.

have an immediate financial incentive to institute tracing systems (Ref. [4]). For example, the
Institute of Food Technologists (IFT) noted in the Product Tracing Pilots report (see page 217,
subtitled “Lack of Standards Results in Fragmented Requirements”) (Ref. [5]) that traceability is
likely to stay in a state of perpetual flux until FDA clearly defines the data requirements and
establishes a framework for full supply chain traceability. IFT found that producers were
reluctant to invest in tracing systems if their fellow producers were not similarly investing, since
tracing is not an isolated exercise.
Further comments received as part of FDA’s New Era for Smarter Food Safety Public
Meeting indicated that there are inconsistencies among suppliers and buyers in terms of the level
of capability for traceability and that food supply chain companies cannot control the
recordkeeping by entities that repackage product further up the supply chain. One comment
from a large trade association indicated that “off the record” conversations with their broad
membership indicated consensus that the time of hoping for voluntary adoption of effective
traceability systems has passed.
In sum, market prices convey most of the necessary information for the ordinary
production and distribution of foods, including the foods on the FTL. 10 However, an actual or
suspected contamination of unknown origin requires more complete and standardized
information as well as the ability to rapidly access and consolidate that information. In order to
protect consumers from further exposure and to find the source and cause of contamination, FDA
must be able to trace food backward and forward through the supply chain. Although the nation’s
food processors, distributors, retail food establishments, importers, and others may benefit from

Prices provide most information about goods and services without the need for buyers and sellers to know much
about each other. However, prices do not always communicate the difference between contaminated versus not
contaminated product in the market, which explains the potential need for government intervention.

such a system, the private costs of creating it would be prohibitively expensive for any single
firm or third-party organization.
As discussed in section C, the proposed rule, if finalized, would address the limitations of
current traceability systems by requiring a rigorous and consistent approach to food tracing
across different industry sectors for more efficient traceability of foods on the FTL. The
proposed rule, if finalized, would enable FDA, its regulatory partners, and industry to better
identify and remove contaminated FTL foods from the marketplace in the case of an outbreak, as
well as to develop mitigation strategies to prevent future contamination.
We request comment on evidence that may support or refute the market failure claims
above, or that may be used to quantify the scope of incomplete internalization of relevant
baseline costs.
C. Purpose of the Proposed Rule
The purpose of this rule is to ensure that contaminated FTL foods covered by this rule
can be swiftly identified and removed from the market to prevent or mitigate a foodborne illness
outbreak. In order to improve FDA’s ability to follow the movement of FTL foods through the
supply chain, the proposed rule, if finalized, would establish traceability recordkeeping
requirements for persons who manufacture, process, pack, or hold FTL foods. Namely, the rule
specifies the data elements and information firms must establish and maintain, along with
information they must send, in certain circumstances, to the next entity in the supply chain. The
core requirements are to establish and maintain records of key data elements (KDEs) associated
with different critical tracking events (CTEs) in a listed food’s supply chain, including the
growing, receiving, transformation, creating, and shipping of the FTL food. Required
information also includes a description of the reference records in which firms keep required
23

tracing information, a list of foods on the FTL they ship, a description of how they assign
traceability lot codes, and other information needed to understand their traceability programs.
The proposed rule, if finalized, would also provide consistent food tracing terminology,
encourage a transition from paper-based recordkeeping to electronic records, and promote a
broader understanding of the data elements needed for efficient traceability and product recall.
This proposed rule would enable FDA and industry to identify the source of an outbreak
or other contamination event, expedite removal of contaminated food from the marketplace, and
prevent additional consumer exposures, as well as develop mitigation strategies to prevent future
contamination. If finalized, this proposed rule would further help the Agency deter and limit the
effects of foodborne outbreaks from FTL foods and thereby improve the safety of the food
supply in the United States.

D. Baseline Conditions
1. Bioterrorism Act of 2002 and the 2004 BT Final Rule Recordkeeping Requirements
We consider the current state of the world as a reasonable approximation of the baseline
(the projected future without the proposed rule) against which to measure the costs and benefits
of regulatory options. 11 Before the enactment of FSMA, FDA implemented recordkeeping
requirements (Subpart J) related to product tracing under authority of the BT Act of 2002. Thus,
the current estimated baseline includes the costs and benefits of the pre-FSMA Establishment
and Maintenance of Records Under the Public Health Security and Bioterrorism Preparedness

However, we note that data available prior to finalization of the rule may show substantial changes relative to the
present — due, for example, to other FSMA regulations increasingly taking effect and to societal changes associated
with the COVID-19 pandemic. We request comment on estimating the baseline trajectory, given the dynamic nature
of the regulatory environment.

and Response Act of 2002 final rule issued in 2004, as estimated in the economic impact analysis
for that rule and as further modified by updated assumptions discussed below. 12
The number of entities covered by Subpart J (which was promulgated pursuant to the BT
Act) was estimated to be 707,672 (including persons who manufacture, process, pack, transport,
distribute, receive, hold, or import food in the U.S.). Records redesign costs were assumed to be
incurred by approximately 101,153 large and small firms two years following issuance of the BT
final rule and by 222,316 very small firms three years following issuance of the BT final rule.
Several types of entities that participate in the food supply system, however, were not covered by
Subpart J. We list and discuss these entities in the next subsection.
The 2004 economic impact analysis of the BT final rule estimated annual and first-year
costs of requiring establishment and maintenance of records to trace the transportation of all food
to both foreign and domestic entities, as well as costs for future entities entering the market each
year. 13 Benefits of the 2004 BT final rule were estimated as the number of averted illnesses due
to improved recordkeeping practices. Nevertheless, in more than ten years since implementation
of these recordkeeping requirements, FDA has learned that there are critical gaps in the
requirements that limit the ability of regulatory agencies to conduct prompt, effective product
tracing, especially in response to foodborne illness outbreaks. These critical gaps, which are
discussed in section III.C of the proposed rule, suggest that the benefits of the 2004 BT rule may
have not been realized and were consequently overestimated. However, as described in the next
section, advances in information technology in the last decade are such that private incentives
have led some entities to implement some degree of food traceability beyond the 2004
requirements, suggesting that annualized costs may have been overestimated in 2004 as well.

12 Federal Register /Vol. 69, No. 236 / Thursday, December 9, 2004 / Rules and Regulations, page 71611.
13 Federal Register / Vol. 69, No. 236 / Thursday, December 9, 2004 / Rules and Regulations, page 71640.

2. Current Industry Practices
For purposes of this analysis, we assume that firms in the food supply system already
adhere to the Subpart J traceability recordkeeping requirements stemming from the BT Act. 14
Subpart J requires that non-transporters of food (persons who own food or who hold,
manufacture, process, pack, import, receive, or distribute food for purposes other than
transportation) maintain records regarding their receipt and release of food. More limited
requirements apply to transporters of food. In accordance with section 414(b) of the FD&C Act,
Subpart J does not apply to:
•

Farms;

•

Restaurants;

•

Fishing vessels are exempt from all of the requirements except for those relating to
records availability;

•

Persons who distribute food directly to consumers are exempt from maintaining
recipient records;

•

Retail food establishments that distribute food to persons who are not consumers are
only required to maintain recipient records if the information is reasonably available;

•

Retail food establishments with ten or fewer full-time -equivalent employees (FTEs)
are exempt from all of the requirements except for those relating to records
availability;

•

Persons who manufacture, process, pack, transport, distribute, receive, hold, or import
food in the United States that is within the exclusive jurisdiction of USDA are

We request comment on whether compliance with the earlier regulations may increase as a result of this rule and,
if so, how to quantify the impact.

excluded from all of the requirements with respect to that food while it is under the
exclusive jurisdiction of USDA;
•

Foreign persons, except those that transport food in the U.S.;

•

Persons who manufacture, process, pack, transport, distribute, receive, hold, or import
packaging, except that persons who manufacture, process, pack, transport, distribute,
receive, hold, or import food are subject to the requirements relating to records
availability with respect to the outer packaging of the food that bears the label and
does not contact the food;

•

Persons who manufacture, process, pack, transport, distribute, receive, hold, or import
food contact substances other than the finished container that directly contacts the
food are exempt from all of the requirements except for those relating to records
availability;

•

Persons who manufacture, process, pack, transport, distribute, receive, hold, or import
the finished container that directly contacts the food are exempt from all of the
requirements except for those relating to records availability; except that persons who
place food directly in contact with its finished container are subject to all of Subpart J
with respect to the finished container that directly contacts that food;

•

Non-profit food establishments are exempt from all of the requirements except for
those relating to records availability;

•

Food for personal consumption; and

•

Persons who receive or hold food on behalf of specific individual consumers and who
are not also parties to the transaction and who are not in the business of distributing
food.

Under these existing requirements, firms must know and record information regarding the
immediate previous sources of the food and the immediate subsequent recipients of the products
they make or distribute or both (commonly referred to as one-up and one-back recordkeeping).
This entails recording the name, address, and telephone number of source and receiver firms and
of transporters; a description of the type of food, and the date it was received or released; and the
quantity of the food, and how it is packaged. Firms covered by Subpart J that manufacture,
process, or pack food also must record a lot or code number or any other identifier when
available, though there is no standardized format (Ref. [5]).
Additionally, firms may already voluntarily conform to traceability standards and best
practices developed by outside groups, such as the consensus standards developed by GS1, an
international non-profit organization that develops and maintains standards for barcodes 15. Other
examples of business communication standards include QR codes 16, data matrices, and radio
frequency identification (RFID) codes. At present, the Foodservice GS1 U.S. Standards
Initiative, which promotes traceability standards, has 130 food service companies among its
membership (Ref. [6]). GS1 has standards and best practices for various entities in the food
supply system. For example, GS1 standards for growers include encoding and communicating
trace lot codes, location identification, and other harvest information in order to link individual
cases of product to harvest sites (Ref. [7]). GS1 standards for subsequent supply chain entities
enable enhanced traceability from grower to retailer and back.
A 2004 report from the USDA Economic Research Service (ERS) addressing the traceability
baseline in the United States found that private sector food firms had already developed

15 https://www.gs1.org/
16 A QR code is a machine-readable code consisting of an array of black and white squares normally for storing
smartphone readable URLS or other information.

substantial capacity to trace by the time the BT final rule was published in 2004 (Ref. [8]).
According to the report, food producers, manufacturers, and retailers were typically keeping
traceability records for a wide range of foods and food attributes including elements concerning
food safety. Recordkeeping systems needed for the Subpart J traceability requirements resembled
the systems that already existed for recording receipts and bills. For many of these firms, one-up,
one-back traceability for a standard set of data elements would require little change to their
existing systems.
Similarly, some affected entities already maintain, to varying degrees, key data elements
required by the proposed rule and would likely incur little cost to comply. However, other
entities may face more substantial changes to their existing recordkeeping systems. For example,
some firms may need to generate and assign traceability lot codes to food on the FTL in
instances where these codes do not already exist. For some growers, new steps may include
recording the growing area coordinates of each lot and collecting additional documentation from
seed vendors. Growers may also need to send additional tracing documentation when shipping
FTL foods. First receivers of FTL foods may face new data-entry costs. Firms that transform
covered products but do not currently link supplier lots (of ingredients) to manufactured lots (of
output) would need to add this step to their recordkeeping process.
We lack information on the extent to which current industry practices align with the
requirements of the proposed rule. Namely, we lack information for identifying the number of
firms by industry sector that would need to substantially alter or replace their current systems in
order to comply with the proposed rule. Hence, for firms and establishments that we believe
manufacture, process, pack, or hold foods on the FTL, we acknowledge that the levels of

upgrades needed to conform with the requirements of the proposed rule may vary widely. We
seek comment on these estimates.
3. Coverage of the Rule
Covered entities (firms or establishments) would incur costs from the proposed rule, if
finalized, to the extent that compliance requires them to change their current practices. Covered
entities are those that manufacture, process, pack, or hold foods that FDA has designated as
requiring additional recordkeeping and placed on the FTL. 17 The proposed traceability
recordkeeping requirements would generally not apply to:
•

Farm sales of food produced and packaged on a farm – if packaging of food
maintains integrity of product and labeling includes name, complete address, and
business phone of farm;

•

Farm sales directly to consumers;

•

Fishing vessels;

•

Transporters of food;

•

Nonprofit food establishments;

•

Food for personal consumption;

•

Certain persons who hold food on behalf of individual consumers;

•

Commingled raw agricultural commodities that are not “covered produce” in the
FSMA Produce Safety Rule;

•

Produce and shell eggs that receive certain types of processing;

17 The list of applicable foods can be updated by publishing a notice in the Federal Register, using the process
described in proposed § 1.1465. See Appendix A for the list as of this writing.

•

Produce that is listed as rarely consumed raw in the FSMA Produce Safety Rule, 21
CFR § 112.2(a)(1);

•

Food that has been subjected to a kill step;

•

Certain small originators, including produce farms that are not covered by the FSMA
Produce Safety Rule under 21 CFR § 112.4(a) and shell egg producers with fewer
than 3,000 lay hens at a particular farm;

•

Retail food establishments with respect to food that they purchase directly from a
farm, provided they document the name and address of the farm; and

•

Farm-to-school and farm-to-institution programs with respect to food that is produced
on a farm and sold to the farm-to-school or farm-to-institution program, provided that
the school food authority or relevant food procurement entity documents the name
and address of the farm.

To estimate the number of covered entities, we use several sources. These include the
U.S. Economic Census, the 2016 Statistics of U.S. Businesses (SUSB), FDA’s Food Facility
Registration Module, and the 2017 Census of Agriculture. All datasets used in this analysis were
the latest available to us as of January 2020.
The Census Bureau’s SUSB publishes the number of firms, establishments, and
employment by firm size and industry on an annual basis. The most recent data available is from
2016. Many firm size standards are based on the number of employees, so the 2016 SUSB
employment size categories are additionally useful for identifying the number of small entities in
each affected industry. SUSB annual data include the number of firms, number of
establishments, employment, and annual payroll for most U.S. business establishments. The data

are tabulated by geographic area, industry, and employment size of the enterprise. The industry
classification is based on 2012 North American Industry Classification System (NAICS) codes.
We assume that the FSMA 204 rule as currently proposed would cover approximately
422,144 firms operating 566,448 establishments, including 22,912 farms, 10,623 manufacturers,
18,686 wholesalers, 3,519 warehouses, and 366,404 retailers. This number includes only
domestic entities (firms or establishments) that manufacture, process, pack, or hold FTL foods
destined for consumption or use in the United States. Table 5 contains a summary and
breakdown of this estimate by NAICS code. 18

Table 5. Number of Affected Entities by Industry Sector
Number
Number of
Type
NAICS Codes
of Firms
Establishments
Farms
22,912
22,947
111219, 111339, 111419, 112310,
/Aquaculture /
112511, 112512, 114111, 114112
Growers
311340, 311351, 311352, 311411,
Manufacturers /
311412, 311421-311423, 311513,
Processors /
10,623
11,557
311520, 311710, 311811-311813,
Packers
311821, 311824, 311911, 311941,
311942, 311991
Wholesalers /
424410, 424420, 424430, 424450,
18,686
24,224
Distributors
424460, 424480, 424490
Warehouse and
3,519
6,880
493110, 493120, 493130
Storage
366,404
500,841
445110, 445120, 445220, 445292,
Retail Food
445230, 445291, 445299, 447110,
Establishments
452910, 454111, 454210, 722310,
(including
722320, 722330, 722410, 722511,
restaurants)
722513-722515

18 https://www.census.gov/eos/www/naics/
In each NAICS code category, only entities that manufacture, process, pack, or hold foods on the FTL would be
affected by this rule.

We estimate that the total number of domestic farms that produce foods on the FTL,
including produce, eggs, and fish, and would thus be affected by the proposed rule is 22,912. 19
This includes 18,918 farms that grow covered foods, including fruits and vegetables such as
leafy greens, cucumbers, tomatoes, peppers, fresh herbs, tropical tree fruits, melons, and sprouts.
We derived the number of farms from the 2017 USDA Census of Agriculture, which includes
farms with on-farm packing, greenhouses, and other originators of covered produce foods (Ref.
[9]).
We estimate that there are approximately 95 covered sprouting operations (included in
the total number of farms in Table 5). There is little information on the size and structure of the
sprout producing sector. We ask for comment on the distribution of sprouting operations across
firm size categories.
The proposed rule also applies to shell egg producers and growers. Shell egg producers
with fewer than 3,000 laying hens at an individual farm, with respect to the shell eggs they
produce, are exempt. This exemption is the same as that found in the Prevention of Salmonella
Enteritidis in Shell Eggs During Production, Storage, and Transportation Final Rule, or the Egg
Safety Final Rule (Ref. [10]). Farmers who sell their eggs directly to consumers are exempt from
all provisions. Sales of eggs directly to consumers include sales of a farmer’s own eggs to
neighbors, at farmers markets, and at roadside stands.
The Egg Safety Final Rule includes recordkeeping requirements related to the prevention
of Salmonella Enteritidis, but not to enhance traceability of shell eggs within distribution
networks. To estimate the number of covered egg producers, we use the National Agricultural

The first domestic entity that takes physical possession of an imported product would be responsible for obtaining
the required information from foreign farms – these firms are affected entities and included within other sector
categories (manufacturers/processors/packers/holders, wholesalers/distributors, or warehouse and storage).

Statistics Service (NASS) 2017 Census of Agriculture to determine the number of farm sites with
layers on hand. NASS estimated that there are 232,500 farms with layers over 20 weeks old in
their inventory (Ref. [9]). There are over 228,000 farms with fewer than 3,000 laying hens,
representing over 99 percent of egg producers but less than one percent of total domestic egg
production. We estimate that out of 3,064 egg farm sites, 1,532 sites would be covered by some
or all parts of the rule. 20
We use FDA’s Food Facility Registration Module (FFRM) biennial registration data for
our estimates of foreign entities affected by this proposed rule. As discussed in more detail in
section II.H, we compare the number of registered foreign facilities in the FFRM to the primary
estimated number of affected domestic establishments minus retail food establishments. We then
use this ratio, approximately 1.95, as a scaling factor to extrapolate from the number of affected
domestic entities in the main analysis. We believe the number of foreign retail food
establishments affected by this rule to be negligible. Thus, we estimate that the rule would cover
approximately 89,977 foreign firms and 118,954 establishments operated by those firms. We do
not have a detailed breakdown of foreign establishments and firms by industry, and instead
assume the same proportional breakdown as in the main analysis. The International Effects
section of this document discusses this process in detail.
We welcome comments on these estimates and our baseline discussion.

20 The NASS Census of Agriculture uses farms with 3,200 birds as its cutoff point for categorization. FDA uses
3,000 birds as its cutoff point for small versus large farms, because this is the measure that is used in other egg and
poultry regulations. To adjust the NASS data, FDA assumes that all flocks are uniformly distributed across the 400
to 3,200 bird category. Using this assumption, 7.1 percent (200 ÷ 2,800) of these farms fall in the over 3,000 bird
category while the remaining 92.9 percent fall in the small farm category.

E. Benefits of the Proposed Rule
The proposed rule, if finalized, is expected to improve FDA’s ability to: (1) quickly and
efficiently trace the movement of listed foods through the supply chain and (2) identify and
remove contaminated food from the marketplace during an outbreak. In the event of a foodborne
outbreak, the ability to trace a food back in the supply chain from the point of sale or service to a
common source is important for identifying contaminated foods or ingredients and removing
those products from the marketplace to prevent additional illnesses. The ability to trace foods
forward can help FDA ensure the removal of all affected products from the marketplace and
understand how the distribution of a food product relates to illnesses or illness clusters,
especially for outbreaks that are challenging to resolve, such as those involving multiple foods
and foods with multiple ingredients. Therefore, if finalized, this rule is expected to result in the
following benefits:
1) Public health benefits from averted foodborne illnesses caused by foods covered by the
proposed rule;
2) Benefits from avoiding broad recalls;
3) Other benefits that we discuss below.
The health-related benefits which are the primary focus of this analysis can only be
realized in the event of an averted or reduced-duration foodborne illness outbreak from foods on
the FTL; benefits from avoiding overly broad recalls may be realized only in the event of a
recall; and the other benefits may be realized regardless of an outbreak. However, the costs
would be incurred by all entities regardless of whether there is an outbreak investigation or recall
underway, and regardless of whether they are implicated in the outbreak.

The benefits accrued from improved product traceback are difficult to quantify. An IFT
report on food tracing pilot projects specifically notes that it was easier to identify costs
associated with system upgrades for firms to implement product tracing than quantify the
benefits of product traceability (Ref. [5]). There is uncertainty in assigning specific benefits from
this rule alone relative to other FSMA and BT rules, which have been implemented over recent
years. 21 FDA has the information about outbreaks and recalls and expects that the proposed rule
would lead to achieving the described benefits. However, we are uncertain about the exact share
of outbreaks that would be averted specifically as a result of this rule in relation to other FSMA
rules. Other recently implemented FSMA rules also aim to avert foodborne illness outbreaks
from FTL foods. Given that the same outbreaks dataset has been used in the past for discussing
benefits of other FSMA rules in those RIAs, we aim to avoid double counting benefits by
predicting that FTL food outbreaks will be averted as a result of this proposed rule alone.
We estimate benefits in two ways. First, we use case studies of four pathogens to estimate
monetized benefits of potential averted illnesses from faster traceback. Second, we rely on three
recall case studies to estimate potential benefits from avoiding overly broad recalls. We welcome
comments on our assumptions and our estimates of the benefits of the proposed rule. We are
particularly requesting any information that would permit the Agency to more accurately
quantify the likely benefits.

21 Since the 2011 enactment of FSMA, FDA has finalized seven major rules implementing FSMA. These rules are
based on the premise that the safety of the food supply is a shared responsibility at many different points in the
global food supply chain for both human and animal food. All FSMA rules combined are designed to make clear
specific actions that must be taken at each of these points to prevent contamination:
https://www.fda.gov/food/guidance-regulation-food-and-dietary-supplements/food-safety-modernization-act-fsma

1. Potential Benefits from the Proposed Rule
i. Establishing a framework for better tracing
This proposed rule would establish a framework to more accurately and efficiently trace
potentially contaminated FTL foods (both domestic and imported) across the U.S. food supply
system to protect the health of consumers. The rule would establish a consistent approach for
product tracing for the different types of products and firms subject to this regulation. The rule
also specifies the data elements and information firms must establish and maintain, along with
information they must send, in certain circumstances, to the next entity in the supply chain.
These linkages of tracing information not only enable FDA to determine the source of a
contaminated food but also provide visibility into the critical affected points of the food supply
system and distribution chains.
Tracing a food back in the supply chain from the point of sale or service to a common
source is important for identifying contaminated foods or ingredients and removing those
products from the marketplace to prevent additional illnesses. Tracing foods forward can help
FDA ensure the removal of all affected products from the marketplace and understand how the
distribution of a food product relates to illnesses or illness clusters, especially for outbreaks that
are challenging to resolve, such as those involving multiple foods and foods with multiple
ingredients.
Improved food product traceback will enhance FDA’s ability to complete traceback
investigations more efficiently, leading to a faster identification and removal of implicated food
products from the market. This should consequently reduce the number of illnesses resulting
from consumption of implicated foods. A shorter traceback investigation would reduce the
duration of an outbreak. A shorter outbreak duration would result in fewer illnesses resulting

from consumption of contaminated foods. In contrast, a delay in a traceback investigation due to
unavailability of adequate traceability records or their poor quality could result in a longer
duration of the outbreak resulting in a larger number of illnesses. Prolonged outbreak response
and product traceback can have devastating public health and economic outcomes because they
increase the total number of illnesses.
The proposed rule, if finalized, would speed up tracing of FTL related outbreak
investigations, because increased and improved records would mean that investigators spend less
time trying to find and analyze information about FTL foods that might have been missing or
incomplete. The proposed rule has the potential to improve traceability as it focuses on quickly
identifying the food that is the source of a foodborne illness outbreak and tracking its movement
throughout the food supply system. By establishing recordkeeping requirements for key data
elements, the rule is designed to increase the rate of successful tracing completions for covered
foods (Ref. [11]).
Implementation of the rule, therefore, is expected to put in place the mechanisms that
should enable us to more swiftly and more precisely trace the source of an outbreak through the
food supply system for foods on the FTL. Improved documentation and recordkeeping practices
should increase the rate and speed of successful tracing completions. Consequently, fewer
foodborne illnesses from foods on the FTL may occur, reducing hospitalizations and deaths.
ii. Better identification of food vehicle
When there are outbreaks that involve foods that are multi-ingredient, it is hard to tease
out what the likely vehicle is if there is not another group of illnesses with exposure to a single
food item for comparison. With the ability to collect linked records more quickly, the
investigative and traceback efforts will move faster and be less burdensome, so less time will be

needed to identify the implicated food or ingredient. Tracing the ingredients and comparing to
see if there are common suppliers could help identify the implicated food vehicle and reduce
public exposure.
The Agency has sometimes been unable to link illnesses to a specific food vehicle due to
inconsistent, unstandardized recordkeeping and the frequent lack of lot tracing during
distribution to specific retail locations. Traceback investigations begin at retail food
establishments to collect data to identify details about products suspected of causing illness,
including information on brands, varieties, and sources of products of interest. More accurate and
detailed data on the products of interest at the retail food establishment (RFE) enables more
refined record collection throughout the rest of the supply chain (Ref. [1]). In 2018, FDA
investigated a cluster of illnesses caused by Cyclospora cayetanensis at small restaurants. We
were unable to obtain enough information to narrow down the products suspected of
contamination (e.g., basil, cilantro, vegetable trays) due to the restaurants’ lack of records
indicating lot numbers received and linking to information throughout the supply chain. In the
absence of more specific data at the retail food establishment, we had to conduct a broader
record collection involving numerous suppliers to ensure that we had sufficient tracing
information to accurately determine what lots likely would have been available for consumption
or purchase at the establishments by the sickened persons.
Information from all RFEs, whether large or small, is integral in investigating foodborne
illness outbreaks, especially since RFEs are most often the first stop in the traceback process.
The examples noted above and in the supporting memo (Ref. [1]) represent just a few of the
outbreaks where availability of “Receiving KDEs” at RFEs would have improved the outcomes
of traceback investigations and recall efficiency. Requiring recordkeeping by RFEs of all sizes

allows for more consistent, organized, and specific information that covers the entire supply
chain and will improve the ability for FDA to investigate foodborne illness outbreaks. The data
available to investigators if all RFEs are covered by this rule would improve the timeliness,
precision, and accuracy of foodborne illness outbreak investigations and thus contribute to
prevention of illness (Ref. [1]).
The improved ability to identify the exact source of contaminated FTL foods may
reduce the time needed to complete tracing investigations when a foodborne illness outbreak
occurs, therefore enabling earlier initiation of a recall. Greater tracing speed can result in faster
and more precise recalls (if the FTL product is still on the market) and other preventive actions
that may reduce the number of illnesses during an outbreak. With the records required by this
proposed rule, the Agency would be able to investigate outbreaks more quickly and may not be
forced to terminate an investigation because of poor or nonexistent records. If finalized, this rule
may reduce both the clinical and other economic burden of foodborne illnesses caused by these
FTL foods outbreaks.
ii. Reducing outbreak duration due to standardization of data elements
When a foodborne illness outbreak occurs, effective traceability programs can lessen the
potential adverse impact of the event by reducing the outbreak duration. This is possible when
the firm can quickly and precisely provide specific traceability information on a suspected
product to regulatory agencies. This information can enable the confirmation of common foods
and ingredients associated with illnesses and help determine which foods and ingredients can be
eliminated from further consideration as possible sources of contamination. Furthermore, being
able to identify the source of a contaminated product quickly enables FDA to conduct timelier

root-cause analysis, which could provide important information to help in understanding how
contamination may have occurred and prevent future outbreaks.
Standardization of data elements is needed to help ensure successful and faster
traceability throughout the supply chain. While some elements of internal product tracing
information are kept by many food producers, manufacturers, distributors, and retailers, the types
of information recorded and maintained, the format in which information is kept, the length of
time information is retained, and the amount of information shared between trading partners
varies among firms. These challenges are further compounded when looking at the traceability of
a product moving through multiple entities in a supply chain.
The IFT report recommended that FDA establish a uniform set of recordkeeping
requirements, representing an enhanced framework from current mandates and practices. The
framework outlined in this report, although not a formally adopted technical standard, has
become a common way of thinking about traceability and is commonly used as a basis for
developing food traceability software and options for food supply entities. As an example of the
influence of the approach, the framework has been referenced in the GS1 Global Traceability
Standard 2.0 (2017) 22, FAO Blockchain Application in Seafood Supply Chains (2020) 23, FAO
Seafood Traceability for Fisheries Compliance (2017) 24, and multiple other research reports, best
practices guidelines, and textbooks. By establishing a more uniform set of recordkeeping
requirements based on IFT’s framework of key data elements and critical tracking events, FDA
can enable tracing software providers, large food supply system entities, and food network
stakeholders to advance system interoperability and standardization. Such a framework enhances

22 https://www.gs1.org/standards/traceability/traceability/2-0
23 http://www.fao.org/3/ca8751en/ca8751en.pdf
24 http://www.fao.org/3/a-i8183e.pdf

the ability of firms within the food system to transmit and exchange information about suspect
ingredients and products involved in investigations and recalls.
In 2016, for example, FDA performed a traceback investigation of frozen strawberries
associated with illnesses of hepatitis A. A lack of traceable lot code information on the products
at the retail level and time understanding the supply chain hindered the traceback investigation.
These two factors, among others, had a significant impact on the speed and efficiency of the
traceback investigation. While a single supplier of strawberries was identified, many of the
complications and delays encountered during the tracing process could have likely been avoided
with the provisions of this rule. Under the proposed provisions of this rulemaking, federal and
state officials would have had access to the traceability lot code generator information at the
retail location. Since there was no manipulation of the product once packaged by the source
manufacturer, access to the traceability lot code generator would have immediately identified the
source and traceable lot codes within the first 24 hours of the initiation of the traceback at the
point of service. In situations where supply chains have secondary customers (i.e., brokers), FDA
must spend time determining where the product physically went. In the case of the hepatitis A
traceback, differentiating these firms increased the time needed to complete the traceback,
ultimately impacting product actions. Because the product was packaged ready for the end-user
and incurred no additional manipulation through its distribution, it would have been unlikely that
contamination occurred at the other points in the supply chain. Knowing this early in the
traceback investigation would have reduced the time necessary to gather information about the
distribution of the product.
Under the existing requirements of Subpart J, it is necessary for FDA to collect
information from each point in the distribution chain to effectively trace a product. Initiation of

the first product recall took approximately one month from the start of the frozen strawberry
traceback based on the need to collect information from every point in the supply chain and
determine where the product physically went. In this case, availability of the traceability lot
codes and traceability lot code generator location identifier at the point of sale would have
allowed FDA to contact the manufacturer of the product or importer immediately while verifying
records across the supply chain simultaneously. Had the recordkeeping requirements of this
proposed rule been in place, a recall would have likely occurred within days of the traceback
being initiated, resulting in a significant time and cost savings.
iii. Reducing the Scale of Outbreaks of Complex and Commingled Products Through
Improved Product Traceability
The IFT traceability pilot report includes a detailed description of a supply chain scenario
where traceability could enhance withdrawal of products from the market even when some
ingredients of the final product could not be linked to the product (Ref. [5]). At the core of this
scenario were two products marketed under the same brand. One was a frozen product (skillet
meal produced by one manufacturer) and the other was a dry version in which the consumer
would add meat (produced by another manufacturer). These two products were manufactured in
two different locations, and the dry product was produced by a co-manufacturer. Additionally,
each product contained “pouches” (separate sauce and peanut pouches in each product, and
whole chilis for the dry product) that were manufactured by additional parties (one produced the
sauce and another pouched the whole dried chilis, with the peanuts packaged by the peanut
supplier).
In this scenario, it was up to the solution provider to use traceability software to
determine the ingredients/lots likely to have been used at both manufacturers. This scenario

demonstrated that the most time-consuming element of the task was harmonizing the six
different data sets into uniform CTEs/KDEs. Once the data was standardized and connected, the
software could display the supply chains and their points of convergence nearly instantly.
This scenario tested the process used to acquire product tracing information, with the
product-brand owner functioning as the main source of the information collected within 24-36
hours. The final product owner was able to provide information to trace back to the specific
ingredients, some of which were traced back to overseas suppliers. Traceability of these products
was completed within 35 hours and the analysis took less than three hours even though some of
the individual ingredients could not initially be linked to the processed food product. Ultimately,
the products that were earmarked for a recall were identified with a degree of precision that
would not have been possible without the availability of traceability records and the standardized
system that allowed this coordination and convergence. The IFT report concludes that, with such
a high level of precision and timeliness, traceability can be used to help resolve outbreaks caused
by illnesses from commingled or more complex products with multiple manufacturers, facilities,
and providers/suppliers. Illnesses from such outbreaks often remain unattributed when reports
are compiled due to the difficulty of identifying the specific food responsible. Thus, improved
traceability can reduce the number of unattributed foodborne illnesses.
iv. Findings of the IFT Report
The 2012 IFT report emphasizes the importance of public health benefits as the primary
outcome from enhanced product traceability. The report provides a description of benefits for
several case studies. The report provides some initial estimates of potential benefits, along with a
framework on how broader estimates of benefits could be calculated. To estimate public health
benefits of food product traceback, the report suggests the following three key assumptions:

(i) Product tracing can reduce the time between identification of an implicated food
causing an outbreak and the identification of the specific product and recall or removal
from commerce. This reduction of disease incidence is assumed to be proportional to the
time reduced through effective whole supply chain product tracing.
(ii) The establishment of seamless product tracing system across the food industry would
result in a near ‘instantaneous’ identification of a food source and the initiation of product
intervention to remove it from commerce and stop public exposure to contaminated
product. Instantaneous traceback gives the theoretical maximum possible reduction in
illnesses.
(iii) Each illness-causing pathogen has a specific incubation period. Any calculation on
benefits of reduced illness due to earlier food vehicle identification and intervention must
consider the specific pathogen incubation period regardless of when the exposure to
contaminated food was stopped.
The IFT report used case studies based on eight outbreaks caused by two pathogens
(Salmonella and Listeria monocytogenes). For lack of reliable data, the report argues that these
case studies can serve as a good guide in identifying and analyzing public health benefits. With a
richer dataset, the methodology can also be replicated and expanded to analyze foodborne illness
outbreaks that could help to provide a more accurate picture of the benefits of improved product
tracing. The scope of the benefits discussed in the IFT report is limited and difficult to generalize
to broader sets of food products and foodborne illness outbreaks, as are some of their
assumptions, such as their estimated cost of Listeria and Salmonella illness of $10,300 and
$17,900 per case, respectively (Ref. [5]). Our analysis uses validated and more generalizable
data sources, including FDA’s foodborne illnesses outbreak data, which consists of 10 pathogens

commonly known to cause FTL-associated illnesses, as compared to only two pathogens
discussed in the IFT report based on a case study from eight outbreaks.
v. Effects of Tracing on Public Health
For public health benefits to be realized, the Agency and industry must take timely
preventive actions. Executing effective and timely recall of contaminated foods is important but
difficult to achieve (Ref. [12]). As the 2010 shell eggs Salmonella contamination illustrates, food
recall is a complex process. According to CDC, the shell eggs outbreak was first reported in May
2010 and the recall was issued in August 2010. However, the outbreak continued until October
2010 when all contaminated food vehicles were identified and recalled (Ref. [13]). Because of
the length of time it took to identify the food vehicle, this outbreak was the largest reported since
the early 1970s when outbreak surveillance was established (Ref. [14] [15]), resulting in many
illnesses that could have been avoided with better tracing.
At the start of the investigation for this outbreak, there was a lack of clusters of illness
with an epidemiologic association to shell eggs. The clusters required more epidemiologic
evidence to be obtained on the consumption of eggs and egg-containing foods in order to have
enough information to begin a traceback investigation. When epidemiological studies cannot
distinguish which of several possible ingredients or foods is clearly implicated in outbreaks,
federal and state authorities often elect to identify the most likely one or two ingredients or foods
in two to three of the clusters in which the best epidemiological information and case histories
are available, and trace those foods to see if there are common suppliers. These are frequently
referred to as “epidemiological tracebacks” and are meant to help inform the epidemiological
investigations (Ref. [16]). This outbreak required an epidemiological traceback to verify that the

food vehicle was in fact shell eggs, and further traceback of additional clusters to verify the
common supplier.
Standardized records linking shipments through the supply chain and back to their
sources would have reduced the time needed to review the necessary information. For some
clusters, the district offices had to return to the firm for additional clarification on linking
incoming to outgoing product shipments, which prolonged the traceback investigation. Also,
many firms did not record this information at all, causing traceback to rely on analysis of
shipment dates alone. This analysis slowed the process for identification of the food vehicle,
common supplier, and affected lots. Ultimately thousands of people were sickened as a result of
this outbreak. The number of illnesses could have been meaningfully reduced if a better tracing
system had been in place.
In sum, the public health benefits of this rule, if finalized, would arise when foodborne
illnesses from foods on the FTL are averted. To assess these benefits, we must therefore place a
value on risk reduction and health-related costs for illnesses that may be averted. However, any
quantified estimates of health benefits from shortening outbreaks would be uncertain as the
following information would be necessary to better estimate the rule’s public health safety
benefits:
•

The baseline risk of foodborne illness attributable to foods covered by the rule;

•

Estimates of lost health as measured by morbidity and mortality attributable to
foodborne illnesses from foods covered by the rule;

•

Value of lost health due to FTL-associated foodborne illnesses;

•

Changes from baseline recordkeeping practices to post-rule practices; and

•

The anticipated effectiveness of the proposed rule in reducing associated foodborne
illnesses.

Therefore, while we believe this proposed rule, if finalized, would produce benefits, the lack of
complete information introduces high uncertainty in estimating such benefits. Our estimates of
these benefits thus include wide ranges based on the numbers of illnesses attributed to outbreaks
in recent history, ranged assumptions in the IFT report about how effectively traceability
measures would have shortened epidemic curves from foodborne illnesses, and ranged
assumptions about the monetary value of lost health due to foodborne illnesses. Section E.2
Public Health Benefits from Averted Illnesses discusses our use of the model from the IFT report
in conjunction with historical epidemic curve data to project the effectiveness of the rule at
reducing foodborne illnesses associated with FTL foods.
vi. Baseline Risk of Foodborne Illnesses Attributable to FTL Foods
As explained in more detail in Appendix B, we estimate that 119,706 cases of foodborne
illnesses per year in the U.S. are caused by known pathogens associated with foods that are on
the FTL. From FDA foodborne illness outbreak data (Ref. [17] [18]), we identified a total of 10
microbial pathogens responsible for the majority of foodborne illnesses attributable to foods
covered by this proposed rule were identified from FDA foodborne illness outbreak data. Of
these pathogens, E. coli accounts for approximately 60 percent of the foodborne illnesses,
Salmonella accounts for approximately 20 percent, and Cyclospora cayetanensis accounts for
approximately 17 percent. In terms of annual hospitalizations, the leading pathogen was
Salmonella, which accounts for 58 percent, followed by Listeria, which accounts for 15 percent,
and E. coli, which accounts for 13 percent of total annual hospitalizations. Other pathogens

within FDA data that cause foodborne illnesses for FTL foods include ciguatoxins, hepatitis A,
norovirus, scombrotoxins, tetrodotoxins, and Vibrio spp.
The foods covered by this rule have a higher public health risk based on the frequency
and severity of outbreaks associated with these foods as well as their frequency of consumption.
Table 6 summarizes the estimated number of illnesses, hospitalizations, and deaths
attributable to foods covered by the proposed rule. As described in Appendix B, to obtain these
numbers, we reviewed FDA’s Coordinated Outbreak Response and Evaluation (CORE) outbreak
data for FTL foods covering the period from 2009 to July 2019. We used multipliers from peerreviewed literature to account for underreporting and misdiagnoses of foodborne illnesses (Ref.
[19] [20]). The methodology outlined in the Scallan et al. follow-up article is used to account for
unspecified and unknown agents 25 (Ref. [21]). Including the roughly 80 percent of illnesses
caused by unknown and unspecified agents, we estimate a total of 598,531 foodborne illnesses
associated with FTL foods (= 119,706 illnesses x 5) (figures rounded to nearest whole number).
Our estimates cover a different time period (2009 – 2019) as compared to Scallan et al. 2011
estimates, which covered the years 2000 – 2008. The distribution of the burden of foodborne
illnesses by different types of pathogens has changed over time, as evidenced by recent CDC
publication which noted that in 2019 there was a 12-fold increase in Cyclospora infections
compared to 2016-2018 (Ref. [22]). Additionally, recent disease outbreaks associated with leafy
greens, fresh herbs, and fresh-cut vegetables – all on the proposed FTL – have been linked to
Cyclospora as evidenced on CDC’s food outbreak records 26. Though we follow these standard

25
As explained in Appendix B, according to Scallan et al., (2011b), apart from foodborne illnesses caused by major
known pathogens, nearly 80% of additional episodes of foodborne illness are caused by unspecified agents,
including known agents about which we lack sufficient data to estimate agent-specific illness. There are also
illnesses caused by known agents that are not yet recognized as causing foodborne illness as well as substances
known to be in food but of unproven pathogenicity, and unknown agents.
26 https://www.cdc.gov/foodsafety/outbreaks/multistate-outbreak/outbreaks-list.html

assumptions provided in peer-reviewed literature, we recognize that sparsity of data results in
uncertainties in our accounting of unspecified and unknown agents. We request comment on how
to best adjust our estimates for unspecified and unknown agents, as well as for underreported and
misdiagnosed agents.
Table 6. Estimated Number of Illnesses, Hospitalizations, and Deaths Attributable to Foods
Covered by the Proposed Rule 27
Annual
Type of Pathogen
Estimates
Hospitalization Deaths Percent total cases
36
0
0
0.03%
Ciguatoxin
20,672
14
0
17.27%
Cyclospora cayetanensis
72,060
53
2
60.02%
E. coli (All strains)
584
39
0
0.49%
Hepatitis A Virus
73
60
12
0.06%
Listeria
918
1
0
0.77%
Norovirus
23,793
238
3
19.88%
Salmonella (All strains)
65
0
0
0.05%
Scombrotoxin
1
0
0
0.00%
Tetrodotoxin
1,504
3
0
1.26%
Vibrio (All strains)
Estimated total from
119,706
408
17
100.00%
known pathogens 28
vii. Economic Burden of Foodborne Illnesses Associated with FTL Foods
We estimate the total burden of foodborne illnesses attributed to FTL foods by
multiplying the estimated annual number of illnesses per pathogen from Table 6 by the updated
burden of illness estimates. We update burden of illness estimates first published in 2015 (Ref.

This table is compiled using FDA’s foodborne illnesses outbreak data that was also used in the risk-ranking
model, which was the basis for designating the FTL.
28
Our estimates of hospitalization and death counts include specific pathogen multipliers according to Scallan’s
methodology. Except for Norovirus, the Scallan methodology doubles both death and hospitalization numbers,
which we accounted for in this table.
Chemicals or substances like ciguatoxins, scombrotoxins and tetrodotoxins were not available in the original
Scallan (2011a) article but have since been established as causes of foodborne illnesses. We obtain underdiagnosis
multipliers for ciguatoxins and scombrotoxins from Pennotti et al (2013) (Pennotti, Scallan, Backer, Thomas, &
Angulo, 2013), and use the same multiplier for tetrodotoxins as for ciguatoxins. We assume their hospitalization and
death rates were constant.

[23]) to 2018 dollar values 29 (Ref. [24]). This burden includes both direct costs and indirect
costs, and accounts for variations in the level of severity of foodborne illnesses. The direct costs
are associated with doctor visits and hospitalization. Indirect costs are from the loss in quality of
life (of which loss in productivity is a subset) as a result of the symptoms and severity of the
foodborne illness. The burden is monetized using the value of a statistical life (VSL) as provided
in HHS guidelines (Ref. [25]). The total economic burden of illness in Minor, et al., 2015, is
therefore estimated by computing and combining for each illness the average monetized acute
health loss, the average monetized secondary health loss (from long-term health effects), the
average monetized loss of life years, and the acute and secondary medical costs. 30 We rely on
these estimates for our analysis and refer the reader to Minor, et al., 2015 for more detailed
discussion of these computations. Table 7 shows the estimated burden of illnesses associated
with outbreaks attributable to foods covered by this rule. We list the common microbial
pathogens associated with covered foods and the estimated average annual number of illnesses
associated with these pathogens. The number of illnesses per pathogen is then multiplied by its
expected burden of illness to produce the economic burden of foodborne illnesses caused by

29. The model updates included revised annual dollar estimates of foodborne illness costs from the 2015 estimates
to 2018 values using BLS deflators. Additionally, new foodborne illness causing agents, which include
scombrotoxin fish poisoning and ciguatera fish poisoning, were also added into the model. These had been missing
in the original model for lack of reliable data. These agents were also never covered in Scallan et al., 2011a which is
largely credited for establishing the basis for most foodborne illness pathogen causing agents. For these multipliers,
we cite Pennotti et al 2013.
30

Minor et al. (2015) present their estimates of cost per illness as applying to the illness episodes that Scallan et al.
(2011a) quantify using data from 2000 to 2008 from several sources. By contrast, the estimates in this RIA are
based on FDA CORE data from 2009 to 2019. While the Scallen et al. primary estimate of 11,407 cases annually
across all food sources is far lower than what is attributed here just to FTL products, Tack et al. (2020) indicate that
compared with 2016-2018, the incidence of Cyclospora increased significantly (1,209%). This increase has been
seen in previous years as well – CDC notes that the incidence of Cyclospora infections increased markedly in 2018,
in part because of large outbreaks associated with produce
(https://www.cdc.gov/mmwr/volumes/68/wr/mm6816a2.htm?s_cid=mm6816a2_w). The increase in Cyclospora
infections might be partly due to increased detection (by more labs using new tests) but also partly due to increased
exposure to this pathogen, particularly contaminated FTL products.

known pathogens. We then multiply by five to account for illnesses caused by
unidentified/unspecified pathogens. If the rule prevented all illnesses estimated as caused by FTL
foods (about 598,500 foodborne illness per year), it would result in annualized public health
benefits of approximately $4.3 billion dollars per year at a seven percent discount rate. We do
not know, however, the extent to which the proposed rule, if finalized, may be able to avert this
estimated number of foodborne illnesses.
To obtain our adjusted estimate of the total annual burden of foodborne illnesses from
outbreaks from foods covered by the proposed rule, we account for exemptions and scale down
our estimates in Table 6. This proposed rule has exemptions for several types of entities and
circumstances and, therefore, not all estimated illnesses may be averted. For example, certain
farms are exempt from some activities because of their size, retail sales volume, or distribution
channel, but are likely responsible for a portion of estimated foodborne illnesses from FTL
foods. For this reason, the next entity in the supply chain after the (exempted) farm will need to
start the record chain by applying the traceability lot code. We lack data on the number of
foodborne illnesses that may continue to result from exempting very small farms. However, if we
account for these illnesses by adjusting our estimates upward by the share of very small farm
sales as in the Regulatory Impact Analysis for the Standards for Growing, Harvesting, Packaging
and Holding of Produce for Human Consumption rule 31 (Ref. [26]), the estimated number of
FTL-related illnesses would still be around 514,700 cases per year (= 598,531 x (1 – 0.14)). The
adjusted annualized burden reduction of all FTL-caused illnesses would still be about $3.74
billion per year (= $4.3 billion x (1 – 0.14)).

31 The economic analysis for this rule estimates that very small farms represent a small proportion (14%) of
produce sales. Because exempting these farms may perpetuate foodborne illnesses, the analysis adjusts the economic
burden of illnesses upward by 14%. We follow the same procedure as an example, noting the small risk of illnesses
associated with products from very small farms due to their limited reach.

We recognize the importance of this analysis considering the impact of other FSMA
regulations in reducing the amount of foodborne illnesses. However, we do not have a good
mechanism to assess whether the other FSMA rules will achieve the health benefits estimated in
the regulatory impact analyses for those rules, particularly since the compliance dates for many
of the other FSMA rules have only passed recently, and for some of the provisions the
compliance dates have not yet passed. While estimated costs and benefits of other FSMA rules
may (or may not) have been incurred entirely, estimated costs of this rule reflect the costs added
by this rule only. Expected health benefits are calculated based on the most recent data on
number of illnesses associated with foods covered by this rule. We ask for comment on best
ways to account for the benefits of other FSMA rules.
Table 7. Estimated Economic Burden of Foodborne Illnesses Associated with Foods
Covered by this Proposed Rule (2018$)

Pathogen
Ciguatoxin
Cyclospora cayetanensis
E. coli (STEC) O157
E-Coli (STEC) non-O157
Hepatitis A Virus
Listeria Monocytogenes
Norovirus
Salmonella non-typhoidal
Salmonella typhoidal
Scombrotoxin
Tetrodotoxin
Vibrio-parahaemolyticus
Vibrio-Cholerae
(i) Subtotal/Weighted Average
(rounded): Known Pathogens
(ii) Unidentified/Unspecified
Pathogens
(iii) Total cases (i) & (ii) (rounded)
Time Period (Years)

Estimated
Monetized Burden
Annual Cases
per Illness
36
$16,208
20,672
$4,022
41,454
$9,376
30,606
$2,266
584
$52,854
73
$1,797,753
918
$442
23,778
$6,563
15
$6,504
65
$504
1
$14,947
1,462
$2,385
42
$1,520
119,706
$7,220

Total: Primary
$581,284
$83,152,546
$388,662,679
$69,362,601
$30,887,800
$131,527,034
$406,061
$156,060,310
$98,867
$32,800
$14,107
$3,486,163
$63,246
$864,335,498

478,825

$7,220

$3,457,115,056

598,531

$7,220

$4,321,450,554
10
53

Estimated Total Burden of Illnesses from FTL Foods
Present Value (3% discount rate)
Present Value (7% discount rate)
Annualized Value (3%, 10 years)
Annualized Value (7%, 10 years)

$43,216,774,918
$36,864,785,600
$30,353,654,258
$4,321,677,492
$4,321,677,492

2. Public Health Benefits from Averted Illnesses
We estimate public health benefits using the model provided in the IFT report (Ref. [5]).
In describing public health benefits related to tracing, the IFT report presents an analysis based
on eight outbreaks. Seven of the outbreaks were from Salmonella infections and one was from
Listeria monocytogenes. Each outbreak provided information on 1) the pathogen associated with
the outbreak; 2) the investigation description; 3) the potential improvement from the estimated
date of the initiation of traceback to the estimated date of recall or other intervention; and 4) total
illnesses and deaths for the duration of the outbreak.
Table 8 below shows the estimated percentage of illnesses prevented assuming 100%
product tracing improvement (a hypothetical maximum of instantaneous traceability) during the
investigation of these foodborne outbreaks. Table 8 includes results from five of the eight case
studies from the IFT report (excluding two involving foods not on the FTL) plus ten additional
case studies using epidemic curve data from CDC and investigation and intervention data from
FDA as explained in Appendix C of this analysis. The additional cases cover outbreaks
associated with four pathogens: Cyclospora, E. coli (STEC), Listeria monocytogenes and nontyphoidal Salmonella. 32

FTL associated outbreaks caused by these four pathogens represent over 90% of all FTL associated illnesses.

Appendix S of the IFT report describes in detail the analytical process and the
applicability of the analysis 33 that we use to estimate the percentage of illnesses that are
potentially preventable with the tracing requirements of this rule. We use the same process in
estimating the percentage of illnesses potentially prevented assuming FDA had 100% tracing
improvement (i.e., instantaneous product tracing) resulting from the tracing requirements from
this rule. However, FDA outbreak investigation processes and outbreak data collection have
changed since the IFT report. In 2011 the Coordinated Outbreak Response and Evaluation
(CORE) Network was created with the purpose of providing a structured process for responding
to an outbreak which includes an outbreak response phase that centers on traceback of product,
removal of product from the marketplace, and investigation of how the outbreak may have
occurred. Due to this established framework, the best consistent date range for post-2011
traceback investigations is the initiation and completion dates of CORE traceability activities, as
described in Appendix C. 34 We therefore use “initiation” and “completion” dates provided by
CORE in estimating the percentage of illnesses potentially prevented for post-2011 outbreaks.
Minimum and maximum preventable illnesses in Table 8 represent variable potential impact of
traceability among case studies involving the same pathogen.

In Appendix S of the IFT report, the estimated number of reduced illnesses potentially prevented is calculated by
using the epidemic curve data for each associated outbreak. Over the outbreak timeline, the IFT report estimates the
number of days (and illnesses) between the initiation of the traceback and the initial or final intervention date
(depending on the outbreak). The number of illnesses over the time period is divided by the total number of illnesses
during the outbreak to obtain the ratio of illnesses potentially prevented assuming 100% tracing improvement (i.e.
instantaneous tracing).
34
Since each outbreak presents unique circumstances, such as availability of product on the market to recall and the
potential for multiple sequential recalls during one outbreak, using the initial date of recall may not represent the
best end date to represent the end of traceability activities. The CORE traceback initiation date represents a point in
time when traceability activities began, and the CORE traceback completion date represents a point in time in which
the traceback activities and interventions such as a recall have ended.

Table 8. Estimated Percentage of FTL Associated Illnesses Preventable with Product Tracing
Improvement.

Year

Commodity

Pathogen

Total
Illnesses per
Epidemic
Curve

Preventable
illnesses

Percentage
of Illnesses
That Are
Preventable

Source

2008

Hot Peppers

Salmonella Saintpaul

1,442

790

55%

2008

Cantaloupe

Salmonella Litchfield

2009

Alfalfa Sprouts

Salmonella Saintpaul

235

31%

2010

Shell eggs
Peanut Butter
and peanut
butter products

Salmonella Enteritidis

3,578

120

Salmonella
Typhimurium

636

188

30%

CDC (1)

Cantaloupe

Salmonella Javiana

163

15%

Appendix C

1,018

200

23%

53
3,578

1
790

2%
55%

E.coli O157: H7
E.coli O157: H7

29
11

4
1

14%
9%

Appendix C
Appendix C

E.coli O157: H7

Appendix C

E.coli O157: H7
Average
Minimum
Maximum

167
68
11
167

28
9
1
28

17%
11%
3%
17%

Appendix C

E.coli O26

34%

Appendix C

E. Coli O145
Average
Minimum
Maximum

26
28
26
29

0
5
0
10

0%
17%
0%
34%

Appendix C

139
139
139
139

69
69
0
69

50%
50%
0% 35
100%

Appendix C

20082009
2019

Percentage range of
cases prevented for
Salmonella

2012
2016
2018
2019

Spinach
Alfalfa sprouts
Romaine
Lettuce
Romaine
Lettuce

Percentage range of
cases prevented for
E.coli O157:H7
2012
2010

Clover sprouts
Romaine
Lettuce

Percentage range of
cases prevented for
other E.coli

2011

Cantaloupe

Percentage range of
cases prevented for LM

Average
Minimum
Maximum

Listeria
monocytogenes
Average
Minimum
Maximum

IFT report,
table 48 and
Appendix C

35 We assume zero for a lower bound because we only have one Listeria monocytogenes outbreak in this analysis.

Year

Commodity

2019

Basil
Leafy Greens,
Cilantro

2013

Percentage range of
cases prevented for
Cyclospora

Pathogen
Cyclospora
cayetanensis
Cyclospora
cayetanensis
Average
Minimum
Maximum

Total
Illnesses per
Epidemic
Curve

Percentage
of Illnesses
That Are
Preventable

Preventable
illnesses

241

Appendix C

631
436
241
631

146
78
9
146

23%
13%
4%
23%

Appendix C

Source

(1) https://www.cdc.gov/salmonella/2009/peanut-butter-2008-2009.html

We extend the IFT analysis by including additional pathogens and using multiple case
studies per pathogen – with the exception of Listeria monocytogenes – to estimate the average
percentage of preventable illnesses by pathogen.
After estimating the number of illnesses that may be prevented with better tracing, we
then multiply the percentage range of preventable illnesses from Table 8 by the estimated
number of annual illnesses for each pathogen in Table 9. The numbers of annual illnesses below
are from the baseline number of illnesses as described in Appendix B, adjusted for unspecified,
underreported, and undiagnosed illnesses (Ref. [19] [21]). We believe that accounting for such
cases is critical because not all illnesses caused by outbreaks are ultimately documented and
attributed to those outbreaks. This approach is consistent with FDA’s past regulatory impact
analyses . Nevertheless, we recognize the scope for uncertainty in these estimates and request
comment on appropriate scaling and adjustment for unattributed illnesses.
Table 9. Estimated Annual Cases of Foodborne Illness That Are Preventable with Product
Tracing Improvement
Estimated Annual Preventable Illnesses
Primary
Minimum
Maximum

Pathogen

Annual
Illnesses

Cyclospora cayetanensis

20,672

2,778

772

4,783

E. coli (STEC) non-O157

41,454

7,147

14,295
57

E. coli (STEC) O157

30,606

3,277

972

5,132

Listeria monocytogenes

Salmonella (non-typhoidal)

23,778

5,389

449

13,027

Subtotal

116,583

18,627

2,192

37,309

Unspecified unknowns25

466,332

74,509

8,769

149,236

Total

582,915

93,136

10,961

186,545

In Table 10, we estimate the burden of foodborne illnesses attributed to FTL foods by
multiplying the estimated total annual number of illnesses from Table 9 by the weighted average
burden per illness (based on the prevalence of preventable illnesses related to each pathogen).
We use the same burden of illness estimates for the selected pathogens as in section E.1.iii.
Hence, the columns in Table 10 rely on the primary, minimum, and maximum possible burden of
illness for the selected pathogens from our burden-of-illness model.
Table 10.- Annual Benefit Based on 84% Improved Product Tracing Time.
Annual Undiscounted Benefit from Tracing Time
Reduction by 84% to 6 Days (Based on Weighted
Burden per Illness)(1)(2)
Pathogen
Primary
Minimum
Maximum
Weighted burden per illness for four
pathogens
Total annual benefit from faster tracing

$9,257

$4,525

$11,048

$722,618,159

$41,570,269

$1,727,263,276

(1) The estimated range of values (primary, minimum and maximum) represent the variability in the valuation
of illness per pathogen.
(2) Foodborne Illnesses caused by Cyclospora Cayetanensis, E. coli (STEC), Listeria monocytogenes and
Salmonella (non-typhoidal).

For discounting purposes, we assume that benefits will begin to accrue in year two after the
publication date of this rule. 36

The costs for discounting purposes assume the publication date of the rule is year 0 and that costs will be incurred
in year 1 (which is one year before the 2-year effective date). Therefore, we assume benefits will begin to accrue on
year 2 (one year after requirements were implemented in year 1).

We estimate the percent improved traceback time resulting from better tracing
requirements using information provided by FDA’s CORE that includes a case study from the
2019 E. coli Romaine lettuce outbreak (Ref. [27]). The outbreak illustrates the difference in time
to identify implicated growers from points of sale (POS) where lot codes were available versus
POS where lot codes were not available. For POS where lot codes were available because
product packaging from a sample that tested positive and matched the outbreak strain was
available from an ill consumer, growers were identified within 24 hours compared to 29 days for
those where no lot code information was available (1 day over 29 days would represent a 96
percent improvement). Although every case is unique, this provides an example of how the
availability of lot code information at the POS could significantly shorten the time in
determining the source of the contaminated product.
Given FDA CORE’s combined years of experience in conducting traceback
investigations associated with foodborne outbreaks at a national level, it is FDA CORE’s expert
judgment that access to lot codes and other key data elements throughout the supply chain would
likely enable FDA to identify common product sources in about five to seven days, for an
average of six days (Ref. [27]). 37 Given that product packaging is often discarded by consumers
and not available to outbreak investigators, the five to seven days estimate assumes that the
product package would not be available. We use this information to estimate the resulting
percent improvement (reduction) based on the median number of days to reach maximum
improvement from our sample outbreaks.

This time period does not account for the time needed for the epidemiologic information (i.e., food exposures)
provided by public health officials to identify POS clusters and be provided to FDA for tracing. Additionally, this
timeframe may vary depending on the complexity of a food’s supply chain. For example, if a food is transformed
multiple times before it reaches an RFE, more time may be needed to identify source information. However, we
account for this in our analysis of outbreak examples as we used the CORE traceback initiation date which
represents a clear point in time when traceability activities began.

The average number of days used for identifying a product source without lot codes is
about 37 (ranging from about 0 to 86), whereas the average number of days needed to identify a
product source from a product with lot codes is six days (ranging between five and seven days).
We estimate that the percent improvement that would result from identifying common product
sources in five to seven days would range between 81% to 87%. We use the middle estimate of
six days which is equivalent to an 84% improvement (Table 10). We estimate that
corresponding (undiscounted) public health benefits would range between $42 million and $1.7
billion per year with a primary estimate of $723 million per year. These benefits are slightly
underestimated as the annual dollar value of the burden associated with outbreaks caused by
these four pathogens represents about 96% of the total annual burden of all FTL associated
illnesses. These benefits may also be overestimated due to uncertainty in adjustments accounting
for under reported, undiagnosed and unspecified illnesses from all FTL-associated illnesses. We
ask for comment on our estimation of health benefits and request comments that would inform
improved estimation of these benefits.
3. Benefits from Avoiding Overly Broad Recalls Quantified for Selected FTL Foods
In addition to the public health benefits discussed above, implementation of more precise
food recalls may result in social benefits realized by avoiding overly broad recalls when
contaminated FTL foods covered by the rule are identified. Market withdrawals and recalls are
expensive and commercial distribution of contaminated food can result in economic harm to
consumers. In the event of a food recall or market withdrawal, the records required by the
proposed regulation may help us to more quickly and accurately locate a violative product that
was commercially distributed, which would make conducting a broad recall unnecessary. Costs
of conducting a recall or market withdrawal include lost sales (lost retail value of product),

expenses associated with notifying retailers and consumers, collection and shipping costs,
disposal costs, and legal costs, among others. 38 In addition to costs of conducting a recall,
aggregate costs include spillover costs to shareholders, competitors, wholesalers, retailers and
customers. While well-established, profit-maximizing food manufacturers and distributors and
retailers may be able to consider in their decisions the costs associated with recalling a product
beyond the value of recalled units to include expenses associated with notifying retailers and
consumers, collection, shipping, disposal and legal costs; there are spillover or negative
externalities associated with a recalled product that may be larger in the aggregate than the losses
of the recalled product to the producer. 39
Although recall of rightly implicated foods is necessary and costly, broad recalls that
involve loosely related to unrelated products can make overall recalls extremely expensive.
According to a survey of companies conducted by the Grocery Manufacturers Association
(GMA), 77 percent of respondents that faced a recall in the past five years estimated the financial
impact of the recall to their company to be up to $30 million, with 23 percent reporting even
higher costs (Ref. [15]). The GMA study suggests that the average cost of recalls to their
members is about $2 billion over a five-year period or an annual loss of about $400 million.
These costs represent an estimated fraction of 0.3 percent of the GMA sample’s annual revenues
and result from business interruption, product disposal costs, customer reimbursement,
transportation, investigation, external professional fees, sanitizing production facilities,

38 One of the steps of a recall process involves disposing or destroying the recalled food product. While it may be
possible for some companies to recover costs by repurposing their recalled products into pet or animal feed or even
fertilizer, this practice is more common with meat producers. In the wake of foodborne illness outbreak-related
recalls, repurposing or diverting recalled foods to recover losses is not a conventional response within our review of
case studies and pilot projects. To the extent any repurposing does occur, the overall costs from lost retail sales
would be defrayed by the value of the repurposed products. We welcome comments on the frequency of such
responses by food supply system participants.
39 Jarrel, Gregg; Peltzman, Sam; The Impact of Product Recalls on the Wealth of Sellers. Journal of Political
Economy, Vol. 93, No. 3 (1985), pp 512-536.

warehouse costs, decreased sales of the brand name product identified, internal time, and other
expenses. This proposed rule, if finalized, may help ensure recalls are conducted in a more
precise manner and unnecessary costs to both industry and consumers are mitigated.
We hereby showcase three case studies from recent food recalls covered under the
proposed rule. The goal is to estimate the impact of avoiding overly broad food product recalls to
society, including the impacts on manufacturers, distributors and retailers of their products, other
firms in the market, and consumers as well as non-users. Better tracing could substantially
reduce the costs from widespread industry losses from an overly broad recall to smaller losses
from a targeted recall, particularly when it comes to the costs of lost sales of undifferentiated
recalled products like the 2008 tomato recall and the more recent 2018 and 2019 leafy green
recalls. Hence, implementation of this proposed rule may help reduce or eliminate the ‘spillover’ impact of recalling uninvolved brands, products or industry segments. This can also help
ensure that the volume of sales of unrelated products is not affected.

a) Multistate Outbreak of Salmonella Saintpaul Infections Linked to Raw Produce
This 2008 outbreak caused almost 1,500 illnesses and was initially attributed
to tomatoes, leading to a recall. The warning for the 2008 tomato recall covered all
red Roma, red plum and red round tomatoes and any other products containing these
raw, red tomatoes. 40 Consumers began to avoid not only the tomatoes included in the
warning, but also all other varietals of tomatoes as well. Even though the FDA
explicitly stated that some varieties were safe, many stores removed them from their

40 Press Release, FDA, FDA Warns Consumers Nationwide Not to Eat Certain Types of Raw Red Tomatoes (June
7, 2008) [hereinafter FDA Recall June 7].

shelves and customers began ordering their customary dishes at restaurants without
tomatoes. 41 Of the sixteen traceback investigations initiated by FDA, four were
discontinued due to lack of records and the remaining 12 tracebacks resulted in no
common growing region, grower, or supplier. Challenges to the tracebacks included
lack of standardized product documentation throughout the supply chain, difficulty in
linking incoming to outgoing shipments, repacking of product, and comingling of
tomatoes. Standardized traceability documentation and linking of shipments
throughout the distribution chain would have decreased the time to complete the
tracebacks and provided timely information that there was no common source of
tomatoes (Ref. [28]). Ultimately the source of the outbreak was later attributed to
jalapeño and serrano peppers produced in Mexico (Ref. [29]). The investigation
showed that jalapeño peppers were a major source of contamination and that serrano
peppers also were a source. 42 Although the recall of red tomatoes and tomato
products was later lifted, the negative impact on red tomatoes and tomato products
significantly affected their sales volumes at the time. In fact, costs to the Florida
tomato industry alone were estimated to be more than $100 million. In Georgia, the
costs to the tomato industry came close to $14 million. 43

41 Salmonella scare hold the tomato, Chicago Tribune (Illinois), June 10, 2008.
https://www.chicagotribune.com/news/ct-xpm-2008-06-10-0806090798-story.html
42 Multistate Outbreak of Salmonella Saintpaul Infections Linked to Raw Produce (FINAL UPDATE)
Centers for Disease Control and Prevention, National Center for Emerging and Zoonotic Infectious Diseases
(NCEZID), Division of Foodborne, Waterborne, and Environmental Diseases (DFWED) Posted August 28, 2008
https://www.cdc.gov/salmonella/2008/raw-produce-8-28-2008.html
43 Reginald L. Brown testifying before the House Committee on Energy and Commerce, Subcommittee on
Oversight and Investigations, The Recent Salmonella Outbreak: Lessons Learned and Consequences to Industry and
Public Health, 110th Cong. 2nd sess., July 31, 2008, http://energycommerce.house.gov/cmte_mtgs/110-oihrg.073108. Brown-Testimony.pdf; “FDA tomato alert costly to Georgia producers”
Southeast Farm Press, September 4, 2008, http://southeastfarmpress.com/vegetables-tobacco/salmonella-warning0905/index.html.

b) Outbreak Investigation of E. coli: Romaine (November 2018- February 2019)
On November 20, 2018, FDA issued a public advisory in response to a multistate outbreak of E. coli O157:H7 linked to romaine lettuce and advised against eating
any romaine lettuce on the market at that time. As a result, producers and distributors
voluntarily withdrew the product from the market. FDA performed a traceback
investigation to determine the source of the romaine lettuce; however, at the time of
the public advisory, FDA did not have enough traceback information to identify the
source of the contamination that would allow conducting a targeted recall. The most
efficient way to ensure keeping contaminated romaine off the market was for industry
to voluntarily withdraw product from the market, and to withhold distribution of
romaine while FDA and state partners conducted a traceback investigation to
determine whether a common supplier or source of contamination could be identified.
By December 13, 2018, FDA was able to refine the traceback investigation
implicating one farm in Santa Barbara which promptly recalled red leaf lettuce, green
leaf lettuce and cauliflower harvested on November 27, 2018, through November 30,
2018. 44 On February 13, 2019, FDA completed its investigation. At the conclusion of
the outbreak, a total of 62 cases (with 25 hospitalizations and no deaths) in 16 states and
Washington DC were associated with this outbreak. 45 Better traceback data would have

allowed FDA to identify the implicated farm in Santa Barbara more quickly, such that
a broad market withdrawal of all romaine lettuce might not have been necessary.

44 https://www.fda.gov/food/outbreaks-foodborne-illness/outbreak-investigation-e-coli-romaine-november-2018
45 People Infected with the outbreak strain of E. Coli O157:H7, by date of illness onset.
https://www.cdc.gov/ecoli/2018/o157h7-11-18/epi.htmlhttps://www.cdc.gov/ecoli/2018/o157h7-11-18/epi.html

c) Outbreak Investigation of E. coli: Romaine (November 2019- January 15, 2020)
In November 2019, FDA, along with CDC, U.S. Department of Agriculture’s
Food Safety Inspection Service (FSIS) and state health authorities, investigated an
outbreak of 167 illnesses of E.coli O157:H7 associated with salads containing
romaine lettuce. The Maryland Department of Health identified a positive sample of
romaine lettuce used in a chicken Caesar salad kit. The contaminated romaine lettuce
was supplied by farms in Salinas, CA. As a result of this positive sample, all chicken
Caesar salad kits containing the positive lot were recalled. Simultaneously, FDA was
investigating two additional outbreaks of E. coli O157:H7 associated with romaine
lettuce and a chopped salad kit containing romaine lettuce. Based on traceback data
available initially, FDA requested that industry voluntarily withdraw romaine grown
in Salinas from the market and is requesting that industry withhold distribution of
Salinas romaine for the remainder of the growing season in Salinas. This was a broad
market withdrawal, because a significant portion of the romaine lettuce consumed in
the United States is grown in Salinas; however, due to a lack of more specific
traceback information, this was the most efficient way to ensure that contaminated
romaine was off the market (Ref. [27]).
We therefore define as a benefit the foregone recall costs incurred by manufacturers,
producers, distributors and retailers from inadvertently being part of a broad scope or
undifferentiated product recall. To provide some examples of the magnitude of such potential
benefits, we estimate the foregone recall costs for three commodities: Romaine lettuce, fresh
salad kits, and fresh tomatoes. We use supermarket scanner data to estimate average sales per
product during a 4-week sales period and a 13-week sales period. We assume the estimated cost

of one product recall is based on the average foregone sales for said product during the recall
period (Columns A and B in Table 11). 46 We choose the time periods assuming four weeks is the
shortest length of a class I recall (our best-case scenario) and 13 weeks as the longest length for a
class I recall (our worst-case scenario) 47. We then use the same data to estimate the sum of
product sales within the same product class for the same four- and 13-week periods (Columns C
and D in Table 11). We estimate the difference in sales between a targeted recall versus a broad
recall as the value of foregone product sales during one recall event (Columns E and F in Table
11). The low and high difference in retail sales in columns E and F represent the estimated low
and high foregone sales of products that were not sold due to a recall.
Table 11. Difference in retail costs between one average product and sum of product
category sales (2018 $)
Category

Average Product Sales
(A) Low
estimate *

(B) High
estimate **

Total Category Sales
(C) Low
estimate *

Pre-cut
$358,396.38
$1,176,222 $374,882,618
Fresh Salad
Mix
Fresh
$538,640
$1,768,684
$73,254,986
LettuceRomaine
Fresh
$111,534
$403,500 $118,226,323
Tomatoes
Jalapeño &
$4,970
$19,180 $118,226,323
Serrano
Peppers
Sum
$1,013,541
$3,367,586 $684,590,249
*Low estimate based on 4-week sales ending 09-27-2014
**High estimate based on 13-week sales ending 09-27-2014

Retail Sales Difference

(D) High
estimate **

(E) Low
estimate *

(F) High
estimate **

$1,230,328,243

$374,524,222

$1,229,152,021

$240,541,021

$72,716,346

$238,772,337

$427,710,361

$118,114,788

$427,306,860

$427,710,361

$118,221,352

$427,691,181

$2,326,289,985

$683,576,709

$2,322,922,399

46 A.C. Nielsen Scantrack data for 52 weeks ending on 09-27-2014.
47 Luis A. Ribera, Marco A. Palma, Michel Paggi, Ronald Knutson, Joseph G. Masabni, and Juan Anciso,
Economic Analysis of Food Safety Compliance Costs and Foodborne Illness Outbreaks in the United States, Hort.
Technology, Vol 22. No. 2 (2012), pp 150 – 156.

To arrive at total costs of a broad recall (columns G and H in Table 12), we multiply the
low and high values from columns E and F in Table 11, above, by 2.5 to account for the
estimated number of produce outbreaks per year. 48 We do not account in this analysis for costs
of a recall beyond the retail value of the recalled product and request comments on this estimate.
49, 50

Total costs of a broad recall in columns G and H include spillover costs of a broad recall,

including the amount over what a narrower recall cost should have been (columns I and J in
Table 12). The difference between conduct of broad recalls and conduct of targeted recalls is
estimated in columns K and L in Table 12 as the difference between total costs of a broad recall
(columns G and H) and costs of a narrower recall (columns I and J).
Table 12. Difference Between Broad and Narrow Recall Costs (millions 2018 $)
Category

Pre-cut Fresh
Salad Mix
Fresh LettuceRomaine
Fresh
Tomatoes
Serrano
Peppers
Sum
(undiscounted)

Total costs of a
broad recall

Total costs of
narrower recall

Social benefit of a targeted recall

(G) Low
estimate
*

(H) High
estimate
**

(I) Low
estimate
*

(J) High
estimate
**

$1,049

$3,442

$1,048

$3,440

$204

$669

$203

$666

$331

$1,196

$331

$1,196

$331

$1,198

$331

$1,198

$1,914

$6,504

$1,913

$6,499

(K) Low
estimate *

(L) High estimate **

48 From FDA outbreak data we estimate an average of approximately 7.3 leafy green/fresh cut related outbreaks per
year and three produce related outbreaks per year over an 11-year period (from 2010 to 2019). We also estimate an
average of two outbreaks from tomato, romaine, and other leafy greens combined over the same 11-year period
using publicly available data. We use the median calculation between two (using publicly available data) and three
(from FDA data) to obtain 2.5 outbreaks per year.
49 Ashley Kern, Dari Duval, George Frisvold, Arizona Leafy Greens: Economic Contributions of the Industry
Cluster, 2015 Economic Contribution Analysis, Agricultural & Resource Economics-University of Arizona: College
of Agriculture and Life Sciences, (2017).
50 In addition to the value of recalled units, costs associated with a recall would have included expenses associated
with notifying retailers and consumers, collection, shipping, disposal and legal costs among others. On the other
hand, some portion of total sales value is not lost in a recall situation; for example, part of a grocery item’s price
represents the opportunity cost of shelf space, and that shelf space can be re-allocated to non-recalled product.

Implementation of the proposed regulation would likely result in a reduction of the number
of recalled or withdrawn products. These estimates only account for the annual difference in
costs of avoiding broad recalls for three commodities for which we have reliable data and
therefore represent the minimum range of potential benefits from this rule. Additional estimated
benefits from reduced recall costs could be expected by including estimates for other recalled
commodities. Other benefits from replacing broad recalls with more targeted recalls would
include reduced costs and resources used by government in terms of conducting outbreak
investigations and recalls for a broader set of products beyond those implicated in an outbreak or
a recall. There are no benefits from removing unimplicated products from the market.
It is possible, but not certain, that both these narrowed-recall benefits and the health and
longevity benefits discussed earlier separately or jointly could be experienced to the extent
quantified in this regulatory impact analysis. It is also possible, on the other hand, but not certain,
that a given instance of baseline contamination would lead to a very broad recall (that could be
narrowed by the proposed rule) or to illnesses (that could be avoided due to the proposed rule) —
but not both. We request comment on these estimates.

4. Other Benefits
There are additional benefits that may arise if this proposed rule is finalized. These
include:
•

Avoidance of costs due to unnecessary preventive actions by consumers;

•

More expedient initiation and completion of recalls;

•

Improvements in supply chain management and inventory control;
68

•

Other supply system efficiencies due to a standardized approach to traceability,
including an increase in transparency and food system trust among food supply
system participants and potential deterrence of fraud.

Broad and long recall windows and not promptly initiated recalls and market withdrawals
and advisories result in additional costs to FDA and industry and may force consumers to take
unnecessary preventive actions. Such costly preventive measures by consumers may involve
throwing away food, stopping their consumption of the suspect food item, or visiting physicians
or emergency rooms to determine if they have been exposed to a pathogen.
Without having prompt access to records that provide timely information, FDA, state,
and local authorities might spend additional time and resources on tracing the source of an
outbreak, initiating broad recalls, and communicating to industry and consumers. The traceability
records required by this rule may help in initiating recalls earlier and shortening recall periods,
which can significantly reduce the costs associated with management and support of recall
activities. Shorter recall periods may also mitigate the loss of future product sales by shortening
the length of negative publicity these products received because of longer recalls.
Another benefit that may be realized should this rule be finalized relates to improved rate
of recovery for contaminated products. When recall orders are issued and include a precise list of
products suspected to be contaminated, the rate of recovery of this product may be substantially
improved under traceability rule. The ability to precisely identify the lot numbers, production
dates and other related information may help those collecting the product achieve a higher
recovery rate than would be otherwise achievable in situations where robust traceability is
unavailable. The Blue Bell ice cream recall is a good example. Since it took so long to be able to
identify the contaminated products and the recalls were issued in piecemeal, the rate of recovery

was only less than ten percent. Of the 492,611,698 Blue Bell products recalled because of
Listeria contamination, only 39,099,448 (or about 8 percent) were recovered and successfully
destroyed (Ref. [30]).
Finally, other benefits may include supply chain management improvements, increase in
transparency and food system trust, and potential deterrence of fraud. The latter benefits may be
realized through a reduction in fraudulent activity or an exit from the market by some because of
increased accountability due to more standardized, improved traceability.

F. Costs of the Proposed Rule
We considered baseline food traceability practices of all covered entities (firms or
establishments) and estimated the incremental costs related to changes in these practices required
by the proposed rule, if finalized. We estimate the following costs of the proposed rule to
industry:
•

one-time labor costs of reading and understanding the rule;

•

one-time capital investment costs of redesigning or upgrading systems to establish,
maintain, and transmit records;

•

one-time costs of training personnel to implement new recordkeeping systems and
establish and maintain records; and

•

recurring costs of establishing, maintaining, retaining, and transmitting additional records
required by this proposed rule.

For discounting purposes, we assume the publication date of the rule is year zero and that costs
will be incurred in year one (which is one year before the two-year compliance date) 51.

1. Main Assumptions of Cost Analysis
We estimate that the costs of this proposed rule, if finalized, would largely arise from
recordkeeping requirements, which represent new and substantive responsibilities for covered
entities. Using the current version of the FTL, we identify the entities and food supply system
sectors affiliated with covered foods. Because each provision of the proposed rule would require
different entities within the food supply system to establish and maintain different sets of records
and key data elements, we list estimated costs by provision. First, all covered entities must
establish and maintain the general traceability records as outlined in § 1.1315. The following
provisions, described in more detail in the sub-sections below, apply differently to entities
depending on whether they grow, receive, transform, create, and/or ship foods on the FTL. If an
entity performs more than one of these activities, it must keep the records required by all relevant
provisions.
The total cost of each statutory requirement thus depends on the number of entities
affected and the additional burden of each requirement relative to baseline practices. In this
proposed rule, FDA specifies the records and information a covered entity must keep but does
not specify the form or system in which those records should be maintained. We expect that, to
the extent possible, firms would satisfy new recordkeeping requirements using their existing
systems or invest in upgraded systems or technology. Furthermore, we assume that firms would

51 The year of publication is year zero and the effective date is year two. For manufacturers to comply with the
requirements of this rule by the effective date (year two), we assume they will begin to incur compliance costs in
year one.

comply with new recordkeeping requirements by modifying existing shipping or purchase
records such as Advance Shipping Notices, Bills of Lading, Invoices, or Purchase Orders.
Moreover, we recognize that some entities may already follow certain recordkeeping
practices as part of their baseline business practices. These entities include those covered by the
Subpart J provisions as well as those following certain industry conventions and best practices.
The likelihood of an entity already adhering to certain recordkeeping practices depends on its
size, industry, and position in the food supply system. Because we lack data to separate the
entities already in full or partial compliance from those who must come into compliance, our
analysis uses a range to account for this uncertainty.
Finally, as discussed in the baseline section and in further detail below, certain entities
covered by the proposed rule may be exempt or partially exempt from new recordkeeping
requirements. From all our cost estimates below, we exclude those farms and other small
originators that would be fully exempt from the proposed rule because of their very small size.
Some entities or types of products are partially exempt, which we describe in further detail
below. Of these partially exempt entities, we exclude farm-to-school and farm-to-institution
entities from our cost estimates because we do not expect these entities to change their current
practices due to the proposed rule. We do not consider in our cost estimates how many entities
covered by the proposed rule transact partially exempt foods (for example, those sold from farms
directly to retail food establishments) because we lack such detailed data.
We conduct our analysis by provision to parallel the structure of the proposed rule. It
covers a ten-year horizon following the proposed rule’s compliance date, and we assume all onetime costs occur in year one. All wage rates come from the Bureau of Labor Statistics,
Occupational Employment Statistics (OES), May 2018, National Industry-Specific Occupational

Employment and Wage Estimates. Hourly wage rates are estimated by doubling the OES
reported mean wage to account for benefits and overhead, in accordance with Department of
Health and Human Services guidelines (Ref. [25]).
2. Costs of Reading and Understanding the Rule
Covered entities would incur one-time costs at the firm level to read and understand the
proposed rule. To estimate the number of firms affected by this provision, we use the 2016
SUSB data (Ref. [31]). We first identify the 2012 NAICS categories of firms that likely
manufacture, process, pack or hold foods on the FTL (Table 13). Because not all firms in each
NAICS category would be affected by this proposed rule, we remove non-covered firms namely
exempt firms and those that we estimate do not likely manufacture, process, pack or hold foods
on the FTL. We estimate that 422,144 firms would be affected by the proposed rule and would
thus need to read and understand the rule, if finalized. Our lower and upper bounds range from
224,679 to 636,809 firms. We request comment on these estimates.
Table 13. NAICS Categories Likely to Manufacture, Process, Pack, or Hold Foods on the
Food Traceability List
2012
Category
NAICS
111219 Other Vegetable (except Potato) and Melon Farming
111339 Other Noncitrus Fruit Farming
111419 Other Food Crops Grown Under Cover
112310 Chicken Egg Production
112511 Finfish Farming and Fish Hatcheries
112512 Shellfish Farming
114111 Finfish Fishing
114112 Shellfish Fishing
311340 Nonchocolate Confectionery Manufacturing
311351 Chocolate and Confectionery Manufacturing from Cacao Beans
311352 Confectionery Manufacturing from Purchased Chocolate
311411 Frozen Fruit, Juice and Vegetable Manufacturing
311412 Frozen Specialty Food Manufacturing
311421 Fruit and Vegetable Canning
311422 Specialty Canning
73

311423
311513
311520
311710
311811
311812
311813
311821
311824
311911
311941
311942
311991
424410
424420
424430
424450
424460
424480
424490
445110
445120
445220
445292
445230
445291
445299
447110
452910
454111
454210
493110
493120
493130
722310
722320
722330
722410
722511
722513
722514
722515

Dried and Dehydrated Food Manufacturing
Cheese Manufacturing
Ice Cream and Frozen Dessert Manufacturing
Seafood Product Preparation and Packaging
Retail Bakeries
Commercial Bakeries
Frozen Cakes, Pies, and Other Pastries Manufacturing
Cookie and Cracker Manufacturing
Dry Pasta, Dough, and Flour Mixes Manufacturing from Purchased Flour
Roasted Nuts and Peanut Butter Manufacturing
Mayonnaise, Dressing, and Other Prepared Sauce Manufacturing
Spice and Extract Manufacturing
Perishable Prepared Food Manufacturing
General Line Grocery Merchant Wholesalers
Packaged Frozen Food Merchant Wholesalers
Dairy Product (except Dried or Canned) Merchant Wholesalers
Confectionery Merchant Wholesalers
Fish and Seafood Merchant Wholesalers
Fresh Fruit and Vegetable Merchant Wholesalers
Other Grocery and Related Products Merchant Wholesalers
Supermarkets and Other Grocery (except Convenience) Stores
Convenience Stores
Fish and Seafood Markets
Confectionery and Nut Stores
Fruit and Vegetable Markets
Baked Goods Stores
All Other Specialty Food Stores
Gasoline Stations with Convenience Stores
Warehouse Clubs and Supercenters
Non-store Retailers: Electronic Shopping
Vending Machine Operators
General Warehousing and Storage
Refrigerated Warehousing and Storage
Farm Product Warehousing and Storage
Food Service Contractors
Caterers
Mobile Food Services
Drinking Places (Alcoholic Beverages)
Full-Service Restaurants
Limited-Service Restaurants
Cafeterias, Grill Buffets, and Buffets
Snack and Nonalcoholic Beverage Bars
74

The preamble and draft regulatory text of the proposed rule contains approximately
44,000 words. Per HHS guidelines on reading speed (Ref. [25]), we estimate that an adult reads
200 to 250 words per minute, with an average speed of 225 words per minute. We divide the
number of words in the preamble and codified by reading speed, producing an estimate of 3.3
hours (= 44,000 / 225 / 60) to read the rule, with a lower bound of 2.9 hours (= 44,000 / 250 / 60)
and an upper bound of 3.7 hours (= 44,000 / 200 / 60). We expect that one employee responsible
for reading the rule would be a supervisor or manager and request comment on the number of
employees who might read the rule per firm. The mean wage of a supervisor or manager is
$18.75 per hour, doubled to $37.50 to account for benefits and overhead. We obtained this wage
information from the 2018 BLS wage data for Office and Administrative Support Occupations.
We thus estimate that the one-time labor cost of reading the rule is approximately $122
per covered firm (= $37.50 x 3.3), with a lower bound of $110 (= $37.50 x 2.9) and an upper
bound of $138 (= $37.50 x 3.67). For all firms covered by the provision, the total labor cost of
reading the rule is $51.6 million (= 422,144 x $122), with a lower bound of $24.7 million (=
224,679 x $110) and an upper bound of $87.6 million (= 636,809 x $138). We request comments
on these estimates. Table 14 summarizes the estimated costs of reading and understanding the
proposed rule.

Table 14. One-time Costs of Reading and Understanding the Proposed Rule (2018$)
Primary
Low
High
422,144
224,679
636,809
Number of affected firms
Average reading speed (words per minute)
225
250
200
Word count of proposed preamble and codified
44,000
44,000
44,000
$37.50
$37.50
Employee hourly wage rate
$37.50
Hours to read
3.3
2.9
3.7
Per firm cost
$122
$110
$138
Total one-time cost of reading and
understanding the rule
$51,595,417 $24,714,642 $87,561,214
75

3. Costs of Capital Investment
We expect the number of firms that may incur costs from capital investment to equal to
the number of firms required to read and understand the proposed rule. However, as discussed
above, some entities may be able to comply without additional capital investments, while others
would need to invest in traceability-related capital. Case studies of prior traceback efforts in
2012 and earlier show a wide range of existing tracing capabilities across sectors and firm size
(Ref. [5]). For example, according to the 2012 IFT study of pilot projects for improving product
tracing, selected large growers, distributors, and processors already had the capacity to scan
KDEs, while only 30-50 percent of selected small distributors, small processors, and large
retailers had this capability at that time. Among the 22 entities surveyed, those that reported
standardized naming capabilities ranged from 0 to 100 percent, indicating that some companies
had no standardized names, while others had partially or fully standardized nomenclature. In
general, traceability technologies, adoption, and implementation have continued to expand since
2012, supported by efforts in multiple industries to integrate standards like those of GS1 into
common practice. 52
Capital investment may include food traceability software, scanners or barcode readers,
barcode printers, and increased data storage (hard disk or cloud storage) to handle the increased
recordkeeping requirements of the proposed rule. These investments would depend on a firm’s
size, role in the supply chain, products, and existing traceability systems, as well as whether the
firm decides to adopt an electronic recordkeeping system as a result of this rule (although the
rule does not require electronic maintenance of records).

52 See, for example, the Produce Traceability Initiative (https://www.producetraceability.org/), the International
Dairy-Deli-Bakery Association (https://www.iddba.org/initiatives/industry-initiatives/food-traceability), and the
National Fisheries Institute (https://www.aboutseafood.com/traceability-and-sustainability-in-the-supply-chain-4/).

The following table summarizes baseline costs of operating product traceability systems
and the additional investments related to specific system improvements, as reported by the 2012
pilot project participants (Ref. [5]). 53 They reported that the total costs for systems currently in
place to capture KDEs at baseline range from tens of thousands to millions of dollars. The costs
of additional improvements in traceability would range from no cost to hundreds of thousands of
dollars.
Table 15: Current and additional costs of traceability-related investments, as reported by
selected entities (Ref. [5])
Large
Large
Small
Large
Small
Large
Type of firm
Grower
Processor
Processor Distributor Distributor
Retailer
Number of
firms surveyed
3
4
3
4
4
4
Baseline cost
to capture
KDEs (manual $350,000$500,000$250,000- $50,000$4.5M
$1.2M
$800,000 $1M
$40,000-$1.5M Unknown
or electronic)
Additional costs by activity
Incoming
KDEs by
electronic data
messages
Supply chain
link
Standardized
naming
Outgoing
KDEs
electronic to
customers
Provide data
summary

Unknown

$0-$15,000

Unknown

$0-$65,000

Unknown
$0$60,000/year

Unknown

$0-$500,000

Unknown

$0-$80,000

$0-$150,000
$5,000$150,000

$2,000$5,000

Unknown

N/A

Unknown

$0-$2,000

$0-$10,000

N/A

Unknown

For further presentation of cost estimates that are higher than those discussed elsewhere in this document, please
see column (b) of Tables 3a and 3b.

Incoming lot
number
information

N/A

$0$60,000/year

$0-$150,000

Unknown

These estimates reported above provide some insight into the types of capital investment
costs surveyed companies needed to improve traceability systems in 2012, such as conversions to
standardized naming and provision of data to supply chain partners. Importantly, not all entities
expected to make the same types of investments, and some entities, like large distributors,
already had the capability to integrate additional traceability practices into existing systems.
While we expect such variability to exist today, we ask for comments on capital investment costs
that would be required as a result of this proposed rule.
There are several important considerations for interpreting the costs from this report and
applying them to the proposed rule. First, the report surveyed selected companies about costs to
update system-wide traceability operations. Companies may have also accounted for training and
system integration costs, in addition to capital, in their estimates. Moreover, these estimates
pertained to pilot projects using specific food products, tomatoes and processed foods. Because
the proposed rule identifies a specific list of foods for enhanced tracing, we expect that
companies are unlikely to overhaul their existing systems; in other words, the incremental costs
of complying with the proposed rule may be significantly less than the cost of a total system
upgrade, particularly because we believe many traceability systems are already in place.
Second, entities surveyed in the report are not fully representative of sectors and sizes
affected by the proposed rule. Most entities reported that they invested in and improved their
tracing systems in the five years prior to the 2012 report (Ref. [5]). The reported costs came from
large companies, which likely rely on relatively sophisticated systems. For example, the large
growers surveyed were in the top two percent of the industry in terms of revenue, and no costs
78

were provided for small and medium growers. The survey also excludes small retailers, which
represent the majority of entities affected by the proposed rule. The study contacted fifteen small
businesses from unreported sectors, most of which reported that their existing tracing capabilities
sufficed and that additional investments would be costly. Others had already invested in tracing
networks that rely on standardized nomenclature.
Third, competition among providers of traceability products and services is likely to
produce lower costs over time (Ref. [32]). Indeed, since the time of the 2012 pilot project report,
technology and vendors have evolved to provide new traceability solutions, software, and offthe-shelf packages that were not available to companies until recent years. Today companies
have access to new tools like web-based systems and blockchain subscriptions. 54 For example,
web and mobile platforms are now able to streamline links between retailers and their wholesale
suppliers. 55 Equipment has also become more accessible and lower-cost. As an indication of this
trend, compared to the reference year of 2007, the consumer price index for computers,
peripherals, and smart home assistants decreased from 74 percent in December 2010 to 38
percent in January 2020. 56
Due to these considerations, we assume that larger entities, such as distribution centers,
are more likely to already have sophisticated systems that can adapt to accommodate foods
affected by the proposed rule. These entities may face incremental costs related to this system
integration as well as increases in printing needs and additional equipment. We assume that these
costs for larger firms are likely to be a fraction of the ranges reported in the table above.

54 See, for example, IBM’s Food Trust: https://www.ibm.com/blockchain/solutions/food-trust.
55 See, for example, BlueCart: https://welcome.bluecart.com/.
56 U.S. Bureau of Labor Statistics, Consumer Price Index for All Urban Consumers: Computers, Peripherals, and
Smart Home Assistants in U.S. City Average [CUSR0000SEEE01], retrieved from FRED, Federal Reserve Bank of
St. Louis; https://fred.stlouisfed.org/series/CUSR0000SEEE01, June 18, 2020.

To better understand potential incremental capital investment costs related to the
proposed rule for small- and medium-sized firms, as opposed to complete system upgrades for
large firms described above, we compiled information on selected products from the literature
and vendors in the table below. We ask for comments and request information that could help us
improve our estimates.
Table 16: Examples of incremental capital investment costs related to the proposed rule
Cost (Mejia et al., 2010)
Cost (2013-2020)
Type of capital investment
*
Non-RFID mobile and desktop printers
$400-$1,500
$140-$1,500 abc
RFID mobile and desktop printers
$400-$1,500*
$1,500 - $5,100 bc
Industrial printers
N/A
$180-$7,100 abd
Labels
$0.005-$0.255/label
$0.004-$0.006/label abe
Handheld mobile computers
$800
$13-$7,000 abc
Barcode scanners
$400-700
$15-$2,000 abc
Software licenses and data hosting for
$1,500-$5,000
$1,500-$5,000 bfc
small-medium company (bundled and
web-based)
Barcode printing software
$90
$0-$30 bg
*
Range does not distinguish between non-RFID and RFID.
a
https://www.thermalprinteroutlet.com/; b https://www.barcodegiant.com/; c
https://lowrysolutions.com/; d https://www.propacksolutions.com/; e GS1 (2013); f
https://pricingnow.com/; g http://www.ptiprint.com/

This table is intended to provide estimates of primary capital expenditures; we lack the
information that would allow us to compile an exhaustive list of itemized costs. As noted above,
we believe not all entities would need to make the same types of investments, depending on their
sector and size. For example, we expect that independent retailers and restaurants may be
increasingly reliant on their suppliers to capture much of the information required by the
proposed rule; we request comment on this assumption (U.S. Food and Drug Administration,
2020). Moreover, some entities may already have the resources they need to comply with the
proposed rule while others will incur additional costs.

Nevertheless, the table shows that the range of costs has widened over time due to greater
product variety, including more affordable options at the low end. For example, (Ref. [33])
report that barcode scanners once ranged from $400-$700; in 2020, scanners ranged from $15$2,000, depending on the level of sophistication needed. Costs of other products, such as
software licenses and data hosting, remained relatively constant over time though these products
have likely improved in quality. Some companies may also have access to free or paid updates to
existing software. We believe necessary investments are likely to vary depending on supply
chain role. For example, we expect retailers would likely need mobile computers and scanners
but not labeling equipment, while growers, transformers, and shippers would likely need labeling
equipment as well.
To comply with the proposed rule, we assume that smaller operations involved may
require relatively limited additional investments, such as a desktop shipping label printer and a
scanner connected to a laptop with label software (Ref. [34]). Using the sources in Table 16
above, a small entity could purchase a standard printer, scanners, and labels for roughly $300,
which could be coupled with free web-based barcode software. A standalone traceability bundle
included label design software with license, label printing, barcode printer, and starting labels for
approximately $1,000. 57
Using the sources above, a medium-sized entity could purchase an industrial printer,
mobile data capture computer, network label print server, and labels for approximately $7,000.
For roughly $6,500, a firm could purchase a kit including a mobile computer, barcode printer, set
of labels, software and licenses. 58 In 2020, licenses for web-based enterprise management

57 https://www.redlineforproduce.com/products/low-cost-label-compliance
58 https://b2b.lowrysolutions.com/shop/product/aidc-pos/pos-solution-bundles/wasp-barcodetechnologies/633808391362/24529041/1

systems that capture, order, receive, transform, and ship ranged from $1,500 to $5,000. (Ref.
[35]) reported that a small-medium grower could expect to make $3,000 in hardware investments
and $2,000-$3,000 in software and licensing to upgrade traceability systems.
To account for this considerable uncertainty in investments by firm and sector, we
assume a wide range of existing tracing capabilities and treat non-restaurants and restaurants
differently. For non-restaurants, we assume that 50 percent of covered firms would incur capital
investment costs, with a lower bound of 25 percent and an upper bound of 75 percent. For
restaurants, we assume that most would rely on suppliers to provide the necessary records, such
that only 5-15 percent of restaurants would need to make their own capital investment as a result
of rule, if finalized. We therefore estimate that 96,644 firms would incur costs of capital
investment, with a lower bound of 21,547 firms and an upper bound of 229,597 firms. We
request comment on these estimates of existing tracing capabilities. We also request comment on
our approach to estimate costs at the level of firm, rather than establishment.
Because the majority of entities affected by the proposed rule are small- and mediumsized firms, we assume that the overall range of additional investment needed skews toward
these firms. Considering variation in needs across firms’ sector, size, and existing capabilities
and synthesizing recent industry reports and vendor estimates with input from subject matter
experts (Ref. [34]), we assume that an affected covered firm would spend between $500 and
$25,000 on all additional capital investments to comply with the proposed rule, with a primary
estimate of $7,500. We recognize that there is scope for substantial variability in investment and
request comment on these assumptions. Specifically, we request comment on capital investment
costs by firm sector, size, and existing capabilities. We also note that, in subsequent sections, this

analysis accounts separately for labor costs and costs of establishing and maintaining records for
each provision.
To obtain the one-time capital investment costs of the proposed rule, we multiply the
capital investment cost per firm by the total number of affected firms. The one-time capital
investment costs of the proposed rule are thus $724.8 million (= 96,644 x $7,500), with a range
of $10.8 million (= 21,547 x $500) to $5.7 billion (= 229,597 x $25,000). Table 17 presents a
summary of the estimated capital investment costs of the proposed rule.

Table 17. One-time Capital Investment Costs of the Proposed Rule (2018$)
Primary
Low
High
96,644
21,547
229,597
Number of affected firms
Per firm cost
$7,500
$500
$25,000
$724,832,681 $10,773,308 $5,739,919,519
Total one-time capital investment costs

4. Costs of Training in New Traceability Practices
All firms covered by the general records provision would incur one-time costs to train
employees and managers in new food traceability practices. These costs may include hiring a
food traceability subject matter expert to train staff in person or via web seminars and the costs
of producing training materials. We assume that the number of firms affected by these training
costs is equal to the number of firms affected by capital investment costs. We therefore estimate
that 96,644 firms would incur these costs, with a lower bound of 21,547 firms and an upper
bound of 229,597 firms.
The labor cost of training depends on its duration and number of participants. Previous
case studies from 2012 contain little information on the labor cost of training employees on new
traceability systems and practices (Ref. [5]). One major food service chain with 20,000
restaurants reported a training cost of $100 per restaurant, but did not specify what this training

involved. One software provider reported a training cost of $1,000 per day, stating that the
number of days could vary based on the size of the operation and nature of changes to processes.
Other entities did not include separable training costs when reporting costs of capital investment,
suggesting that the labor cost of training may be unknown, included in capital investment (such
as a software package that includes training), or that new traceability training may replace
current training that becomes obsolete.
We therefore assume that some covered firms may need to conduct new education of key
personnel in traceability practices. This training will prepare firms to adapt their systems and
identify necessary capital investments. We expect operational changes on a day-to-day basis to
be disseminated through ongoing meetings and trainings, such that firms will not face additional
costs of training all employees. We request comment on this assumption.
We use an industry-standard online training 59 as a proxy for the cost of additional
training and assume that between one and five key supervisors or managers responsible for
guiding traceability practices per firm will undergo this training. We recognize that needs may
vary across firms and request comment on this assumption. For a half-day training, the cost
ranges from $450 for one employee to $2,125 for five employees, with a primary estimate of
$1,350 for three employees. We also account for the cost of labor time spent in training. Using
the 2018 BLS wage data for Office and Administrative Support Occupations, the time cost of
labor is equal to the hourly wage of these employees is $37.50, doubled to account for benefits
and overhead (= $18.75 x 2).
Accounting for the cost of training as well as the labor cost of trainees, we estimate that
the one-time training costs per firm are approximately $1,800 (= $1,350 + (4 x 3 x $37.50)), with

We use the GS1 Standards for Food Traceability Online Certificate Course.

a lower bound of $600 (= $450 + (4 x 1 x $37.50)) and an upper bound of $2,875 (= $2,125 + (4
x 5 x $37.50)). The total one-time training costs of the proposed rule for all firms are thus $174
million (= 96,644 x $1,800), with a lower bound of $13 million (= 21,547 x $400) and an upper
bound of $660 million (= 229,597 x $2,875). We request comments on these estimates. Table 18
presents a summary of the estimated training costs of the proposed rule.

Table 18. One-time Costs of Training in New Traceability Practices (2018$)
Primary
Low
High
96,644
21,547
229,597
Number of affected firms
Hours to train
4
4
4
Cost of training
$1,350
$450
$2,125
Trainee (employee) hourly labor cost
$37.50
$37.50
$37.50
Number of trainees per firm
3
1
5
Per firm cost
$1,800
$600
$2,875
$173,959,861
$12,928,972
$660,083,977
Total one-time training costs

5. Recordkeeping Requirements
The proposed rule, if finalized, would require certain persons to establish and maintain
records related to the growing, receiving, transforming, creating, and shipping of FTL foods. We
estimate the new costs of each recordkeeping requirement, taking into account such persons’
current recordkeeping practices. Relevant records contain KDEs associated with different CTEs
in a food supply system. These required records and data elements, which vary across the types
of entities in the food supply system, are described in detail below.
To estimate the recordkeeping costs of the proposed rule, the time to perform the various
recordkeeping functions, the frequency of recordkeeping by record type, and the average time
spent keeping records by record type, we use several sources, including industry standards such
as the Global Data Synchronization Network, Institute of Food Technologists publications, along
with FDA’s Evaluation of Recordkeeping Costs for Food Manufacturers. As in previous
85

sections, all wage rates come from the Bureau of Labor Statistics, Occupational Employment
Statistics (OES) from May 2018.
Some entities perform multiple CTEs and would be subject to more than one
recordkeeping provision. Each provision outlines the KDEs necessary to effectively trace a
product based on the CTE an entity performs (e.g., receiving, transformation, shipping). Not all
KDEs are relevant for each CTE; however, entities that perform multiple CTEs would be
required to maintain all KDEs that pertain to the CTEs they perform. For example, an entity that
receives a food on the FTL and then transforms and ships it would need to keep records of KDEs
relevant to receiving, transforming, and shipping.
We estimate recordkeeping burdens by CTE because each entity must comply with all
requirements in its relevant CTE. Some KDEs (e.g., traceability lot code, traceability product
identifier) are required by multiple CTEs; however, no two CTEs contain exactly the same KDE
requirements. We lack information about which entities may be able to use the same KDE for
multiple CTEs. Because we cannot account for these potential time savings, our total
recordkeeping burdens by CTE may be overestimated.
a. Exempt Entities (§ 1.1305)
As discussed above, our analysis excludes those farms and other small originators that
would be fully exempt from the proposed rule because of their very small size. Specifically,
farms that are not covered under the produce safety rule in accordance with 21 CFR 112.4(a) are
exempt from this proposed rule with respect to the produce they grow. Shell egg producers with
fewer than 3,000 laying hens are exempt with respect to the shell eggs they produce. Other
originators of food are exempt if the average annual monetary value of the food sold during the
previous 3-year period was no more than $25,000 (on a rolling basis), adjusted for inflation using

2019 as the baseline year for calculating the adjustment. Furthermore, when all eggs produced at
a particular farm receive a treatment 60 in accordance with 21 CFR 118.1(a)(2), the rule does not
apply to those shell eggs.
There are other exemptions in the proposed rule that we do not specifically identify in our
cost analysis because we lack the data. As a result, our costs may be overstated. We discuss these
additional exemptions below and request comment on potentially affected entities.
For instance, farms are exempt with respect to FTL foods produced on the farm that are
sold directly to a consumer by the owner, operator, or agent in charge of the farm. Also, the
proposed rule does not apply to FTL foods that are produced and packaged on a farm if the
packaging of the food remains in place until the food reaches the consumer, and the packaging
maintains the integrity of the product and prevents subsequent contamination or alteration of the
product and maintains labeling that includes the name, address, and phone number of the farm.
Produce FTL foods that receive commercial processing to adequately reduce the presence of
microorganisms of public health significance are exempt from the recordkeeping requirements if
the conditions set forth in 21 CFR 112.2(b) are met (regarding the commercial processing
exemption to the produce safety rule). Produce that is listed as rarely consumed raw in 21 CFR
112.2(a)(1) is also exempt. The proposed rule also has partial exemptions related to certain
commingled raw agricultural commodities on the FTL and FTL food produced through the use
of fishing vessels.
Retail food establishments are partially exempt with respect to FTL foods that are
produced on a farm and sold directly to the retail food establishment by the owner, operator, or

Treatment (or treated) means a technology or process that achieves at least a 5-log destruction of SE for shell
eggs, or the processing of egg products in accordance with the Egg Products Inspection Act

agent in charge of that farm. The only records retail food establishments must maintain in such
cases are the name and address of the source farm.
Other entities not subject to this rule’s recordkeeping requirements include transporters of
FTL foods; non-profit food establishments; persons who manufacture, process, pack, or hold
FTL foods for personal consumption; and persons who hold FTL foods on behalf of specific
individual consumers, provided that these persons are not parties to the transaction involving the
FTL food they hold or are not in the business of distributing food. These entities are not included
in our counts of covered establishments.
Under proposed § 1.1305(g), farm-to-school and farm-to-institution programs would be
partially exempt from the proposed recordkeeping requirements. When such programs purchase
a food on the FTL from a farm, the applicable requirement would be for the school food
authority or relevant food procurement entity to establish and maintain a record documenting the
name and address of the farm that was the source of the food. We believe that this is the same
location description data element that is typically stored in distribution and shipping
recordkeeping systems. This record must be maintained for 180 days, which is the same retention
period proposed for retail food establishments purchasing foods on the FTL directly from farms.
This partial exemption applies to farm-to-school and farm-to-institution entities that
operate under various Federal, state, and local jurisdictions facilitating these food distribution
programs. These programs include, but are not limited to, programs in which farms sell food (1)
directly to schools under competitive procurement, (2) to competitively procured food
distributors, (3) to vendors in the USDA Fresh Fruit and Vegetable Program operated by the
Department of Defense’s Defense Logistics Agency, and (4) to the USDA’s Agricultural
Marketing Service for use in the National School Lunch Program.

Intermediaries facilitate most of these sales between originators and the school or
institution (Ref. [36]). We assume these intermediaries, in the form of school food authorities or
relevant food procurement entities that obtain foods on the FTL directly from farms, already
keep the name and address of the farm as part of their conventional recordkeeping systems and
existing regulatory mandates, and so no new costs would be incurred by these firms. We include
food service contractors and other private procurement food distributors among our counts of
entities covered by this proposed rule.
There are some schools and institutions that purchase foods on the FTL directly from
farms and not through any of these intermediaries or procurers. The 2015 USDA Farm to School
Census indicates that approximately 22 percent of participating school food authorities
purchased foods directly from farms or farmer cooperatives (Ref. [37]). School food authorities
purchase for one or many schools. Of the 42,711 reported schools that had farm-to-school
activities in 2013-2014, this 22 percent would represent about 9,396 schools. We assume these
schools, in the form of their school food authorities, already keep the name and address of the
farm as part of their conventional recordkeeping systems and existing regulatory mandates, and
so no new costs would be incurred by these schools. We invite comment on these assumptions.
We do not have additional information that would allow us to estimate the precise
number of participating non-school institutions (e.g., hospitals, colleges, prisons) not already
counted in our coverage discussion of food service contractors, wholesalers, and distributors. We
ask for comments regarding estimates of the number of such institutions.
b. Traceability Program Records (§ 1.1315)
The proposed rule, if finalized, would require entities that manufacture, process, pack, or
hold foods on the FTL to maintain general records of traceability and make them available to

FDA upon request. Entities affected by this provision would therefore incur one-time and
recurring annual costs at the establishment level associated with establishing and maintaining
general traceability records. These records would include a description of the reference records
in which an establishment maintains required information, an explanation of where the
information appears in the records, and, if applicable, a description of how reference records for
different critical tracking events are linked. Additional records would include a list of foods on
the FTL the establishment ships, including the traceability product identifier and traceability
product description for each food, and a description of how the establishment establishes and
assigns traceability lot codes to affected foods.
Some establishments may already follow practices that meet some requirements in this
provision. To account for this uncertainty, we assume that 50 percent of establishments would
incur costs related to changing their current recordkeeping practices to fulfill the requirements of
this provision, with a lower bound of 25 percent and an upper bound of 75 percent. As in the
previous sections, we apply a different assumption to restaurants, assuming 5-15 percent would
incur costs. We estimate the costs of this provision on a per establishment basis and conclude
that 130,063 establishments would incur costs related to this provision, with range of 30,107 to
306,464 establishments. We request comment on these estimates.
We assume that establishments covered by this provision would face one-time costs to
develop procedures for recordkeeping and to set up traceability program records, as well as
recurring annual costs to keep records up to date as inputs are added, deleted, or changed.
Establishing traceability program records would involve setting up master data in a standardized
format, such as a report, spreadsheet template, or software. Although some of the costs
associated with establishing traceability program records may not be affected by the volume of

business done by the establishment, we assume that the costs associated with establishing these
records will increase with the number of different FTL products handled by the establishment
and with the number of lots of FTL foods that the establishment handles annually.
Because we lack information on the number of unique FTL products handled by
establishments, and because there is no clear relationship between number of products and
number of lots, we model only number of lots in accounting for variation in recordkeeping
burden. Namely, we scale the time to establish general records based on number of lots.
Furthermore, as we do not know the particular numbers of lots typical to each industry sector, we
use a range of numbers of lots in our model to reflect variability in both sector and establishment
size. Based on discussion with FDA subject matter experts, we assume that the average
establishment (by type and size) has approximately 1,000 lots of FTL foods that they originate,
create, or transform annually, with a lower bound of 500 lots and an upper bound of 2,000 lots
(Ref. [34]). We apply this assumption to subsequent provisions. We request comments on these
lot-based assumptions and on other appropriate measures for calculating the burden of general
records, such as the number of different products handled by an average firm.
To calculate the labor cost of establishing and maintaining general records, we use the
hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one employee
at each covered establishment. Based on discussion with FDA subject matter experts, we assume
that the one-time burden to establish general records by setting up master data ranges from one to
three minutes per lot. We assume the time burden to maintain these records with periodic updates
is lower, from ten to 20 seconds per lot (Ref. [34]). We request comments on these assumptions.
We multiply the annual number of affected FTL lots, the associated time burden of
establishing and maintaining a record in accordance with the provision, and the employee labor

cost to yield the one-time and recurring costs per establishment. We estimate that the one-time
cost per affected establishment is $1,250 (= 1,000 x 0.03 x $37.50) and the recurring cost is $156
(= 1,000 x 0.004 x $37.50). The total one-time cost of this provision for all affected
establishments is $162.6 million (= 130,063 x $1,250) and the annual recurring cost is roughly
$20.3 million (=130,063 x $156). Table 19 presents a summary of the estimated recordkeeping
costs of this provision with lower and upper bounds.

Table 19. One-time and Recurring Recordkeeping Costs of the General Records, 1.1315
Provision (2018$)
Primary
Low
High
130,063
30,107
306,464
Number of affected establishments
Number of FTL lots per
establishment
1,000
500
2,000
Time to of establish records (hours)
0.03
0.02
0.05
Time to of maintain records (hours)
0.004
0.003
0.006
Labor cost of hourly employee
Per establishment one-time cost
One-time cost of general records
Per establishment recurring cost
Recurring cost of general records

$37.50
$1,250
$162,578,510
$156
$20,322,314

$37.50
$313
$9,408,569
$52
$1,568,095

$37.50
$3,750
$1,149,240,155
$417
$127,693,351

c. Records of Growing a Food on the Food Traceability List (§ 1.1325)
The proposed rule, if finalized, would require entities who grow foods on the FTL to
establish and maintain records containing and linking the traceability lot code of the food to
growing area coordinates. Growers of sprouts would be required to maintain additional records;
their requirements are discussed below. Affected growers would incur recurring costs associated
with establishing and maintaining these records, to be made available to FDA upon request.
In addition to the traceability lot code and growing area coordinates for each lot that all
growers must establish and maintain, sprout growers would need to request documentation on
the incoming seeds from their suppliers and record this information into their work order
92

systems. For each received seed lot, this includes the location identifier and location description
of the grower of seeds for sprouting, the associated seed lot code assigned by the seed grower,
and the date of seed harvesting; the location identifier and location description of the seed
conditioner or processor, the associated seed lot code assigned by the seed conditioner or
processor, and the date of conditioning or processing; the location identifier and location
description of the seed packinghouse (including a re-packer), the associated seed lot code
assigned by the seed packinghouse, and the date of packing (and repacking, if applicable); the
location identifier and location description of the seed supplier; a description of the seeds,
including the seed type or taxonomic name, growing specifications, volume, type of packaging,
and antimicrobial treatment; the seed lot code assigned by the seed supplier, including the master
lot and sub-lot codes, and any new seed lot code assigned by the sprouter; the date of receipt of
the seeds by the sprouter; and for each seed lot code, the sprout traceability lot code(s) and the
date(s) of production associated with that seed lot code.
Some sprout growers may already be following some or all of these recordkeeping
requirements as discussed in the 2017 FDA draft guidance for the sprout operations industry
(Ref. [38]) or as recommended by good agricultural practices. We estimate that sprout growers
not already performing certain recordkeeping activities would incur new recurring recordkeeping
costs for the records outlined above. In this analysis, we use the inventory of sprout farms and
operations used by the FDA’s Office of Regulatory Affairs. Excluding very small sprout
growers, this internal inventory counts 95 sprout growers. To account for uncertainty, we halve
and double this number to approximate a range of 48 to 190 sprout growers covered by this
provision and add these ranges to the total number of affected growers. We request comment on
this estimate.

As in previous sections, we estimate the total number of growers affected by identifying
NAICS categories likely to grow foods on the Food Traceability List (Table 20) and removing
exempt and non-covered growers. We lack information on how many growers already keep such
records as a part of their current business practices. To account for this uncertainty, we assume
that 50 percent of these growers would need to change their practices, with a range of 25 to 75
percent. We therefore estimate that in total 9,459 growers, including sprout growers, would incur
costs related to this provision, with a lower bound of 482 and an upper bound of 26,964. Because
our primary, low, and high estimates of sprout growers represent 0.54%, 0.14%, and 1.62% of all
growers, we estimate that 51 sprout growers, with a lower bound of 13 and an upper bound of
153, would have to change their current practices to comply with this provision. We request
comment on these estimates.
Table 20. NAICS Categories of Entities that Likely to Grow Foods on the Food
Traceability List
2012
Category
NAICS
111219 Other Vegetable (except Potato) and Melon Farming
111339 Other Non-citrus Fruit Farming
111419 Other Food Crops Grown Under Cover

Growers covered by this provision would incur annually recurring recordkeeping costs
related to establishing and maintaining traceability records for specific lots of food, to be made
available to FDA upon request. We estimate that each covered grower would establish and
maintain records required by this provision for each lot of FTL food that it grows. Applying our
earlier lot assumptions, we assume that each grower affected by this provision grows 1,000 lots
of FTL foods annually, with a lower bound of 500 lots and an upper bound of 2,000 lots. We
request comments on these assumptions.

To calculate the labor cost of establishing and maintaining these records, we use the
hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one employee
at each covered establishment. For non-sprout growers we estimate that the time burden of
establishing and maintaining a record for this provision ranges from one to three minutes. For
sprouting growers, we double this burden. We request comments on these estimates.
We multiply the number of affected FTL lots per grower, the associated time burden of
maintaining a record in accordance with the provision, and the employee labor cost to yield the
per establishment cost of maintaining records of growing FTL foods. For non-sprout growers we
estimate that the recurring cost of establishing and maintaining records for this provision is
$1,250 (= 1,000 x 0.03 x $37.50), with a lower bound of $313 (= 500 x 0.02 x $37.50) and an
upper bound of $3,750 (= 2,000 x 0.05 x $37.50). For sprouting growers, we estimate that the
recurring cost of establishing and maintaining records for this provision is $2,500 (= 1,000 x
0.07 x $37.50), with a lower bound of $625 (= 500 x 0.03 x $37.50) and an upper bound of
$7,500 (= 2,000 x 0.10 x $37.50).
For both non-sprouting and sprouting growers, we multiply the per establishment cost of
establishing and maintaining records for this provision by the number of establishments affected
by this provision to estimate the total cost of establishing and maintaining records of growing
FTL foods. We estimate that this annually recurring cost is approximately $11.9 million (=
(9,408 x $1,250) + (51 x $2,500)), with a lower bound of $0.2 million (= (469 x $312.50) + (13 x
$625)) and an upper bound of approximately $101.7 million (= (26,811 x $3,750) + (153 x
$7,500)). Table 21 presents a summary of the estimated recordkeeping costs of this provision.
Table 21. Recurring Recordkeeping Costs of the Records of Growing FTL Food, 1.1325
Provision (2018$)
Primary
Low
High
Non-Sprout Firms
95

9,408
1,000

469
500

26,811
2,000

0.03

0.02

0.05

$37.50
$1,250
Sprout Firms
51
Number of affected firms
Annual number of FTL lots per firm
1,000
Time to establish and maintain records for
sprouting firms (hours)
0.07

$37.50
$313

$37.50
$3,750

13
500

153
2,000

0.03

0.10

$37.50
$625
$154,624

$37.50
$7,500
$101,688,031

Number of affected firms
Annual number of FTL lots per firm
Time to establish and maintain records
(hours)
Labor cost of hourly employee
Per firm cost

Labor cost of hourly employee
Per firm cost
Total recurring cost

$37.50
$2,500
$11,887,851

d. Records to Be Kept by First Receivers of Foods on the FTL (§ 1.1330)
The proposed rule, if finalized, would require entities that are the first receivers of FTL
foods to establish and maintain records on traceability. We estimate that entities affected by this
provision would incur annual recurring costs at the establishment level of establishing and
maintaining such records. These records would link the traceability lot code of the food received
to the location identifier and location description of the originator, cooler, and packer, as well as
contact information for the harvester; records would also include the date and time of harvesting,
cooling, and packing. For FTL foods that were obtained from a fishing vessel, first receivers
must establish and maintain records containing and linking the traceability lot code of the
received seafood product to the name of the vessel, its unique identifier or license number (both
if available), its licensing country (if any), a point of contact for the vessel, and the harvest date
range and locations for the trip during which the seafood was caught. Finally, first receivers that
receive an FTL food without a traceability lot code would need to establish and maintain a
traceability lot code.
96

As in previous sections, we estimate the total number of first receivers affected by
identifying NAICS categories likely to grow foods on the Food Traceability List (Table 22) and
removing exempt and non-covered establishments. 61 We lack information on how many first
receivers already keep such records as a part of their current business practices. To account for
this uncertainty, we assume that 50 percent of these establishments would need to change their
practices, with a range of 25 to 75 percent. We therefore approximate that about 12,700
establishments would incur costs related to this provision, with a lower bound of 3,491
establishments and an upper bound of 25,791 establishments. We request comment on these
estimates.
Table 22. NAICS Categories Likely to Be First Receivers of Foods on the Food Traceability
List
2012
Category
NAICS
311340 Nonchocolate Confectionery Manufacturing
311351 Chocolate and Confectionery Manufacturing from Cacao Beans
311352 Confectionery Manufacturing from Purchased Chocolate
311411 Frozen Fruit, Juice and Vegetable Manufacturing
311412 Frozen Specialty Food Manufacturing
311421 Fruit and Vegetable Canning
311422 Specialty Canning
311423 Dried and Dehydrated Food Manufacturing
311520 Ice Cream and Frozen Dessert Manufacturing
311710 Seafood Product Preparation and Packaging
311811 Retail Bakeries
311812 Commercial Bakeries
311813 Frozen Cakes, Pies, and Other Pastries Manufacturing
311824 Dry Pasta, Dough, and Flour Mixes Manufacturing from Purchased Flour
311941 Mayonnaise, Dressing, and Other Prepared Sauce Manufacturing
311942 Spice and Extract Manufacturing
311991 Perishable Prepared Food Manufacturing
424410 General Line Grocery Merchant Wholesalers
424420 Packaged Frozen Food Merchant Wholesalers
424460 Fish and Seafood Merchant Wholesalers
6161. This provision may affect a small amount of retail food establishments, but we cannot estimate this number
due to data limitations. We ask for comment on the number of these entities.

424480
424490
445220
445230
493110
493120
493130

Fresh Fruit and Vegetable Merchant Wholesalers
Other Grocery and Related Products Merchant Wholesalers
Fish and Seafood Markets
Fruit and Vegetable Markets
General Warehousing and Storage
Refrigerated Warehousing and Storage
Farm Product Warehousing and Storage

First receivers covered by this provision would incur recurring annual costs of
establishing and maintaining these records, to be made available to FDA upon request. We
estimate that each affected establishment would do that for each affected lot of FTL food. We
estimate that each first receiver receives approximately 1,000 lots of FTL foods annually, with a
lower bound of 500 lots and an upper bound of 2,000 lots. We request comments on these
estimates.
To calculate the labor cost of establishing and maintaining these records, we use the
hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one employee
at each covered establishment. We estimate that the time burden of establishing and maintaining
a record for this provision is approximately two minutes per record, with a lower bound of
approximately one minute and an upper bound of approximately three minutes. We request
comments on these estimates.
We multiply the number of affected FTL lots per first receiver, the associated per record
time burden, and the employee labor cost to yield the per establishment cost of establishing and
maintaining records required by this provision. We estimate that this recurring per first receiver
cost is $1,250 (= 1,000 x 0.03 x $37.50), with a lower bound of $313 (= 500 x 0.02 x $37.50)
and an upper bound of $3,750 (= 2,000 x 0.05 x $37.50). We multiply the per first receiver cost
by the number of affected establishments to estimate the total cost of establishing maintaining
first receiver records. We estimate that this annually recurring cost is $15.9 million (= 12,700 x
98

$1,250), with a lower bound of approximately $1.1 million (= 3,491 x $313) and an upper bound
of approximately $96.7 million (= 25,791 x $3,750). Table 23 presents a summary of the
estimated recordkeeping costs of this provision.
Table 23. Recurring Recordkeeping Costs of First Receiver Records, Provision 1.1330
(2018$)
Primary
Low
High
12,700
3,491
25,791
Number of affected establishments
Number of FTL lots per establishment
1,000
500
2,000
Time to establish and maintain records
(hours)
0.03
0.02
0.05
$37.50
$37.50
Labor cost of hourly employee
$37.50
Per establishment cost
$1,250
$313
$3,750
$15,875,215
$1,091,053
$96,715,230
Total recurring cost
e. Records for Receipt of Foods on the Food Traceability List (§ 1.1335)
The proposed rule, if finalized, would require receivers of foods on the FTL to establish
and maintain records containing and linking the traceability lot code of the food to the location
identifier and location description of the immediate previous source of an FTL food; its entry
number (if imported); location identifier and location description for where food was received,
and the receiving date and time of receipt; quantity and unit of measure of the food received;
traceability product identifier and description; location identifier and description and POC for the
traceability lot code generator; reference record type and record number for the receipt of the
food; and the name of the transporter.
As in previous sections, we estimate the total number of receivers affected by identifying
NAICS categories likely to grow foods on the Food Traceability List (Table 24) and removing
exempt and non-covered establishments. We lack information on how many receivers already
keep such records as a part of their current business practices. To account for this uncertainty, we
assume that 50 percent of these establishments would need to change their practices, with a range

of 25 to 75 percent. We estimate therefore that approximately 265,610 establishments would
incur costs related to this provision, with a lower bound of approximately 75,481 and an upper
bound of approximately 587,043. We request comment on these estimates.
Table 24. NAICS Categories Likely to Receive Foods on the Food Traceability List
2012
Category
NAICS
311340 Nonchocolate Confectionery Manufacturing
311351 Chocolate and Confectionery Manufacturing from Cacao Beans
311352 Confectionery Manufacturing from Purchased Chocolate
311411 Frozen Fruit, Juice and Vegetable Manufacturing
311412 Frozen Specialty Food Manufacturing
311421 Fruit and Vegetable Canning
311422 Specialty Canning
311423 Dried and Dehydrated Food Manufacturing
311513 Cheese Manufacturing
311520 Ice Cream and Frozen Dessert Manufacturing
311710 Seafood Product Preparation and Packaging
311811 Retail Bakeries
311812 Commercial Bakeries
311813 Frozen Cakes, Pies, and Other Pastries Manufacturing
311821 Cookie and Cracker Manufacturing
311824 Dry Pasta, Dough, and Flour Mixes Manufacturing from Purchased Flour
311911 Roasted Nuts and Peanut Butter Manufacturing
311941 Mayonnaise, Dressing, and Other Prepared Sauce Manufacturing
311942 Spice and Extract Manufacturing
311991 Perishable Prepared Food Manufacturing
424410 General Line Grocery Merchant Wholesalers
424420 Packaged Frozen Food Merchant Wholesalers
424430 Dairy Product (except Dried or Canned) Merchant Wholesalers
424450 Confectionery Merchant Wholesalers
424460 Fish and Seafood Merchant Wholesalers
424480 Fresh Fruit and Vegetable Merchant Wholesalers
424490 Other Grocery and Related Products Merchant Wholesalers
445110 Supermarkets and Other Grocery (except Convenience) Stores
445120 Convenience Stores
445220 Fish and Seafood Markets
445292 Confectionery and Nut Stores
445230 Fruit and Vegetable Markets
445291 Baked Goods Stores
445299 All Other Specialty Food Stores
447110 Gasoline Stations with Convenience Stores
100

452910
454111
454210
493110
493120
493130
722310
722320
722330
722410
722511
722513
722514
722515

Warehouse Clubs and Supercenters
Non-store Retailers: Electronic Shopping
Vending Machine Operators
General Warehousing and Storage
Refrigerated Warehousing and Storage
Farm Product Warehousing and Storage
Food Service Contractors
Caterers
Mobile Food Services
Drinking Places (Alcoholic Beverages)
Full-Service Restaurants
Limited-Service Restaurants
Cafeterias, Grill Buffets, and Buffets
Snack and Nonalcoholic Beverage Bars

Receivers of FTL foods covered by this provision would incur recurring annual
recordkeeping costs of establishing and maintaining these records, to be made available to FDA
upon request. Each covered receiver would establish and maintain records for each lot of
received FTL foods. We assume that all shipments sent by shippers must be received by
receivers, and we estimate that each receiver receives from roughly 2,000-8,000 shipments per
year on average. We request comment on this assumption, which we arrive at by dividing the
number of shipments sent by shippers by the total number of receivers. We note that entities
responsible for shipping, such as distribution centers, wholesalers, and warehouses, may
maintain receiving records for clients. We do not make an assumption about how these costs are
shared.
To calculate the labor cost of establishing and maintaining these records, we use the
hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one employee
at each covered establishment. We estimate that the time burden of establishing and maintaining
receiver records ranges from 15 to 20 seconds per shipment. We request comments on these
estimates.
101

We multiply the number of affected FTL food lots per receiver, the associated time
burden, and the employee labor cost to yield the per receiver cost of this provision. We estimate
that the recurring per receiver cost of this provision is $370 (= 2,367 x 0.004 x $37.50), with a
lower bound of $137 (= 1,312 x 0.003 x $37.50) and an upper bound of $869 (= 4,169 x 0.006 x
$37.50). We multiply the per receiver cost by the number of establishments affected by this
provision to estimate the total cost of establishing and maintaining receiver records. We estimate
that this annually recurring cost is $98.2 million (= 265,610 x $370), with a lower bound of
$10.3 million (= 75,481 x $137) and an upper bound of $510 million (= 587,043 x $869). Table
25 presents a summary of the estimated recordkeeping costs of this provision.

Table 25. Recurring Recordkeeping Costs of Receiver Records, Provision 1.1335 (2018$)
Primary
Low
High
265,610
75,481
587,043
Number of affected establishments
Number of FTL shipments per
2,367
1,312
4,169
establishment
Time to establish and maintain
records (hours)
0.004
0.003
0.006
$37.50
$37.50
Labor cost of hourly employee
$37.50
$370
$137
$869
Per establishment cost
$98,230,051
$10,316,392
$509,907,986
Total recurring cost
f. Records of Transformation of Foods on the Food Traceability List (§ 1.1340)
The proposed rule, if finalized, would require entities who transform foods on the FTL to
establish and maintain traceability records containing and linking the new traceability lot code of
the food produced through transformation to the information described below. We estimate that
entities affected by this provision would annually incur recurring costs at the establishment level
of establishing and maintaining records, to be made available to FDA upon request. For the
food(s) on the FTL used in transformation, these records would include the information on the
traceability lot codes, the traceability product identifier and product description, and the quantity
102

of each traceability lot. For the food produced through transformation, these records would
include information on the location identifier and description of where food was transformed
(and the date transformation was completed), the new traceability product identifier and product
description for the food to which the new traceability lot code applies, the quantity and unit of
measure of food produced through transformation for each new traceability lot code, and the
reference record type and record number for the documents containing the information on the
foods used in, and produced through, transformation.
As in previous sections, we estimate the total number of transformers affected by
identifying NAICS categories likely to transform foods on the Food Traceability List (Table 26)
and removing exempt and non-covered establishments. We lack information on how many
transformers already keep such records as a part of their current business practices. To account
for this uncertainty, we assume that 50 percent of these establishments would need to change
their practices, with a range of 25 to 75 percent. We therefore estimate that 5,244 establishments
would incur costs related to this provision, with a lower bound of 1,219 and an upper bound of
11,022 establishments. We request comment on these estimates.
Table 26. NAICS Categories Likely to Transform Foods on the FTL
2012
Category
NAICS
311340 Nonchocolate Confectionery Manufacturing
311351 Chocolate and Confectionery Manufacturing from Cacao Beans
311352 Confectionery Manufacturing from Purchased Chocolate
311411 Frozen Fruit, Juice and Vegetable Manufacturing
311412 Frozen Specialty Food Manufacturing
311421 Fruit and Vegetable Canning
311422 Specialty Canning
311423 Dried and Dehydrated Food Manufacturing
311513 Cheese Manufacturing
311520 Ice Cream and Frozen Dessert Manufacturing
311710 Seafood Product Preparation and Packaging
311811 Retail Bakeries
103

311812
311813
311821
311824
311941
311942
311991

Commercial Bakeries
Frozen Cakes, Pies, and Other Pastries Manufacturing
Cookie and Cracker Manufacturing
Dry Pasta, Dough, and Flour Mixes Manufacturing from Purchased Flour
Mayonnaise, Dressing, and Other Prepared Sauce Manufacturing
Spice and Extract Manufacturing
Perishable Prepared Food Manufacturing

Each person affected by this provision would incur recurring annual recordkeeping costs
for establishing and maintaining these records, to be made available to FDA upon request. We
estimate that each affected establishment would do that for each new traceability lot of food
produced through transformation of FTL food(s). Using our previous lot assumptions, we
estimate that each such establishment transforms approximately 1,000 lots of FTL foods
annually, with a lower bound of 500 lots and an upper bound of 2,000 lots. We request
comments on these estimates.
To calculate the labor cost of establishing and maintaining these records, we use the
hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one employee
at each covered establishment. We estimate that the time burden of establishing and maintaining
a record for this provision is approximately two minutes per record, with a lower bound of
approximately one minute and an upper bound of approximately three minutes. We request
comments on these estimates.
We multiply the number of affected FTL lots per covered establishment, the associated
time burden, and the employee labor cost to yield the per establishment cost of establishing and
maintaining records required by this provision. We estimate that this recurring annual per
establishment cost is $1,250 (= 1,000 x 0.03 x $37.50), with a lower bound of $313 (= 500 x 0.02
x $37.50) and an upper bound of $3,750 (= 2,000 x 0.05 x $37.50). We multiply the per
establishment cost by the number of affected covered establishments to estimate the total cost of
104

establishing and maintaining the required food transformation records. We estimate that this
annually recurring cost is $6.6 million (= 5,244 x $1,250), with a lower bound of $0.4 million (=
1,219 x $313) and an upper bound of $41.3 million (= 11,022 x $3,750). Table 27 presents a
summary of the estimated recordkeeping costs of this provision.
Table 27. Recurring Recordkeeping Costs of Food Transformation Records, Provision
1.1340 (2018$)
Primary
Low
High
Number of affected establishments
5,244
1,219
11,022
Number of FTL lots per establishment
1,000
500
2,000
Time to establish and maintain records
(hours)
0.03
0.02
0.05
$37.50
$37.50
Labor cost of hourly employee
$37.50
Per establishment cost
$1,250
$313
$3,750
Total recurring cost
$6,554,420
$381,066
$41,330,739
g. Records of Creation of Foods on the Food Traceability List (§ 1.1345)
The proposed rule, if finalized, would require entities who create foods on the FTL to
establish and maintain records containing and linking the traceability lot code of the food created
to the location identifier and location description for where a food was created, the creation date,
the traceability product identifier and product description, the quantity and unit of measure of the
food created, and the reference record type and record number for the documents containing the
previously-stated information. We estimate that entities affected by this provision would incur
recurring annual costs at the establishment level of establishing and maintaining these records, to
be made available to FDA upon request.
As in previous sections, we estimate the total number of creators affected by identifying
NAICS categories likely to create foods on the Food Traceability List (Table 28) and removing
exempt and non-covered establishments. We lack information on how many creators already
keep such records as a part of their current business practices. To account for this uncertainty, we

105

assume that 50 percent of these establishments would need to change their practices, with a range
of 25 to 75 percent. We estimate therefore that approximately 222 establishments would incur
costs related to this provision, with a lower bound of approximately 45 establishments and an
upper bound of approximately 552 establishments. We request comment on these estimates.
Table 28. NAICS Categories Likely to Create Foods on the Food Traceability List
2012
Category
NAICS
311513 Cheese Manufacturing
311911 Roasted Nuts and Peanut Butter Manufacturing
311991 Perishable Prepared Food Manufacturing

Each affected establishment would need to establish and maintain records for each lot of
created FTL food. We approximate that each affected establishment creates 1,000 lots of FTL
foods annually, with a lower bound of 500 lots and an upper bound of 2,000 lots. We request
comments on these estimates.
To calculate the labor cost of establishing and maintaining these records, we use the
hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one employee
at each covered establishment. We estimate that the time burden of establishing and maintaining
a record for this provision is approximately one to three minutes per lot. We request comments
on these estimates.
We multiply the number of created FTL lots per affected establishment, the associated
time burden, and the employee labor cost to yield the per establishment cost of maintaining
records of creating FTL foods. We estimate that this recurring annual cost is $1,250 (= 1,000 x
0.03 x $37.50), with a lower bound of $313 (= 500 x 0.02 x $37.50) and an upper bound of
$3,750 (= 2,000 x 0.05 x $37.50). We multiply the per establishment cost by the number of
establishments affected by this provision to estimate the total cost of establishing and
106

maintaining records of creating FTL foods. We estimate that this annually recurring cost is
$278,000 (= 222 x $1,250), with a lower bound of $14,000 (= 45 x $313) and an upper bound of
$2.1 million (= 552 x $3,750). Table 29 presents a summary of the estimated recordkeeping costs
of this provision.
Table 29. Recurring Recordkeeping Costs of Creating FTL Food, Provision 1.1345
(2018$)
Primary
Low
High
Number of affected establishments
222
45
552
Number of FTL lots per establishment
1,000
500
2,000
Time to establish and maintain records (hours)
0.03
0.02
0.05
$37.50
$37.50
Labor cost of hourly employee
$37.50
Per establishment cost
$1,250
$313
$3,750
Total recurring cost
$278,018
$14,050
$2,069,114
h. Records to Be Kept and Sent for Shipment of Foods on the Food Traceability List (§
1.1350)
The proposed rule, if finalized, would require entities who ship foods on the FTL to
establish, maintain, and send (to the immediate subsequent recipient (other than a transporter) of
the food) certain records related to the covered food, including information on the location
identifier and location description for the immediate subsequent recipient of the food; its entry
number (if imported); the location identifier and location description of the place from which the
food was shipped, and date and time the food was shipped; the traceability lot code; the quantity
and unit of measure of the food shipped; the traceability product identifier and product
description for the food; and the location identifier, location description, and point of contact for
the traceability lot code generator. Entities who ship FTL foods would also be required to
establish and maintain (but not send forward) records of the reference record type and record
number for the shipment of each traceability lot of the food; and the name of the transporter.
When sending records, farms must also include a statement that they are a farm and additional
107

information on the location identifier and location description of the originator, cooler, and
packer of the transported FTL food, as well as the name, point of contact, and phone number of
the harvester, and the dates and times of harvesting, cooling, and packing. Entities affected by
this provision would incur recurring annual costs associated at the establishment level of
establishing, maintaining, and sending these records.
As in previous sections, we estimate the total number of shippers affected by identifying
NAICS categories likely to ship foods on the Food Traceability List (Table 30) and removing
exempt and non-covered establishments. We lack information on how many shippers already
keep and send such records as a part of their current business practices. To account for this
uncertainty, we assume that 50 percent of these establishments would need to change their
practices, with a range of 25 to 75 percent. We therefore estimate that approximately 29,593
establishments would incur costs related to this provision, with a lower bound of approximately
5,908 establishments and an upper bound of approximately 68,641 establishments. We request
comment on these estimates.
Table 30. NAICS Categories Likely to Ship Foods on the Food Traceability List
2012
Category
NAICS
111219 Other Vegetable (except Potato) and Melon Farming
111339 Other Non-citrus Fruit Farming
111419 Other Food Crops Grown Under Cover
112310 Chicken Egg Production
112511 Finfish Farming and Fish Hatcheries
112512 Shellfish Farming
114111 Finfish Fishing
114112 Shellfish Fishing
311340 Nonchocolate Confectionery Manufacturing
311351 Chocolate and Confectionery Manufacturing from Cacao Beans
311352 Confectionery Manufacturing from Purchased Chocolate
311411 Frozen Fruit, Juice and Vegetable Manufacturing
311412 Frozen Specialty Food Manufacturing
311421 Fruit and Vegetable Canning
108

311422
311423
311513
311520
311710
311811
311812
311813
311821
311824
311911
311991
424410
424420
424430
424460
424480
424490
493110
493120
493130

Specialty Canning
Dried and Dehydrated Food Manufacturing
Cheese Manufacturing
Ice Cream and Frozen Dessert Manufacturing
Seafood Product Preparation and Packaging
Retail Bakeries
Commercial Bakeries
Frozen Cakes, Pies, and Other Pastries Manufacturing
Cookie and Cracker Manufacturing
Dry Pasta, Dough, and Flour Mixes Manufacturing from Purchased Flour
Roasted Nuts and Peanut Butter Manufacturing
Perishable Prepared Food Manufacturing
General Line Grocery Merchant Wholesalers
Packaged Frozen Food Merchant Wholesalers
Dairy Product (except Dried or Canned) Merchant Wholesalers
Fish and Seafood Merchant Wholesalers
Fresh Fruit and Vegetable Merchant Wholesalers
Other Grocery and Related Products Merchant Wholesalers
General Warehousing and Storage
Refrigerated Warehousing and Storage
Farm Product Warehousing and Storage

Each covered shipper would establish, maintain, and send records for each shipment lot
of the FTL food. Compared to other shippers affected by this provision, we assume distribution
centers, wholesalers, and warehouses collect information at a faster pace (from ten to 20 seconds
per shipment to record and ten to 20 seconds per shipment to send). Based on discussions with
FDA subject matter experts, we assume these establishments have a higher volume of
transactions because they service a larger number of clients (Ref. [34]). We assume these
establishments service between ten and 100 clients, with a primary estimate of 35. We multiply
our previous lot assumptions (500 to 2,000, with a primary estimate of 1,000) by this distribution
to calculate the total time spent to establish, maintain, and send shipping records. Separately, we
assume that the remaining establishments in this category, such as growers and manufacturers,
collect and record information at a slower speed (from one to three minutes per shipment to

109

record and one to three minutes per shipment to transmit). We request comments on these
estimates.
To calculate the labor cost of establishing, maintaining, and sending these records, we use
the hourly wage rate of $37.50 (= $18.75 x 2 to account for benefits and overhead) for one
employee at each covered establishment. We multiply the shipping time burden and the
employee labor cost to yield the per establishment cost of establishing, maintaining, and sending
shipping-related records. For distribution centers, wholesalers, and warehouses, the primary per
establishment cost is $15,104 per year (= 48,333 shipments x 0.008 hours x $37.50). For all other
entities, this per establishment cost is $2,500 (= 1,000 shipments x 0.06 hours x $37.50).
We multiply the per establishment cost by the number of affected establishments to
estimate the total recordkeeping cost related to shipping FTL foods. We estimate that this
annually recurring cost is $191.2 million (= 12,657 x $15,000) for distribution centers,
wholesalers, and warehouses and $42.3 million for all other entities (=16,936 x $2,500). Table
31 presents a summary of the estimated recordkeeping costs of this provision, including lower
and upper bounds.
Table 31. Recurring Recordkeeping Costs of Shipping FTL Foods, Provision 1.1350
(2018$)
Primary
Low
High
Distribution Centers, Wholesalers, and Warehouses
Number of affected establishments
12,657
4,060
24,404
Number of affected shipments per
establishment
48,333
24,167
96,667
Time to establish and maintain records
(hours)
0.004
0.003
0.006
Time to send records (hours)
0.004
0.003
0.006
Recordkeeping employee hourly labor cost
$37.50
Per establishment cost of recordkeeping
$15,104
$191,167,560
Total shipping cost (annual)
All Other Entities
16,936
Number of affected establishments

$37.50
$5,035
$20,440,229

$37.50
$40,278
$982,952,378

1,849

44,236
110

Number of affected shipments per
establishment
Time to establish and maintain records
(hours)
Time to send records (hours)
Recordkeeping employee hourly labor cost
Per establishment cost of recordkeeping
Total shipping cost (annual)
Total shipping cost (annual) for all
establishments

1,000

500

2,000

0.03
0.03

0.02
0.02

0.05
0.05

$37.50
$2,500
$42,340,338
$233,507,899

$37.50
$625
$1,155,325
$21,595,554

$37.50
$7,500
$331,772,336
$1,314,724,71
4

6. Non-Quantified Costs
The information flows brought about by the proposed rule may prompt new protective
actions — for example, in farming, manufacturing or cooking processes—that themselves would
have costs. These costs have not been quantified because we lack information on both the nature
and likelihood of these behavior changes; we therefore request comment on these potential costs.
Indeed, there is a likely correlation between these costs’ occurrence and the realization of health
and longevity benefits attributable to this rule. One of the challenges of such attribution, for both
health and longevity benefits and this category of costs, is the lag in data availability as other
FSMA regulations continue to take effect. 62
Other provisions of the proposed rule, if finalized, may generate costs that we cannot
estimate quantitatively. There may be costs incurred by FDA to review petitions requesting
modified requirements or exemptions (§1.1380), adopt modified requirements or grant
exemptions on our own initiative (§1.1385), decide that modified requirements or exemptions
should be revised or revoked (§1.1400), receive and respond to waiver petitions (§1.1435), waive

As noted in section II.D, above, the outcomes of earlier FSMA regulations should be taken into account in the
characterization of this proposal’s regulatory baseline.

111

requirements on our own initiative (§1.1440), and determine that a waiver should be modified or
revoked (§1.1450).
For provisions concerning petitions, costs to FDA could include time spent reviewing and
responding, as well as publishing notices of decisions in the Federal Register. For provisions that
allow FDA to modify requirements and grant exemptions and waivers on our own initiative,
costs could include time spent making these determinations and publishing notices of decisions
in the Federal Register. Because we cannot estimate the number of petitions FDA would receive
if this proposed rule is finalized, we cannot estimate the costs of these provisions.
Unquantified costs incurred by industry may include the time to create and file petitions
for modifications, exemptions, and waivers. Preparing and submitting these documents would
present a one-time cost per firm per food product. However, we believe that firms would pursue
this course of action only if it is in their best financial interests and thus that these costs would
not increase the total costs of this rule.
If finalized, other provisions of the proposed rule may generate one-time costs for
industry to meet FDA requests. One such requirement would be for entities (at the establishment
level) to generate certain records in the form of a sortable electronic spreadsheet upon request by
FDA. FDA would make such a request when necessary to help FDA prevent or mitigate a
foodborne illness outbreak, or to assist in the implementation of a recall, or to otherwise address
a threat to the public health. The number of times establishments would generate this electronic
spreadsheet depends on the number of times FDA would request said record. Some
establishments may already keep records in a sortable electronic spreadsheet format while other
may not. Because electronic sortable spreadsheets are commonly used for other general business
purposes, such as for taxes, we estimate that any additional costs from this requirement would be

112

negligible. Moreover, because these spreadsheets would be needed only in the event of an
investigation, we assume such costs would be incurred infrequently and unpredictably. We seek
comment on these assumptions.
Other one-time costs would result from time spent completing and submitting petitions
for modified requirements or exemptions (§ 1.1370), petitions for waiver for a type of entity
(§ 1.1425), or waiver requests for an individual entity (§ 1.1415). Because we cannot estimate
the number of persons or entities that will submit petitions and request waivers if the proposed
rule is finalized, we cannot estimate the costs associated with these actions. However, we believe
that because petitions and waiver requests are voluntary, firms would only incur such costs if
they expect that this choice is in their best financial interest.
Finally, we note that a request from FDA to produce a sortable spreadsheet (under the
circumstances described above) will be withdrawn when necessary to accommodate a religious
belief of a person asked to provide such a spreadsheet. Because this does not require any further
action from persons or entities requesting a waiver other than stating a religious reason, we
believe any such additional costs to be negligible.
7. Summary of Costs
Table 32 summarizes our estimates of the one-time and recurring costs of the proposed
rule. We estimate that the total one-time costs of the proposed rule, if finalized, would be
approximately $1,113 million, with a lower bound of $58 million and an upper bound of $7.6
billion. We estimate that the total recurring costs of the proposed rule, if finalized, would be
approximately $387 million per year, with a lower bound of $35 million per year and an upper
bound of $2.2 billion per year.
Table 32. Total Costs of the Proposed Rule (millions 2018$)
One-time Costs
Primary
Low

High
113

Reading and Understanding the Rule
Capital Investment
Training
§ 1.1315 General Recordkeeping
Total One-time Costs
Annually Recurring Costs
§ 1.1315 General Recordkeeping
§ 1.1325 Growing Recordkeeping
§ 1.1330 First Receiver Recordkeeping
§ 1.1335 Receiver Recordkeeping
§ 1.1340 Transformation Recordkeeping
§ 1.1345 Creation Recordkeeping
§ 1.1350 Shipping Recordkeeping
Total Recurring Costs

$52
$725
$174
$163
$1,113

$25
$11
$13
$9
$58

$88
$5,740
$660
$1,149
$7,637

$20
$12
$16
$98
$7
$0.3
$234
$387

$2
$0.2
$1
$10
$0.4
$0.01
$22
$35

$128
$102
$97
$510
$41
$2
$1,315
$2,194

We estimate that in the first year after the proposed rule becomes effective, total costs
would be $1,500 million dollars, with a lower bound of $93 million and an upper bound of
approximately $9.8 billion. In subsequent years, the annual cost of the proposed rule would
decrease to $387 million, with a lower bound of $35 million and an upper bound of
approximately $2.2 billion. We estimate that the total costs of the proposed rule over ten years
would be approximately $5.0 billion, ranging from a lower bound of $409 million and an upper
bound of approximately $29.6 billion.
A summary of the estimated ten-year stream of costs of the proposed rule is presented in
Table 33. The present value of total estimated costs is approximately $3.8 billion at a seven
percent discount rate and $4.4 billion with a three percent discount rate. The ten-year annualized
value of costs is $535 million with a seven percent discount rate and $513 million with a three
percent discount rate.
Table 33. Ten-Year Timing of the Costs of the Proposed Rule (millions 2018$)
Year
Primary
Low
High
1
$1,500
$93
$9,831
2
$387
$35
$2,194
3
$387
$35
$2,194
4
$387
$35
$2,194
114

5
6
7
8
9
10
Total Costs of the Proposed Rule
Present Value of Total Costs (3%)
Present Value of Total Costs (7%)
Annualized Value of Costs (3%)
Annualized Value of Costs (7%)

$387
$387
$387
$387
$387
$387
$4,980
$4,379
$3,756
$513
$535

$35
$35
$35
$35
$35
$35
$409
$356
$301
$42
$43

$2,194
$2,194
$2,194
$2,194
$2,194
$2,194
$29,578
$26,131
$22,548
$3,063
$3,210

G. Distributional Effects
If this proposed rule is finalized, we do not expect any significant distributional effects,
although the rule would potentially affect a significant proportion of entities (firms or
establishments) that manufacture, process, pack, or hold foods that appear on the FTL. The rule
does not require any additional tasks to specific types of entities compared to other affected
entities other than establishing, maintaining, and sending records that would enable traceability
of the covered foods. We therefore do not expect any significant distributional effects as a result
of enforcing this rule should it be finalized.
H. International Effects

This section estimates costs for foreign entities (firms or establishments) who
manufacture, process, pack, or hold foods listed on the FTL and who offer such foods for export
to the U.S. market. While it is possible that there might be a small number of foreign entities that
sell food in the United States and meet the definition of “retail food establishment,” we assume
that the number of such entities affected by this rule is negligible. We request comment on this
assumption.

115

We estimate that total one-time costs to foreign entities range from approximately $10.3
million to $3,494 million, with a primary estimate of $461 million. One-time costs occur in year
one. We estimate that recurring costs to foreign entities (in each year) range from approximately
$17.1 million to $1,149 million dollars, with a primary estimate of $198 million dollars. At a
seven percent discount rate, our primary estimate of the present value of costs to foreign entities
over ten years is approximately $1,821 million, and our primary estimate of the annualized value
of these costs is approximately $259 million, ranging from $18.4 million to $1.6 billion. These
cost estimates are costs to foreign entities only. To the extent that these costs are passed on to
U.S. entities, U.S. consumers and U.S. firms that purchase products from foreign entities may
experience higher costs. However, requirements of this rule affect all domestic entities in the
same manner regardless of whether their suppliers are domestic or foreign and their estimated
costs are estimated in section F of this analysis. We face uncertainty concerning the portion of
foreign producers’ compliance costs that may be passed on to U.S. consumers. It is possible that
U.S. producers may also pass on some of their compliance costs to consumers of U.S. exports in
other countries. We expect that U.S. and foreign producers alike will attempt to pass their costs
onto consumers in markets where demand is least elastic. Bekkers et al. (2013) estimate that the
pass-through rate from world food prices to final consumption prices is about 12 percent in high
income countries (Ref. [39]), but this research does not speak to the extent to which countryspecific costs (such as those imposed as a result of a regulation) are passed through to consumers
or supply chain entities in that country.
To estimate these costs of the proposed rule on foreign entities, we extrapolate from the
main cost estimates by comparing the number of foreign facilities in FDA’s Food Facility
Registration Module (FFRM) to the primary estimated number of domestic establishments minus

116

retail food establishments. As stated above, we assume that the number of foreign retail food
establishments affected by this rule is negligible. Hence, we exclude domestic retail food
establishments and their costs from the extrapolation to foreign entities. With 212,404 foreign
food facilities, the ratio of foreign establishments to domestic establishments is approximately
1.95. Since the numerator of this ratio is all foreign facilities in the FFRM, the denominator of
this ratio is likewise all 109,000 domestic non-RFE establishments in the NAICS codes relevant
to the rule, and not just the ones (65,607) that we have estimated to manufacture, process, pack,
or hold FTL foods in section II.D.iii “Coverage of the Rule.” We then use this ratio as a scaling
factor to extrapolate from counts of entities in the main analysis, namely the counts of affected
entities. This implicitly assumes that the same proportion of registered foreign firms are affected
by the rule (i.e., manufacture, process, pack, or hold FTL foods) as the proportion of domestic
firms.
Thus, we estimate the count of foreign establishments affected by the rule to be about
1.95 times the main count of establishments affected by the rule (or approximately 127,925 =
65,607 x 1.95), and likewise for the count of affected firms (approximately 108,686 = 55,740 x
1.95). We do not have a detailed breakdown of these foreign establishments and firms by
industry, and instead assume the same proportional breakdown as in the main analysis. In other
words, for each cost element described in sections II.F.ii – II.F.v, we simply exclude RFEs and
then scale the remaining number of affected entities by 1.95. Hence, we estimate foreign costs
by: 1) multiplying the domestic firm and establishment counts from the main analysis by 1.95
(after excluding RFEs) to obtain foreign counts, and 2) replacing domestic wage estimates with
foreign wage estimates in labor cost calculations.

117

We generally expect foreign wages to be lower than domestic wages. To estimate wages
for foreign employees and supervisors, we take weighted average wages across the top twenty
foreign countries by value of food exports to the United States (weighted by value of food
exports) (Ref. [40]). 63 This produces a foreign employee wage of $6.91 per hour and a foreign
supervisor wage of $15.63 per hour, which we double to account for benefits and overhead.
Additionally, we adjust estimates of the time to read and understand the rule accounting
for potentially lower English language proficiency and internet accessibility among some foreign
facilities. 64 For example, in learning about the requirements of this proposed rule, we assume
that entities from English speaking countries will spend a comparable amount of time as
domestic entities. Entities from countries with low English proficiency but with high internet
access may spend more time learning about the rule than domestic entities because they may
need to have internet access to translate the rule. Entities from countries with both low English
proficiency and low internet access may spend even more time learning about the rule (and incur
higher costs) than entities from countries with high English proficiency and/or internet access.
We estimate that foreign establishments on average would spend 1.61 hours for every hour that
domestic establishments spend reading the rule. This estimate is the average of an estimated
English proficiency weighted average of 1.67 hours and an estimated internet usage weighted
average of 1.56 hours. To account for language proficiency differences, we use information from
a report titled “Education First English Proficiency Index” (EF EPI) published in November of
2018. 65 This report ranks countries by the average level of English language skills amongst

63 Except for Vietnam and Ecuador, for which the OWW dataset lacked data.
64 FDA does not require that records be maintained in English. Instead, we require that an English translation be
provided upon request.
65 EF English Proficiency Index – Comparing English skills between countries – EF EPI. Ef.com. Retrieved on
2018 -11-16. http://www.ef.edu/epi/

118

adults using data collected via English tests available over the internet. To account for country
differences in internet accessibility, we use 2018 internet user percentage estimates by country. 66
We request comment on these estimates.
I. Uncertainty and Sensitivity Analysis
There are several sources of uncertainty associated with both the costs and benefits of the
proposed rule. Most of these sources of uncertainty arise from inadequate data or information
related to implementation of the proposed rule, if finalized.
We have very limited information regarding the number of entities (firms or
establishments) that would be affected by the proposed rule. As there is no specific single
registry for entities that currently manufacture, process, pack, or hold foods on the FTL, it is
difficult to establish the total number of such affected entities. Moreover, should the rule be
finalized, the Food Traceability List could potentially change through addition and/or deletion of
foods. We are therefore unable to predict the foods that will in the future be on the FTL, and the
number of entities that might become covered or cease to be covered. Instead, we more broadly
consider Census Bureau industry classifications that we expect to encompass FTL products and
related entities. Our estimates rely on assumptions about the share of each industry that would be
affected by the proposed rule, if finalized. These share assumptions are based mostly on receipts
data from the 2012 Economic Census.
Additionally, our cost estimates reflect uncertainty about the degree to which entities
already satisfy the requirements of the proposed rule. While in recent years the food industry has
adopted new technologies, including those related to product traceability, we are unable to

119

exactly quantify the extent of this change. For example, because of the volume and diversity of
existing State and local recordkeeping requirements, our cost estimates do not account for State
or local regulations that require similar recordkeeping practices for small sectors of the food
economy. Overall, we expect the exclusion of these other regulatory factors to have a very small
effect on our cost computations.
However, apart from technological change that corresponds to regulatory requirements
(e.g., in response to FSMA or to State or local regulations), entities have been gradually adopting
technologies for business reasons, including but not limited to enhanced inventory management
and supply system efficiencies. To account for this, our main cost estimates use three different
assumptions regarding the percentage of estimated traceability costs that entities would incur due
to the rule: twenty-five percent, fifty percent, and seventy-five percent. Note that we only
consider an average percentage over all covered entities. Particularly given the number of
industries likely to be affected, we lack data to estimate the percentage of remaining traceability
investment needed by each industry. As there may be significant variation between industries in
size, supply chain complexity, and current practice, there may accordingly be significant
uncertainty that our estimates do not capture.
We also face significant uncertainty in relation to the expected benefits of the proposed
rule. While we expect the rule to improve public health by reducing the size and impact of
outbreaks, we are unable to determine the magnitude of this improvement. Our uncertainty
regarding the benefits of the rule primarily stems from the complexity of predicting the health
benefits of shorter foodborne disease outbreaks.
Concerning our baseline scenario, we are also uncertain about the current number of
foodborne illnesses caused by foods covered by the FTL rule. Our baseline estimates are

120

primarily based on 2009-2019 FDA outbreak data. During this period, numerous initiatives have
taken effect, including publication of several FSMA-related rules intended to help prevent
foodborne illness outbreaks, including outbreaks from foods on the FTL. We lack data on the
extent to which these rules might have changed our baseline.
To address some of the uncertainties outlined above, we show estimates using wide
ranges for the share of covered firms within each industry. We also allow the share of
traceability investment needed among covered firms to range from twenty-five percent to
seventy-five percent. Finally, our monetized benefits reflect a wide uncertainty range because
they were calculated using information from a small (possibly under-representative) number of
outbreaks.

J. Analysis of Regulatory Alternatives to the Proposed Rule
We considered five different regulatory alternatives in addition to the proposed and coproposed options. We compare these regulatory alternatives to proposed Option 2 only. We
estimate costs of regulatory alternatives keeping in mind that FDA is also proposing that the
compliance date for any final rule would be two years after the date the final rule is effective. We
estimate cost of all regulatory alternatives as if they have the same effective and compliance
dates as the proposed rule.
The five alternatives we consider are Alternative a): No new regulatory action as the
baseline for determining the costs and benefits of other alternatives including the proposed rule;
Alternative b): Expand the FTL rule to cover all foods instead of the proposed rule; Alternative
c): Expand coverage for very small farms; Alternative d): Include foods transformed or created at
the retail level; and Alternative e): Extend compliance date to three years. Table 34 shows a
121

detailed summary of the costs associated with each regulatory alternative and the change in the
estimated costs relative to those associated with the proposed rule.

Alternative a: No Action
We treat the alternative of taking no new regulatory action as the baseline for determining
the costs and benefits of other alternatives. In choosing an appropriate baseline, OMB Circular
A-4 recommends considering a wide range of factors, including market evolution, changes in
external factors affecting expected benefits and costs, changes in regulations promulgated by the
agency and the degree of compliance by regulated entities with other regulations. In choosing a
baseline, we assume costs and benefits of the BTA food tracing requirements are already
accounted for (although benefits have been either overestimated or not fully realized). As such,
if FDA pursued Alternative a, there would be no additional costs or benefits under this
alternative.

Alternative b: Expand the proposed rule to include all foods at covered firms

This option evaluated a scenario where traceability would be a requirement for all foods
and not just for FTL foods. Although this option is not legally viable under current law, ideally
such a requirement for the industry may be a logical step to further improve public health. It has
been argued that the risk of food contamination is based more on handling practices rather than
on the type of commodities (Ref. [41]). Some food traceability experts have argued that a
traceability platform that focuses on a select list of foods runs the risk of becoming outdated,
especially when new outbreak vehicles emerge. Thus far, firms that have chosen to invest in

122

traceability programs have generally done so for their entire inventory, maximizing their
investment, rather than focusing on a specific subset of products. While many firms will not
necessarily require additional traceability investments, we recognize that building a food
traceability platform with wide scope and depth will likely be more feasible for larger businesses
than for small businesses who constitute nearly 90 percent of firms (Ref. [42]).
Costs of including foods that are not on the FTL would affect mostly small businesses.
We estimate the annualized cost of Alternative b would be $656 million, which is about $121
million more than the estimated cost of the proposed rule (Table 34).
Any additional benefits from including other foods beyond the FTL would depend on the
extent to which other foods not on the FTL are involved in outbreaks. We are unable to quantify
any additional benefits from this option and request comments on the benefits of including other
foods beyond those on the FTL.
Alternative c: Expand coverage for very small farms and other originators.

We evaluate a scenario where implementation of the proposed rule will include over
22,000 very small farms and other originators which in the current proposed rule would be
exempt. Including these additional firms will increase the annualized cost of the rule to $549
million (using a seven percent discount rate) which is roughly about $14 million more than the
estimated cost of the rule. While the extension of the rule to include certain small farms and
originators increases costs to these small entities, including them could also increase the benefits
of the rule by improving public health. The expected additional benefits of such coverage would
depend on the proportion of FTL foods from these very small farms and originators entering the
market and may or may not offset the additional social costs of such a broad rule. If we assume a

123

small market share of FTL foods are from very small farms and originators, any additional
estimated benefits of this option would also be small. We request comments on this assumption.
Alternative d: Include foods transformed or created at the retail level

Under this alternative, the proposed traceability recordkeeping requirements would apply
to the transformation or creation of food performed by retail food establishments such as grocery
stores and convenience stores (but not restaurants, who transform food for individual diners). We
considered the additional recordkeeping requirements for non-restaurant retail food
establishments if they create a food on the FTL that they sell directly to a consumer or if they
transform a food on the FTL to food they sell directly to a consumer. For example, if the retail
food establishment prepared fresh salsa using tomatoes and fresh herbs, it would be required to
keep records for the lot numbers and other traceability product identifiers for the tomatoes and
herbs, as well as records for the fresh salsa, as outlined by the requirements of § 1.1340 and §
1.1345.
FDA has tentatively concluded that to require non-restaurant retail food establishments to
keep records of these transformation and creation events would be too burdensome and not
necessary in order to address credible foodborne illness outbreak threats. We also do not
anticipate this alternative having an impact on reducing the impact of overly broad recalls or
market withdrawals and therefore assume estimated benefits of this option to be same as the
estimated benefit of the proposed rule (Table 34). We request comments on this assumption.
Alternative e: Extend compliance date to three years
This alternative extends the compliance date of the rule to three years following the
effective date of the final regulation. Under this Option 1 we assume that the rule would publish
124

in year 0, costs would begin in year 2 instead of year 1, and benefits would begin in year 3
instead of year 2. Compared to Option 2, delaying the compliance date would reduce the burden
on covered entities by shifting costs into the future. Health benefits would also be lower under
this alternative because they would begin one year later than under Option 2.
Table 34. Summary of Costs by Regulatory Alternatives Compared to Option 2 (the
Proposed Rule) ($ millions)
Costs
Regulatory Option - Domestic
Facilities
Proposed Rule Option 2
Regulatory Alternative a: No Action
Change from Proposed Rule
Regulatory Alternative b: Expand
coverage beyond FTL foods
Change from Proposed Rule
Regulatory Alternative c: Expand
coverage for very small farms and
other originators
Change from Proposed Rule
Regulatory Alternative d: Include
foods transformed or created at the
retail level
Change from Proposed Rule
Regulatory Alternative e: Extend
compliance date to three years
Change from Proposed Rule

One-Time Costs

Recurring Costs
($/year)

Annualized
Total (7%)

$1,113

$387

$535

($1,113)

($387)

($535)

$1,514

$455

$656

$401

$68

$121

$1,167

$394

$549

$54

$14

$1,113

$452

$600

$65

$1,113

$387

$474

($61)

III. Initial Small Entity Analysis
The Regulatory Flexibility Act requires Agencies to analyze regulatory options that
would minimize any significant impact of a rule on small entities. Because some small firms may
have annualized costs (over ten years at a seven percent discount rate) that exceed one percent of
125

their annual revenue, we find that the proposed rule will have a significant economic impact on a
substantial number of small entities. This analysis, as well as other sections in this document,
serves as the Initial Regulatory Flexibility Analysis, as required under the Regulatory Flexibility
Act.
A. Description and Number of Affected Small Entities
The entities in this small entity analysis are firms. The Small Business Administration
(SBA) publishes size standards for industry categories of firms defined by North American
Industry Classification System (NAICS) codes. SBA defines each NAICS code’s small business
threshhold either in terms of sales revenue or number of employees of a firm. Using the 2019
SBA size standards 67 in conjunction with the SUSB counts of firms in each NAICS code by
revenue and employment size, 68 we estimate the numbers of covered small firms by industry.
Overall, about 89.7 percent of firms coverd by this rule count as small businesses by SBA
standards. Table 35 shows estimated counts of covered small firms by NAICS code.

67 Small Business Association. Table of Size Standards. Aug 19, 2019. Available from:
https://www.sba.gov/document/support--table-size-standards
68 We use the 2012 SUSB, the last release that contained revenue data, and inflate revenues to 2018 dollar values
using the GDP deflator.
Census Bureau. 2012 SUSB Annual Datasets by Establishment Industry. Updated 2015. Available from:
https://census.gov/data/datasets/2012/econ/susb/2012-susb.html

126

Table 35. Entities affected by the proposed rule
2012
NAICS
Code

111219

111339

111419

NAICS
Industry
Description
Other
Vegetable
(except Potato)
and Melon
Farming 69
Other Noncitrus Fruit
Farming69
Other Food
Crops Grown
Under Cover69

112310

Chicken Egg
Production 70

112511

Finfish
Farming and
Fish
Hatcheries 71

112512

Shellfish
Farming71

114111

Finfish Fishing

114112

Shellfish
Fishing

Firm type

Number
of
Covered
Small
Firms

6,604

Annual
Revenue
($millions)

SBA Size
Revenue
Standard
per Firm
($millions
($million
or no. of
s)
employees)

$927

$0.14

Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs
Farms/Aquac
ulture/Growe
rs

271

$38

$0.14
$1

220

$31

$0.14
$1

1,375

$4,024

$2.93
$16.5

292

$82

$0.28
$1

311

$77

$0.25
$1

586

$610

$1.04
$22.00

447

$256

$0.57
$6.00

69 We base the small entity count and revenue estimate for this industry on Tables 47 and 48 of the regulatory
impact analysis for the prior Produce Rule (Ref. [43]): https://www.fda.gov/media/94153/download
70 The SBA defines chicken and egg producers to be small if their total revenues are less than $16.5 million. A
producer that receives $0.85 per dozen eggs (the midpoint of seasonally adjusted December 2018 and December
2019 market egg prices) and has layers that produce 265 eggs per year would have to have over 879,000 layers in
production to earn revenues of over $16.5 million. Because only about 320 farms fall into the category of 100,000 or
more layers, more than 99 percent of the farms with more than 3,000 layers are considered small by SBA standards
and account for roughly 76 percent of all production. Multiplying this production amount (76 percent) by the value
of all egg production in 2018 ($10,586,262,000) (Ref. [44]) gives $8,045,559,120 in total revenue for small farms. If
99 percent of egg farms (2,777 x 0.99 = 2,749) are considered small by SBA standards, then we estimate the average
revenue per SBA small egg farm to be $2,926,722.
71 We use Table 9 from the 2018 USDA Census of Aquaculture to compute the weighted average revenue of small
farms (less than $1 million in sales) by fish type (baitfish, food fish, crustaceans, and mollusks). We then combine
categories by weighted average of types of small farms (Finfish includes food fish and Shellfish includes
Crustaceans and Mollusks). Finally, we multiply the average revenue by the total number of small farms to obtain
the total revenue for all small farms.

127

311340

311351

311352

311411

311412

311421

Nonchocolate
Confectionery
Manufacturing
Chocolate and
Confectionery
Manufacturing
from Cacao
Beans
Confectionery
Manufacturing
from
Purchased
Chocolate
Frozen Fruit,
Juice and
Vegetable
Manufacturing
Frozen
Specialty Food
Manufacturing
Fruit and
Vegetable
Canning

311422

Specialty
Canning

311423

Dried and
Dehydrated
Food
Manufacturing

311513

Cheese
Manufacturing

311520

311710

Ice Cream and
Frozen Dessert
Manufacturing
Seafood
Product
Preparation
and Packaging

311811

Retail Bakeries

311812

Commercial
Bakeries

311813

Frozen Cakes,
Pies, and Other
Pastries
Manufacturing

Manufacturin
g/Processing/
Packing

245

$5,813

$23.73
1,000

$3,090

$42.91

Manufacturin
g/Processing/
Packing

1,250

670

$5,716

$8.53

Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing

1,000

$7,943

$86.33
1,000

151

$7,367

$48.79
1,250

476

$21,900

$46.01
1,000

$6,823

$106.61
1,250

138

$5,935

$43.01
750

156

$18,143

$116.30
1,250

230

$5,278

$22.95
1,000

248

$5,649

$22.78
750

4,214

$2,222

$0.53
500

1,923

$26,323

$13.69
1,000

134

$3,983

$29.73
750

128

311821

311824

311911

311941

311942

311991

424410

424420

424430

424450

424460

424480

Cookie and
Cracker
Manufacturing
Dry Pasta,
Dough, and
Flour Mixes
Manufacturing
from
Purchased
Flour
Roasted Nuts
and Peanut
Butter
Manufacturing
Mayonnaise,
Dressing, and
Other Prepared
Sauce
Manufacturing
Spice and
Extract
Manufacturing
Perishable
Prepared Food
Manufacturing
General Line
Grocery
Merchant
Wholesalers
Packaged
Frozen Food
Merchant
Wholesalers
Dairy Product
(except Dried
or Canned)
Merchant
Wholesalers
Confectionery
Merchant
Wholesalers
Fish and
Seafood
Merchant
Wholesalers
Fresh Fruit and
Vegetable
Merchant
Wholesalers

Manufacturin
g/Processing/
Packing

157

$6,567

$41.83
1250

211

$7,221

$34.22

Manufacturin
g/Processing/
Packing

750

106

$6,080

$57.36
750

164

$5,982

$36.48

Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing
Manufacturin
g/Processing/
Packing

750

141

$2,033

$14.42
500

300

$2,116

$7.05
500

1,561

$26,983

$17.29

Wholesalers/
Distributors

250

1,539

$29,023

$18.86

Wholesalers/
Distributors

200

824

$13,135

$15.94

Wholesalers/
Distributors

200

1,142

$8,547

$7.48
200

1,285

$8,758

$6.82

Wholesalers/
Distributors

100

3,384
Wholesalers/
Distributors

$34,480

$10.19
100

129

424490

445110

Other Grocery
and Related
Products
Merchant
Wholesalers
Supermarkets
and Other
Grocery
(except
Convenience)
Stores

445120

Convenience
Stores

445220

Fish and
Seafood
Markets

445292

Confectionery
and Nut Stores

445230

Fruit and
Vegetable
Markets

445291

Baked Goods
Stores

445299

447110

452910

454111

454210

493110

493120

All Other
Specialty Food
Stores
Gasoline
Stations with
Convenience
Stores
Warehouse
Clubs and
Supercenters
Non-store
Retailers:
Electronic
Shopping
Vending
Machine
Operators
General
Warehousing
and Storage
Refrigerated
Warehousing
and Storage

7,945

$62,160

$7.82

Wholesalers/
Distributors

250

21,691

$58,995

$2.72

Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Warehouses
and Storage
Warehouses
and Storage

$35.00

12,769

$11,135

$0.87
$32.00

1,003

$1,187

$1.18
$8.00

1,253

$742

$0.59
$8.00

1,411

$1,805

$1.28
$8.00

2,015

$957

$0.48
$8.00

2,587

$1,843

$0.71
$8.00

21,087

$74,323

$3.52
$32.00

$0.36
$32.00

7,954

$16,124

$2.03
$41.50

2,033

$1,894

$0.93
$12.00

2,523

$16,733

$6.63
$30.00

360

$1,322

$3.67
$30.00

130

493130

Farm Product
Warehousing
and Storage

722310

Food Service
Contractors

722320

Caterers

722330

Mobile Food
Services

722410

Drinking
Places
(Alcoholic
Beverages)

722511

Full-Service
Restaurants

722513

722514

722515

LimitedService
Restaurants
Cafeterias,
Grill Buffets,
and Buffets
Snack and
Nonalcoholic
Beverage Bars

Warehouses
and Storage
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts
Retail Food
Establishmen
ts

239

$808

$3.38
$30.00

1,482

$2,063

$1.39
$41.50

7,604

$6,083

$0.80
$8.00

2,369

$652

$0.28
$8.00

18,104

$9,211

$0.51
$8.00

116,204

$99,411

$0.86
$8.00

83,421

$77,268

$0.93
$12.00

3,731

$3,010

$0.81
$30.00

31,711

$15,109

$0.48
$8.00

B. Description of the Potential Impacts of the Rule on Small Entities
In the main cost section of this economic impact analysis, we present low, medium, and
high cost estimates for each provision of the rule. We estimate that costs to small firms would
fall primarily between our low and primary estimates. We do not attempt to characterize the
variability within this range, and hence present the low and primary estimates without suggesting
any particular distribution of the costs between them. We do not suggest, for example, that the
average of the low and primary estimates is particularly representative of small firms. In this
small entity analysis, we call these estimates our lower and upper bounds.

131

Table 36 breaks down our estimate of one-time cost per covered small entitity across
broad categories of industries. Though aggregate totals are displayed for these broad categories,
the underlying analysis in this section accounts for applicable provisions at the level of each
NAICS code. We assume that all covered small entities would incur one-time costs to read and
understand the rule. Some covered small entities would also incur a one-time capital investment
cost and a one-time cost of training, as well as recurring annual recordkeeping costs. We expect
one-time per firm compliance costs to range from about $300 to $5,500 for small farms and
small manufacturers, $300 to $5,700 for small wholesalers, $300 to $6,100 for small warehouses,
and $100 to $2,200 for small retailers.
Table 36. One-time per firm compliance costs of the proposed rule
Lower Bound
Upper Bound
Farms /Aquaculture / Growers
$287
$5,543
Manufacturers / Processors / Packers
$290
$5,597
Wholesalers / Distributors
$297
$5,727
Warehouse and Storage
$321
$6,139
Retail Food Establishments
$113
$2,178

Using the same breakdown, Table 37 shows estimated cost per covered small entities,
anualized over ten years at a seven percent discount rate. We expect annualized costs to range
from about $400 to $5,300 for small farms, $500 to $6,300 for small manufacturers, $500 to
$6,000 for small wholesalers, $500 to $5,600 for small warehouses, and $0 to $600 for small
retailers.
Table 37. Annualized per firm compliance costs of the proposed rule (over ten years, seven
percent discount rate)
Lower Bound
Upper Bound
Farms /Aquaculture / Growers
$416
$5,286
Manufacturers / Processors / Packers
$520
$6,339
Wholesalers / Distributors
$522
$6,057
Warehouse and Storage
$472
$5,618
Retail Food Establishments
$45
$582
132

We use the SUSB 72 to estimate the magnitude of costs as a percent of the revenues of
covered small firms. We consider costs per firm exceeding one percent of annual revenues to be
a substantial impact. Table 38 shows our estimate of the one-time compliance costs as a
percentage of revenue for small firms, broken down by broad industry categories. We expect
one-time costs as a percentage of annual revenue to range from about 0.1% to 0.9% for small
farms overall, 0% to 0.2% for small chicken egg farms, 0.1% to 2.4% for other small farms, 0%
to 0% for small manufacturers, 0% to 0.1% for small wholesalers, 0% to 0.1% for small
warehouses, and 0% to 0.2% for small retailers.
Table 38. One-time per firm compliance costs as a percentage of small firm annual revenue
Costs as a percent
Costs as a percent
of revenue
of revenue
(Lower Bound)
(Upper Bound)
Farms /Aquaculture / Growers
0.05%
0.93%
Chicken egg production
0.01%
0.19%
Other farms
0.12%
2.40%
Manufacturers / Processors / Packers
0.00%
0.04%
Wholesalers / Distributors
0.00%
0.06%
Warehouse and Storage
0.01%
0.10%
Retail Food Establishments
0.01%
0.19%
Using the same categorical breakdown, Table 39 shows the annualized values of our
estimates of compliance costs over ten years at a seven percent discount rate, again as a
percentage of the revenues of covered small firms. Over ten years, at a seven percent discount
rate, we expect annualized costs as a percentage of annual revenue to range from about 0.07% to
0.9% for small farms overall, 0% to 0.2% for small chicken egg farms, 0.2% to 2.3% for other

For small farms, we estimate revenues based on the Produce Rule economic impacts analysis, the USDA National
Agricultural Statistics Service, and the USDA Census of Agriculture. We describe these estimates in footnotes 69,
70, and 71.

133

small farms, 0% to 0% for small manufacturers, 0% to 0.1% for small wholesalers, 0% to 0.1%
for small warehouses, and 0% to 0.1% for small retailers.
Table 39. Annualized per firm compliance costs as a percentage of annual revenue (ten
years, seven percent discount rate)
Costs as a percent
Costs as a percent
of revenue
of revenue
(Lower Bound)
(Upper Bound)
Farms /Aquaculture / Growers
0.07%
0.88%
Chicken egg production
0.01%
0.18%
Other farms
0.18%
2.28%
Manufacturers / Processors / Packers
0.00%
0.04%
Wholesalers / Distributors
0.01%
0.06%
Warehouse and Storage
0.01%
0.09%
Retail Food Establishments
0.00%
0.05%

In Table 40, we estimate that the total costs of the proposed rule per covered small firm
over ten years ranges from approximately $900 to $10,700. At a seven percent discount rate, the
present value of the total costs of the proposed rule per covered small firm ranges from
approximately $700 to $8,100. Discounted at three percent, the present value of the total costs of
the proposed rule per covered small firm ranges from approximately $800 to $9,400. Discounted
at either seven or three percent over ten years, the estimated annualized value of costs of the
proposed rule per small firm ranges from approximately $100 to $1,100. We request comment on
our small entity analysis. In particular, we recognize that there may be variability in the cost
burden for very small firms and request comment and data on how to best characterize very
small firms by sector.
Table 40. Costs of the proposed rule per small firm (over ten years)
Lower Bound
Upper Bound
Total Costs of the Proposed Rule
$883
$10,700
Present Value of Total Costs (7%)
$651
$8,108
Present Value of Total Costs (3%)
$769
$9,428
Annualized Value of Costs (7%)
$93
$1,154
Annualized Value of Costs (3%)
$90
$1,105

134

C. Alternatives to Minimize the Burden on Small Entities
As the vast majority (roughly ninety-three percent) of covered firms qualify as small
entities, the analysis of regulatory alternatives for covered firms described in Section J above
effectively describes the effects of the alternatives on small entities. Delaying the compliance
date for small entities would delay the implementation of the proposed rule for the vast majority
of FTL products. While the postponement of capital investments and labor expenses for
compliance would reduce the present value of costs of the proposed rule, it would also reduce the
present value of the health benefits.

IV. Co-proposed Option 1
A. Summary
For purposes of estimating the economic impact of co-proposed Option 1 (full exemption
for small retail food establishments (RFEs)), we use the same assumptions used to estimate the
economic impact of Option 2 (exemption for small RFEs from the requirement to provide FDA,
under certain circumstances, with an electronic sortable spreadsheet of requested tracing
information). The rule under Option 1 would fully exempt RFEs with ten or fewer full-time
equivalent employees from the proposed rule; all provisions are otherwise the same. Option 2
would only exempt such RFEs from the requirement that, under certain circumstances, entities
provide an electronic sortable spreadsheet to FDA upon request. Our economic analysis of
impacts of Option 2 is described in sections II and III. In this section IV, we summarize
estimates for co-proposed Option 1 and compare them to estimates for Option 2. Table 41 shows

135

the overall reduction in covered firms and establishments and reduction in annualized costs for
both Options.
Table 41. Comparison of Impacts Between Option 1 and Option 2
Measure
Covered Retail Food Firms
Covered Retail Food
Establishments
Costs (Annualized at 7%, 10
years), Millions 2018$
International Costs that May
Be Passed Through to US
Supply Chain and Consumers
(Annualized at 7%, 10 years),
Millions 2018$
Public Health Benefits
(Annualized at 7%, 10 years),
Millions 2018$
Benefits from Avoiding Overly
Broad Recalls, (Annualized at
7%, 10 years), Millions 2018$
Cost per Illness, 2018$

366,404

132,551

Difference
(Option 1 Option 2)
(233,853)

500,841

266,246

(234,595)

$535

$411

($124)

Up to $259

($0)

$626

$567

($59)

$1,658 - $5,634

($0)

$7,220

($0)

Option 2

Option 1

Costs are lower in Option 1, relative to Option 2, because fewer RFEs would need to
comply with the proposed rule. However, if RFEs with 10 or fewer full-time equivalent
employees are exempt from Subpart S requirements, the timeliness, precision, and accuracy of
traceability efforts can be impacted and non-quantified benefits, such as enhancement of our
ability to narrow the number of lots in a recall and the ability of retail food establishments with
10 or fewer full-time equivalent employees to have the data necessary to quickly identify and
remove contaminated products from shelves, will be lessened in comparison to Option 2. The
importance of RFEs, including small RFEs, to the efficiency and expediency of traceback and
recalls is detailed in a memorandum by staff of FDA’s Center for Food Safety and Applied
Nutrition, Inclusion of Retail Establishments of all Sizes Under FSMA Section 204 (Ref. [1]).
136

To the extent that, in some outbreaks where small RFEs provide the only or best cluster
identification for traceback, our estimates relying on simply scaling by sales volume may
underestimate the full impact of exempting small RFEs.

B. Benefits
We provide a detailed description of benefits of this rule in sections I.B and II.E of this
analysis. In this section we describe the difference in the benefits between the two options.
As explained in section E.3, we estimate benefits for co-proposed option 1 in the same
manner as estimated for the proposed option 2.
Using the SUSB data, we estimate that RFEs with ten or fewer full-time equivalent
employees account for approximately 9.5 percent of total RFE revenues. Using percent of
revenue as a proxy for the percent of product that these RFEs handle – and in turn the percent of
foodborne illnesses related to these RFEs – we estimate the benefits under Option 1 by scaling
down our benefit estimates under Option 2 by 9.5 percent. We therefore estimate that
(undiscounted) public health benefits from Option 1 would range between $38 million and $1.6
billion per year with a primary estimate of $654 million per year. Additional benefits from
avoiding overly broad recalls could be the same as under option 2, ranging between $1.7 billion
and $5.6 billion.

C. Costs
Using the SUSB data, we estimate that 47 percent of all RFEs (approximately 420,000
RFEs) have ten or fewer full-time equivalent employees. We estimate that Option 1 would cover
in total approximately 188,291 firms (between 98,099 and 282,969 firms) and 331,858
137

establishments (between 180,280 and 494,115 establishments). Option 2 would in total cover
approximately 422,144 firms (between 224,679 and 636,809 firms) and 566,448 establishments
(between 307,246 and 849,095 establishments).
Table 42 compares the costs of Option 1 and Option 2; Table 42 summarizes the onetime and recurring costs of Option 1. Because Option 1 affects fewer entities, it would result in
lower total costs than Option 2. The largest reduction in costs in Option 1 is due to fewer
restaurants incurring costs of capital investment, which accounts for about 77 percent of the total
difference (Table 42). In total, Option 1 avoids $548 million in one-time costs ($1,113 million $565 million = $548 million) and $51 million in recurring costs ($387 million - $336 million =
$51). Many recurring costs result from provisions that do not apply to RFEs, and our estimates of
these costs are thus the same under both options.

Table 42. Comparison of Costs between Option 1 and Option 2 by Provision (undiscounted,
Millions 2018$)
Difference
One-time Costs
Option 2
Option 1
(Option 1 –
Option 2)
Reading and Understanding the
$52
$23
($29)
Rule
Capital Investment

$725

$356

($369)

Training

$174

$85

($89)

§ 1.1315 General Recordkeeping

$163

$101

($62)

$1,113

$565

($548)

§ 1.1315 General Recordkeeping

$20

$13

($8)

§ 1.1325 Growing Recordkeeping

$12

§ 1.1330 First Receiver
Recordkeeping

$16

$15

($1)

Total One-time Costs
Annually Recurring Costs

138

§ 1.1335 Receiver Recordkeeping

$98

$56

$(42)

§ 1.1340 Transformation
Recordkeeping

§ 1.1345 Creation Recordkeeping

§ 1.1350 Shipping Recordkeeping

$234

Total Recurring Costs

$387

$336

($51)

$1,500

$901

($599)

Total Costs

Table 43 below summarizes the low, primary and high undiscounted cost estimates of
Option 1 by provision. One-time costs for Option 1 would range from $23 million to $3,582
million. Recurring (annual) costs under Option 1 would range from $30 million to $1,917
million.

Table 43. Total Costs of Option 1 (Millions 2018$)
One-time Costs
Primary
Reading and Understanding the Rule
$23
Capital Investment
$356
Training
$85
§ 1.1315 General Recordkeeping
$101
Total One-time Costs
$565
Annually Recurring Costs
§ 1.1315 General Recordkeeping
$13
§ 1.1325 Growing Recordkeeping
$12
§ 1.1330 First Receiver Recordkeeping
$15
§ 1.1335 Receiver Recordkeeping
$56
§ 1.1340 Transformation Recordkeeping
$7
§ 1.1345 Creation Recordkeeping
$0.3
§ 1.1350 Shipping Recordkeeping
$234
Total Recurring Costs
$336

Low

High
$11
$5
$6
$6
$28

$39
$2,759
$317
$700
$3,816

$1
$0.2
$1
$6
$0.4
$0.01
$22
$30

$78
$102
$93
$287
$41
$2
$1,315
$1,917

Table 44 shows the timing of costs of Option 1 over ten years. The total costs of Option 1
are $1,055 million less than Option 2 ($4,980 million - $3,925 million = $1,055 million).
Annualized at a seven percent discount rate, the costs of Option 1 are $124 million less than the

139

costs of Option 2 ($535 million - $411 million = $124 million). At a three percent discount rate,
the costs of Option 1 are $113 million less ($513 million - $400 million = $113 million).

Table 44. Ten-Year Timing of the Costs of Option 1 (Millions 2018$)
Year
Primary
Low
1
$901
$58
2
$336
$30
3
$336
$30
4
$336
$30
5
$336
$30
6
$336
$30
7
$336
$30
8
$336
$30
9
$336
$30
10
$336
$30
Total Costs of Co-proposed Option 1
$3,925
$330
Present Value of Total Costs (3%)
$3,415
$285
Present Value of Total Costs (7%)
$2,888
$238
Annualized Value of Costs (3%)
$400
$33
Annualized Value of Costs (7%)
$411
$34

High
$5,733
$1,917
$1,917
$1,917
$1,917
$1,917
$1,917
$1,917
$1,917
$1,917
$22,990
$20,061
$17,033
$2,352
$2,425

D. International Effects
We expect international costs of the proposed rule under co-proposed Option 1 to be the
same as under Option 2. While it is possible that there might be a small number of foreign
entities that sell food in the United States and meet the definition of “retail food establishment,”
we assume that the number of such entities affected by this rule is negligible. We request
comment on this assumption.

E. Regulatory Alternatives
In Section II.J we considered five different regulatory alternatives to Option 2; in this
section IV.E we compare these regulatory alternatives to co-proposed Option 1. We estimate
140

costs of regulatory alternatives, noting that FDA is also proposing that the compliance date for
any final rule would be two years after the date the final rule is effective. We estimate costs of all
regulatory alternatives as if they have the same effective and compliance dates as Option 2.
The five alternative options we consider are Alternative a): No new regulatory action as
the baseline for determining the costs and benefits of other alternatives; Alternative b): Expand
the FTL rule to cover all foods in affected industries; Alternative c): Expand coverage to include
very small farms; Alternative d): Include foods transformed or created at the retail level; and
Alternative e): Extend compliance date to three years. Table 45 shows a detailed summary of the
costs associated with each regulatory alternative, and the change in the estimated costs relative to
those associated with the co-proposed Option 1.

Table 45. Summary of Costs by Regulatory Alternatives Compared to Option 1 (millions
2018$)
Costs
Regulatory Option Domestic
Recurring Costs Annualized
Facilities
One-Time Costs
($/year)
Total (7%)
Co-proposed Option 1
Regulatory Alternative a: No
Action
Change from Option 1
Regulatory Alternative b: Expand
coverage beyond FTL foods
Change from Option 1
Regulatory Alternative c: Expand
coverage for very small farms and
other originators
Change from Option 1

$565

$336

$411

($565)

($336)

($411)

$1,514

$454

$656

$949

$118

$245

$1,167

$394

$549

$603

$58

$138

141

Regulatory Alternative d: Include
foods transformed or created at the
retail level
Change from Option 1
Regulatory Alternative e: Extend
compliance date to three years
Change from Option 1

$1,113

$452

$600

$565

$336

$411

$565

$336

$362

($49)

F. Initial Small Entity Analysis
The Regulatory Flexibility Act requires Agencies to analyze regulatory options that
would minimize any significant impact of a rule on small entities. In Option 2 we found that,
because some small firms may have annualized costs (over ten years at a seven percent discount
rate) that exceed one percent of their annual revenue, the proposed rule would have a significant
economic impact on a substantial number of small entities. The following analysis of Option 1
described in this sub-section and the analysis of Option 2 in Section III as well as other sections
in this document, serve as the Initial Regulatory Flexibility Analysis, as required under the
Regulatory Flexibility Act.
For this analysis we report the difference in the impact to retail food establishments
between Option 2 and Option 1. Table 46 summarizes the annualized difference in costs between
Option 2 and Option 1 per small retail food establishment. The lower bound annualized cost per
small retail food establishment is greater under Option 1 ($57) compared to Option 2 ($45)
because the average cost per retail food establishment is greater under Option 1.
Table 46. Annualized Costs of Option 1 per Small Retail Food Establishment Firm (Ten
years, Seven Percent Discount Rate, Millions 2018$)
Costs as a
Costs as a
Retail Food
Lower
percent of
Upper
percent of
Establishments
Bound
revenue
Bound
revenue
(Lower)
(Upper)
142

Option 1
Option 2
Difference

$57
$45
$12

0.01%
0.004%
0.006%

$679
$582
$97

0.07%
0.05%
0.02%

As the vast majority (roughly ninety-three percent) of covered firms qualify as small
entities, the analysis of regulatory alternatives for covered firms described in Section II.J above
effectively describes the effects of the alternatives on small entities. Delaying the compliance
date for small entities would delay the implementation of the proposed rule for the vast majority
of FTL products. While the postponement of capital investments and labor expenses for
compliance would reduce the present value of costs of the proposed rule, it would also reduce the
present value of the health benefits. Under Option 1, we provide more flexibility to small entities
by exempting all retail food establishments with ten or fewer full-time equivalent employees.

143

V. References

[1]
[2]

[3]

[4]

[5]
[6]
[7]

[8]

[9]
[10]
[11]
[12]
[13]

[14]
[15]

K. Irvin, K. Blickenstaff, A. Fields and K. Vierk, Memorandum of Record, U.S. Food and
Drug Administration - Center for Food Safety and Applied Nutrition, 2020.
Centers for Disease Control and Prevention, "Multistate Outbreaks of Shiga toxinproducing Escherichia coli O26 Infections Linked to Chipotle Mexican Grill Restaurants
(Final Update)," 1 February 2016. [Online]. Available:
https://www.cdc.gov/ecoli/2015/o26-11-15/index.html.
Public Health Agency of Canada, "Public Health Notice Update - Outbreak of E.coli
infections with possible link to leafy greens," 25 May 2015. [Online]. Available:
https://www.canada.ca/en/public-health/services/public-health-notices/2015/public-healthnotice-update-outbreak-e-coli-infections-possible-link-leafy-greens.html.
K. Appold, "Challenges of Preventing Leafy Green Outbreaks," 2019. [Online]. Available:
https://www.foodqualityandsafety.com/article/fresh-produce-outbreaks-challengesprevention/?singlepage=1. [Accessed 31 August 2020].
J. McEntire and T. Bhatt, "Pilot Projects for Improving Product Tracing along the Food
Supply System – Final Report," Institute of Food Technologists, Illinois, Chicago, 2012.
GS1 US, "Foodservice GS1 US Standards Initiative," 2020. [Online]. Available:
https://www.gs1us.org/industries/foodservice/initiative.
GS1 US, "Improving Traceability and Food Safety with GS1 Standards in Fresh Foods,"
2012. [Online]. Available:
https://www.gs1us.org/DesktopModules/Bring2mind/DMX/Download.aspx?Command=Co
re_Download&EntryId=593&language=en-US&PortalId=0&TabId=134.
E. Golan, B. Krissoff, F. Kuchler, L. Calvin, K. Nelson and G. Price, "Traceability in the
U.S. Food Supply: Economic Theory and Industry Studies," Agricultural Economic Report,
vol. 830, 2004.
U.S. Department of Agriculture National Agricultural Statistics Service, "Census of
Agriculture," 2017.
U.S. Food and Drug Administration, "Prevention of Salmonella Enteritidis in Shell Eggs
During Production, Storage, and Transportation; Final Rule," Federal Register, 2009.
J. Zhang and T. Bhatt, "A guidance document on the best practices in food traceability,"
Comprehensive Reviews in Food Science and Food Safety, pp. 1074-1103, 2014.
Grocery Manufacturers Association, "Recall execution effectiveness:Collaborative
approaches to improving consumer safety and confiedence," GMA, Washington, DC, 2010.
Centers for Disease Control and Prevention, "Multistate Outbreak of Human Salmonella
Enteritidis Infections Associated with Shell Eggs," 9th October 2010. [Online]. Available:
https://www.cdc.gov/salmonella/2010/shell-eggs-12-2-10.html.
H. Roberts and S. R. Veil, "Health literacy and crisis: Public relations in the 2010 egg
recall," Public Relations Review, pp. 214-218, 2016.
Grocery Manufacturers Association, "Capturing recall costs:measuring and recovering the
losses," GMA, Washington DC, 2011.
144

[16] K. Smith, B. Miller, K. Vierk, I. Williams and C. Hedberg, "Product tracing in
epidemiologic investigations of outbreaks due to commercially distributed food items -utility, application, and considerations," 2015.
[17] U.S. Food and Drug Administration, CORE Outbreak Dataset (1999-2019).
[18] J. Blankenship, M. Bazaco, K. Irvin, K. Blickenstaff and A. Fields, Memorandum of
Record, U.S. Food and Drug Administration - Center for Food Safety and Applied
Nutrition, 2020.
[19] E. Scallan, R. M. Hoekstra,, F. J. Angulo, R. V. Tauxe, M. A. Widdowson, S. L. Roy, J. L.
Jones and P. M. Griffins, "Foodborne illness acquired in the United States--major
pathogens.," Emerging Infectious Disease, pp. 7-15, 2011a.
[20] R. Pennotti, E. Scallan, L. Backer, J. Thomas and F. Angulo, "Ciguatera and Scombroid
Fish Poisoning in the United States," Foodborne Pathogens and Disease, vol. 10, no. 12,
pp. 1059-1066, 2013.
[21] E. Scallan, P. M. Griffins, F. J. Angulo, R. V. Tauxe and R. M. Hoekstra, "Foodborne
illness acquired in the United States -- unspecified agents.," Emerging Infectious Disease,
pp. 16-22, 2011b.
[22] D. M. Tack, L. Ray, P. M. Griffin, P. R. Cieslak, J. Dunn, T. Rissman, R. Jervis, S.
Lathrop, A. Muse, M. Duwell, K. Smith, M. Tobin-D'Angelo, D. Vugia, J. Z. Kufel, B. J.
Wolpert, R. Tauxe and D. C. Payne, "Preliminary Incidence and Trends of Infections with
Pathogens Transmitted Commonly Through Food- Foodborne Disease Active Surveillance
Network, 10 U.S. Sites, 2016-2019," Morbidity and Mortality Weekly Report (MMWR), pp.
509-514, 2020.
[23] T. Minor, A. Lasher, K. Klontz, B. Brown, C. Nardinelli and D. Zorn, "The per case and
total annual costs of foodborne illness in the United States," Risk Analysis, pp. 1125-39,
2015.
[24] L. R. Scharff, "The Economic Burden of Foodborne Illness in the United States," in Food
Safety Economics: Incentives for a Safer Food Supply, Gewerbestrasse 11, 6330 Cham,
Switzerland, Springer International Publishing AG,, 2018, pp. 123-142.
[25] U.S. Department of Health and Human Services, "Guidelines for Regulatory Impact
Analysis," 2016.
[26] U.S. Food and Drug Administration, "FDA guidance for industry: Standards for the
growing, harvesting, packing, and holding of produce for human consumption: What you
need to know about the FDA regulation-Small entity compliance guide," FDA, Silver
Spring, MD, 2017.
[27] A. Fields, M. Salter and k. Irvin, Memorandum of Record, U.S. Food and Drug
Administration - Center for Food Safety and Applied Nutrition, 2020.
[28] C. Behravesh, R. K. Mody, J. Jungk and et.al., "2008 Outbreak of Salmonella Saintpaul
Infections Associated with Raw Produce," N Engl J Med 2011, vol. 364, pp. 918-27, 2011.
[29] L. A. Ribera, M. A. Palma, M. Paggi, R. Knutson, J. G. Masabni and J. Anciso, "Economic
Analysis of Food safety Compliance Costs and Fodborne Illness Outbreaks in the United
States," Hort Technology 22 (2), pp. 150-156, 2012.
[30] Eastern Research Group (ERG), "Recall Cost Case Studies," Eastern Research Group, Inc,
Massachussettes, 2020.
145

[31] U.S. Census Bureau, "2016 Statistics of U.S. Businesses (SUSB)".
[32] J. C. McEntire, S. Arens, M. Bernstein, B. Bugusu, F. F. Busta, M. Cole, A. Davis, W.
Fisher, S. Geisert, H. Jensen and B. Kenah, "Traceability (product tracing) in food systems:
An IFT report submitted to the FDA, Volume 1: Technical aspects and recommendations,"
Comprehensive Reviews in Food Science and Food Safety, vol. 1, no. 9, pp. 92-158, 2010.
[33] C. Mejia, J. McEntire, K. Keener, M. K. Muth, W. Nganje, T. Stinson and H. Jensen,
"Traceability (product tracing) in food systems: an IFT report submitted to the FDA,
Volume 2: Cost consideration and implications.," Comprehensive Reviews in Food Science
and Food Safety, pp. 159-175, 2010.
[34] U.S. Food and Drug Administration, "Memorandum: Economic Analysis for 204," U.S.
Food and Drug Administration, Office of Food Policy and and Response, Traceability &
Emerging Technology, 2020.
[35] GS1, "Integrated Traceability in Fresh Foods: Ripe Opportunity for Real Results," 2013.
[36] L. Christensen, B. Jablonski, L. Stephens and A. Joshi, "Evaluating the economic impacts
of farm-to-school procurement: An approach for primary and secondary financial data
collection of producers selling to schools," Journal of Agriculture, Food Systems, and
Community Development, vol. 8, no. Suppl 3, pp. 73-94, 2018.
[37] U.S. Department of Agriculture Food and Nutrition Services (USDA FNS), "2015 Farm to
School Census," 2015. [Online]. Available: https://farmtoschoolcensus.fns.usda.gov/about.
[Accessed January 2020].
[38] U.S. Food and Drug Administration, "Draft Guidance for Industry: Compliance with and
Recommendations for Implementation of the Standards for the Growing, Harvesting,
Packing, and Holding of Produce for Human Consumption for Sprout Operations," 2017.
[39] E. Bekkers, M. Brockmeier, J. Francois and F. Yang, "Pass Through, Food Prices and Food
Security," Working Paper, 15 April 2013.
[40] R. B. Freeman and R. H. Oostendorp, Occupational Wages around the World (OWW)
Database: Update for 1983-2008 [Data file and documentation]., Available from NBER
website: https://www.nber.org/oww/, 2012.
[41] A. Uçar, M. V. Yilmaz and F. P. Çakıroğlu, "Food Safety - Problems and Solutions," in
Significance, Prevention and Control of Food Related Diseases, Croatia, InTech, 2016.
[42] S. Charan and A. Panghal, "Importance of Traceability in Food Supply Chain for Brand
Protection and Food Safety Systems Implementation," Annals of Biology, pp. 111-118,
2018.
[43] U.S. Food and Drug Administration, "Regulatory Impact Anaysis: Standards for the
Growing, Harvesting, Packing, and Holding of Produce for Human Consumption Final
Rule," 2015.
[44] USDA - National Agricultural Statistics Service, "Poultry - Production and Value 2018
Summary," USDA, 2019.
[45] Associated Press, "Recall price tag $1 billion: And that could just be the beginning,
producers set to testify March 11, 2009," 11 March 2009. [Online]. Available:
http://www.nbcnews.com/id/29634279/ns/business-going_green/t/peanut-industry-recallprice-tag-billion/#.XZ47KmZ7mUK.

146

[46] Bureau of Labor Statistics, "Occupational Employment and Wages, May 2018 19-1012
Food Scientists and Technologists," 2018.
[47] T. Bhatt and J. Zhang, "Food product tracing technology capabilities and interoperability,"
Food Science, pp. 28-33, 2013.
[48] A. J. Collart, "The food safety modernization act and the marketing of fresh produce,"
Choices: Agriculture and Applied Economics Association, pp. 1-7, 2016.
[49] D. Sparling and B. Sterling, "Food traceability: understanding business value," RCM
Technology Canada, pp. 1-8, 2004.
[50] U.S. Food and Drug Administration, "Methodology Excerpt Risk-Ranking Model for
Product Tracing FSMA Section 204 (U.S. Code#2223)," FDA, College Park, MD, 2019.
[51] S. Hoffman, M. B. Batz and G. J. Morris Jr., "Annual cost of illness and quality adjusted
life years losses in the United States due to 14 foodborne pathogens.," Food Protection, pp.
1292-302, 2012.
[52] J. A. Painter, T. Ayers, R. V. Tauxe, R. M. Hoekstra, C. R. Braden, F. J. Angulo and P. M.
Griffin, "Attribution of foodborne illnesses, hospitalizasions and hdeaths to food
comodities by using outbreak data, United States, 1998 - 2008.," Emerging Infectious
Disease, pp. 407-415, 2013.
[53] S. Pouliot and D. A. Summer, "Treaceability, liability and incentives for food safety and
quality," American Journal of Agriculture Economics, pp. 15-27, 2008.
[54] U.S. Food and Drug Administration, "Standards for the Growing, Harvesting, Packing and
Holder of Produce for Human Consumption (FRIA)," 13 November 2015. [Online].
Available: https://www.fda.gov/media/94153/download.
[55] USDA National Agricultural Statistics Service, "Chickens and Eggs," USDA, 2020.
[56] Office of the Assistant Secretary for Planning and Evaluation (ASPE), "Guidelines for
Regulatory Impact Analysis," Department of Health and Human Services (HHS),
Washington DC, 2016.
[57] Institute of Food Technologists, "Pilot Projects for Improving Product Tracing along the
Food Supply System - Final Report," Institute of Food Technologists, Chicago, IL, 2012.

147

VI. Appendices
A. Food Traceability List (FTL)
The list of applicable foods might change prior to publication of the final rule. After
publication of the final rule, the list can be updated using the procedure set forth in proposed
§ 1.1465, if that provision is finalized. As of this writing, the list includes the following foods
(Table A.1).
Table A.1. Tentative Food Traceability List
Food Traceability List
Cheeses, other than hard cheeses

Description
Includes all soft ripened or semi-soft cheeses,
and fresh soft cheeses that are made with
pasteurized or unpasteurized milk

Shell eggs

Shell egg means the egg of the domesticated
chicken
Includes all types of tree nut and peanut butters;
does not include soy or seed butters

Nut butter

Cucumbers

Includes all varieties of cucumbers

Herbs (fresh)

Includes all types of herbs, such as parsley,
cilantro, basil, etc.
Includes all types of leafy greens, such as
lettuce, (e.g., iceberg, leaf and Romaine
lettuces), kale, chicory, watercress, chard,
arugula, spinach, pak choi, sorrel, collards,
endive, etc.
Includes all types of melons, such as cantaloupe,
honeydew, watermelon, etc.

Leafy greens, including fresh-cut leafy greens

Melons
Peppers
Sprouts
Tomatoes
Tropical tree fruits

Fruits and Vegetables (fresh-cut)

Includes all varieties of peppers
Includes all varieties of sprouts
Includes all varieties of tomatoes
Includes all types of tropical tree fruit, such as
mango, papaya, mamey, guava, lychee, jackfruit,
starfruit, etc.
Includes all types of fresh-cut fruits and
vegetables

148

Finfish, including smoked finfish

Crustaceans
Mollusks, bivalves

Ready-to-eat deli salads

Includes all finfish species, such as cod,
haddock, Alaska pollack, tuna, mahi mahi,
mackerel, grouper, barracuda, salmon, etc.;
except does not include siluriformes fish, such as
catfish
Includes all crustacean species, such as shrimp,
crab, lobster, crayfish, etc.
Includes all species of bivalve mollusks, such as
oysters, clams, mussels, etc.; does not include
scallop adductor muscle.
Includes all types of ready-to-eat deli salads,
such as egg salad, potato salad, pasta salad,
seafood salad, etc.; does not include meat salads

149

B. Methodology Used to Estimate the Number of Illnesses
To obtain the number of illnesses, hospitalizations, and deaths reported in Table 5, we
rely on FDA Coordinated Outbreak Response and Evaluation (CORE) data. We report these
CORE data, covering the 10.5-year period from 2009 through July 2019, in columns 1-4 of Table
B.1 in this appendix. We do not use CDC outbreak data for our estimates because those data
include illnesses resulting from improper food handling, as well as illnesses associated with
foods not regulated by FDA. Note that CORE data include more illnesses than those attributable
to FTL food products. These include adverse reactions and fungus-related illnesses. We therefore
use only the subset of CORE data on foodborne illness outbreaks associated with FTL foods. 73
Based on these data, of 31 known foodborne illness-causing pathogens, 10 are commonly
associated with foods currently designated by FDA on the FTL. We list these pathogens in Table
B.1.
To account for underreporting as well as underdiagnosing of foodborne illnesses, we
apply Scallan et al. (2011a) multipliers for each pathogen, presented in columns 5, 6, and 7 of
Table B.1. Column 5 contains Scallan et al.’s underdiagnosis multipliers specifically for
hospitalizations and deaths, taken from the authors’ Technical Appendix 3. Columns 6 and 7
contain the multipliers for underreporting and underdiagnosis of illnesses, also taken from Table
3.5 of the authors’ Technical Appendix 3. Columns 8, 9, and 10 show the resulting estimates of
the total number of cases, hospitalizations, and deaths from FTL foods over the 10.5-year period
from 2009 through July 2019. Scallan et al. (2011a) did not provide illnesses underreporting
multipliers for six pathogens (Ciguatoxins, Listeria, Norovirus, Salmonella, Scombrotoxins, and

73 We focus only on 2009-2019 outbreaks from this data set. Additionally, we subtract outbreaks related to
adverse reactions, fungus, and those attributed to non-FTL foods, leaving 191 out of 557 outbreaks that we used for
our analysis.

150

Tetrodotoxins), and for three of these pathogens (Ciguatoxins, Norovirus, and Scombrotoxins)
the authors did not provide underdiagnosis multipliers. Furthermore, at least three foodborne
illness-causing agents (Ciguatoxins, Scombrotoxin and Tetrodotoxins) did not have
hospitalization and death underdiagnosis multipliers, for which we assume a value of one. We
obtain underdiagnosis multipliers for illnesses from Ciguatoxins and Scombrotoxin from Pennotti
et al. (2013) (Ref. [20]), and use the same multiplier for Tetrodotoxins as for Ciguatoxins.
Following Scallan et al.’s (2011) treatment of other pathogens, we assumed 100 percent
reporting for pathogens without underreporting multipliers and 100 percent diagnosis for
pathogens without underdiagnosis multipliers. 74 To estimate the number of annual illnesses,
hospitalizations, and deaths caused by each pathogen, we divide columns 8, 9, and 10 of Table
B.1 by 10.5 years. Columns 11, 12, and 13 provide the resulting annual estimates. We estimate
that these pathogens cause 119,706 illnesses, 408 hospitalizations, and 17 deaths annually via
consumption of FTL-products.
Following Scallan et al. (2011b), we multiply the CORE number of illnesses from FTL
foods by a factor of five to account for the roughly 80 percent of cases caused by unspecified
agents. According to Scallan et al. (2011b) and CDC 75, nearly 80 percent of foodborne illnesses,
53 percent of hospitalized foodborne illnesses, and 58 percent of deaths from foodborne illnesses
result from unspecified or unknown pathogens. Following Scallan et al. (2011b), we multiply the
CORE number of illnesses from FTL foods by a factor of five to account for the roughly 80
percent of cases caused by unspecified agents; we also multiply the number of hospitalization
and deaths by 2.13 (= 1/(1 - 0.53)) and 2.28 (= 1/(1 - 0.58)) respectively to account for

74 For Norovirus, we assumed the same overall case count underdiagnosis multiplier as for non-typhoidal
Salmonella, which is 29.3.
75 https://www.cdc.gov/foodborneburden/2011-foodborne-estimates.html

151

unspecified agents. Not considering the burden caused by unspecified and unknown pathogens
will result in substantial underestimation of FTL caused illnesses. We assume the same ratios for
unidentified to identified cases due to FTL foods. The assumption is consistent with FDA’s past
regulatory impact analyses, including Standards for the Growing, Harvesting, Packing, and
Holding of Produce for Human Consumption of 2015. This assumption was made because
outbreak data on unidentified pathogens, specifically their associated food commodity, is
extremely sparse. The approach presumes that the percentage of identified illnesses, across all
pathogens, attributable to FTL products would be lower than the percentage of illnesses from
unidentified pathogens attributable to same products. The last row of Table B.1 present estimates
of total illnesses, hospitalizations, and deaths from listed foods after scaling for unspecified
and/or unknown agents. We estimate that in total about 598,500 foodborne illnesses are
associated with FTL foods (= 119,706 cases from specified pathogens and 478,700 cases are
illnesses from unspecified/unidentified pathogens). We use these estimates in our Preliminary
Regulatory Impact Analysis. We also scale up the number of hospitalizations and deaths to
account for unspecified agents and estimate that in total 869 hospitalizations (= 408 / (1 - 0.53))
and 39 deaths (= 16.6 / (1 - 0.58)) are caused by FTL foods annually.
Due to the sparsity of outbreak data on unspecified agents, as well as on underreporting
and underdiagnosis of foodborne illnesses, our estimates are subject to assumptions described
above. We request comment on our approach to estimating the number of illnesses,
hospitalizations, and deaths associated with FTL foods.

152

Appendix Table B.1: Illnesses, Hospitalizations, and Deaths Attributable to Illness-Causing Pathogens Associated with FTL
Foods
(1)
Number of
FTL
Related
Outbreaks

Ciguatoxin
Cyclospora
E. coli (STEC)
O157
E-Coli (STEC)
non-O157
Hepatitis A
Virus
Listeria
Norovirus
Salmonella
non-typhoidal
Salmonella
typhoidal
Scombrotoxin
Tetrodotoxin
Vibrio-para
VibrioCholerae
Total from
specified
pathogens
Total
including
unspecified/un
identified
pathogens

Raw Count of FTL Illnesses

(2)
Cases

(3)
Hospitalizations

8
8

38
2,612

4
75

(4)
Death
s
0
0

654

241

118

613

203

20
8

334
329

314
5

62
0

8,521

1,242

21
1
8

56
1
98

2
1
14

0
0
0

191

13,398

2,148

(5)
Hospitalizatio
n and Death
Multipliers
due to
Underdiagnos
is

1
2
2
2
2
2
1.5
2
2
1
1
2
2

(6)
Scallan
Underrep
orting
Multiplier

(7)
Scallan/
Pennotti
Underdia
gnosis
Multiplier

Estimated Total (2009-2019)

1
1

9.91
83.1

(8)
Total
Cases
377
217,057

25.5

26.1

435,270

25.5

106.8

321,361

1.1

9.1

6,136

1
1

2.3
29.3

768
9,640

29.3

249,665

13.3

160

1
1
1.1

12.21
9.91
142.4

684
10
15,351

1.1

33.1

437

(9)
Hospital
ization
4
150
482

(10)
Deaths

Estimated Annual Total
(2009-2019)

(12)
Hospita
lization
0
14
46

(13)
Deaths

(11)
Total
Cases
36
20,672
41,454

30,606

406

584

628
8
2,484

124
28

73
918
23,778

60
1
237

12
0
3

2
1
28
-

65
1
1,462
42

0
0
3
0

0
0
0
0

1,256,915

4,287

174

119,706

408

6,284,575

9,120

414

598,531

869

0
0
2

C. Outbreak Case Studies Used in Estimation of Public Health Benefits
This dataset represents 15 foodborne outbreaks from 2007 – 2019 coordinated by FDA’s
Emergency Coordination Response Team (ECRT) (2007 – 2010) and FDA’s Coordinated
Outbreak Response and Evaluation (CORE) Network (2011 – 2019), yielding 15 Public Health
Benefit Case Studies. Each outbreak included in this analysis:
•

involved a major pathogen/contaminant (Salmonella, Escherichia coli (STEC), Listeria
monocytogenes, or Cyclospora cayetanensis);

•

involved FDA-regulated food(s) from the Food Traceability List (FTL) that was
identified as the outbreak vehicle and/or contaminated product;

•

involved a formal traceback investigation that was coordinated by FDA;

•

resulted in voluntary or enforced product interventions;

•

and resulted in public communications issued by FDA and/or CDC.
Definitions:
•

Year – The reported year that an outbreak was evaluated/investigated by FDA.

•

Pathogen/Contaminant – The identified pathogen or contaminant associated with an
outbreak according to the case definition, as defined by CDC.

•

Species/Serotype – The species/serotype(s) that corresponds to the reported pathogen
as determined by CDC.

•

Commodity – The item(s) identified by FDA as the outbreak vehicle and/or
contaminated product.

•

Response Start Date – The date that a given outbreak was transferred to a CORE
Response Team.

•

Traceback Initiation Date – The date that represents when FDA’s traceback
investigation began.
o This date represents when the first traceback information request was issued
for record collection.

•

Traceback Completion Date – The date that represents when FDA’s traceback
investigation (including the review of collected records and documentation of
findings) ended.
o This date represents when the last record was received by FDA for the
traceback investigation.

•

Response End Date – The date that a given outbreak was closed by a CORE
Response Team.

•

Final CDC Publication Date – The date of publication for the final outbreak web
posting or corresponding update issued by CDC.

•

Final CDC Web Post Link – The link to the final outbreak web posting or
corresponding update issued by CDC.
o For these case studies, the epidemiologic data that was used for the analysis
included the final case count, hospitalization, and death totals that were
publicly reported for a given outbreak.

Limitations:
The outbreaks used for the Public Health Benefit Case Studies were selected to represent
significant outbreaks involving some of the commodities identified on the FTL. It should be
noted that these cases studies do not represent all foodborne outbreaks investigated or traceback

155

investigations conducted from 2008 to 2019, nor do they represent all occurrences of an FTL
product being implicated as the cause of an outbreak during that timeframe. Also, several
outbreaks that were used for these case studies were excluded from the risk ranking analysis as
this effort was performed as of 6/25/2020.
Another limitation of this analysis is the use of publicly reported epidemiologic data
(case count, hospitalizations, deaths). For some outbreaks, the publicly reported values may
differ from the final values internally reported by FDA and/or CDC, including the data that was
used for the risk ranking analysis. Specifically, for outbreaks associated with Cyclospora
cayetanensis, lack of a validated molecular subtyping methodology made it difficult to
differentiate historical outbreaks that may have been occurring concurrently but were associated
with different products. This led to challenges regarding the attribution of epidemiologic data to
those distinct outbreaks and/or commodities.
Additionally, the level of FDA documentation that was readily available for the outbreak
investigations included in this analysis varied, especially for those outbreaks that occurred before
CORE was established in 2011. The Traceback Initiation and Completion Dates are estimations
that best represent when the traceback investigations started and ended based on data pulled from
varying sources documenting each outbreak (e.g., email correspondence, outbreak summary
documents, etc.). The values in the dataset represent the best data currently available for
comparing the investigational elements of interest across these outbreaks.

156

Table C.1.: Outbreak Case Studies Used for Estimation of Public Health Benefits
Year
2008
2008
2009
2010
2010
2011
2012
2012

Pathogen/
Species/Serotype
Contaminant
Litchfield
Salmonella
Salmonella
Saintpaul
Saintpaul
Salmonella
Enteritidis
Salmonella
E. coli
O145
monocytogenes
Listeria
O26
E. coli
O157:H7
E. coli

Commodity
Cantaloupe
Hot Peppers
Alfalfa Sprouts
Shell Eggs
Romaine Lettuce
Cantaloupe
Clover Sprouts
Spinach
*Leafy Greens

Response Start
Date
3/4/2008
5/28/2008
2/26/2009
7/26/2010
4/21/2010
9/7/2011
2/3/2012
11/1/2012

FDA Traceback
Completion Date
4/10/2008
7/17/2008
4/30/2009
8/31/2010
5/11/2010
11/23/2011
2/17/2012
11/29/2012

7/11/2013

9/26/2013

8/13/2013
2/19/2016
11/15/2018

11/7/2013
3/22/2016
12/17/2018

Response End
Date
4/15/2008
8/5/2008
5/1/2009
9/3/2010
5/11/2010
12/11/2011
5/22/2012
12/21/2012

12/2/2013

Cyclospora

cayetanensis

2016
2018

E. coli
E. coli

O157:NM
O157:H7

2019

Cyclospora

cayetanensis

Basil

7/11/2019

7/15/2019

7/30/2019

2/3/2020

9/30/2019

2019
2019

E. coli
Salmonella

O157:H7
Javiana

Romaine Lettuce
Cantaloupe

11/12/2019
12/6/2019

11/18/2019
12/6/2019

12/13/2019
1/8/2020

3/16/2020
3/18/2020

1/15/2020
2/18/2020

2/18/2016
11/9/2018

*Represents two concurrent outbreak investigations attributed to the same pathogen/contaminant, but different commodities.

12/16/2013

Final CDC
Publication Date
4/2/2008
8/28/2008
5/8/2009
12/2/2010
5/21/2010
8/27/2012
4/3/2012
12/20/2012

2013

*Cilantro
Alfalfa Sprouts
Romaine Lettuce

7/11/2013

FDA Traceback
Initiation Date
3/5/2008
6/1/2008
3/2/2009
8/4/2010
4/27/2010
9/11/2011
2/7/2012
11/1/2012

4/5/2016
3/25/2019

3/25/2016

Final CDC Web Post Link
https://www.cdc.gov/salmonella/2008/cantaloupes-4-2-2008.html
https://www.cdc.gov/salmonella/2008/raw-produce-8-28-2008.html
https://www.cdc.gov/salmonella/2009/raw-alfalfa-sprouts-5-8-2009.html
https://www.cdc.gov/salmonella/2010/shell-eggs-12-2-10.html
https://www.cdc.gov/ecoli/2010/shredded-romaine-5-21-10.html
https://www.cdc.gov/listeria/outbreaks/cantaloupes-jensen-farms/index.html
https://www.cdc.gov/ecoli/2012/O26-02-12/index.html
https://www.cdc.gov/ecoli/2012/o157h7-11-12/advice-consumers.html
https://www.cdc.gov/parasites/cyclosporiasis/outbreaks/investigation-2013.html
https://www.cdc.gov/ecoli/2016/o157-02-16/index.html

1/9/2019

https://www.cdc.gov/ecoli/2018/o157h7-11-18/index.html
https://www.cdc.gov/parasites/cyclosporiasis/outbreaks/2019/weekly/index.htm
l
https://www.cdc.gov/ecoli/2019/o157h7-11-19/index.html
https://www.cdc.gov/salmonella/javiana-12-19/index.html

File Type	application/pdf
File Title	Requirements for Additional Traceability Records for Certain Foods PRIA
Author	Food and Drug Administration
File Modified	2020-11-18
File Created	2020-09-11