Final Regulatory Impact Analysis

DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration

Standards for the Growing, Harvesting, Packing
and Holding of Produce for Human
Consumption
Docket No. FDA-2011-N-0921

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

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

Executive Summary
The rule establishes science-based minimum standards for the safe growing, harvesting,
packing, and holding of produce on farms. The rule addresses microbiological risks from
certain routes of contamination, including workers, agricultural water, biological soil
amendments of animal origin, buildings, tools and equipment and sanitation, and wild
and domesticated animals. The rule also includes specific requirements for sprouts.
Using a science-based framework, we characterized the public health risks associated
with the consumption of produce and are establishing specific provisions that address the
risks of microbial contamination from these routes of contamination. The primary
benefits of the provisions in this rule are an expected decrease in the incidence of
illnesses related to microbial contamination of produce. Annualizing benefits over the
first ten years after the effective date of this final rule at seven percent, benefits are
expected to derive from averting approximately 331,964 illnesses per year (362,059 at
three percent), valued at $925 million annually ($976 million at three percent). Similarly,
annualized costs, estimated at seven percent, are expected to be approximately $366
million annually ($387 million at three percent).Additionally, annualized costs for foreign
farms are estimated to be approximately $138 million annualized at seven percent ($146
million at three percent).

Page 1

Table of Contents
I. Introduction and Summary .............................................................................................. 3
A. Introduction ................................................................................................................ 3
B. Summary of Costs and Benefits ................................................................................. 4
C. Comments on the Preliminary Regulatory Impact Analysis and Our Responses ...... 5
II. Final Regulatory Impact Analysis ................................................................................ 29
A. Background .............................................................................................................. 29
B. Need for Regulation ................................................................................................. 34
C. Purpose of the Rule .................................................................................................. 36
D. Inputs and Assumptions ........................................................................................... 36
E. Benefits of the Rule .................................................................................................. 48
F. Costs of the Rule....................................................................................................... 63
1. Personnel and Training (Subpart C) ..................................................................... 63
2. Health and Hygiene (Subpart D)........................................................................... 64
3. Agricultural Water (Subpart E) ............................................................................. 65
4. Biological Soil Amendments (Subpart F) ............................................................. 75
5. Domesticated and Wild Animals (Subpart I) ........................................................ 76
6. Growing, Harvesting, Packing, and Holding Activities (Subpart K) ................... 77
7. Equipment, Tools, Buildings, and Sanitation (Subpart L) .................................... 78
8. Sprouts (Subpart M).............................................................................................. 79
9. Recordkeeping (Subpart O) .................................................................................. 82
10. Administrative Provisions................................................................................... 98
11. Corrective Steps .................................................................................................. 99
12. Variances........................................................................................................... 100
13. Summary of Costs ............................................................................................. 100
G. Distributional Effects ............................................................................................. 104
H. International Effects ............................................................................................... 105
I. Uncertainty and Sensitivity Analysis ..................................................................... 106
J. Analysis of Regulatory Alternatives to the Rule .................................................... 109
III. Final Small Entity Analysis ...................................................................................... 114
A. Description and Number of Affected Small Entities ............................................. 114
B. Description of the Potential Impacts of the Rule on Small Entities ....................... 116
C. Alternatives to Minimize the Burden on Small Entities ........................................ 118
IV. References................................................................................................................. 120

Page 2

I. Introduction and Summary
A. Introduction
FDA has examined the impacts of the final rule under Executive Order 12866,
Executive Order 13563, the Regulatory Flexibility Act (5 U.S.C. 601-612) and the
Unfunded Mandates Reform Act of 1995 (Public Law 104-4). Executive Orders 12866
and 13563 direct Agencies 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). The Agency believes that
this final rule will be an economically significant regulatory action as defined by
Executive Order 12866.
If a rule has a significant economic impact on a substantial number of small
businesses, the Regulatory Flexibility Act requires Agencies to analyze regulatory
alternatives that would minimize any significant impact of a rule on small entities. FDA
has determined that this final rule will have a significant economic impact on a
substantial number of small entities.
Section 202(a) of the Unfunded Mandates Reform Act of 1995 requires that
Agencies prepare a written statement, which includes an assessment of anticipated costs
and benefits, before proposing “any rule that includes any Federal mandate that may
result in the expenditure by state, local, and tribal governments, in the aggregate, or by
the private sector, of $100,000,000 or more (adjusted annually for inflation) in any one
year.” The current threshold after adjustment for inflation is $141 million, using the most

Page 3

current (2013) Implicit Price Deflator for the Gross Domestic Product. FDA does expect
this final rule to result in any 1-year expenditure that will meet or exceed this amount.

B. Summary of Costs and Benefits
The requirements of the final Standards for the Growing, Harvesting, Packing,
and Holding of Produce for Human Consumption regulation (Produce Safety rule, the
final rule, or the rule) will lead to higher costs for both the industry and consumers than
the current state of no new regulatory action. As described in the preamble, the final rule
includes requirements for covered domestic and foreign farms engaged in the growing,
harvesting, packing, and/or holding of one or more raw agricultural commodities
(RACs) 1 that are covered produce. The final rule also requires covered domestic and
foreign farms to train their employees; take certain measures related to employees’ health
and hygiene; monitor, understand, and take certain measures related to their agricultural
water; assess for domesticated and wild animals activity in areas used for covered
activities; take certain measures during growing, harvesting, packing, and holding
activities; and take certain measures relating to sanitation, including cleaning and
sanitizing equipment and tools, and appropriately maintaining buildings. In addition, the
rule establishes certain requirements for the growing, harvesting, packing, and holding of
sprouts. Farms will be required, to take appropriate corrective actions, and maintain
certain records, including records that document these corrective actions. The affected
farms will incur costs to comply with this final regulation. Depending on how the farms

1

When discussing Raw Agricultural Commodities (RAC), we refer to RACs covered by the rule unless
otherwise noted.

Page 4

in the affected markets respond to these requirements, some of the costs may ultimately
be borne by consumers as prices rise. The higher prices, however, will likely not be
sufficient to fully offset the costs borne by farms.
Table 1 summarizes the costs and benefits of the Produce Rule. More detail on
these estimates is provided in the relevant sections of this document, specifically benefits
come from Table 6 and costs come from Table 37.
Table 1: Summary of Benefits and Costs of Final Rule (in millions)
Discount
Rate
Annualized Benefits over
10 years

3%
7%

NPV of Benefits over 10
years

3%
7%

Annualized Costs over 10
years

3%
7%

NPV of Costs over 10
years

3%
7%

Primary Estimate

Low Estimate

High Estimate

$976

$748

$1,195

$925

$710

$1,132

$8,322

$6,381

$10,190

$6,498

$4,988

$7,950

$387

$319

$425

$366

$301

$401

$3,304

$2,717

$3,624

$2,571

$2,113

$2,817

In addition to the costs presented in Table 1, we estimate there will also be costs
incurred by foreign farms shipping RACs to the U.S. We estimate a total annualized cost
to foreign farms shipping produce RACs to the US of $136 million annually, using a 7
percent discount rate ($146 million using a 3 percent discount rate).

C. Comments on the Preliminary Regulatory Impact Analysis and Our
Responses
FDA’s proposed rule “Standards for the Growing, Harvesting, Packing, and
Holding of Produce for Human Consumption” (78 FR 3504; the 2013 proposed rule) was

Page 5

published on January 16, 2013 and its comment period ended November 22, 2013. In
addition, we published a supplemental notice to the proposed rule on September 29, 2014
(79 FR 58434) and its comment period ended December 15, 2014. (We refer to both of
these documents collectively as “the proposed produce safety rule.”) We prepared a full
“Preliminary Regulatory Impact Analysis” in connection with the proposed and
supplemental rule. We also included sections titled “Costs and Benefits” and “Initial
Regulatory Flexibility Analysis” in the preamble to the proposed rule (76 FR 19192 at
19220-19225). In the following paragraphs, we describe and respond to the comments
we received on our analyses of the impacts presented in those sections.

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) Several commenters express concern about the magnitude of the cost
of the rule. Specifically, they state that the rule would: cost farmers over half of their
profits; put an unfair financial burden on small and medium farms; cause many farms to
go out of business; deny farmers access to local food markets by making it harder to
diversify (e.g., a small strawberry operation that is part of a large non-produce farm may
be subject to the rule even if the specific sales of strawberries are below the exemption
cutoff); and prevent new farmers from starting to farm.
Response 1) FDA recognizes that the cost of this rulemaking is not
inconsequential. However, we believe the need for a safer food supply warrants such
expenditures. In our analyses, we find that the average cost of the rule for very small
farms is $2,885 per farm, while the average value of produce sales is $85,616. Similarly,

Page 6

we find the average cost for small farms is $15,265 per farm, while the average value of
produce sales is $358,814 per farm. We do not believe that this rule will in any way
hinder farmers’ access to local markets. In fact, exemptions are set up in such a way as to
encourage sales of produce locally (Ref. 1). We have revised our provisions related to
coverage of the rule (see 112.4, which establishes the monetary threshold based on sales
of produce (rather than sales of food)), and the rule, as finalized, will not hinder the
diversity or force those farms that have a relatively small amount of produce grown on
their farm to exit the industry. Finally, we recognize that these costs will affect farmers
entering into the industry, but we believe that all new farmers should be practicing safe
food practices, especially in the cases where the produce is likely to be consumed raw.
See also section III of the rule.
Comment 2) Several commenters state that the proposed produce safety rule will
have additional health costs because, by being disadvantageous to small and local farms,
they will reduce access to fresh, local, and healthy food. Commenters also suggest that
FDA needs to consider large scale crop losses, harm to soil and the municipal water
supply, and ecological impacts brought on by the water testing requirements, in its cost
analyses.
Response 2) FDA does not believe that this rule will reduce access to produce. In
fact, exemptions are set up in such a way as to encourage sales of produce locally (Ref. 1).
Additionally, FDA has conducted an assessment of impacts of the rule on the human
environment of the United States, and prepared an environmental impact statement (Ref.
2). According to the EIS, “providing that any pesticide that is EPA-registered and is
handled and applied in accordance with labeling requirements should not result in significant
environmental impacts to vegetation, wildlife, and wetland resources. However, such

Page 7

applications may result in short-term minimal to moderate impacts on these resources
particularly if applied preceding substantial periods of precipitation which may increase runoff. Such impacts would be intermittent and acute.” It further states that “if approved
products are used in accordance with labeling requirements, chemical contamination is not
expected to pose a human health risk.” In terms of soil, the EIS states, “…chloride is not
adsorbed by soils and moves readily with the soil-water; is taken up by the crop; moves in the
transpiration stream; and accumulates in the leaves. The chemical reactions that occur when
chlorine and organic matter are exposed to each other also produce toxic and carcinogenic
by-products. The use of antimicrobials, however, would not be expected to exceed the
threshold that would be toxic to crops, as long as labeling requirements are followed for
application purposes, and adverse effects to crops from overexposure to chemical treatments
should not occur.”

Comment 3) Several commenters state that the water testing requirements will be
overly costly to farms using water from creeks, streams and rivers,.
Response 3) We acknowledge that there is a cost to testing water; however, we
believe that the testing is important given the significant risk of foodborne illnesses
presented by agricultural water as a potential route of contamination. Numerous changes
have been made to make the requirements for agricultural water more flexible (see
section XIII of the rule) and we have attempted to account for those flexibilities within
this analysis. In total we estimate that agricultural water provisions, as written in the final
rule, will cost approximately $37 million dollars annually, which represents an average
cost to a single farm of approximately $1,058 per year.

Page 8

Comment 4) One comment states that FDA did not compare less costly
alternatives, such as establishing labeling requirements to instruct consumers to wash
produce.
Response 4) We believe that such an approach would be ineffective at reducing
the human health burden associated with contaminated produce, and therefore we did not
analyze the cost of such an approach. There are already a number of education campaigns
currently in progress, or that have been completed, which try to stress safe food handling
practices to the consumer. However, these are not completely effective in reducing
foodborne illness. We also note that establishing new labeling requirements does have the
potential to involve significant costs, especially where no label is currently required, such
as for many produce RACs.
Comment 5) Several commenters state that the costs of water testing are
particularly burdensome for operations with multiple water sources.
Response 5) The water testing provisions have been revised. The most
burdensome testing regimen is associated with the use of untreated surface water that is
used during growing of covered produce (other than sprouts) using a direct water
application method. If farms use untreated surface water source(s) for this purpose, they
will generally need to perform, for each source, an initial survey of 20 samples and
recurring annual samples of five per year, which is estimated to cost approximately $692
annualized over 10 years. The rule includes a provision allowing sharing of water testing
data under certain circumstances (§ 112.47(a)(2)). This will allow some farms to reduce
testing costs by sharing testing data.

Page 9

Comment 6) Several commenters state that customers may require partial or full
compliance with the produce rule standards even for operations that may be otherwise
exempt, therefore causing these operations to incur the costs of the produce rule.
Response 6) FDA recognizes that some costs may potentially be incurred by
farms not covered by this rule that are not required by FDA. To our knowledge, however,
there is no data on which to base a reasonable estimate of these costs not directly
attributable to the rule. Uncovered farms that incur these costs likely do so in order to
maintain market share and thus maximize revenues. We include the costs for farms not
covered or otherwise exempt for maintaining paperwork related to certain produce
exempt from the Produce Safety rule, and costs of complying with modified requirements
for those farms eligible for a qualified exemption with modified requirements. Anything
done by a farm to comply with aspects of the rule from which they are officially exempt
would likely be performed to preserve market share and/or profitability.
Comment 7) Several commenters state that the FDA should not assume small and
very small farms only operate three months out of the year, and that large farms operate
only 6 months per year and harvest, pack or hold produce only 90 days. Some suggest
increasing season estimates for all farms depending on the region.
Response 7) We agree that the original time estimates for very small, small, and
large farms may have been underestimated for some farms. Therefore, we have increased
our estimates of operating days for very small farms to 100, small farms to 150, and large
farms to 200. This is not to say that these farms do not carry on operations year round,
but, for our costs analysis, we are primarily concerned with those times when the
harvested or harvestable portions of the produce are exposed on the farm. Additionally,

Page 10

because we do not explicitly examine farms by region, but are tasked with the average
costs to all farms operating within the US, applying regional differences to operations is
not possible for this analysis.
Comment 8) Several commenters state that FDA should include costs for farm
mixed-type facilities in their cost-benefit analyses.
Response 8) We currently estimate the cost to all farms that meet the current farm
definition. The analysis of the costs and benefits of the produce rule is not affected by
whether or not a covered farm is also a facility subject to the Preventive Controls for
Human Food (PCHF) Rule. If a farm is covered under the produce rule, then it must
adhere to the rule. If that farm is also a facility subject to the PCHF rule, then the costs it
incurs by adhering to the PCHF rule will be accounted for in the cost and benefit analysis
of the PCHF rule (Ref. 3).
Comment 9) Several commenters state that FDA should analyze how the costs of
the rules will be passed on to consumers (e.g., via increased prices).
Response 9) FDA estimates the costs to industry and society as a whole but does
not estimate who will actually incur those costs (e.g., farms, intermediaries, retail
establishments, or end consumers). This is largely due to the lack of quantifiable data on
the issue. However, the total costs of this rule ($560 million, as shown in Table 34) when
fully implemented represent approximately 1.5 percent of the total value of produce sold
in the US ($38 billion). Additionally, the total cost to foreign farms that ship to the US is
$211 million (as estimated in Section H, International Effects), once the rule is fully in
effect, meaning that the total cost of this rule, foreign and domestic, represents
approximately 2 percent of the total value of the US produce market. This means that

Page 11

even if the total costs of compliance were passed on to consumers, which are highly
unlikely, it would represent a price increase of only 2 percent.
Comment 10) Several commenters state that “FDA disguises the first-year costs
of the regulations by annualizing them over 7 years for depreciation,” which “ignores the
issue of whether the farmer has the money to comply in the first year to begin with, as
well as the fact that many small farmers do not have sufficient income to make
depreciation cycles relevant.”
Response 10) FDA annualizes cost in accordance with Circular A-4 and
Executive Order 12866 (Ref. 4;5). This is not to ‘disguise’ the costs, but rather to
illustrate the likely costs of financing larger purchases over the long term. However, to
illustrate the complete first year costs, not annualized over any time horizon, we also
present these costs in Table 34. Summary of Costs for the Produce Safety Rule (in
millions)
Comment 11) One commenter states that FDA’s estimated rental value of $359
per acre for a full year is too small.
Response 11) This estimate was based on the best data that we could find on crop
land values for the proposed rule. However, because certain requirements related to
biological soil amendments of animal origin have been removed from the final rule,
related costs estimates have also been removed from this analysis and the rental value of
land no longer enters any of our calculations of costs to a farm.
Comment 12) One commenter states that there are not any EPA approved water
treatments, and that farmers would either have to stop irrigating (which will lead to crop
damage) or turn to public water sources, which can be more expensive.

Page 12

Another

comment adds that the “cost required to invest in a groundwater pump can be significant
and initial costs can be substantial. In 2013, in many parts of the West, drilling and
developing a new groundwater irrigation well costs between $100,000 and $500,000 to
supply water to 120 acres of productive farm land”
Response 12) As discussed in section XIII of the final rule, in § 112.45, we are
providing for different options that a covered farm can consider when agricultural water
is found to be not safe or of adequate sanitary quality for its intended use and/or does not
meet the relevant microbial quality criteria in § 112.44(a) or (b), and treatment is only
one of those options. We anticipate that covered farms will consider and implement the
flexible options provided in §§ 112.45(a) and (b) and 112.49, as appropriate, prior to or in
conjunction with considering whether to treat water to ensure that it meets the applicable
requirements for its intended use. Indeed, we believe some of these options are likely to
be more feasible than the option to treat water. Moreover, covered farms will have two
additional years (beyond the date of compliance for the remainder of the rule) to comply
with many of the water provisions of this rule for covered activities involving covered
produce (except sprouts), which is intended to help farms to consider and implement
measures that are most appropriate for their operations.
Comment 13) Several commenters state that the Clean Water Act statistics do not
provide a good estimate of how much irrigation water would fail to meet the EPA
recreational water standard. They state that there is no information in the report about
which of the water sources that don’t meet the standards are used for irrigation, how
much irrigation water is drawn from impaired sources, and groundwater usage.

Page 13

Response 13) We agree that EPA’s Clean Water Act statistics do not provide
precisely the measurements we would prefer to estimate the amount of water that is likely
to fail to meet the microbial water quality criteria in § 112.44(b); however, in the absence
of another source, we believe this to be the most comprehensive and nationally
representative source of data available. Because commenters did not provide any
additional data or sources of data on this topic, and because we were unable to find any
new or additional sources, we retain this as our source for estimates of water quality in
the final analysis.
Comment 14) Several commenters state that there is no analysis of the cost of
imposing microbial water quality criteria.
Response 14) The costs of imposing microbial water quality criteria are realized
through treatment of water used in growing or post-harvest activities (an estimate
affected by the number of farms we estimate that will be able to use other methods to
meet the microbial water quality criteria, such as reinspection/correction and reliance on
die-off or removal rates). These costs are presented in Table 18 and Table 19 of the
analysis.
Comment 15) Several commenters argue with FDA’s cost analysis by providing
counterexamples, which primarily referred to one farm, one specific region, or one
specific crop.
Response 15) For a national analysis of the costs and benefits of this rule we are
not able to comprehensively account for farms by commodities or agricultural region. We
are aware that there are differences in needs and resources across different farms, and as
such we attempt to provide a national average estimate that reflects this variety.

Page 14

Commodities and regions of production are taken into account when constructing our
costs estimates whenever there are data which allow us to do so.
Comment 16) Several commenters state that FDA needs to account for travel and
staff/lab time in the costs of water testing.
Response 16) We explicitly account for these costs in the original analysis. Table
43 estimates the 0.5 hours of farm labor and 1 hour of laboratory travel time labor per
sample (Ref. 6) This represented a total cost of a single water test of $87.30; for the final
analysis we have increased this estimate to $110 per sample. The hourly estimate is
retained in the final analysis; however, wage rates have changed from those presented in
the PRIA.
Comment 17) One commenter states that FDA underestimates the costs associated
with subpart E (Agricultural Water), and offers their own estimation, which states that the
minimum cost for compliance with the rule, including testing and the associated, time,
labor and other incurred costs, would be $7,912 for a single surface water supply source
(regardless of farm size). They state that FDA’s initial economic estimate for a very small
farm was $4,697, which was less than 60 percent of the cost they estimated.
Response 17) We have re-evaluated the costs associated with Subpart E,
Agricultural Water. Our final estimate indicates that water testing will cost an average of
$1,058 per year. While this is somewhat below the commenter’s average costs, we
believe it represents the most accurate estimate utilizing the most recent and applicable
data sources.
Comment 18) Several commenters express concern that the costs of water testing
requirements will fall disproportionately on small farmers and farms in remote areas. For

Page 15

example, it may be more costly for a single-operator farm to spend time on testing.
Farms in remote areas may have trouble accessing a lab, and may need to pay extra
expenses to ship samples to far away labs.
Response 18) We include the cost of shipping samples to labs when one is not
nearby. We then average the costs of a local laboratory sample and shipped sample
together to produce one average cost of laboratory testing across farms. The original
estimate was provided in Table 43 of the PRIA (Ref. 6;7) and is retained here in the final
analysis. See also section IV.G. of the final rule where we address comments about
reducing burden on small farms.
Comment 19) One commenter states that this rule “will impose substantial
economic burdens upon American citizens which will not be imposed upon foreign
producers. Consequently, foreign produce will be less expensive than produce grown in
the United States.”
Response 19) This rule applies equally to domestically-produced and imported
produce. Covered entities in the United States and abroad must adhere to the same
standards. As such, we do not agree that it will disadvantage United States farms as
compared to foreign farms.
With respect to enforcement, FDA intends to use the resources at its disposal to
ensure that both domestic and foreign producers are following the requirements of the
rule. As discussed in Subpart Q of the rule, our strategy to ensure the safety of produce,
both domestically produced and originating from foreign farms, will focus on education,
training, and guidance to achieve compliance. This will include outreach to foreign
governments. We will also work to provide education and assistance in local languages to

Page 16

reach farmers exporting covered produce into the United States, including by working
with organizations and other sources of information that are familiar and accessible to the
produce farming community (such as alliances, international organizations, universities,
trade associations, foreign partners, Joint Institute of Food Safety And Nutrition, and
federal agencies (such as United States Agency for International Development and
United States Department of Agriculture), among others).
Inspections will also play a key role. Under the FD&C Act, FDA has authority to
inspect produce farms and can take enforcement action when needed, such as to prevent
significant hazards from entering the food supply or in response to produce safety
problems. While FDA is not in a position to inspect every foreign farm that produces
food for consumption in the United States, the inspections FDA is able to conduct will be
bolstered by other efforts, such as the final Foreign Supplier Verification Program rule
establishing subpart L of 21 CFR part 1. The FSVP regulation establishes requirements
for importers to verify that imported food (including produce) is produced in compliance
with the produce safety regulation or is produced in accordance with processes and
procedures that ensure the same level of public health protection as is required in the
United States.
Comment 20) One commenter references data from the USDA, which estimated
that the average net farm income for farmers nationally was 10 percent of sales in 2011,
and argues that the estimation implies that for a farm with less than $250,000 in annual
sales, complying with the Produce Safety rule requirements may consume more than half
of their profits.

Page 17

Response 20) We have found sources from the USDA that confirm the fact that,
for many farms, farming is not the primary source of income (Ref. 8), and that, in general,
roughly 90 percent of farm income comes from off farm sources (Ref. 9). However,
these statistics refer to total farm income, while our cost estimates are based on sales of
produce. We do not include any other farm income sources in our estimations of farms
that are covered by this rulemaking; produce sales alone are what determines coverage
throughout the analysis.
Comment 21) One comment suggests that FDA has not considered the fact that
FSMA regulations are different from USDA GAP (or other third party) audits. Some
suggest that FDA allow the use of GAP.
Response 21) See section IV.F. of the final rule where we address comments
about existing industry guidelines and certification programs. Where requirements are
different for farms already performing GAPs we have estimated the cost for a change in
practice. However, if farms are already conducting the required activities through GAPs
or some other agreement, we have attempted to remove previously incurred costs out of
our analysis.
Comment 22) One comment states that FDA’s cost analysis does not differentiate
between costs across crops or across production regions.
Response 22) This is true. For a national analysis of the costs and benefits of this
rule we are not able to differentiate farms by commodities or agricultural regions. We are
aware that there are differences in needs and resources across different farms, and as such
we attempt to provide a national average estimate that reflects this variety. Although, the
costs are not differentiable by these factors both commodities and regions of production

Page 18

are taken into account when constructing our costs estimates whenever there are data
which allow us to do so.
Comment 23) One commenter states that FDA’s estimates do not match with
current average costs for the produce sector, and cite things such as “outdated wage rates
and inconsistent application of wage rates throughout, and “a lack of cost estimates for
replacing tools and equipment that were not able to be brought into compliance with
FDA’s proposed rule.” Another commenter offers an alternative estimation based on
more recent BLS data.
Response 23) In an attempt to more accurately reflect the true costs to farms,
FDA has updated its wage rates to 2013 levels according to the BLS. Additionally, we
now apply a one hundred percent overhead to all wages to more accurately account for
the indirect costs of labor which may be incurred. The rule requires that certain
tools/equipment must be of adequate design, construction, and workmanship to enable
them to be adequately cleaned and properly maintained, and requires keeping
tools/equipment clean and in sanitary condition. We expect the replacement of
tools/equipment as a result of this rule to be rare, however, as such requirements are
sufficiently flexible to accommodate many types of equipment and tools.
Comment 24)

A few commenters offer their own estimates of the costs of the

produce rule. They state that these estimates are based in “more accurate and current
data,” and on their own independent research (e.g., interviews). Specifically, they assume
that: 1) labor costs are higher, based on updating wage rates from 2000 to 2012; 2)
average cost of water sampling is higher, based on a higher expected cost of analysis; 3)
covered farms would test their water more frequently (weekly), based on a higher

Page 19

expected frequency of irrigation; 4) large farms have two irrigation water distribution
systems to inspect, based on the assumption that larger farms may have more irrigation
facilities than smaller ones; 5) farm owners or managers are responsible for
recordkeeping, due to potential liability issues; 6) record keeping hours are much longer,
based on interviews with industry associations; and 7) the time per acre it takes to comply
with the rule is higher, based on the fact that FDA’s costs are calculated using an
expected minimum that does not apply to all farms. Overall, these commenters state that
FDA needs to perform a more detailed, crop-specific analysis, and not make
generalizations for all products and regions. They also suggest that a sensitivity analysis
could be beneficial.
Response 24) These analyses provide a number of suggestions for improving the
analysis and we have incorporated changes where the data were nationally applicable and
relevant. Additionally, we do provide a sensitivity analysis both in this document and in
the original PRIA. In response to the individual suggestions: 1) we have updated wage
rates to 2013, which more accurately reflects the costs that may be incurred by farmers
today; 2) similar to the 25 percent increase in wage rates (from 50 percent overhead to
100 percent), we have increased the estimated cost of a single water test by
approximately 25 percent; 3) the weekly testing frequency originally proposed for certain
water sources and uses in the 2013 proposed rule have been removed from the final water
testing requirements in favor of a tiered testing frequency that results in less frequent
testing; therefore we do not estimate that any weekly water testing will occur; 4) we have
doubled the time estimated for large farms to inspect their agricultural water systems; 5)
while it is true that the owner, operator, or agent in charge of the farm will be responsible

Page 20

for keeping records, we believe that the actual people creating the records will typically
be the farm’s workers; 6) while some records may take longer to produce from scratch,
we believe, based on a study of industry recordkeeping practices from Economics
Research Group (Ref. 7) that our estimated recordkeeping burden is close to accurate;
and 7) we believe the time costs estimated throughout the document represent a
reasonable average by estimated farm size. Finally, it should be noted that a crop by crop
analysis was not feasible given the large number of individual crops covered and the
nature of farms that grow multiple crops on the same acres; therefore, we believe our
approach, estimating costs to the average covered farm based on inputs, is the most
logical way to estimate compliance costs with this rule.
Comment 25) One commenter states that on page 6 of the report, the Farm
Supervisor Mean Wage Rate is calculated as $30.26 per hour, while in the section on
agricultural water testing, a wage rate of $30.83 per hour is used instead.
Response 25) We have simplified our analysis to incorporate only those wage
rates discussed in section 2. Additionally we have updated wage rates to 2013, which
more accurately reflects the costs that may be incurred by farmers today.
Comment 26) One commenter states that the probability of other significant
events that could impact produce farms and create a need to prevent contamination from
sewage is ignored. For example, the commenter notes hurricanes and tornadoes could
both generate problems with sewage and septic systems, but the cost of monitoring after
these events is not included.
Response 26) We agree that these events can have a significant impact on the
actions a farm may take to prevent contamination of their produce. Our analysis of the

Page 21

cost of the rule, however, takes into account average current farming practices. We are
not able to estimate the probability of a natural disaster followed by the expected cost of
contamination reduction.
Comment 27) According to one commenter, FDA assumes that it takes one
minute to clean and sanitize one tool, and there is one tool per farm job, but sometimes
more than one tool is used or it takes longer than 1 minute to clean the tool.
Response 27) For a national analysis of the costs and benefits of this rule, we are
not able to differentiate our estimates based on individual cases (i.e., individual jobs). We
are aware that there are differences in needs and resources across different jobs, and as
such we attempt to provide a national average estimate that reflects this variety. While
some tools may take longer to clean, others will take a much shorter time, and certain
jobs may not even require a tool at all (e.g., harvesting by hand).
Comment 28) One commenter stated that feedback from several produce industry
groups suggests that their crops would require additional irrigation beyond 0.77 acre feet
per growing season, and that the amount of water needed from planting to harvest varies
significantly by crop.
Response 28) In Table 49 of the original PRIA, we estimate that it takes
approximately 2.16 acre/ft. of water to irrigate a single acre using direct water application
techniques. Because this estimate comes directly from the 2008 Farm and Ranch
Irrigation Survey (FRIS), we retain it in the final analysis (Ref. 10). Additionally,
because we do not explicitly examine farms by crop, but are instead tasked with
providing the average costs to all farms operating within the US, applying crop-based
differences to operations is not possible for this analysis. Finally, our estimate is very

Page 22

similar to that found by the U.S. Geological Survey, which states the national average
application rate for irrigated water in 2005 was 2.35 acre-feet per acre (Ref. 11). This
estimate is not preferred because it is not as current, but it provides further support for
our retained estimate.
Comment 29) One commenter states that FDA’s estimates of the number of not
covered and exempt farms by sales class is difficult to verify and analyze because the
data does not come from a publicly available source.
Response 29) We get our data to estimate the number of not covered and exempt
farms from the National Agricultural Statistics Service’s Census of Agriculture, which is
publicly available. Summary tables are available at the Census of Agriculture’s website
(Ref. 12), which allow the public to see the data in summary format. Anyone can apply
for access to the micro-data (Ref. 13), which will allow for a full, independent analysis.
Due to data restrictions and disclosure concerns, we are not able to provide the full data
set ourselves.
Comment 30) One commenter suggests that FDA should consider using a value of
eight hours of additional training in food safety, which greatly increases the cost.
Response 30) Table 112 from the PRIA estimates that farm operators are involved
in food safety training for a total of eight hours, seven in training and one additional for
travel time. These time estimates are retained in the final analysis; however, wage rates
have been updated to more accurately reflect the current state of the industry. We do not
believe that it will be necessary to further train each worker for eight hours in food safety,
once the manager/operator has received the more comprehensive training.
Comment 31) One commenter asks how FDA will determine if a farm is exempt.

Page 23

Response 31) We are adding a new provision § 112.7 to establish certain
recordkeeping requirements in relation to a qualified exemption. Records required under
this provision will assist farms in determining whether they are eligible for a qualified
exemption and will assist FDA in verifying eligibility.
Comment 32) Several commenters state that a specific type of produce (e.g.,
apples) has never been associated with food borne illness outbreaks, which means that, in
the case of this type of produce, in the commenters’ view, there are no benefits from the
rule. Some suggest that FDA should look at comparative benefits by type of produce.
Others say that grouping high and low risk commodities together in our analysis distorts
the risk, and therefore the benefits estimation. In addition, several commenters state that a
specific part of the rule (e.g., agricultural water testing) will provide no benefit.
Response 32) Although certain commodities have never been implicated in an
outbreak during the time period analyzed, there are numerous outbreaks which occurred
in association with produce commodities that had previously not been implicated in an
outbreak. These cases are of great public health concern and failing to take into account
the sporadic nature of foodborne illness may miss a large potential threat to public health.
Table 8 provides a pathway specific breakdown of the implicated causes of outbreak
illnesses. Additionally, the rule focuses on the potential routes of contamination of
produce, and covers specific practices, procedures, and processes on a farm, all of which
may present significant risk, regardless of the commodity grown, harvested, packed, or
held at the farm. See discussion in section IV of the rule.
Comment 33) Several commenters state that FDA has not provided “real”
evidence of a public benefit to this rule.

Page 24

Response 33) The estimation of benefits are based on the most accurate and upto-date data on produce related foodborne illness. Additionally, the estimates of
effectiveness are based on a number of studies, citing experts in produce related
foodborne illness, which all point to these safety measures having a measureable effect
on the number of produce related foodborne illness.
Comment 34) Many sources state that FDA hasn’t done a cost-benefit analysis for
the supplier program. Comments suggest that FDA doesn’t present any information as to
how that program will affect farms, especially those already affected by the produce rule.
Response 34) We interpret these comments to be referring to requirements of the
PCHF and FSVP rules, not this produce safety rule. There are only a few specific
requirements in this rule that relate to entities in a farm’s supply chain other than the farm
itself, and we do not consider any of these requirements to constitute a “supplier program.”
The relevant provisions are: § 112.2(b)(2) for produce eligible for exemption because it
receives commercial processing to adequately reduce pathogens (requiring certain
disclosures to, and written assurances from, a farm’s customers related to such
processing); § 112.60(b)(1) for treated biological soil amendments of animal origin
received from third parties (requiring covered farms to keep certain documentation
related to the third party’s treatment and handling of such materials); § 112.142(b)(2)
relating to seeds or beans used for sprouting that may be contaminated with a pathogen
(requiring sprouting operations to report that information to seed/bean suppliers under
certain circumstances); and §§ 112.142(e) and 112.150(b)(1) allowing sprouting
operations to rely on prior treatment of seeds or beans for sprouting conducted by a
grower, distributor, or supplier with appropriate documentation. The costs and benefits of

Page 25

these provisions have been included in our analysis for this rule. The costs and benefits
associated with the supplier programs in FDA’s PCHF and FSVP regulations are
discussed in the FRIAs related to those rules.
Comment 35) Many cite the benefits of diversification, and say FSMA should
incentivize diversification, not discourage it. Similar comments are made about the
benefits of organic food, rich top soil, etc.
Response 35) While FDA believes there may be benefits to the farmer and
farmland of diversification of crops and organic farming, to our knowledge, there are no
quantifiable impacts on the human health burden associated with produce from these two
activities. Additionally, the primary goal of our integrated approach to this rule was to not
single out any specific crop or to limit diversification of crops in any way. See section
IV.I. of the rule.
Comment 36) One comment states that no real cost-benefit analysis has been done
because we perform a qualitative risk analysis. This comment further suggests that we
have not complied with Executive Order 13563, which directs agencies to assess all costs
and benefits of available regulatory alternatives.
Response 36) The Qualitative Assessment of Risk (QAR) is only one piece of
information that helped to inform both the rule and the quantified estimation of benefits.
FDA believes that we have fulfilled all the requirements for a complete regulatory impact
analysis required under the pertinent Executive Orders.
Comment 37) Several commenters suggest that FDA significantly overestimated
the benefits of the proposed rule, and made “unjustified leaps of logic”. Specifically,
they state that applying Scallan’s multiplier to estimate foodborne illness leads to an

Page 26

overestimation of foodborne illnesses attributed to produce, and that our estimates were
significantly higher than Scallan’s (Ref. 31). They suggest that FDA’s use of this
multiplier is unjustified, and that we should look at more than one study. They also
criticize FDA’s use of a “shaky survey” to estimate the effectiveness of the rule, as well
as the fact that FDA extrapolates to all produce some results based on the leafy greens
and tomato industries, which are associated with the highest number of outbreaks.
Response 37) FDA does not believe that it has overestimated the benefits of this
rule. We acknowledge that some assumptions were made when data were less than robust,
specifically when estimating the ‘unidentified’ burden of illnesses. To alleviate this
concern we provide a more conservative estimate, which reduces our estimated number
of unidentified illnesses. To get this number, we multiply the total number of estimated
preventable illnesses attributable to FDA regulated produce by 4 to obtain a number of
unidentified illnesses which is consistent with Scallan, et al., who estimate that
unidentified illnesses make up about 80% of all foodborne illnesses. Additionally, we
only implicitly, not directly, apply Scallan et al.’s multiplier in the estimation of
quantified benefits. We use only the annual incidence of foodborne illness by pathogen
to compute the number of annual illnesses associated with produce, although this does
implicitly have a pathogen multiplier that is estimated by Scallan using active and passive
surveillance.
Comment 38) One commenter suggests that FDA has overestimated the benefits
of the rule, and proposes omitting Fresh Cut produce from the benefits, as well as
unidentified illnesses, which may be “too speculative.” They offer their own estimates,
which suggest that the costs will overtake the benefits with the omission of Fresh Cut and

Page 27

unidentified illnesses. Other commenters recommend removing Fresh Cut produce from
the estimation of illnesses due to RACs, and state that Fresh Cut produce most likely is
contaminated outside of the farm and in the processing facility.
Response 38) FDA agrees that Fresh Cut should be omitted from the benefits
analysis of the produce rule. We have, therefore, moved Fresh Cut from this FRIA
related to the produce safety rule to the cost-benefit analysis related to PCHF rule. In
terms of the unidentified illnesses, we have refined our estimation to be more
conservative in terms of the number of unidentified illnesses.

However, we have

included an alternative calculation of benefits without unidentified illness in Table 11,
which shows that omitting unidentified illnesses does not drastically change the benefits,
and does not cause the costs to overtake the benefits.
Comment 39) One commenter states that many covered farms in North Carolina
have made significant capital outlays in equipment appropriate to the scale of their
operations, and will incur significant expenses in order to retrofit existing infrastructure.
The commenter requests that FDA grandfather capital equipment for an additional seven
years.
Response 39) We realize that replacing capital equipment, which typically has a
long lifetime, would pose a significant burden to farmers; however, the rule has been
written in a way that we expect to minimize such needs. The rule is not prescriptive as to
the nature of tools or equipment used in covered activities by covered farms and,
therefore, as long as relevant tools and equipment are of adequate design, construction,
and workmanship to enable them to be adequately clean and properly maintained, it will
not be necessary to replace a farm’s tools or equipment to comply with this rule. To that

Page 28

end we have estimated the cost of cleaning current capital equipment, rather than the
replacement value. Additionally, to provide increased flexibility to all farms, we stagger
compliance dates (see section XXIV of the rule).
Comment 40 ) One commenter states that the PRIA should reflect net profit
instead of sales.
Response 40) We prefer sales rather than net profit because sales data serve as a
proxy for total produce volume on a farm. Although we realize this is an imperfect
measure, net profits could significantly understate the volume of food leaving any
particular farm. Additionally, data on sales is easily observable and shared by many
farmers, where information on profits is not.

II. Final Regulatory Impact Analysis
A. Background
Table 2 presents a side-by-side comparison of the estimated costs of the proposed
rule and updated estimated costs of the final rule. To present a valid comparison, we
have updated the (previously published) estimated costs of the proposed rule using the
latest data and techniques. Estimated total steady state costs to domestic operations,
using a 7 percent discount rate over 10 years, are $530 million for the proposed rule, and
$560 million for the final rule.
Table 2. Comparison of Costs of the Rulemaking across Data Sources (in millions)
Original Analysis
With Updated Data

Cost Sections
Personnel Qualifications and training

$124.12

Final Analysis
$187.38

Health and Hygiene

$141.87

$135.61

Agricultural water

$58.94

$37.07

Page 29

Biological soil amendments of animal origin

$2.47

Domesticated and wild animals

$9.19
$37.78

$15.86

Growing, harvesting, packing, and holding activities

$0.52

$2.25

Equipment, tools, buildings, and sanitation

$72.99

$118.69

Sprouting operations

$7.51

$6.77

Recordkeeping

$40.18

$27.49

Administrative cost to learn the rule

$34.31

$23.25

Corrective steps

$2.01

$3.25

Variances

$0.10

$0.11

Total Costs (annual in millions)

$529.51

$560.19

Net present value (7 percent)

$2,929

$15,992

$417
$366
Annualized costs (7 percent)
Note: This table utilizes two different timing scenarios when calculating NPV. For the original analysis
with updated data large farms are given an extra year for compliance, small farms are given two years, and
very small farms are given three. The timing for the current analysis is more complex, and fully laid out in
Table 4 of this analysis. Additionally the new timing allows farms more time to implement requirements,
thus lessening the burden when discounted.

Using the steady-state comparison illustrated in Table 2, the final rule has
estimated costs ($560.16 million annually) that are 21.9 percent higher than the estimated
costs of the proposed rule ($459.56 million annually). This 21.9 percent increase in
estimated costs is attributable to the changes in the provisions of the rule from the
proposal to the final stage. Between the publication of the proposed rule and the final
rule, however, we updated some of the data and techniques used to estimate costs. We
have updated wage data, updated the way we account for overhead costs in relation to
wages, updated data on the number of operations affected by the rule, and we adopted
new techniques for modeling some of the provisions, based on comments and other
information gathered since the publication of the proposed rule. The published estimate
of the annualized costs of the proposed rule was $459.56 million using a 7 percent
discount rate (Ref. 6) The adjusted estimate of $529.51 million in annual costs of the
proposed rule in Table 2 above reflects a 15.2 percent increase compared to the previous

Page 30

estimate, and this 15.2 percent increase is attributable to changes in the data and
techniques used in our cost estimation, not changes in the provisions of the rule.
One significant cause for the increase in our estimated steady state cost is the
change in our estimate of costs of labor hours. Following DHHS guidelines, we
corrected our estimate for computing overhead costs to include a 100 percent adjustment
relative to the money wage, rather than the 50 percent adjustment used in the original
estimates. New DHHS guidelines, for computing labor costs recommend (based on
general industry data) benefits plus other overhead costs equal 100 percent of pre-tax
wages (Ref. 14). This correction results in a roughly 13.3 percent ($66 million) increase
in estimated costs. We also updated the base year for computing wage rates from 2010 to
2013, the most recent year for which the Bureau of Labor Statistics has complete wage
rate data. This update alone results in a 2.9 percent ($15.8 million) increase in costs. The
sum effect of the two updates to the wage estimates results in a roughly 16 percent ($81.8
million) change in estimated annualized costs.
We obtained more recent data for the farm count from the USDA, National
Agriculture Statistical Service’s (NASS) 2012 Census of Agriculture (Ref. 15) Our
estimate of the total farms covered decreases from the 40,496 estimated in 2007 to 35,029
using the latest census numbers. The new farm count results in a 9 percent (roughly
$55.2 million) net decrease in costs.
Based on data and information gathered from and in response to public comments,
as well as other new sources, we changed the way we modeled the cost estimates of a
number of provisions. For example, we have increased the estimate for the number of
operational days where the harvested or harvestable portion of produce is exposed,

Page 31

increased the time estimated to inspect agricultural water sources and systems, decreased
the time estimated for farms not covered or eligible for a qualified exemption to read and
learn about the rule, and increased the average estimated cost of environmental testing for
sprout operations and water testing for all covered farms. In addition, some of the
proposed provisions in the 2013 proposed rule and the supplemental notice have changed
for this final rulemaking. For example, the inclusion of an allowance for microbial die-off
in relation to use of agricultural water during growing of covered produce (other than
sprouts) using a direct water application method, has allowed us to reduce some of the
burden to farmers. These adjustments led to changes in total estimated costs. The net
effect of all of these changes from the proposed rule is a roughly 16.1 percent increase
(almost $73.4 million) in total estimated costs.
The combined effect of updating and correcting our method for estimating
overhead costs, using the most recent baseline for calculating wage rates, the most recent
farm count, and other adjustments to estimates based on public comment and changes to
the regulatory requirements, change the steady state estimate of total domestic costs of
the proposed rule from approximately $459.56 million (the originally published estimate
with no update to wages or data) to $560.16 million, a 21.9 percent increase.
We use the revised wage rates, most recent base year, the revised farm count, and
other adjustments throughout our analysis of the final rule.
The estimated benefits of the proposed rule and the updated estimated benefits of
the final rule also differ. In all, the estimated number of prevented illnesses decreases by
about two-thirds from the proposed rule to the final rule, while the total estimated
benefits increase by about one-third. This somewhat counterintuitive change is due to an

Page 32

increase in the dollar costs of illnesses, combined with new data and estimation methods
for the number of illnesses.
The final rule uses a higher VSL and QALD than the proposed rule. The new
VSL values are taken from Robinson and Hammitt (2015) (Ref. 16). They present a VSL
of $9 million and a QALD value of $1,260, whereas the proposed rule uses a VSL of $7.9
million and a QALD value of $586. The updated values of both QALD and VSL lead to
increases in the quantified burden of illnesses. The increase in QALD implies
particularly large increases for illnesses that last for long periods of time, while the
increase in VSL leads to greater increases when the percentage fatality rate associated
with a particular illness is high.
An increase in data range, combined with a more conservative estimate of
unidentified goods, leads to an increase in more burdensome illnesses, but a decrease in
less burdensome illnesses (i.e., unidentified illnesses). The data used in the final rule
covers 2003 through 2012, while the data in the proposed rule only covers 2003 through
2008. Because 2008 through 2012 saw the relative incidence of outbreaks associated
with produce RACs rise, our estimated number of illnesses, which is based on the ratio of
reported FDA-regulated produce RAC outbreaks to total CDC identified illnesses in the
same time period, increased. This increase, however, was somewhat offset by the large
decrease in unidentified illnesses. In the final rule, we employ the more conservative
estimate, of the two published in the original analysis, of unidentified illnesses, which
have a very low estimated cost per illness. This change strictly drives the number of
unidentified illnesses down. We also omit outbreak illnesses associated with Fresh Cut

Page 33

products, as they are now addressed in the Preventive Controls rule, which further
reduces the estimated number of illnesses.

B. Need for Regulation
The need for this rule stems from a market failure caused by the asymmetric
information associated with the safe production and consumption of raw agricultural
commodities that are covered produce. If covered farms do not apply the socially optimal
level of food safety practices, they create a potentially harmful situation for consumers,
which is largely unobservable to consumers. There is not a sufficiently significant direct
link between poor produce safety practices and food-related illnesses, which suggests that
food safety is not an experience good (product for which characteristics, such as quality
or price, are difficult to observe in advance, but can be ascertained upon consumption);
with rare exceptions, the link between consumption of raw agricultural products and
experiencing a food-related illness cannot be determined by consumers.
This final rule aims to reduce the effects of the information asymmetry by
requiring certain science-based minimum standards for the safe growing, harvesting,
packing, and holding of covered produce across all covered farms, thereby reducing
foodborne illnesses from this source.
Using a science-based framework we characterize the magnitude of the public
health risks associated with the consumption of produce, and establish specific standards
that address the risks of microbial contamination from significant agricultural inputs
(labor, water, biological soil amendments of animal origin, and tools and equipment),
unsanitary conditions in buildings, and wild and domesticated animals, as well as the
risks of microbial contamination in the production of sprouts. We provide a framework to
Page 34

evaluate the effectiveness of the rule for addressing the public health risks associated
with biological hazards in produce.
We define thresholds for different farm size categories that will be covered, with
each farm size category linked to a quantitatively defined level of public exposure to risk.
We estimate the costs of each provision by farm size.
The rule also responds to lower-than-socially-optimal private incentives to
provide safe practices. These are a result of uncertainties in the individual farm’s
understanding of the magnitude of the public health risk from the consumption of
produce grown on their farm, as well as the effectiveness of measures and controls at
addressing that risk. At this point in time, public health surveillance is often unable to
determine whether an illness resulted from a foodborne pathogen or which particular food
or food category may have served as the vehicle for the pathogen that caused the illness.
It is also frequently unable to identify the specific farm or practice implicated in a
produce-associated outbreak. This may result in the underestimation by producers of the
costs to society from consuming produce and may cause them to discount the value of
food safety practices and to provide less-than-the-socially optimal amount.
In addition, this rule responds to a statutory mandate in Section 419 of the Federal
Food, Drug, and Cosmetic Act requiring that the Secretary of HHS adopt a regulation
setting forth those procedures, processes, and practices that the Secretary determines to
minimize the risk of serious adverse health consequences or death, including those
determined to be reasonably necessary to prevent the introduction of known or
reasonably foreseeable hazards into fruits and vegetables, and to provide reasonable

Page 35

assurances that the produce is not adulterated under Section 402 of the Federal Food,
Drug, and Cosmetic Act.

C. Purpose of the Rule
The rule establishes science-based minimum standards for the safe growing,
harvesting, packing, and holding of produce on farms. The rule addresses microbiological
risks from certain routes of contamination, including workers, agricultural water,
biological soil amendments of animal origin, and tools and equipment, unsanitary
conditions in buildings, and wild and domesticated animals during growing, harvesting,
packing, and holding activities of covered produce, including sprouts.

D. Inputs and Assumptions
The following section outlines some of the standard information utilized
throughout the remainder of the analysis. First, we present all standard cost estimates and
assumptions that allow us to calculate the costs of implementation at the farm level. This
section includes things like standard labor costs and data sets used to inform estimates
and assumptions. Next, we provide information on the coverage of the analysis and how
it relates to the US produce industry as a whole. Finally, we provide some information on
the timing of both costs and benefits of this regulation. Detailed discussion of how these
estimates and data are used to estimate industry costs are included in the detailed analysis
of costs section.
Measuring Costs

Page 36

We measure costs based on the best available information from government,
industry, and academic sources. We list some common conventions used throughout the
cost analysis here.
All wage rates used come from the Bureau of Labor Statistics (BLS),
Occupational Employment Statistics, May 2013, National Industry-Specific Occupational
Employment and Wage Estimates, under NAICS 11 – Agriculture, Forestry, Fishing, and
Hunting (Ref.17). Wages are increased by 100 percent to account for overhead.
•

Farm Operator or Manager Mean Wage Rate: Our estimate for the mean hourly
wage rate for a farm operator or manager is $72.12 including fringe benefits and
other overhead. We derive our estimate from the BLS mean hourly wage rate for
Farmers, Ranchers, and Other Agricultural Managers working in the agriculture
industry as shown in (Ref.17) of $36.06 and we add 100 percent for fringe
benefits and other overhead costs ($36.06) for a total estimate of $72.12.

•

Farm Supervisor Mean Wage Rate: Our estimate for the mean hourly wage rate
for farm supervisors is $42.74 including fringe benefits and other overhead.

We

derive our estimate from the BLS mean hourly wage rate for First-Line
Supervisors of Farming, Fishing, and Forestry Workers as shown in (Ref.17) of
$21.37 and we add 100 percent for fringe benefits and other overhead costs
($21.37) for a total estimate of $42.74
•

Farm Worker (Nonsupervisory) Mean Wage Rate: Our estimate for the mean
hourly wage rate for farm workers (nonsupervisory) is $18.56 including fringe
benefits and other overhead. We derive our estimate from the BLS mean hourly
wage rate for Farmworkers and Laborers, Crop, Nursery, and Greenhouse as

Page 37

shown in (Ref.17) of $9.28 and we add 100 percent for fringe benefits and other
overhead costs ($9.28) for a total estimate of $18.56.
We use the 2012 Census of Agriculture farm-level database to derive the total
number of domestic farms (including greenhouses) that grow produce, the number of
produce acres operated, the amount of labor employed, and their food sales; to estimate
the number of farms that are eligible for the qualified exemption created by section 419(f)
of the FD&C Act; and to create estimates of the rates of specific food safety practices
currently being undertaken by farms (current industry practices). (Ref.18)
We use FDA’s Operational and Administrative System for Import Support
(OASIS) database to estimate the number of foreign farms that will be covered by the
rule. (Ref.19)
We use the following surveys and literature where possible to create estimates of
the rates of specific food safety practices currently being undertaken by farms (current
industry practices):
•

1999 Fruit and Vegetable Agricultural Practices Survey (FVAP) (Ref.20)

•

Farm Food Safety Practices: A Survey of New England Growers (Ref.21)

•

Growers’ Compliance Costs for the Leafy Greens Marketing Agreement and
Other Food Safety Programs (Ref.22)

•

USDA Agricultural Marketing Service (AMS) Fresh Produce Audit Verification
Program, including commodity-specific audits for the tomato and mushroom
industries (Ref.23).

•

Food safety regulations and marketing agreements: Florida Tomato Regulation
(Florida Rule 5G-6.011) (Ref.24), and the Leafy Greens Marketing Agreements in

Page 38

California (Ref.25) and Arizona (Ref.26) (together, sometimes referred to as
“LGMA”).
•

National Agricultural Workers Survey (NAWS), U.S. Department of Labor,
Public Access Database, 1989 to 2006, for years 2005 to 2006 to estimate the
number of workers that are employed on multiple farms, and the number of
workers employed by farm task; it is also used to create estimates of the rates of
specific food safety practices currently being undertaken by farms (current
industry practices) (Ref.27)
We annualize any one time costs over 10 years at discount rates of 7 percent and 3

percent. For ease of reading, in the main document, we report only results derived from
the 7 percent discount rate. In the sensitivity analysis and summary sections, we also
report results derived from the 3 percent discount rate
To classify farms that are covered by the rule by size, we identified farms as very
small when they generate over $25K but no more than $250K annually in produce sales,
small when they generate over $250K but no more than $500K annually in produce sales,
and large when they generate more than $500K annually in produce sales.
We estimate that very small farms operate 100 days out of the year where the
edible portion of produce may be exposed, small farms operate 150 days, and large farms
operate 200 days (non-consecutive). 2
We estimate that the farm operator or manager is the person responsible for
training on all farms.

2

This estimate is based on annual planting data from USDA (Ref.18). This estimate is based on annual
planting data from USDA (Ref.18).

Page 39

For the purposes of this analysis, we use the term post-harvest activities to refer to
all covered activities that occur after produce is removed from the growing area. We note
that for the purposes of the rule, the term “harvesting” is broad enough to encompass
some of these activities. We do not use the term “harvesting” in the same sense here but
rather use it to refer only to removing produce from the growing area.
We use FDA’s Evaluation of Recordkeeping Costs for Food Manufacturers,
February 13, 2007, for our estimates for the hours necessary to perform the various
recordkeeping functions, for our estimate of the frequency of recordkeeping by record
type; and the average minutes spent keeping records by record type. Recordkeeping
estimates in this report are based on expert opinion and an extensive literature review
(Ref.7).
Coverage of the Analysis
1. All Farms
The rule applies to covered farms that grow covered produce including fruits and
vegetables such as berries, leafy greens, herbs, and sprouts. It applies equally to farms
located domestically and farms in foreign countries exporting covered produce to the US.
There are approximately 121,116 farms in the U.S. that grow produce for sale excluding
sprouting operations, which we analyze separately (Ref.18). This number was derived
using the 2012 Census of Agriculture and includes farms with on-farm packing,
greenhouses, farms eligible for qualified exemption (§ 112.5), farms that grow covered
produce for commercial processing (§ 112.2(b)), and farms that are not covered by the
rule (§ 112.4). We estimate that there are approximately 475 sprouting operations, which
include farms eligible for qualified exemptions, and sprouting operations that are not

Page 40

covered by the final rule. Sprouting operations will be considered in the sprouts section.
We estimate that there are 70,395 foreign farms that will offer covered produce for
import into the U.S., which includes farms eligible for qualified exemptions, and farms
that are not covered by the final rule (Ref.19). This number was estimated using the
number of foreign produce manufacturers in the OASIS database from fiscal year 2008,
and multiplying it by the ratio of domestic farms to domestic manufacturers in the U.S.
2. Eligibility for Exemption and Corresponding Modified Requirements
The rule identifies certain farms and certain produce that are eligible for
exemptions provided certain requirements are met. The eligibility for an exemption is
established under two criteria: (1) the monetary value of all food sold on the farm and
direct marketing of a portion of the food, and (2) produce that receives commercial
processing that adequately reduces the presence of microorganisms of public health
significance (e.g. a microbial kill-step). Farms, or produce, that qualify for either
exemption are subject to a subset of the administrative provisions of the regulation, which
are discussed in detail in the summary of records section of this analysis.
a. Monetary value of all food sold and direct farm marketing (“Qualified
Exemption”)
Farms are eligible for a qualified exemption if the average value of their food
sales over the last 3 years was less than $500,000 and if more than 50 percent of their
food sales were direct sales to qualified end-users as that term is defined in the rule (see
§§ 112.3(c), 112.5, 112.6, and 112.7). “Food” is defined in § 112.3(c) and Section 201(f)
of the Federal Food, Drug, and Cosmetic Act. In order to estimate the number of farms
that meet this qualification, we use data from the 2012 Census of Agriculture. We

Page 41

estimate that there are approximately 3,134 total farms, including 171 sprouting
operations, eligible for the qualified exemption after accounting for farms that are not
covered, which is explained in part c. of this section, “Coverage of the Analysis”.
b. Commercially processed produce
Produce that is commercially processed in a manner so as to adequately reduce
pathogens is eligible for exemption from the rule provided that certain required steps are
taken (see § 112.2(b)). Processing of low acid or acidified foods (in compliance with
applicable FDA regulations in Parts 113 and 114) and processing of juice (in compliance
with applicable FDA regulations in Part 120) are examples of eligible processing
methods. Produce that is destined for the frozen or fresh-cut markets is typically not
eligible since there is generally no adequate reduction of pathogens in the processing
method.
We estimate the number of farms whose covered produce would qualify for this
exemption using production information, specifically the amount sold to fresh versus
processed markets, available in published reports for citrus, non-citrus, berries,
vegetables, and tree nuts from the 2012 Census of Agriculture (Ref. 15). There are
approximately 3,199 farms whose produce would qualify for this exemption, after
accounting for farms that are not covered, and farms that do not also grow other covered
produce. Farms that grow covered produce that is eligible for the commercial processing
exemption and that also grow other covered produce will be subject to the regulation only
with respect to their other covered produce.
3. Farms and produce not covered

Page 42

Farms not covered by the regulation are those with an average annual monetary
value of produce sold during the previous three-year period of $25,000 or less (see §
112.4). Produce that is rarely consumed raw, such as beets, potatoes, sweet corn, and
sweet potatoes, is also not covered by the rule (the rule includes an exhaustive list of such
produce, from which we have provided only a few examples here) (see § 112.2(a)(1)). A
farm that only grows these commodities, and does not also grow covered produce, will
not be subject to the regulation. Farms that grow these commodities and covered produce
will be subject to the regulation only with respect to their covered produce. Produce for
personal or on-farm consumption is also not covered by the regulation (see § 112.2(a)(2)).
A farm that only grows produce for personal or on-farm consumption, and does not also
grow covered produce, will not be subject to the regulation. Farms that grow produce for
personal or on-farm consumption and covered produce will be subject to the regulation
only with respect to their covered produce.
The USDA National Commission on Small Farms defines a small farm as a
family farm with less than $250,000 total monetary value of food a year (Ref.28). The
Commission’s recommendation was based on the reasoning that these farms are the
likeliest to exit the industry, and have the greatest need to improve net farm incomes
since they receive only 41 percent of all gross sales revenue, but make up 94 percent of
all U.S. farms (Ref.28). We use the $250,000 monetary value of produce threshold for
the upper end of our very small farm category. Covered produce farms below this
threshold make up 17 percent of produce acres, and 87 percent of all produce farms. We
use the monetary value cutoff of $500,000 from the qualified exemption for direct farm
marketing in § 419(f) of the FD&C Act as the upper end of our small farm category.

Page 43

Farms below this $500,000 threshold make up 24 percent of produce acres and 92 percent
of all produce farms. Farms that are not covered because they have no more than
$25,000 in average annual monetary value of produce make up about 5 percent of
produce acres, but 62 percent of all produce farms.

d. Summary of Farms Eligible for Exemption, Farms Not Covered, and Produce
Not Covered
Table 3 shows the total number of domestic farms, the number of covered and
exempt/not covered farms, and a breakdown of the number of farms that are eligible for
a qualified exemption and that are not covered by the rule. All farm numbers are
calculated from the NASS 2012 Census of Agriculture (Ref.18). Not accounting for
sprouts, we estimate that there are a total of 21,666 farms that would be eligible for the
qualified exemption, and 18,381 of those farms generate $25,000 or less in produce sales
and therefore are not covered. Similarly, we estimate that there are a total of 4,153 farms
all of whose covered produce would be eligible for the commercially processed produce
exemption, and 954 of these farms generate $25,000 or less in produce sales and
therefore are not covered. We estimate there are 16,190 farms not covered because they
grow produce that is rarely consumed raw, and 11,518 of those farms generate $25,000 or
less in produce sales and therefore are not covered. Lastly, there are 44,078 farms not
covered under this rule because they generate $25,000 or less in produce sales and
therefore are not covered. After accounting for those farms that are eligible for a qualified
exemption and also generate $25,000 or less in produce sales and therefore are not
covered, we estimate that a total of 86,087 farms (21,666 + 4,153 + 16,190 + 44,078) are

Page 44

not covered under the rule. The numbers for sprouting operations are covered in the
sprouts section.
Table 3. Breakdown of Covered and Exempt Farms
$25K or less
monetary value of
produce produced

very small

small

large

Total

Total Produce Farms
Total Produce Acres

74,931
410,319

30,952
1,050,000

5,128
580,969

10,105
6,380,000

121,116
8,422,103

Qualified exemption farms
% total produce acres
Exempt produce –
commercially processed
% total produce acres
Not covered produce - rarely
consumed raw
% total produce acres

18,381
1%

3,015
1%

270
1%

-

21,666
3%

954

1,991

448

760

4,153

1%

2%

1%

8%

12%

11,518

3,165

454

1053

16,190

1%

2%

1%

9%

14%

-

-

-

2%

-

-

-

2%

-

22,781
7%

3,956
4%

8,292
58%

35,029
70%

Not covered farms –$25,000
or less monetary value of
produce
% total produce acres
Total Covered Farms
% total produce acres

44,078

40,078

The 21,666 ‘qualified exemption’ farms, who have less than $500K in average
annual monetary value of food sales over a rolling 3-year period and sell over half of
their food directly to qualified end-users, account for about 3 percent of all US produce
acreage. The 74,931 farms that generate $25,000 or less in produce sales, account for
only 5 percent of all domestic produce acreage, but for 62 percent of all farms that grow
produce. They have average produce sales of $6,539 per farm and grow an average of
5.5 produce acres. After accounting for farms that would not be covered because they
grow produce that is rarely consumed raw or that receives commercial processing,
qualified exemption farms still account for about 2 percent of all covered domestic
Page 45

produce acreage. After accounting for the farms that are eligible for a qualified
exemption or that grow produce that is rarely consumed raw or commercially processed,
then the leftover 44,078 not covered farms only account for about 2 percent of all
domestic produce acreage. In total, the rule covers about 29 percent of all domestic
produce farms, and about 94 percent of all domestic produce acres that are not dedicated
to growing commodities rarely consumed raw or that will receive commercial processing.
Timing of Costs and Benefits
Because the timing of the rule’s compliance dates varies across provisions, by
farm size, and for sprouts, it is necessary to discount these costs and benefits accordingly,
as neither will be realized immediately. Table 4 presents the timing of all costs and
benefits as they accrue across farm sizes for the first ten years after publication of this
final rule. Zero costs and benefits are estimated to be incurred by covered farms in the
first two years following publication, because all farms are given two years to implement
the provisions of the rule (except with regard to sprouts, discussed separately below). In
addition to this, all small farms are given an additional year and very small farms are
given two additional years to implement the required provisions. Finally, all farms,
regardless of size are given an additional two years from their specific compliance date to
implement certain required water provisions (except with regard to sprouts).
In addition, the timing for sprout operations is different from other farms. Large
sprouting operations have one year to comply with the rule, small sprouting operations
have two years, and very small sprouting operations have three years, with no additional
time for any particular provisions.

Page 46

Finally, qualified exempt farms will have to begin complying with the record
retention requirement for records supporting eligibility in § 112.7(b) upon the effective
date of the rule, and with the modified requirement in § 112.6(b)(1) on January 1, 2020.
Otherwise, qualified exempt small farms will have three years to comply with the
remaining modified requirements in §§ 112.6 and 112.7, and very small qualified exempt
farms will have four years. We do not explicitly estimate a cost to keeping the records
required by 112.7(b), as we expect that such records would be kept under normal
business practices.

Year 6

Year 7

Year 8

Year 9

Year 10

FCB

FCB

FCB

FCB

FCB

FCB

FCB

Year 5

CBLW

Full
Costs/
Benefits
(FCB)

Year 4

Year 3

Year1

Farms

Year 2

Table 4. Timing of Produce Costs and Benefits

--

--

Costs/
Benefits
Less
Water1
(CBLW)

--

Costs/
Benefits
Less
Water1
(CBLW)

CBLW

Full
Costs/
Benefits
(FCB)

CBLW

Full
Costs/
Benefits
(FCB)

FCB

FCB

FCB

FCB

FCB

FCB

FCB

FCB

FCB

FCB

FCB

Covered Farms
Very
Small

Small

--

--

--

--

Costs/
Benefits
--Less
Large
Water(C
BLW)
Covered Sprout Operations
Very
Small
Sprouts
Small
Sprouts

Large
Sprouts

--

--

Full
Costs/
Benefits
(FCB)

--

--

Full
Costs/
Benefits
(FCB)

FCB

FCB

FCB

FCB

FCB

FCB

FCB

--

Full
Costs/
Benefits
(FCB)

FCB

FCB

FCB

FCB

FCB

FCB

FCB

FCB

--

Page 47

Exempt Farms
Very
-Small
Exempt
Small
Exempt

--

--

--

--

Full
Costs
(FC)

FC

FC

FC

FC

FC

--

--

Full
Costs
(FC)

FC

FC

FC

FC

FC

FC

Full
Costs
FC
FC
FC
FC
FC
FC
FC
(FC)
Note: Certain water testing-related provisions are delayed by two years from initial compliance dates.
Large
Exempt

--

--

Throughout the remainder of this document, we estimate the annual costs of
compliance across farm sizes and provisions, as well as the benefits that are likely to
occur; these are the primary estimates presented in the benefits or specific costs
calculations. Following this, to reflect the nature of the way these costs and benefits will
be realized, we take a net present value (NPV) over these 10 years for both costs and
benefits, and we annualize them according to the table above, using both a 3 and 7
percent discount rate. Both costs and benefits are discounted in the same manner to
provide easily comparable annualized estimates.

E. Benefits of the Rule
The primary benefits of the provisions in this rule are an expected decrease in the
incidence of illnesses relating to produce from microbial contamination. For the purpose
of this analysis, we develop a conceptual framework that describes how implementing
this rule is likely to reduce the level of foodborne illness.
1. Baseline Risk of Foodborne Illness 3

3

The estimated burden of illness and subsequent estimations of rule benefits include illnesses occurring in
the U.S. tied to imported produce. We do not attempt to estimate the benefits that would accrue due to the
mitigation of produce related illness in other countries due to improvements in the safety of U.S. exports or
produce grown and consumed in other countries on farms covered by the rule. A rough estimate of costs

Page 48

To estimate the number of baseline illnesses attributable to produce from
microbial contamination only, we begin with only those outbreaks we can directly
attribute to FDA-regulated produce that has suffered microbial contamination. Table 5
presents all outbreaks, organized by produce commodity and pathogen, which can be
linked to microbial contamination of produce raw agricultural commodities (RAC) other
than sprouts, and sprouts (treated separately), based on illnesses recorded in FDA’s
outbreak database (Ref. 29). This does not include Fresh Cut (FC), which are not RACs.
In total, there are 69 outbreaks, 7,050 illnesses, and 46 deaths in the FDA database
attributable to FDA-related produce. This averages out to about 7 outbreaks, 705
illnesses, and 4.6 deaths per year observed in the outbreak database.
The data span of 2003-2012 is utilized for this analysis because it represents the
most current, and comprehensive data available. We are unable to look at years beyond
2012, because the full outbreak data, from CDC, has not been completely collected,
sorted, cleaned, and made available for public use. Additionally, collection methods by
both FDA and CDC have improved vastly in recent years, and data further back may be
more subject to underreporting biases. It is important to note that our data span differs
from that of the PRIA (Ref. 6), which uses the years, 2003-2008. This drives up the raw
numbers of outbreaks, cases, hospitalizations, and deaths in this final RIA, but does not
necessarily impact our annual estimates. The fact that the years 2008 through 2012 saw a
higher relative incidence of FDA covered RAC attributable illnesses than the previous
years does drive up the ratio of reported FDA RAC outbreaks to total CDC identified

can be found in the Unfunded Mandates section.

Page 49

illnesses. The implications of extending the outbreak data to 2012 are further discussed
in the Uncertainty and Sensitivity Analysis section.
Table 5. FDA Outbreak Data, 2003-2012
Outbreak Data Attributed to Produce RACs Other Than Sprouts 2003-2012
Commodity
Agent
Outbreaks
Cases
Hospitalizations

Deaths

berries
berries
green onion
herb
herb
leafy greens

Cyclospora
Salmonella
Hepatitis A
Cyclospora
E. coliO157:H7
Cyclospora

2
2
1
2
1
1

67
20
919
622
108
38

2
1
128
1
8
0

0
0
3
0
0
0

leafy greens
leafy greens
melon
melon
melon
nut*

E. coliO157:H7
Salmonella
Listeria monocytogenes
Salmonella
Shigella sonnei
E. coliO157:H7

3
1
1
8
1
1*

60
15
147
514
56
8*

15
1
143
140
3
3*

0
0
33
6
0
0*

nut

Salmonella

2

95

12

1

other
other
tomato
unknown
RAC Total

Cyclospora
Salmonella
Salmonella
Salmonella

2
6
8
6

172
1925
661
860

0
370
80
132

0
2
0
0

48

6287

1039

45

Outbreak Data Attributed to Sprouts, 2003-2012
sprout
E. coliO157: NM (H-)
3
36
3
0
sprout
E. coliO157:H7
2
27
5
0
sprout
E. coliO26
1
29
7
0
sprout
Listeria monocytogenes
1
20
16
0
sprout
Salmonella
14
651
56
1
Sprout Total
21
763
87
1
Total
69
7050
1126
46
Note: The E. Coli nut outbreak is associated with hazelnuts, which are not covered by the final rule (they
are exempt as rarely consumed raw under § 112.2(a)(1)). Therefore we do not include this outbreak in
calculating the estimated benefit of the rule.

Table 6 presents the estimation of the total number of illnesses attributable to
produce RACs other than sprouts based on FDA outbreak data combined with CDC
outbreak data (Ref. 30) and applied to Scallan, et al.’s estimate of the total number of
foodborne illnesses (Ref.31). To estimate the number of total illnesses associated with
Page 50

FDA regulated produce, we employ a two-step calculation, fully explained in the
Preliminary Regulatory Impact Analysis (Ref. 6): First, to determine the percent of illness
attributable to produce we examine FDA specific outbreak data and the whole universe of
identified pathogen illnesses, accounting for all outbreaks associated with an identified
food vehicle. Dividing the number of observed FDA-regulated produce-associated
illnesses by the total outbreak illnesses, gives us the percentage attributable to FDAregulated produce. This number is then multiplied by Scallan, et al.’s estimate of the total
annual incidence of each specific foodborne pathogen (Ref.31). This step corrects for
numerous downward biases in the CDC database of illnesses such as under-reporting and
under-identification of a foodborne illness. Multiplying the percentage attributable to
FDA-regulated produce by the annual incidence yields the annual estimated illnesses
attributable to FDA-regulated produce.
Dividing the number of produce acres associated with covered farms by the
number of produce acres more susceptible to contamination resulting in preventable
illness (i.e., produce that is not commercially processed or rarely consumed raw), we find
that approximately 94.2 percent of produce acres associated with preventable illness are
covered by the produce rule. This means that 5.8 percent of produce associated with
illnesses potentially preventable by the rule is exempt or not covered. If the marginal risk
of illnesses associated with a unit of output were distributed uniformly across farms
within a given commodity, 4 then we could see a total reduction in preventable illnesses of

4

There has been no evidence to suggest that the marginal risk of illness from a unit of output on large farm
is smaller or larger than the marginal risk of illness from a unit of output on a small farm.

Page 51

about 5.8 percent, or to 130,398 (138,424 x [1-.058]) for produce RACs other than
sprouts and 52,888 (56,145 x [1-.058]) for sprouts. 5
We multiply the total number of estimated preventable illnesses attributable to
FDA regulated produce (130,398+52,888 = 183,826) by 4 to obtain 733,146 unidentified
illnesses. This creates a ratio of identified to unidentified illnesses that is consistent with
Scallan, et al., who estimate that unidentified illnesses make up about 80% of all
foodborne illnesses (Ref.31). Using this calculation methodology, the total number of
preventable foodborne illnesses caused by microbial contamination of FDA-regulated
produce is estimated to be 916,432 (183,826+733,146, rounded). This is the more
conservative of the two estimation methods presented in the PRIA (Ref. 6), which
reduces our estimate of total unidentified illnesses.
Table 6. Estimated Number of Illnesses
Estimated Number of Illnesses Attributable to Produce RACs other than sprouts
Estimated
Identified
Annual
Percentage
FDA RAC
Cases
Agent
Foodborne
Attributable
(2003-2012) (2003Illnesses
to RACs
2012)
(Scallan)
Salmonella
4,090
36,790
11.12%
1,072,450

Estimated
Annual
Illnesses
Attributable to
RACs
119,226

Shigella sonnei
Listeria monocytogenes

56
147

3,044
361

1.84%
40.72%

154,053
1,680

2,834
684

Hepatitis A
Cyclospora cayatenensis
E.coli, STEC0157

919
899

1,250
1,109

73.52%
81.06%

1,665
13,906

1,224
11,273

168
6,279

3694
46,349

4.55%
13.56%

69,972
1,438,692

3,182
138,424

Estimated
Annual
Foodborne
Illnesses
(Scallan)

Estimated
Annual
Illnesses
Attributable
to Sprouts

Total Identified RAC

Estimated Number of Illnesses Attributable to sprouts

Agent

FDA
Sprouts
(20032012)

Identified
Cases
(20032012)

Percentage
Attributable
to Sprouts

5

We do not consider there to be a significant drop in benefits due to the exclusion of produce rarely
consumed raw or produce headed for commercial kill step processing, as such produce can be expected to
receive treatment to reduce risk from biological hazards and is therefore considered to present lower risk
than other types of produce.

Page 52

Salmonella
Listeria monocytogenes
E.coli, STEC0157
E.coli, STEC non 0157

651
20
63
29

36,790
361
3,694
101

1.77%
5.54%
1.71%
28.71%

1,072,450
1,680
69,972
124,966

18,977
93
1,193
35,881

Total Identified sprouts

763

46,349

1.65%

1,438,692

56,145

We estimate the monetized value of reducing foodborne illnesses from produce
by multiplying the annual number of illnesses per pathogen by the estimated cost
(including willingness-to-pay for longevity and avoided pain and suffering) per case. The
estimated cost per case is a pathogen specific estimate of dollar burden a typical case of
this particular foodborne illness places on an individual, which comes from Minor et al
(2014) (Ref. 32). Our estimated costs per illness are higher than those in the PRIA
because we utilize a higher Value of Statistical Life (VSL), $9 million, and a higher
QALD estimate, $1,260, for all pathogens (Ref. 16). Table 7 presents the burden of
illness attributable to microbial contamination of FDA-regulated produce RACs other
than sprouts and sprouts. Column two contains the total number of preventable illnesses
attributable to FDA-regulated produce, previously calculated. This number is multiplied
by the expected dollar loss per case, to give the annual cost of each pathogen in the US
population. Taken together, we estimate that the total cost of the illnesses linked to all
items of produce is approximately $2.5 billion. As discussed below, these figures are not
the expected benefits associated with the provisions in this rule. We expect that the rule
would eliminate only some portion of illnesses linked to produce and so would have
lower real-world benefits.
Table 7. Estimated Dollar Burden of Illnesses
Estimated Dollar Burden Attributable to Produce RACs other than sprouts

Page 53

Agent

Est. Annual
Illnesses
Attributable
to RACs

Salmonella
Shigella sonnei
Listeria monocytogenes
Hepatitis A
Cyclospora cayatenensis
E.coli, STEC0157
Total RAC Identified

119,226
2,834

% produce
acres
associated
with
preventable
illness
94.2%
94.2%

684
1,224
11,273
3,182
138,575

94.2%
94.2%%
94.2%
94.2%
94.2%

Total RAC Unidentified

-

521,592

-

651,990

Total RAC

Estimated Dollar Burden Attributable to sprouts
% produce
Est. Annual
acres
Illnesses
associated
Agent
Attributable
with
to RACs
preventable
illness
18,977
94.2%
Salmonella
93
94.2%
Listeria monocytogenes
1,193
94.2%
E.coli, STEC0157
35,881
94.2%
E.coli, STEC non 0157
56,145
94.2%
Total Sprouts Identified
Total Sprouts Unidentified
Total Sprouts

Est.
Preventable
Attributable
Illnesses

Expected
Dollar Loss
per Case

Covered
Dollar
Burden
(millions)

112,311
2,670
645
1,154
10,620
2,998
130,398

$6,015
$3,323

$676
$9

$1,574,670
$46,704
$4,056
$11,631

$1,015
$54
$43
$35
$1,831

$409

$214

Est.
Preventable
Attributable
Illnesses
18,977
93
1,193
35,881
52,888
211,554
264,442

$2,045

Expected
Dollar Loss
per Case

Covered
Dollar
Burden
(millions)

$6,015
$1,574,670
$11.631
$2,253
$409

TOTAL

$108
$138
$13
$76
$335
$87
$421
$2,466

2. Produce Rule Model of Risk Reduction
We examine the overall effectiveness of the regulation in reducing human
foodborne illnesses. To do this, we estimate the public health benefits of the produce
regulation provisions in two distinct ways: as a whole and by pathways of contamination.
We specify eight pathways of contamination: Agricultural Water for growing and harvest
activities; Agricultural Water for postharvest activities; Biological Soil Amendments;
Page 54

Worker Health and Hygiene in growing and harvest activities; Worker Health and
Hygiene in postharvest activities; Domesticated and Wild Animals; Equipment, Tools,
Buildings, and Sanitation in growing and harvest activities; and Equipment, Tools,
Buildings, and Sanitation in postharvest activities. These pathways come from the
Qualitative Assessment of Risk (QAR), which defines five routes of contamination:
Water, Soil Amendments, Animals, Worker Health and Hygiene, and Equipment and
Buildings (Ref. 33). We split Water, Worker Health and Hygiene, and Equipment and
Buildings into two separate pathways each, based on timing (growing and harvest versus
postharvest activities), for a total of eight pathways. These eight pathways are addressed
by an Expert Elicitation, the results of which are used to assign risk reduction values to
each pathway (Ref. 34).
We estimate the change in the probability of produce contamination as a function
of the relative likelihood of contamination from each specific pathway and the
effectiveness of the rule in reducing the risk of produce contamination within a specific
pathway of contamination. This change in the probability of contamination is then
applied to the current baseline of preventable foodborne illnesses attributable to FDAregulated produce. Based on current scientific literature, expert elicitation, census data,
research, and outbreak investigations, we can estimate the range of measureable
effectiveness of the produce safety regulation on the current burden of illness as a whole
(Ref.34;35;36;37). Additionally, these data are stratified to examine the effect amongst
specific commodities, or contamination pathways.
Table 8 presents the associated illnesses and mean relative weights and
effectiveness used in the model, as well as the calculation of the percentage reduction in

Page 55

contamination, by pathway and for the rule as a whole. For more detailed information on
how the weights and effectiveness values are assigned, see the PRIA and relevant sources
(Ref. 6;34;36;37). Because the weights and the effectiveness values are based on the
average values of distributions, we acknowledge the uncertainty they introduce. We
account for this in our uncertainty analysis of benefits in Section II, subsection I,
(formerly addressed in section IV, subsection H, subsection 3 in the PRIA). In the
uncertainty analysis, we run Monte Carlo simulations in which the values of the weights
and effectiveness, among others, vary based on our calculated parameters of their
distributions (mean, 5th percentile, 95th percentile). This allows us to calculate low and
high estimates of the benefits, taking into account the possible uncertainty of the weights
and effectiveness values.
To translate this percentage reduction in farm contamination to human health
outcomes, we estimate that a reduced probability of contamination will result in a
corresponding reduction in the expected number of illnesses. This means that roughly a
56 percent reduction in contamination will similarly reduce costs of illnesses. We apply
this percentage reduction to the average cost of illness, specific to produce-associated
illnesses, to estimate the overall benefits of the rule through illness prevention. We can
also use these assumptions to examine potential benefits of this rule by contamination
pathway. These calculations are also presented in Table 8.
Table 8. Mean Reduction in Risk of Contamination/ Benefits by Pathway
Mean Reduction in Risk of Contamination/ Benefits by Pathway attributable to Produce RACs other
than sprouts
Contamination
Covered Dollar
Likelihood of
Effectiveness Reductio
Benefits
Pathway
Burden
Being the Path
of Controls
n in Risk
(millions)
(millions)
of
Contamination
Agricultural Water
$2,045
16.32%
54.49%
8.89%
$182
(growing/harvest)

Page 56

Agricultural Water
(postharvest)
Biological Soil
Amendments
Worker Health and
Hygiene
(growing/harvest)
Worker Health and
Hygiene (postharvest)
Domesticated and Wild
Animals
Equipment, Tools,
Building and Sanitation
(growing/harvest)
Equipment, Tools,
Buildings and
Sanitation (postharvest)
Total

$2,045

14.37%

72.55%

10.42%

$213

$2,045

13.81%

65.62%

0.7%*

$15

$2,045

15.62%

66.04%

10.32%

$211

$2,045

15.20%

73.50%

11.17%

$228

$2,045

14.09%

58.04%

8.18%

$167

$2,045

4.18%

56.71%

2.37%

$49

$2,045

6.42%

67.97%

4.36%

$89

56.43%

$1,154

Mean Reduction in Risk of Contamination/ Benefits by Pathway attributable to sprouts
Contamination
Covered Dollar
Likelihood of
Effectiveness Reductio
Benefits
Pathway**
Burden
Contamination
of Controls
n in Risk
(millions)
(millions)
Agricultural Water
$421
16.32%
54.49%
8.89%
$38
(growing/harvest)
Agricultural Water
$421
14.37%
72.55%
10.42%
$44
(postharvest)
Biological Soil
$421
13.81%
65.62%
Amendments
Worker Health and
$421
15.62%
66.04%
10.32%
$44
Hygiene
(growing/harvest)
Worker Health and
$421
15.20%
73.50%
11.17%
$47
Hygiene (postharvest)
Domesticated and Wild
$421
14.09%
58.04%
8.18%
$35
Animals
Equipment, Tools,
$421
4.18%
56.71%
2.37%
$10
Building and Sanitation
(growing/harvest)
Equipment, Tools,
$421
6.42%
67.97%
4.36%
$18
Buildings and
Sanitation (postharvest)
Total
55.71%
$234
*The estimated effectiveness of Biological Soil Amendments has changed from the PRIA, because certain
proposed requirements for this section have been removed in the rule (see § 112.56(a)(1)(i)). See below for
a full explanation of the calculations. ** We do not have data to estimate risk reduction due to sprout
specific contamination pathways and therefore analyze the same pathways for sprouts as we do for other
produce..

From the table, we see that Agricultural Water for growing and harvest activities
is estimated to be the most important pathway of contamination, at about 16 percent. This

Page 57

is followed by Worker Health and Hygiene in postharvest activities (16 percent), Worker
Health and Hygiene in growing and harvest activities (15 percent), and Domestic and
Wild Animals (14 percent). Equipment, Tools, Buildings, and Sanitation in growing and
harvest activities represents the lowest contamination pathway, accounting for only about
4 percent overall. 6
We also see that the rule is estimated to do the best job of controlling risk of
contamination for Worker Health and Hygiene in postharvest (ph) activities, about a 74
percent reduction. This is followed closely by controls on Agricultural Water used in
postharvest activities (ph), estimated to have around 73 percent effectiveness in reducing
the associated risks of contamination. Controlling Agricultural Water used for growing
and harvest (g/h) activities is estimated to have the lowest effectiveness, at about 55
percent.
Provisions covering worker health and hygiene in postharvest (g/h) activities are
estimated to have the most impact on overall contamination, reducing it by an estimated
11 percent. Provisions covering Equipment, Tools, Buildings, and Sanitation in growing
and harvest (g/h) activities are estimated to contribute the least, at only about a 2 percent
reduction in contamination.
Taken together, this adds up to about a 56.43 percent reduction in risk of
contamination for produce RACs other than sprouts, and 55.71 percent reduction risk of
6

The number of outbreaks attributed to Equipment, Tools, Buildings, and Sanitation may be biased for a
few reasons. When it is implicated in the data, outbreaks are typically associated with multiple
contamination pathways, forcing the illnesses to be split amongst them, lowering the overall share of
illnesses attributable to this specific pathway. Additionally, problems with things like sanitation or tools
may be incorrectly attributed to another category, like worker health and hygiene. It could be that a worker
improperly washes their hands or cleans their tools because sufficient hand-washing facilities or cleaning
materials were not provided; however, when a resulting outbreak is recorded, only worker contact may be
cited as a contamination pathway. With the current data available, these are only speculations, and we
assign illnesses based only on the observable data.

Page 58

contamination for sprouts. Note, in Table 8, we only account for a very small reduction in
risk associated with our requirements related to Biological Soil Amendments because
certain proposed requirements that we accounted for in the PRIA have now been
eliminated from the rule (see § 112.56(a)(1)(i)). The originally estimated benefits
attributable to Biological Soil Amendments would have contributed an approximate $226
million in additional benefits (or 9.06 % of all foodborne illnesses attributable to FDA
RACs). We estimate that the remaining provisions will produce smaller costs and
benefits than previously estimated. Since the use of most Biological Soil Amendments of
Animal Origin in growing covered root crops is prohibited by the rule (because it is not
possible to minimize the potential for contact between soil amendments and root crops,
only amendments that meet the requirements of 112. 55(a) may be used in growing
covered root crops), we turn our focus to root crop farms. The proportion of covered
non-sprout farms that grow root vegetables is 8% (Ref. 15). Therefore, we estimate that
the benefits associated with the remaining requirements of BSA are 0.7% (9.06% x 8%)
of all foodborne illnesses attributable to FDA regulated produce RACs other than sprouts,
or approximately $15 million.
We are unable to account for the provisions specific to sprouts, namely batch
testing, seed treatment, and environmental monitoring because we are unable to parse out
their individual effects beyond what has already been done for all covered produce.
However, Ding and Fu (2013) (Ref. 38) and Montville and Schaffner (2004) (Ref. 39),
suggest that these sprout-specific provisions are effective in reducing or preventing
contamination. Therefore, our estimates likely represent a low estimate of the reduction
in risk of foodborne illnesses attributable to sprouts.

Page 59

Table 9 shows the estimated reduction in illnesses that may be attributable to the
regulation, shown both in illnesses averted and total dollar costs attributable to those
avoided illnesses. The overall benefits are higher than those in the PRIA, yet the number
of illnesses prevented is lower than that of the PRIA. This is mainly attributable to the
higher annual incidence of identified outbreaks associated with produce RACs other than
sprouts and sprouts. Combined with a more conservative estimate of unidentified goods,
which have a very low estimated cost per illness, we estimate a lower number of total
illnesses, which have a higher average costs per illness.

Table 9. Summary of Annual Benefits of Produce Regulation
Reduction
in Risk

Illnesses
Attributable to
Produce Covered
by this Rule

Illnesses
Prevented

Cost
Per
Illness

Total
Benefits
(in
millions)

Produce RACs other than sprouts

56.43%

651,990

367,949

$3,136

$1,1154

Sprouts

55.71%

264,442

147,321

$1,593

$235

916,432

515,269

Total

$1,389

We estimate that this rule may prevent, when fully implemented, about 515,269
illnesses, with an associated benefit of approximately $1.4 billion, annually. Furthermore,
the effectiveness of the rule may increase over time as farms learn by doing. However,
these benefits of this rule will not be immediately realized, nor will they be uniformly
implemented, due to the staggered nature of compliance times. Table 10 presents the
annual values of benefits as they are estimated to occur.

Covered Farms

Page 60

Year 10

Year 9

Year 8

Year 7

Year 6

Year 5

Year 4

Year 3

Year 2

Farms

Year1

Table 10. Timing of Produce Benefits (in millions)

Very
Small
Small

0

0

0

0

$90

$90

$137

$137

$137

$137

0

0

0

$50

$50

$76

$76

$76

$76

$76

Large

0

0

$620

$620

$942

$942

$942

$942

$942

$942

$28

$28

$28

$28

$28

$28

$28

$15

$15

$15

$15

$15

$15

$15

$191

$191

$191

$191

$191

$191

$191

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Covered Sprout Operations
Very
0
0
0
Small
Sprouts
Small
0
0
$15
Sprouts
Large
0
$191
$191
Sprouts
Exempt and Not Covered Farms
Very
0
0
0
Small
Exempt
Small
0
0
0
Exempt
Large
0
0
0
Exempt

The annualized benefits in Table 10 are calculated based on timing of produce
costs and benefits schedule shown in Table 4. For example, in year 2, full benefits are
realized from large sprout operations (roughly $191 million, which is calculated as the
total benefits attributable to sprouts operations multiplied by the percentage of covered
farms that fall into the large category). Because no other farms are affected, no other
benefits are being realized in year 2. This means that the total benefits realized in year 2
are roughly $191 million. In year 3, full benefits are realized from large sprout
operations ($191 million). Also in year 3, full benefits are realized from small sprout
operations ( $15 million, which is calculated as the total benefits attributable to sprouts
operations multiplied by the percentage of covered farms that fall into the small category),
and benefits minus those related to certain water provisions, are realized from large
covered, non-sprout operations (roughly $620 million, which is calculated as the total
benefits attributable to non-sprout operations, less the benefits attributable to certain
Page 61

water provisions, multiplied by the percentage of covered farms that fall into the large
category). This means that the total benefits in year 3 are roughly $826 million. This
continues, and in year 7, all benefits are realized, continuing on through our examined
timeline. Adding over the different operation types and sizes for year 7 yields our full
benefit estimation of roughly $1.4 billion. This is also the case for year 8, year 9, and
onward.
Next, we annualize estimates of the benefits below in Table 11. In this estimate,
we take into account the time that different sized farms have to comply with the rule, as
well as the different compliance times (notably, for agricultural water provisions, the
initial survey testing requirement for untreated surface water used for direct water
application during growing for produce other than sprouts, and certain related provisions,
are subject to the earlier compliance dates). Estimates are annualized over 10 years.
Table 11. Net Present Value and Annualized Benefits of Produce Regulation
Annualized Quantified
Illnesses

Annualized Monetized
Benefits (millions)

3,181,093
2,494,785

$8,322
$6,498

362,059
331,964

$976
$925

Annualized @ 3 percent over 10 years

72,411

$854

Annualized @ 7 percent over 10 years

66,392

$809

Net present value at 3 percent
Net present value at 7 percent
Annualized Values
Annualized @ 3 percent over 10 years
Annualized @ 7 percent over 10 years
Excluding Unidentified Illnesses

Annualizing benefits over the first ten years after publication of the rule, benefits
are expected to be approximately 362,059 illnesses averted per year, valued at $976
million annually.

Page 62

F. Costs of the Rule
With the data available we have attempted to accurately estimate the baseline
safety practices of the produce industry, and the costs related to the changes in those
practices as required by the rule. We utilize the most current and representative data
available.
We estimated most of the costs of the rule in the PRIA (which accompanied the
2013 proposed rule) and supplemental analysis (which accompanied the supplemental
notice), which contain detailed explanations of all calculations (Ref. 6) Where costs have
not changed substantially from those presented in either the proposed or supplemental
analysis, we do not present those detailed estimates here. Instead, we provide the
summary tables of the relevant Subpart, noting that only wages and farm counts have
changed, while underlying methodology and requirements remain constant.

1. Personnel and Training (Subpart C)

We did not receive substantial comments on the cost estimates for Personnel and
Training requirements; therefore, we have not altered the underlying methodology from
those originally proposed and estimated in the PRIA. In addition, our changes to the
proposed requirements in finalizing subpart C do not affect our cost estimates. Thus, we
present only summary statistics of estimates utilizing more current wage information and
farm counts. Table 12 provides the total cost for Personnel and Training; for full
information on how these costs are estimated please refer to Tables 112-115 of the
original PRIA (Ref. 6). The underlying estimates of this section have not changed;
however, these requirements are almost exclusively reliant on labor hours so the increase

Page 63

in wage rates has increased the costs. Also, based on public comments we increased the
wage rate of the training official from a supervisor to operator level, which accounts for
the majority of the increase in costs from those presented in the PRIA.
Table 12. Total costs for personnel qualifications and training (in thousands)
Outside Training
Management Personnel Food Safety
Training
Personnel Food Safety Training
Ensuring Personnel Compliance with
Training
Total Costs Accrued to Farms
(Annualized)

Very Small

Small

Large

Total

$2,975

$517

$714

$4,205

$880

$465

$940

$1,986

$4,118

$2,637

$7,576

$14,330

$33,171

$50,760

$82,932

$166,863

$41,143

$54,078

$92,162

$187,383

2. Health and Hygiene (Subpart D)

We did not receive substantial comments on the cost estimates for Health and
Hygiene requirements; therefore, we have not altered the underlying methodology from
those originally proposed and estimated in the PRIA. In addition, our changes to the
proposed requirements in finalizing subpart D do not affect our cost estimates. 7 Thus, we
present only summary statistics of estimates utilizing more current wage information and
farm counts. Table 13 provides the total cost for Personnel and Training; for full
information on how these costs are estimated please refer to Tables 35 – 39 of the
original PRIA (Ref. 6)
Table 13. Total Cost for Health and Hygiene (in thousands)
Very Small
Costs to exclude ill workers
Costs to wash and dry hands
thoroughly
Costs to avoid contact with animals

Small

Large

Total

$1,808

$723

$5,845

$8,377

$12,653

$10,176

$82,090

$104,919

$121

$98

$676

$896

7

There is new language that requires jewelry to be removed or covered and prohibits eating, chewing gum,
or consuming tobacco in certain areas. We estimate that farms are largely already in compliance with this
language and therefore do not present new estimates.

Page 64

Costs to wash hands before glove
use and maintain/replace gloves
Costs to inform, ensure compliance
by, and have toilets for visitors
Total Costs (annual)

$380

$306

$2,467

$3,153

$13,144

$2,282

$2,835

$18,261

$28,107

$13,585

$93,914

$135,606

3. Agricultural Water (Subpart E)

Agricultural water has undergone the most changes due to changes in
requirements from those proposed, public comments, and updated data. Therefore, we lay
out all estimates related to Agricultural water below. The most significant impacts on the
estimated costs from those presented in the proposed analysis are: increased our
assumption about the time it takes for farms to conduct a water system inspection based
on public comments; reduced the number of annual tests a farm must conduct due to
changes in the rule’s requirements; increased the number of farms that are required to
conduct water testing, as this requirement does not apply to only farms with post-harvest
activities; and allowed for die-off as a means to avoid water treatment, due to changes in
the rule’s requirements. Although some of these changes served to increase the costs of
the Agricultural Water requirements, such as broader application of water testing and
increased time to inspect water systems, the overall impact of these changes serves to
reduce the costs of the Agricultural Water requirements, where changes in the rule’s
requirements have led to the largest reductions in costs.
We estimate the cost of inspecting water systems, in accordance with § 112.42,
for the proportion of covered farms that are not currently conducting inspections; we find
that 22,781 very small, 3,956 small, and 8,292 large farms will need to implement
inspections. We estimate that very small and small farms will take four hours annually to
inspect agricultural water systems and that large farms will take eight hours annually, this

Page 65

estimate is based on data cited in the PRIA (Ref. 6) and public comments received on the
same document. We multiply these time burdens by the average farm operator wage rate
and estimate an annual per farm inspection cost of $288 for very small and small farms,
and $342 for large farms. Table 14 presents the total cost of inspecting water systems.
These estimates are largely taken from the PRIA (Ref. 6) with the exception of
hours to inspect which has been increased in response to comments.
Table 14.

Cost of inspecting water systems
Very Small

Number of covered farms

Small

Large

22,781

3,956

8,292

Rate of current practice

1.30%

0.60%

3.78%

Number of farms that need to inspect

22,485

3,932

7,979

4.00

4.00

8.00

$72.12

$72.12

$42.74

$288.48

$288.48

$341.92

$6,486,429

$1,134,380

$2,728,030

Hours to inspect
Farm operator wage rate
Annual cost of inspection per farm
Total annual cost of inspection

Total
35,029
34,396

$10,348,838

We estimate the cost of sampling and testing untreated surface water for covered
farms when the water is used in a direct application method during growing of covered
produce (other than sprouts), in accordance with § 112.46(b). We estimate that 42 percent
of irrigated farms use untreated surface water for the relevant purpose (direct water
application during growing produce other than sprouts) (Ref. 40). This results in 7,703
very small farms, 1,512 small farms, and 3,339 large farms that must conduct untreated
surface water testing. We estimate that the cost of collecting a water sample, including
collection, shipping costs, analysis, and travel is $110. In the initial two years of sampling,
we estimate that farms will collect 10 samples annually to develop a microbial water
quality profile, and then collect five samples annually to update their microbial water
quality profile using a 20-sample rolling dataset (see § 112.46(b)(1)(i)(A) and

Page 66

(b)(2)(i)(A)) at a per farm cost of $550 (five samples at $110 each). Additionally, it may
be necessary for farms to take a total of 20 new samples starting in any given year to
develop a new water quality profile, if the farm has determined or has reason to believe
that its microbial water quality profile no longer represents the quality of its water, in
accordance with § 112.46(b)(3)(i)(A). We estimate that 7.5 percent of farms using
untreated surface water will need to take 20 new samples starting in any given year to
develop a new water quality profile.
Table 15 presents the total costs of testing untreated surface water used for the
relevant purpose. We estimate that the total costs of testing surface water are $7.9 million
for very small farms, $1.6 million for small farms, and $3.4 million for large firms,
totaling to $12.9 million. These estimates are from the PRIA (Ref. 6) with the exception
of the testing frequency which we have updated in finalizing the rule.
Table 15. Costs of Sampling and Testing Untreated Surface Water used in Direct
Application During Growing Produce (Other than Sprouts)
Very small
18,262

Small
3,585

Large
7,916

Percent of farms that use surface water
Number of farms that must perform initial
survey

42.18%

42.18%

42.18%

7,703

1,512

3,339

Cost of collecting sample

$110.00

$110.00

$110.00

5

5

5

Number of irrigated farms

Baseline survey testing frequency*

Total
29,763

12,554

Annually recurring cost of 5 tests
$550.00
$550.00
$550.00
Percent of farms that will need to develop
new water quality profile
7.5%
7.5%
7.5%
Testing frequency (20 samples – 5 already
estimated for all farms)
15
15
15
Cost of 20 annual sample testing for 7.5%
of farms
$3,013,230
$591,525
$1,306,140
$4,910,895
Cost of 5 annual sample testing for all
farms
$4,927,653
$967,344
$2,135,982
$8,030,978
Total cost of sampling and testing
untreated surface water
$7,940,883
$1,558,869
$3,442,122
$12,941,873
Note: The initial survey of 20 samples must be in place before farms can comply with some of the other
annual requirements for agricultural water that relate to the microbial water quality profile developed from
the initial survey. For untreated surface water, testing for this will begin in year 3 for large farms, year 4 for
small farms, and year 5 for very small farms.

Page 67

We estimate the cost of sampling and testing untreated groundwater for covered
farms when the water is used in a direct application method during growing of covered
produce (other than sprouts), in accordance with § 112.46(b). Assuming that 32 percent
of covered farms use groundwater for the relevant purpose (direct water application
during growing produce other than sprouts) (Ref. 40), 5,811 very small farms, 1,141
small farms, and 2,519 large farms must test their untreated groundwater. We estimate
that the cost of collecting a water sample is $110 and in the first year, all farms will
collect four samples (see § 112.46(b)(1)(i)(B)), at a cost of $440 per farm. In subsequent
years, most farms will collect one sample annually (see § 112.46(b)(2)(i)(B)), at a cost of
$110 per farm per year. Additionally, it may be necessary for farms to take a total of 4
new samples in any given year to develop a new water quality profile, if the farm has
determined or has reason to believe that its microbial water quality profile no longer
represents the quality of its water, in accordance with § 112.46(b)(3)(i)(B). We estimate
that 5 percent of farms using untreated ground water will need to collect four new
samples in any given year to develop a new water quality profile. Table 15 presents the
costs of testing untreated groundwater used for the relevant purpose. We estimate that the
total costs of testing groundwater are $1.3 million for very small farms, $246 thousand
for small farms, and $542 thousand for large farms, totaling to $2.0 million.
Table 16. Costs of sampling and testing untreated groundwater used in Direct
Application During Growing Produce (Other than Sprouts)
Very small
Number of irrigated farms
Percent of farms that use ground water
Number of farms that must test

Small

Large

Total

18,262

3,585

7,916

29,763

31.82%
5,811

31.82%
1,141

31.82%
2,519

31.82%
9,471

4

4

4

Initial testing frequency

Page 68

Initial testing cost (year 1)

$440.00

$440.00

$440.00

1

1

1

$110.00

$110.00

$110.00

5%

5%

5%

3

3

3

NPV (at 3%)

$1,259

$1,259

$1,259

NPV (at 7%)

$1,081

$1,081

$1,081

Annualized costs (at 3%)

$148

$148

$148

Annualized costs (at 7%)
Cost of testing for farms testing 4 times per
growing season or year

$154

$154

$154

$401,764
$849,652

$78,870
$166,795

$174,152
$368,297

$654,786
$1,384,744

$1,251,416

$245,665

$542,449

$2,039,530

Annual testing frequency
Annual testing cost
Percent of farms that will need to develop new
water quality profile
Testing frequency (4 samples – 1 already
estimated for all farms)

Cost of testing for farms testing once annually
Total cost of testing ground water

We estimate the cost of sampling and testing untreated ground water when used
for certain uses specified in § 112.44(a) (including, for example, water used as sprout
irrigation water, and water applied in a manner that directly contacts covered produce or
food-contact surfaces during or after harvest), in accordance with § 112.46(c). All
covered farms and sprouting operations that use untreated ground water for such purposes
(i.e., farms that do not use water exempt from testing under § 112.46(a) such as public
(e.g., municipal) water sources meeting the established criteria in that section or water
treated in accordance with the requirements of § 112.43) must conduct water sampling
and testing. We estimate that 41 percent of sprouting operations use untreated ground
water for sprout irrigation, and that 30 very small, 25 small, and 62 large sprouting
operations must therefore test their untreated groundwater in accordance with § 112.46(c).
We estimate that 32 percent of farms use ground water for other purposes identified in §
112.44(a) (other than sprout irrigation water) and 26 percent of these farms use water
exempt from testing under § 112.46(a), and 1.3 percent of very small farms, 0.6 percent
of small farms, and 3.8 percent of large farms are already conducting water sampling and
Page 69

testing (20;Ref. 40). The remaining proportion of non-sprout farms and sprouting
operations includes 5,292 very small farms, 942 small farms, and 1,896 large farms. We
estimate that the cost of collecting and testing a water sample is $110 and that all farms
required to conduct these tests will test an average of 1.5 times per year (the midpoint
between 1 and 2 samples). This estimated average is derived from the required testing
frequency in § 112.46(c), which requires at least 4 tests in the first year, allowing one
test per year thereafter if the results meet the quality criterion, with required resumption
of 4 tests per year if any annual test fails to meet the quality criterion. Table 17 presents
the total costs of water sampling and testing for farms that use water for §
112.44(a)activities. We estimate that the total costs of water sampling and testing are
$873 thousand for very small farms, $155 thousand for small farms, and $313 thousand
for large farms, totaling to $1.3 million.
Table 17. Cost of sampling and testing untreated ground water for § 112.44(a)
purposes
Very small

Small

Large

Total

Total number of farms
Number of sprout operations that
use untreated ground water

22,781

3,956

8,292

35,029

30

25

62

117

Total number of farms
Percent of non-sprout farms that
use ground water
Number of non-sprout farms that
use ground water

22,811

3,981

8,354

35,146

31.82%

31.82%

31.82%

7,279

1,283

2,700

Rate of practice for water treatment

1.30%

0.60%

3.78%

Percent of farms using public water
Number of farms that must test
under the rule

26.0%

26.0%

26.0%

5,292

942

1,896

1.5

1.5

1.5

$110.00

$110.00

$110.00

$873,183

$155,432

$312,879

Testing frequency
Testing cost
Total costs of water sampling and
testing

Page 70

$1,341,495

All covered irrigated farms that do not use public water sources exempt from
testing and that use water for purposes in § 112.44(b) may choose to conduct water
treatment to meet the microbial quality criteria (see § 112.45(b)(3)). Treatment of water
is one of multiple options provided in § 112.45(b) to meet the microbial quality criteria in
§ 112.44(b). Farms may use the option to treat water, for example, if the farm is not able
to take advantage of the provisions for microbial die-off and/or microbial removal,
provided in § 112.45(b)(1), or the provision for re-inspection and corrections in §
112.45(b)(2). We estimate 22,025 farms (or 74 percent of covered irrigated farms) will
conduct testing. We also estimate that 48 percent of irrigated farms use application
methods where the water is intended to contact covered produce and 33 percent use
application methods where the water is likely to contact covered produce; these include
farms growing commodities such as cantaloupe, honeydew, other melons (including
Canary, Crenshaw and Persian), pineapple, strawberries, summer squash (such as patty
pan, yellow and zucchini), and watermelon (10;Ref. 15;40). We calculate the number of
farms that use direct water application methods by adding the proportions and
multiplying by the number of farms that must conduct testing, and estimate that this
includes 10,946 very small farms, 2,149 small farms, and 4,745 large farms, or 17,840
farms in total. We divide the number of operating days per year across farm size by 360
and multiply this proportion by the average number of irrigated acres for very small,
small, and large farms and estimate that there are 122,817 irrigated acres for very small,
131,080 irrigated acres for small, and 2,746,960 irrigated acres for large farms. We
estimate that 2.4 percent of irrigated acres do not meet the microbial quality criteria (Ref.
6) and that approximately 80 percent of all farms can use the die-off provisions in §

Page 71

112.45(b)(1) or the re-inspection and correction provisions in § 112.45(b)(2), leaving 590
acres on very small farms, 629 acres on small farms, and 13,185 acres on large farms that
may treat their water to meet the microbial quality criteria. We estimate there to be 2.16
acre-feet of water per acre and multiply (Ref. 40) this by the number of acres to be
treated, resulting in 1,273 acre-feet for very small farms, 1,359 acre-feet for small farms,
and 28,480 acre-feet for large farms. We estimate that the current rate of practice for
water treatment is 1.3 percent for very small farms, 0.6 percent for small farms, and 3.8
percent for large farms, resulting in 1,257 acres on very small farms, 1,351 acres on small
farms, and 27,404, acres on large farms to be treated (Ref. 20) We multiply acres by our
estimated treatment costs per acre-foot ($543 for very small farms, $289 for small farms,
and $32 for large firms) to find total costs. Table 18 presents total costs of water
treatment to meet the microbial quality criteria. We estimate that the total costs of
treatment are $682,449 for very small farms, $390,405 for small farms, and $876,925 for
large farms, totaling to $1,949,779.
Table 18. Water treatment to meet microbial quality criteria of GM of 126 CFU /
100 mL and STV of 410 CFU / 100 mL
Very small
Number of covered irrigated farms
Percent of farms that use public water
Number of farms that test water
Percent of farms using agricultural water
intended to contact covered produce
Percent of farms using agricultural water
likely to contact covered produce
Number of farms using direct water
application
Percent of season when produce is present
Farms with irrigated acreage using direct
water application methods, weighted by
percentage of season when produce is present
Average irrigated acres
Irrigated acres using direct water application

Small

18,262
26%
13,514

3,585
26%
2,653

77,916
26%
5,858

48%

48%

48%

33%

33%

33%

10,946

2,149

4,745

33%

50%

83%

3,612

1,074

3,952

34

122

695

122,817

Page 72

Large

Total
29,763
22,025

17,840

8,639

methods

131,080

2,746,960

2.4%
2,948

2.4%
3,146

2.4%
65,927

Acres that must be treated

80%
590

80%
629

80%
13,185

Acre-ft of water per acre

2.16

2.16

2.16

Rate of current practice

1,273
1.3%

1,359
0.6%

28,480
3.8%

Acres that will treat

1,257

1,351

27,404

Treatment costs per acre-ft

$543

$289

$32

$682,449

$390,405

$876,925

Percent of farms that do not meet quality
criteria
Acres to be treated
Percent where die-off until harvest or storage
is an option

Acre-ft of water to be treated

Total cost

$1,949,779

All covered farms that use water for purposes in § 112.44(a) that is not public
water exempt from testing may choose to conduct water treatment to meet the microbial
quality criterion. Treatment of water is one of multiple options provided in § 112.45(a) to
meet the microbial quality criterion in § 112.44(a) (see § 112.45(a)(1)(ii)). Farms may
use the option to treat water, for example, if the farm is not able to take advantage of the
provisions for re-inspection and corrections in § 112.45(a)(1)(i). We estimate that 15.2
percent of water does not meet quality criteria of no detectable E. coli (6;10;20;40;Ref.
41) The number of farms requiring treatment is calculated by multiplying the number of
farms using water for § 112.44(a) purposes by the percent of farms that do not meet
quality criteria and by the portion of farms that do not use public water exempt from
testing. This yields 2,534 very small farms, 446 small farms, and 906 large farms that
may treat. We estimate that one-time capital costs will be $2,441.34 for very small farms,
$3,678.13 for small farms, and $3,567.78 for large farms and that annual operating costs
will be $117 for very small farms, $1,099 for small farms, and $6,714 for large farms(Ref.
6;41;42;43) We add annualized one-time capital costs and annual operating costs and
multiply by the number of farms that initially test and then treat water to estimate total

Page 73

costs of $1.2 million for very small farms, $724 thousand for small farms, and $6.5
million for large farms, totaling to $8.4 million. Table 19 presents the total costs of water
treatment to meet the microbial quality requirement in § 112.44(a).
Table 19. Water treatment to meet quality criterion of no detectable E. coli for
purposes in § 112.44(a)
Very small
22,781

Small
3,956

Large
8,292

26.0%

26.0%

26.0%

30

25

62

117

16,888

2,952

6,198

26,038

15.2%
2,567

15.2%
449

15.2%
942

3,958

Current rate of practice

1.3%

0.6%

3.8%

Number of farms that test

2,534

446

906

One-time capital costs

$2,441.34

$3,678.13

$3,567.78

Annualized costs (3%)

$286.20

$431.19

$418.25

Annualized costs (7%)

$347.59

$523.68

$507.97

Operating cost per year

$117.26

$1,099.32

$6,713.74

$1,177,771

$723,886

$6,546,385

Number of covered farms
Percent of farms using public water
Number of sprout operations that use
untreated ground water
Number of farms subject to microbial
testing requirements in § 112.46(c) (to meet
§ 112.44(a) criterion)
Percent contaminated
Number of farms that require treatment

Total costs for water treatment

Total
35,029

3,886

$8,448,015

Table 20 presents a summary of the costs of the agricultural water provisions.
Excluding recordkeeping, the total cost of the water provisions is $18 million for very
small farms, $4 million for small farms, and $14 million for large farms, totaling to $37
million.
Table 20. Summary of the costs of the agricultural water provisions (in thousands)
Description
Inspection and maintenance of
agricultural water systems
Cost of testing untreated surface water
used in direct application during growing
for produce other than sprouts
Cost of testing untreated ground water
used in direct application during growing
for produce other than sprouts
Cost of testing untreated ground water
used for 112.44(a) purposes (including
sprout irrigation water)

Very small

Small

Large

Total

$6,486

$1,134

$2,728

$10,349

$7,941

$1,559

$3,442

$12,942

$1,251

$246

$542

$2,040

$873

$155

$313

$1,341

Page 74

Water treatment to meet criteria of GM
of 126 CFU / 100 mL or STV of 410
CFU / 100 mL for direct application
during growing of produce other than
sprouts
Treatment to meet criteria of no
detectable E. coli for 112.44(a) purposes,
including sprout irrigation water
Total cost by size category
Cost per farm

$682

$390

$877

$1,950

$1,178

$724

$6,546

$8,448

$18,412

$4,209

$14,449

$37,070

$808

$1,064

$1,742

$1,058

4. Biological Soil Amendments (Subpart F)

The minimum application intervals for biological soil amendments of animal
origin, which we proposed in the 2013 proposed rule, have been removed from the rule.
We estimate that removing these application intervals will remove an overwhelming
majority of all costs originally estimated. Therefore, we have eliminated the original costs
estimates attributed to Biological Soil Amendments of animal origin attributable to this
rulemaking. There are still recordkeeping requirements related to Biological Soil
Amendments, and those costs are presented in the Recordkeeping (Subpart O) section of
this analysis.
In addition, the use of Biological Soil Amendment of Animal Origin in growing
covered root crops is prohibited unless the amendment meets the requirements of
112.55(a). Therefore, the costs of root crop farms that use BSA of animal origin
switching to permissible soil amendments are presented in Table 21. Using data from the
NASS Agricultural Census, we estimate that approximately eight percent of covered
farms grow root crops (Ref. 15), and 15 percent of total farms apply any type of BSA
(Ref. 6;20). Therefore, we estimate that 273 very small farms (22,781 farms x 8 percent x
15 percent), 47 small farms (3,956 farms x 8 percent x 15 percent), and 100 large farms

Page 75

(8,292 farms x 8 percent x 15 percent) will incur a cost of switching amendment types. 8
From the PRIA, we estimate that the average cost of switching to commercial chemically
treated compost is $1,600 for very small farms, $6,600 for small farms, and $17,300 for
large farms, and we expect that a switch to permissible amendments for covered root
crops (such as amendments not containing materials of animal origin, or BSAs treated to
meet the § 112.55(a) microbial standard) will represent a comparable cost. 9 In total, we
estimated that the cost of switching away from most BSAs for root crops is
approximately $2.5 million, annually.
Table 21. Cost to root crop farms of switching from compost or raw manure of
animal origin
Very small
Number of farms

Total cost by category

Large

Total

22,781

3,956

8,292

8%

8%

8%

1,822

316

663

15%

15%

15%

273

47

100

$1,600

$6,600

$17,300

$437,395

$313,315

$1,721,419

Percent of farms that grow root crops
Number of root crop farms
Percent of farms using biological soil
amendments of any type
Number of root crop farms using biological soil
amendments
Average cost of switching to treated BSAs that
meet the microbial standard in § 112.55(a) or
other permissible amendments

Small

35,029

2,802

420

$2,472,130

5. Domesticated and Wild Animals (Subpart I)

We did not receive substantial comments on cost estimates for Domesticated and
Wild Animals; therefore, we have not altered the underlying methodology from those

8

We recognize that there may be more efficient means of meeting the requirements for an individual farm,
such as chemical treatment or switching to a vegetative manure source; however, either of these activities
would likely be utilized as a cost savings measure if they are employed instead of purchasing commercial
compost. Therefore, our average costs estimates may be viewed as somewhat higher than those that are
likely to be realized by individual farms.
9
Costs are calculated without taking into account opportunity or time costs of searching for new suppliers
or rewriting contracts.

Page 76

originally proposed and estimated in the PRIA. The rule’s requirements have been altered
in two key ways that reduce the cost estimated for Domesticated and Wild Animals. First,
assessment requirements have been limited to only operational days where the
harvestable portion of the product is present. This is a reduction from year round
monitoring estimated in the PRIA. Additionally the waiting period requirement related to
grazing animals has been removed completely from the rule and thus all of the associated
costs have been removed. Table 22 provides the total cost for Domesticated and Wild
Animals; for full information on how these costs are estimated please refer to Tables 82 –
83 of the original PRIA (Ref. 6).
Table 22. Cost for Domesticated and Wild Animals
Very small
Number of produce farms

Small

Large

22,781

3,956

8,292

Per-acre monitoring cost increase

3.36

3.36

3.36

Increase in cost per affected farm

$378

$1,260

$2,520

Percent of year in operation

27%

41%

55%

$2,359,238

$2,048,449

$11,449,775

Total cost per category

Total
35,029

$15,857,462

6. Growing, Harvesting, Packing, and Holding Activities (Subpart K)

We did not receive substantial comments on the cost estimates for Growing,
Harvesting, Packing, and Holding Activities; therefore, we have not altered the
underlying methodology from those originally proposed and estimated in the PRIA. In
addition, our changes to the proposed requirements in finalizing subpart K do not affect
our cost estimates. Thus, we present the estimates utilizing more current wage
information and farm counts. Table 23 provides the total cost for Growing, Harvesting,
Packing, and Holding Activities. These requirements are reliant on labor hours so the
increase in wage rates has increased the costs. Additionally, based on public comments
we have revised the number of operational days upwards to 100 for very small farms, 150
Page 77

for small farms, and 200 for large farms (up from 45, 45, and 90), which increases the
estimated costs. Finally, in the PRIA we estimated that only farms with post-harvest
activities would incur costs of Growing, Harvesting, Packing, and Holding Activities;
however, we now estimate that all farms with reusable food contact surfaces will need to
clean and sanitize. All of these changes have substantially increased the cost estimates of
Growing, Harvesting, Packing, and Holding Activities.

Table 23. Cost of Cleaning and Sanitizing Food Contact Surfaces
Very small

Small

Large

Number of Farms
Percentage of farms with reusable food contact
surfaces
Number of farms with reusable food contact
surfaces
Percentage of farms that do not clean/sanitize
food contact surface
Number of farms that need to clean/sanitize food
contact surface
Time to clean/sanitize (hours)

22,781

3,956

8,292

18%

18%

18%

4,101

712

1,493

30%

30%

30%

2,870

498

1,045

0.17

0.25

0.25

Non-supervisor wages
Labor cost to clean/sanitize a food contact
surface
Cost of sanitizer per farm job

$18.56

$18.56

$18.56

$3.16

$4.64

$4.64

$0.05

$0.05

$0.05

$3.21

$4.69

$4.69

100

150

200

$321

$704

$938

$920,023

$350,664

$980,015

Daily per farm cost to clean/sanitize
Operational harvest days
Annual per farm cost to clean/sanitize food
contact surfaces
Total cost to clean/sanitize food contact
surfaces

Total
35,029

$2,250,701

7. Equipment, Tools, Buildings, and Sanitation (Subpart L)

We did not receive substantial comments on cost estimates for Equipment, Tools,
Buildings, and Sanitation requirements; therefore, we have not altered the underlying
methodology from those originally proposed and estimated in the PRIA. In addition, our
changes to the proposed requirements in finalizing subpart L do not affect our cost
estimates. Thus, we present only summary statistics of estimates utilizing more current
Page 78

wage information, farm counts, and operational days where the harvested or harvestable
portion of produce is exposed. Table 24 provides the total cost for Equipment, Tools,
Buildings, and Sanitation; for full information on how these costs are estimated please
refer to Tables 88 – 94 of the original PRIA (Ref. 6). These requirements are almost
exclusively reliant on labor hours so the increase in wage rates has increased the costs.
Additionally, based on public comments we have revised the number of operational days
upwards to 100 for very small farms, 150 for small farms, and 200 for large farms (up
from 45, 45, and 90), which greatly increases the costs of these sections.
Table 24. Summary of Equipment, Tools, Buildings, and Sanitation Costs (in
Millions)
Total cost to clean and sanitize tools
Total cost to clean machinery
Total cost of pest control
Total cost to provide toilets and hand washing
Total cost to prevent sewage contamination
Total cost to dispose litter and land drainage
Total cost of trash removal
Total costs of equipment, tools, buildings, and
sanitation

Very small
$5.44
$7.15
$0.75
$3.05
$0.01
$3.09
$0.06

Small
$6.27
$3.39
$0.51
$1.05
$0.00
$2.69
$0.02

Large
$22.86
$24.22
$1.07
$12.25
$0.02
$24.88
$0.04

Total
$34.57
$34.76
$2.33
$16.34
$0.03
$30.66
$0.11

$19.49

$13.91

$85.29

$118.69

8. Sprouts (Subpart M)

We did not receive substantial comments on cost estimates for Sprouts
requirements; therefore, we have not altered the underlying methodology from those
originally proposed and estimated in the PRIA. In addition, our changes to the proposed
requirements in finalizing subpart M do not affect our cost estimates related to subpart M,
other than those captured in other parts of this document. Thus, we present only summary
statistics of estimates utilizing more current wage information and farm counts. Table 26

Page 79

provides the total cost for Sprouts; for full information on how these costs are estimated
please refer to Tables 102 – 107 of the original PRIA (Ref. 6).
Table 25 presents updated costs to conduct batch tests related to sprouts. The
initial estimate has not changed substantially from those presented in the PRIA. We
estimate that it costs approximately $147 to test each batch of sprouts for E. Coli
O157:H7 and Salmonella, and there are approximately 3,710 batches from the 74 very
small sprouting operations, 2,976 batches from the 60 small sprouting operations, and
33,623 batches from the 151 large sprouting operations. We estimate that batch testing
for E. Coli O157:H7 and Salmonella will cost approximately $5 million, annually. New
language has been added to the rule which requires sprouting operations to hold their
batches while awaiting the test results. We estimate holding costs as a function of the
total value of sprouts produced by the operation. We estimated that very small sprouting
operations generate total revenue of $70 thousand annually, small sprouting operations
generate revenue of $300 thousand annually, and large sprouting operations generate
annual revenue of approximately $600 thousand annually (Ref. 44). We estimate that
very small operations will need to hold 25 percent of their product while awaiting test
results, small operations will hold 10 percent of their product, and large operations will
only need to hold 5 percent of their product. Additionally, commonly cited holding costs
in the manufacturing literature are 25% of the total value. This yields an annual holding
cost for very small sprouting operations of $43,750 ($70 thousand x .25 x .25), small
operations of $7,500 ($300 thousand x .10 x .25), and large operations of $30,000 ($600
thousand x .05 x .1), and a total estimate of approximately $81 thousand. There is also a
requirement that sprout operations take appropriate action to prevent any food that is

Page 80

adulterated under section 402 of the Federal Food, Drug, and Cosmetic Act (Ref. 44)
from entering commerce; however, we do not estimate any additional costs to this
language as any such product is already illegal to sell. Finally, we add 10 percent on to
the bottom line to account for language which requires batch testing for additional
pathogens if and when certain criteria are met. In total we estimate that batch testing of
sprouts will cost approximately $5 million dollars annually.
Table 25. Total costs to test each batch of sprouts for E. coli O157:H7, Salmonella
species, and additional pathogens as applicable
Number of sprouting operations
Number of batches
Testing costs
Rate of industry practice
Total cost by size category
Average Sales Volume
Inventory Holding Cost
Additional Holding Time
Per Facility Cost of Holding Product Awaiting
Test Results
Rate of industry practice
Total Cost of Holding Product Awaiting Test
Results
Percent needing to be held
Inventory Holding Cost
Inventory Holding Cost
Addition for additional pathogen testing costs
Additional pathogen testing costs
Total cost of E. coli O157:H7 and
Salmonella batch testing, holding,
prevention, and additional pathogen tests

Very small
74
3,710
$545,444
55%
$245,450
$70,000
25%

Small
60
2,976
$437,532
55%
$196,889
$300,000
25%

Large
151
33,623
$4,943,253
55%
$2,224,464
$600,000
25%

14%

14%

14%

$2,500
55%

$10,714
55%

$21,429
55%

$83,250
25%

$289,286
10%

$1,456,071
5%

25%
$323,750
10%
$56,920

25%
$450,000
10%
$64,689

25%
$1,132,500
10%
$335,696

$626,120

$711,578

$3,692,660

Total
285

$2,666,803

$1,828,607

$1,906,250
$457,305

$5,030,358

There are new requirements for sprout producers to establish a written corrective
action plans as part of their environmental monitoring plan and written sampling plans;
however, these costs are presented in the recordkeeping section of this analysis rather
than the sprout requirements.
Table 26. Summary of the Total Costs of the Sprouts Provisions
Page 81

Costs to disinfect seeds
Costs to implement an environmental
monitoring plan
Costs for a specified protocol for collecting
environmental samples and testing for L.
sp., or L. monocytogenes
Cost of E. coli O157:H7 and Salmonella
batch testing, holding, prevention, and
additional pathogen tests
Total costs of the sprouts provisions

Very small
$79,190

Small
$63,523

Large
$717,683

Total
$860,396

$117,957

$164,759

$588,495

$871,212

$795

$644

$1,622

$3,061

$626,120

$711,578

$3,692,660

$5,030,358

$824,062

$940,504

$5,000,461

$6,765,027

9. Recordkeeping (Subpart O)

Farms will incur recordkeeping costs related to demonstrating qualified
exemption status; the commercial processing exemption; the agricultural water
provisions; the biological soil amendments of animal origin provisions; cleaning
equipment, tools, buildings, and sanitation; sprouting operations; and food safety training.
We present detailed costs for the recordkeeping activities required for agricultural water
and new provisions for sprouting operations; however, the other records have not
changed substantially from the PRIA (though there have been some changes to
recordkeeping, discussed in greater detail in the Paperwork Reduction Act analysis), and
we therefore present in this section only summary statistics of the remainder of
recordkeeping activities. For more on the full methodology please refer to the PRIA (Ref.
6).
We estimate that farms will incur recordkeeping costs pertaining to the water
provisions (under Subpart O and § 112.50), including keeping records of inspection of
water systems (§ 112.50(b)(1)), test results of untreated surface water (§ 112.50(b)(2)),
test results of untreated ground water (§ 112.50(b)(2)), scientific information supporting
adequacy of water treatment methods (§ 112.50(b)(3)), water treatment monitoring
results (§ 112.50(b)(4)), documentation of corrective actions including use of microbial
Page 82

die-off or removal rates (§ 112.50(b)(6)) and scientific data relied on for such rates
between harvest and end of storage (§ 112.50(b)(5)), use of public water sources (§
112.50(b)(7)), data to support any alternatives (including alternative microbial quality
criteria, alternative microbial die-off rates and maximum time intervals, or alternative
minimum numbers of samples for initial and annual surveys in testing untreated water
used for direct water application in growing produce other than sprouts) (§ 112.50(b)(8)),
and analytical methods used in lieu of those incorporated in the rule (§ 112.50(b)(9)).
We estimate that all covered farms not currently keeping such records will
maintain records of inspection of water systems (§ 112.50(b)(1)) and that the time burden
is one hour annually. We multiply the farm operator wage rate by the time burden and
annual frequency and estimate the costs of water inspection records are $1.6 million for
very small farms, $284 thousand for small farms, and $341 thousand for large farms.
From earlier estimates of water testing, we estimate that there are a total of 26,038
farms that use untreated ground water will incur the costs maintaining records of their
results from testing the water for 0 detectable generic E. coli (§ 112.50(b)(2)). We
estimate that the time burden of recordkeeping is 0.33 hours and that the annual
frequency of recordkeeping is estimated to be 2 times. We multiply the farm operator
wage rate by the time burden and the annual frequency and estimate the costs of surface
water testing records are $804 thousand for very small farms, $141 thousand for small
farms, and $175 thousand for large farms.
From earlier estimates of water testing, we estimate that 12,544 farms (those that
use untreated surface water less the percentage estimated to use public water sources)
will incur costs maintaining records of their results from testing the water for GM of 126

Page 83

CFU / 100 mL and STV of 410 CFU / 100 mL Generic E. coli (§ 112.50(b)(2)). We
estimate that the time burden of recordkeeping is 0.33 hours and that the annual
frequency of recordkeeping is estimated to be 10 times in the first two years and 5 times
in subsequent years. We multiply the farm operator wage rate by the time burden and the
net present value of the annual frequency over ten years and estimate the costs of surface
water testing records are $1.2 million for very small farms, $226 thousand for small
farms, and $296 thousand for large farms.
From earlier estimates of water testing, we estimate that 9,471 farms (those that
use untreated ground water less the percentage estimated to use public water sources) will
incur costs maintaining records of results from testing the water for GM of 126 CFU /
100 mL and STV of 410 CFU / 100 mL Generic E. coli (§ 112.50(b)(2)). We estimate
that the time burden of recordkeeping is 0.33 hours and that the annual frequency of
recordkeeping is 4 times in the first year and once in subsequent years. We multiply that
farm operator wage rate by the time burden and the net present value of the annual
frequency over ten years and estimate the costs of ground water testing records $194
thousand for very small farms, $38 thousand for small farms, and $50 thousand for large
farms.
We estimate that 20 percent of farms that treat water to meet quality criteria of
GM of 126 CFU / 100ml or STV of 410 CFU /100ml and 50 percent of farms that treat
water to meet quality criterion of no detectable E. coli (a total of 5,547 farms) will
maintain records of the adequacy of their water treatment methods (§ 112.50(b)(3)). We
estimate that 5,547 will maintain records, with a one-time burden of 0.5 hours. We
multiply the farm operator wage rate by the number of farms, the hourly time burden, and

Page 84

estimate that the costs of maintaining records of data to support method adequacy are
$194 thousand for very small farms, $38 thousand for small farms, and $50 thousand for
large farms. Because this is a onetime cost, we then annualize over 10 years.
From earlier estimates of water testing, we estimate that all farms that treat their
water (an estimated total of 5,547 farms) will maintain records of the results of water
treatment monitoring (§ 112.50(b)(4)), with an annual time burden of one hour. We
multiply the farm operator wage rate by the number of farms, the hourly time burden, and
the annual frequency and estimate that the costs of maintaining records of water
treatment monitoring are $250 thousand for very small farms, $47 thousand for small
farms, and $61 thousand for large farms.
Farms that rely on a microbial die-off or removal rate to determine a time interval
between harvest and end of storage, including other activities such as commercial
washing, to achieve a calculated log reduction of generic E. coli in accordance with §
112.45(b)(1)(ii), must have documentation of the scientific data or information they rely
on to support that rate (§ 112.50(b)(5)). We estimate that 25 percent of all farms that rely
on die-off, 3,661 (17,840 farms from table 18 of the FRIA x 80 percent that rely on die
off + 371 irrigated farms subject to a corrective action x 25 percent) would generate these
records for postharvest die-off intervals. It is estimated that two recordkeepers for each of
3,661 farms will spend .5 hour one-time on this documentation, estimated to consist of
gathering and maintaining the documentation of scientific data and information. We
multiply the farm operator wage rate by the number of farms, the hourly time burden, and
estimate that the costs of maintaining records of data to support microbial die-off are

Page 85

$162 thousand for very small farms, $32 thousand for small farms, and $41 thousand for
large farms. Because this is a onetime cost, we then annualize over 10 years.
When covered farms take corrective actions in accordance with § 112.45, they
must maintain certain required records (§ 112.50(b)(6)), including keeping certain
records about specific time intervals or log reductions applied. We calculate that 14,643
farms will incur the costs of documentation of any corrective actions taken in accordance
with § 112.45, including any time intervals or calculated log reductions applied.
Therefore, it is estimated that 1 recordkeeper on each of the 14,643 farms will spend an
average of 0.5 hours per year on recordkeeping related to corrective actions applied. The
total costs of corrective action recordkeeping, including microbial die-off or removal
records, is $325 thousand for very small farms, $63 thousand for small farms, and $83
thousand for large farms.
All covered farms that use public water sources exempt from testing, such as
municipal water, will maintain certain required records related to those public water
systems (§ 112.50(b)(7)). We estimate that 9,108 farms (the number of farms using
public water systems such as municipal water sources) will need to keep these records
and that the time burden is 0.33 hours annually (Ref. 6;10;40) We multiply the farm
operator wage by the proportion of farms that use municipal water and estimate that
public water system recordkeeping costs are $141 thousand for very small farms, $24
thousand for small farms, and $30 thousand for large farms.
Section 112.50(b)(8) requires all farms that choose to rely on an alternative under
§ 112.49 to have documentation of the scientific data or information they rely on to

Page 86

support that alternative. There are four types of alternatives that may be employed
according to 112.49(a)-(d).
Section 112.49(a) provides for an alternative microbial quality criterion (or
criteria) using an appropriate indicator of fecal contamination, in lieu of the microbial
quality criteria in § 112.44(b). Farms must maintain records supporting any such
alternative microbial criteria they use (§ 112.50(b)(8)). We estimate that approximately
8,757 farms that irrigate (35,029 total farms x 25 percent) will generate these alternative
records. We estimate each farm will spend half an hour one time on this documentation.
We multiply the farm operator wage by the number of farms and estimate that this
alternative microbial quality criterion recordkeeping costs are $205 thousand for very
small farms, $36 thousand for small farms, and $44 thousand for large farms. Because
this is a onetime cost, we then annualize over 10 years.
Section 112.49(b) provides for an alternative microbial die-off rate and an
accompanying maximum time interval, in lieu of the microbial die-off rate and maximum
time interval in § 112.45(b)(1)(i). Farms must maintain records supporting any such
alternative die off rate and maximum time interval they use (§ 112.50(b)(8)). We estimate
that approximately 3,661 farms that irrigate (14,643 total farms x 25 percent) will
generate these alternative records. We estimate each farm will spend half an hour one
time on this documentation. We multiply the farm operator wage by the number of farms
and estimate that this alternative microbial die-off rate recordkeeping costs are $81
thousand for very small farms, $16 thousand for small farms, and $21 thousand for large
farms. Because this is a onetime cost, we then annualize over 10 years.

Page 87

Section 112.49(c) provides for an alternative minimum number of samples used in
the initial survey for an untreated surface water source, in lieu of the minimum number of
samples required under § 112.46(b)(1)(i)(A). Farms must maintain records supporting
any such alternative sampling rate they use (§ 112.50(b)(8)). We estimate that
approximately 2,551 farms that utilize surface water (12,554 irrigated farms that use
surface water less the percentage estimated on public water sources x 20 percent) will
generate these alternative records. We estimate that 1,541 very small farms, 302 small
farms, and 668 large farms will develop one record that will take 0.5 hours to complete.
In total, we estimate that this recordkeeping will cost very small farms $56 thousand,
small farms $11 thousand, and large farms $14 thousand. Because this is a onetime cost,
we then annualize over 10 years.
Section 112.49(d) provides for an alternative minimum number of samples used
in the annual survey for an untreated surface water source, in lieu of the minimum
number of samples required under § 112.46(b)(2)(i)(A). Farms must maintain records
supporting any such alternative sampling rate they use (§ 112.50(b)(8)). We estimate that
approximately 2,551 farms that utilize surface water (12,554 irrigated farms that use
surface water less the percentage estimated on public water sources x 20 percent) will
generate these alternative records. We estimate that 1,541 very small farms, 302 small
farms, and 668 large farms will develop one record that will take 0.5 hours to complete.
In total, we estimate that this recordkeeping will cost very small farms $56 thousand,
small farms $11 thousand, and large farms $14 thousand. Because this is a onetime cost,
we then annualize over 10 years.

Page 88

All farms that are required to test their agricultural water in compliance with §
112.46 must have documentation of any analytical methods that they choose to use for
such testing in lieu of the methods that are incorporated by reference in § 112.151 (§
112.50(b)(9)). It is not known how many farms will use other analytical methods;
however, to the extent that they do this it will likely be as a cost savings measure.
Therefore, we do not include any cost of recordkeeping for 112.50(b)(9) here. This is
acknowledged in the PRA analysis.
Table 27 presents the recordkeeping costs of the water provisions. We estimate
that the total costs of recordkeeping are $4.5 million for very small farms, $0.83 million
for small farms, and $1.0 million for large farms, totaling to $6.4 million.
Table 27. Recordkeeping Costs of the Water Provisions
Very small
Farm operator wages
Inspection of water systems
(§ 112.50(b)(1))
Number of farms

Small

Large

Total

$72.12

$72.12

$42.74

22,485

3,932

7,979

Time burden

1

1

1

Frequency

1

1

1

Total inspection recordkeeping costs
$1,621,607
Initial and annual tests for 0 detectable Generic E. coli
(§ 112.50(b)(2))
Number of farms
16,888

$283,595

$341,004

2,952

6,198

2

2

Time burden

2

34,396

$2,246,206

26,038

Frequency
0.33
0.33
0.33
Baseline recordkeeping costs of testing
ground water for 0 detectible generic E.
$803,869
$140,515
$174,835
$1,119,219
coli
Initial and annual tests of surface water for GM of 126 CFU / 100 mL and STV of 410 CFU / 100 mL
Generic E. coli
(§ 112.50(b)(2))
Number of farms
7,703
1,512
3,339
12,554
Time burden

0.33

0.33

0.33

Frequency
Baseline recordkeeping costs of testing
surface water for GM 126 CFU/STV 410
CFU/100 mL generic E. coli

6.29

6.29

6.29

$1,153,122

$226,369

$296,218

Page 89

$1,675,708

Initial and annual tests of ground water for GM of 126 CFU / 100 mL and STV of 410 CFU / 100 mL
Generic E. coli
(§ 112.50(b)(2))
Number of farms
5,811
1,141
2,519
9,471
Time burden

0.33

0.33

0.33

Frequency
1.4
1.4
1.4
Baseline recordkeeping costs of testing
ground water for GM 126 CFU/STV 410
$193,618
$38,009
$49,737
$281,365
CFU/100 mL generic E. coli
Cost of records of data to support adequacy of a treatment method used to satisfy § 112.43(a)(1) and
(a)(2)
(§ 112.50(b)(3))
Number of farms
3,473
654
1,420
5,547
Time burden

0.5

0.5

0.5

Frequency
1
1
Recordkeeping costs of data to support
$125,228
$23,588
method adequacy
NPV (@7%)
$17,830
$3,358
Cost of records of results of water treatment monitoring records
(§ 112.50(b)(4))
Number of farms
3,473
654

1
$30,346

179,161

$4,321

$25,509

1,420

5,547

Time burden

1

1

1

Frequency

1

1

1

$250,455

$47,175

$60,692

Recordkeeping costs of water treatment

358,322

NPV (@7%)
$35,659
$6,717
$8,641
$51,017
Cost of records of data to support microbial die-off/max time interval between harvest and end of storage
or removal during activities such as commercial washing
(§ 112.50(b)(5))
Number of farms
2,251
440
970
3,661
Time burden

0.5

0.5

0.5

Frequency
2
2
2
Recordkeeping costs of data to support die$162,339
$31,727
$41,454
off or maximum time interval
Costs of records for corrective actions under § 112.45, including die-off or removal use
(§ 112.50(b)(6))
Number of farms
9,004
1,760
3,880
Time burden

$235,520

14,643

1

1

1

Frequency
0.5
Recordkeeping costs for corrective actions,
$324,677
including die-off or removal use
Costs of records related to public water systems
(§ 112.50(b)(7))
Number of covered irrigated farms
5,923

0.5

0.5

$63,454

$82,909

$471,039

1,029

2,156

9,108

0.33

0.33

0.33

1

1

1

Time burden
Frequency

Page 90

Recordkeeping cost of public water systems
$140,966
$24,479
$30,408
$195,853
Scientific data or information you rely on to support any alternative that you establish and use in
accordance with § 112.49(a)
(§ 112.50(b)(8))
Number of farms
5,695
989
2,073
8,757
Time burden

0.5

0.5

0.5

Frequency
1
1
1
Recordkeeping cost of data to support
$205,371
$35,663
$44,300
$285,334
alternatives
NPV (@7%)
$29,240
$5,078
$6,307
$40,625
Scientific data or information you rely on to support any alternative that you establish and use in
accordance with § 112.49(b)
(§ 112.50(b)(8))
Number of farms
2,251
440
970
3,661
Time burden

0.5

0.5

0.5

Frequency
1
1
1
Recordkeeping cost of data to support
$81,169
$15,863
$20,727
$117,760
alternatives
NPV (@7%)
$11,557
$2,259
$2,951
$16,766
Scientific data or information you rely on to support any alternative that you establish and use in
accordance with § 112.49(c)
(§ 112.50(b)(8))
Number of farms
1,541
302
668
2,511
Time burden

0.5

0.5

0.5

Frequency
1
1
1
Recordkeeping cost of data to support
$55,553
$10,906
$14,271
$80,730
alternatives
NPV (@7%)
$7,910
$1,553
$2,032
$11,494
Scientific data or information you rely on to support any alternative that you establish and use in
accordance with § 112.49(d)(§ 112.50(b)(8))
Number of farms
1,541
302
668
2,511
Time burden
Frequency
Recordkeeping cost of data to support
alternatives
NPV (@7%)
Total recordkeeping costs of the water
provisions

0.5

0.5

0.5

1

1

1

$55,553

$10,906

$14,271

$80,730

$7,910

$1,553

$2,032

$11,494

$4,510,303

$828,664

$1,042,849

$6,381,815

Sprouting operations will incur one-time and recurring recordkeeping costs
(Subpart O and § 112.150).

Page 91

One-time recordkeeping costs include an environmental monitoring plan (§
112.150(b)(2)) with a one-time burden of 7 hours for very small farms, 12 hours for small
firms, and 17 hours for large firms (Ref. 3) not already estimated to be performing these
actions. These time burdens are multiplied by the number of sprouting operations and the
wage rate for farm operators ($72.12 for very small and small farms, $42.74 for large
farms) to estimate a total one-time cost of $123,379.
One-time recordkeeping costs also include an irrigation water sampling plan (§
112.150(b)(3)) with a one-time burden of 8 hours per sprouting operation not already
performing these actions. These time burdens are multiplied by the number of sprouting
operations and by the farm operator wage rate to estimate a one-time irrigation water
sampling plan recordkeeping cost of $79,944.
Sprout operations are required to have documentation of any analytical methods
used in lieu of the methods for both environmental testing and batch testing that are
incorporated by reference in §§ 112.152 and 112.153 (§ 112.150(b)(5)). It is not known
how many sprout operations will use other analytical methods; however, to the extent that
they do this it will likely be as a cost savings measure. Therefore, we do not include any
cost of recordkeeping for 112.50(b)(5) here. This is acknowledged in the PRA analysis.
In addition, § 112.144(c) requires sprout operations to conduct testing for additional
pathogens when certain conditions are met, and § 112.150(b)(5) requires sprouting
operations to have documentation of any analytical methods used for such testing because
there is no specific method for such testing incorporated by reference in § 112.152 or
112.153. It is not known if or when there will be a pathogen(s) meeting the relevant
criteria; however, it is estimated that one 2 hour record will fulfill this requirement,

Page 92

estimated as the time needed to establish a new testing routine. These time burdens are
multiplied by the number of sprouting operations and by the farm operator wage rate to
estimate a one-time record of analytical testing method recordkeeping cost of $19,986.
One-time environmental monitoring plan, irrigation water sampling plan, and
additional pathogen analytical test method recordkeeping costs total to $56,251 for very
small operations, $59,023 for small operations, and $108,036 for large operations. Table
28 presents these totals annualized at 7 percent for 10 years, estimated at $8,009 for very
small operations, $8,404 for small operations, and $15,382 for large operations, totaling
to $31,794.
Table 28. One-time Recordkeeping Costs for Sprouts
One-time recordkeeping costs

Very small
operations

Environmental monitoring plan (§ 112.150(b)(2))
Number of sprout operations
Time burden
Frequency
Recordkeeping cost of environmental
monitoring
NPV (@7%)

Small
Operations

Large
Operations

46
7

37
12

94
17

1

1

1

$23,162

$32,194

$68,022

$123,379

$3,298

$4,584

$9,685

17,566

8
1

37
8
1

94
8
1

$26,471

$21,463

$32,011

$79,944

$3,769

$3,056

$4,558

11,382

Total

177

Irrigation water sampling plan(§ 112.150(b)(3))
Number of sprout operations
Time burden
Frequency
Recordkeeping cost of water sampling
plan
NPV (@7%)

46

177

Record of analytical method for additional pathogen testing(§§ 112.150(b)(5), 112.44(c))
Number of sprout operations
Time burden
Frequency
Recordkeeping cost of analytical method
NPV (@7%)
Total one-time recordkeeping costs by
size category
Annualized one-time recordkeeping

46

177

2
1
$6,618
$942

37
2
1
$5,366
$764

94
2
1
$8,003
$1,139

$19,986
2,846

$56,251

$59,023

$108,036

$223,309

$8,009

$8,404

$15,382

$31,794

Page 93

costs by size category (7 percent for 10
years)

We estimate that sprouting operations not already performing certain
recordkeeping activities will incur recurring recordkeeping costs, including
documentation of seed treatment (§ 112.150(b)(1)), environmental monitoring plan annual maintenance (§ 112.150(b)(2)), environmental monitoring test results (§
112.150(b)(4)), spent irrigation water sampling plan – annual maintenance (§
112.150(b)(3)), spent irrigation water test results (§ 112.150(b)(4)), and documentation of
corrective actions taken under §§ 112.142(b) and (c), 112.146, and 112.148 (§
112.150(b)(6)).
We estimate that records of documentation of seed or bean treatment (including
documentation of previous treatment by a third party) ((§ 112.150(b)(1)), will need to be
documented by 128 sprouting operations not already performing these activities. This
record will need to be made 50 times for small and very small operations, and 223 times
for large operations, based on the number of batches. We estimate that this record will
take approximately 12 minutes to make (20 percent of one hour). These time burdens
multiplied by the number of sprouting operations and by the farm operator wage rate to
estimate an annual record of seed treatment recordkeeping cost of $173,015.
Environmental monitoring plan- annual maintenance recordkeeping (§
112.150(b)(2)) will need to be documented by 177 sprouting operations not already
performing these activities. This record will need to be made once annually by each
operation. We estimate that this record will take approximately 9 minutes to make (15
percent of one hour). These time burdens are multiplied by the number of sprouting

Page 94

operations and by the farm operator wage rate to estimate an annual environmental
monitoring plan- annual maintenance recordkeeping cost of $1,499.
Environmental monitoring test result records (§ 112.150(b)(4)) will need to be
documented by 128 sprouting operations not already performing these activities. This
record will need to be made 60 times for very small operations, 120 times for small
operations, and 180 times for large operations, based on the number of tests conducted.
We estimate that this record will take approximately 10 minutes to make (17 percent of
one hour). These time burdens are multiplied by the number of sprouting operations and
by the farm operator wage rate to estimate an annual environmental monitoring test result
recordkeeping cost of $153,088.
Spent irrigation water sampling plan – annual maintenance recordkeeping (§
112.150(b)(3)) will need to be documented by 177 sprouting operations not already
performing these activities. This record will need to be made once for each operation. We
estimate that this record will take approximately one hour to make. These time burdens
are multiplied by the number of sprouting operations and by the farm operator wage rate
to estimate an annual spent irrigation water sampling plan – annual maintenance
recordkeeping cost of $9,993.
Spent irrigation water test results records (§ 112.150(b)(4)) will need to be
documented by 128 sprouting operations not already performing these activities. This
record will need to be made 125 times for very small and small operations, and 558 times
for large operations, based on batches. We estimate that this record will take
approximately 9 minutes (15 percent of one hour) to make. These time burdens are

Page 95

multiplied by the number of sprouting operations and by the farm operator wage rate to
estimate an annual spent irrigation water test results recordkeeping cost of $324,403.
Documentation of corrective actions taken under §§ 112.142(b) and (c), 112.146,
and 112.148 (§ 112.150(b)(6)) will need to be documented by 285 sprouting operations.
This record will need to be made once for each corrective action. We estimate that this
record will take approximately 30 minutes (50 percent of one hour) to make. These time
burdens are multiplied by the number of sprouting operations and by the farm operator
wage rate to estimate an annual corrective action recordkeeping cost of $8,059.
Each of these time burdens is multiplied by the hourly wage rate for farm
operators at very small, small, and large operations. Table 29 presents the recurring
recordkeeping costs for the sprouts provisions. We estimate the total recurring
recordkeeping costs for sprouts are $100,016 for very small operations, $100,956 for
small operations, and $469,085 for large operations.
Table 29. Recurring Recordkeeping Costs for Sprouts
Recurring recordkeeping
Very small
costs
operations
Documentation of seed treatment (§ 112.150(b)(1))

Small
Operations

Large
Operations

Total

Number of sprout operations

33

27

68

Time burden

50

50

223

Frequency
0.20
0.20
Recordkeeping cost of seed
$24,016
$19,472
treatment
Environmental monitoring plan – annual maintenance(§ 112.150(b)(2))
Number of sprout operations
Time burden

46
1

Frequency
0.15
Recordkeeping cost of
environmental monitoring $496
annual maintenance
Environmental monitoring test results(§ 112.150(b)(4))

37

0.20
$129,527

$173,015

94

177

1

1

0.15

0.15

$402

$600

$1,499

128

Number of sprout operations

33

27

68

Time burden

60

120

180

0.17

0.17

0.17

Frequency

Page 96

128

Recordkeeping cost of
environmental monitoring test
$24,496
$39,724
$88,868
results
Spent Irrigation water sampling plan –annual maintenance(§ 112.150(b)(3))
Number of sprout operations

$153,088

46

37

94

1

1

1

Frequency
1
Recordkeeping cost of water
sampling plan - annual
$3,309
maintenance
Spent irrigation water test results(§ 112.150(b)(4))

1

1

$2,683

$4,001

$9,993

27

68

128

Time burden

Number of sprout operations

33

177

Time burden

125
125
558
Frequency
0.15
0.15
0.15
Recordkeeping cost of spent
irrigation water test results
$45,030
$36,511
$242,862
$324,403
Recordkeeping costs of corrective actions taken under §§ 112.142(b) and (c), 112.146, and 112.148 (§
112.150(b)(6))
Number of sprout operations
74
60
151
285
Time burden
Frequency
Recordkeeping cost of spent
irrigation water test results
Total recurring
recordkeeping costs by size
category

1

1

1

0.50

0.50

0.50

$2,668

$2,164

$3,227

$8,059

$100,016

$100,956

$469,085

$670,057

Table 30 presents a summary of recordkeeping costs. The total costs of
recordkeeping are $16 million for very small farms, $4.2 million for small farms, and
$7.3 million for large farms, totaling to $27.5 million for all farms.
Table 30. Summary of Recordkeeping Costs (annually, in thousands)
Recording activity
Qualified exempt farms labeling and documentation
(§ 112.7)
Agricultural water
(§ 112.50)
Biological soil amendments of animal origin
(§ 112.60)
Equipment, tools, buildings, and sanitation
(§ 112.140)
Sprouting operations
(§ 112.150)
Training
(§ 112.30)
Documentation relating to commercial processing
exemption

Very Small

Page 97

Small

Large

Total

$5,239

$469

$0

$5,709

$4,510

$829

$1,043

$6,382

$184

$32

$40

$256

$4,829

$2,620

$5,492

$12,941

$108

$109

$484

$702

$1,069

$186

$227

$1,482

$13

$3

$3

$18

(§ 112.2(b)(4))
Total cost (annual in thousands)

$15,951

$4,249

$7,290

$27,490

10. Administrative Provisions

We did not receive substantial comments on the cost estimates for Administrative
Provisions; therefore, we have not altered the underlying methodology from those
originally proposed and estimated in the PRIA. In addition, our changes to the proposed
requirements in finalizing those provisions do not affect our cost estimates. Thus, we
present the estimates utilizing more current wage information and farm counts. Table 31
provides the total cost for Administrative Provisions.
In total we estimate that learning about the rule will cost all farms approximately
$23 million, annualized at 7 percent over ten years. These costs are comprised of all
qualified exempt and non-covered farms spending 4 hours with the rule, which was
lowered from 10 hours estimated in the PRIA based on public comment and feedback
from public meetings. Very small covered farms are estimated to spend 40 hours with the
rule, and small and large covered farms spend 40 hours with the rule as well as 40 hours
of legal review (for a total of 80 hours); these estimates have not been altered from those
originally proposed.
Table 31. Total Costs of Reading and Learning about the Rule Requirements
Number of qualified
exempt and non-covered
farms
Farm operator wage
Time reading and learning
rule
Per farm learning cost
Cost to learn about the rule
Number of covered farms
Farm Operator Wage

Exempt

Very Small

Small

Large

74,931

30,952

5,128

10,105

Total

121,116
$42.74

$72.12

$72.12

$42.74

4

4

4

4

$171

$288
$8,929,032.
96
22,781
$72.12

$288

$171

$1,479,325

$1,727,551

3,956
$72.12

8,292
$42.74

$12,810,204
0

Page 98

35,029

Time reading and learning
rule
Legal analyst wage
Time reading and learning
rule
Per farm learning cost
Cost to learn about the rule
Total One Time Cost
Costs annualized over 10
years

40

40

40

$96.00

$96.00

40

40

$2,885
$65,718,629

$6,725
$26,603,309

$5,550
$46,017,283

$12,810,204

$74,647,662

$28,082,634

$47,744,834

$1,823,885

$10,628,148

$3,998,335

$6,797,790

$163,285,334
$23,248,158

11. Corrective Steps

Although the requirements have not changed dramatically from those proposed in
the original rule, our estimates of Corrective Steps have increased from those originally
provided. Primarily in response to comments received on the economic analysis, we have
doubled the frequency at which we estimate that corrective actions may occur. Otherwise,
we generally retain our costs methodology from those in the PRIA. The analysis include
all steps taken under 112.45, for example, when agricultural water is not safe/adequate or
fails to meet a microbial standard, and all the steps required in subpart M for sprouters
when they get an environmental positive or a batch pathogen positive (required under
112.146 and 148). Our changes to the proposed requirements for corrective actions were
in relation to the requirements for agricultural water and sprouts. Thus, we present only
summary statistics of estimates utilizing more current wage information and farm counts.
Table 32 provides the total cost for Corrective Steps related to agricultural water and
sprouts; for full information on how these costs are estimated please refer to Tables 119 –
120 of the original PRIA(Ref. 6).
Table 32. Summary of Costs of Corrective Steps (in thousands)
Very Small

Small

Large

Total

Failed standards Directed to Agricultural Water

$412

$97

$260

$770

Failed standards Directed to Sprouts

$322

$336

$1,818

$2,476

Page 99

Total Costs of Corrective Steps (annual)

$735

$433

$2,078

$3,246

12. Variances

We did not receive substantial comments on the cost estimates for Variances;
therefore, we have not altered the underlying methodology from those originally
proposed and estimated in the PRIA. In addition, our changes to the proposed
requirements in finalizing subpart P do not substantively affect our cost estimates. Thus,
we present the estimates utilizing more current wage information and a slightly increased
number of applicants, to account for the allowance for tribal applications. Table 33
provides the total cost for Administrative Provisions.
Table 33. Total Costs of Preparing and Reviewing Initial Petition
Cost Components
Hours to complete petition

80

Wage (GS 14.1)

$75.62

Cost to complete petition

$6,049.60

Hours to internally review

40

Wage (GS 15.3)

$94.88

Cost to internally review petition

$3,795.20

Cost to complete & review

$9,844.80

Hours for FDA review

80

Wage (GS 13.7)

$76.79

Cost for FDA review

$6,143.20

Total individual cost of petition

$15,988

Potential number of applicants

7

Total Cost of Preparing and Reviewing Final Petition

$111,916

13. Summary of Costs

The total costs by standard in the rule and other sections are summarized in Table
34 by farm size. The “not covered” category only includes the 74,931 farms that
generate an average annual monetary value of produce sold of $25,000 or less. All farms

Page 100

either covered or not by the rule would incur the costs to learn the rule. In addition to
learning the rule, the 30,952 covered by the rule would incur the costs of implementing
the standards directed to personnel health and hygiene; agricultural water; domesticated
and wild animals; growing, harvesting, packing, and holding activities; equipment, tools,
buildings, and sanitation; personnel qualifications and training; sprouts (only for sprout
farms); and recordkeeping.
Farms that are eligible for a qualified exemption would incur costs to not only
learn the rule and retain documentation demonstrating their eligibility for the qualified
exemption, but also costs to change labels if necessary or otherwise disclose their name
and complete business address at the point of sale. For farms that grow, harvest, pack, or
hold produce that receives commercial processing that adequately reduces the presence
of microorganisms of public health significance, costs will be incurred in making
required disclosures and receiving and maintaining records of written assurances from
customers. The costs to these farms of these requirements are included in the total
recordkeeping costs of the rule.
The estimates in Table 34 are reported in millions for ease of readability with the
exception of the average cost per farm estimates, which are reported with no abbreviation.
Table 34. Summary of Costs for the Produce Safety Rule (in millions)
Cost Sections
Personnel Qualifications and
training
Health and Hygiene
Agricultural water
Biological soil amendments of
animal origin
Domesticated and wild animals
Growing, harvesting, packing,
and holding activities
Equipment, tools, buildings, and

Not
Covered

Very Small

Small

Large

Total

$0.00

$41.14

$54.08

$92.16

$187.38

$0.00

$28.11

$13.59

$93.91

$135.61

$0.00

$18.41

$4.21

$14.45

$37.07

$0.00

$0.44

$0.31

$1.72

$2.47

$0.00

$2.36

$2.05

$11.45

$15.86

$0.00

$0.92

$0.35

$0.98

$2.25

$0.00

$19.49

$13.91

$85.29

$118.69

Page 101

sanitation
Sprouting operations

$0.00

$0.82

$0.94

$5.00

$6.77

Recordkeeping
Administrative cost to learn the
rule
Corrective steps

$5.71

$10.71

$3.78

$7.29

$27.49

$1.82

$10.63

$4.00

$6.80

$23.25

$0.00

$0.73

$0.43

$2.08

$3.25

Variances

$0.00

$0.00

$0.00

$0.11

$0.11

Total Costs (annual in millions)

$7.53

$133.76

$97.65

$321.24

$560.19

Average Cost per farm

$101

$5,872

$24,683

$38,741

$15,992

The costs of the rule may decrease over time as farms learn by doing. However,
these costs of this rule will not be immediately realized, nor will they be uniformly
implemented, due to the staggered nature of compliance times. Table 35 presents the
annual estimates of costs as they are estimated to occur.

Year1

Year 2

Year 3

Year 4

Year 5

Year 6

Year 7

Year 8

Year 9

Year 10

Table 35. Timing of Produce Costs (in millions)

Covered Farms
Very
$0
Small
Small
$0

$0

$0

$0

$115

$115

$133

$133

$133

$133

$0

$0

$92

$92

$97

$97

$97

$97

$97

$0

$0

$302

$302
$316
$316
Covered Sprout operations

$316

$316

$316

$316

$0

$0

$0

$1

$1

$1

$1

$1

$1

$1

$0

$0

$1

$1

$1

$1

$1

$1

$1

$1

$0

$5

$5

$5

$5

$5

$5

$5

$5

Farms

Large
Very
Small
Sprouts
Small
Sprouts
Large
Sprouts

$5
Exempt Farms

Very
Small
$0
$0
$0
$0
$7
$7
$7
$7
$7
$7
Exempt
Small
$0
$0
$0
$1
$1
$1
$1
$1
$1
$1
Exempt
Large
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Exempt
Note: Summing across a single year gives a single year cost of full may not match the actually estimated
cost of this rulemaking due to rounding errors in this table, which is meant for illustrative purposes.

Page 102

Next, we annualize estimates of the costs below in Table 36. In this estimate, we
take into account the time that different sized farms have to comply with the rule, as well
as the different compliance times for agricultural water provisions and for activities
relating to sprouts. Estimates are annualized over 10 years. We estimate that the
annualized costs of the final rule would be approximately $368 million per year using a
discount rate of 7 percent over 10 years. The average cost per covered farm is $10,351.
We note that within size categories costs borne by individual farms will diverge widely
from the averages reported here, depending upon whether or not the farm is already in
compliance with most of the provisions of the rule.
Table 36. Summary of Costs for the Produce Safety Rule Considering Time to
Comply with the Rule (in millions)
Cost Sections
Personnel Qualifications and
training
Health and Hygiene
Agricultural water
Biological soil amendments of
animal origin
Domesticated and wild
animals
Growing, harvesting, packing,
and holding activities
Equipment, tools, buildings,
and sanitation
Sprouting operations
Recordkeeping
Administrative cost to learn
the rule
Corrective steps

Not Covered

Very
Small

Small

Large

Total

$0.00

$21.30

$33.87

$68.44

$123.61

$0.00

$14.55

$8.51

$69.74

$92.80

$0.00

$6.48

$1.87

$7.76

$16.11

$0.00

$0.23

$0.16

$0.89

$1.28

$0.00

$1.22

$1.28

$8.50

$11.01

$0.00

$0.48

$0.22

$0.73

$1.42

$0.00

$10.09

$8.71

$63.33

$82.14

$0.00

$0.52

$0.70

$4.34

$5.55

$4.24

$5.55

$2.37

$5.41

$17.57

$1.35

$5.50

$2.50

$5.05

$14.41

$0.00

$0.38

$0.27

$1.54

$2.19

$0.00

$0.08

$0.08

$60.47

$235.82

$368.17

$15,285.87

$28,438.88

$10,350.83

Variances
$0.00
$0.00
Total Costs (annual in
$5.59
$66.29
millions)
Average Cost per farm*
$74.65
$2,910.02
Note: Average costs values not reported in millions.

Page 103

Annualizing costs over the first ten years after publication of this final rule, costs
are expected to be approximately at $368 million annually at 7 percent and $389 million
at 3 percent.
Table 37. Net Present Value and Annualized Costs of the Produce Safety Rule (in
millions)
Net present value at 3
percent
Net present value at 7
percent
Annualized at 3 percent
over 10 years
Annualized at 7 percent
over 10 years

Exempt

Very Small

Small

Large

Total

$37

$613

$550

$2,104

$3,304

$28

$462

$424

$1,657

$2,571

$4

$72

$65

$247

$387

$4

$66

$60

$236

$366

$16,304

$29,749

$11,059

$15,265

$28,452

$10,449

Average Cost Per Farm at
3 percent
$58
$3,155
Average Cost Per Farm at
7 percent
$53
$2,885
Note: Average costs values not reported in millions.

G. Distributional Effects
We do not expect that the rule will have any adverse distributional effects on any
one specific party. That is, depending on how the farms in the affected markets respond
to these requirements, some of the costs may ultimately be borne by consumers as price
increases. The higher prices, however, will likely not be sufficient to fully offset the costs
borne by food establishments. As an overly simple example, if 100 percent of the costs
of this rule were passed along directly to consumers this would increase the market price
for fresh produce by only 2.1 percent ($231+ foreign costs + $560 domestic costs million
divided by $38 billion). Additionally, it is highly unlikely that any one party, either
consumers or industry, will bear the entire burden of costs from compliance with this rule.
Rather, the costs will likely be shared amongst all parties based on numerous factors such
Page 104

as the relative price elasticity of the produce market and producers’ ability to set prices in
the marketplace.

H. International Effects
For the FRIA, we retained the methodology for the number of foreign farms that
will be covered by our rule based on the latest number of foreign farms shipping produce
to the US. As with domestic farms, we adjust these numbers based on new data sources.
Our estimate for the total number of foreign farms exporting produce to the US is
approximately 45,000. Of those farms exporting RACs to the US, we estimate that
approximately 13,000 might incur compliance costs to continue exporting to the US. 10
Because we lack survey data about baseline foreign farms’ food safety practices and the
likely costs to incorporate all the changes to comply with the rule, we estimate the costs
by assuming that the average costs will be the same for foreign and domestic farms; they
will have the same proportion of baseline practices and the same proportion of farms not
covered or eligible for an exemption. Applying the average annualized cost of the rule
for domestic farms of roughly $10,000 per farm using a 7 percent discount rate ($11,000
at a 3 percent discount rate) yields an estimated total annualized cost to foreign
operations of $136 million ($146 million using a 3 percent discount rate). Additionally,
those farms that are exempt from or not covered by the rule are estimated to incur the
same average costs of domestic exempt or non-covered farms. Applying the average
annualized cost of the rule for domestic farms of roughly $53 per farm using either a 7
10

This estimate is derived from the total number of entities importing RACs from OASIS data (45,000)
multiplied by the percent of domestic farms that are covered by this rulemaking, 29 percent (35,029
covered farms divided by 121,116 total farms). The methodology has not changed from the proposed
analysis but both sources of data are now updated.

Page 105

percent discount rate ($58 using a 3 percent discount rate) yields an estimated total
annualized cost to exempt or non-covered foreign operations of $1.7 million using a 7
percent discount rate ($1.8 million using a 3 percent discount rate). Together, we estimate
an annual cost to foreign farms shipping produce RACs to the US of $138 million
annualized, using a 7 percent discount rate ($146 million using 3 percent).
This analysis may overstate or understate the true cost to foreign farms. From our
OASIS data, we know that foreign operations will often only send a small fraction of
their total production to the US and therefore our estimate is likely the upper bound
estimate. If average foreign wage rates are significantly lower than average US wage
rates, if total production costs are lower, or if some foreign farms simply cease to ship
their products to the US because of the regulatory compliance costs, the total costs to
foreign farms might be significantly less. Conversely, if fewer foreign farms are already
preforming some of the required activities, or if average foreign wage costs are higher,
then the total costs to foreign farms could be higher.

I. Uncertainty and Sensitivity Analysis
1. Costs
A source of uncertainty is our FVAP survey (Ref. 20) The survey is older data,
from 1999, and it is highly likely that the produce industry has made significant
improvements in safety measures since it was originally conducted. There has been a
growing industry wide understanding of the benefits of safe food handling practices and
more and more establishments are adopting some food safety controls. If the survey
overstates the number of operations that lack our controls today by 25 percent, to account
Page 106

for trends in industry practices, the total costs of the rule would decline to $301 million as
shown in Table 38.
In addition, it could be that farm food safety practices have actually decreased
since this survey was conducted. Therefore we additionally lower the percentage
compliance rates by 10% to more fully capture the variability inherent in this analysis.
We adjust compliance percentages downwards somewhat less than we adjusted upwards,
because we believe that it is much less likely that farms have regressed in their safety
activities since the survey was conducted. If the survey understates the number of
operations that lack practices compliant with part 112 today by 10 percent, the total costs
for the final rule would rise to $401 million as shown in Table 38.
The costs of the water provisions are another source of uncertainty we address in
our sensitivity analysis. We raise water provision compliance rates by 25 percent in our
low estimate and decrease them to zero percent in our high estimate. In addition, because
the costs to treat water are somewhat more uncertain than some other cost estimates, we
also lower water treatment costs to $32 in our low estimate and raise water treatment
costs to $543 in our high estimate, to capture the full potential range of marginal water
treatment costs. Because water costs represent about 6.6 percent of the total costs of the
rule, substantial changes such as doubling or halving them would only result in a 6.6
percent increase or a 3.3 percent decrease in the total costs of the rule.
Table 38. Sensitivity Analysis of Costs (in millions)
Low

High

Annualized at 3 percent

$319

$425

Annualized at 7 percent

$301

$401

2. Benefits

Page 107

Previously presented benefits are mean values derived from multiple data ranges
and distributions. In order to more fully characterize the expected benefits of this rule and
highlight the uncertainty built into this estimation, we present ranges for estimates. Our
primary outcomes of interest are presented below in Table 39. For simplicity of
interpretation, we only examine the total outcomes, but all estimates previously presented
were derived from multiple distributions, including the annual incidence, full costs per
pathogen, and efficacy estimates. In our sensitivity analysis below, we run Monte Carlo
simulations in which these values vary based on our calculated parameters of their
distributions (mean, 5th percentile, 95th percentile). This allows us to calculate low (5th
percentile) and high (95th percentile) estimates of the benefits.
Table 39. Sensitivity Analysis of Benefits (in millions)
Illnesses
Low
Annualized at 3 percent
Annualized at 7 percent

273,227
250,212

High

Benefits (millions)
Low

High

449,626

$748

$1,195

412,504

$710

$1,132

Another source of uncertainty in the estimation of benefits is the data on reported
outbreaks associated with FDA-regulated produce RACs. The incidence of reported
outbreaks varies by year, with some periods of time experiencing more of these outbreaks
than others. Because our estimated number of total outbreaks related to FDA regulated
produce RACs is calculated as the ratio of reported FDA regulated produce RAC
outbreak illnesses to total CDC identified illnesses, the variability in the reported FDA
regulated produce RAC outbreak illnesses may lead to an overestimation or
underestimation of the total outbreaks related to FDA regulated produce RACs. If the
data span used encompasses a time period with a relatively low incidence of reported

Page 108

FDA regulated produce RAC outbreak illnesses, it may lead to an underestimation of the
total outbreaks related to FDA regulated produce RACs, while if it encompasses a time
period with a relatively high incidence of reported FDA regulated produce RAC outbreak
illnesses, it may lead to an overestimation of the total outbreaks related to FDA regulated
produce RACs.
For example, if we examine only the time frame available for the PRIA, 20032008, our total estimated benefits would be slightly below $900 million, as opposed to
the $1.4 billion in steady state benefits we currently estimate; a reduction of
approximately 35 percent. Additionally, if we were to exclude the year with the most
total reported illnesses attributable to FDA RACs, 2011, our total estimate of benefits
would fall by approximately 42 percent, to approximately $810 million, annually.
Conversely, if we were to exclude the year with the least total reported illnesses, 2007,
our total estimate of benefits would rise by approximately 8 percent, to approximately
$1.5 billion, annually.
3. Net Benefits
Finally, we compare the range of estimate benefits to the range of estimate costs.
This information is presented in Table .
Table 40. Sensitivity Analysis of Net Benefits (in millions)
Benefits
Costs
Net Benefits

Low
$1,059

Mean
$1,389

High
$1,719

$301
$758

$366
$1,023

$390
$1,329

J. Analysis of Regulatory Alternatives to the Rule
FDA identified and assessed several regulatory alternatives including: (1) relying

Page 109

on non-regulatory solutions, (2) a lower or higher monetary value threshold for farms not
covered under the rule, (3) longer or shorter compliance periods, and (4) reduced
requirements.
1. Non-regulatory Solutions
In the absence of FSMA, under this alternative, FDA could rely on some or all of the
following:
•

voluntary recommendation of some or all provisions of the regulation,

•

current or enhanced State and local enforcement of existing state or local laws to
bring about a reduction of potential harm from contaminated produce, or

•

the tort system, with litigation or the threat of litigation serving to bring about the
goals of the rule.
The advantage of this alternative is that it is already in place and the produce

industry generally understands the requirements in the rule. The disadvantage of this
alternative is that the regime lacks several of the most important provisions of the rule
that have the potential to prevent avoidable foodborne illnesses that we estimate are
worth approximately $976 million per year.
By voluntarily introducing procedures, establishments that do so demonstrate that
their expected private economic benefits will exceed their private costs. Voluntary
adoption of any practices will occur when it is profitable to do so. Although many
establishments have adopted some food safety practices in order to meet the public
demand for safer produce, numerous surveys show that many farms have not adopted the
practices that provide socially optimal levels of food safety.
Public and private health agencies, consumer groups, competitors, trade

Page 110

organizations or other independent parties could publicize the risks from produce not
grown, harvested, packed or held using appropriate practices and allow consumers to
decide for themselves about the risks of adulteration. The weakness of this approach is
that independent organizations cannot discover food safety hazards until after consumers
are sickened. In the absence of the produce safety standards, the burden of monitoring
safety practices fall more heavily on consumers.
Finally, FSMA requires that we issue a Produce Safety regulation. Therefore, this
is not a legally viable alternative.
2. Lower or Higher Monetary Value Threshold for Farms not Covered
The rule does not cover farms with $25,000 or less in annual produce sales. As
this monetary value threshold falls, the number of farms not covered will fall. Table 41
shows the costs and benefits for a monetary value threshold of $10,000 in annual produce
sales.
Table 41. Lower Monetary Value Threshold for Farms not Covered
Annualized Costs
Annualized Benefits

7%
$460
$940

3%
$489
$991

Conversely, as this monetary value threshold rises, the number of farms not
covered rises. Table 42 shows the costs and benefits for a monetary value threshold of
$100,000 in annual produce sales.
Table 42. Higher Monetary Value Threshold for Farms Not Covered
Annualized Costs
Annualized Benefits

7%
$345
$899

3. Shorter or Longer Compliance Periods

Page 111

3%
$364
$938

The rule could have established shorter compliance periods, such as one year for
farms of all sizes. With a one year compliance period, the affected farms would need to
begin the process of compliance immediately. With a one-year compliance period, the
costs increase to $438 million, and smaller farms with fewer resources must adopt the
requirements in a time period that does not allow them to adopt the requirements
correctly or fully, which might add to their costs and not add to public health. Moreover,
FSMA establishes certain minimum compliance periods, so this is not a legally viable
option. Table 43 shows the benefits and costs under this option.
Table 43: One-year Compliance Period
Annualized Costs
Annualized Benefits

7%
$435
$1,089

3%
$450
$1,125

The rule could have established a longer compliance period for all affected farms,
such as three years for large farms and a corresponding extra year for all other farms.
With a three -year compliance period, the affected farms would have more time to
implement the produce safety standards required by the rule. With a three-year
compliance period, the costs decrease to $308 million as smaller operations with fewer
resources are able to implement the requirements in a time period that would allow them
to adopt them correctly or fully.
Table 44. One Extra Year Compliance Period (3 years for Large Farms)
Annualized Costs
Annualized Benefits

7%
$307
$771

3%
$331
$830

4. Fewer Requirements
Under this Option, the rule could establish less extensive requirements. Several
provisions could be combined to provide a less extensive set of standards than those in
Page 112

the rule. Certain prevention measures could be separated and put forth as stand-alone
regulations; for example, requirements regarding agricultural water could be issued as a
separate rule. As an alternative, certain provisions could be eliminated altogether; for
example, as shown in Table 45, eliminating provisions related to domesticated and wild
animals and growing, harvesting, packing, and holding activities would reduce the cost of
the rule by nearly $12 million; however, potential benefits would also be reduced by
about $154 million. Another alternative shown in Table 45 is eliminating provisions
related to agricultural water for growing or harvest pathway activities, which would
reduce the cost of the rule by nearly $16 million; however, potential benefits would also
be reduced by about $127 million (annualized at 3 percent).
It is not possible to present each combination of provisions as separate options;
however, the individual effects of the various on-farm prevention measures can be seen in
the summary of costs and benefits. Dropping measures would, individually, generate
lower costs than the integrated program outlined in the rule. However, we also expect that
dropping measures would, individually, lead to the number of illnesses prevented being
lower than in the integrated program outlined in the text.
Table 45. Fewer Requirements
Eliminating provisions related to domesticated and wild animals and growing, harvesting, packing, and
holding activities
7%
3%
Annualized Costs

$354

$374

Annualized Benefits

$778

$822

Eliminating provisions related to agricultural water for growing or harvest pathway activities
7%

3%

Annualized Costs

$351

$371

Annualized Benefits

$808

$849

5. Summary of Alternatives

Page 113

Table 46 summarizes the costs and benefits of the rule and under several
regulatory alternatives.
Table 46. Summary of Regulatory Alternatives (Present Values, $ million)
Alternative
Lower monetary value threshold for farms
not covered
Higher monetary value threshold for farms
not covered
One-year compliance period for all farms
Three-year compliance period for all farms
Fewer requirements: domesticated and wild
animals
Fewer requirements: agricultural water
The Rule, as finalized

Incremental
Total
Incremental
Total
Incremental
Total
Incremental
Total
Incremental
Total
Incremental
Total
Incremental
Total

Costs
at 3%
$102
$489
-$23
$364
$63
$450
-$56
$331
-$13
$374
-$16
$371
-$387

Benefits
at 3%
$15
$991
-$38
$938
$149
$1,125
-$146
$830
-$154
$822
-$127

Costs
at 7%
$94
$460
-$21
$345
$69
$435
-$59
$307
-$12
$354
-$15

Benefits
at 7%
$15
$940
-$26
$899
$164
$1,089
-$154
$771
-$147
$778
-$117

$849
-$976

$351
-$366

$808
-$925

Note: incremental costs and benefits are relative to previously-listed alternative.

III. Final Small Entity Analysis
The Small Business Regulatory Flexibility Act requires agencies to analyze
regulatory options that would minimize any significant impact of a rule on small entities.
Small entities have fewer resources to devote to regulatory compliance and, therefore,
may be more affected by regulatory compliance costs. The agency finds that the rule will
have a significant economic impact on a substantial number of small entities.

A. Description and Number of Affected Small Entities
The Small Business Administration defines farms involved in crop production as
“small” if their total revenue is less than $750,000 (Ref. 45). Approximately 95 percent
of all farms that grow covered produce are considered small by the SBA definition, and

Page 114

these farms account for 62 percent of covered produce production. Exempting all of
these small entities would substantially reduce the expected health benefit of the rule.
As described in the preamble, section 419(a)(3)(F) of the FD&C Act requires
FDA to define the terms “small business” and “very small business.” For purposes of
this rule, FDA has defined a small business as a farm that is covered by the rule whose
average annual monetary value of produce, on a rolling basis, sold during the previous
three-year period is no more than $500,000, and that is not a very small business. FDA
has defined a very small business in part 112, as a farm that is covered by the rule and
whose average annual monetary value of produce, on a rolling basis, sold during the
previous three-year period is no more than $250,000. See § 112.3(b). The definitions for
small business and very small business exclude farms that are not subject to the rule per §
112.4(a), that is, farms with $25,000 or less in average annual monetary value of produce
sold. Approximately 3,956 farms that are covered by the rule are considered small
businesses under the rule, and these farms account for 5 percent of covered produce.
Approximately 22,781 farms that are covered by the rule are considered very small
businesses under the rule, and these farms account for 9 percent of covered produce.
The rule reduces the burden on small entities in part through the use of
exemptions: certain small entities are eligible for a qualified exemption based on average
monetary value of food sold and direct sales to qualified end users (§ 112.5). The rule
additionally reduces the burden on small entities by not covering farms with $25,000 or
less of average annual monetary value of produce sold (§ 112.4(a)). The rule additionally
provides all farms flexibility for alternative practices to be used for certain specified
requirements related to agricultural water, provided the farm has adequate scientific

Page 115

support (see §§ 112.12 and 112.49). The rule also provides for States, Tribes, and foreign
countries to submit a request for a variance for one or more requirements of the rule. To
be granted, the procedures, processes, and practices to be followed under the variance
must be reasonably likely to ensure that the produce is not adulterated under Section 402
of the Act and to provide the same level of public health protection as the requirements of
the rule.
Farms (except sprout operations) defined as small businesses have 3 years to
comply with most provisions of the rule after the effective date of the rule, and farms
(except sprout operations) defined as very small businesses have 4 years. There is also an
additional 2-year compliance period beyond the respective compliance date for certain
requirements related to agricultural water. See section XXIV of the rule.
Table 47 summarizes the total number of domestic farms covered by the rule, the
percentage of covered farms and produce they account for, and their average annual
monetary value of food sold by size. For purposes of the small business analysis,
Columns 2 and 3 of the table identify the farms that meet our definition of a very small
and small business, respectively.
Table 47. Covered Farms in the Rule
Number of covered farms
Percentage of covered farms
Percentage of produce acres
Average annual monetary value of food

Very
Small
22,781
66%
9%
$86,000

Small

Large

Total

3,956
11%
5%
$360,000

8,292
23%
60%
$3,450,000

35,029
100%
74%
$882,000

B. Description of the Potential Impacts of the Rule on Small Entities
The costs to implement the rule will vary across farms as their current practices
vary, and farms whose practices, processes, or procedures are not already in compliance
Page 116

with the requirements will bear the costs for compliance. If a farm’s profit margin is
significantly reduced after the regulatory costs are subtracted from its pre-regulatory
revenues, then the farm will be at risk of halting production of the crops that it deems too
costly to grow, pack, harvest, and hold. Regulatory cost burdens tend to vary across
different-sized farms. Farm size is an important determinant of regulatory impacts and
for determining business risk. Small entities with above average costs of doing business
will be at a competitive disadvantage. Some small entities might determine that their
new expected costs are likely to exceed their revenues.
This may be especially true for small sprouting operations, whose average costs
of compliance may be higher due to the additional requirements on their production. We
estimate that average revenues for very small sprouting operations are approximately
$49,000 and small sprouting operations are $67,000. Average costs to very small and
small sprouting operations estimated to be approximately $17,000, or approximately 36
and 26 percent of revenues for very small and small sprouting operations, respectively.
These costs are in addition to the other applicable costs of the rule for sprouting
operations.
Table 48 shows the average costs and average upfront costs of implementing the
requirements of the rule (annualized at 7 percent over 10 years) as a percentage of the
average annual monetary value of food sales per very small and small farm. For
comparison, we include the results for large farms. Average costs make up 3 percent of
the average food sales for very small farms and 4 percent for small farms. Small and
very small farms whose practices, processes, or procedures are not already in compliance

Page 117

with a significant portion of the requirements will incur a larger cost than the average
shown.
Table 48. Average Costs of Implementing Proposed Rule as Percentage of Food
Sales by Farm Size
Very
Small

Small

Large

All Farms

Average costs of implementing provisions in the
$2,885
$15,265
$28,452
$10,449
proposed rule
Average upfront costs of implementing
provisions in the proposed rule
$5,027
$23,382
$36,396
$14,525.69
Average annual monetary value of food sold
$86,000
$360,000
$3,450,000
$882,000
Average costs percentage of average annual
3%
4%
1%
1%
monetary value of food sold
Note: Because of the timing of the rule, farms will incur upfront costs in different years. Average upfront
costs to firms are estimated here by calculating the average cost for farms of different sizes based on the
first year in which they incur costs. Additionally, this estimate does not include the costs of the water
provisions as these costs are further delayed for farms of all sizes.

C. Alternatives to Minimize the Burden on Small Entities
In the final rule, we have introduced several provisions for regulatory relief for
small entities. The most important are the modified requirements for businesses that
qualify for a “qualified exemption.” In addition, small and very small businesses have
additional time to comply with the requirements: small businesses (except sprout
operations) have three years and very small businesses (except sprout operations) have
four years to come into compliance after the effective date of the final rule. This is an
additional 12 months or 24 months, respectively, beyond the time given to larger
operations to comply with this rule. We have also provided for extended compliance
dates for certain agricultural water requirements for all covered farms with respect to
covered produce other than sprouts. See section XXIV of the rule.
The final rule provides substantial cost relief to small businesses. We identified
two other options for regulatory relief that were not adopted.
a. Longer compliance period for small businesses

Page 118

Small entities may find it more difficult to learn about and implement the
requirements than it will be for large entities. Lengthening the compliance period for
small businesses beyond the additional time we currently allow would provide some
additional regulatory relief by allowing small businesses to take advantage of increases in
industry knowledge and experience in implementing these regulations. A longer
compliance period will allow additional time to learn about the requirements of the rule,
to hire or train workers, to take samples for their initial water quality survey, to purchase
new or replacement equipment, to arrange financing and for any other initial expenditure
of time, effort and money. It will also delay the impact of the annual costs of compliance.
The annualized costs savings from the delay are estimated to be approximately $70
million.
b. Fewer Requirements
The alternative to only require certain provisions and not require others (for
example, not require small businesses to comply with the standards related to personnel
qualifications and training or those related to agricultural water) would reduce average
costs for small businesses. Under this alternative, the costs for all small businesses would
be reduced from $175 million to $94 million, annualized.

Page 119

IV. References
Reference List
(1) Food Safety Modernization Act Sec. 419 (f) Exemption for Direct Farm
Marketing. 1-4-2011.
Ref Type: Unenacted Bill/Resolution
(2) Draft Environmental Impact Statement (EIS) for the Proposed Rule: Standards
for Growing, Harvesting, Packing, and Holding of Produce for Human
Consumption. 1-9-2015.
Ref Type: Unenacted Bill/Resolution
(3) FDA. Food Safety Modernization Act: Current Good Manufacturing Practice
and Hazard Analysis and Risk-Based Preventive Controls for Food For Animals
(Final Rule). 2015.
Ref Type: Report
(4) The White House Office of Management and Budget. Circular A-4. 2003.
Ref Type: Report
(5) Executive Order 12866. 1993.
Ref Type: Report
(6) FDA. Analysis of Economic Impacts - Standards for the Growing, Harvesting,
Packing and Holding of Produce for Human Consumption (PRIA).
http://www.fda.gov/downloads/Food/GuidanceRegulation/FSMA/UCM334116.
pdf . 2013.
Ref Type: Online Source
(7) FDA, ERG. Evaluation of Recordkeeping Costs for Food Manufacturers. 2-12007.
Ref Type: Report
(8) USDA. Farm Household Income (Historical). 11-25-2014. 4-21-2015.
Ref Type: Report
(9) USDA. FARM FAMILY INCOME. 2004. 4-21-2015.
Ref Type: Report
(10) USDA. 2007 Census of Agriculture Farms and Ranch Irrigation Survey, Vol 3,
Special Studies Part 1. 2008.
Ref Type: Online Source
(11) U.S.Geological Survey. Irrigation Water Use. 2014.
Ref Type: Online Source

Page 120

(12) USDA. 2012 Census Publications. 2015.
Ref Type: Online Source
(13) USDA. National Agricultural Statistics Service: Contact Us. 2013.
Ref Type: Online Source
(14) United States.Office of Management and Budget. Guidelines and discount rates
for benefit-cost analysis of federal programs. 2002 Available from:
URL:http://purl.access.gpo.gov/GPO/LPS46031
(15) USDA, NASS. 2012 Census of Agriculture.
http://www.agcensus.usda.gov/Publications/2012/Getting_Started/Guide_to_Ce
nsus_Products/ . 2012.
Ref Type: Online Source
(16) Lisa A.Robinson, James K.Hammitt. Valuing Reductions in Fatal Illness Risks:
Implications of Recent Research. Health Econ 2015.
(17) United States.Bureau of Labor Statistics. Occupational employment statistics
occupational employment and wage estimates. 2000 Available from:
URL:http://www.bls.gov/oes/oes%5Fdata.htm
(18) United States.National Agricultural Statistics Service.Washington Field Office.
Washington office of USDA's NASS Washington statistics. 2004 Available
from: URL:http://www.nass.usda.gov/Statistics_by_State/Washington/index.asp
(19) United States.Food and Drug Administration. Operational and Administrative
System for Import Support (OASIS) database. 2009.
Ref Type: Report
(20) United States.National Agricultural Statistics Service., United
States.Agricultural Statistics Board. Fruit and vegetable agricultural practices 1999. 2001 Available from: URL:http://purl.access.gpo.gov/GPO/LPS89119
(21) Cohen N, Hollingsworth CS, Olson RB, Laus MJ, Coli WM. Farm Food Safety
Practices: A Survey of New England Growers. FOOD PROTECTION
TRENDS 2005; 25(5):363-370.
(22) Hardesty SD, Kusunose Y. Growers' Compliance Costs for the Leafy Greens
Marketing Agreement and Other Food Safety Programs. 9-1-2009.
Ref Type: Unpublished Work
(23) Washington (State). USDA good agricultural and good handling practices : an
audit verification program for the fresh fruit and vegetable industry. USDA
good agricultural and good handling practices : an audit verification program for
the fresh fruit and vegetable industry . 1999.
Ref Type: Journal (Full)

Page 121

(24) Florida Administrative Regulation (Florida Rule 5G-6.011) Tomato Inspection.
https://www.flrules.org/gateway/ruleNo.asp?id=5G-6.011 . 2007.
Ref Type: Online Source
(25) California Leafy Greens Marketing Agreement.
http://www.caleafygreens.ca.gov/sites/default/files/LGMA%20Accepted%20Fo
od%20Safety%20Practices%2008.04.2010_0.pdf . 2012.
Ref Type: Online Source
(26) Arizona Leafy Greens Marketing Agreement.
http://www.azlgma.gov/userfiles/Arizona%20GAPS%20%20metrics%2008_01_2010(2).pdf . 2012.
Ref Type: Online Source
(27) U.S.Department of Labor. National Agricultural Workers Survey.
http://www.doleta.gov/agworker/naws.cfm . 2006.
Ref Type: Online Source
(28) USDA National Commission on Small Farms. A time to act : a report of the
USDA National Commission on Small Farms. [Washington, D.C.]: The
Commission; 1998.
(29) Gubernot D. Updated Version of the Outbreaks Database. 2-20-2015.
Ref Type: Personal Communication
(30) CDC. Foodborne Outbreak Online Database (FOOD). 4-16-2015. 4-16-2015.
Ref Type: Online Source
(31) Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson M-A, Roy SL et al.
Foodborne illness acquired in the United States-Major pathogens. Emerg Infect
Dis Emerging Infectious Diseases 2011; 17(1):7-15.
(32) Minor T, Lasher A, Klontz K, Brown B, Nardinelli C, Zorn D. The Per Case
and Total Annual Costs of Foodborne Illness in the United States. Risk Analysis
2014;n/a.
(33) FDA. Draft Qualitative Assessment of Risk to Public Health from On-Farm
Contamination of Produce. 2012.
Ref Type: Report
(34) ERG. The Effectiveness of Harvest and Post-Harvest Measures for Reducing E.
Coli on Leafy Green Production—FINAL. 3-20-2009.
Ref Type: Report
(35) California Tomato Farmers Food Safety Audit Database.
http://www.californiatomatofarmers.com/member/login.php . 11-2-2011.
Ref Type: Online Source

Page 122

(36) ERG. Cost Effectiveness of Practices Intended to Prevent Tomato Related
Foodborne Illness—FINAL. 3-18-2009.
Ref Type: Report
(37) FDA. Technical Appendix: Estimation of Contamination Risk Mitigated Based
on External Expert Elicitation Studies on Leafy Greens and Tomatoes. 2012.
Ref Type: Report
(38) Hongliu Ding, Tong-Jen Fu. Assessing the public health impact and
effectiveness of risk interventions associated with Salmonella contamination in
sprouts. 2014.
Ref Type: Report
(39) Rebecca Montville, Donald W.Schaffner. Analysis of published sprout seed
sanitization studies shows treatments are highly variable. ournal of Food
Protection 2004.
(40) USDA. 2012 Census of Agriculture Farm and Ranch Irrigation Survey, Vol 3,
Special Studies Part 1. 2013.
Ref Type: Online Source
(41) United States.Environmental Protection Agency.Office of Water. U. S.
Environmental Protection Agency, Office of Water. 1998 Available from:
URL:http://www.epa.gov/water/
(42) Leslie Snowden-Swan, John Piatt, Ann Lesperance. Disinfection Technologies
for potable Water and Wastewater Treatment: Alternatives to Chlorine and Gas.
1998.
Ref Type: Report
(43) USDA. Glyphosphate-Tolerant Alfalfa Events J101 and J163: Request for
Nonregulated Status. 2009.
Ref Type: Report
(44) The Packer's Produce Universe. Sprouts. 2013.
Ref Type: Online Source
(45) U.S.Small Business Administration. Table of Small Business Size Standards
Matched to North American Industry Classification System Codes. 2014.
Ref Type: Online Source

Page 123