Substances Generally Recognized as Safe FRIA

DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration

Substances Generally Recognized as Safe
Final Rule
Docket No. FDA-1997-N-0020 (formerly 97N-0103)

Final Regulatory Impact Analysis
Final Regulatory Flexibility Analysis
Final Unfunded Mandates Reform Act Analysis

Economics Staff
Office of Planning
Office of Policy, Planning, and Legislation
Office of the Commissioner

Table of Contents
I. Introduction and Summary

3

A. Introduction
B. Summary of Costs and Benefits

3
3

II. Final Regulatory Impact Analysis

5

A. Background and Need for Regulation
B. Comments
C. Baseline
D. Costs of the Rule
E. Benefits of the Rule
F. Distributional Effects
G. International Effects
H. Uncertainty and Sensitivity Analysis
I. Analysis of Regulatory Alternatives to the Rule
III. Final Regulatory Flexibility Analysis

6
7
10
12
16
18
18
19
24
24

2

I. Introduction and Summary
A. Introduction
For this final regulatory impact analysis, we use the following terms: “GRAS substance”
refers to a substance not subject to premarket approval as a food additive under the Federal Food,
Drug, and Cosmetic Act (FD&C Act) because the substance is generally recognized as safe
(GRAS) under the conditions of its intended use in food for humans or animals under Section
201(s) of the FD&C Act (21 U.S.C. 321(s)) and the criteria for eligibility for classification as
GRAS in 21 CFR 170.30 or 21 CFR 570.30; “GRAS conclusion” refers to a conclusion that a
substance is GRAS under the conditions of its intended use within the meaning of Section 201(s)
of the FD&C Act and the criteria for eligibility for classification as GRAS in 21 CFR 170.30 or
21 CFR 570.30; and “GRAS petition” refers to a GRAS affirmation petition that had been
provided for under 21 CFR 170.35(c) or 21 CFR 570.35(c).
The final rule amends certain criteria in the regulations related to the GRAS provision of
the FD&C Act, and replaces a voluntary GRAS petition process with a voluntary GRAS
notification procedure. A substance must be GRAS or otherwise excepted from the definition of
a food additive before adding it to human or animal food without approval as a food additive.
Once you conduct a review of safety information and of general recognition of safety by
qualified experts, and if you reach a GRAS conclusion, you can decide to submit a GRAS notice
to us for our evaluation. We clarify the procedure that you can follow to notify us of your GRAS
conclusion, the type of documentation we expect you to submit with a GRAS notice to support
your GRAS conclusion, and how we will respond to your notice.
B. Summary of Costs and Benefits

3

The final rule will eliminate the petition process to affirm a substance is GRAS under the
conditions of its intended use and replace the petition process with a GRAS notification
procedure. The level of effort required by a firm to reach a conclusion that a substance is GRAS
under the conditions of its intended use remains unchanged by the final rule. We estimate that
from 350 to 480 notifiers may voluntarily decide to participate in the GRAS notification
procedure and submit on average from 60 to 75 GRAS notices each year. These notifiers will
incur one-time administrative costs to read and understand the rule and revise their standard
operating procedures. In addition, from 40 to 45 firms with outstanding GRAS affirmation
petitions may also incur one-time costs to prepare and submit a GRAS notice. Because the final
rule will require that notifiers submit more information to support their GRAS conclusion than
they currently provide with their notices, we include annual incremental costs for this additional
information.
We estimate that over 10 years with a 7 percent discount rate, the present value of the
total costs of the final rule range from $0.9 million to $3.3 million; with a 3 percent discount rate,
the present value of the total costs range from $0.9 million to $3.4 million. The annualized costs
of the rule range from $0.1 million to $0.4 million with a 7 percent discount rate and range from
$0.1 million to $0.5 million with a 3 percent discount rate. We do not quantify the benefits of
the final rule, but assume that firms will spend the same level of effort for a GRAS petition and a
GRAS notice to reach a GRAS conclusion. Thus, firms will voluntarily participate in the GRAS
notification procedure when they expect to receive a non-negative private benefit. Clarifying the
information we expect from firms that reach a GRAS conclusion will allow us to complete our
evaluation of a GRAS notice within the timelines specified in the final rule. We anticipate that
the more timely evaluation of GRAS notices, compared to the rulemaking process for GRAS

4

petitions, will create an incentive for firms to submit GRAS notices. The net benefit of the final
rule will depend on the quantity of GRAS notices submitted and the difference in the values
firms assign to a GRAS notice and a GRAS petition.
In table 1, we provide the Regulatory Information Service Center/Office of Information
and Regulatory Affairs Consolidated Information System accounting information.

Table 1: Summary of Benefits, Costs and Distributional Effects of Final Rule
Category

Benefits

Costs

Transfers

Effects

Annualized
Monetized
$millions/year
Annualized
Quantified
Qualitative

Primary
Estimate

Low
Estimate

High
Estimate

The final rule will reduce the
delay to evaluate industry
submissions on GRAS
substances.
$0.3
$0.1
$0.4
$0.3
$0.1
$0.5

Annualized
Monetized
$millions/year
Annualized
Quantified
Qualitative
Federal
Annualized
Monetized
$millions/year
From/ To
From:
Other
Annualized
Monetized
$millions/year
From/To
From:
State, Local or Tribal Government: No effect
Small Business: No effect
Wages: No effect
Growth: No effect

5

Year
Dollars

Units
Discount
Rate
7%
3%

Period
Covered

7%
3%

2014
2014

7%
3%
7%
3%
7%
3%

To:

To:

7%
3%

10 years
10 years

Notes

II. Final Regulatory Impact Analysis
We have examined the impacts of the final rule under Executive Order 12866, Executive
Order 13563, the Regulatory Flexibility Act (5 U.S.C. 601-612), 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). We have
developed a comprehensive Economic Analysis of Impacts that assesses the impacts of the final
rule. We believe that this final rule is not a 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. The final rule replaces the voluntary
GRAS petition process with a voluntary GRAS notification procedure. Similar to the petition
process, we expect that profit-maximizing firms will only submit the GRAS notice when the
private benefits equal or exceed the costs of the GRAS notice, regardless of the size of the firm.
Because small firms face the same voluntary business decision as large firms, we certify that the
final rule will not 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 $144 million, using the most current (2014) Implicit Price Deflator for the Gross

6

Domestic Product. This final rule will not result in an expenditure in any year that meets or
exceeds this amount.
A. Background and Need for Regulation
In 1997, we published a proposed rule that would amend our regulations regarding GRAS
substances for use in human food or in food for animals. At that time, firms could voluntarily
participate in a GRAS petition process. Of the 93 GRAS petitions (for substances intended for
use in human food) that we filed between 1972 and 1996, we completed rulemaking for 24 of
those petitions 1. Because this process requires notice and comment rulemaking, the time to
review and issue regulations in response to these petitions ranged from 1.4 years to 19.8 years
and averaged 7.9 years.
In the proposed rule, we invited interested persons who determine 2 that a use of a human
food substance is GRAS to notify us of those determinations as described in the proposed rule
(62 FR 18938 at 18954; the “Interim Pilot program”). Shortly after publication of the proposed
rule, we began accepting and evaluating GRAS notices for human food substances. By the end
of 2015, we had filed over 600 GRAS notices and sent 474 GRAS “no questions letters”, 17
“insufficient basis letters”, and 97 “cease to evaluate letters.” We make this information publicly
available on our website in a GRAS Inventory. In 2010, we published a notice in the Federal
Register (2010 notice) reopening the comment period for the proposed rule to update comments
because of the length of time that had elapsed since the publication of the proposed rule and
because we had identified a number of issues within the scope of the proposed rule that may
require further clarification. In June of that same year, we also launched an Interim Pilot
1

We note that some of these GRAS petitions came to completion through issuance of a food additive regulation
rather than a GRAS affirmation regulation.
2
We note that the final rule refers to GRAS “conclusions” rather than “GRAS determinations.”

7

program for animal food substances similar to the one for human food substances (75 FR 31800).
By the end of 2015, we had sent 7 GRAS “no questions letters”, 5 “insufficient basis letters”, and
6 “cease to evaluate letters” for GRAS notices for substances intended for use in animal food.
The final rule will replace the voluntary GRAS petition process with the voluntary GRAS
notification procedure and we are therefore ending our Interim Pilot program. The rule will
provide more clarity to firms considering whether a substance is GRAS under the conditions of
its intended use and considering submitting GRAS notices to us. Our GRAS petition process
required notice and comment rulemaking, a time-consuming process that created a disincentive
for firms to participate. We judge that the GRAS notification procedure provides a strong
incentive for firms to voluntarily participate in our program, and allows us to respond to GRAS
notices in a timely manner.
B. Comments
The agency received several public comments to the 1997 proposed rule and to the 2010
notice reopening the comment period for the proposed rule. Most comments address specific
aspects of the proposed rule or responded to specific questions we asked. A few of these
comments quote parts of the preliminary regulatory impact analysis (PRIA) or address broad
issues raised in the PRIA. However, none of the comments on the PRIA provide sufficient
information that we can use to revise our preliminary analysis. We group comments by general
topic: Potential costs savings of a GRAS notification procedure; the value of a GRAS notice; and
potential incentives or disincentives of the GRAS notification procedure.
Comment 1) Some commenters agree with our statement in the PRIA that the companies
spend the same effort to gather information that supports their GRAS conclusions for a GRAS
notice and a GRAS petition, and thus the cost savings from the rule would be modest. Other
8

commenters stated that the GRAS notification procedure would take less time than the GRAS
petition process, decreasing costs for both us and industry. One commenter noted that by setting
regulatory deadline for our responses to GRAS notices, the proposed rule would reduce
uncertainty and eliminate the long delays associated with the GRAS petition process.
Response 1) The GRAS petition process required rulemaking, a lengthy and time
consuming process. Comparing the GRAS notification procedure to the GRAS petition process,
we note that companies must devote the same level of effort to generate information sufficient to
reach a GRAS conclusion. Of the 93 GRAS petitions (for substances intended for use in human
food) that we filed between 1972 and 1996, we completed rulemaking for 24 of those petitions.
The time to complete rulemaking for those 24 GRAS petitions ranged from 1.4 years to 19.8
years and averaged 7.9 years. Results from our Interim Pilot program for human food show that,
on average, we sent “no questions letters,” “cease to evaluate letters,” and “insufficient basis
letters” for GRAS notices in 200 days after we filed a notice. Although we can’t quantify the
value of faster decisions, we expect that companies benefit from the much faster GRAS
notification procedure. Moreover, the final rule specifies reasonable timelines for our responses
to GRAS notices based on our real-world experience with the Interim Pilot programs.
Comment 2) Some commenters assert that food manufacturers would place less value on
a GRAS notice than a GRAS petition and would not participate in the program.
Response 2) Companies can use a food substance that is GRAS under the conditions of
its intended use without notifying us; the final rule does not change this. The final rule will
replace the GRAS petition process with the GRAS notification procedure. We disagree that
companies place less value on the GRAS notification procedure. Even though some companies
may prefer the GRAS petition process over the GRAS notification procedure, we see no

9

evidence that the GRAS notification procedure creates a disincentive for companies to
participate. Moreover, from 1998 through 2015, we filed over 600 GRAS notices for human
food substances in our Interim Pilot program. Such a large number of notices demonstrates that
notifiers gain some private benefit from GRAS notices and that notifiers value the GRAS
notification procedure and consider it a viable and cost-effective alternative to the GRAS petition
process.
Comment 3) One comment expressed concern that the GRAS notification procedure
would hurt consumer confidence in food products.
Response 3) We disagree that the GRAS notification procedure would hurt consumer
confidence, because the rule doesn’t change the statutory eligibility criteria for classification of a
substance as generally recognized as safe for its intended use in human or animal food. Notifiers
may still use a substance they conclude is GRAS without FDA evaluation or concurrence. The
success of the voluntary GRAS notification procedure has actually encouraged notifiers to
submit GRAS notices to us, increasing our awareness of the composition of our food supply and
the dietary exposure to GRAS substances. In addition, notified substances include substances
that are intended to address food safety problems (e.g., antimicrobial substances and substances
intended to reduce acrylamide formation) and public health issues (e.g., substances that would
reduce levels of sodium chloride in food). We conduct a substantive evaluation of the GRAS
notices that we receive to evaluate the basis for the notifier’s GRAS conclusion. We strive to
make the process transparent and have posted information about the human and animal GRAS
inventories on our website. We will continue to make this information accessible to food
manufacturers and consumers.

10

Comment 4) Some commenters stated that the more timely GRAS notification procedure,
including a regulatory timeframe for FDA to complete its evaluation of a GRAS notice, creates
an incentive for firms to participate. Some commenters supported the GRAS notification
procedure, but expressed concern that the success of the program depends on our timely response
to notices.
Response 4) We agree that the timely evaluation of GRAS notices will create an
incentive for companies to participate. In the final rule we clarify the information we expect
notifiers to submit in support of their GRAS conclusions. Removing uncertainty about our
expectations and setting realistic timelines for our evaluation of GRAS notices will further
encourage companies to participate in the GRAS notification procedure.
C. Baseline
In the proposed rule, we invited interested persons who determine that a human food
substance is GRAS under the conditions of its intended use to participate in the Interim Pilot
program. We explained that we would administer the notices as described in the proposed rule
(i.e., we would acknowledge receipt of the notice, respond in writing to the notifier, and make
publicly accessible a copy of all GRAS exemption claims and our response). We filed the first
GRAS notice for human food substances in 1998. By the end of 2015, we had filed over 600
GRAS notices for substances intended for use in human food and sent response letters for about
95 percent of these GRAS notices. Table 2 shows a breakdown of the response letters by
response time for GRAS notices for substances intended for use in human food. We sent more
than 60 percent of these response letters within 180 days; we sent more than 85 percent of these
response letters within 270 days. Even though we sent 17 insufficient basis letters since 1998,
we have not sent any of these letters since 2011.

11

Table 2. Response Letters for GRAS Notices for Substances Intended for Use in Human Food
by Response Time1 and Type of Letter
Response
Response Response
Response
Total
Time from
Percent of
Time from Time from
Time
Number of
181 Days
Letters by
1 Day to 91 Days to
More than Letters by
to 270
Type
2
90 Days
180 Days
270 Days
Type
Days
Cease to
evaluate letters
Insufficient basis
letters
No questions
letters

1

Total Number of
Letters by
Response Time
Percent of
Letters by
Response Time

41

33

8

15

97

16%

2

10

1

4

17

3%

20

268

129

57

474

81%

63

311

138

76

588

100%

11%

53%

23%

13%

100%

Response time for “cease to evaluate letters” equals the number of days between the date we filed the
GRAS notice and the date we stopped evaluating the GRAS notice; for the other two types of response
letters, response time equals the number of days between the date we filed the GRAS notice and the date
of our response letter to the GRAS notice.
2
We count the response letter for GRN13 as an “insufficient basis letter”.
Numbers may not sum due to rounding.

We established an Interim Pilot program for substances intended for use in animal food in
2010 and filed the first GRAS notice for a substance intended for use in animal food in
December 2010. By the end of 2015, we had filed 18 GRAS notices. Table 3 shows a
breakdown of the response letters by response time for GRAS notices for substances intended for
use in animal food. We sent about 28 percent of these response letters within 270 days; we sent
more than 95 percent of these response letters within 360 days. Moreover, we have not sent any
insufficient basis letters since 2012.

12

Table 3. Response Letters for GRAS Notices for Substances Intended for Use in Animal Food
by Response Time1 and Type of Letter
Response Response
Response
Response
Total
Time from Time from
Percent of
Time from
Time
Number of
181 Days 271 Days
Letters by
1 Day to
More than Letters by
to 270
to 360
Type
180 Days
360 Days
Type
Days
Days
Cease to
evaluate letters

1

3

Insufficient basis
letters
No questions
letters
Total Number of
Letters by
Response Time
Percent of
Letters by
Response Time

1

2

6

33%

5

5

28%

5

1

7

39%
100%

2

3

12

1

18

11%

17%

67%

6%

100%

1

Response time for “cease to evaluate letters” equals the number of days between the date we filed the
GRAS notice and the date we stopped evaluating the GRAS notice; for the other two response letters,
response time equals the number of days between the date we filed the GRAS notice and the date of our
response letter to the GRAS notice.

D. Costs of the Rule
The final rule will eliminate the petition process to affirm that a substance is GRAS under
the conditions of its intended use and replace the GRAS petition process with a GRAS
notification procedure. The level of effort required by a firm to reach a GRAS conclusion
remains unchanged by the final rule. Once firms reach their GRAS conclusions, they can
voluntarily decide to submit a GRAS notice to us. However, when a firm decides to submit a
GRAS notice to us, the firm must follow the requirements specified in the final rule. Because
the final rule will require some additional reporting compared to current practices for GRAS
notices (e.g., a certification statement, and identification of any information that the notifier
13

views as confidential), we estimate incremental annual costs for these additional requirements.
In addition to annual costs, notifiers will incur one-time costs to read and understand the rule,
and to revise their standard operating procedures (SOPs). Some firms have outstanding GRAS
petitions, for which we will be closing the docket. We anticipate that if these firms decide to
submit a GRAS notice they will use the information from their GRAS petition in their GRAS
notice. We include one-time costs for these firms to prepare and submit a GRAS notice for the
substance in their GRAS petition.
1. Who is affected by the Final Rule?
Using the North American Industry Classification System (NAICS), the government
classifies firms by their primary type of business operation. Firms most likely to submit a GRAS
notice fall into three basic industries. These include Food Manufacturing (NAICS code 311000),
Basic Chemical Manufacturing (NAICS code 325100), and Other Chemical Product and
Preparation Manufacturing (NAICS code 325900). Based on GRAS notices we filed during our
two Interim Pilot programs, as shown in table 4, we estimate that from 350 to 480 notifiers will
incur costs to comply with the final rule.
Table 4. Estimated Number of Affected Notifiers
Number of Human Food Substance Notifiers
Number of Animal Food Substance Notifiers
Total Number of Affected Notifiers

Lower Bound
Estimate
340
10
350

Upper bound
Estimate
460
20
480

2. Unit Costs
We use 2014 occupation specific wages from the US Bureau of Labor Statistics for these
three industries to calculate the average hourly wage for managers and clerical employees. We
selected types of employees likely to perform the actions required by the final rule. For

14

managers, we used the average hourly wages for business and financial operations (occupation
13-0000) and compliance officers (occupation 13-1041) as proxies; for clerical workers, we used
hourly wages for general office clerks (occupation 43-9061). As shown in table 5, the average
hourly wage for managers equals $66.60 with a 100-percent overhead; the average hourly wage
for clerical employees shown in table 6 equals $32.26 with a 100-percent overhead.
Table 5. Average Hourly Wages for Managers
NAICS Code

Average
Hourly Wage

Loaded
Average
Hourly Wage

Compliance Officers

311000

$29.04

$58.08

Business and Financial Operations

311000

$29.75

$59.50

Compliance Officers

325100

$39.44

$78.88

Business and Financial Operations

325100

$36.44

$72.88

Compliance Officers

325900

$33.04

$66.08

Business and Financial Operations

325900

$32.09

$64.18

$33.30

$66.60

Occupation

Total Average Hourly Wage

Table 6. Average Hourly Wages for Clerical Workers
Occupation

NAICS
Code

Average
Hourly
Wage

Office Clerks General

311000

$14.71

Loaded
Average
Hourly
Wage
$29.42

Office Clerks General

325100

$17.51

$35.02

Office Clerks General

325900

$16.17

$58.34

$16.13

$32.26

Total Average Hourly Wage

3. One-time costs
For all affected notifiers, we expect that they will spend time reading and understanding
the requirements of the final rule and revising SOPs for preparing and submitting GRAS notices.
15

Based on the length of the final rule and using average reading speed ranging from 200 to 250
words per minute as recommended by the Department of Health and Human Services, it will take
about 5.4 to 6.8 hours for one person to read and understand the rule. Due to the number of
provisions in the final rule, we anticipate that 2 or 3 managers will perform this task. As shown
in table 7, each affected notifier will incur a one-time cost to read and understand the rule that
range from about $720 to $1,350.
Table 7. Cost to Read and Understand the Final Rule

1

Lower Bound
Estimate
5.4
2
$66.60
$720.24

Number of Hours
Number of Managers
Hourly Wage1
Cost to Read and Understand the Final Rule
Source: Tables 5 and 6.

Upper bound
Estimate
6.8
3
$66.60
$1,350.45

We anticipate that notifiers will revise their SOPs to conform to the requirements of the
final rule. We estimate that notifiers will spend from 10 hours to 60 hours to perform this task.
Because revising and writing SOPs require effort from both clerical workers and managers, we
use a weighted average hourly wage of about $58.00 (25 percent for clerical workers and 75
percent for managers). As shown in table 8, each notifier will incur a one-time cost to revise and
write SOPs that range from about $580 to $3,480.
Table 8. Cost to Revise Standard Operating Procedures

1

Number of Hours
Weighted Average Hourly Wage1
Cost to Revise SOPs

Lower Bound
Estimate

10
$58.02
$580.15

Upper bound
Estimate

60
$58.02
$3,480.90

Source: Tables 5 and 6. Average weighted wage of $58.02 = (0.25 x 32.26 + .75 x $66.60).

16

Firms with outstanding GRAS petitions may choose to submit GRAS notices and
incorporate the information previously included in their petitions. To account for
the additional effort by these firms, we include the one-time cost to prepare and submit a GRAS
notice for up to 45 outstanding GRAS petitions for substances intended for use in human food.
As shown in table 9, we estimate that notifiers will spend between 170 and 190 hours divided
evenly between clerical workers and managers. With a weighted average wage of $49.43 (50
percent for clerical workers and 50 percent for managers), the one-time cost to prepare and
submit a GRAS notice will range from about $8,400 to $9,150 for each outstanding GRAS
petition.
Table 9a. Estimated Number of Outstanding GRAS Petitions

Lower
bound

GRAS petitions for substances intended for use in human food

Upper
bound
40

45

Table 9b. Cost to Prepare a GRAS Notice for an Outstanding GRAS Petition
Number of Hours
Weighted Average Hourly Wage1
1

Cost to Prepare a GRAS Notice

170

175

180

185

$49.43

$49.43

$49.43

$49.43

$8,403.10

$8,650.25

$8,897.40

$9,144.55

Source: Tables 5 and 6. Average weighted wage of $49.43 = (0.5 x 32.26 + 0.5 x $66.60).

4. Annual Costs
In addition to these one-time costs, the final rule will require that firms submit some
additional information to support their GRAS conclusion with their notices. Some examples of
the type of additional information we expect firms to submit include a certification statement or
identification of any information that the notifier views as confidential. We estimate that
notifiers will submit between 60 and 75 GRAS notices each year, 40 to 50 GRAS notices for

17

substances intended for use in human food and 20 to 25 GRAS notices for substances intended
for use in animal food. Although uncertain, we estimate that notifiers will spend between 5 more
hours and 20 more hours to prepare and submit each GRAS notice than they currently spend.
With a weighted average wage of $49.43 (50 percent for clerical workers and 50 percent for
managers), notifiers will incur incremental costs that range from $250 to $990 for each GRAS
notice submitted annually.
5. Total Costs of the Final Rule
Annual costs range from $0.01 million to $0.07 million with an average of $0.04 million.
The one-time costs range from $0.8 million to $2.7 million with an average one-time cost of $1.8
million. Over 10 years with a 7 percent discount rate, the present value of the total costs of the
final rule range from $0.9 million to $3.3 million; with a 3 percent discount rate, the present
value of the total costs range from $0.9 million to $3.4 million. The annualized costs of the rule
range from $0.1 million to $0.4 million with a 7 percent discount rate, and range from $0.1
million to $0.5 million with a 3 percent discount rate.
E. Benefits of the Rule
Although we lack sufficient data to quantify the benefits of the final rule, we include a
qualitative discussion of the expected benefits. Notifiers have voluntarily participated in the
Interim Pilot program for substances intended for use in human food since 1998; notifiers have
voluntarily participated in the Interim Pilot program for substances intended for use in animal
food since 2010. By the end of 2015, we had filed over 600 GRAS notices for substances
intended for use in human food and 18 GRAS notices for substances intended for use in animal
food. The response to these programs suggests that notifiers consider the voluntary GRAS
notification procedure a viable alternative to the voluntary GRAS petition process.
18

Through the GRAS notices we have already received, we have increased our awareness
of the composition of the nation’s food supply and the dietary exposure to GRAS substances,
which helps us to ensure the safe use of substances added to food. The ongoing submission of
GRAS notices provides evidence that our response to a GRAS notice can support the marketing
of a food substance by the regulated industry. Notified substances include substances that are
intended to address food safety problems (e.g., antimicrobial substances and substances intended
to reduce acrylamide formation) and public health issues (e.g., substances that would reduce
levels of sodium chloride in food). In addition, the letters we issue responding to GRAS notices
demonstrate that we inform notifiers of any scientific or regulatory issues that call into question a
notifier’s conclusion of GRAS status, and stakeholders have ready access to those letters. As
discussed in Response 81 of the preamble, we intend to increase the transparency of our response
letters when a notifier asks us to cease to evaluate a GRAS notice.
In the years since we published the proposed rule, we have in part focused resources on
postmarket actions with respect to the GRAS program, such as issuing a declaratory order
announcing our final determination that there is no longer a consensus among qualified experts
that partially hydrogenated oils are GRAS for any use in human food, as well as issuing warning
letters regarding the use of caffeine as an added ingredient in alcoholic beverages. We also have
taken other important public health actions with respect to substances used in food on the basis
of the GRAS provision of the FD&C Act. For example, we recently announced an initiative to
establish voluntary short-term and long-term goals for sodium reduction in a variety of identified
categories of foods to address the excessive intake of sodium in the current population and
promote improvements in public health (81 FR 35363, June 2, 2016). In addition, we recently
held a public meeting in which we invited public comment on what should be included, changed,

19

or even excluded from our guidance entitled “Guidance for Industry and Other Stakeholders:
Toxicological Principles for the Safety Assessment of Food Ingredients” (79 FR 64603, October
30, 2014); that guidance is intended to help interested parties understand our expectations
regarding how to determine which toxicity studies are appropriate and regarding the design,
conduct, and reporting of the results of toxicity studies and applies to assessing the safety of
GRAS substances.
The final rule clarifies the criteria that notifiers must use when they consider whether a
substance is GRAS under the conditions of its intended use. Making our expectations more clear
will reduce uncertainty about the procedure and encourage notifiers to continue their
participation in the voluntary GRAS notification procedure. Making our expectations more clear
also will help notifiers prepare GRAS notices and will allow us to complete our evaluation
within the timelines specified in the final rule.
Based on the response to the Interim Pilot programs, we assume that industry benefits
from a more timely evaluation under the voluntary GRAS notification procedure than under the
voluntary GRAS petition process. For example, during the Interim Pilot program for human
food substances, we took an average of 200 days after filing a notice to send 588 response letters;
we took an average of 7.9 years to complete rulemaking for 24 GRAS petitions. As expected
with any pilot program, the Interim Pilot program for substances intended for use in human food
evolved over time as parties discovered more efficient ways to comply with the program.
Although the Interim Pilot program for animal food has existed for a much shorter time
than the program for human food, we see similar time savings with the GRAS notification
procedure. Under this Interim Pilot program, we took an average of 276 days after filing a notice
to send 18 response letters; we took an average of 4.9 years to complete rulemaking for the 3
20

previous GRAS petitions. We assume that notifiers will voluntarily participate in the GRAS
notification procedure when the notifier expects to receive some benefit from the participation.
The final rule clarifies our expectations about the information notifiers should submit with a
GRAS notice for a substance intended for use in animal food. We expect that by removing some
uncertainty that may have discouraged participation by notifiers in the animal food industry, the
final rule will encourage participation of notifiers in the GRAS notification procedure for
substances intended for use in animal food.
F. Distributional Effects
We estimate that the final rule will have no distributional effects.
G. International Effects
Foreign notifiers who voluntarily choose to submit GRAS notices to us must comply with
the same requirements as domestic notifiers. Based on the GRAS notices filed since 1998, about
45 percent of the notifiers come from foreign establishments located in 29 countries. Table 10
shows a distribution of foreign notifiers by country; notifiers from Japan, Canada, The
Netherlands and China represent about 50 percent of all foreign notifiers. Because the final rule
will not change the incentives for foreign notifiers to submit GRAS notices to us, we anticipate
that the final rule will have no international effects.
Table 10. Number and Share of Foreign Notifiers by Country
Number of Share of
Foreign
Country
Country
Foreign
Notifiers
Notifiers
Japan
34
20% South Korea
Canada
27
16% Hong Kong
The Netherlands
14
8% Sweden
China
11
6% Thailand
21

Number
of Foreign
Notifiers
3
2
2
2

Share of
Foreign
Notifiers
2%
1%
1%
1%

Germany
Spain
United Kingdom
Australia
France
India
Israel
Denmark
Belgium
Finland

10
9
7
6
5
5
5
4
3
3

New Zealand

3

Norway

3

6%
5%
4%
4%
3%
3%
3%
2%
2%
2%

USA and China
Chile
Iceland
Ireland
Italy
Luxembourg
Mexico
Peru
Singapore
Switzerland
The Netherlands
2%
and Germany
2% USA and Germany

2
1
1
1
1
1
1
1
1
1

1%
1%
1%
1%
1%
1%
1%
1%
1%
1%

1

1%

1

1%

H. Uncertainty
The final rule replaces the voluntary GRAS petition process with a voluntary GRAS
notification procedure. The incentives to voluntarily participate in the GRAS notification
procedure remain unchanged. Firms will participate when they expect to receive a private
benefit. For our final regulatory impact analysis, we base our estimates on the number of firms
that submitted GRAS notices since 1998 and the number of GRAS petitions currently
outstanding. To account for uncertainty, we used ranges for the number of affected firms and the
incremental burden of the final rule once firms voluntarily decide to submit a GRAS notice.
As shown in table 11, the number of GRAS notices filed for substances intended for use
in human food varies year to year, but on average has increased over time. The total number of
notices strongly suggests that firms will continue to participate in the GRAS notification
procedure for substances intended for use in human food.

Table 11. Number of GRAS Notices for Human Food Substances Filed by Filing Year
Number of Notices
Number of Notices
Filing Year
Filing Year
Filed
Filed
22

1998
1999
2000
2001
2002
2003
2004
2005
2006

12
23
30
28
26
23
20
26
30

2007
2008
2009
2010
2011
2012
2013
2014
2015

20
36
36
55
45
41
43
69
51

In comparison to GRAS notices for substances intended for use in human food, we had
filed fewer than 20 GRAS notices by the end of 2015 for substances intended for use in animal
food, with more than half of these filed in 2011. In addition, as shown in table 12, we filed no
GRAS notices in 2015. The final rule clarifies our expectations and the type of information
notifiers should include in their GRAS notices. This will remove some uncertainty that may
have discouraged firms from submitting GRAS notices for substances intended for use in animal
food in the past. Although uncertain, we expect that the number of GRAS notices for animal
food substances will increase as firms gain experience with the GRAS notification procedure.

Table 12. Number of GRAS Notices for Animal Food Substances Filed by Filing Year
Filing Year
Number of Notices Filed
2010
2
2011
11
2012
1
2013
1
2014
3
2015
0

I. Analysis of Regulatory Alternatives to the Rule
Although we received no comments on the regulatory alternatives discussed in our
preliminary regulatory impact analysis, several comments suggested that we retain the GRAS
23

affirmation petition process in addition to the GRAS notification procedure. Commenters
suggested that they could use our GRAS affirmation to promote their substances in certain
markets. This alternative would come with substantial costs—because the GRAS affirmation
petition process requires rulemaking, which can take many years —without generating any social
benefits. Although we lack information to estimate the costs and benefits of this alternative, we
expect that the net social costs of this alternative would greatly exceed the net social costs of the
final rule.
III. Final Regulatory Flexibility Analysis
The Regulatory Flexibility Act requires a Regulatory Flexibility Analysis (RFA) unless
the Agency can certify that the final rule would have no significant impact on a substantial
number of small entities. The Small Business Administration (SBA) establishes thresholds for
small entities by North American Industry Classification System (NAICS); the SBA considers
small any entity below these thresholds.
The final rule will affect all small entities that voluntarily participate in the GRAS
notification procedure. We expect that this will primarily affect entities in the food
manufacturing and chemical manufacturing industries. For manufacturing industries, SBA uses
employment size to determine firm size. The threshold number of employees varies from 500
employees to 1,250 employees depending on the particular manufacturing sector. Although this
may overestimate the number of small entities affected by the final rule, we use 500 employees
as the cut-off between small and large entities. As shown in tables 13 and 14, we consider at
least 98 percent of the food manufacturing establishments and 98 percent of chemical
manufacturing establishment as small. However, the rule will not affect a substantial number of
entities in these industries.
24

Table 13. Food Manufacturing Establishments by Employment Size (NAICS 311)

Number of
Establishments
Total value of
shipments ($ mil)
Average per
Establishment Value
of Shipments ($ mil)

Fewer than
20
Employees

20 to 99
Employees

100 to 499
Employees

500 or
more
Employees

17,458

5,017

2,618

526

25,619

$33,335

$140,709

$344,067

$221,162

$739,272

$1.9

$28.0

$131.4

$420.5

$28.9

Industry
Total

Percent of
Small

98%

Source: 2012 Economic Census of the United States, EC1231SG2: Manufacturing: Summary Series:
General Summary: Industry Statistics for Subsectors and Industries by Employment Size: 2012

Table 14. Chemical Manufacturing Establishments by Employment Size (NAICS 325)

Number of
Establishments
Total value of
shipments and
receipts for services
($ mil)
Average per
Establishment
Receipts ($m mil)

Fewer than
20
Employees

20 to 99
Employees

100 to 499
Employees

500 or
more
Employees

7,851

3,888

1,367

211

13,317

$29,370

$154,230

$336,760

$264,939

$785,300

$3.7

$39.7

$246.3

$1,255.6

$59.0

Industry
Total

Percent of
Small

98%

Source: 2012 Economic Census of the United States, EC1231SG2: Manufacturing: Summary Series:
General Summary: Industry Statistics for Subsectors and Industries by Employment Size: 2012

We expect that profit-maximizing firms will only submit the GRAS notice when the
private benefits equal or exceed the costs of preparing and submitting the GRAS notice,
regardless of the size of the firm. Affected notifiers will incur one-time costs to read and
understand the final rule and revise their SOPs. We estimate that these costs range from $1,300
to $4,830 for each notifier. Some notifiers will incur annual reporting costs ranging from $250
to $1,000. Even with our upper bound cost estimate, the per notifier annualized costs equal less
than $1,700. This accounts for less than 0.3 percent for the smallest manufacturing
establishments that employ fewer than 5 employees. Because affected notifiers will incur
25

minimal costs compared to their annual revenues, we certify that the final rule will not have a
significant economic impact on a substantial number of small entities.

26