Download:
pdf |
pdfNutrition Facts/Serving Size Combined Final RIA
REGULATORY IMPACT ANALYSIS FOR FINAL RULES ON:
“FOOD LABELING: REVISION OF THE NUTRITION AND
SUPPLEMENT FACTS LABELS”
DOCKET NO. FDA-2012-N-1210
AND
“FOOD LABELING: SERVING SIZES OF FOODS THAT CAN
REASONABLY BE CONSUMED AT ONE EATING OCCASION; DUALCOLUMN LABELING; UPDATING, MODIFYING, AND ESTABLISHING
CERTAIN REFERENCE AMOUNTS CUSTOMARILY CONSUMED;
SERVING SIZE FOR BREATH MINTS; AND TECHNICAL
AMENDMENTS”
DOCKET NO. FDA-2004-N-0258 (formerly Docket No. 2004N-0456)
REGULATORY FLEXIBILITY ANALYSIS
PAPERWORK REDUCTION ACT OF 1995
1
Nutrition Facts/Serving Size Combined Final RIA
Executive Summary
The Nutrition Facts Label Final Rule amends our labeling regulations for foods to provide
updated nutrition information on the label to assist consumers in maintaining healthy dietary
practices. The Serving Size Final Rule amends our regulations on serving sizes based on newer
consumption data and other recent scientific evidence to provide updated serving size
information on the label to assist consumers in maintaining healthy dietary practices. These final
rules’ impacts are characterized by substantial uncertainty, with annualized benefits shown in our
sensitivity analysis to potentially range between $0.2 and $2 or $5 billion, and annualized costs
ranging between $0.2, $0.5 and $0.8 billion, over the next twenty years (2014 dollars) at seven
percent interest. This analysis will be made available at:
http://www.fda.gov/AboutFDA/ReportsManualsForms/Reports/EconomicAnalyses/.
2
Nutrition Facts/Serving Size Combined Final RIA
Table of Contents
I. Analysis of Impacts ................................................................................................................... 5
A.
Introduction ...................................................................................................................... 5
B.
Summary of the Final Rules ............................................................................................. 6
C.
Summary of Costs and Benefits ....................................................................................... 9
D.
Need for Regulation ....................................................................................................... 12
E.
Single Analysis for the Two Final Rules ....................................................................... 17
F.
Coverage of the Rule and Industry Overview ................................................................ 17
G.
Comments on the Preliminary Regulatory Impact Analysis (PRIA) and the
Supplemental PRIA and Our Responses................................................................................... 21
H.
Costs and Benefits of Regulatory Options – Detailed Analysis..................................... 45
1. Option 1: No new federal regulatory action ......................................................................... 46
2. Option 2: The final rules that would give small manufacturers a 3-year compliance period
and large manufacturers a 2-year compliance period ............................................................. 46
a. Costs ................................................................................................................................. 48
b. Benefits ............................................................................................................................ 73
3. Option 3: The final rules, but with a 2-year compliance period for all manufacturers ....... 89
a. Costs ................................................................................................................................. 89
b. Benefits ............................................................................................................................ 94
4. Option 4: The final rules, but with a 3-year compliance period for all manufacturers ....... 94
a. Costs ................................................................................................................................. 94
b. Benefits ............................................................................................................................ 98
5. Option 5: The final rules, but with a 4-year compliance period for all manufacturers ....... 99
a. Costs ................................................................................................................................. 99
b. Benefits .......................................................................................................................... 103
6. Summary of Estimated Costs and Illustratively Calculated Benefits by Regulatory Option
................................................................................................................................................. 103
II. Regulatory Flexibility Analysis........................................................................................... 105
A.
Introduction .................................................................................................................. 105
B.
Economic Effects on Small Entities ............................................................................. 105
C.
Regulatory Options ...................................................................................................... 106
D.
Summary ...................................................................................................................... 107
3
Nutrition Facts/Serving Size Combined Final RIA
III. PRA of 1995 ........................................................................................................................ 107
A.
Nutrition Facts Label Rule ........................................................................................... 107
B.
Serving Size Rule ......................................................................................................... 116
4
Nutrition Facts/Serving Size Combined Final RIA
I. Analysis of Impacts
A.
Introduction
FDA (FDA or we) has examined the impacts of the final rules 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). The final rules (collectively, the
nutrition labeling final rules or the final rules) include the following:
1. Title: Food Labeling: Revision of the Nutrition and Supplement Facts Labels. (Docket
No. FDA-2012-N-1210)
2. Title: Food Labeling: Serving Sizes of Foods That Can Reasonably Be Consumed At
One Eating Occasion; Dual-Column Labeling (DCL); Updating, Modifying, and Establishing
Certain Reference Amounts Customarily Consumed (RACCs); Serving Size for Breath Mints;
and Technical Amendments. (Docket No. FDA-2004-N-0258 (formerly Docket No. 2004N0456)).
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 are publishing two final rules on
nutrition labeling in the Federal Register. We have developed one final Regulatory Impact
Analysis (RIA) that assesses the impacts of the two final rules taken together. We believe that
the final rules, taken as a whole, are 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. Additional costs per entity from the
5
Nutrition Facts/Serving Size Combined Final RIA
final rules are small, but not negligible, and as a result we find that the final rules, taken as a
whole, 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 issuing
"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
Domestic Product. These final rules, taken as a whole, would result in an expenditure that meets
or exceeds this amount.
B.
Summary of the Final Rules
Food Labeling: Revision of the Nutrition and Supplement Facts Labels (the Nutrition Facts
Label rule)
The purpose of this final rule is to amend our labeling regulations for foods to provide
updated nutrition information on the label to assist consumers in making informed purchasing
decisions. As such, we are:
Updating the list of nutrients that are required or permitted to be declared.
Updating the Daily Reference Values (DRVs) and Reference Daily Intakes (RDIs) based
on current dietary recommendations from consensus reports.
Establishing a DRV for added sugars and requiring the declaration of the percent Daily
Value (DV) for added sugars.
Providing a definition of dietary fiber.
6
Nutrition Facts/Serving Size Combined Final RIA
Establishing DRVs and RDIs for nutrients in foods1 purported for infants up to 12
months, young children (1 through 3 years), and pregnant and lactating women.
Modifying the categorization of the subpopulation of children ages 2 through 4 years to
children 1 through 3 years of age.
Requiring that under certain circumstances manufacturers make and keep records
sufficient to verify the label declaration for the amount of added sugars, dietary fiber,
soluble fiber, insoluble fiber, vitamin E, and folate/folic acid in products and provide
these records upon request by FDA, during an inspection, for official review and
photocopying or other means of reproduction.
Increasing the prominence of calories and the serving size information; right-justifying
the quantitative amounts of the serving size information if adequate space is available;
removing the declaration of “Calories from fat”; declaring added sugars as an indented
listing, “Includes ‘X’ g Added Sugars,” directly beneath the listing for “Total Sugars”;
declaring the quantitative amounts - in addition to percent DVs - of mandatory vitamins
and minerals, and for voluntarily declared vitamins and minerals, when declared,
permitting the voluntary declaration of quantitative amounts and percent DVs; modifying
the footnote; and adding a horizontal line directly beneath the “Nutrition Facts” or
“Supplement Facts” heading.
Food Labeling: Serving Sizes of Foods That Can Reasonably Be Consumed At One Eating
Occasion; DCL; Updating, Modifying, and Establishing Certain RACCs; Serving Size for
Breath Mints; and Technical Amendments (the Serving Size rule)
1
Unless otherwise specified, the word food or foods refers to conventional foods and dietary supplements.
7
Nutrition Facts/Serving Size Combined Final RIA
The purpose of this final rule is to amend our regulations on serving sizes based on newer
consumption data and other current scientific evidence. As such, we are requiring that:
A product that is packaged and sold individually that contains less than 200 percent of the
applicable reference amount be considered a single-serving container, and the entire
content of the product be labeled as one serving, including products that have RACCs of
100 g (or 100 mL) or larger.
The serving size of a product in a multi-serving package containing discrete units in
which each discrete unit contains at least 200 percent of the applicable reference amount
be the amount that approximates the reference amount.
Products that are packaged and sold individually and that contain at least 200 percent and
up to and including 300 percent of the applicable reference amount, as well as discrete
units in multi-serving packages that contain at least 200 percent and up to and including
300 percent of the applicable reference amount, provide an additional column within the
Nutrition Facts label that lists the quantitative amounts and percent DVs for the entire
package or discrete unit, as applicable, as well as the column listing the quantitative
amounts and percent DVs for a serving that is less than the entire package or discrete
unit, as applicable (i.e., the serving size derived from the RACC).
RACCs for certain food product categories be updated based on newer food consumption
data available from the National Health and Nutrition Examination Survey (NHANES),
and that additional RACCs be modified or established where appropriate.
The label serving size for breath mints be “1 unit.”
A number of technical revisions be made to help clarify the serving size requirements in
§ 101.9 and § 101.12.
8
Nutrition Facts/Serving Size Combined Final RIA
C.
Summary of Costs and Benefits
Summary of Costs
Costs of complying with the final rules may include undertaking a minor label change or
an extensive or major label redesign; reformulating products to continue to make certain health
claims and nutrient content claims or as a result of the requirement to display new information on
the product label; recordkeeping costs associated with making and keeping records sufficient to
verify the label declaration for the amounts of added sugars, dietary fiber, soluble fiber, insoluble
fiber, vitamin E, and folate/folic acid; research and additional labeling costs related to the new
dietary fiber definition; and labeling costs associated with future Universal Product Code (UPC)
growth. Each is discussed in turn.
Labeling costs have been summed across an estimate of the total number of products for
which label changes may result from the final rules using FDA’s Labeling Cost Model (Ref. 1),
developed by Research Triangle Institute (RTI). We estimate that there would be over 30,000
manufacturers and close to 800,000 UPCs for which label changes may result after
implementation of the final rules. The estimated present value (PV) of the labeling costs of the
final rules, over the next 20 years2 and assuming the same compliance dates for both rules,
ranges from $1,405 million to $4,578 million, with an estimated mean3 cost of $2,720 million
2
We believe that 20 years is an appropriate time horizon to show the likely implications of the final rules. A tenyear time horizon might underrepresent the full effects of the rules due to limited or no benefits accruing prior to the
compliance dates. A time frame of more than 20 years begins to be rather speculative. After twenty years, it is
likely that food, drug, and supplement technologies, as well as medical treatments, will have changed enough so that
the potential benefits derived from these rules may be less impactful. Moreover, even though only 20 years’ worth
of changed consumption behavior is included in the analysis, the projected health impacts of the rule extend over the
entire lives of the affected consumers, and technology will likely have changed even further over that overall time
horizon than over the first two decades following rule finalization.
3
Throughout this document, “mean” refers to the mean of a distribution of values, not to the mean of the low and
high values.
9
Nutrition Facts/Serving Size Combined Final RIA
(2014$)4, using a 3 percent discount rate, and ranges from $1,331 million to $4,369 million, with
an estimated mean cost of $2,570 million (2014$), using a 7 percent discount rate. The
uncertainty in the estimates is due largely to the highly variable costs of printing methods
utilized by manufacturers (Ref. 1).
Reformulation costs have been summed across an estimate of the total number of
formulas for which reformulation is projected as a result of the final rules using FDA’s
Reformulation Cost Model (Ref. 2), developed by RTI. We estimate the PV of the reformulation
costs of the final rules over the next 20 years to be between $651 million and $2,682 million,
with an estimated mean cost of $1,446 million (2014$). The uncertainty in the estimates is due
to the highly variable costs of the various steps in the reformulation process undertaken by
manufacturers (Ref. 2).
The PV of recordkeeping costs over the next 20 years is estimated to be $21.4 million at a
3 percent discount rate and $21.3 million at a 7 percent discount rate (2014$). Related to the
new definition of dietary fiber, the PV of study costs over the next 20 years is estimated to be
roughly $4 million (2014$), and the PV of additional labeling costs over the next 20 years is
estimated to range from $176 million to $1,101 million, with a mean estimate of $430 million
(2014$).5 Finally, the PV of labeling costs associated with future UPC growth over the next 20
years is estimated to range from $66 million to $198 million, with a mean estimate of $135
million, using a 3 percent discount rate, and from $35 million to $107 million, with a mean
estimate of $72 million, using a 7 percent discount rate.
4
By 2014$ we mean that the amounts are in 2014 dollars.
The PVs of reformulation, fiber study, and additional labeling costs are equivalent at 3 or 7 percent because we
estimate that these one-time costs are incurred upon publication of the rules instead of at the end of the compliance
period.
5
10
Nutrition Facts/Serving Size Combined Final RIA
To summarize, the PV of total costs over the next 20 years ranges from $2,323 million to
$8,584 million, with a mean estimate of $4,756 million (2014$), using a 3 percent discount rate.
Using a 7 percent discount rate, the PV of total costs over the next 20 years ranges from $2,218
million to $8,284 million, with a mean estimate of $4,543 million (2014$).
Summary of Benefits
The Nutrition and Supplement Facts labels contain nutrient content information that can
help people maintain healthy dietary practices. The growth in the prevalence of obesity and
diabetes and high rates of chronic diseases such as heart disease and stroke in the United States
has elevated the treatment and prevention of these diseases to a top public health concern and a
national priority. The final rules will help consumers maintain healthy dietary practices by (i)
better aligning the information provided on the Nutrition and Supplement Facts labels with new
data on consumption, scientific evidence on the relationship between nutrition and chronic
disease, and dietary recommendations and (ii) improving the design and content of the Nutrition
and Supplement Facts labels to make relevant label information more salient and easy to
understand so that consumers can make more informed consumption decisions. The final rules
may also potentially prompt industry to reformulate products to maintain health and nutrient
content claims, and reformulate products based on new label information required by the final
rules.
To illustrate the quantification of the benefits of the final rules, we extrapolated from the
welfare effects estimated in a retrospective study on the impact of the Nutrition Labeling and
Education Act of 1990 (NLEA) by Abaluck (2011) (Ref. 3). Abaluck (2011) measured
consumer welfare gains as the willingness-to-pay (WTP) for nutrient content based on revealed
preference data, i.e., data on food consumption and prices. We lack direct evidence with which
11
Nutrition Facts/Serving Size Combined Final RIA
to scale Abaluck’s estimates of the effect of NLEA in a manner that precisely reflects the
impacts of other changes in nutrition labels; therefore, throughout this analysis we incorporate
uncertainty ranges so as to assess the sensitivity of benefits estimates to key analytic inputs. The
extent of the uncertainty is highlighted by the wide range of numerical benefits results; the PV
over the next 20 years ranges from $2.8 billion to $77.7 billion, with a mean estimate of $33.1
billion (2014$), at a discount rate of 3 percent, and ranges from $1.9 billion to $52.5 billion, with
a mean estimate of $22.3 billion (2014$), at a discount rate of 7 percent.6
Table 1 - Summary of the Primary Sensitivity Analysis of the Costs and Benefits of the Final
Rules (in billions of 2014$)
Benefits
Benefits Benefits
Costs
Costs
Costs
(Low)
(Mean)
(High)
(Low)
(Mean)
(High)
PV
$2.8
$77.7
$2.3
$8.6
3%
$33.1
$4.8
$1.9
$52.5
$2.2
$8.3
7%
$22.3
$4.5
Annualized Amount
$0.2
$0.6
3%
$0.2
$2.2
$5.2
$0.3
$0.2
$0.8
7%
$0.2
$2.1
$5.0
$0.4
Notes: Costs estimates reflect an assumption that the rules have the same compliance date. Compliance period is 36
months for small businesses and 24 months for large businesses. For purposes of this analysis, we consider a small
business to be a business with annual food sales of less than $10 million, and a large business to be a business with
annual food sales of $10 million or more. Costs include relabeling, recordkeeping, fiber study, additional labeling,
future UPC growth labeling, and reformulation costs. Annualized Amount = Amount / Annualizing Factor. 3
percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
D.
Need for Regulation
Information failure, a well-established type of market failure, can provide an economic
rationale for the mandatory disclosure of nutrition information. The government does not
necessarily have to intervene to address a market failure from a lack of information. However,
when individuals find collecting information costly, time-consuming, or both, the revealed
private demand for information may differ from the socially optimal level of information.
6
There is substantial uncertainty regarding the impacts of the two nutrition labeling rules. For a full discussion of
the uncertainty, please see the Welfare Estimates – Primary Sensitivity Analysis section of this Regulatory Impact
Analysis.
12
Nutrition Facts/Serving Size Combined Final RIA
Before NLEA, consumers lacked access to nutrition information for specific product categories.
After promulgating a series of regulations implementing NLEA, FDA began using mandatory
nutrition information disclosure as a regulatory tool to address information asymmetries
regarding the nutritional content of packaged foods. Given that consumers have limited time,
attention, and resources for seeking out new information, FDA, through our requirements to
provide Nutrition and Supplement Facts labels, requires that manufacturers convey relevant
nutrition information to consumers. This information, which consumers value (Ref. 3), can be
used by consumers to better inform choices at the point of purchase.
To the extent that there is information failure with respect to nutrition information, the
revisions to the Nutrition and Supplement Facts labels may provide relief. If the revised labels
provide fewer barriers to obtaining nutrition information, they may further assist consumers in
making healthier dietary choices by facilitating access to, and understanding of, the information
about the underlying nutrient amounts in prepackaged foods. This is because, as discussed in the
preambles of the Nutrition Facts Label and Serving Size rules, the revised labels will reflect
more recent science. Thus, the final rules may reduce information failure arising from lack of
consumer access to sufficient information, and understanding of that information, and therefore
may assist consumers in making healthy dietary choices.
Changes in labeling may also assist consumers by making the long-term health
consequences of consumer food choices more salient and by providing contextual cues of food
consumption. We note that the behavioral economics literature suggests that distortions internal
to consumers (or internalities) due to time-inconsistent preferences, myopia or present-biased
preferences, visceral factors (e.g., hunger), or lack of self-control can also create the potential for
13
Nutrition Facts/Serving Size Combined Final RIA
policy intervention to improve consumer welfare (Refs. 4 - 7).7 In a study that examines one of
the possible factors that drive obesity, Ruhm (2012) finds that standard economic models of
rational preferences and optimal consumption, which emphasize changes in the price of calorie
consumption and expenditure as the primary causes of obesity, have a limited ability to explain
the rapid and continuing increase in the prevalence of obesity (Ref. 4). The author suggests that
we can characterize decisions related to eating and body weight as an interaction between a
“deliberative system,” where individuals trade off the “utility from current food intake against
the associated monetary expense and disutility of future weight gains to achieve a constrained
optimum,” and an “affective system,” which “responds to cues and stimuli but does not consider
long-term effects of current actions” (Ref. 4).8 Akerlof (1991) proposes that when consumers
face repeated decisions with a short span of time in between each decision, e.g., choosing food
items or meals, and consumers give the present benefits of consumption undue salience relative
to their future costs, then small deviations from the utility maximizing (rational) level of
consumption can quickly accumulate into large mistakes (Ref. 8).9
Consistent with predictions based on models of bounded rationality, consumers can
systematically make suboptimal dietary choices because they discount future health
consequences relative to immediate benefits more than they would if they chose according to
7
An individual has time inconsistent preferences if his or her welfare-maximizing choice is different when
considered at different times. For example, an individual has time inconsistent preferences when making food
choices if he or she chooses to consume a relatively unhealthy but highly palatable item at lunchtime in spite of
having packed, in the morning, a healthier and less palatable alternative.
8
In the behavioral economics and psychology literatures the dual decision maker systems are also referred to as the
reflective and automatic, long-run and short-run, or cold and hot systems (or selves).
9
Several other behavioral economics models exist. These models can account for the seemingly irrational behaviors
of overeating and continually postponing efforts at weight loss by incorporating the effects of visceral factors,
present-biased preferences, heuristics, and other factors that influence decision making (Refs. 4, 6 - 9).
14
Nutrition Facts/Serving Size Combined Final RIA
their underlying or true preferences, leading them to regret their decisions at a later date.10 To
the extent that some form of intrapersonal market failure characterizes diet-related decisions,
changes in labeling may assist consumers by making the long-term health consequences of
consumer food choices more salient and by providing contextual cues of food consumption. For
example, the DCL format cues the consumer into the caloric consequences of consuming the
entire package or discrete unit, as applicable. Similarly, the DCL format and the added sugars
declaration may encourage individuals to consume less energy dense and nutrient poor foods by
making the nutritional attributes of these products more salient and reducing the undue salience
consumers place on the utility of consuming these products.
Consumer research supports the importance of salience and cues in immediate
consumption decisions. Some research has found evidence that increased portion sizes of food
products may have contributed to the obesity epidemic (Ref. 11). Because larger package and
portion sizes have been shown to be correlated with increased food consumption, the
“supersizing” trend may represent one of the drivers of the obesity epidemic. One study suggests
that package and portion size influence consumers’ intake of food products (Ref. 12). Some
argue that current labeling practices, which allow packages containing multiple servings per
container to only list the nutrient content per serving, confuse consumers who think that the
nutrition information on the product label refers to the entire package (Refs. 12 - 14). The DCL
provision of the final Serving Size rule, which requires that certain packages and discrete units of
food present the nutrition information per serving as well as based on the nutrition information
for the entire package or discrete unit, as applicable, may provide a contextual cue to consumers
10
Bounded rationality refers to the idea that there are limits (e.g., limited time, limited computing power, limited
information) to the ability of people to engage in rational decision making. Because of such limits, individuals
typically use heuristics or rules of thumb to simplify the decision-making process, but they often sacrifice judgment
accuracy for the reduction in cognitive effort in systematic ways (Ref. 10).
15
Nutrition Facts/Serving Size Combined Final RIA
that the package or discrete unit contains more than one serving and may aid consumers in
calculating the amount of calories and nutrients in a particular package. Increasing the font size
of and bolding the word “Calories” may also increase the salience of the caloric content of foods.
Thus, the DCL format and larger font may increase the usefulness of nutrition information on
packaged foods. Additionally, given that the Nutrition Facts Label final rule will require the
declaration of added sugars and that the 2015 Dietary Guidelines Advisory Committee
(“DGAC”) and 2015-2020 Dietary Guidelines for Americans (“DGA”) recommend limiting
calories from added sugars, we expect that some producers may reformulate their products to
reduce added sugars.
Overall, major predicted elements of the consumer and industry response to the nutrition
labeling final rules include:
Increased knowledge of the nutrient content of packaged foods, which may help
consumers make healthier food choices; and
Increased ease of nutrition label use from the availability of more easy-to-understand
information for products that bear DCLs.
These effects may help reduce the information failure and increase the salience of the
information on food packages, thereby assisting consumers in maintaining healthy dietary
practices.
Other major predicted elements of the consumer and industry response to the nutrition
labeling final rules include:
Greater disclosure of the nutrient content of existing packaged foods, which may give
firms an incentive to provide additional items with healthier formulations; and
16
Nutrition Facts/Serving Size Combined Final RIA
Potential reformulation of products to reduce added sugars or change amounts of added
vitamins and minerals based on current recommendations.
E.
Single Analysis for the Two Final Rules
We prepared this RIA for the two final rules together, since the two final rules involve
some form of a label change. Promulgating the requirements for the final label changes together
would allow manufacturers to undertake all of the revisions to the Nutrition or Supplement Facts
label under one label change. For example, for manufacturers that would be required to make
label changes under the Serving Size and Nutrition Facts Label rules, it would be more efficient
to make changes to their product’s serving size and to the other information on the Nutrition or
Supplement Facts label at the same time. If promulgated and made effective at different points
in time, these final rules could require manufacturers to make two label changes: one for the
changes required under the final Serving Size rule and one for the changes required under the
Nutrition Facts Label final rule. Two distinct label changes could result in greater cost.
F.
Coverage of the Rule and Industry Overview
The nutrition labeling final rules, together, cover all manufacturers of foods that are
required by FDA to provide nutrition labeling. This represents over 30,000 manufacturers that
manufacture close to 800,000 UPCs. We estimate that approximately 99 percent of the covered
firms are small businesses as defined by the Small Business Administration (SBA). SBA’s
definition of a small business, as it is applied to food manufacturers, includes those entities that
have less than 500 employees.
Table 2 - Industry Coverage Breakdown By Final Rules
Final Rule
Type of Label Change
Nutrition Facts Label
Conventional Food
Minor
Conventional Food
Major
Dietary Supplements
Minor
17
Covered UPCs
375,532
5,297
64,194
Nutrition Facts/Serving Size Combined Final RIA
Dietary Supplements
Total Nutrition Facts Label UPCs
Serving Size
Conventional Food
Conventional Food
Conventional Food
Total Serving Size UPCs
Major
191
445,214
Minor
Major
Extensive
255,852
76,363
2,817
335,032
780,246
Total UPCs
Table 2 presents the industry coverage breakdown by the two final rules and the type of
label changes or re-designs each entails. A minor label change is defined as a one-color/printing
plate change that does not require a label redesign (Ref. 1). Examples include one or more of the
following: changes to the net quantity statement; minimal changes to the Nutrition or
Supplement Facts panel; minimal changes to an ingredient list; the addition of a toll-free number;
and minimal changes to a claim, caution statement, or disclaimer on the back or side of a
package (affecting one color/plate) (Ref. 1). A major label change is defined as a multiple
color/printing plate change that requires a label redesign (Ref. 1). Examples include changes to
the name of the product; changes to the standard of identity or fanciful (trade) name for a food
product; the requirement for a product to provide a Nutrition or Supplement Facts panel when
none was previously required; substantial changes to an ingredient list; substantial changes to or
elimination of a claim; the addition of or substantial changes to a caution statement; and the
addition of or substantial changes to a disclaimer (Ref. 1). An extensive label change is defined
as a major format change that requires a change to the product packaging to accommodate
labeling information (Ref. 1). Examples include the addition of a peel-back label and increases
in the package surface area for labeling information.
18
Nutrition Facts/Serving Size Combined Final RIA
The Nutrition Facts Label rule will require significantly more label changes and will
cover more UPCs and manufacturers than will the Serving Size rule. A substantial majority of
the required labeling changes under the Nutrition Facts Label rule would be minor. However,
certain changes in nutrient DVs and the new definition of dietary fiber may require some
products to either reformulate to continue to make a related health or nutrient content claim or to
remove the health or nutrient content claim from their label altogether, the latter which entails a
major label change. While, as discussed in the Nutrition Facts Label rule and Serving Size rule,
we may reevaluate claim requirements in light of the changes we are making to be less stringent
than what is set forth in these rules, for the purposes of this analysis, we assume that no changes
to claim requirements are made. To the extent that requirements for certain nutrient content
claims and health claims may change in response to the changes made in the Nutrition Facts
Label and Serving Size rules before the compliance dates for these rules, this analysis could
overestimate the costs to manufacturers in complying with the Nutrition Facts Label and Serving
Size rules.
Under the Serving Size rule, certain products would have to undertake minor label
changes related to the updating, modifying, and establishing of RACCs for certain products, and
also related to the change in the definition of a single-serving container and the change in serving
size for a product in a multi-serving package in discrete units in which each discrete unit contains
at least 200 percent of the RACC. Additionally, for certain products (i.e., those that are
packaged and sold individually that contain at least 200 percent and up to and including 300
percent of the applicable RACC and those containing discrete units that contain at least 200
percent and up to and including 300 percent of the applicable RACC) the DCL provisions of the
Serving Size rule would require manufacturers to undertake a major label change. Also, certain
19
Nutrition Facts/Serving Size Combined Final RIA
changes in RACCs will require manufacturers to either reformulate some products to continue to
make a related health or nutrient content claim or to remove the health or nutrient content claim
from their label altogether, the latter which entails a major label change. Finally, the change in
the definition of a single-serving container to require products that have RACCs of 100 g or 100
mL or larger and that are packaged such that they contain more than 150 percent but less than
200 percent of the RACC to be defined as a single-serving container (versus offering such
products the option of being labeled as one or two servings) will require some manufacturers to
increase the package size of their products if they want to continue to make a health or nutrient
content claim, which entails an extensive labeling change.11,12
Under the combined effects of both rules, some products will be required to bear a
Nutrition Facts panel when such information was not previously required.13 The addition of a
Nutrition Facts panel when none was required previously entails a major labeling change.
11
Consider, as an example, UPC X. UPC X has a RACC of 100 g, contains 0.4 g of fat per RACC, and as packaged
contains 160 g of product (proportionally, UPC X contains 0.64 g of fat per 160 g of product). UPC X is currently
labeled as containing 2 servings. Under preexisting 21 CFR 101.62(b), UPC X would be permitted to make a fatfree claim on its label, as UPC X contain less than 0.5 g of fat per RACC and per labeled serving (UPC X contains
0.4 g of fat per RACC and 0.32 g of fat per labeled serving). Under the final rules, UPC X would be required to be
labeled as a single serving and so would contain 0.4 g of fat per RACC but 0.64 g of fat per labeled serving. Thus,
under the final rules, UPC X would no longer satisfy the requirements to make a fat-free claim. If UPC X changed
its package size to, for example, contain 450 g of product, then UPC X would be able to keep its fat free claim.
12
Even though the cost of reformulation is less than half the cost of increasing the package size, it is unlikely that in
the situation where the change to the serving size requirements affects the ability of a manufacturer to make a claim,
that manufacturers of affected products would choose reformulation in lieu of increasing the package size, as
reformulation would result in some package sizes having a different formulation, and therefore a potentially
different taste, than other package sizes of the same general product. It is also unlikely that in this instance
manufacturers would choose the relatively cheaper option of claim removal in lieu of increasing the package size, as
this would result in some package sizes carrying the claim and others not carrying the claim despite the fact that all
package sizes would contain the same formulation of the product.
13
Such products may include, for example, unsweetened instant coffee and tea, which may no longer contain
“insignificant” amounts of potassium due to the combined effects of the Nutrition Facts label and Serving Size final
rules, and bottled water, which may no longer contain “insignificant” amounts of sodium under the Nutrition Facts
final rule, the Serving Size final rule, or both (see 21 CFR 101.9(j)(4)). While, as noted in the Serving Size final
rule, we are declining to amend the definition of “insignificant” in 21 CFR 101.9(j)(4) at this time, we intend to
consider the applicability of an exemption from nutrition labeling requirements in a future rulemaking with respect
to certain products and, until such time as we have had the opportunity to consider such matters further, intend to
consider the exercise of enforcement discretion with respect to mandatory nutrition labeling on products that would
have been exempt from nutrition labeling requirements under § 101.9(j)(4) prior to the effective date of the two
20
Nutrition Facts/Serving Size Combined Final RIA
It is important to note that there is significant overlap in terms of which food
manufacturers will be required to make labeling changes or reformulate their products as a result
of both final rules. For example, many food manufacturers that will have to update some of the
nutrition information for their products will also have to update the serving size information for
their products. Additionally, all food manufacturers that will be affected by changes that result
from implementation of the Serving Size final rule will also be required to change their labeling
as a result of the Nutrition Facts label final rule. This overlap, as previously mentioned in Section
D of this document, allows firms to avoid a significant portion of the costs that would otherwise
be incurred, by undertaking all of the required label changes via one label change.
G.
Comments on the Preliminary Regulatory Impact Analysis (PRIA) and the Supplemental
PRIA and Our Responses
FDA’s proposed rules “Food Labeling: Revision of the Nutrition and Supplement Facts
Labels” (Docket No. FDA-2012-N-1210) and “Food Labeling: Serving Sizes of Foods That Can
Reasonably Be Consumed at One-Eating Occasion; Dual-Column Labeling; Updating,
Modifying, and Establishing Certain Reference Amounts Customarily Consumed; Serving Size
for Breath Mints; and Technical Amendments” (Docket No. FDA-2004-N-0258) were published
on March 3, 2014, and their comment period ended on August 1, 2014. In addition, we
published a supplemental notice to the proposed rule “Food Labeling: Revision of the Nutrition
and Supplement Facts Labels” titled “Food Labeling: Revision of the Nutrition and Supplement
Facts Labels; Supplemental Proposed Rule To Solicit Comment on Limited Additional
Provisions” on July 27, 2015 (Docket No. FDA-2012-N-1210) and its comment period ended on
rules. However, to avoid underestimating the potential costs to industry that may result from the combined effects
of the two rules, our analysis in this document considers the costs to industry assuming manufacturers of products
that were previously exempt from nutrition labeling requirement were to make labeling changes to include nutrition
labeling.
21
Nutrition Facts/Serving Size Combined Final RIA
October 23, 2015. We prepared PRIAs in connection with the proposed rules and supplemental
notice. In this section, we summarize and respond to the comments we received on those PRIAs.
We have numbered each comment to help distinguish between different comments. The number
assigned to each comment is purely for organizational purposes and does not signify the
comment’s value, importance, or the order in which it was received.
(Comment 1) Some comments stated that manufacturers will need to analyze products related to
the mandatory declarations of potassium, vitamin D, and added sugar.
(Response 1) In this final RIA, analytical testing costs are included in our labeling cost
estimates. Currently, there are no analytical methods that can distinguish between added
and naturally occurring sugars. In such cases, manufacturers must make and keep certain
written records to verify their declaration of added sugars on the Nutrition or Supplement
Facts label. We accounted for these kinds of recordkeeping costs in the PRIA and
continue to account for them in this final RIA.
(Comment 2) Some comments expressed concern that under the proposed rules manufacturers
will no longer be allowed to use the word “folate” on the Supplement Facts label, but
instead will be required to use the term “folic acid.” The comments stated that as a result
companies currently using dietary folate in their products would need to reformulate to
remove the folate and replace it with folic acid.
(Response 2) Under the final rules, both “folate” and “folic acid” will be required to appear on
the Supplement Facts label.
(Comment 3) Some comments stated that certain changes to RACCs and DVs called for by the
proposed rules will require manufacturers to reformulate to maintain health and nutrient
content claims tied to these RACCs and DVs.
22
Nutrition Facts/Serving Size Combined Final RIA
(Response 3) We accounted for these types of reformulation costs in the PRIA (see p. 29) and
continue to account for them in this final RIA.
(Comment 4) With regards to the dietary fiber citizen petition process, some comments stated
that if the proposed process is kept, it is unlikely that FDA will have reviewed all fibers
and notified manufacturers in time for compliance with the 2 year time limit, thus
requiring manufacturers of products with extrinsic fibers to change the label twice, once
to comply with the proposed rules and a second time once the fiber is approved.
(Response 4) In this final RIA, we estimate the cost of an additional label change for the
approximately 20 percent of UPCs which we estimate might have to change their label
twice related to the citizen petition review process.
(Comment 5) Some comments stated that the FDA’s proposed changes to labeling singleserving containers will make some foods ineligible to make health or nutrient content
claims. Comments noted that there will be instances in which products in large packages
(i.e., those greater than 200 percent of the RACC) will be able to make a health or
nutrient content claim while the smaller package size version (between 150 and 200
percent of the RACC) of the same product will not be able to make such a claim because
the latter is considered a single serving, thus requiring manufacturers to either
reformulate or repackage their product to continue to make the claim, relabel to remove
the claim, or remove the product from the market.
(Response 5) In this final RIA, we account for this possibility very conservatively14 by
estimating the cost of an extensive label change (repackaging) for all UPCs which both
(i) contain a health or nutrient content claim and (ii) are packaged such that they contain
14
“Conservative” cost estimates are cost estimates which, because of data and other limitations, are likely
overestimated.
23
Nutrition Facts/Serving Size Combined Final RIA
more than 150 percent and less than 200 of the RACC. However, we do not believe that
the changes to the single-serving container requirements will affect a lot of UPCs that
make nutrient content claims or health claims, because some UPCs containing more than
150 percent and less than 200 percent of the RACC are already labeled as one serving,
and further, eligibility to make a nutrient content claim or health claim is generally
dependent on the RACC for a product (not the serving size).
(Comment 6) Some comments questioned our choice of a 2-year compliance period given that
the PRIA showed that net benefits are maximized under a 4-year compliance period.
(Response 6) Under the final rules, the compliance period is 36 months for small businesses and
24 months for large businesses. (As noted previously, for purposes of this analysis, we
consider a small business to be a business with annual food sales of less than $10 million,
and a large business to be a business with annual food sales of $10 million or more.) We
believe that this provides industry time to revise labeling to come into compliance with
the new labeling requirements while balancing the need for consumers to have the
information in a timely manner.
(Comment 7) Some comments expressed concern that the proposed rules might increase the
amount of space presently occupied by the Nutrition Facts Panel.
(Response 7) The Nutrition Facts label will occupy more space on a product’s package in the
case of DCL and in cases when a Nutrition Facts label was not previously required (see
footnote 8). The costs associated with DCL were captured in the PRIA, and we continue
to capture them in the final RIA. The costs associated with providing a Nutrition Facts
label when one was not previously required are captured in this final RIA.
24
Nutrition Facts/Serving Size Combined Final RIA
(Comment 8) Some comments expressed concern that the footnote will not be finalized until
after the publication of the final rule, which could require an additional labeling change.
(Response 8) The footnote language is included in the final rule. Therefore, manufacturers
would be able to include the footnote language in the same label change as any other of
the required changes under the two rules.
(Comment 9) Some comments expressed concern that with respect to the use of nutrient content
claims and health claims on products with DCL, the FDA is proposing to require a
statement accompanying the claim clarifying that the level of the nutrient that is the
subject of the claim is based on the RACC. The commenters expressed concern that this
could result in additional labeling costs.
(Response 9) The labeling costs associated with this (major label change) are subsumed under
the labeling costs associated with adding a DCL (major label change), which were
captured in the PRIA and continue to be captured in the final RIA. Other concerns
relating to the requirement to add a statement accompanying such claims are addressed in
the Serving Size final rule.
(Comment 10) Some comments have requested that FDA articulate in the final rules how FDA
intends to monitor and measure any public health impacts of the final rules for the
purpose of eventually conducting a retrospective analysis.
(Response 10) Although no new data collection required by this rulemaking would necessarily
help to facilitate a retrospective review, FDA does expect to monitor and review the
effects of these final rules, as emphasized in Executive Order 13563. This will most
likely occur through analysis of public data sources, such as NHANES.
25
Nutrition Facts/Serving Size Combined Final RIA
(Comment 11) One comment stated that in the “Need for Regulation” section of the PRIA, we
described a market failure that might exist but did not demonstrate that the market failure
actually does exist.
(Response 11) We have identified the problems that the rule is intended to address and assessed
the significance of those problems consistent with Section 1(b)(1) of Executive Order
12866, which requires us to identify the problem that the rule intends to address as well
as assess the significance of that problem (Executive Orders 13258 and 13422, which
previously amended Executive Order 12866, were revoked by Executive Order 13497 on
January 30, 2009). In the “Need for Regulation” section of the PRIA, we identified the
problems of information failure and intrapersonal market failure that the proposed rules
would intend to address, if finalized, and we continue with this approach in the final RIA.
The PRIA further explained that the information failure results from a lack of consumer
access to certain nutrition information, stating that “[g]iven that consumers have limited
time, attention, and resources for seeking out new information, the Nutrition Facts label
attempts to convey relevant nutrition information to better inform choices at the point of
purchase.” We assessed the significance of intrapersonal market failure utilizing citations
from the behavioral economics literature and citations of consumer research that support
the importance of salience and cues in consumption decisions. We continue to assess the
significance of information failure and intrapersonal market failure in these ways in the
final RIA.
(Comment 12) Some comments stated that the increase in the Vitamin D DV combined with
current Vitamin D fortification limits could jeopardize the ability of some manufacturers
26
Nutrition Facts/Serving Size Combined Final RIA
to make related nutrient content and health claims, forcing them to remove such claims
from their labels.
(Response 12) We estimated in the PRIA, and continue to estimate in the final RIA, the number
of UPCs which will either relabel to remove a health or nutrient content claim or
reformulate to maintain a health or nutrient content claim as a result of the various DV
changes described in the proposed rules. We estimated in the PRIA, and continue to
utilize this estimate in the final RIA, a 50 percent/50 percent relabel/reformulate split
before assigning costs to these UPCs. We asked for comment on these estimates. The
comments we received are discussed later in this document.
(Comment 13) Some comments expressed concern that the increase in the coffee RACC size
could bring Potassium levels above the two percent threshold, positing the first and only
criterion on which the industry would need to change packaging to add a Nutrition Facts
Panel.
(Response 13) In the final RIA, we estimate the cost of adding a Nutrition Facts label from
scratch for products, including certain coffee, tea, and bottled water products, that would
be required to add a Nutrition Facts label based on the final rules.15 As noted in the
Serving Size final rule, however, we intend to consider the applicability of an exemption
from nutrition labeling requirements in a future rulemaking with respect to certain
products, and until such time as we have had the opportunity to consider such matters
further, we intend to consider the exercise of enforcement discretion with respect to
mandatory nutrition labeling products that would have been exempt from certain nutrition
labeling under § 101.9(j)(4) prior to the effective date of the final rules. This final RIA,
15
See footnote 13.
27
Nutrition Facts/Serving Size Combined Final RIA
however, nonetheless considers the costs to industry that would result from adding a
Nutrition Facts label to certain coffee, tea, and bottled water products that previously
were not required to provide a Nutrition Facts label.
(Comment 14) A comment questioned our estimate regarding the percentage of affected UPCs
which will reformulate to maintain health and nutrient content claims related to various
changes described in the proposed rules (50 percent), as well as our estimate regarding
the percentage of affected UPCs which will reformulate to reduce added sugar content in
response to the added sugar declaration described in the proposed rules (5 to 6 percent),
stating that we provide no basis for these estimates.
(Response 14) While some affected products will be relabeled and some will undergo
reformulation, we have identified no data with which we are able to project the percent of
products that will be relabeled and the percent that will undergo reformulation. In the
absence of data, the PRIA estimated that half of the affected products will be relabeled
and the other half will be reformulated, and we asked for comment on these estimates in
lieu of reliable data about how manufacturers might respond to the proposed rules.
Neither this comment nor any other information we received in response to the PRIA
provided further basis upon which to adjust these assumptions, and so we retain our
initial assumptions in this final RIA.
(Comment 15) A comment stated that each of the changes to the Nutrition Facts label on its own
represents a minor label change, but taken together such changes represent a major label
change.
(Response 15) When the changes to the Nutrition and Supplement Facts labels described in the
proposed rules are such that only one new printing plate must be cut for the black printing
28
Nutrition Facts/Serving Size Combined Final RIA
of the label, such a change is defined by the FDA Labeling Cost Model as a minor label
change. Thus, multiple revisions required by the two rules that require only one new
printing plate to be cut for the black printing of the label are considered a minor label
change.
(Comment 16) A comment stated that under the Serving Size rule, we incorrectly only included
conventional foods in our cost analysis.
(Response 16) None of the provisions of the Serving Size rule affect dietary supplements. The
changes to DCL requirements appear in 21 CFR 101.9, and under 21 CFR 101.9(j)(6),
dietary supplements are exempt from the requirements of 21 CFR 101.9 and must be
labeled in compliance with 21 CFR 101.36. Additionally, the final amendments to 21
CFR 101.12 did not alter the requirements for dietary supplements.
(Comment 17) A comment stated that the new breath mint serving size requirement was omitted
from our cost analysis.
(Response 17) For purposes of this analysis, the labeling costs associated with the new breath
mint serving size requirement were captured in Table 4 of the PRIA under “Conventional
Food,” “Changes Due to RACC Proposals” and are captured similarly in the final RIA.
(Comment 18) A comment stated that in the PRIA the estimate of UPC counts is dated and
insensitive to changes in future UPC counts. The comment further stated that this is
inconsistent with how population is treated in the benefits section of the PRIA, saying
that while it may be true that the Labeling Cost Model is based on 2008 data, it should be
possible to scale costs based on actual and projected growth of UPC labels since 2008.
(Response 18) We estimated labeling costs using 2008 data in the proposed rules and 2012 data
in the final rules (the Labeling Cost Model has been updated so that the conventional
29
Nutrition Facts/Serving Size Combined Final RIA
food and dietary supplement UPC counts are now based on 2012 data and labeling costs
are estimated in the final RIA using this updated version of the Labeling Cost Model).
We acknowledge that the labeling costs of the final rules do depend in part on the
number of new UPCs in the future which will need to change their labels because of the
final rules. According to the FDA Labeling Cost Model, the addition of a Nutrition or
Supplement Facts panel to a previously unlabeled UPC, such as to a new UPC, represents
a major label change. New UPCs under the final rules would have had to add a label
under NLEA, which is to say that new UPCs under the final rules would have had to
make a major label change under NLEA. Thus, the labeling costs for future UPCs
subject to minor and major label changes under the final rules would not generally be
considered a cost of the final rules. However, future UPC growth is relevant for UPCs
that would need to bear a Nutrition Facts Panel for which none was previously required
and UPCs which would require an extensive label change. We estimate that a fixed
fraction of future UPCs fall into these categories, and we estimate associated labeling
costs for UPCs affected by these requirements over the next 20 years.
(Comment 19) A comment stated that it is likely that the single-serving and DCL provisions of
the proposed rules will create an incentive for food firms with UPCs in these categories
to create new package sizes. The comment stated that, for example, if a package of
cookies has 350 percent of the RACC, the firm may want to increase the package size to
over 400 percent to avoid the DCL requirement. The comment concluded by stating that
we ignored such costs.
(Response 19) According to the FDA Labeling Cost Model, the costs of increasing package size
are roughly forty times the cost of adding a DCL. Thus, we estimate that most
30
Nutrition Facts/Serving Size Combined Final RIA
manufacturers would choose to add a DCL to their existing package in lieu of increasing
the size of their existing package. We captured DCL costs in the PRIA and continue to
capture them in the final RIA. We also note that the example of the package size
provided in the comment would no longer be required to provide a DCL on a per-package
basis, as the Serving Size final rule will require a second column of nutrition information
on packages containing at least 200 percent and up to and including 300 percent of the
RACC (see § 101.9(b)(12)), subject to certain exceptions.
(Comment 20) A comment stated that an important cost of each regulation is the cost to
government from enforcing that regulation. The comment concluded by saying that we
did not account for such costs in the PRIA.
(Response 20) According to the most recent version (9/17/2003) of the Office of Management
and Budget’s (OMB’s) Circular A-4, these costs, like all categories of impacts, should be
included in the analysis only when they are significant. Based on internal data, we
estimate the current cost of labeling enforcement efforts to be roughly $1.3 million per
year (2014$), and we do not expect a significant increase in such efforts as a result of the
final rules. Thus, in the case of the proposed rules, as well as the final rules, we do not
consider these costs to be significant. Therefore, they are not captured in the PRIA or in
this final RIA.
(Comment 21) A comment stated that throughout the cost section of the PRIA, we do not
explain the criteria used to determine the categorization of UPC data, saying that this lack
of transparency makes it difficult to assess the validity of the cost estimates. The
comment requested more detail.
31
Nutrition Facts/Serving Size Combined Final RIA
(Response 21) This comment refers in part to our description of the Labeling Cost Model in the
PRIA. Greater detail regarding the Labeling Cost Model is provided in the Labeling Cost
Model report, which we cited in the PRIA and made available on regulations.gov, and
continue to cite in the final RIA and make available on regulations.gov. We provide
additional detail in this final RIA related to how we obtained our estimates of the number
of UPCs affected by the various provisions of the final rules (e.g., source of data, richer
descriptions of steps in the estimation process).
(Comment 22) Some comments expressed concern about the fact that the Abaluck paper cited in
the benefits section of the PRIA is unpublished and, therefore, has not undergone a
formal peer review process.
(Response 22) The comment period serves as a form of review of the Abaluck paper by
providing the public with time to provide comments on it. In this instance, the comment
period provided the public with several months to comment on the Abaluck paper that
forms the basis of our analysis of the benefits of the proposed rules. We provide
responses to the comments we received on the Abaluck paper later in this section.
(Comment 23) One commenter stated that by relying on Abaluck to estimate the benefits of the
proposed rules, we are assuming that the proposed rules and NLEA have similar impacts.
(Response 23) We disagree. Our starting point for estimating the benefits of the proposed rules
was the benefits associated with the original Nutrition and Supplement Facts label that
was mandated by NLEA that were estimated by Abaluck. We then acknowledged that
the proposed rules and NLEA do not have similar impacts, stating on p. 48 of the PRIA
that “. . . the proposed changes to the Nutrition Facts label would increase the available
nutrition information on food labels, but by an amount much smaller than the changes
32
Nutrition Facts/Serving Size Combined Final RIA
brought about by NLEA” before proceeding to adjust Abaluck’s benefits estimates
downward to reflect the effect of the proposed rules by the factors outlined on pp. 48 – 56
of the PRIA. We retain this approach in the illustrative calculations that appear in the
final RIA.
(Comment 24) Some comments expressed concern that we did not individually estimate the
costs and benefits of each of the changes described in the proposed rules.
(Response 24) The way in which we estimated the benefits of the proposed rules is such that we
were unable to disentangle the marginal benefit of each of the changes described in the
proposed rules. This continues to be true for our illustration of the benefits of the final
rules. We were unable to measure the marginal cost of each individual change described
in the proposed rules because the various costs associated with the proposed rules are too
interdependent to allow us to do so, and this continues to be true for our estimate of the
costs of the final rules. For example, a UPC subject to DCL would need to undergo a
major label change assuming the product had not previously provided a DCL on a
voluntary basis, the cost of which would also cover the various changes to the Nutrition
and Supplement Facts Panels described in the proposed rules. As a second example, a
UPC would need to undergo a minor label change whether the UPC makes one or all of
the various changes to the Nutrition and Supplement Facts Panels described in the
proposed rules.
(Comment 25) Some comments expressed disagreement with the s1 parameter of our benefits
model, saying that s1 appears to be based entirely on quantity of change, not quality.
(Response 25) We agree that our s1 measure is based entirely on quantity, and not quality, of
change, the implication of which is that each label change receives equal weight. We
33
Nutrition Facts/Serving Size Combined Final RIA
took such an approach in the PRIA because, due to data and other limitations, we were
unable to estimate quality weights associated with each label change. In light of such
limitations, we believed that this approach represented a reasonable approximation. We
retain this approach in the illustrative calculations that appear in the final RIA but note
that uncertainty regarding s1 is primarily why we maintain significant uncertainty ranges
around all calculations.
(Comment 26) One comment expressed concern that Abaluck’s sample is limited to women
aged 19 to 50 who are the primary meal preparers in their households, stating that we
address this issue in the PRIA by calculating separate benefits estimates for women only
but that these estimates are not then used in the discussion of net benefits.
(Response 26) To estimate benefits, and thus net benefits, we extrapolated to the entire U.S.
population because the proposed rules are expected to affect the entire U.S. population,
not just women aged 19 to 50 who are the main meal planners in each household. We
continue with this approach in the final RIA. However, we acknowledged in the PRIA,
and continue to acknowledge in the final RIA, that Abaluck generates his WTP estimates
from a sample of women 19 to 50 years of age who are the main meal planners in each
household. Because we do not know for certain whether such a sample generalizes to the
population as a whole, as a sensitivity analysis we estimated in the PRIA, and continue to
illustrate in the final RIA, the benefits of the proposed rules based just on a female
subpopulation.
(Comment 27) One comment stated that regarding the annual welfare gain per person values
taken from the Abaluck paper for the ΔW parameter of the benefits model, we wisely use
the structural equation estimates that Abaluck notes are a better fit but, without comment,
34
Nutrition Facts/Serving Size Combined Final RIA
only uses two of the structural models in its analysis, omitting two other models with
lower values ($32.10 and $28.30).
(Response 27) We relied just on Abaluck’s Model 1 and Model 2 estimates in the PRIA, and
continue to do so in the final RIA, due to the uncertainty surrounding the proper
specification of Model 3 and Model 4. Model 3 and Model 4 disaggregate total fats into
unsaturated and saturated fats, but according to Abaluck, because only total fats were
listed on nutrition labels prior to NLEA, it is not clear what information, if any, we
should assume label users possessed about the content of saturated or unsaturated fats in
labeled foods prior to NLEA, and such prior information is a component of Abaluck’s
WTP estimates.
(Comment 28) One comment stated that regarding the PRIA benefit model’s USE parameter,
we adjust for changes in use over time by taking the ratio of post-rule label usage (77
percent in the 2009-2010 NHANES plus expected increase due to the rule) to usage soon
after NLEA (65 percent, unreferenced in the PRIA). The comment stated that these two
data sources may not be compatible, given that NHANES changed the mode of asking
this question after 2006, saying that in NHANES 2005-2006 the question was asked in
person and the respondent was shown a Nutrition Facts Panel, but starting with the 20072008 collection, the question was asked over the phone without a visual. The result, the
comment said, was an increase from 53 percent to 72 percent usage in 2 years, which the
comment believes is likely due to question format rather than to actual usage change.
The comment concluded by stating that if this is being incorporated into the analysis,
then we are likely overestimating changes in usage.
35
Nutrition Facts/Serving Size Combined Final RIA
(Response 28) Label use soon after NLEA is not relevant to our benefits estimates. More
specifically, we estimated the benefits of the proposed rules, and illustrate the benefits of
the final rules, using the formula BtLabel = ΔWLabel x s1 x USE x (1 – USDA) x POPt. Note
that s1 = USERight After NLEA Publication*(%UPCsSingle Column Label*%Single Column Label
Content Changed + %UPCsDual Column Label*%Dual Column Label Content Changed) and
USE = USECurrent / USERight After NLEA Publication. USERight After NLEA Publication mathematically
cancels. Thus, only current use is relevant to our estimate of the benefits of the proposed
and final rules. In this final RIA, we continue to estimate current label use using 20092010 NHANES data, as these continue to be the most recent data from which this
information is available. Furthermore, in this final RIA, we have updated our discussion
of benefits so as to remove any mention of USERight After NLEA Publication since it is not
relevant to our benefits illustrations.
(Comment 29) One comment stated that our statement in the PRIA that “Antonuk and Block
(2006) found a 13.9 percent increase in attention to the label when switching from singlecolumn labeling (SCL) to DCL format” is not true, saying that this appears to be the case
for non-dieters only and that the effect is much smaller for dieters, but that the difference
in both cases is not likely to be significant (though it is not tested in the paper).
(Response 29) In the PRIA, we adjusted recent label use using this amount (13.9 percent) to
estimate label use once the proposed rules take effect. However, in light of our
confirmation that the differences in attention to the label when switching from the SCL to
the DCL format are statistically significant neither for non-dieters nor for dieters, we do
not make this adjustment in the final RIA.
36
Nutrition Facts/Serving Size Combined Final RIA
(Comment 30) A comment stated that Abaluck’s re-evaluated approach completely ignores the
hedonic elements of food consumption and assumes that consumers only purchase food
for its healthfulness. The comment said that it is possible that the entire discrepancy
between Abaluck’s WTP and re-evaluated estimates is due to a rational tradeoff between
taste, cost, and healthfulness, not to uninternalized nutrition benefits. The comment
states that it is also possible that some of the discrepancy is due to uninternalized
nutrition benefits but that we give no direct evidence to support this claim. The comment
concluded by saying that given that the scope of unintentional nutrition benefits is
entirely speculative, use of the Abaluck quantitative estimate, which the comment says
assumes that 100 percent of the discrepancy is due to unintentional nutrition benefits, is
not justifiable.
(Response 30) In this final RIA, we no longer rely on Abaluck’s re-evaluated estimates, which
were used to estimate other sources of benefits in the PRIA, because such estimates
reflect what consumers should be willing to pay for NLEA (based on medical evidence
on the effect of diet on health) versus what they actually are willing to pay for NLEA, the
latter which we used as the basis of our primary benefits estimates in the PRIA and which
we continue to use as the basis of our benefits illustrations in the final RIA.
(Comment 31) A comment stated that the results of the Abaluck study are highly dependent on
assumptions about price responsiveness. The comment stated that changes in
consumption figures are converted to dollar values using price estimates from the
Continuing Survey of Food Intakes by Individuals (CSFII) and price elasticity
(responsiveness) estimates from other studies, but that it is impossible to assess whether
37
Nutrition Facts/Serving Size Combined Final RIA
the responsiveness values used are credible or sufficiently tailored to product categories,
however, because they are not cited.
(Response 31) The results of the Abaluck study are dependent on price responsiveness
estimates. The comment incorrectly asserted that it is impossible to assess whether the
responsiveness values used are credible or sufficiently tailored to product categories
because they are not cited. The Abaluck study cites on p. 19 a 2010 American Journal of
Public Health paper titled “The Impact of Food Prices on Consumption: A Systematic
Review of Research on the Price Elasticity of Demand for Food” by Tatiana Andreyeva,
Michael Long, and Kelly Brownell as his source for price elasticities for each product
group.
(Comment 32) A comment stated that one stage in the calculation of welfare gains from NLEA
involves scaling to account for prior knowledge and that Abaluck estimates that, prelabeling, consumers already had knowledge of 20 percent of the information included on
NLEA labels. The comment says that this estimate is cited to a study on Starbucks
customers, but that no such estimates are given in the cited paper and that Abaluck gives
no explanation for how the information from the cited paper is used.
(Response 32) The comment misunderstands the study on Starbucks customers. In this study,
the authors conducted a survey in which Starbucks customers were asked prior to the
adoption of menu calorie labeling how many calories they believed were in various
Starbucks food items. Regressing these estimates of calorie content per gram on actual
calorie content per gram produced a coefficient of 0.2, or 20 percent, suggesting that
consumers estimate the calorie content of food items to be about 20 percent of the actual
calorie content. Abaluck describes how he uses this estimate on p. 21 of his paper.
38
Nutrition Facts/Serving Size Combined Final RIA
(Comment 33) A comment stated that Abaluck artificially constrains the mathematical modeling
used in the study by assuming that only benefits, not adverse outcomes, can result from
providing additional nutritional information. The comment says that Abaluck’s model
does not allow for an increase in the consumption of other food by more than the
decrease in the unhealthy foods avoided, even though this behavior has been observed in
many studies. The comment cites Qing Yang, “Gain Weight By Going Diet? Artificial
Sweeteners and the Neurobiology of Sugar Cravings,” Yale J. Biol. Med., June 2010,
83(2), 101-108 and also states that the phenomenon that decreases in one area are
eclipsed by more consumption in another is well known in behavioral economics as “the
Jevons effect.”
(Response 33) Abaluck’s satiation constraint states that individuals, on average, consume a
constant weight of food in each day. That is to say, any decline in the grams consumed in
some product groups must be offset by an increase in the grams consumed in other
product groups, healthy or otherwise. Therefore, we disagree that Abaluck’s model does
not allow for an increase in consumption of calories from other food by more than the
decrease in unhealthy foods avoided. Based on the satiation constraint, Abaluck could
have found that labeling led individuals to substitute towards foods with more calories
within product groups, which would have been consistent with the phenomenon
described in the Yang paper cited by the comment, and would have made them worse off.
Instead, Abaluck found the opposite.
(Comment 34) A comment stated that the Abaluck paper assumes that consumers will
automatically improve their dietary choices in response to additional information, even
though the only two empirical studies cited in his paper (Elbel, Kersh, Brescoll, and
39
Nutrition Facts/Serving Size Combined Final RIA
Dixon (2009) and Bollinger, Leslie, and Sorensen (2010)) show just the opposite, that
nutritional labeling in restaurants was not effective in altering dietary choices.
(Response 34) We disagree with the comment’s characterization of the findings of the study
conducted by Bollinger, Leslie, and Sorensen (2010), which finds a modest but
statistically significant impact of calorie labeling on food consumption (6 percent
reduction in calorie consumption). Abaluck also cites Mathios (2000), who finds that
post NLEA there was a significant decline in sales for the highest fat salad dressings, as
well as Variyam and Cawley (2006) and Variyam (2008), who examine the impact of
NLEA and label use on nutrient consumption and BMI and find decreases in obesity rates
among non-Hispanic white women. Only Elbel et al. (2009), who look at menu labeling
in fast food restaurants in New York, do not find an impact.
(Comment 35) A comment stated that there is no empirical evidence in Abaluck that any
improvement in dietary choices actually occurred but rather that it is only a computation
of what the benefits in monetary terms would be if people had actually behaved as
Abaluck’s mathematical model assumes that they might in response to certain
information. The comment stated that Abaluck explicitly concedes that his is a
theoretical computation of what the monetary benefits would be if people made healthier
dietary choices, not an empirical demonstration that any particular labeling regime will in
fact result in healthier dietary choices, via the statement made in the conclusion of his
paper that “[t]he value reported here for the potential welfare gains from consuming
healthier foods is by no means definitive . . ..”
40
Nutrition Facts/Serving Size Combined Final RIA
(Response 35) This comment references Abaluck’s re-evaluated estimates. To illustrate the
benefits of the final rules in the final RIA, we rely on Abaluck’s WTP estimates, not his
re-evaluated estimates.
(Comment 36) A comment stated that to the extent that any dietary benefits from prior labeling
requirements actually occurred, Abaluck attributes most of them to the “calories from
fat” labeling which, according to the comment, FDA is now proposing to ban. The
comment asserted that the flaw in Abaluck’s methods are demonstrated by the fact that
Abaluck attributes most of the alleged benefits that he calculates to a requirement that the
comment says that many, including FDA, believe to have backfired.
(Response 36) To estimate the benefits of the proposed rules in the PRIA, we relied on the
welfare estimates presented in Table 7 of the Abaluck paper, and continue to do so to
illustrate the benefits of the final rules in this final RIA. We note that the labeling
benefits attributable to the labeling declarations related to fat are limited to the “total fat”
declaration and not the “calories from fat” declaration. While we acknowledge that the
total fat declaration contributes to the dietary benefits from prior labeling requirements, it
is not the only contributor. Underlying these welfare estimates are the estimates
presented in Table 5 of the Abaluck paper of peoples’ WTP for NLEA nutrient
information not just for total fat but for many other nutrients as well. We disagree that
Abaluck’s methods are flawed in this regard.
(Comment 37) Some comments expressed concern that the projected public health benefits are
overestimated due to consumers’ misunderstanding of the proposed label revisions.
Specifically, many comments asserted that many consumers will be confused regarding
41
Nutrition Facts/Serving Size Combined Final RIA
the focus on “added sugars” and that, therefore, the desired public health outcomes will
not be attained.
(Response 37) Clarifying language regarding added sugars has been added to the Nutrition Facts
label in the final rule. FDA believes that this clarifying language will reduce the
confusion found among some consumers in consumer research. In addition, educational
initiatives with partners such as other federal government agencies within the Department
of Health and Human Services and United States Department of Agriculture (USDA),
state health departments, health professional organizations, food manufacturers, retailers,
and non-profit organizations, could increase consumers’ knowledge and effective use of
the new food label.
(Comment 38) Some comments raised concerns that projected reformulations would either not
materialize or not lead to desired improvements in nutritional quality because (1)
consumers will avoid more nutritious products with relatively more added sugars or (2)
products may be reformulated to include less added sugars but be less nutritious overall.
(Response 38) It is possible that as a result of the final rules some consumers may avoid more
nutritious products with relatively more added sugars in favor of one of the following:
less nutritious products with more added sugars, less nutritious products with less added
sugars, or more nutritious products with less added sugars. However, due to data
limitations, we are unable to quantify the cumulative change in benefits associated with
these effects.
As noted in both the Nutrition Facts Label rule and Serving Size rule, the
purpose of the amended requirements in these rules is to promote the public health and
provide consumers access to factual information that will help them understand the
42
Nutrition Facts/Serving Size Combined Final RIA
nutrient content of food products; while any product reformulation that might result in
improvements in the nutritional profile of foods would be a beneficial result, it is not the
purpose of the regulations, as these comments suggest. We have nonetheless considered
potential improvements to the public health as a potential benefit of the rules. To the
extent this comment suggests that we overestimated benefits that would result from the
potential reformulation of products, however, we disagree. In the PRIA, we estimated
reformulation costs, but due to data limitations did not estimate benefits associated with
product reformulation. We continue with this approach in this final RIA. The issues
raised in this comment are therefore irrelevant to our calculations.
(Comment 39) Some comments noted that specific consumer products that include significant
amounts of added sugars for palatability would be unduly affected by the focus on added
sugars.
(Response 39) We did not consider the potential economic impact to such manufacturers
differently from our evaluation of other manufacturers. The declaration of added sugars
and the percent DV on the label allow consumers to understand the sources of added
sugars in their diet and make healthy dietary choices, which can include some amounts of
added sugars. Education about the consumption of limited amounts of added sugars in
the diet, including to improve the palatability of certain foods, are also planned. Industry
may also reformulate products to include less added sugars (for example, by using other
sweeteners or using alternate blends). Also, any specific losses, should they occur, would
be offset in the aggregate by any gains incurred in turn by producers of substitute
products.
43
Nutrition Facts/Serving Size Combined Final RIA
(Comment 40) A comment suggests that the 2015 Supplemental PRIA does not include reliable
data on costs that small businesses will face in implementing the proposed regulations,
especially those related to reformulation.
(Response 40) We disagree. The Regulatory Flexibility Analysis associated with the 2015
supplemental PRIA used the same factors evaluated in the Regulatory Flexibility
Analysis contained in the PRIA, which includes an estimate of the cost to all businesses,
including small businesses, that would result from the need to relabel, reformulate, and
maintain records as a result of these final rules. The comment provided no basis upon
which to adjust these cost estimates. Furthermore, we have no reason to believe that the
data we used to produce these estimates are not reliable. The Regulatory Flexibility
Analysis contained in this final RIA likewise includes an estimate of the cost of the rules
to all businesses, including small businesses, that is comprised of updated relabeling,
reformulation, and recordkeeping costs, as well as study costs related to the new
definition of dietary fiber.
(Comment 41) A comment notes that the 2015 Supplement does not include an analysis of the
costs involved with the professional skills necessary for record-keeping and reports
related to the added sugars disclosure.
(Response 41) The costs associated with making and keeping records to verify the amount of
added sugars declared on the Nutrition Facts label were captured in the PRIA (beginning
on p. 27) and, thus, were also captured in the 2015 supplemental PRIA. We continue to
capture such costs in this final RIA.
44
Nutrition Facts/Serving Size Combined Final RIA
(Comment 42) A comment references the Regulatory Flexibility Act in expressing concern that
the initial regulatory flexibility analysis does not assess other federal rules that may
overlap or conflict with the new label requirements.
(Response 42) We are not aware of any other federal rules that overlap or conflict with the
requirements in this regulation. Therefore, we did not identify any such rules in the RFA
section of the PRIA or this final RIA.
(Comment 43) A comment suggested that no estimates are provided on the number of small
businesses subject to alternate compliance dates, or exempt from the rule altogether, if
such guidance will be provided.
(Response 43) This comment correctly asserts that we did not conduct a full economic analysis
for alternative regulatory options (e.g., certain exemptions and different compliance
dates) that we suggested were possible regulatory approaches in the PRIA. The final
rules and this Final RIA, however, clarify that no small businesses are exempted from the
rule and analyze costs pertaining to the compliance dates selected for both small
businesses and large businesses. For purposes of this analysis, we consider a small
business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Under the final
rules, the compliance period is 36 months for small businesses and 24 months for large
businesses. We estimate that about 95 percent of food manufacturers (i.e., close to
30,000 manufacturers) have annual food sales of less than $10 million.
H.
Costs and Benefits of Regulatory Options – Detailed Analysis
We have identified five regulatory options for the final rules:
1.) No new federal regulatory action;
45
Nutrition Facts/Serving Size Combined Final RIA
2.) The final rules that would give small manufacturers a 3-year compliance period and large
manufacturers a 2-year compliance period;
3.) The final rules, but with a 2-year compliance period for all manufacturers;
4.) The final rules, but with a 3-year compliance period for all manufacturers;
5.) The final rules, but with a 4-year compliance period for all manufacturers; and,
Requiring different compliance dates for the Nutrition Facts label rule and the Serving
Size rule is a policy choice that could be made. Therefore, in our discussions of regulatory
alternatives (2) through (6), we present impact estimates with both sub-options: the same and
differing compliance dates.
1. Option 1: No new federal regulatory action
This option serves as our baseline. We define costs and benefits relative to this baseline.
By definition, the baseline has no costs and no benefits.
2. Option 2: The final rules that would give small manufacturers a 3-year compliance period and
large manufacturers a 2-year compliance period
Under this option, some manufacturers would be required to undertake an extensive label
change due to the effect of the changed definition of a single-serving container on the
permissibility of certain health and nutrient content claims, while other manufacturers would
need to undertake a major redesign of their labels to include a Nutrition Facts Panel that had not
previously been required, to satisfy the requirements of DCL for packages of certain sizes, or to
remove a health or nutrient content claim in response to certain changes in the DVs, RACCs, or
the new definition of dietary fiber. All manufacturers would need to undertake a minor label
change due to certain changes in the nutrition information that must be declared on Nutrition and
Supplement Facts labels (e.g., mandatory declaration of “Added Sugars,” increasing the
46
Nutrition Facts/Serving Size Combined Final RIA
prominence of calories), and some manufacturers would need to undertake a minor label change
due to the update, modification, and establishment of certain RACCs or to the change in the
definition of a single-serving container. Some manufacturers may choose to reformulate their
product as a result of new mandatory label declarations or for the purpose of continuing to make
certain health or nutrient content claims which their preexisting formulation would no longer be
eligible to make as a result of changes to the regulations. Also, some manufacturers will incur
study costs and additional labeling costs (minor label change) related to the new dietary fiber
definition, and some will incur recordkeeping costs associated with the requirement that under
certain circumstances manufacturers make and keep records sufficient to verify the label
declaration for the amounts of added sugars, dietary fiber, soluble fiber, insoluble fiber, vitamin
E, and folate/folic acid. Finally, we estimate future labeling costs for UPCs based on the
assumptions that some manufacturers will undertake (1) an extensive label change due to the
effect of the changed definition of a single-serving container on the permissibility of certain
health and nutrient content claims and (2) a major redesign of their labels to include a Nutrition
Facts Panel that had not previously been required. The requirements of the final rules are
described in greater detail in the individual final rules published in the Federal Register.
Costs and benefits under this regulatory option are estimated using a weighted average of
costs and benefits using a 2-year compliance period for large businesses and costs and benefits
using a 3-year compliance period for small businesses. The respective weights used are the
percentage of sales earned by large businesses (which for purposes of this analysis we consider
to be businesses earning $10 million or more in annual food sales) and the percentage of sales
earned by small businesses (which for purposes of this analysis we consider to be businesses
earning less than $10 million in annual food sales). Using Nielsen data, as well as the
47
Nutrition Facts/Serving Size Combined Final RIA
classifications of large and small businesses above, we estimate that 95 percent of sales are
earned by large businesses and 5 percent of sales are earned by small businesses.
a. Costs
i.) Relabeling Costs
In order to comply with the final rules, some manufacturers would have to change some
of their labels for the Serving Size final rule, whereas all manufacturers would have to make
changes to all of their labels to comply with changes resulting from the Nutrition Facts Label
final rule (e.g., increasing the prominence of calories). Relabeling costs were estimated using
FDA’s Labeling Cost Model (Ref. 1). The model, which was built based on discussions with
trade associations and product manufacturers, provides estimates of the costs of making labeling
changes for a range of food products. Because of the number of steps involved in changing the
information on food packaging and labeling, the entire labeling change process generally takes
several months (Ref. 1).
Labeling costs, which include labor, materials, inventory (discarded inventory and
disposal costs), recordkeeping, and in certain cases recurring costs associated with package size
increases, are first calculated on a per-UPC basis and then aggregated across each product
category, and are calculated separately as low, mean, and high cost estimates (Ref. 1). To
determine the UPC counts in each product category, the model utilizes 2012 Nielsen ScanTrack
data (Ref. 1). The model allows the user to select the types of products that would be covered
under the regulation, the type of label change (extensive, major, or minor, as described in detail
in Section E) that would be required under the regulation, and the compliance period (3 months
to 60 months, in three month increments) (Ref. 1).
48
Nutrition Facts/Serving Size Combined Final RIA
It is important to note that within the data on broad categories of food, there are products
that are regulated by USDA (e.g., Salisbury steak frozen dinner) (Ref. 15). We are not able to
separate out from the total number of UPCs in each category the number of UPCs that are
regulated by USDA, and which therefore are not covered by the final rules. Based on a random
sample of UPCs taken from Nielsen Weekly Scanner data, we estimate the percentage of UPCs
which fall under USDA jurisdiction to be 5 percent and build this into our analysis.16
Available data show that most products that are voluntarily relabeled are relabeled in a
two- to five-year cycle, with private-label17 products less likely to be relabeled in any given year
than branded products (Ref. 1). Manufacturers who can coordinate a required labeling change
(regulatory labeling change) with a planned voluntary labeling change (non-regulatory labeling
change) will incur lower costs associated with the required labeling change than they would
otherwise (Ref. 1). Longer compliance periods increase the proportion of required labeling
changes that can be coordinated with planned voluntary labeling changes (Ref. 1). However,
even if manufacturers can coordinate a required labeling change, the FDA Labeling Cost Model
includes costs of administrative and recordkeeping activities associated with labeling changes
because manufacturers still incur costs associated with understanding the regulation, determining
their response, tracking the required change throughout the labeling change process, and
reviewing and updating their records of product labels (Ref. 1). Other types of costs, though,
such as prepress, graphic design, and engraving plates or cylinders, are not attributable to the
16
Nielsen Weekly Scanner data consist of weekly purchase and pricing data generated from participating retail store
point-of-sale systems in all U.S. markets. See http://www.nielsen.com/us/en.html for more information.
17
Branded products make their way to store shelves by way of branded food manufacturers and distributors (e.g.,
Hunt’s ketchup, French’s mustard). Private-label products make their way to store shelves either by way of in-house
manufacturing or manufacturers who specialize in the manufacture of private-label products (e.g., Wal-Mart’s
“Great Value” product line).
49
Nutrition Facts/Serving Size Combined Final RIA
regulation if the required labeling change is coordinated with a planned voluntary label change
(Ref. 1).
With a 2-year compliance period, the Labeling Cost Model estimates that manufacturers
of 74 percent of private-label conventional food products would have to undertake a labeling
change that is uncoordinated with their regulatory scheduled changes. For branded conventional
food products we estimate no uncoordinated labeling changes. For dietary supplements, 78
percent of branded products and 84 percent of private-label products would have to undertake an
uncoordinated labeling change.
With a 3-year compliance period, the Labeling Cost Model estimates that 43 percent of
private-label conventional food products would have to undertake an uncoordinated labeling
change. For branded conventional food products, we estimate no uncoordinated labeling
changes. For dietary supplements, 55 percent of branded products and 69 percent of privatelabel products would have to undertake an uncoordinated labeling change.
The number of UPCs that would have to undertake coordinated versus uncoordinated
labeling changes for each of the final rules is illustrated in Table 3. Table 3 also shows the
number of UPCs that can undertake one label change to satisfy the requirements of all of the
final rules.
The final Nutrition Facts Label rule covers all packaged food subject to FDA regulation,
as reflected in the total UPC count of 780,246. As described earlier, all food UPCs will have to
undergo various changes to the Nutrition and Supplement Facts labels, as applicable. In
addition, based on analyses conducted using a combination of Gladson and Mintel data, we
estimate that under the final Nutrition Facts Label rule, some products will need to relabel to
50
Nutrition Facts/Serving Size Combined Final RIA
remove a health or nutrient content claim in response to certain DV changes.18 Additionally,
based on analysis of the combined Gladson/Mintel data, we estimate that under the final
Nutrition Facts Label rule, some products will need to relabel to remove a health or nutrient
content claim in response to the new definition of dietary fiber. Further detail regarding these
analyses is provided in Subsection ii (Reformulation Costs).
The final Serving Size rule is updating, modifying, and establishing RACCs for some
foods. Using Nielsen data, we identified UPCs affected by amendments to the RACC
requirements and estimate that these UPCs will need to undergo a label change. In addition, the
final Serving Size rule is changing the definition of a single-serving container such that all food
UPCs containing less than 200 percent of the RACC must be labeled as a single-serving
container. As a result, those food UPCs with RACCs of 100 g or 100 mL or larger and that are
packaged such that they contain more than 150 percent but less than 200 percent of the RACC
will no longer have the option of being labeled as either one or two servings (as they did under
the preexisting requirements). Some of these food UPCs will need to change their label so that
the label lists just one serving. Using Nielsen data, we identified UPCs packaged such that the
UPC contains between 150 percent and 200 percent of the RACC, and conservatively estimate
that these UPCs will need to undertake such a label change. Further, under the final Serving Size
rule, some food products will be required to include a DCL (i.e., certain food products in
packages that contain at least 200 percent and up to and including 300 percent of the RACC,
certain food products in discrete units in which each discrete unit weighs at least 200 percent and
up to and including 300 percent of the RACC). We identified, using Nielsen data, those UPCs
18
The Gladson Nutrition Database and the Mintel Global New Products Database are commercial label databases
which contain information gathered since approximately the year 2000 on product type, brand, ingredients, and
verbatim wording manufacturers use on product labels. For more information, see http://www.mintel.com and
http://www.gladson.com.
51
Nutrition Facts/Serving Size Combined Final RIA
packaged such that the UPC contains at least two times the RACC but no more than three times
the RACC, and estimate that these UPCs will require a DCL. Using Gladson and Mintel data,
we estimate that some UPCs will need to be relabeled to remove a health or nutrient content
claim in response to certain RACC changes. Further detail regarding this analysis is provided in
Subsection ii (Reformulation Costs). Changes in the RACCs of certain products under the final
Serving Size rule and/or modified nutrient labeling requirements under the Nutrition Facts final
rule (e.g., mandatory labeling of potassium, a change in the daily value of sodium) would require
that certain products bear a Nutrition Facts panel for which none was previously required. Based
on comments we have received, it appears that most products for which a Nutrition Facts panel
will be required when such labeling was not required previously are certain coffee, tea, and
bottled water products. Using Nielsen data, we identified the total number of coffee, tea, and
water UPCs in the marketplace and conservatively estimate that none required a Nutrition Facts
panel previously and that all will require a Nutrition Facts panel under the combined effects of
the two regulations.19 For purposes of this analysis, because we have no way of reliably
allocating affected UPCs across the Nutrition Facts Label and Serving Size rules, we assign all of
these UPCs to the Serving Size rule. This has no effect on total labeling costs under the same
compliance date scenario, which is the scenario that is consistent with the final rules, and
conservatively overestimates total labeling costs under the different compliance dates scenario.
Finally, the change in the definition of a single-serving container mandated by the final Serving
Size rule to now require products that both (i) have RACCs of 100 g or 100 mL or larger and (ii)
are packaged such that they contain more than 150 percent but less than 200 percent of the
19
As noted previously in this document, the Serving Size final rule announces that we intend to consider the
exercise of enforcement discretion with respect to with respect to mandatory nutrition labeling on products that
would have been exempt from nutrition labeling under § 101.9(j)(4) prior to the effective date of the two rules.
52
Nutrition Facts/Serving Size Combined Final RIA
RACC to be defined as a single-serving container, versus offering such products the option of
being labeled as one or two servings, will require some of these products to increase their
package size to continue to make a health or nutrient content claim. Using Nielsen and Mintel
data, we identified UPCs for products that both (1) are packaged such that the UPC contains
between 150 percent and 200 percent of the RACC and (2) make a health or nutrient content
claim. Based on these data, we conservatively estimate that certain UPCs will need to increase
their package size to continue to make certain health or nutrient content claims.
Table 3 - Industry Coverage by Final Rule - Coordinated vs. Uncoordinated
Final Rule
Source of
Label
Change
Type of
Label
Change
Affected UPCs
Affected UPCs
Different Compliance Dates
Same Compliance Date
Uncoordinated
Coordinated
Uncoordinated
Coordinated
Minor
200,756
509,808
106,099
269,433
Major
494
985
494
985
Major
1,236
2,582
1,236
2,582
Minor
51,231
12,963
51,231
12,963
Major
57
15
57
15
Major
95
24
95
24
Nutrition Facts
Label (NFL)
Conventional
Food
Conventional
Food
Conventional
Food
Dietary
Supplements
Dietary
Supplements
Dietary
Supplements
Various
Changes to the
NFL
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Dietary Fiber
Definition
Various
Changes to the
NFL
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Dietary Fiber
Definition
Total NFL
UPCs
Serving Size
(SS)
Conventional
Food
Conventional
Food
780,246
Changes Due
to RACC
Amendments
Change in
Definition of
445,214
Minor
31,048
111,578
31,048
111,578
Minor
29,536
83,690
29,536
83,690
53
Nutrition Facts/Serving Size Combined Final RIA
Single-Serving
Size Container
Conventional
Food
Conventional
Food
Conventional
Food
Conventional
Food
DCL
Claim Removal
Related to
RACC Change
First Time
Nutrition Facts
Panel Due To
RACC
Proposals
Package Size
Change to
Retain Claims
as an Effect of
Change of
Definition of
Single-Serving
Container
Major
11,268
33,204
11,268
33,204
Major
224
514
224
514
Major
8,967
22,186
8,967
22,186
Extensive
2,817
-
2,817
-
335,032
Total SS UPCs
335,032
1,115,278
780,246
Total
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of
this analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a
large business to be a business with annual food sales of $10 million or more.
From Table 3 it can be seen that if the rules have different compliance dates, then the
covered products under the final Serving Size rule (335,032 UPCs) would have to undergo two
label changes—one for the final Nutrition Facts label rule and one for the final Serving Size
rule—or incur the cost of accelerating the second label change so as to coordinate the two
revisions. However, a UPC subject to multiple label changes under one rule only has to apply a
one-time change under that rule. For example, under the final Nutrition Facts label rule, a UPC
subject both to changes to the Nutrition and Supplement Facts labels that affect all UPCs (e.g.,
changes to the listing of required nutrients) and claim removal related to the new dietary fiber
definition was counted only under the latter change, because we considered major label changes
to subsume minor label changes for the purpose of this economic analysis. When both of the
final rules have the same compliance date, the covered products under the final Serving Size rule
can satisfy the requirements of both the final Nutrition Facts Label rule and the final Serving
54
Nutrition Facts/Serving Size Combined Final RIA
Size rule under one label change. For example, a UPC subject both to changes to the Nutrition
and Supplement Facts labels that affect all UPCs (e.g., changes to the listing of required
nutrients) under the final Nutrition Facts Label rule and DCL under the final Serving Size rule
was counted only under the final Serving Size rule because, as noted previously, we considered
major label changes to subsume minor label changes for the purpose of this economic analysis.
Label cost estimates in 2014 dollars are presented in Table 4. We first compare the cost
of the final rules if they have different compliance dates with the cost under the same compliance
date.20 Such a comparison illustrates the reduction in label costs that firms would experience if
they undertook all of the label changes mandated by the final rules under one label redesign. The
labeling cost of the final rules, if they had different compliance dates, would be between $3.2
billion and $3.4 billion at the mean estimate. This translates into per UPC costs (at the mean) of
between $4,101 and $4,358 per UPC.
If the final rules have the same compliance date, the cost savings are significant. As
described in the “Same Compliance Date” columns in Table 4, the labeling cost of the final rules
would be reduced by about 20 percent, to between $2.6 billion and $2.7 billion, at the mean
estimate. This translates into per-UPC costs (at the mean) of between $3,332 and $3,460.
Again, these cost savings arise because firms that are affected by both of the final rules are now
able to undertake just one label change to comply with the requirements of the two final rules.
Table 4 - Summary of Relabeling Costs of the Final Rules (in millions of 2014$)
Final Rule
Source of
Label
Change
Type of
Label
Change
Different Compliance Dates
20
Same Compliance Date
The Nutrition Facts Label rule and the Serving Size rule have the same compliance date. Thus, the “Different
Compliance Dates” labeling cost estimates are mainly for illustration purposes. That said, these estimates are likely
conservative; had the final rules had different compliance dates, the dates would likely have been close enough
together so that manufacturers would have still been able to undertake just one label change to comply with the
requirements of the two final rules.
55
Nutrition Facts/Serving Size Combined Final RIA
Low
Mean
High
Low
Mean
High
Minor
$607
$1,340
$2,535
$321
$708
$1,339
Major
$3
$7
$12
$3
$7
$12
Major
$8
$17
$31
$8
$17
$31
Minor
$116
$253
$467
$116
$253
$467
Major
$0.3
$1
$1
$0.3
$1
$1
Major
$1
$1
$2
$1
$1
$2
$322
$421
$530
$322
$421
$530
$1,057
$2,040
$3,578
$771
$1,408
$2,382
Minor
$102
$231
$442
$102
$231
$442
Minor
$96
$209
$393
$96
$209
$393
Major
$76
$162
$303
$76
$162
$303
Major
$2
$3
$6
$2
$3
$6
Major
$78
$147
$256
$78
$147
$256
Extensive
(3%)
$280
$560
$796
$280
$560
$796
Extensive
(7%)
$206
$410
$587
$206
$410
$587
Total SS (3%)
$634
$1,312
$2,196
$634
$1,312
$2,196
Total SS (7%)
$560
$1,162
$1,987
$560
$1,162
$1,987
TOTAL NFL +
SS
(3%)
$1,691
$3,352
$5,774
$1,405
$2,720
$4,578
Nutrition Facts
Label (NFL)
Conventional
Food
Conventional
Food
Conventional
Food
Dietary
Supplements
Dietary
Supplements
Dietary
Supplements
Various Changes to
the NFL
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Dietary Fiber
Definition
Various Changes to
the NFL
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Dietary Fiber
Definition
Analytical Testing Costs
Total NFL
Serving Size (SS)
Conventional
Food
Conventional
Food
Conventional
Food
Conventional
Food
Conventional
Food
Conventional
Food
Conventional
Food
Changes Due to
RACC
Amendments
Change in
Definition of
Single-Serving
Container
DCL
Claim Removal
Related to RACC
Change
First Time Nutrition
Facts Panel Due To
RACC Proposals
Package Size and
Claim Effects of
Change of
Definition of
Single-Serving
Container
Package Size and
Claim Effects of
Change of
Definition of
Single-Serving
Container
56
Nutrition Facts/Serving Size Combined Final RIA
TOTAL NFL +
SS
(7%)
Annualized
(3%)
Annualized
(7%)
$1,617
$3,202
$5,565
$1,331
$2,570
$4,369
$114
$225
$388
$94
$183
$308
$153
$302
$525
$126
$243
$413
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of this
analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Analytical testing costs associated with the
mandatory declarations of potassium and vitamin D are conservatively estimated across all conventional food and
dietary supplement UPCs using the FDA Labeling Cost Model. The cost of an extensive label change includes
recurring annual costs associated with increased package sizes. Annualized Amount = Amount / Annualizing Factor.
3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
With the same compliance date, it can be seen that the labeling cost estimates of the final
rules range from $1.4 billion to $4.6 billion using a 3 percent discount rate, and from $1.3 billion
to $4.4 billion using a 7 percent discount rate. Table 4 also presents annualized labeling costs,
using both a 3 percent and a 7 percent discount rate.
The final Nutrition Facts Label rule requires that under certain circumstances
manufacturers make and keep records to verify the amount of added sugars, dietary fiber, soluble
fiber, insoluble fiber, vitamin E and folate/folic acid in products. Thus, manufacturers must
maintain records sufficient to verify the label declaration for these nutrients for at least 2 years
after introduction or delivery for introduction of the food into interstate commerce and must
provide these records upon request by FDA, during an inspection, for official review and
photocopying or other means of reproduction. The records required to be retained could be
nutrient database analyses, recipes or formulations, batch records, or other appropriate
verification documentation that a manufacturer has that verifies the nutrient content in the final
product. We estimate that manufacturers currently maintain the above-mentioned records in the
normal course of doing business. Thus, the time burden to the covered food manufacturer would
be to review these records to verify the amount of such nutrients in a food and to make such
records available to appropriate regulatory officials upon request. Referring to the Paperwork
57
Nutrition Facts/Serving Size Combined Final RIA
Reduction Act (PRA) analysis in Section III of this final RIA, we estimate that manufacturers
will incur 187,726 recordkeeping hours initially and 216 recordkeeping hours on an annual
recurring basis. According to the Bureau of Labor Statistics National Industry-Specific
Occupational Employment and Wage Estimates, the mean hourly wage of an operations manager
in the food manufacturing industry is $55.59 in 2014 dollars. We increase this cost by 100
percent to account for benefits and overhead, making the total cost of time $111.18. Thus, total
recordkeeping costs, discounted over 20 years and in 2014 dollars, are estimated to be
approximately $21.4 million using a 3 percent discount rate and roughly $21.3 million using a 7
percent discount rate.
Under the final Nutrition Facts Label rule, manufacturers of food products that contain an
isolated or synthetic non-digestible carbohydrate that does not meet the new definition of dietary
fiber will not be able to declare the isolated or synthetic non-digestible carbohydrate as dietary
fiber on the Nutrition or Supplement Facts label. However, manufacturers can submit a citizen
petition to request the agency amend the definition of “dietary fiber” to include the isolated or
synthetic non-digestible carbohydrate as a “dietary fiber” for purposes of nutrition labeling. In
addition, if the isolated or synthetic non-digestible carbohydrate is the subject of an authorized
health claim, FDA would consider the carbohydrate to be a dietary fiber with a beneficial
physiological effect to human health and would amend the definition of “dietary fiber” to include
the carbohydrate as a listed dietary fiber. If a citizen petition is submitted and the request to
amend the definition is granted, or if the isolated or synthetic non-digestible carbohydrate is
subject to an authorized health claim, the isolated or synthetic non-digestible carbohydrate is
considered to meet the definition of dietary fiber and the definition would be amended to include
58
Nutrition Facts/Serving Size Combined Final RIA
the dietary fiber in the listing of dietary fibers that must be included in the total amount of dietary
fiber declared on the Nutrition or Supplement Facts label when present in a food.
FDA estimates that there are currently about 35 isolated or synthetic non-digestible
carbohydrates that are added to food. Based on a review of the relevant literature, FDA found
that certain isolated or synthetic fibers have a demonstrated beneficial physiological effect to
health in the final rule, and includes such fibers in the definition for dietary fiber. Based on
FDA’s review, the scientific evidence for approximately 28 isolated or synthetic carbohydrates is
generally mixed, or appears to be insufficient to demonstrate a beneficial physiological effect to
human health. However, it is not clear whether there may be additional data or information
concerning the beneficial health effects of these non-digestible carbohydrates that interested
persons have and are not yet publically available. Therefore, FDA is not making a determination
on whether these other non-digestible carbohydrates meet the definition of “dietary fiber” in the
final rule and plans to provide an opportunity for comment on the available scientific evidence
for these non-digestible carbohydrates. Thus, FDA intends to publish a separate notice to seek
comment on the available scientific data on these non-digestible carbohydrates to determine if it
should consider additional non-digestible carbohydrates to be added to the list of dietary fibers.
Although FDA intends to publish a separate notice to seek comment on the available scientific
data on these non-digestible carbohydrates to determine if it should consider additional nondigestible carbohydrates to be added to the list of dietary fibers, for purposes of the economic
analysis we estimate additional studies will be needed for the approximately 28 remaining
isolated or synthetic non-digestible carbohydrates. Thus, we estimate the cost associated with
conducting 84 studies, three for each of the 28 remaining isolated or synthetic non-digestible
59
Nutrition Facts/Serving Size Combined Final RIA
carbohydrates.21 FDA estimates that to conduct such a study would require approximately 480
labor hours (= 2 scientists x 20 hours/week x 12 weeks) and about $10,000 in material costs.
According to the Bureau of Labor Statistics National Industry-Specific Occupational
Employment and Wage Estimates, the mean hourly wage of a biological scientist is $38.08 in
2014 dollars. We increase this cost by 100 percent to account for benefits and overhead, making
the total cost of time $76.16 per hour. Thus, we estimate that the total cost of conducting 84
studies is approximately $3.9 million in 2014 dollars (= (480 x $76.16 + $10,000) x 84).22
It is possible that a request to amend the definition of “dietary fiber” to add a new fiber,
or the submission of a health claim for an isolated or synthetic non-digestible carbohydrate may
be submitted to FDA prior to the end of the final Nutrition Facts Label rule’s compliance period,
but not granted or authorized until after the end of the final Nutrition Facts Label rule’s
compliance period. In such instances, food products which contain such an isolated or synthetic
non-digestible carbohydrate would have to undertake two labeling changes – one to comply with
the final rules and an additional labeling change, after a citizen petition is granted or a health
claim is authorized, to declare the isolated or synthetic non-digestible carbohydrate as dietary
fiber on the Nutrition or Supplement Facts label. We estimate that at most approximately 20
percent of UPCs that contain an isolated or synthetic non-digestible carbohydrate might have to
undertake such an additional labeling change. Using Gladson and Mintel data, we estimate that
close to 60,000 UPCs contain an isolated or synthetic non-digestible carbohydrate. Thus, we
estimate that at most close to 12,000 (= 0.2 x 60,000) UPCs might have to undertake this type of
21
We have no way of estimating the number of isolated or synthetic non-digestible carbohydrates that will be
discovered in the future.
22
Estimated fiber study costs are small relative to the estimated total cost of the final rules. Thus, to the extent that
we underestimated the number of studies per fiber, it will have a negligible impact on the estimated total cost of the
final rules.
60
Nutrition Facts/Serving Size Combined Final RIA
additional labeling change at a cost, estimated using the FDA Labeling Cost Model and in 2014
dollars, of between $176 million and $1,101 million, with a mean estimate of $430 million.23
The above costs are summarized in Table 5 below. Table 5 also presents annualized
costs, using both a 3 percent and a 7 percent discount rate.
Table 5 – Summary of Recordkeeping Costs and Study and Additional Labeling Costs
Related to the New Definition of Dietary Fiber (in millions of 2014$)
Low
Mean
High
Recordkeeping Costs
$21
$21
$21
Study Costs Related to New Definition of
$4
$4
$4
Dietary Fiber
Additional Labeling Costs Related to New
$176
$430
$1,101
Definition of Dietary Fiber
$201
$455
$1,126
Total
Annualized (3%)
$14
$31
$76
Annualized (7%)
$19
$43
$106
Notes: These costs do not depend on compliance period. Annualized Amount = Amount / Annualizing Factor. 3
percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
According to the FDA Labeling Cost Model, the addition of a Nutrition or Supplement
Facts panel to a previously unlabeled UPC, such as to a new UPC, represents a major label
change. New UPCs under the final rules would have had to add a label under NLEA, which is to
say that new UPCs under the final rules would have had to make a major label change under
NLEA. Thus, the labeling costs for future UPCs subject to minor and major label changes under
the final rules would not generally be considered a cost of the final rules. However, future UPC
growth is relevant for UPCs that would need to bear a Nutrition Facts Panel for which none was
previously required and UPCs which would require an extensive label change. We estimate that
23
The additional labeling change represents a minor labeling change (Ref. 1). In light of the physiological benefits
associated with consuming dietary fiber, we estimate that manufacturers, upon receiving post-compliance period
approval of their dietary fiber citizen petition or health claim petition, would relabel as quickly as possible. Thus,
labeling costs are estimated using the “3 Months or Less” compliance period option in the FDA Labeling Cost
Model. However, it should be noted that each amendment that adds a new fiber to the definition will also have a
compliance period, which will provide manufacturers with additional time, and may allow them to coordinate their
fiber-related label change with other future planned label changes.
61
Nutrition Facts/Serving Size Combined Final RIA
a fixed fraction of future UPCs fall into these categories, and we estimate associated labeling
costs for UPCs affected by these requirements over the next 20 years. Looking at Nielsen data,
we estimate that food UPCs have historically grown by about 1.3 percent per year. Lacking data
or other information that indicates that historical and future UPC growth rates will diverge, we
use 1.3 percent as our estimate of future annual growth in food UPCs. Using this annual growth
rate in food UPCs, as well as the FDA Labeling Cost Model, we estimate labeling costs for
UPCs affected by the above requirements over the next 20 years of between $66 million and
$198 million, with a mean estimate of $135 million, using a 3 percent discount rate, and between
$35 million and $107 million, with a mean estimate of $72 million, using a 7 percent discount
rate (2014$). These costs are summarized in Table 6 below. Table 6 also presents annualized
costs, using both a 3 percent and a 7 percent discount rate.
Table 6 – Summary of Labeling Costs Associated With Future UPC Growth (in millions of
2014$)
Type of
Label
Low
Mean
High
Change
Present Value (3%)
First Time Nutrition Facts Panel Due
Major
$6
$14
$26
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$60
$121
$172
Serving Container
$66
$135
$198
Total
Present Value (7%)
First Time Nutrition Facts Panel Due
Major
$4
$10
$18
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$31
$62
$89
Serving Container
$35
$72
$107
Total
Annualized (3%)
$4
$9
$13
Annualized (7%)
$3
$7
$10
Notes: Annualized amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
62
Nutrition Facts/Serving Size Combined Final RIA
ii.) Reformulation Costs
The final rules could motivate food manufacturers to reformulate their products in
response to the requirements in these final rules. Incentives to reformulate can be categorized
into three groups: (i) reformulations to maintain health and nutrient content claims motivated by
DV, RACC changes and changes in the definition of dietary fiber, (ii) reformulations motivated
by the mandatory labeling of the quantity of added sugars, and (iii) reformulations motivated by
the reduction in the Vitamin B12 DV. We estimate reformulation costs associated with each
group in turn.24
Cost of Reformulation of Food to Maintain Health Claims and Nutrient Content Claims
The final rules could affect manufacturers of food products who currently make certain
nutrient content claims or health claims authorized under the Federal Food, Drug, and Cosmetic
Act (FD&C Act) on their product’s label. The use of a claim may add value to a product’s brand
by providing additional nutritional information that may improve consumer perception of a
product compared to that of a competitor’s product. Under the FD&C Act and FDA regulations,
RACCs (and sometimes serving sizes as well) are to be used to determine whether a product
meets the criteria for a nutrient content claim or health claim. The changes the final rules make
to product label requirements (specifically, changes in the RACCs and changes in the DVs
leading to changes in the declared percent DV and changes in the definition of dietary fiber so as
to exclude certain isolated or synthetic fibers from the definition) could cause a product currently
making a health or nutrient content claim to become ineligible to make the claim. In the absence
of regulations changing the requirements for nutrient content claims and health claims, as
24
It should be noted that labeling costs associated with product reformulation (e.g., changing nutrient amounts,
changing ingredient lists) are subsumed under the labeling costs estimated in Subsection i (Labeling Costs) of this
RIA.
63
Nutrition Facts/Serving Size Combined Final RIA
discussed above, this would require manufacturers of these products to either remove the claim
from their product’s label, reformulate, or repackage in order to continue to make the claim.25 It
is difficult to predict how the requirements of the final rules would influence manufacturers’
decisions to reformulate, repackage, or remove the claim from the product label. We therefore
estimate that, given the requirements of the final rules, some manufacturers would reformulate
while others would remove the claim from their product’s label.
To estimate reformulation costs, we gathered data to provide a representative sample of
the total number of food products from the Gladson and Mintel databases. To determine which
UPCs would need to either reformulate to continue to make health or nutrient content claims or
relabel to remove such claims from their label related to RACC changes, the data were refined to
identify UPCs with new or changing RACCs. Label data were aggregated for each food/nutrient
category specifically targeted by the final rules through a change in RACC. For each product,
the current nutrition and serving size data were used to calculate new hypothetical nutrition data
25
To illustrate, consider these examples:
1) Brand X Low Fat Ice Cream – Brand X currently packages its product in pint-sized cartons, with 3
grams of fat per ½ cup. The Serving Size final rule increases the RACC of ice cream from ½ cup to 2/3 cup.
Pursuant to 21 CFR 101.62(b)(2), under preexisting RACCs, this product is eligible to make a “low fat” claim on its
label since it contains 3 grams or less of total fat per RACC, assuming it satisfies the other requirements of
paragraph (b)(2). By increasing the RACC, Brand X would contain more than 3 grams of fat per RACC, rendering
it ineligible to keep the “low fat” claim.
2) Brand Y Greek Yogurt – Brand Y currently sells yogurt in 24 oz. packages, with 8 oz. per serving. It
also contains 22 percent of the DV of calcium and, pursuant to 21 CFR 101.54(b), currently qualifies for the nutrient
content claim “excellent source of calcium” since it contains 20 percent or more of the calcium DV per RACC. The
Serving Size final rule decreases the RACC of yogurt to 6 oz., which brings the percent DV of calcium down to less
than 20 percent. Brand Y is no longer permitted to claim the product is an “excellent source of calcium,” and must
either reformulate to keep the claim, or change the claim to “good source” of calcium, the latter which, pursuant to
21 CFR 101.54(c), requires that the product contains 10 to 19 percent of the calcium DV per RACC.
3) Brand Z Breakfast Bar – Pursuant to 21 CFR 101.54(b), in order to make an “Excellent Source of Fiber”
claim, Brand Z breakfast bars must contain 20 percent or more of the fiber DV (DV = 28 grams) per RACC (RACC
= 40 grams). Brand Z breakfast bars currently contain 6 grams of dietary fiber per 40 gram bar (21 percent of DV
per RACC) and, thus, are currently eligible to make the “Excellent Source of Fiber” claim on their label. The source
of this dietary fiber, however, is a fiber known as Fiber A, an isolated/synthetic fiber which does not meet the
definition of dietary fiber under the Nutrition Facts final rule. Thus, under the Nutrition Facts final rules, Brand Z
breakfast bars must either reformulate their breakfast bars using natural dietary fiber or an isolated/synthetic dietary
fiber that meets the definition of dietary fiber to continue to make the “Excellent Source of Fiber” claim, or remove
the claim from the breakfast bar label.
64
Nutrition Facts/Serving Size Combined Final RIA
under the final RACCs. The data were then filtered down to only products with label health
claims (excluding qualified health claims) and nutrient content claims as stated in Appendices AD of the FDA Food Labeling Guide (Ref. 16) (e.g., “good source,” “low sodium,” “Vitamin C
added,” etc.). Products currently making label claims but rendered ineligible after the
introduction of the RACC changes were flagged as needing either relabeling to remove the claim
or reformulation to continue to make the claim. An equivalent refinement methodology was
used to estimate the number of UPCs for which, related to DV changes, either reformulation
would be required to continue to make a health or nutrient content claim about the product or
relabeling would be required to remove the claim from the label.
To estimate the number of UPCs which, related to the new definition of dietary fiber,
would need to be reformulated for the product to continue to bear a health or nutrient content
claim, or would need to be relabeled to remove the claim from their label, the data were refined
to identify UPCs containing at least one isolated or synthetic fiber. The data were then further
refined to identify which among these UPCs contain a fiber-related health or nutrient content
claim. Because of data limitations, we conservatively estimate that all UPCs that both contain at
least one isolated or synthetic fiber and that carry a fiber-related health or nutrient content claim
will need to either reformulate to continue to make such a claim or relabel to remove such a
claim from their label.
The Gladson and Mintel databases do not provide estimates of formula counts26. To
estimate formula counts, we relied on UPC and formula count data from the FDA Labeling Cost
Model (Ref. 1). The FDA Labeling Cost Model estimates both total UPCs as well as total
formulas for various food product categories. For each product category, we multiply the ratio
26
To understand the distinction between UPCs and formulas, Soda X has one “taste” (formula), but comes in a
variety of sizes (e.g., 12 ounce can, 2-liter bottle), each of which represents a unique UPC.
65
Nutrition Facts/Serving Size Combined Final RIA
of Formulas to UPCs estimated using the FDA Labeling Cost Model by the UPC counts
estimated using the Gladson and Mintel databases. The UPC and resulting formula counts are
illustrated below in Tables 7 and 8.
Table 7. Estimates by Product Category of the Number of UPCs and Formulas with a
Health or Nutrient Content Claim Affected by the RACC and DV Changes Mandated by
the Final Rules
Product Category
UPCs
Formulas
Baked Goods
570
468
Baking Ingredients
Beverages
Breakfast Foods
Candy and Gum
Dairy Foods
Dressings and Sauces
Entrees
Fruits and Vegetables
Pizza
Seafood
Side Dishes and Starches
Snack Foods
Soups
Dietary Supplements
Total
80
894
950
96
808
48
90
260
20
144
170
238
72
144
4,584
64
602
648
68
620
42
80
200
16
108
146
176
62
122
3,422
Table 8. Estimates by Product Category of the Number of UPCs and Formulas with a
Health or Nutrient Content Claim Affected by the New Definition of Dietary Fiber
Product Category
UPCs
Formulas
Baked Goods
1,514
1,242
Baking Ingredients
Beverages
Breakfast Foods
Candy and Gum
Condiments/Dips/Spreads
Dairy Foods
Desserts
Dressings and Sauces
Entrees
Fruits and Vegetables
122
992
1,976
126
70
804
28
72
592
104
66
98
668
1,348
92
58
616
22
62
530
80
Nutrition Facts/Serving Size Combined Final RIA
Infant Foods
Pizza
Seafood
Side Dishes and Starches
Snack Foods
Soups
Sweeteners
Dietary Supplements
Total
188
58
50
204
530
200
4
238
7,872
152
52
38
174
392
176
4
200
6,004
Reformulation costs are estimated using the FDA Reformulation Cost Model (Ref. 2).
The FDA Reformulation Cost Model, the development of which was based on an expert panel of
individuals who previously oversaw product reformulation at major food manufacturing
companies or who currently provide formulation consulting services to small and large food
manufacturers, estimates the costs to food manufacturers of reformulating foods based on
variations in (i) food product complexity (some products are more easily reformulated than
others), (ii) company size (larger companies put substantially more effort into reformulation than
smaller companies), (iii) reformulation types (reformulation of a non-critical minor ingredient, of
a critical minor ingredient, and of a major ingredient) and activities (determination of response to
regulation; project management; product reformulation/process modification; packaging
assessment and development; product and package performance testing; sensory evaluation;
analytical testing; production scale-up; discarding of unused inventory of raw materials,
packaging, and labels; and updating product records), and (iv) compliance period (costs are
higher for shorter compliance periods because if the compliance period is short, manufacturers
will incur increased costs for overtime labor, additional staffing, and rush charges with vendors
67
Nutrition Facts/Serving Size Combined Final RIA
and suppliers).27 A summary of per-formula costs by food product complexity, company size,
reformulation type, and reformulation activity is provided on pp. 3-47 to 3-52 of the FDA
Reformulation Cost Model. We estimate that the ingredients in question generally play a noncritical, minor role in product formulation.28
If we rely on the formula counts in Tables 7 and 8, we will likely overestimate the cost of
reformulation. First, the estimates of the number of affected formulas are likely high. More
detailed data would allow us to better refine our estimate of the number of affected formulas, but
we are not aware of any such data. Second, not all manufacturers would choose to reformulate
their product to continue to make the health or nutrient content claim, but rather might instead
choose to remove the health or nutrient content claim from their product’s label. If a
manufacturer decides to remove a claim instead of reformulating, we estimate that any loss in
sales directly attributable to the removal of the claim would be transferred to a substitute product,
with negligible value lost to society. Thus, in such a case, the relevant cost to consider is the
relabeling cost associated with printing new labels that do not contain the claim (which we
already capture in the labeling cost estimation section of this Final RIA). In lieu of reliable data
regarding manufacturers’ response to the final rules that would cause us to weight the split one
way or the other, we estimate that 50 percent of the Table 7 and 8 UPCs would choose to
reformulate and 50 percent would choose to relabel (we previously asked for comment on this
estimate, but received none that provided a basis upon which to adjust the estimate). This brings
the total formula count to 4,713 (0.5 × 3,422 + 0.5 × 6,004). Using this formula count and the
27
Product reformulation may result in higher ongoing production costs for food manufacturers. Such costs can only
partially be passed along to consumers, as demand for food is not perfectly inelastic (Ref. 17). However, due to data
limitations, we are unable to quantify either the total costs or how the costs are distributed across manufacturers and
consumers.
28
We previously asked for comment on our estimate that the ingredients in question generally play a non-critical,
minor role in product formulation, but received none that provided a basis upon which to adjust it.
68
Nutrition Facts/Serving Size Combined Final RIA
FDA Reformulation Cost Model, we estimate total reformulation costs associated with
maintaining health and nutrient content claims related to RACC and DV changes and the new
definition of dietary fiber of between $156 million and $642 million, with a mean estimate of
$346 million (2014$).
Cost of Reformulation of Foods That Significantly Contribute Added Sugars to Diets
The final rules, particularly the label disclosures of added sugars and the percent DV for
added sugars, may increase the visibility of nutrients that, based on current scientific evidence,
consumers should limit. These label changes may motivate some food manufacturers to
reformulate existing products. To estimate the number of formulas that would be reformulated
as a result of these label changes, we use a similar methodology to that which was used in the
previous section.
To identify products that significantly contribute added sugars to diets, we considered the
evidence supporting the 2015 DGA, as well as the Gladson and Mintel databases. The identified
products span the product categories listed in Table 8. We estimate reformulation costs using
both the FDA Labeling Cost Model and the FDA Reformulation Cost Model. The FDA
Labeling Cost Model provides formula counts for each of the identified products. These formula
counts, by product category, are reported in Table 9. A total of 113,968 formulas were identified
as significantly contributing added sugars to diets.
Table 9. Estimates of the Number of Formulas, by Product Category, That Significantly
Contribute Added Sugars to Diets
7.5%
9%
Product Category
Total Formulas
Reformulated Reformulated
Baked Goods
20,897
1,567
1,881
Beverages
25,584
1,919
2,303
Breakfast Foods
5,578
418
502
Condiments/Dips/Spreads
9,853
739
887
Dairy Foods
24,270
1,820
2,184
Dressings and Sauces
11,956
897
1,076
69
Nutrition Facts/Serving Size Combined Final RIA
Entrees
Infant Foods
Pizza
Total
10,502
1,901
3,427
113,968
Average:
788
143
257
8,548
9,403
945
171
308
10,257
The labeling changes for added sugars would likely have a variety of effects29 on
manufacturers regarding their decision to reformulate voluntarily. The required disclosures of
added sugars and the percent DV for added sugars on Nutrition and Supplement Facts labels
under the final rules is new, and so would require an entirely new line item, providing a set of
information that was not previously available.
We estimate that 7.5 to 9 percent of formulas that significantly contribute added sugars to
diets will be reformulated. In lieu of reliable data to precisely predict the extent of such
reformulation, we acknowledge that the actual rate of reformulation may be higher or lower than
this range (we previously asked for comment on this estimate, but received none that provided a
basis upon which to adjust the estimate). Referring back to Table 9, an estimated range of 7.5 to
9 percent reformulation yields total formula counts of between 8,548 and 10,257 formulas, with
an average formula count of 9,403 formulas. Using this average formula count and the FDA
Reformulation Cost Model, we estimate total reformulation costs associated with foods that
significantly contribute added sugars to diets of between $296 million and $1,221 million, with a
mean estimate of $658 million (2014$).
Cost of Reformulation of Foods Associated With Reduction in Vitamin B12 DV
29
We acknowledge that the requirement to enlarge and bold the calorie content on package labels may have an
impact. There are studies that suggest that increasing or bolding the font could lead to increases in attention of
readers (Ref. 18). However, a study by Lando and Lo (2013) (Ref. 19) concludes that enlarged font size for calories
did not independently affect label usability. Without further study, it is unclear how large the impact of increasing
prominence of calories on the label will be; hence we do not attribute costs or benefits to reformulation based on
calorie bolding alone.
70
Nutrition Facts/Serving Size Combined Final RIA
Under the final rules, the DV for vitamin B12 is being reduced from 6.0 mcg to 2.4 mcg.
This new DV may incent some manufacturers who currently fortify their products with vitamin
B12 to reformulate their products to reduce B12 amounts, for two reasons. First, those
manufacturers who currently fortify their products to 100 percent of the DV will be able to do so
under the final rules for less cost by using 3.6 mcg per serving less of vitamin B12. Second, those
manufacturers who currently fortify their products to 100 percent of the DV, at 6.0 mcg per
serving, will have a DV of 250 percent (= 6.0 mcg / 2.4 mcg) if they continue to fortify at 6.0
mcg per serving under the final rules. Some consumers may find this unappealing.
Manufacturers of ready-to-eat breakfast cereals are those most likely to fortify their products
with vitamin B12 (Ref. 20) and, thus, reformulate their products to reduce B12 amounts.30 We
estimate reformulation costs using both the FDA Labeling Cost Model and the FDA
Reformulation Cost Model. Using the FDA Labeling Cost Model, we estimate the number of
ready-to-eat breakfast cereal formulas to be 5,578. Using this formula count and the FDA
Reformulation Cost Model, we estimate total reformulation costs associated with reducing B12
amounts in ready-to-eat breakfast cereals of between $199 million and $819 million, with a mean
estimate of $442 million (2014$).
Summary of Reformulation Costs
Table 10 presents a summary of estimated reformulation costs. We estimate total
reformulation costs (in 2014 dollars) of between $651 million and $2,682 million, with a mean
estimate of $1,446 million. Table 10 also presents annualized reformulation costs, at 3 percent
and 7 percent.
30
Some vitamin B12-deficient individuals may not rely solely on fortified foods such as cereal to supplement their
diet with vitamin B12, but rather rely on dietary supplements in pill form. The majority of dietary supplements
contains levels of vitamin B12 well in excess of 100 percent of the DV, though, and thus are unlikely to be
reformulated.
71
Nutrition Facts/Serving Size Combined Final RIA
Table 10 - Summary of Reformulation Costs of the Final Rules (in millions of 2014$)
Category
Low
Mean
High
Claims
$156
$346
$642
Added Sugars
$296
$658
$1,221
B12
$199
$442
$819
$651
$1,446
$2,682
Total
Annualized (3%)
$44
$97
$180
Annualized (7%)
$61
$137
$253
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of this
analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Annualized Amount = Amount /
Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing
factors are calculated by summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t =
20).
Summary of Total Cost of the Final Rules
Table 11 summarizes the PV and annualized total cost of the final rules, using both a 3
percent and 7 percent discount rate.
Table 11 - Summary of Total Costs of the Final Rules (in millions of 2014$)
Low
Mean
PV (3%)
Relabeling
$1,405
$2,720
Recordkeeping
$21
$21
Fiber Studies
$4
$4
Additional Labeling
$176
$430
Future UPC Growth Labeling
$66
$135
Reformulation
$651
$1,446
$2,323
$4,756
Total
PV (7%)
Relabeling
$1,331
$2,570
Recordkeeping
$21
$21
Fiber Studies
$4
$4
Additional Labeling
$176
$430
Future UPC Growth Labeling
$35
$72
Reformulation
$651
$1,446
$2,218
$4,543
Total
Annualized (3%)
$156
$320
Annualized (7%)
$209
$429
High
$4,578
$21
$4
$1,101
$198
$2,682
$8,584
$4,369
$21
$4
$1,101
$107
$2,682
$8,284
$577
$782
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of this
analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Analysis assumes that the Nutrition Facts
Label and Serving Size rules will have the same compliance date. Annualized Amount = Amount / Annualizing
72
Nutrition Facts/Serving Size Combined Final RIA
Factor. 3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are
calculated by summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20). PVs of
fiber study, additional labeling, and reformulation costs are equivalent at 3 or 7 percent because we conservatively
estimate that these one-time costs are incurred upon publication of the rules instead of at the end of the compliance
period.
b. Benefits
Background
A large body of scientific research has consistently confirmed the significant effect of
diet on health, quality of life, and longevity (Refs. 21 – 23). In the United States, high intakes of
total energy (kilocalories), saturated fat, trans fat, and sodium, and low intakes of fruits,
vegetables, whole gains, and dairy products correlate with an increased risk of various chronic
health conditions (such as cardiovascular disease (CVD), obesity, diabetes, and osteoporosis)
that can impair the quality of life and decrease longevity. Treating these health conditions comes
at a considerable expense (Refs. 21 – 23). Chronic diseases, such as heart disease, cancer and
stroke are the leading causes of death and disability in the United States, and account for 70
percent of all deaths in the United States (Ref. 21). In 2005, 133 million Americans, almost one
out of every two adults, had at least one chronic illness (Ref. 21). An estimated 37 percent of
Americans suffer from CVD (Ref. 22), 11 percent of individuals 20 years and older have
diabetes, 35 percent of adults have pre-diabetes (Ref. 24), and an estimated 41 percent of the
population will receive a diagnosis of cancer during their lifetime (Ref. 25). While the causes of
these chronic diseases are multifactorial, having a poor diet contributes to excess morbidity and
mortality (Ref. 26), and numerous nutrients affect chronic disease risk.
Although many sources of nutrition information currently exist, nutrition labeling gives
consumers a combination of information and reminders that accompany foods at the points of
purchase and consumption. For that reason, nutrition information represents an important tool for
providing information to assist consumers in maintaining healthy dietary practices. The final
73
Nutrition Facts/Serving Size Combined Final RIA
rules (1) better align the information provided on Nutrition and Supplement Facts labels with
new data on consumption, scientific evidence on the relationship between nutrition and chronic
diseases, and dietary recommendations, (2) improve the design and content of the Nutrition and
Supplement Facts labels such that relevant information is more salient and easy to understand for
the purpose of informing consumption decisions, and (3) could have the effect of prompting
industry to reformulate products to maintain health claims and nutrient content claims and
reformulate products based on new label information required by the final rules.
In addition to alerting consumers to calorie and nutrient content, major predicted
elements of the consumer and industry response to the nutrition labeling final rules include:
Increased knowledge by consumers of the nutrient content of packaged foods,
which may help them make healthier food choices;
Increased ease of nutrition label use from the availability of more easy-to-
understand information for products that bear DCLs;
Greater transparency of the nutrient content of existing packaged foods, which
may give firms an incentive to provide additional items with healthier formulations; and
Potential reformulation of products to reduce sodium and added sugars or
increase other vitamins and minerals.
These responses could potentially result in reduced consumption of less nutritious food
products. Note that consumers may offset any reduction in their consumption of less nutritious
food products with consumption of unlabeled and less nutritious meals or snacks. Consumers
substitute between nutrient sources when attempting to modify their food choices (Ref. 27).
Because we lack any data or information on how consumers would substitute between foods in
74
Nutrition Facts/Serving Size Combined Final RIA
response to the labeling changes, the benefits estimates in this analysis may over- or understate
the realized effects of the final rules.
Nutrition labeling provides consumers with information they can use to compare products
and build a healthy diet that conforms to federal dietary recommendations, their nutritional
preferences, or both. The costs of consuming a poor diet include the value of the quality of life
lost to illness and other sources of disutility, such as taking medications every day, as well as the
value of years of life lost (YLL) from premature death. The costs of consuming a poor diet also
include the net lifetime cost of treating the diseases caused or exacerbated by poor diet. We
expect that the final rules will lead to changes in the prevalence and/or intensity of label use, as
well as some product reformulation, which may result in changes in nutrient intake. The benefits
of the final rules would come from consumer welfare gains primarily due to increases in health
and longevity generated by improvements in overall diet.
Welfare Estimates—Primary Sensitivity Analysis
We illustrate the welfare gains (benefits) of the final rules, BtLabel, using the formula:
BtLabel = ΔWLabel x s1 x USE x (1 – USDA) x POPt
(1)
where ΔWLabel represents the change in consumer welfare from NLEA, s1 is the weighted average
change in label content associated with the final rules relative to the 1993 rules that implemented
NLEA, USE represents estimated current use of the Nutrition Facts label, USDA represents the
percentage of labeled food regulated by the USDA Food Safety Inspection Service (FSIS), and
POPt is the population of the United States in period t. Equation 1 encapsulates our illustration
of the welfare gains from consumers using the information on the updated Nutrition and
Supplement Facts labels.
Changes in Consumer Welfare from NLEA (“ΔWLabel” )
75
Nutrition Facts/Serving Size Combined Final RIA
To illustrate the effect of the final rules on consumer welfare, we extrapolate from the
welfare effects estimated in a retrospective study on the impact brought about by NLEA. Using
data on women 19 to 50 years of age who are the main meal planner in their household on the
prevalence of package labels, label use, and food intake from the FDA Food Labeling and
Package Survey (FLAPS), the Diet and Health Knowledge Survey (DHKS), and the CSFII,
Abaluck (2011) estimates the change in nutrient intake attributable to changes brought about by
NLEA in nutritional information content across foods.31 Abaluck (2011) bases his estimates on a
structural model of food demand that accounts for substitution effects in food consumption,
differences in demand elasticities across products, and heterogeneity in the use and knowledge of
nutrition information.
Abaluck (2011) measures the consumer welfare gains as the WTP for nutrient content
based on revealed preference data, i.e., food consumption and prices. This hinges on the idea that
when labeling reveals the true marginal cost of consumption, an individual responds to that
information by internalizing the health costs as if they have experienced a change in the price of
that good. Then one can compare the change in nutrient intake based on changes attributable to
NLEA to the equivalent price change that would have to occur to produce the same response
given that preferences and tastes also influence the demand for food. One can then use the
difference in the perceived price of consumption before and after receipt of the information to
31
Abaluck (2011) uses the estimated percent of annual sales of packaged foods that labeled foods represent from the
report “Status of Nutrition Labeling of Processed Foods: 1995” by O’Brien (1995) (Ref. 28), who used the FLAPS
survey for various years. The FLAPS survey is created using a multistage sampling plan to select a representative
sample of food products from the retail packaged food supply. The FLAPS data provide comprehensive labeling
information for food products in the United States. Despite the fact that the FLAPS data set was created using a
sampling scheme biased towards the highest selling products within a product category, thus creating a not
completely random sample of products in the marketplace, this data set is the most robust, nationally representative
data set to measure the increase in labeling over the time period prior to and directly following the implementation
of NLEA. However, due to the collection methods of FLAPS, the estimated change in the prevalence of labeling
due to NLEA may be over- or understated. Further information about FLAPS can be found in Brecher et al. (2000)
(Ref. 29) and Ferguson (2013) (Ref. 30).
76
Nutrition Facts/Serving Size Combined Final RIA
value the measured change in nutrient intake. We refer to these estimates as the WTP estimates.
In Table 12, we summarize these estimates and convert them to 2014 dollars. Abaluck (2011)
finds that NLEA led to an average increase in consumer welfare of $61 (2014$) per year per
label user.
Table 12 - Annual Welfare Gains Per Person Based on Abaluck (2011)
Annual Welfare Gain Per Person
1990$
NLEA
WTP
2014$
Model 1
Model 2
Model 1
Model 2
Mean
$41
$33
$67
$55
$61
Notes: These estimates can be found in Table 7 of Abaluck (2011). Models 1 and 2 are different specifications of
Abaluck’s model of WTP for nutrient content. Model 1 estimates the WTP for calories, sodium, and cholesterol
and Model 2 disaggregates calories into protein, non-fiber carbohydrates (e.g., sugars), dietary fiber, and total fat.
The annual welfare gains per person presented here are the same for children and adults.
The estimates generated by Abaluck (2011) (that is, the mean WTP estimates of welfare
gains from NLEA, $61) represent welfare gains and, thus, are of the appropriate form for
estimating the benefits of the final rules. However, we must calibrate these estimates to the
expected effects of the final rules. For example, the changes to the Nutrition and Supplement
Facts labels mandated by the final rules would increase the available nutrition information on
food labels, but by an amount much smaller than the changes brought about by the 1993 rules
that implemented NLEA.32
We can use the estimated welfare gains from NLEA in Table 12 as the basis of the
illustrative calculations of benefits from the final rules. The WTP estimates tie directly to
revealed preference data and represent a plausible lower bound for the welfare gains from
NLEA.
Effect of Final Rules Relative to the 1993 Rules that Implemented NLEA (“s1”)
32
E.g., “Food Labeling: Mandatory Status of Nutrition Labeling and Nutrient Content Revision, Format for
Nutrition Label,” 58 FR 2079 (Jan. 6, 1993); “Food Labeling; Reference Daily Intakes and Daily Reference
Values,” 58 FR 2206 (Jan. 6, 1993); and “Food Labeling: Serving Sizes,” 58 FR 2229 (Jan. 6, 1993).
77
Nutrition Facts/Serving Size Combined Final RIA
We cannot use the estimated welfare gains associated with the 1993 rules that
implemented NLEA in Abaluck (2011) for the estimation of the benefits of the final rules
directly because the NLEA estimate would overstate the incremental effects of the nutrition
labeling final rules. This is because the 1993 rules that implemented NLEA added nutrition
labels to previously unlabeled products and revised nutrition labels on previously labeled
products, whereas the nutrition labeling final rules involve only modifications to the existing
label. That is, the estimated welfare gain from Abaluck (2011) represents the estimated welfare
gain that resulted from generally no requirement to provide a Nutrition Facts panel on packaged
foods to the requirement to provide a Nutrition Facts panel, and the estimated welfare gain
resulting from the final rules would represent impacts that would result from changes to existing
requirements.
The final rules represent a significant modification of existing Nutrition and Supplement
Facts labels. Changes to Nutrition and Supplement Facts labels mandated by the final rules are
summarized in the preambles to the final Nutrition Facts Label and final Serving Size rules
published in the Federal Register.
The Nutrition Facts Label and Serving Size final rules would change approximately 92
percent of UPCs.33 Of these UPCs, all would be required to change the first or only (as
applicable) column of nutrition information. Under the combined final rules, approximately 6
percent of UPCs will be required to include a second column of nutrition information.
We lack direct evidence with which to scale Abaluck’s estimates of the effect of NLEA
in a manner that precisely reflects the impacts of other changes in nutrition labels; therefore,
throughout this analysis we incorporate uncertainty ranges so as to assess the sensitivity of
33
We exclude dietary supplements from this percentage estimate, as we believe that the benefits from the final rules
are overwhelmingly attributable to changes in the consumption of conventional food.
78
Nutrition Facts/Serving Size Combined Final RIA
benefits estimates to key analytic inputs (especially the factor used to scale the NLEA results for
the present final rules). We estimate that the change in label content associated with the
inclusion of the first or only (as applicable) column of nutrition information is uniformly
distributed between 16 percent and 60 percent, with a mean of 38 percent (= [16 + 60] / 2) and
that the change in label content associated with the addition of a second column of nutrition
information is uniformly distributed between 0 percent and 50 percent, with a mean of 25 percent
(= [0 + 50] / 2). Weighting these amounts by the share of total UPCs that SCL and DCL
products represent (92 percent) and the share of total UPCs that DCL products represent (6
percent), respectively, produces a mean effect of the final rules of 36.5 percent (= [0.38 × 0.92] +
[0.25 × 0.06]) of the estimated welfare gains associated with the 1993 rules that implemented
NLEA in Abaluck (2011).
Given the assumptions presented above, we will scale the estimated welfare gain
associated with the 1993 rules that implemented NLEA in Abaluck (2011) by a range of values
(denoted s1) to capture the fact that the final rules will result in a welfare gain somewhat lower
than that which resulted after implementation of the 1993 nutrition labeling rules. We use a
uniform distribution with a minimum of 0, a maximum of 0.365, and a mean of 0.183 (= [0 +
0.365] / 2) for s1, which implies that the effect of the final rules lies between 0 and 36.5 percent
of the estimated welfare gains associated with the 1993 rules that implemented NLEA in
Abaluck (2011).34, 35 Our choice of zero as a lower bound is intended to reflect the uncertainty
34
We acknowledge that s1 is based entirely on quantity, and not quality, of change, the implication of which is that
each label change receives equal weight. We take such an approach because, due to data and other limitations, we
are unable to estimate quality weights associated with each label change. We previously asked for comment on our
s1 estimate, but received none that provided a basis upon which to adjust it.
35
At the means of the benefits model parameters, the breakeven estimate of s 1, defined as the value of s1 that
produces an estimate of benefits that equals the costs of the final rules, is 0.0262 using a 3 percent discount rate and
0.0371 using a 7 percent discount rate. Thus, an s1 of at least 3.72 percent would produce an estimate of benefits
that would exceed the estimated costs of the final rules.
79
Nutrition Facts/Serving Size Combined Final RIA
surrounding the impact of the final rules relative to the 1993 rules that implemented NLEA.
However, we do not expect that these rules will have zero impact on consumer behavior. In fact,
Abaluck (2011) demonstrates that new or improved information on the label of food products
can result in a substantial change to consumer behavior. We use zero as an absolute minimum to
capture the entire range of uncertainty, to allow for the possibility of even a very small effect.
Food Label Use (“USE”)
The estimated welfare gains associated with the 1993 rules that implemented NLEA in
Abaluck (2011) do not take into account label use.36 Data from the 2009-2010 NHANES show
that 79 percent of respondents used the Nutrition Facts label at least “Sometimes.” A change in
labeling regulations (and the educational messages that accompany the change) via the final rules
will likely increase label use beyond 79 percent, but we have no way of quantifying this likely
increase. Thus, following the implementation of the final rules we estimate that the prevalence
of label use would remain at its current level of 79 percent. Therefore, in the model, USE = 0.79.
USDA Regulated Food Labels (“USDA”)
The estimated welfare effects of the 1993 rules that implemented NLEA from Abaluck
(2011) may also capture the effect of labeling regulations simultaneously issued by USDA. The
USDA FSIS regulates the labeling of certain meat, poultry, and egg products. The USDA
labeling regulations for packaged foods mirror the FDA regulations almost exactly. We estimate
that approximately 16.4 percent of the calories an average American consumes daily come from
foods regulated by the USDA (Ref. 31). Therefore, to accommodate for the possibility that the
welfare effects capture the benefits of labeling on some USDA-regulated products, we scaled the
36
Because Abaluck (2011) generates a per capita WTP from revealed preference estimation there is no need to
explicitly address label usage. The estimates generated will only show changes in consumer purchase behavior if the
newly required label motivated some change. Because we are unable to observe pre and post data on purchases, we
must make some inferences about those consumers who currently do, and will continue to, use the label.
80
Nutrition Facts/Serving Size Combined Final RIA
benefits from nutrition labeling to reflect the fact that between zero and 32.8 percent of daily
calories come from products that would not be affected by the final rules. That is, USDA ~ U(0,
0.328), with a mean of 0.164 (= [0 + 0.328] / 2), or, equivalently, (1 – USDA) ~ U(0.672, 1),
with a mean of 0.836 (= [0.672 + 1] / 2).
Stream of Benefits (“Bt”)
The WTP estimates implicitly capture and reflect the fact that individuals discount the
benefits stemming from the effects of their current diet on their future health status. In other
words, the full measured benefits of the final rules are summarized in a value (WTP) that is
simultaneous with the timing when manufacturers comply with the final rules. We adjust the
annual stream of benefits from the final rules for the projected growth in the total population in
the United States over the next 20 years37 from the U.S. Census Bureau’s National Population
Projections.
Compliance Period (“ct”)
The welfare gains from the final rules illustrated according to the steps described above
reflect the full annual impact of the regulations. However, industry would need time to comply
with the regulations and reformulate products. Thus, it would take several years after the
publication date of the final rules, depending on the compliance date, for consumers to realize
the full annual welfare gains.
We estimate that the percentage of UPCs in compliance at time t,
denoted ct, equals 100 percent if time t falls on or after the compliance date or equals the
percentage of UPCs which can coordinate a scheduled label change with a required label change
(Ref. 1) if time t falls before the compliance date. Table 13 illustrates the relationship between
the compliance period and the percentage of UPCs which are able to coordinate a scheduled
37
We use 2015 – 2034 population data. Thus, benefits might be slightly underestimated due to population growth,
which is a conservative approach.
81
Nutrition Facts/Serving Size Combined Final RIA
label change with a required label change. For example, for a 2-year compliance period, ct = 0 in
the first year of the final rules, ct = 0.08 in the second year of the final rules, and ct = 1 in the
third year of the final rules through the twentieth year.
Table 13 – Percentage of UPCs Able to Coordinate a Scheduled Label Change with a
Required Label Change by Compliance Period
Compliance Period
% Who Can Coordinate
0 Months
0%
12 Months
8%
24 Months
67%
36 Months
80%
48 Months
98%
The following equation gives the formula for the PV of this stream of benefits, B,
discounted at a rate of r percent per year with, again, the percent of UPCs in compliance at time t
equal to ct.38
PV(B) =
∑ ct [Bt / (1 + r)t]
(2)
Benefits Illustrated
Using the @Risk software (Ref. 32), we carried out a simulation with 10,000 iterations to
illustratively calculate the PV of the benefits from the final rules over the next 20 years. Each
iteration of the simulation randomly draws a value for s139 and (1 – USDA), which each have a
uniform distribution with their respective minima and maxima, and calculates the PV of the
stream of benefits over the next 20 years using Equations (1) and (2). Table 14 displays the
results of this simulation.
Table 14 - Summary of Benefits Simulation Results for the Entire U.S. Population (in
billions of 2014$)
Low
Mean
High
38
We take t = 1 to be the first year of the rule, t = 2 to be the second year of the rule, and so on.
Each iteration of the simulation also randomly draws a value for the change in label content associated with SCL
and DCL products, each of which are components of the s1 calculation and are estimated to have a uniform
distribution with their respective minima and maxima.
39
82
Nutrition Facts/Serving Size Combined Final RIA
PV (3%)
PV (7%)
$2.8
$1.9
$33.1
$22.3
$77.7
$52.5
Annualized (3%)
Annualized (7%)
$0.2
$0.2
$2.2
$2.1
$5.2
$5.0
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of this
analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Estimates reflect total U.S. population
(children and adults). Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7
percent annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the
discount rate to the power of the year (t = 1 through t = 20).
Depending on the values the parameters s1 and (1 – USDA) take, and based on the WTP
welfare gain estimate and other modeling assumptions, the PV of the stream of illustratively
calculated benefits for the total U.S. population over the next 20 years ranges from $2.8 billion to
$77.7 billion with a mean estimate of $33.1 billion, using a 3 percent discount rate, and from
$1.9 billion to $52.5 billion with a mean estimate of $22.3 billion, using a 7 percent discount rate
(2014$). Table 14 also presents annualized benefits, using both a 3 percent and 7 percent
discount rate.
Other Sources of Benefits
Reformulation
Manufacturers may reformulate to reduce the amounts of added sugars because the
Nutrition Facts Label rule will require manufacturers to list added sugars and the percent DV for
added sugars on the Nutrition or Supplement Facts label (increased visibility). However, limited
data are available to quantify the effect of reformulation to reduce added sugars on measures of
health. Thus, we do not quantify the potential benefits from reformulation to reduce added
sugars.
Under the Nutrition Facts Label rule, the DV for vitamin B12 is being reduced from 6.0
mcg to 2.4 mcg. This new DV may incent manufacturers who currently fortify their products
with vitamin B12 to reformulate to reduce B12 amounts. The risk of developing a vitamin B12
83
Nutrition Facts/Serving Size Combined Final RIA
deficiency increases with age, with the elderly more likely to develop a vitamin B12 deficiency
because they are at risk for both malabsorption and malnutrition (Ref. 33). Because vitamin B12
deficiency may contribute to certain health problems (Refs. 33 – 34), reformulation to reduce the
amount of vitamin B12 per serving may potentially have a negative health impact on a small
portion of the elderly population. However, there is limited data to quantify the effect of vitamin
B12 reformulation on health outcomes. Thus, we do not quantify the potential negative effect on
benefits from vitamin B12 reformulation. Such an effect on benefits would likely be very small,
though, for two reasons. First, only approximately 3.2 percent of persons aged 51 years or older
currently have a vitamin B12 deficiency (Ref. 33), which translates into an even lower percentage
of the United States population as a whole. Second, it is unlikely that vitamin B12 deficient
individuals rely solely on fortified foods such as cereal to supplement their diet with vitamin B12,
but rather rely on dietary supplements in pill form, the majority of which contain levels of
vitamin B12 well in excess of 100 percent of the DV and, thus, are unlikely themselves to
reformulate.
Finally, manufacturers may also reformulate in response to certain amendments to the
RACC regulations, certain DV changes, and changes in the definition of dietary fiber so that they
can continue to make certain nutrient content claims or health claims authorized under the FD&C
Act. Such reformulations may generate benefits to society. For example, the Serving Size final
rule increases the RACC for ice cream from ½ cup to ⅔ cup. Ice cream products that currently
make a low-fat claim must, among other requirements, contain 3 g or less of total fat per RACC.
Thus, products which currently make a low-fat claim and contain, for example, 3 g of fat per ½
cup contain, proportionally, 4 g of fat per ⅔ cup. Therefore, under the Serving Size final rule,
manufacturers of such products would need to lower the amount of total fat in their ice cream to
84
Nutrition Facts/Serving Size Combined Final RIA
3 g or less per ⅔ cup in order to continue to make a low-fat claim as a result of the Serving Size
final rule. Reductions in fat, especially saturated fat, are associated with a reduced risk of
cardiovascular disease (Ref. 35). In addition, under the final rules, only an isolated or synthetic
non-digestible carbohydrate which has a physiological effect that is beneficial to human health,
and listed as a dietary fiber in the definition of “dietary fiber,” may be declared on the Nutrition
or Supplement Facts label. Thus, under the Nutrition Facts final rule, manufacturers of products
which both contain a carbohydrate that has not been shown to have beneficial physiological
effect to human health, and which make fiber-related health or nutrient content claims that they
wish to keep making, will need to reformulate their products so as to contain amounts of dietary
fiber that have been shown to have beneficial physiological effects to human health (the actual
amount of such dietary fiber called for depends on the claim in question). Limited data,
however, are available to quantify the effects of such reformulations on measures of health.
Thus, we do not quantify the potential benefits from such reformulations.
Benefits of Reduced Morbidity
Changes in label use could reduce the risk of morbidity and prolong life to the extent
consumers use such changed label information to maintain healthy dietary practices. Research
has demonstrated links between diet and excess body weight (overweight and obesity), CVD
(which includes CHD, heart attack, stroke and high blood pressure), type 2 diabetes (or noninsulin dependent diabetes mellitus), some cancers, cognitive decline, osteoporosis, and dental
disease (Ref. 23; Refs. 36 - 38). Each of these conditions may cause some degree of disability,
impairment, discomfort, and anxiety among sufferers, and may also involve a significant amount
of time for daily treatment or management. However, due to data limitations, we are unable to
85
Nutrition Facts/Serving Size Combined Final RIA
directly quantify the effect of the final rules on reduced morbidity. However, such effects are
implicitly captured in our WTP estimates.
Medical Costs
We have not fully quantified the effect the final rules may have on medical spending in
this analysis because, due to data limitations, we are unable to directly estimate the effect of the
final rules on the incidence of diet-related disease. If, however, the requirements in the final
rules help improve diet quality and reduce the prevalence of chronic diet-related diseases, then
consumers and other payers would spend less on medical treatment for these diseases (the
portion consumers spend on themselves is implicitly included in our WTP estimates; any portion
borne by the rest of society has not been quantified). For example, preventing obesity, and
avoiding the increased medical costs associated with it, could generate significant long-run
savings for publicly funded programs like Medicare, Medicaid, and Social Security Insurance
(disability). One report estimates that preventing obesity would save $684 (2014$) per Medicaid
recipient per year and $2,025 (2014$) per Medicare recipient per year (Ref. 39).
The medical costs associated with diet-related diseases and conditions increase medical
spending for persons afflicted with these diseases and conditions. These diseases and conditions,
however, also reduce life expectancy. Persons with longer life expectancy incur medical costs
for more years, and may incur very large expenses. Therefore, we do not know in advance if the
present value of lifetime medical expenses is on average higher for individuals with diet-related
conditions.
Offsetting Utility Loss
As with morbidity effects and medical savings borne by consumers themselves, there is a
potential impact of the final rules that is theoretically implicitly captured in the WTP estimates:
86
Nutrition Facts/Serving Size Combined Final RIA
offsetting loss of utility (which is an economics term sometimes described as enjoyment,
usefulness or satisfaction). Consumers may prefer the taste of relatively less nutritious foods, so
when they switch consumption to other products or reformulated versions of the same products,
utility loss will offset some portion of their health and longevity gains. Similarly, healthy food
may require greater preparation time than unhealthy food, in which case there would be a time
cost attributable to the final rules.
Additional Sensitivity Analysis
The primary source of variation in the benefits presented in this analysis stems from the
uncertainty surrounding the parameter s1, which translates the welfare gain estimates associated
with the 1993 rules that implemented NLEA in Abaluck (2011) into the illustrated welfare gains
from the final rules, to include the change in label content associated with SCL and DCL
products, each of which are parameters of the s1 calculation, and the parameter (1 – USDA),
which ensures that we are only capturing the benefits of labeling associated with FDA-regulated
products. As described in detail above, we perform simulations that reflect an assumption that
each of these parameters takes a range of equally likely values (that each parameter has a
uniform distribution over some range).
Another source of uncertainty in the benefits comes from the fact that Abaluck (2011)
generates his WTP estimates of welfare gains from a data set that contains only women of 19 to
50 years of age who are the main meal planner in their household from the DHKS and CSFII.
Using an identical methodology to that which is used to produce the illustrative benefits results
in Table 14, we illustrate the benefits from the final rules for women only.40 Table 15 contains
40
Given the fairly broad age range of women in Abaluck’s sample, we feel that it is reasonable to include all women
in our sensitivity analysis here. However, using the same female subsample used by Abaluck (women aged 19 to
87
Nutrition Facts/Serving Size Combined Final RIA
our illustrations of the benefits from the final rules if only women received benefits. We
illustratively calculate that the PV of the stream of benefits for women ranges from $1.5 billion
to $39.6 billion with a mean of $16.8 billion, using a 3 percent discount rate, and from $1.0
billion to $26.6 billion with a mean of $11.4 billion, using a 7 percent discount rate (2014$).
Table 15 also presents annualized benefits, at both a 3 percent and a 7 percent discount rate.
Table 15 - Summary of Benefits Simulation Results for Women in the U.S. (in billions of
2014$)
Low
Mean
High
PV (3%)
$1.5
$16.8
$39.4
PV (7%)
$1.0
$11.3
$26.6
Annualized (3%)
Annualized (7%)
$0.1
$0.1
$1.1
$1.1
$2.6
$2.5
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of this
analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Estimates reflect total female population in
the U.S. Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
A third source of uncertainty has to do with the Abaluck (2011) annual welfare gains per
person, which were previously summarized in Table 12 of this RIA. Underlying these estimates
is the full set of Model 1 and Model 2 estimates presented in Table 5 of Abaluck (2011) of
peoples’ WTP for NLEA nutrient information. All of the WTP coefficients estimated in Model 1
are statistically significant at conventional levels (10 percent or better), and some of the WTP
coefficients estimated in Model 2 are statistically significant at conventional levels.41 Using an
identical methodology to that which is used to produce the benefits results in Table 14, we
50) produces mean benefits of $6.8 billion (2014$) using a 3 percent discount rate and $4.6 billion (2014$) using a 7
percent discount rate.
41
We expect that, all else equal, when individuals learn that a food product has more cholesterol or sodium than
they previously thought, they will not increase their consumption of the food product. Thus, a one-tailed test of the
type Ho: Coefficient = 0, Ha: Coefficient < 0 is used to assess the statistical significance of the cholesterol and
sodium coefficients in Models 1 and 2. Two-tailed tests are used to assess the statistical significance of the
remaining coefficients in Models 1 and 2.
88
Nutrition Facts/Serving Size Combined Final RIA
illustrate the benefits using the Model 1 annual welfare gain per person of $67 illustrated in
Table 12 and, being extremely conservative, a Model 2 annual welfare gain per person of $0,
which produces an average annual welfare gain per person of $34 (2014$). Table 16 contains
our illustration of the benefits using an average annual welfare gain per person of $34 instead of
the $61 that was used in the primary illustration (2014$). We illustratively calculate that the PV
of the stream of benefits in this case ranges from $1.6 billion to $43.3 billion with a mean of
$18.4 billion, using a 3 percent discount rate, and from $1.1 billion to $29.0 billion with a mean
of $12.4 billion, using a 7 percent discount rate (2014$). Table 16 also presents annualized
benefits, at both a 3 percent and a 7 percent discount rate.
Table 16 - Summary Benefits Simulation Results for the Entire U.S. Population (in billions
of 2014$)
Low
Mean
High
PV (3%)
$1.6
$18.4
$43.3
PV (7%)
$1.1
$12.4
$29.0
Annualized (3%)
Annualized (7%)
$0.1
$0.1
$1.2
$1.2
$2.9
$2.7
Notes: Compliance period is 36 months for small businesses and 24 months for large businesses. For purposes of this
analysis, we consider a small business to be a business with annual food sales of less than $10 million, and a large
business to be a business with annual food sales of $10 million or more. Estimates reflect total U.S. population
(children and adults) and rely on an average annual welfare gain per person of $34 (2014$). Annualized Amount =
Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The
annualizing factors are calculated by summing the inverse of 1 plus the discount rate to the power of the year (t = 1
through t = 20).
3. Option 3: The final rules, but with a 2-year compliance period for all manufacturers
a. Costs
i.) Relabeling Costs
A 2-year compliance period for all manufacturers would give small manufacturers less
time to comply with the final rules and would thus, as compared with the 3-year compliance
period, increase the labeling costs associated with the final rules. This increase in costs would
89
Nutrition Facts/Serving Size Combined Final RIA
result from small firms having to undertake a greater amount of labeling changes uncoordinated
with regularly scheduled changes.
Label costs under this scenario are summarized in Table 17. With the same compliance
date, the labeling cost estimates of the final rules under a 2-year compliance period range from
$1,421 million to $4,621 million with a mean estimate of $2,748 million, using a 3 percent
discount rate, and from $1,347 million to $4,412 million with a mean estimate of $2,599 million,
using a 7 percent discount rate (2014$). Table 17 also presents annualized labeling costs, using
both a 3 percent and a 7 percent discount rate.
Table 17 - Summary of Relabeling Costs – 24 Month Compliance Period (in millions of
2014$)
Source of
Label
Change
Final Rule
Nutrition Facts
Label (NFL)
Conventional
Food
Conventional
Food
Conventional
Food
Dietary
Supplements
Dietary
Supplements
Dietary
Supplements
Various Changes to
the NFL
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Dietary Fiber
Definition
Various Changes to
the NFL
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Dietary Fiber
Definition
Analytical Testing Costs
Type of
Label
Change
Different Compliance Dates
Same Compliance Date
Low
Mean
High
Low
Mean
High
Minor
$620
$1,362
$2,569
$327
$720
$1,358
Major
$3
$7
$12
$3
$7
$12
Major
$8
$17
$31
$8
$17
$31
Minor
$117
$255
$472
$117
$255
$472
Major
$0.3
$1
$1
$0.3
$1
$1
Major
$1
$1
$2
$1
$1
$2
$322
$1,071
$421
$2,064
$530
$3,617
$322
$778
$421
$1,422
$530
$2,406
Minor
$105
$235
$448
$105
$235
$448
Minor
$98
$213
$399
$98
$213
$399
Major
$78
$165
$306
$78
$165
$306
Total NFL
Serving Size (SS)
Conventional
Food
Conventional
Food
Conventional
Food
Changes Due to
RACC
Amendments
Change in
Definition of
Single-Serving
Container
DCL
90
Nutrition Facts/Serving Size Combined Final RIA
Conventional
Food
Claim Removal
Related to RACC
Change
First Time Nutrition
Facts Panel Due To
RACC Proposals
Package Size and
Claim Effects of
Change of
Definition of
Single-Serving
Container
Package Size and
Claim Effects of
Change of
Definition of
Single-Serving
Container
Major
$2
$3
$6
$2
$3
$6
Major
$80
$150
$260
$80
$150
$260
Extensive
(3%)
$280
$560
$796
$280
$560
$796
Extensive
(7%)
$206
$411
$587
$206
$411
$587
Total SS (3%)
$643
$1,326
$2,215
$643
$1,326
$2,215
Total SS (7%)
$569
$1,177
$2,006
$569
$1,177
$2,006
$1,714
$3,390
$5,832
$1,421
$2,748
$4,621
$1,640
$3,241
$5,623
$1,347
$2,599
$4,412
$115
$228
$392
$95
$185
$311
$155
$306
$531
$127
$245
$417
Conventional
Food
Conventional
Food
Conventional
Food
TOTAL NFL +
SS
(3%)
TOTAL NFL +
SS
(7%)
Annualized
(3%)
Annualized
(7%)
Notes: Analytical testing costs associated with the mandatory declarations of potassium and vitamin D are
conservatively estimated across all conventional food and dietary supplement UPCs using the FDA Labeling Cost
Model. The cost of an extensive label change includes recurring annual costs associated with increased package
sizes. Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
Related recordkeeping costs are estimated to be $21.4 million at a 3 percent discount rate
and $21.3 million at a 7 percent discount rate (2014$). Related to the new definition of dietary
fiber, study costs are estimated to be roughly $4 million (2014$), and additional labeling costs42
are estimated to range from $176 million to $1,101 million with a mean estimate of $430 million
42
The additional labeling change represents a minor labeling change (Ref. 1). Labeling costs are estimated using
the “3 Months or Less” compliance period option in the FDA Labeling Cost Model (we conservatively estimate that
manufacturers, upon receiving post-compliance period approval of their dietary fiber citizen petition or health claim
petition, would relabel as quickly as possible).
91
Nutrition Facts/Serving Size Combined Final RIA
(2014$). These costs are summarized below in Table 18. Table 18 also presents annualized
costs, using both a 3 percent and a 7 percent discount rate.
Table 18 – Summary of Recordkeeping Costs and Study and Additional Labeling Costs
Related to the New Definition of Dietary Fiber (in millions of 2014$)
Low
Mean
High
Recordkeeping Costs
$21
$21
$21
Study Costs Related to New Definition of
$4
$4
$4
Dietary Fiber
Additional Labeling Costs Related to New
$176
$430
$1,101
Definition of Dietary Fiber
$201
$455
$1,126
Total
Annualized (3%)
$14
$31
$76
Annualized (7%)
$19
$43
$106
Notes: These costs do not depend on compliance period. Annualized Amount = Amount / Annualizing Factor. 3
percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
Labeling costs associated with future UPC growth are estimated to range from $66
million to $198 million, with a mean estimate of $135 million, using a 3 percent discount rate,
and from $35 million to $107 million, with a mean estimate of $72 million, using a 7 percent
discount rate (2014$). These costs are summarized below in Table 19. Table 19 also presents
annualized costs, using both a 3 percent and a 7 percent discount rate.
Table 19 – Summary of Labeling Costs Associated With Future UPC Growth (in millions of
2014$)
Type of
Label
Low
Mean
High
Change
Present Value (3%)
First Time Nutrition Facts Panel Due
Major
$6
$14
$26
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$60
$121
$172
Serving Container
$66
$135
$198
Total
Present Value (7%)
First Time Nutrition Facts Panel Due
Major
$4
$10
$18
to RACC
Package Size and Claim Effects of
Extensive
$31
$62
$89
Change of Definition of Single92
Nutrition Facts/Serving Size Combined Final RIA
Serving Container
Total
Annualized (3%)
Annualized (7%)
$35
$4
$3
$72
$9
$7
$107
$13
$10
Notes: Annualized amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
ii.) Reformulation Costs
Reformulation costs under a 2-year compliance period are summarized in Table 20. A 2year compliance period gives small manufacturers less time to reformulate their products and
would thus, as compared with the 3-year compliance period, increase reformulation costs
associated with the final rules (as discussed earlier in the RIA in describing the FDA
Reformulation Cost Model, reformulation costs are higher for shorter compliance periods
because if the compliance period is short, manufacturers will incur increased costs for overtime
labor, additional staffing, and rush charges with vendors and suppliers). Total costs of
reformulation due to requirements affecting eligibility to make health claims and nutrient content
claims and voluntary reformulation due to increased visibility of added sugars and B12 range
from $655 million to $2,698 million with a mean estimate of $1,454 million (2014$). Table 20
also presents annualized reformulation costs, using both a 3 percent and a 7 percent discount
rate.
Table 20 - Summary of Reformulation Costs – 24 Month Compliance Period (in millions of
2014$)
Category
Low
Mean
High
Claims
$157
$348
$646
Added Sugars
$298
$662
$1,228
B12
$200
$444
$824
$655
$1,454
$2,698
Total
Annualized (3%)
$44
$98
$181
Annualized (7%)
$62
$137
$255
Notes: Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
93
Nutrition Facts/Serving Size Combined Final RIA
b. Benefits
We illustratively calculate that under a 2-year compliance period the PV of the stream of
benefits for the total U.S. population over the next 20 years ranges from $2.8 billion to $77.8
billion with a mean of $33.2 billion, using a 3 percent discount rate, and from $1.9 billion to
$52.6 billion with a mean of $22.4 billion, using a 7 percent discount rate, as illustrated in Table
21 (2014$). Table 21 also presents annualized benefits, using both a 3 percent and a 7 percent
discount rate.
Table 21 - Summary of Benefits Simulation Results for Entire U.S. Population - 24 Month
Compliance Period (in billions of 2014$)
Low
Mean
High
PV (3%)
$2.8
$33.2
$77.8
PV (7%)
$1.9
$22.4
$52.6
Annualized (3%)
Annualized (7%)
$0.2
$0.2
$2.2
$2.1
$5.2
$5.0
Notes: Estimates reflect total U.S. population (children and adults). Annualized Amount = Amount / Annualizing
Factor. 3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are
calculated by summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
4. Option 4: The final rules, but with a 3-year compliance period for all manufacturers
a. Costs
i.) Relabeling Costs
A 3-year compliance period would give large manufacturers more time to comply with
the final rules and would thus decrease the labeling costs associated with the final rules as
compared with a 2-year compliance period. This decrease in costs would result from large firms
having to undertake fewer labeling changes uncoordinated with their regularly scheduled
labeling changes.
Label costs under this scenario are summarized in Table 22. With the same compliance
date, the labeling cost estimates of the final rules under a 3-year compliance period range from
94
Nutrition Facts/Serving Size Combined Final RIA
$1,100 million to $3,777 million with a mean estimate of $2,210 million, using a 3 percent
discount rate, and from $1,026 million to $3,568 million with a mean estimate of $2,060 million,
using a 7 percent discount rate (2014$). Table 22 also presents annualized labeling costs, using
both a 3 percent and a 7 percent discount rate.
Table 22 - Summary of Relabeling Costs – 36 Month Compliance Period (in millions of
2014$)
Final Rule
Source of
Label
Change
Type of
Label
Change
Different Compliance Dates
Same Compliance Date
Low
Mean
High
Low
Mean
High
Minor
$359
$925
$1,881
$190
$489
$994
Major
$2
$5
$9
$2
$5
$9
Major
$5
$12
$23
$5
$12
$23
Minor
$92
$206
$386
$92
$206
$386
Major
$0.3
$1
$1
$0.3
$1
$1
Major
$0.4
$1
$2
$0.4
$1
$2
$322
$421
$530
$322
$421
$530
$781
$1,571
$2,832
$612
$1,135
$1,945
Minor
$62
$163
$337
$62
$163
$337
Minor
$55
$142
$290
$55
$142
$290
Major
$49
$117
$230
$49
$117
$230
Major
$1
$2
$4
$1
$2
$4
Major
$42
$92
$176
$42
$92
$176
Extensive
(3%)
$279
$559
$795
$279
$559
$795
Nutrition Facts
Label (NFL)
Various Changes
to the NFL
Claim Removal
Conventional Food
Related to DV
Change
Claim Removal
Related to New
Conventional Food
Dietary Fiber
Definition
Dietary
Various Changes
Supplements
to the NFL
Claim Removal
Dietary
Related to DV
Supplements
Change
Claim Removal
Dietary
Related to New
Supplements
Dietary Fiber
Definition
Analytical Testing Costs
Conventional Food
Total NFL
Serving Size (SS)
Conventional Food
Conventional Food
Conventional Food
Conventional Food
Conventional Food
Conventional Food
Changes Due to
RACC
Amendments
Change in
Definition of
Single-Serving
Container
DCL
Claim Removal
Related to RACC
Change
First Time
Nutrition Facts
Panel Due To
RACC Change
Package Size and
Claim Effects of
Definition of
Single-Serving
95
Nutrition Facts/Serving Size Combined Final RIA
Container
Conventional Food
Package Size and
Claim Effects of
Definition of
Single-Serving
Container
Extensive
(7%)
$205
$409
$586
$205
$409
$586
Total SS (3%)
$488
$1,075
$1,832
$488
$1,075
$1,832
Total SS (7%)
$414
$925
$1,623
$414
$925
$1,623
$1,269
$2,646
$4,664
$1,100
$2,210
$3,777
$1,195
$2,496
$4,455
$1,026
$2,060
$3,568
$85
$178
$313
$74
$149
$254
$113
$236
$421
$97
$195
$337
TOTAL NFL + SS
(3%)
TOTAL NFL + SS
(7%)
Annualized
(3%)
Annualized
(7%)
Notes: Analytical testing costs associated with the mandatory declarations of potassium and vitamin D are
conservatively estimated across all conventional food and dietary supplement UPCs using the FDA Labeling Cost
Model. The cost of an extensive label change includes recurring annual costs associated with increased package
sizes. Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
Related recordkeeping costs are estimated to be $21.4 million at a 3 percent discount rate
and $21.3 million at a 7 percent discount rate (2014$). Related to the new definition of dietary
fiber, study costs are estimated to be roughly $4 million (2014$), and additional labeling costs43
are estimated to range from $176 million to $1,101 million with a mean estimate of $430 million
(2014$). These costs are summarized below in Table 23. Table 23 also presents annualized
costs, using both a 3 percent and a 7 percent discount rate.
Table 23 – Summary of Recordkeeping Costs and Study and Additional Labeling Costs
Related to the New Definition of Dietary Fiber (in millions of 2014$)
Low
Mean
High
Recordkeeping Costs
$21
$21
$21
Study Costs Related to New Definition of
$4
$4
$4
Dietary Fiber
Additional Labeling Costs Related to New
$176
$430
$1,101
Definition of Dietary Fiber
43
The additional labeling change represents a minor labeling change (Ref. 1). Labeling costs are estimated using
the “3 Months or Less” compliance period option in the FDA Labeling Cost Model (we conservatively estimate that
manufacturers, upon receiving post-compliance period approval of their dietary fiber citizen petition or health claim
petition, would relabel as quickly as possible).
96
Nutrition Facts/Serving Size Combined Final RIA
$201
$14
$19
Total
Annualized (3%)
Annualized (7%)
$455
$31
$43
$1,126
$76
$106
Notes: These costs do not depend on compliance period. Annualized Amount = Amount / Annualizing Factor. 3
percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
Labeling costs associated with future UPC growth are estimated to range from $66
million to $198 million, with a mean estimate of $135 million, using a 3 percent discount rate,
and from $35 million to $107 million, with a mean estimate of $72 million, using a 7 percent
discount rate (2014$). These costs are summarized below in Table 24. Table 24 also presents
annualized costs, using both a 3 percent and a 7 percent discount rate.
Table 24 – Summary of Labeling Costs Associated With Future UPC Growth (in millions of
2014$)
Type of
Label
Low
Mean
High
Change
Present Value (3%)
First Time Nutrition Facts Panel Due
Major
$6
$14
$26
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$60
$121
$172
Serving Container
$66
$135
$198
Total
Present Value (7%)
First Time Nutrition Facts Panel Due
Major
$4
$10
$18
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$31
$62
$89
Serving Container
$35
$72
$107
Total
Annualized (3%)
$4
$9
$13
Annualized (7%)
$3
$7
$10
Notes: Annualized amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
ii.) Reformulation Costs
Reformulation costs under a 3-year compliance period are summarized in Table 25.
Compared with a 2-year compliance period, a 3-year compliance period would decrease
97
Nutrition Facts/Serving Size Combined Final RIA
reformulation costs associated with the final rules (as discussed earlier in the RIA in describing
the FDA Reformulation Cost Model, reformulation costs are higher for shorter compliance
periods because if the compliance period is short, manufacturers will incur increased costs for
overtime labor, additional staffing, and rush charges with vendors and suppliers). Total costs of
reformulation due to requirements affecting eligibility to make health claims and nutrient content
claims and voluntary reformulation due to increased visibility of added sugars and B12 range
from $575 million to $2,377 million with a mean estimate of $1,280 million (2014$). Table 25
also presents annualized reformulation costs, using both a 3 percent and a 7 percent discount
rate.
Table 25 - Summary of Reformulation Costs – 36 Month Compliance Period (in millions of
2014$)
Category
Low
Mean
High
Claims
$137
$305
$566
Added Sugars
$261
$580
$1,078
B12
$177
$395
$733
$575
$1,280
$2,377
Total
Annualized (3%)
$39
$86
$160
Annualized (7%)
$54
$121
$224
Notes: Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
b. Benefits
We illustratively calculate that under a 3-year compliance period the PV of the stream of
benefits for the total U.S. population over the next 20 years ranges from $2.8 billion to $76.1
billion with a mean of $32.4 billion, using a 3 percent discount rate, and from $1.9 billion to
$51.1 billion with a mean of $21.7 billion, using a 7 percent discount rate, as illustrated in Table
26 (2014$). Table 26 also presents annualized benefits, using both a 3 percent and a 7 percent
discount rate.
Table 26 - Summary of Benefits Simulation Results for Entire U.S. Population - 36 Month
98
Nutrition Facts/Serving Size Combined Final RIA
Compliance Period (in billions of 2014$)
PV (3%)
PV (7%)
Low
$2.8
$1.9
Mean
$32.4
$21.7
High
$76.1
$51.1
Annualized (3%)
Annualized (7%)
$0.2
$0.2
$2.2
$2.0
$5.1
$4.8
Notes: Estimates reflect total U.S. population (children and adults). Annualized Amount = Amount / Annualizing
Factor. 3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are
calculated by summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
5. Option 5: The final rules, but with a 4-year compliance period for all manufacturers
a. Costs
i.) Relabeling Costs
A 4-year compliance period for all manufacturers would decrease the labeling costs
associated with the final rules, as compared to a 2- or 3-year compliance period. This decrease
in costs would result from fewer firms having to undertake labeling changes that are
uncoordinated with their regulatory scheduled changes.
Label costs under this scenario are summarized in Table 27. With the same compliance
date, it can be seen that the labeling cost estimates of the final rules under a 4-year compliance
period range from $800 million to $2,797 million with a mean estimate of $1,636 million, using
a 3 percent discount rate, and from $726 million to $2,588 million with a mean estimate of
$1,487 million, using a 7 percent discount rate (2014$). Table 27 also presents annualized
labeling costs, using both a 3 percent and a 7 percent discount rate.
Table 27 - Summary of Relabeling Costs – 48 Month Compliance Period (in millions of
2014$)
Final Rule
Source of
Label
Change
Type of
Label
Change
Different Compliance Dates
Same Compliance Date
Low
Mean
High
Low
Mean
High
$139
$494
$1,133
$73
$261
$599
Nutrition Facts
Label (NFL)
Conventional Food
Various Changes to
the NFL
Minor
99
Nutrition Facts/Serving Size Combined Final RIA
Claim Removal
Related to DV
Change
Claim Removal
Related to New
Conventional Food
Dietary Fiber
Definition
Dietary
Various Changes to
Supplements
the NFL
Claim Removal
Dietary
Related to DV
Supplements
Change
Claim Removal
Dietary
Related to New
Supplements
Dietary Fiber
Definition
Analytical Testing Costs
Conventional Food
Major
$1
$2
$5
$1
$2
$5
Major
$2
$6
$13
$2
$6
$13
Minor
$43
$107
$214
$43
$107
$214
Major
$0.1
$0.3
$1
$0.1
$0.3
$1
Major
$0.2
$0.4
$1
$0.2
$0.4
$1
$322
$421
$530
$322
$421
$507
$1,031
$1,897
$441
$798
$530
$1,36
3
Minor
$27
$96
$221
$27
$96
$221
Minor
$22
$78
$179
$22
$78
$179
Major
$18
$61
$138
$18
$61
$138
Major
$0.3
$1
$2
$0.3
$1
$2
Major
$13
$44
$99
$13
$44
$99
Extensive
(3%)
$279
$558
$795
$279
$558
$795
Extensive
(7%)
$205
$409
$586
$205
$409
$586
Total SS (3%)
$359
$838
$1,434
$359
$838
Total SS (7%)
$285
$689
$1,225
$285
$689
$866
$1,869
$3,331
$800
$1,636
$792
$1,720
$3,122
$726
$1,487
$58
$126
$224
$54
$110
$188
$75
$162
$295
$69
$140
$244
Total NFL
Serving Size (SS)
Conventional Food
Conventional Food
Conventional Food
Conventional Food
Conventional Food
Conventional Food
Conventional Food
TOTAL NFL + SS
(3%)
TOTAL NFL + SS
(7%)
Annualized
(3%)
Annualized
(7%)
Changes Due to
RACC
Amendments
Change in
Definition of
Single-Serving
Container
DCL
Claim Removal
Related to RACC
Change
First Time Nutrition
Facts Panel Due To
RACC Proposals
Package Size and
Claim Effects of
Change to
Definition of
Single-Serving
Container
Package Size and
Claim Effects
Change to
Definition of Single
Serving Container
Notes: Analytical testing costs associated with the mandatory declarations of potassium and vitamin D are
100
$1,43
4
$1,22
5
$2,79
7
$2,58
8
Nutrition Facts/Serving Size Combined Final RIA
conservatively estimated across all conventional food and dietary supplement UPCs using the FDA Labeling Cost
Model. The cost of an extensive label change includes recurring annual costs associated with increased package
sizes. Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
Related recordkeeping costs are estimated to be $21.4 million at a 3 percent discount rate
and $21.3 million at a 7 percent discount rate (2014$). Related to the new definition of dietary
fiber, study costs are estimated to be roughly $4 million (2014$) and additional labeling costs 44
are estimated to range from $176 million to $1,101 million with a mean estimate of $430 million
(2014$). These costs are summarized below in Table 28. Table 28 also presents annualized
costs, using both a 3 percent and a 7 percent discount rate
Table 28 – Summary of Recordkeeping Costs and Study and Additional Labeling Costs
Related to the New Definition of Dietary Fiber (in millions of 2014$)
Low
Mean
High
Recordkeeping Costs
$21
$21
$21
Study Costs Related to New Definition of
$4
$4
$4
Dietary Fiber
Additional Labeling Costs Related to New
$176
$430
$1,101
Definition of Dietary Fiber
$201
$455
$1,126
Total
Annualized (3%)
$14
$31
$76
Annualized (7%)
$19
$43
$106
Notes: These costs do not depend on compliance period. Annualized Amount = Amount / Annualizing Factor. 3
percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
Labeling costs associated with future UPC growth are estimated to range from $66
million to $198 million, with a mean estimate of $135 million, using a 3 percent discount rate,
and from $35 million to $107 million, with a mean estimate of $72 million, using a 7 percent
discount rate (2014$). These costs are summarized below in Table 29. Table 29 also presents
annualized costs, using both a 3 percent and a 7 percent discount rate.
44
The additional labeling change represents a minor labeling change (Ref. 1). Labeling costs are estimated using
the “3 Months or Less” compliance period option in the FDA Labeling Cost Model (we conservatively estimate that
manufacturers, upon receiving post-compliance period approval of their dietary fiber citizen petition or health claim
petition, would relabel as quickly as possible).
101
Nutrition Facts/Serving Size Combined Final RIA
Table 29 – Summary of Labeling Costs Associated With Future UPC Growth (in millions of
2014$)
Type of
Label
Low
Mean
High
Change
Present Value (3%)
First Time Nutrition Facts Panel Due
Major
$6
$14
$26
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$60
$121
$172
Serving Container
$66
$135
$198
Total
Present Value (7%)
First Time Nutrition Facts Panel Due
Major
$4
$10
$18
to RACC
Package Size and Claim Effects of
Change of Definition of SingleExtensive
$31
$62
$89
Serving Container
$35
$72
$107
Total
Annualized (3%)
$4
$9
$13
Annualized (7%)
$3
$7
$10
Notes: Annualized amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
ii.) Reformulation Costs
Reformulation costs under a 4-year compliance period are summarized in Table 30. As
compared to a 2- or 3-year compliance period, a 4-year compliance period would decrease
reformulation costs associated with the final rules (as discussed earlier in the RIA in describing
the FDA Reformulation Cost Model, reformulation costs are higher for shorter compliance
periods because if the compliance period is short, manufacturers will incur increased costs for
overtime labor, additional staffing, and rush charges with vendors and suppliers). Total costs of
reformulation due to requirements affecting eligibility to make health claims and nutrient content
claims and voluntary reformulation due to increased visibility of added sugars and B12 range
from $575 million to $2,377 million with a mean estimate of $1,280 million (2014$). Table 30
102
Nutrition Facts/Serving Size Combined Final RIA
also presents annualized reformulation costs, using both a 3 percent and a 7 percent discount
rate.
Table 30 - Summary of Reformulation Costs – 48 Month Compliance Period (in millions of
2014$)
Category
Low
Mean
High
Claims
$137
$305
$566
Added Sugars
$261
$580
$1,078
B12
$177
$395
$733
$575
$1,280
$2,377
Total
Annualized (3%)
$39
$86
$160
Annualized (7%)
$54
$121
$224
Notes: Annualized Amount = Amount / Annualizing Factor. 3 percent annualizing factor = 14.88. 7 percent
annualizing factor = 10.59. The annualizing factors are calculated by summing the inverse of 1 plus the discount rate
to the power of the year (t = 1 through t = 20).
b. Benefits
We illustratively calculate that under a 4-year compliance period the PV of the stream of
benefits for the total U.S. population over the next 20 years ranges from $2.7 billion to $75.1
billion with a mean estimate of $32.0 billion, using a 3 percent discount rate, and from $1.8
billion to $50.2 billion with a mean estimate of $21.3 billion, using a 7 percent discount rate, as
illustrated in Table 31 (2014$). Table 31 also presents annualized benefits, using both a 3
percent and a 7 percent discount rate.
Table 31 - Summary of Benefits Simulation Results for Entire U.S. Population - 48 Month
Compliance Period (in billions of 2014$)
Low
Mean
High
PV (3%)
$2.7
$32.0
$75.1
PV (7%)
$1.8
$21.3
$50.2
Annualized (3%)
Annualized (7%)
$0.2
$0.2
$2.2
$2.0
$5.0
$4.7
Notes: Estimates reflect total U.S. population (children and adults). Annualized Amount = Amount / Annualizing
Factor. 3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are
calculated by summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
6. Summary of Estimated Costs and Illustratively Calculated Benefits by Regulatory Option
103
Nutrition Facts/Serving Size Combined Final RIA
Estimated costs and illustratively calculated benefits, by regulatory option, are
summarized below in Table 32 and Table 33. Table 32 presents results assuming that the final
rules are enacted together. Table 33 presents results assuming that the final rules are enacted
separately.
Table 32 - Summary of Net Benefits by Regulatory Option If Final Rules Have Same
Compliance Date (in billions of 2014$)
PV
Option
1 - No New Federal
Regulatory Action
2 - Final Rules
3 - Final Rules - 2 Year
Compliance Period
4 - Final Rules - 3 Year
Compliance Period
5 - Final Rules - 4 Year
Compliance Period
Discount
Rate
Benefits
3%
Annualized
Costs
Net
Benefits
Benefits
Costs
Net
Benefits
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
7%
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
3%
$33.1
$4.8
$28.3
$2.2
$0.3
$1.9
7%
$22.3
$4.5
$17.8
$2.1
$0.4
$1.7
3%
$33.2
$4.8
$28.4
$2.2
$0.3
$1.9
7%
$22.4
$4.6
$17.8
$2.1
$0.4
$1.7
3%
$32.4
$4.1
$28.3
$2.2
$0.3
$1.9
7%
$21.7
$3.9
$17.8
$2.0
$0.4
$1.6
3%
$32.0
$3.5
$28.5
$2.2
$0.2
$2.0
7%
$21.3
$3.3
$18.0
$2.0
$0.3
$1.7
Notes: Benefits and costs reflect mean estimates. Costs include relabeling, recordkeeping, fiber study, additional
labeling, future UPC growth labeling, and reformulation costs. Annualized Amount = Amount / Annualizing Factor.
3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
Table 33 - Summary of Net Benefits by Regulatory Option If Final Rules Have Different
Compliance Dates (in billions of 2014$)
PV
Option
1 - No New Federal
Regulatory Action
2 - Final Rules
3 - Final Rules - 2 Year
Compliance Period
4 - Final Rules - 3 Year
Compliance Period
Discount
Rate
Benefits
3%
Annualized
Costs
Net
Benefits
Benefits
Costs
Net
Benefits
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
7%
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
3%
$33.1
$5.4
$27.7
$2.2
$0.4
$1.8
7%
$22.3
$5.2
$17.1
$2.1
$0.5
$1.6
3%
$33.2
$5.4
$27.8
$2.2
$0.4
$1.8
7%
$22.4
$5.2
$17.2
$2.1
$0.5
$1.6
3%
$32.4
$4.5
$27.9
$2.2
$0.3
$1.9
7%
$21.7
$4.3
$17.4
$2.0
$0.4
$1.6
104
Nutrition Facts/Serving Size Combined Final RIA
5 - Final Rules - 4 Year
Compliance Period
3%
$32.0
$3.7
$28.3
$2.2
$0.2
$2.0
7%
$21.3
$3.5
$17.8
$2.0
$0.3
$1.7
Notes: Benefits and costs reflect mean estimates. Costs include relabeling, recordkeeping, fiber study, additional
labeling, future UPC growth labeling, and reformulation costs. Annualized Amount = Amount / Annualizing Factor.
3 percent annualizing factor = 14.88. 7 percent annualizing factor = 10.59. The annualizing factors are calculated by
summing the inverse of 1 plus the discount rate to the power of the year (t = 1 through t = 20).
II. Regulatory Flexibility Analysis
A.
Introduction
We have examined the economic implications of the final rules as required by the
Regulatory Flexibility Act (5 U.S.C. §§ 601-612). If a rule has a significant economic impact on
a substantial number of small entities, the Regulatory Flexibility Act requires agencies to analyze
regulatory options that would lessen the economic effect of the rule on small entities consistent
with statutory objectives. We conclude that the final rules will have a significant economic
impact on a substantial number of small entities.
B.
Economic Effects on Small Entities
1. Number of Small Entities Affected
For the purposes of the Regulatory Flexibility Act analysis, we use the SBA’s definition
of a small business as it applies to the relevant economic sectors, in this case, North American
Industry Classification System (NAICS) 311, 312 and 325. SBA generally defines a small food
manufacturer as one that has 500 or fewer employees. Using Dun and Bradstreet data, we
estimate that roughly 99 percent of food manufacturers, or about 30,970 food manufacturers,
have 500 or fewer employees and are affected by the final rules.
2. Costs to Small Entities
The final rules will result in costs to small business. We cannot estimate the exact cost
per small entity because we do not know how many UPCs on average are owned by small
entities as defined using the SBA definition. However, we estimate that the final rules would
105
Nutrition Facts/Serving Size Combined Final RIA
cost roughly $6,096 per UPC (2014$). Therefore, a small firm owning one to three UPC’s
would incur a cost of between roughly $6,096 to $18,288 (2014$). This is likely a significant
overestimate, though, as the share of sales controlled by small firms is typically small, and much
smaller than the share of UPCs controlled by small firms (Ref. 2). The cost of the rule per entity
(including large firms) is approximately $152,031 (2014$). This number likely significantly
overstates the cost per small entity because the share of firms which are small businesses is
typically large and the share of sales controlled by small firms typically small (Ref. 2).
C.
Regulatory Options
The economic impact of the final rules on small entities is reduced in two ways. First, we
note that the targeted exemption from labeling that currently exists for some small businesses
will continue to be available. Currently, certain small businesses whose products do not sell
more than 100,000 units may apply for a 12-month exemption from certain nutrition labeling
requirements for that particular product, and the business has the option to reapply for a
continuation of this exemption (21 CFR 101.9(j)(18). Currently, there are about 3,000 small
businesses registered with FDA for a small business nutrition labeling exemption. On average we
grant labeling exemptions to approximately 10,000 products per year.
Second, in response to comments we received on the proposed rules,45 for firms earning
less than $10 million in annual food sales, which, based on Nielsen data we estimate to cover
approximately 95 percent of all food manufacturers and 48 percent of all food UPCs, we have
increased the compliance period from 2 years to 3 years. We estimate that this extended
compliance period reduces the cost per UPC for covered firms from roughly $640 per UPC to
about $545 per UPC, and the cost per covered firm from approximately $8,062 per firm to
45
A summary of and our responses to these comments are provided in the final rules published in the Federal
Register.
106
Nutrition Facts/Serving Size Combined Final RIA
around $6,864 per firm (2014$) because with a longer compliance period there will be fewer
uncoordinated labeling changes and manufacturers will have less of a need for overtime labor
and additional staffing and to pay rush charges to vendors and suppliers related to reformulation.
D.
Summary
Under the Regulatory Flexibility Act (5 U.S.C. 606(b)), we conclude that the final rules
will have a significant economic impact on a substantial number of small entities.
III. PRA of 1995
We are publishing two final rules on nutrition labeling in the Federal Register. The two
final rules will have the same effective date and compliance date. Thus, a manufacturer is likely
to be able to coordinate the required label changes in the two separate rules. However, we
estimate the hour burden of the final rules as if the rules have proceeded as a unique change to
the labeling regulations.
A.
Nutrition Facts Label Rule
The Nutrition Facts Label final rule contains information collection provisions that are
subject to review by OMB under the PRA. A description of these provisions is given in this
section with an estimate of the annual reporting, recordkeeping, and third-party disclosure
burden. Included in the estimate is the time for reviewing instructions, searching existing data
sources, gathering and maintaining the data needed, and completing and reviewing each
collection of information.
Title: Record Retention, Reporting, and Third-Party Disclosure Requirements for the
Declaration of Added Sugars, Dietary Fiber, Soluble Fiber, Insoluble Fiber, Vitamin E and
Folate/Folic Acid
Recordkeeping Requirements
107
Nutrition Facts/Serving Size Combined Final RIA
Description of Respondents: The likely respondents to this information collection are
manufacturers of retail food products marketed in the United States whose products contain (1) a
mixture of naturally occurring and added sugars or (2) a mixture of non-digestible carbohydrates
that do and do not meet the definition of dietary fiber. The likely respondents to this information
collection also include manufacturers of retail food products marketed in the United States
whose products contain (1) mixtures of different forms of vitamin E or (2) both folate and folic
acid.
Description: The Nutrition Facts Label rule requires that under certain circumstances
manufacturers make and keep certain records to verify the amount of added sugars when a food
product contains both naturally occurring sugars and added sugars, isolated or synthetic nondigestible carbohydrates that do not meet the definition of dietary fiber, different forms of
vitamin E, and folate/folic acid declared on the Nutrition Facts or Supplement Facts label, which
is the amount in the finished food product. Manufacturers are required to provide such records
to an appropriate regulatory official upon request during inspection. Manufacturers are also
required to maintain the records to verify the label declaration of the aforementioned nutrients
for a period of 2 years after introduction or delivery for introduction of the food into interstate
commerce. Manufacturers of food products that contain an isolated or synthetic non-digestible
carbohydrate that are not listed in the definition of dietary fiber will have the option of
submitting a citizen petition to FDA to request the agency amend the definition of “dietary fiber”
to include the carbohydrate as a listed dietary fiber, by demonstrating the physiological benefits
of the isolated or synthetic non-digestible carbohydrate to human health. In addition, if the
isolated or synthetic non-digestible carbohydrate is the subject of an authorized health claim,
FDA would consider the carbohydrate to be a dietary fiber with a beneficial physiological effect
108
Nutrition Facts/Serving Size Combined Final RIA
to human health and would amend the definition of “dietary fiber” to include the carbohydrate as
a listed dietary fiber. If the citizen petition is granted, or if the isolated or synthetic nondigestible carbohydrate is the subject of an authorized health claim, then the non-digestible
carbohydrate is considered to meet the definition of dietary fiber and the definition would be
amended to include the dietary fiber in the listing of dietary fibers that must be included in the
total amount of dietary fiber declared on the Nutrition or Supplement Facts label by food
manufacturers who manufacture food products that contain the isolated or synthetic nondigestible carbohydrate. The record requirements are necessary because analytical methods are
not available that would allow us to differentiate between naturally occurring and added sugars,
non-digestible carbohydrates (soluble or insoluble) that do and do not meet the definition of
dietary fiber, the various forms of vitamin E, and folate or folic acid in order to quantify the
amount of added sugars, dietary fiber, soluble fiber, insoluble fiber, vitamin E or folate/folic acid
in the final food product. For the nutrients described above for which there are no analytical
methods available to verify the label declaration, we must rely on information known only to the
manufacturer, e.g., analyses of nutrient databases, the food’s formulation or recipe, batch
records, or other records, to determine whether their product contains the declared amount of the
nutrient and is in compliance with the requirements of §§ 101.9(g) and 101.36(f).
We require that firms make and keep certain records necessary to verify the amount of
the nutrients in the finished food product. The Nutrition Facts Label rule does not specify what
records must be used to verify the amounts of these nutrients, but does specify the information
that the records must contain. The Nutrition Facts Label rule would require manufacturers to,
upon request during an inspection, provide FDA with the records that contain the required
information for each of these nutrients to verify the amount of the nutrient declared on the label.
109
Nutrition Facts/Serving Size Combined Final RIA
These records may include analyses of nutrient databases, recipes or formulations, information
from recipes or formulations, batch records, or other appropriate verification records that contain
the required information to verify the nutrient content in the final product.
We estimate the burden of this collection of information as follows:
Table 1.--Estimated Annual Recordkeeping Burden1
Type of Declaration/
CFR Section
Added Sugars/
§ 101.9(c)(6)(iii)2
Dietary Fiber/
§ 101.9(c)(6)(i)2
Soluble Fiber/
§ 101.9(c)(6)(i)(A)2
Insoluble Fiber/
§ 101.9(c)(6)(i)(B)2
Dietary Fiber/
§ 101.9(c)(6)(i)
Vitamin E/
§ 101.9(c)(8)3
Folate/Folic Acid/
§ 101.9(c)(8)3
No. of
Recordkeepers
No. of Records
per
Recordkeeper
Total
Annual
Records
Average Burden
per Recordkeeping
Total
Hours
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
28
1
28
1
28
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
Total
187,726
Total Initial Hours
187,726
New Products
216
1
Total Recurring Hours
216
1
216
216
Total Burden Hours
187,942
110
Nutrition Facts/Serving Size Combined Final RIA
1
There are no capital costs or operating and maintenance costs associated with this collection of information.
These estimates are likely to be large overestimates, as not all manufacturers will need to keep records for added
sugars, dietary fiber, and soluble and insoluble fiber. Manufacturers will only need to keep records for products with
both added and naturally occurring sugars and products with non-digestible carbohydrates (soluble or insoluble) that
do and do not meet the definition of dietary fiber.
3
These estimates are likely to be large overestimates, as not all manufacturers will need to keep records for vitamin
E and folate/folic acid. The declaration of vitamin E and folate/folic acid is not mandatory unless a health or
nutrient content claim is being made or these nutrients are directly added to the food for enrichment purposes.
2
Based on our experience with food labeling regulations, we believe that the new records
that would be required to be retained by the final rules are records that a prudent and responsible
manufacturer uses and retains as a normal part of doing business, e.g., analyses of nutrient
databases, recipes or formulations, batch records, or other records. Thus, the recordkeeping
burden of the final rules consists of the time required to identify and assemble the records for
copying and retention. Based on our previous experience with similar recordkeeping
requirements, we estimate the recordkeeping burden of the Nutrition Facts Label rule to be 1
hour per manufacturer as estimated in Table 1.
Under the Nutrition Facts Label rule, the declarations for added sugars, dietary fiber,
soluble fiber, and insoluble fiber are mandatory, and we conservatively estimate that all roughly
31,283 food manufacturers would incur this recordkeeping burden and that the required
recordkeeping would be 1 hour per manufacturer. We estimate that there are currently
approximately 28 isolated or synthetic non-digestible carbohydrates that are not listed in the
definition of dietary fiber.46 Once a citizen petition is filed by a manufacturer related to a
particular isolated or synthetic non-digestible carbohydrate is granted or denied, or the
carbohydrate is the subject of an authorized health claim, and the dietary fiber is listed in the
definition of dietary fiber, the use of the dietary fiber as an ingredient in any food product must
be included in the total amount of dietary fiber declared in nutrition labeling for such product.
46
We have no way of estimating the number of isolated or synthetic non-digestible carbohydrates that will be
discovered in the future.
111
Nutrition Facts/Serving Size Combined Final RIA
Thus, it is estimated that 28 manufacturers would incur a recordkeeping burden associated with
filing a citizen petition to amend the listing of dietary fiber related to an isolated or synthetic
non-digestible carbohydrate that is not currently listed in the definition of dietary fiber and that
the required recordkeeping would be 1 hour per manufacturer. The declaration of vitamin E and
folate/folic acid is not mandatory unless a health or nutrient content claim is being made or these
nutrients are directly added to the food for enrichment purposes. However, we conservatively
estimate that all roughly 31,283 food manufacturers would incur this recordkeeping burden and
that the required recordkeeping would be 1 hour per manufacturer.
It is hard to predict with certainty the exact number of newly introduced products that
would be covered under the Nutrition Facts Label rule each year, but based on the industry
growth rate estimated using U.S. Census Bureau Business and Industry data, we estimate that
number to be about 216. Thus, we estimate that about 216 new products would be affected by
the Nutrition Facts Label rule, and that the required recordkeeping would be 1 hour per product,
for an annual recurring recordkeeping burden of 216 hours (216 × 1). Adding the burden from
new products to the burden for existing products results in a total of 187,942 recordkeeping
burden hours for the covered establishments under the Nutrition Facts Label rule, as reported in
Table 1.
Reporting Requirements
Description of Respondents: The likely respondents to this information collection are
manufacturers of retail food products marketed in the United States whose products contain (1) a
combination of both naturally occurring and added sugars (2) a mixture of non-digestible
carbohydrates that do and do not meet the definition of dietary fiber, soluble fiber, and insoluble
fiber. The likely respondents to this information collection also include manufacturers of retail
112
Nutrition Facts/Serving Size Combined Final RIA
food products marketed in the United States whose products contain (1) mixtures of different
forms of vitamin E or (2) both folate and folic acid if a health or nutrient content claim is being
made or these nutrients are directly added to the food for enrichment purposes.
Description: Under the Nutrition Facts Label rule, we require that firms provide records
upon request during an inspection that they use to verify the declared amounts of added sugars,
dietary fiber, soluble fiber, insoluble fiber, vitamin E, and folate/folic acid on the Nutrition Facts
or Supplement Facts label.
The reporting requirement is necessary because, at the present time, analytical methods
are not available that would allow us to differentiate between naturally occurring and added
sugars, non-digestible carbohydrates that both do and do not meet the definition of dietary fiber,
soluble fiber, and insoluble fiber, the various forms of vitamin E, and folate or folic acid in order
to quantify the amount of added sugars, dietary fiber, vitamin E or folate/folic acid in the final
food product. For these foods, we must rely on information known only to the manufacturer to
assess compliance with the qualifying amount of nutrient. The food manufacturer would
assemble and provide the records to FDA regulatory officials upon request during an inspection.
We would review the records to verify the label declaration and assess compliance.
Table 2.--Estimated Annual Reporting Burden1
Type of Declaration/
CFR Section
Added Sugars/
§ 101.9(c)(6)(iii)2
Dietary Fiber/
§ 101.9(c)(6)(i)2
Soluble Fiber/
No. of
Respondents
No. of Responses
per Respondent
Total Annual
Responses
Average
Burden per
Response
Hours
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
113
Total
Nutrition Facts/Serving Size Combined Final RIA
§ 101.9(c)(6)(i)(A)2
Insoluble Fiber/
§ 101.9(c)(6)(i)(B)2
Vitamin E/
§ 101.9(c)(8)3
Folate/Folic Acid/
§ 101.9(c)(8)3
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
31,283
1
31,283
1
31,283
Total
187,698
Total Initial Hours
187,698
New Products
216
1
216
1
Total Recurring Hours
216
216
Total Burden Hours
187,914
1
There are no capital costs or operating and maintenance costs associated with this collection of information.
These estimates are likely to be large overestimates, as not all manufacturers will need to keep records for added
sugars, dietary fiber, and soluble and insoluble fiber. Manufacturers will only need to keep records for products with
both added and naturally occurring sugars and products with non-digestible carbohydrates (soluble or insoluble) that
do and do not meet the definition of dietary fiber.
3
These estimates are likely to be large overestimates, as not all manufacturers will need to keep records for vitamin
E and folate/folic acid. The declaration of vitamin E and folate/folic acid is not mandatory unless a health or
nutrient content claim is being made or these nutrients are directly added to the food for enrichment purposes.
2
Based on our experience with food labeling regulations, we believe that the records that
would be required to be provided to FDA, upon request, are records that a prudent and
responsible manufacturer uses and retains as a normal part of doing business, e.g., analyses of
nutrient databases, recipes or formulations, batch records, or other records. Thus, the reporting
burden to the food manufacturer consists of the time required to assemble and provide the
records to appropriate regulatory officials. Based on our previous experience with similar
reporting requirements, we estimate the reporting burden of the Nutrition Facts Label rule to be 1
hour per response, as estimated in Table 2.
114
Nutrition Facts/Serving Size Combined Final RIA
We do not expect to request records from all covered manufacturers to assess
compliance, but for the purpose of this analysis the number of respondents is conservatively
estimated to be all covered establishments. We estimate the number of responses per record
keeper to be 1 and the hourly burden per response to be 1 hour. Built into the estimate of 1 hour
is the range from 0 hours, for some covered manufacturers that do not need to maintain records,
to a larger number of hours for some covered manufacturers, such as those who produce
fermented foods, that may require more time to gather or produce the necessary records. As
shown in Table 2, the initial reporting burden for covered establishments is 187,698 hours. Also,
in accordance with our previous estimate of the number of newly introduced products that would
be covered by the requirements to be 216, we estimate the recurring reporting burden hours to be
216. Adding the burden from new products to the initial hours results in a total of 187,914
reporting burden hours for the covered establishments under the Nutrition Facts Label rule, as
estimated in Table 2.
Third-Party Disclosure Requirements
Description of Respondents: Respondents to this collection of information include
manufacturers of food products. We estimate the burden of this collection of information as
follows:
Table 3.--Estimated Annual Third Party Disclosure Burden1
CFR
Section
101.9 and
101.36
Number of
Respondents
No. of
Disclosures
per
Respondent
31,283
26
Total Annual
Disclosures
Average
Burden per
Disclosure
Total
Hours
Total
Capital
Costs (in
billions of
2014$)
813,358
2
1,626,716
$2.47
1. There are no operating and maintenance costs associated with this collection of information.
We have estimated that the burden associated with the Nutrition Facts Label rule would
be a burden created by the need for food manufacturers to revise their nutrition labels. We
115
Nutrition Facts/Serving Size Combined Final RIA
estimate that the third party disclosure burden would be approximately 2 hours per disclosure,
for a total burden of 1,626,716 hours.
Third party disclosure burden for manufacturers:
The incremental time burden for reviewing labels to assess how to bring them into
compliance with the requirements of the Nutrition Facts Label rule has been estimated to be 1
hour per label. These requirements do not generate any recurring burden per label because
establishments must already print packaging for food products as part of normal business
practices, and must disclose required nutrition information under NLEA.
Each label redesign would require an estimated 1 additional hour, making the total
burden hours to be 2 hours in burden per UPC.
We estimate that about 31,283 manufacturers representing about 813,358 UPCs, with an
average disclosure of 26 (813,358/31,283), would be covered under the Nutrition Facts Label
rule. The total number of responses is equal to the total number of UPCs being changed.
Multiplying the total number of responses by the hours per response gives the total burden hours
(Table 3, Column 6). Based on the RIA, we have estimated the capital cost to be $2.47 billion
(2014$).
B.
Serving Size Rule
This rule contains information collection provisions that are subject to review by OMB
under the PRA. A description of these provisions is given in this section with an estimate of the
annual third-party disclosure burden. Included in the estimate is the time for reviewing
instructions, searching existing data sources, gathering and maintaining the data needed, and
completing and reviewing each collection of information.
116
Nutrition Facts/Serving Size Combined Final RIA
Title: Third-Party Disclosure Requirements for Serving Sizes of Foods That Can Reasonably Be
Consumed At One Eating Occasion; Dual-Column Labeling; Updating, Modifying and
Establishing Certain RACCs; Serving Size for Breath Mints; and Technical Amendments
Reporting Requirements
None.
Recordkeeping Requirements
None.
Third-Party Disclosure Requirements
Description of Respondents: The respondents to this information collection are
manufacturers of retail food products marketed in the United States.
Description: In major part, the Serving Size rule revises §§ 101.9 and 101.12 to (1)
amend the definition of a single serving, (2) require a second column of nutrition information per
package for products that contain at least 200 and up to and including 300 percent of the
applicable RACCs, as well as per unit for discrete units in multi-serving packages in which each
unit contains at least 200 percent and up to and including 300 percent of the applicable RACCs,
(3) update, modify, and establish RACCs for certain food products, (4) make several technical
amendments to the regulations for serving sizes, and (5) change the label serving size for breath
mints to “1 unit.” These revisions, in many instances, will require changes to the nutrition
information that is presented on the Nutrition Facts label of retail food products. Preexisting §§
101.9 and 101.12 are approved by OMB in accordance with the PRA under OMB control
number 0910-0381. The Serving Size rule will modify the information collection associated
with preexisting §§101.9 and 101.12 by adding to the burden associated with the collection by
requiring the following manufacturers to make changes to their product labels: those whose
117
Nutrition Facts/Serving Size Combined Final RIA
retail food products are labeled with a serving size that is inconsistent with the provisions of the
final rule, and those whose retail food products would be required to use DCL.47 The nutrient
information disclosed on labels of retail food products is necessary to inform purchasers of the
nutritional value of the food.
We estimate the burden of this collection of information as follows:
Table 1.--Estimated Annual Third Party Disclosure Burden1
21 CFR Section
No. of
Respondents
No. of
Disclosures per
Respondent
Total Annual
Disclosures
Average
Burden per
Disclosure
Total
Hours
Total
Capital
Costs (in
billions of
2014$)
101.9 and 101.12
13,452
25
336,300
2
672,600
$1.00
672,600
$1.00
1,000
$0.01
1,000
$0.01
673,600
$1.01
Total Initial Hours and Capital Costs
New Products
500
1
500
2
Total Recurring Hours and Capital Costs
Total Burden Hours and Capital Costs
1
There are no operating and maintenance costs associated with this collection of information.
Under §§ 101.9 and 101.12, some manufacturers of retail food products would need to
make a labeling change to modify the serving sizes and other nutrition information based on
changes to what products may be or are required to be labeled as a single serving or based on
updated, modified, or established RACCs. Additionally, some manufacturers would need to
change their product labels to add a second column of nutrition information per package or per
discrete unit as part of the Nutrition Facts label. The third-party disclosure burden consists of the
47
Included in this burden are the labeling costs that result from changes in the eligibility to bear nutrient content
claims or health claims (e.g., the cost of removing a claim from labeling or adding a required disclaimer).
118
Nutrition Facts/Serving Size Combined Final RIA
setup time required to design a revised label and incorporate it into the manufacturing process.
The third-party disclosure burden for the Serving Size rule is estimated in Table 1.
Based upon our knowledge of food labeling, we estimate that the affected manufacturers
would require 2 hours per product to modify the label’s Nutrition Facts panel. We estimate that
it would take an affected manufacturer 1 hour to review a label to assess how to bring it into
compliance with the requirements of the Serving Size rule. Each label redesign would require an
estimated 1 additional hour per UPC, for a total of 2 hours per UPC.
We estimate that about 13,452 manufacturers would initially be affected by the Serving
Size rule and that about 336,300 products would initially be required to be relabeled, for an
average of 25 (336,300/13,452) products per respondent. The total initial third-party disclosure
burden of 672,600 hours is reported in Table 1. The final column of Table 1 gives the estimated
initial capital cost of the relabeling associated with the Serving Size rule. Based on the RIA, we
estimate the initial capital cost to be approximately $1 billion (2014$).
The Serving Size rule generates recurring burdens related to the requirement that some
manufacturers undertake an extensive label change due to the effect of the changed definition of
a single-serving container on the permissibility of certain health and nutrient content claims and
also to the requirement that some manufacturers undertake a major redesign of their labels to
include a Nutrition Facts Panel that had not previously been required.48 We estimate that about
500 new products would be affected by these requirements each year, and that the required third
party disclosure burden would be 2 hours per product, for an annual recurring third party
48
The Serving Size rule does not otherwise generate any recurring burdens because establishments must already
print packaging for food products as part of normal business practices and must disclose required nutrition and
serving size information under NLEA.
119
Nutrition Facts/Serving Size Combined Final RIA
disclosure burden of 1,000 hours. Based on the RIA, we estimate the annual recurring capital
cost to be approximately $0.01 billion (2014$).
Adding the recurring burden from new products to the initial burden for existing products
results in a total of 673,600 third party disclosure burden hours and $1.01 billion (2014$) in
capital costs as reported in Table 1.
The information collection provisions in the Nutrition Facts Label rule and the Serving
Size rule have been submitted to OMB for review as required by Section 3507(d) of the PRA of
1995.
Before the effective date of the Serving Size final rule and the Nutrition Facts Label final
rule, we will publish a notice in the Federal Register announcing OMB’s decision to approve,
modify, or disapprove the information collection provisions in the rules.
An Agency may not conduct or sponsor, and a person is not required to respond to, a
collection of information unless it displays a currently valid OMB control number.
120
Nutrition Facts/Serving Size Combined Final RIA
References
1
RTI International. “2014 FDA Labeling Cost Model.” Prepared by Mary K. Muth,
Samantha Bradley, Jenna Brophy, Kristen Capogrossi, Michaela C. Coglaiti, and Shawn
A. Karns. Contract No. HHSF-223-2011-10005B, Task Order 20, August 2015.
2
RTI International. “Reformulation Cost Model.” Prepared by Mary K. Muth, Samantha
Bradley, Jenna Brophy, Kristen Capogrossi, Michaela Coglaiti, Shawn Karns, and
Catherine Viator. Contract No. HHSF-223-2011-10005B, Task Order 20, August 2015.
3
Abaluck J. What Would We Eat if we Knew More: The Implications of a Large-Scale
Change in Nutrition Labeling. 10-17-2011.
4
Ruhm CJ. Understanding overeating and obesity. J Health Econ 2012; 31/6: 781-796.
5
O'Donoghue T, Rabin M. Optimal sin taxes. Journal of Public Economics 2006; 90/10:
1825-1849.
6
Loewenstein G. Out of Control: Visceral Influences on Behavior. Organizational
Behavior and Human Decision Processes 1996; 65/3: 272-292.
7
O'Donoghue T, Rabin M. Doing It Now or Later. American Economic Review 1999;
89/1: 103-124.
8
Akerlof GA. Procrastination and obedience. American Economic Review 1991; 81/2: 1.
9
Epstein LH, Saelens BE. Behavioral Economics of Obesity: Food Intake and Energy
Expenditure. In: Bickel WK, Vuchinich RE (eds). Reframing Health Behavior Change
With Behavioral Economics. Lawrence Erlbaum Asociates, Inc.: Mahwah, NJ, 2000, pp.
295-314.
10
Conlisk J. Why Bounded Rationality? Journal of Economic Literature 1996; 34/2: 669700.
11
Lin J, Lee JY, Yen ST. Do dietary intakes affect search for nutrient information on food
labels? Social Science & Medicine 2004; 59/9: 1955-1967.
12
Nielsen SJ, Popkin BM. Patterns and trends in food portion sizes, 1977-1998. JAMA: The
Journal of the American Medical Association 2003; 289/4: 450-453.
13
Wansink B, Kim J. Bad Popcorn in Big Buckets: Portion Size Can Influence Intake as
Much as Taste. Journal of Nutrition Education and Behavior 2005; 37/5: 242-245.
14
Wansink B. Environmental factors that increase the food intake and consumption volume
of unknowing consumers. Annual Review of Nutrition 2004; 24/1: 455-479.
121
Nutrition Facts/Serving Size Combined Final RIA
15
Post, R., Budak, C., Canavan, J., Duncan-Harrington, T., Jones, B., Jones, S., MurphyJenkins, R., Myrick, T., Wheeler, M., White, P., Yoder, L., Kegley, M. “A Guide to
Federal Food Labeling Requirements for Meat and Poultry Products”. The Labeling and
Consumer Protection Staff, Office of Policy, Program, and Employee Development, Food
Safety and Inspection Service, U.S. Department of Agriculture, 2007.
16
U.S. Department of Health and Human Services. (2009). Guidance for Industry: A Food
Labeling Guide, 3rd Edition, Washington, DC.
17
Hubbard RG and O’Brien AP. Microeconomics. 3rd Edition. 2010. Boston, MA,
Pearson.
18
Pieters, R., and M. Wedel. Attention Capture and Transfer in Advertising: Brand,
Pictorial, and Texas-Size Effect, Journal of Marketing, 2004; 68:36-50.
19
Lando, A. and S. Lo. Single-Larger-Portion-Size and Dual Column Nutrition Labeling
May Help Consumers Make More Healthful Choices. Journal of the Academy of
Nutrition and Dietetics. 2013;113:241-250.
20
National Institutes of Health Office of Dietary Supplements. Dietary Supplement Fact
Sheet: Vitamin B12. Accessed on November 18, 2013 at
http://ods.od.nih.gov/factsheets/VitaminB12-HealthProfessional/.
21
Danaei G, Ding EL, Mozaffarian D, Taylor B, Rehm J, Murray CJL et al. The
Preventable Causes of Death in the United States: Comparative Risk Assessment of
Dietary, Lifestyle, and Metabolic Risk Factors. PLoS Med 2009; 6/4: e1000058.
22
Ford ES, Bergmann MM, Boeing H, Li C, Capewell S. Healthy lifestyle behaviors and
all-cause mortality among adults in the United States. Preventive Medicine 2012; 55/1:
23-27.
23
Mokdad AH, Marks J, Stroup DF, Gerberding JL. Actual causes of death in the united
states, 2000. JAMA: The Journal of the American Medical Association 2004; 291/10:
1238-1245.
24
Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity in the United States,
2009-2010. NCHS Data Brief 2012;/82: 1-8.
25
Bibbins-Domingo K, Chertow GM, Coxson PG, Moran A, Lightwood JM, Pletcher MJ et
al. Projected Effect of Dietary Salt Reductions on Future Cardiovascular Disease. N Engl
J Med 2010; 362/7: 590-599.
26
National Center for Health Statistics. Health, United States, 2011: With Special Feature
on Socioeconomic Status and Health. Hyattsville, MD. 2012.
122
Nutrition Facts/Serving Size Combined Final RIA
27
Antonuk B, Block LG. The Effect of Single Serving Versus Entire Package Nutritional
Information on Consumption Norms and Actual Consumption of a Snack Food. Journal
of Nutrition Education and Behavior 2006; 38/6: 365-370.
28
O’Brien, T. “Status of Nutrition Labeling of Processed Foods: 1995.” Center for Food
Safety and Applied Nutrition, U.S. Food and Drug Administration.
29
Brecher, SJ, Bender, MM, Wilkening, VL, McCabe, NM, and EM Anderson. Status of
Nutrition Labeling, Health Claims, and Nutrient Content Claims for Processed Foods:
1997 Food Label and Package Survey. Journal of the American Dietetic Association
2000; 100: 1057-1062.
30
Ferguson, MS. Statistical Support Memo for Justifying Using Historical FLAPS Data.
FDA Internal Memorandum 2013.
31
National Cancer Institute. Food Sources of Energy Among U.S. Population, 2005 – 2006.
Risk Factor Monitoring and Methods. Control and Population Sciences. National Cancer
Institute; 2010.
32
Palisade Corporation. @Risk for Excel. Version 6.1.2: Professional Edition. 2013.
33
Evatt, ML, Mersereau, PW, Bobo, JK, Kimmons, J, and J Williams. Why Vitamin B12
Deficiency Should Be on Your Radar Screen. Course WB1349 Prepared for the National
Center on Birth Defects and Developmental Disabilities. Centers for Disease Control and
Prevention. February 11, 2010.
34
Stabler, S. Vitamin B12 Deficiency. The New England Journal of Medicine 2013; 368(2):
149-160.
35
U.S.Department of Agriculture and U.S.Department of Health and Human Services.
Dietary Guidelines for Americans, 2010. 2010. Washington, DC.
36
World Health Organization. "Diet, nutrition, and the prevention of chronic diseases.
Report of a WHO Study Group," Geneva: World Health Organization; 1990. pp.1-204.
37
Knopman D, Boland LL, Mosley T, Howard G, Liao D, Szklo M et al. Cardiovascular
risk factors and cognitive decline in middle-aged adults. Neurology 2001; 56/1: 42-48.
38
Rogers PJ. A healthy body, a healthy mind: long-term impact of diet on mood and
cognitive function. The Proceedings of the Nutrition Society 2001; 60/1: 135-143.
39
Brill, A. The Long-Term Returns of Obesity Prevention Policies. Robert Wood Johnson
Foundation. April, 2013. Available at: http://www.rwjf.org/content/dam/farm/reports/reports/2013/rwjf405694.
123
| File Type | application/pdf |
| File Modified | 2016-05-17 |
| File Created | 2016-05-17 |