FDA has examined the impacts of the final rule under Executive Order 12866 and the Regulatory Flexibility Act. Executive Order 12866 directs 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 effects; distributive impacts; and equity). The Regulatory Flexibility Act (Pub. L. 96-354) requires analyzing options for regulatory relief for small businesses. The Unfunded Mandates Reform Act requires (in Section 202) that agencies prepare an assessment of anticipated costs and benefits before proposing any rule that may result in an annual expenditures by State, local and tribal governments, in the aggregate, or by the private sector, of $100,000,000, (adjusted annually for inflation). That Act also requires (in Section 205) that the agency identify and consider a reasonable number of regulatory alternatives and, from these alternatives, select the least costly, most cost-effective, or least burdensome alternative that achieves the objective of the rule. Even though FDA finds that the costs of this final rule may be below $100 million a year, estimating these costs is a difficult task involving uncertainties. This analysis, together with the preamble published in the Federal Register and supporting analysis and materials, constitutes a final RIA. Therefore, FDA has treated the final rule as an economically significant regulatory action under Executive Order 12866. Consequently, the agency has completed this full Regulatory Impact Analysis which demonstrates that this rule is consistent with the principles set in the Executive Order and in these two statutes. In addition, this document has been reviewed by the Office of Management and Budget as an economically significant regulatory action under Executive Order 12866. FDA has concluded that the net benefits of this rule (benefits minus costs) are largest for the regulatory option selected as specified by Executive Order 12866 FDA has also concluded that, pursuant to the Unfunded Mandates Act, the regulatory option selected is the least burdensome option to accomplish the goal of controlling all physical, chemical and microbiological hazards reasonably likely to be present in seafood. As a part of the preamble to the proposed regulation, FDA published a summary of the Preliminary Regulatory Impact Analysis (PRIA) and placed on file with FDA's Docket Managements Branch the complete PRIA.
FDA has fully reviewed the information on which the PRIA was based,
the comments on the PRIA, and other available information on the
costs and benefits of HACCP for the seafood industry. Based on
this review, FDA has arrived at two estimates of the costs in this
final rule as well as upper and lower estimates of benefits. The
estimates of costs and benefits are summarized in table 1.
Table 1.
SUMMARY OF TOTAL COSTS AND BENEFITS
(in millions of dollars)
YEAR
COSTS FROM
FDA MODELS
COSTS
ADJUSTED
FROM NMFS
MODEL
BENEFITS
LOWER
BENEFIT
UPPER
1
$69
$162
$73
$108
2
$42
$91
$73
$108
3
$41
$83
$85
$156
4
$38
$80
$87
$158
TOTAL*
$677
$1,482
$1,435
$2,561
* The total defines the total discounted costs and benefits beyond
the fourth year and discounted at 6%.
Preliminary Regulatory Impact Analysis (PRIA)
As a component of the proposal, FDA published a summary of a PRIA
with preliminary estimates of costs and benefits. The PRIA
included two sources of cost data. The first source included
reported costs of implementing HACCP by companies with a total of
64 processing plants. The second source was a NMFS study that was
based on a report prepared for the National Marine Fisheries
Service (NMFS). The total domestic costs that were estimated in
the PRIA were those that were derived from the NMFS study and were
estimated to be $139 million, or about $24,000 per plant in the
first year. The costs reported by industry in the PRIA were in the
$1,000 to $5,000 range annually for the majority of the firms.
Benefits that were quantified were denoted as safety benefits from
reducing the incidence of foodborne disease and health benefits
resulting from increased consumption of seafood. Safety benefits
were reported to be between $15 and $75 million per year whereas
potential health benefits, deriving from an increase in consumer
confidence in the safety of seafood, were between $3 and $14
billion.
The benefits of this final regulation include a reduction of foodborne disease associated with seafood consumption, a cost saving resulting from a more efficient HACCP certification program for exported seafood, a reduction in federal and state enforcement actions and potential health benefits deriving from improved consumer confidence in the safety of seafood. FDA, in consultation with CDC, has estimated that there are approximately 114,000 cases of foodborne illness each year associated with seafood. This represents an increase in the estimate provided in the PRIA of 33,000 cases (upperbound 80,000 cases). Most of the increase results from new information provided to FDA increasing the numbers for Hepatitis A, Norwalk virus (12,400 to 100,000) and Vibrio vulnificus. HACCP is expected to reduce between 20,000 and 59,000 of all cases each year by the third year of its operation resulting in benefits ranging between $45 to $116 million. However, most commenters who responded to the PRIA on this subject felt that reducing illnesses associated with seafood could not be affected by mandatory HACCP for processors. Comments on this subject noted that shellfish was already controlled by the Interstate Shellfish Sanitation Commission and that most problems were associated with harvesting illegally from closed waters. As enforcement of shellfish harvesting in these waters is a state function, commenters felt that federal HACCP for processors would not affect illnesses from this source. Other commenters noted that, for all kinds of seafood, the majority of problems arise from recreational harvesting or abuse by consumers or restaurants, all sources of problems outside of the scope of this rule. FDA has explained that, in fact, harvesting from illegal waters and recreational harvest which finds its way into commercial channels will be affected by the source control and recordkeeping provisions of the rule. In addition, the FDA scientific panel (a group of internal experts composed of epidemiologists, microbiologists and a seafood industry expert) took these factors into account in it's initial estimate of reduced cases. In addition, the Model Food Code, while not part of this regulation, addresses mishandling by restaurants. One particular problem that many commenters addressed was the problem of Vibrio vulnificus associated with Gulf of Mexico oysters. FDA believes that better processing controls such as icing and individual quick freezing (IQF) will reduce the number of illnesses by up to half by the third year of operation. Another benefit of this rule includes $20 million per year due to a less expensive FDA federally overseen HACCP program than the NMFS HACCP program. Because exporters must have a federally overseen HACCP program to continue to export to EU, Japan or Australia, this represents a savings to society. These benefits were qualitatively mentioned in the PRIA but not quantified. FDA fully expects that, as HACCP is implemented in the seafood industry, the need for Federal enforcement activity outside of verification inspections and general surveillance will be reduced because of better compliance by the industry. This reduction could be perhaps by as much as half of the current number of enforcement actions. Although only qualitatively mentioned in the PRIA, this benefit has been estimated to be $22 million due to fewer recalls, detentions, seizures and fewer injunctions against processors of non-molluscan seafood and seafood products such as finfish and crabmeat. States also will have fewer enforcement actions although FDA does not have sufficient data with which to quantify this benefit. In the proposed rule, FDA estimated large benefits from improvements in consumer confidence in seafood, between $3 and $14 billion. These benefits would be realized as consumers who were previously concerned about the safety of seafood became confident via the HACCP program that seafood was safe. Greater confidence in seafood would cause consumers to increase their consumption of lower fat/lower cholesterol seafood products and decrease consumption of higher fat meat and poultry. The overall decrease in fat consumed would cause a decrease in cases of coronary heart disease and cancer. Commenters pointed out some of the weaknesses in this argument including the possibility that increased prices that will result from this rule may more than offset the increase in demand for seafood and that seafood as prepared may not be much lower in fat, if at all, than meat and poultry as prepared. FDA continues to believe that this rule will promote increased confidence in seafood products although FDA cannot estimate any increase in seafood prices. Because of the many variables associated with the link between increased consumer confidence leading to increased consumption of seafood and the final link to coronary heart disease and cancer, FDA is not quantifying this benefit in this final RIA. Finally, FDA reports other benefits not quantified such as improved operational efficiencies and employee morale and decreased spoilage of finfish due to extended shelf-life. For example, Processors have confirmed to FDA the possibility that there may be "re-engineering" types o f benefits associated with these regulations.
The final rule covers both foreign and domestic seafood
manufacturers, repackers, warehouses, and U.S. importers.
Establishments that might generally be described as common
carriers, grocery stores, supermarkets, food service
establishments, or other types of retail establishments are not
covered by this final rule. Although distributors (unless they are
warehouses) and harvesting vessels are also not covered directly by
the final rule, they will be affected in those cases where
processors adopt HACCP controls that ensure that proper harvesting
and handling practices are carried out.
The range of costs for an average plant are reported in the chart
below.
Table 2
Average Per Plant Cost Estimates
Year
Adjusted NMFS
Model
FDA Model
Plants
First Year
$23,000
$6,400
Subsequent years
$13,000
$4,800
As table 2 indicates, there is still uncertainty surrounding the
cost estimates. There is no single source of data that FDA has
found to be entirely satisfactory for developing an estimate of
costs for these regulations. Because there is not yet an adequate
database of actual costs from firms that have implemented HACCP,
and no database of costs from firms that have fully implemented
these final regulations, many of the estimates for this program
were modeled. FDA has found that different models produce
different cost estimates. Consequently, this RIA contains a range
of costs as derived from two different models.
Firstly, in estimating the cost of this rule, FDA has relied
heavily on estimates and information contained in reports prepared
under a grant from the National Marine Fisheries Service (NMFS) of
the National Oceanic and Atmospheric Administration. The approach
used to furnish estimates of the costs and economic impacts of the
Model Seafood Surveillance Project (MSSP) HACCP-based program on
manufacturers is based on data and cost estimates collected at the
plant level. NMFS randomly selected, for field visits, a sample of
130 plants from the seafood manufacturers listed in the NMFS
database. These plants were not operating under HACCP. The costs
estimated were based on private contractor estimates of the costs
of the HACCP program designed by the National Marine Fisheries
Program. While there are many similarities between the MSSP and
the requirements of this regulation, there are some significant
differences, including coverage of nonsafety hazards such as
decomposition and fraud (which FDA did not include in its own
estimate taken from the MSSP study). Thus, this source of
information consisted of data, in terms of the distribution of
processing types and information on current conditions in the
industry (plant, equipment and practices relative to HACCP), and
engineered estimates on needed plant and equipment changes as well
as HACCP costs expenditures.
In areas where FDA had better data than that used in the reports
for NMFS, it used this information including data from its field
surveys on current practices or conditions in the industry in
general. For example, the NMFS estimated that there were 2,800
domestic seafood processing plants in 1989, whereas FDA counted
4,846 plants in 1992 for both the PRIA and this RIA. Another
source of uncertainty in the analysis is that the sample taken from
a universe of 2,800 plants may not be representative of a sample
that could be drawn from 4,486 plants. Where gaps in the NMFS
estimates exist that could not be filled in by information from FDA
field surveys, a number of assumptions detailed throughout this
analysis have been made for the purposes of this economic
assessment, this was also done in the PRIA. Average plant costs
estimated from this source are estimated to be $23,000 in the first
year and $13,000 in subsequent years.
In addition to the cost estimate based on the MSSP modeling, FDA
has provided an alternative model that gives the basis for a second
cost estimate for these regulations. In this model, FDA's seafood
experts estimated the cost of compliance with these regulations by
creating two hypothetical small seafood processors that the agency
believes to be representative of a significant portion of the
seafood industry.
One of the model processors is assumed to be in substantial
compliance with existing GMP requirements. Therefore, the costs
experienced by that firm are attributable exclusively to the
establishment and maintenance of a HACCP system. The other
processor has GMP deficiencies that the agency believes are
commonly displayed by seafood processors, although no major plant
or equipment changes such as those reported in the MSSP report
($40-45 million) are assumed to be required. This model processor
is identical to the other processor except for the GMP deviations.
The costs calculated for this second firm represent costs
associated with the establishment and maintenance of HACCP as well
as costs associated with the correction and monitoring of
sanitation conditions.
Reports received by FDA on the actual costs of implementing HACCP
experienced by a number of seafood processors generally corroborate
the results of this modeling across the seafood industry. The
majority of firms that could estimate their start up costs
indicated costs in the $1,000 to $5,000 range. The rest appear to
be roughly equally divided between costs that are below and above
this range. Whereas some firms reported costs less than $1,000
others reported costs in excess of $20,000. In subsequent years,
some firms reported costs that exceed first year costs.
While the models were limited to certain types of small operations,
the firms for which FDA has received cost information that
represent a cross section of processing operation types, including
canned, fresh, frozen, smoked/salted, molluscan shellfish, and
cooked, ready-to-eat products as well as warehouses and repacking
operations. While this information is neither sufficiently
detailed nor complete enough to calculate costs of the proposed
regulations that will be incurred by the entire seafood industry,
it indicates that the FDA modeling is reasonable and that the cost
estimate for the regulations as extrapolated from the FDA modeling
should be included in this RIA.
In order to estimate an average plant cost from these theoretical
models FDA has assumed, based on its inspection experience, that
20% of the firms are similar to the model plant that needs to make
some GMP improvements and that 80% of the firms are similar to the
plant that is in compliance with existing GMPs. The agency has
also assumed that the cost of compliance of large firms is equal to
that of small firms. Average plant costs estimated in this manner
are $6,400 in the first year and $4,800 in subsequent years.
FDA used the two hypothetical models of small plants to represent
all small processors. The total cost of this regulation using this
hypothetical method of cost modeling is $75 million in the first
year and $28 million in the fourth year and beyond. However,
implicit in these calculations is the assumption that the small
plant models of two plants that (1) cut and package tuna that is
received frozen (no hazards) and (2) distribute frozen orange
roughy fillets (a product for which there are hazards) can
represent the universe of different types of processors such as raw
molluscan shellfish, cooked shrimp, surimi, smoked fish or breaded
fish processors. The agency recognizes the difficulty with these
projections. Nonetheless, the modeling demonstrates that
processors may have costs that are significantly below the averages
estimated by means of the MSSP report.
Virtually all commenters who addressed the PRIA's cost estimates,
including the MSSP estimates, stated their belief that the cost
estimates in the PRIA were underestimated. However, based on
agency experience, FDA assumes that most commenters who submit
comments on costs are on the high end of the distribution.
Based on the FDA model, total domestic costs are estimated to be
$64 million in the first year, $39 million in the second year, $41
million in the third year and $38 million in the fourth and
following years. Total foreign processor costs passed on to U.S.
consumers are estimated to be $5 million in the first year and $3
million in the second year and zero beyond the second year.
Based on the NMFS model taken from the sample plants, total
domestic costs estimated for harvesters, manufacturers, importers,
repackers, and warehouses will be $144 million in first year, $71
million in second year, $83 million in third year and $80 million
in the fourth and following years. Total foreign processor costs
passed on to U.S. consumers are estimated to be $18 million in
first year and $10 million in second year. The agency expects that
there will be no costs to foreign processors attributable to this
rule after the second year.
There are several major differences between the preliminary
economic analysis and this final analysis of this regulation.
Firstly, cost estimates have been made for several items for which
no cost estimates were previously made. They include cost
estimates of verification, analytical testing and controls for V.
vulnificus. Secondly, the agency has reduced its estimates of
costs for several items from the estimates previously made. They
include costs to importers, costs of writing HACCP plans, costs to
processors that export and foreign plant costs.
FDA has also completed a final Regulatory Flexibility Analysis that
examines the impacts of the rule on small businesses. Although
much of the costs of compliance with this rule are fixed costs, and
therefore affect small businesses disproportionately, FDA believes
that various factors will help small firms comply without undue
economic hardship. These factors include providing two years to
comply from the date of publication of this rule (applies to all
firms, regardless of size), provision of a hazard guide which can
be used to create a HACCP plan, and other organizational help such
as the Seafood Alliance. First year average cost estimates for
small firms range from as low as $5,600 to $12,400 in the FDA model
plant estimates (depending upon the extent of compliance with
GMP's) to $23,000 estimated in the MSSP model. However, some
plants with no GMP deficiencies and no hazards may have minimal
costs whereas others with severe plant and equipment problems may
have much higher costs. FDA estimates that 140 small firms will go
out of business for every per 1% increase in the price of seafood
although FDA is unable to estimate the actual price increase that
will result from this regulation. If demand increases for seafood,
this will mitigate some of the price effects.
In estimating the cost of this rule, FDA has relied heavily on estimates and information contained in reports (the Kearney reports) prepared for the National Fisheries Education and Research Foundation, Inc. (NFERF) by A.T. Kearney, Inc. (Contract No. NA88AA-H-SK006). NFERF, which is associated with the National Fisheries Institute, commissioned the reports after receiving a grant from the National Marine Fisheries Service (NMFS) of the National Oceanic and Atmospheric Administration. These reports estimated the incremental costs beyond those now being incurred by seafood processors for implementing the specific HACCP-based system developed for the National Marine Fisheries Service's Model Seafood Surveillance Project (MSSP). The approach taken by A.T. Kearney, Inc. to furnish estimates of the costs and economic impacts of the MSSP HACCP-based program on manufacturers is based on estimates and data collected at the plant level. To perform in-plant, on-site analysis, NMFS randomly selected for field visits a sample of 130 plants from the seafood manufacturers listed in the NMFS database. These plants were not operating under HACCP. Information on sanitation, processing and economic conditions was collected from each individual plant. Controls now in place and records currently being kept were taken into account. Estimates were then made of the costs to these firms to install and implement MSSP HACCP systems. The numbers of small and large firms were estimated using the percentages of firms in various industry segments with less than $1,000,000 in annual gross revenue ($2,000,000 in annual gross revenue for plants manufacturing breaded, cooked or raw shrimp) reported in the Kearney reports. The primary benefits estimated in the PRIA were the safety benefits, the reduction in foodborne disease caused by seafood consumption. The estimates derived for these benefits used, among other sources, a report commissioned by FDA from Research Triangle Institute entitled "Estimating the Value of Consumer's Loss from Foods Violating the FD&C Act." This model had five components: 1) identify adverse health effect(s) from a violation, 2) describe impacts on health, 3) estimate changes in health status, 4) estimate utility losses, and 5) estimate dollar values of utility losses. FDA has used the same methodology in this final RIA. The first component of the model is the annual incidence of foodborne disease that results from seafood consumption. Next, each case of illness was "valued" by examining consumers' loss in terms of medical expenses, lost productivity and pain and suffering. Where loss of life was an outcome of a foodborne disease, $5 million was used to value that outcome. Morbidity losses and medical expenses were directly estimated. Lost productivity and pain and suffering were indirectly estimated as time endured (days, weeks) being sick in relative states (mild, moderate, severe or death) of discomfort and inability to work. FDA then used an internal panel of experts to determine, hazard by hazard, the number of cases of illness expected to be avoided by the proposed regulation each year. In the final rule, FDA has consulted with CDC on these estimates. Multiplying each avoided case by the cost of the illness gives the total safety benefit to be derived by this rule. For example, FDA has estimated that between 4,000 and 6,000 cases of scombrotoxin poisonings will be avoided as vessels and processors take better care to ice certain kinds of finfish such as tuna and mackerel to prevent histamine formation. Scombrotoxin poisoning is generally a mild illness that lasts only a day or so. Costs of lost productivity and pain and suffering are valued at approximately $340 per case with total benefits estimated to be between $1,354,377 and $2,031,566.
FDA received approximately 230 comments on issues involving the
PRIA which are summarized and analyzed below.
MARKET FAILURE
No comments were received that directly addressed the issue of
whether or not the food safety issues addressed in this proposal
are adequately addressed by the seafood market. Some commenters
did address this indirectly and those comments are discussed in the
preamble to the final rule. FDA continues to believe that seafood
safety is a problem that should be addressed by Federal regulation.
ESTIMATION OF COSTS AND BENEFITS
Both costs and benefits of HACCP are difficult to predict, and this
difficulty is reflected in the comments. Costs and benefits of
HACCP will depend on several factors. First, the first edition of
the FDA Fish and Fishery Products Hazards and Controls Guide (the
Guide) that the agency will publish prior to the effective date of
but after publication of this rule will help firms to do a hazard
analysis and create their HACCP plan. Second, enforcement of HACCP
by both Federal and State regulators presents a different kind of
challenge from more prescriptive types of regulations and thus
makes cost estimation more difficult. Third, and most importantly,
because processors are responsible for operating their own HACCP
systems, the numerous decisions that domestic and foreign producers
make will ultimately determine the costs and benefits of HACCP.
These decisions reflect changing information and technologies.
Thus, unlike traditional regulation, where there is either an
established performance level or required technology, HACCP is a
flexible, interactive system that will evolve as conditions
warrant. This system presents greater challenges for a benefit-cost analysis, w
hich must predict the actions of affected entities.
With this HACCP program in place, a retrospective look will help
with future efforts at mandating HACCP for other sectors of the
food industry and for better estimating future costs and benefits
of HACCP programs.
Regulatory Options
The agency raised and received comment on a number of regulatory
options in the PRIA. The most significant two options raised were
regulating only high risk products or the most serious hazards and
providing regulatory relief for small businesses. The first option
is inconsistent with the objective of this regulation, to control
all physical, chemical or microbiological hazards reasonably likely
to be present in seafood products. Although FDA has not granted
relief only for small business, FDA has extended the compliance
date for all firms from one to two years. COSTS
As explained more fully below, FDA used modeling techniques and the
experience reported by seafood firms to produce estimates that are
in general agreement and that are approximately one fourth of those
in the PRIA, ranging from a total of $677 million to $1.4 billion
(discounted). Moreover, these estimates are also consistent with
the estimates made for NMFS of implementing HACCP by small
processors. Even though FDA found that the costs of this final
rule may be below $100 million, estimating these costs is a
difficult task involving many uncertainties. For example, the
agency recognizes that the final rule may affect in a material way
seafood processors such that FDA has treated this final rule as a
significant regulatory action under Executive Order 12866.
Consequently, the agency has completed a full Regulatory Impact
Analysis.
While there is some experience with HACCP in the seafood industry,
that experience is limited, and one may be uncertain about being
able to generalize from it to all processors because the first
firms to initiate HACCP are likely to be those who derive the
greatest economic advantage from implementing it. In the PRIA, FDA
was reluctant to rely only on the results of the limited experience
with HACCP in the seafood industry. Thus, FDA balanced the
reports of some seafood firms, which showed that the costs of HACCP
were low, with a study of the costs of HACCP that had been done
under contract with NMFS. This study showed significantly higher
costs (as reflected in the range of cost estimates summarized
above) but had several limitations in terms of its applicability to
this cost analysis. For example, none of the plants that were the
subjects of the study had actually implemented HACCP, and the
system whose costs were studied was significantly broader in scope
than the system embodied in Part 123 (this final rule). FDA had to
extract the relevant requirements in order to produce cost
estimates. Moreover, some of the controls in the NMFS study are
more expensive than controls that processors could select under the
FDA program. Despite these facts, the cost estimates in the PRIA
were based on the results of the NMFS study because FDA considered
it to represent the best evidence available.
FDA used three separate estimates of the costs in the PRIA to
ensure that the uncertainties in the estimates of the costs that
would be imposed by adoption of this regulation were accounted for.
The first is the cost comments that were received on the PRIA. FDA
took advantage of them whenever they were supported by data. The
second is modeling by FDA experts based on their experience with
hazards and controls in the seafood industry, working with aquatic
species and the public health problems that they present,
inspecting and studying both seafood plants and low acid canned
food (LACF) plants (which have operated under HACCP principles for
almost two decades), and participating in the FDA-NMFS seafood
pilot. The results of this modeling are detailed below. The third
source is information that FDA received about firms that have
actually implemented HACCP.
Adjusted Cost Estimates from the NMFS Report
Table 3 summarizes the specific cost estimates that FDA arrived at
using data from the NMFS model with cost refinements received from
commenters and FDA seafood industry experts. The refinements in
the estimates from the PRIA are fully explained below. These
adjusted cost estimates from the NMFS model result in per plant
costs for domestic manufacturers of $23,000 in the first year and
$13,000 in subsequent years. Total costs for compliance with these
regulations using the adjusted NMFS data are shown in table 3.
Table 3
Disaggregated Costs from Adjusted NMFS Model
First Year Costs
Domestic manufacturers and importers: $112 million
Major plant repair and renovation: $ 13 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
Harvesters for rejected raw product: $ 1 million
Shellfish vessels: $ 3 million
Foreign processors: $ 18 million
First Year TOTAL: $162 million
Second Year Costs
Domestic manufacturers: $ 65 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
Shellfish vessels: $ 1 million
Foreign processors: $ 10 million
Second Year TOTAL: $ 91 million
Third Year Costs
Domestic manufacturers: $ 65 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
IQF Shellfish plants: $ 3 million
Third Year TOTAL: $ 83 million
Fourth Year (and subsequent years) Costs
Domestic manufacturers: $ 65 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
Fourth Year TOTAL: $ 80 million
Total Discounted Costs
Beyond the fourth year and discounted at 6%, the costs are $1,482
million. The agency requested that commenters supply data on a number of
cost categories for which it did not have sufficient data to make
estimates, as well as on all other aspects of the analysis. It is
important to remember that these comments and FDA's response to
them generally involve either costs that the agency estimated from
the NMFS model or that the agency was not able to estimate in the
PRIA. For this RIA, the agency has revised the cost estimate from
the NMFS model partly on the basis of these comments and partly on
the basis of differences between the proposal and the final
regulations. As mentioned above, this estimate is then compared to
a second cost estimate, based on the FDA modeling as supported by
reported industry costs. A discussion of the comments received on
costs in the PRIA follows.
Cost of Creating a HACCP Plan
A number of commenters stated that FDA underestimated the amount of
time (24 and 72 person hours, respectively for simple and complex
plans) required to develop a HACCP plan. In the PRIA all plants
producing cooked crustaceans; smoked, cured and dried fish; stuffed
specialty products; and surimi analog products were assumed to have
complex plans. Another commenter said that FDA's estimates were
"about right." One commenter from the National Marine Fisheries
Service stated that, based on their experience with the NMFS HACCP-based Inspect
ion Program, "it takes a significant amount of time to
develop a workable, effective HACCP plan." One large seafood
processor that produces a variety of products stated that FDA's
estimates should be doubled to 48 person hours for simple plans and
144 hours for complex plans "if the [plan developer| has been
trained in HACCP and has a working knowledge of HACCP before
creating a program."
The NMFS commented that, based on its experience, the development
of a HACCP plan takes considerable time and effort. FDA has stated
in this final rule that HACCP plans are not required when the
conclusion of a hazard analysis is that there is no reasonable
likelihood of a safety hazard to be controlled. However, FDA
believes that NMFS' comment is due primarily to the fact that the
NMFS' HACCP plans cover quality and economic adulteration issues as
well as safety measures required by this rule. FDA believes that
the seafood industry is dominated by firms that would have
relatively simple HACCP plans if safety were the only issue. This
is the primary reason why these costs were reduced from the NMFS
estimate in FDA's analysis of the costs of HACCP. Moreover, the
agency is convinced that two factors will have a bearing on the
time necessary to develop a HACCP plan. First, FDA's Guide will
enable the rapid development of plans in many cases. Second, the
training in HACCP being developed by the Seafood HACCP Alliance
(the Alliance) is expected to include the drafting of actual HACCP
plans. Processors with simple plans should have a major portion of
their plans already drafted by the completion of the course.
However, FDA is convinced by the majority of commenters which
stated that FDA had underestimated the time to prepare a HACCP
plan. Thus, FDA is revising its estimates to 24 person hours for
simple plans and 144 hours for complex plans. FDA believes that
the vast majority of plants will be developing simple HACCP plans.
FDA anticipates that complex HACCP plans will be confined to 25% of
the plants in the following categories: cooked crustaceans;
smoked, cured and dried fish; stuffed specialty products; and
surimi analog products. Plans for the other 75% of plants in those
categories, as well as for all plants in other categories, should
be relatively simple.
One comment from an Extension/Sea Grant Seafood Processing
Specialist addressed FDA's assertions that Sea Grant seafood
specialists would be available to act as inexpensive consultants
and processing authorities helping small processors develop HACCP
plans and assisting them in verification activities. The commenter
stated that "Local area Sea Grant Extension offices do not usually
house seafood specialists and therefore will not be a good source
of process authorities unless new staff are hired and unless those
new staff are attached to some research group which has the ability
and budget for testing and research. Inland processors have very
little access to Sea Grant Extension offices."
However, the agency remains convinced that Sea Grant will provide
such services to processors as it has committed itself to do
through its participation in the Alliance. Sea Grant seafood
specialists are, in fact, located in many seafood processing areas.
Nonetheless, the cost to society is approximately the same whether
it is paid by processors or by taxpayers for the support of the Sea
Grant program. The agency erred by not including an estimate of
the cost of assistance from Sea Grant in the PRIA. If the cost to
Sea Grant of providing assistance to processors is $500 per plant
and one half of the small plants (1923 plants) per year receive
assistance from Sea Grant Seafood Specialists, then the annual
social cost of assistance from this source will be approximately $1
million. While this shifts a portion of the cost from processors
and importers to Sea Grant, there is no net effect on the overall
estimated costs of this regulation.
Plants Already Operating under HACCP
For the purpose of estimating costs in the PRIA, the agency assumed
that none of the 4,846 plants were operating under HACCP. FDA is
revising that assumption based on the fact that 439 plants now
shipping to the EU will need to have HACCP systems in place by
January 1, 1996. The cost of establishing such systems would be
incurred by these processors irrespective of the issuance of these
regulations. Additionally, as of July 1995, 80 processors were
operating under the NMFS HACCP program and 90 more were expected to
be by the end of 1995 for a total of 170. Allowing for an assumed
50% overlap with the processors that ship to the EU, this adds an
additional 85 plants to the 439 plants now shipping to the EU that
will implement HACCP by the end of the year regardless of the
implementation these regulations. This means that a total of 524
plants (439 + 85) may be expected to have HACCP in place by the end
of 1995.
For the purposes of this analysis, FDA has assumed that all of
these plants fall into the category that the agency has classified
as large, although, in fact, some exporters could actually be
classified as small. Based on this assumption, the number of large
plants in each category of seafood manufacturer has been reduced by
a factor of 524/1000 (the fraction of large manufacturers that will
have implemented HACCP by January 1, 1996). The agency recognizes
that even these plants may need to make some minor changes based on
the requirements of this rule. These changes are estimated to be
$1,000 per plant per year, or the equivalent of 15 minutes of
managerial labor time at $15 per hour per day.
Cost of Corrective Actions
A number of commenters challenged FDA's assumption that the annual
per plant cost will be $1,000 for taking corrective actions when
deviations from critical limits occur. A few commenters indicated
that critical limit deviations occur daily. A few other commenters
indicated that the cost of responding to deviations from critical
limits would be approximately $10,000 in the early phase of
implementation because of destroyed or reworked product, analytical
tests, production down-time, and labor time spent investigating the
cause, severity, and consequence of the deviations. A few
commenters indicated that costs could be even higher for large
plants processing products with a high retail value in cases where
product must be destroyed. However, one commenter from one of the
largest meat and seafood processing companies in the US stated
that, after operating under HACCP for a number of years, the cost
of responding to critical limit deviations is approximately $1,000
per year.
FDA agrees with the comment that pointed out that the cost of
taking corrective actions will be correlated with the value of a
firm's output. Thus, if a relatively large firm reports annual
corrective action costs of $1,000, it is reasonable, all other
things being equal, to assume that smaller firms will have lower
annual costs for corrective actions.
In response to the comments that stated that deviations from
critical limits occur every day, the agency observes that the
monitoring and recordkeeping features of HACCP are intended, in
part, to enable processors to spot trends in processing parameters
that, if properly responded to, will prevent deviations from
occurring every day. FDA expects that deviations will become much
less frequent under a well-running HACCP system.
Nonetheless, based on the data submitted, the agency is revising
its estimate of the cost of corrective actions. The agency has
estimated that the average cost of responding to critical limit
deviations will be $2,000 per plant in the first year, and $1,000
in the following years. FDA believes that firms that experience
$10,000 in corrective action costs in a year have significant
control problems that must be remedied regardless of whether HACCP
is remedied or not.
Cost of Verification
Several commenters challenged the absence of a cost associated with
changes in the HACCP plan once the original plan had been put in
place. One comment, from a very large seafood processor, stated
that "The Agency has also underestimated the costs of verifying the
adequacy of a HACCP plan. Verification activities may include
shelf-life testing, pathogen challenge testing, and certain in-line
and finished product testing. Manufacturers often modify product
formulations, processes and production equipment. It is common
that at least one such change takes place each year. We feel that
the HACCP-related cost of verification activities is approximately
$5,000 initially, and $1,000 thereafter for each product/production
line combination on an annual basis."
The agency agrees that HACCP plans will need to be verified, and
possibly altered, every time a plant produces a new product or
substantially changes its production process. Depending upon the
circumstances, plans may also have to be verified after taking a
corrective action. Whether the plan would then require alteration
that would involve notable cost to the processor would depend on
whether the hazards and controls for those hazards had changed.
Based on FDA's own inspection experience, not all plants would be
expected to make such changes in a given year. Many, if not most,
would be expected to process the same products from the same
sources for many years.
In the final rule, the agency is requiring that the HACCP plan be
reviewed annually as part of the verification activity. As the
preamble to the final regulations points out, much of the annual
verification can be informal. The regulations are deliberately
flexible on this point in order to allow processors to develop
verification systems that are appropriate to their operations.
FDA's preliminary cost estimates were based on information in the
agency's databases that seafood plants produce, on average, two
different types of products (for example, shucked oysters and
cooked crabs). Using this data and the per product/production line
estimates supplied by the commenter, the verification cost for the
average plant with two product/production line combinations would
be estimated to be $10,000 in the first year and $2,000 thereafter.
However, based on the changes made in this final rule, the agency
now estimates that the per plant cost of verification activities is
approximately $1,000 per year. In the agency's judgement, this
amount should be sufficient for verification of the HACCP plan
covering the entire plant. The agency believes that the size and
complexity of the types of seafood produced by the commenter are an
explanation for the higher reported costs. For example, shelf-life
testing and pathogen challenge testing should not be necessary for
most products.
Verification - Cost of Analytical Testing
In the PRIA, FDA made no estimate of the cost of analytical testing
but asked for and received numerous comments on this cost. The
comments can be grouped into two sets, those relating to analytical
testing of seafood in general, and those relating to testing for
mercury in particular species.
In regard to testing seafood in general, many commenters stated
that it is not unusual for processors to receive raw products from
many different suppliers in the same year. Commenters stated that
the per sample testing cost ranges from $50 to $450. Several
commenters stated that, on average, the cost per sample is $250.
All of the commenters that remarked on this issue expressed concern
that the suggested testing frequencies described in the Guide would
place a financial burden on seafood processors significant enough
to threaten their viability. Based on firm costs calculated by
several commenters, testing costs ranged from $5,340 annually for
a small plant with an undocumented number of species and suppliers
to $22,500 annually for a small plant with three suppliers of ten
species.
As previously mentioned, finished product testing is identified in
the final rule as a verification procedure at the option of the
processor. FDA believes that many processors will choose other,
less costly verification options. Where testing proves necessary,
FDA believes that the frequency of testing will vary greatly (e.g.,
from several samples per year to verify the performance of a new
supplier to one sample per lot as is currently practiced by the
canned tuna industry for histamine analysis). Many of these
analyses will not be new costs as processors are now doing some
testing. The testing recommendations in the draft Guide are being
reviewed and cost considerations, as described in the comments,
will be carefully considered and balanced against public health
needs before making recommendations for testing in the first
edition of the Guide. Table 4 identifies the ten most commonly
consumed species of fish and their likely analytical testing needs
for: industrial and agricultural chemical contaminants (Chemical),
methyl mercury (MeHg), marine toxins (Toxin), histamine, food and
color additives used in feed (Additive), parasites, aquaculture
drugs (Drug), and microbial pathogens associated with the harvest
water (Pathogen). The table does not reflect analytical testing
that may be performed to verify the control of hazards that are
introduced during processing, such as pathogens introduced during
the processing of a cooked, ready-to-eat product. On average, FDA
expects that processors will collect and analyze four additional
samples per year at a total cost of $1,000 per plant per year.
Table 4
Likely Analytical Testing Relating to the Control of Species-Related Hazards i
n the Ten Most Commonly Consumed Types of Fish
Types
Chemical
MeHg
Toxin
Histamine
Additive
Parasite
Drug
Pathogen
tuna
No
No1
No
Yes2
No
No
No
No3
shrimp
No4
No
No
No
Yes5
No
No4
No4
cod
No
No
No
No
No
No
No
No
pollock
No
No
No
No
No
No
No
No
flatfish
No6
No
No
No
No
No
No
No
clams
Yes7
No
Yes7
No
No
No
No
Yes7
catfish
Yes8
No
No
No
No
No
Yes8
No
salmon
No9
No
No
No
No9
No10
No9
No3
crabs
No
No
No
No
No
No
No
No
scallops
No
No
Yes11
No
No
No
No
No3
1 Not likely for most tuna species. However, methyl mercury
testing may be performed as a verification function in some large
species of tuna.
2 Histamine testing may be performed as part of a system of
monitoring (as it is now in the canned tuna industry), corrective
action, or verification.
3 Not likely for most tuna, salmon, or scallop products.
However, pathogen testing may be performed as a verification
function in tuna, salmon, or scallop intended for raw consumption.
4 Not likely for wild caught shrimp. However, chemical
contaminants, drug residue and salmonella testing may be performed
as a verification function in aquacultured shrimp where there is
reason to suspect a problem.
5 Sulfite testing may be performed as a raw material monitoring
method or may be performed as a verification function in wild
caught shrimp.
6 Not likely for most flatfish. However, chemical contaminant
testing may be performed as a verification function for near-shore
species of flatfish.
7 Chemical contaminant, toxin, and/or microbiological testing
may be performed as a verification function for clams. However,
such testing is routinely performed by State Shellfish Control
Agencies as part of their growing water classification efforts, and
it is likely that most processors will rely upon this effort.
8 Chemical contaminants and drug residue testing may be
performed as a verification function in aquacultured catfish where
there is reason to suspect a problem.
9 Not likely for wild caught salmon. However, chemical
contaminants, food and color additives, and drug residue testing
may be performed as a verification function in aquacultured salmon
where there is reason to suspect a problem.
10 Not likely for most salmon products. However, parasite
testing may be performed as a verification function for salmon
intended for raw consumption.
11 Toxin testing may be performed as a verification function for
whole scallops. However, such testing is routinely performed by
State Shellfish Control Agencies, and it is likely that most
processors will rely upon this effort.
Cost of Consumer Complaint Evaluation
A number of commenters stated that setting up and operating
consumer complaint evaluation systems could range from $25,000 to
$100,000 per year. The agency estimated $1,000 for this purpose.
In the final rule the agency has changed its expectations for the
handling of consumer complaints so that consumer complaint handling
and evaluation should not be as costly as it would have been had
the provisions in the proposal been finalized. Plants will be
expected to evaluate consumer complaints for their relevance to
food safety concerns, however, plants need not have detailed
written operating procedures as to how that will be done. The
agency also notes that many processors never receive consumer
complaints, primarily because many products lack brand names.
Moreover, most processors of branded products already evaluate
consumer complaints on a regular basis. The agency believes that,
given these facts and the changes in the regulation, the average
per plant cost of the additional activities required by this rule
is $100 annually.
Cost of Monitoring and Recordkeeping and Hiring Personnel
As more fully discussed in the preamble of the final rule, many
commenters addressed the monitoring and recordkeeping requirements.
Some provided unsubstantiated estimates of the cost of these
provisions of the rule. Approximately half of the commenters,
however, agreed with FDA's estimate of approximately 650 person
hours per plant per year. In the absence of additional data, the
agency is retaining its original estimate.
Numerous commenters stated that the amount of extra labor needed
for monitoring and recordkeeping would cause processors to have to
hire another employee. Some of the commenters on this topic stated
that the agency had not estimated the cost of hiring a full-time
employee. Other commenters stated that they could not afford to
hire an additional employee and that the need to do so would cause
financial hardship. FDA acknowledges that some processors may need
to hire more workers as a result of this rule. However, the FDA
believes that, in many cases, employees already working at or near
critical control points (CCPs) can monitor those points as they
should be doing now in most cases and by only making a brief
notation of the result of that monitoring. In most cases only
simple recordkeeping is needed.
If processors choose to hire workers to work more hours than the
additional hours needed to satisfy the monitoring and recordkeeping
requirements of this rule, then it would presumably be because the
processors decide to employ workers for reasons beyond the
requirements of this regulation. Therefore the agency rejects the
comments that the cost of this regulation is more than the
additional hours estimated for monitoring and recordkeeping.
Cost of Temperature Monitoring and Recording Equipment
Many commenters stated that temperature recording devices cost
$1,000 each to purchase and install. Based on information cited in
the preamble to the proposed rule, FDA's preliminary cost analysis
assumed that 35% of the plants would need to purchase and install
two temperature recording devices costing $1,000 each. The agency
recognizes that some plants may need to install more than two
devices, while other plants may only need to install one device.
FDA believes that the preliminary cost estimates appropriately
accounted for the cost of temperature monitoring and recording
devices. Nonetheless, no plant is required to have a temperature
recording device under these regulations so long as they can
effectively monitor their CCPs through other means (e.g., high
temperature alarms for coolers). This is one example of how
processors have a range of choices available to them that makes it
difficult to assign a single overall cost for these regulations
with certainty.
Cost of Record Storage Space
One commenter requested that FDA include an estimate of the rental
cost of space needed to keep records required by this rule. In its
preliminary cost estimate the agency included costs for storage
cabinets for records in cases where the reports for NMFS indicated
that they were necessary. The agency acknowledges that large
plants may generate large volumes of records. However, FDA
encourages processors to keep simple records wherever appropriate,
and that complex and lengthy notations should be the exception, not
the rule. In any event, in the absence of data in these comments
with which to make new estimates of costs, the agency finds that it
has no basis to alter its previously estimated cost of record
storage. Furthermore, the NMFS study was based on a form of HACCP
which included quality and economic adulteration that would
generate far more records than the HACCP system being mandated in
these regulations (i.e., safety only).
In addition, it should be noted that the final regulations have
added some flexibility to records storage in remote locations by
deleting the proposed requirement that records from remote
processing facilities be returned to those facilities during the
processing season. Now records from remote sites may be
transferred to more convenient locations under certain
circumstances and may remain at those locations. Therefore, FDA
concludes that the costs of record storage are minimal.
Cost of Training
Several commenters stated that to ensure that they could operate in
compliance with the regulation, each plant would need two HACCP
trained individuals. FDA does not rule out the possibility that
some firms will need more than one HACCP trained individual. Such
firms may be among the larger and more complex seafood operations.
However, the final regulation differs from the proposal in two key
respects that should affect the number of trained individuals that
processors need. First, as discussed in detail in the preamble of
the final rule, the agency is requiring, among other things, only
that a HACCP trained individual review the monitoring records of
CCPs weekly rather than before shipment of the product to which the
records relate. This modification from the proposal should reduce
the need for a trained individual to be on site continually and
should lower costs. Second, the final regulation no longer
requires that each processor employ at least one trained
individual, only that certain functions be performed by a trained
individual. Thus, processors who do not need a HACCP trained
individual on site all the time may purchase the services of a
trained third party on an as needed basis. Shared purchasing of
third parties should also decrease the average number of trained
individuals needed per processor. Therefore, FDA estimates that,
on average, small processors will need to have only one individual
trained per plant. The agency assumes, for the purpose of this
analysis that all large plants have already employed HACCP trained
individuals.
In terms of the number of employees per plant that must be trained,
several commenters believed the agency's estimate of $900 to train
a single individual in HACCP to be too low. No additional data
were submitted by these commenters. The agency notes that the most
recent quoted prices for HACCP courses have increased, but does not
find that its original estimates need to be increased significantly
to reflect these price increases. Thus, based on the cost cutting
measures discussed in the final rule regarding the Alliance
Steering Committee curriculum, and the lack of additional data from
commenters, the agency continues to believe that training a single
individual by a cost-conscious processor will not exceed $1,000.
Several commenters stated that for HACCP to be successful, all
plant employees should receive some type of training in HACCP. The
agency agrees that HACCP works best when plant employees working at
CCPs are taught the purpose of HACCP, their particular
responsibilities, and how they fit into the HACCP system. Such
training need not be as extensive as for the individuals fulfilling
the role of the HACCP-trained individual identified in the rule.
In its preliminary cost estimate, FDA based its estimate of
training needs on the judgement of the inspectors involved in
producing the reports for NMFS. From the sample of plants in each
seafood category, FDA extrapolated to training needs for the entire
seafood industry. Nonetheless, although the cost of training
production workers in HACCP responsibilities was not separately
described in the preliminary cost analysis, those costs were
included in the above estimate.
Cost of Major Plant Repair, Renovation and Reconstruction
Several commenters criticized FDA for rejecting the estimate of the
reports for NMFS that a total of $40 to $45 million would need to
be spent for major plant repair, renovation and reconstruction.
The agency stated that it believed that $13 million was a more
accurate estimate. The commenters all described their anticipated
need for major plant repairs with the highest estimated cost being
$200,000. The NMFS studies, on the other hand, reported only two
plants that needed major repairs -- one with estimated costs of
$441,000 and the other with estimated costs of $500,000. The
highest amount provided by a commenter represents only 45% of the
lower cost of the two plants ($441,000) described in the NMFS
reports. The agency is satisfied that the two plants described in
the NMFS reports were gross anomalies, and the agency stands by its
initial judgement that the cost of major plant repair, renovation
and reconstruction is a one-time cost of $13 million.
Cost of Creating Sanitary Zones
Many commenters stated that the creation of sanitary zones in
existing plants could cost as much as $200,000. These commenters
described the cost of building new walls and purchasing separate
refrigerators for cooked and raw products. In the final rule, the
agency is not requiring the creation of sanitary zones and includes
no such costs for creating sanitary zones in its cost estimate.
Importer Costs
Many commenters stated that they believed that FDA had
underestimated the costs to importers. One commenter stated that
"Many small seafood firms import products from over 25 countries"
and that they cannot afford to provide the surveillance necessary
to ensure compliance. Some commenters estimated that annual costs
could go as high as $200,000 per importer. No data was included
with these comments to support the estimates, and FDA believes that
this estimate is unrealistically high. Nonetheless, FDA has
considered the various comments received concerning importers.
Importers are now required only to conduct verification activities
(unless they also meet the definition of a processor), and HACCP
training is not required for importers. In fact, FDA plans to
include in the forthcoming first edition of the Guide specific
materials related to importer's verification procedures.
Although FDA believes that US importers are likely to incur some
additional costs from implementation of the regulation, the
requirements to importers are much less than were originally
proposed. In the absence of an active, functioning and enforceable
MOU between the FDA and the country from which the products are
being imported, importers must have written product specifications
to ensure that the product is not adulterated under section 402(a)
or 402(c) of the Federal Food Drug and Cosmetic Act and take one or
more affirmative steps to ensure that the products being imported
were processed under HACCP and sanitation controls. Most importers
already have written product specifications covering safety and
quality. FDA estimates that importers that do not currently have
written purchasing specifications could develop them in about two
hours of managerial time per product. In many cases, importers
could use existing Federal tolerances, action levels and guidelines
relating to safety standards, as described in the preamble to the
final rule. On average the cost of writing and updating product
specifications could be conservatively estimated at $500 per
importer. The regulation describes several affirmative steps that
importers may take. The least costly option described in the
regulation is for the importer to maintain on file a copy, in
English, of the foreign processor's HACCP plan and a written
guarantee from the foreign processor that the imported product is
being processed in accordance with that HACCP plan. The cost of
taking this step should be very small.
Several commenters questioned the agency's estimate of the number
of importers. These commenters stated that it appeared that the
agency had only included importer establishments and left out
processors who also import. These commenters stated that some
processors purchase raw materials directly from foreign suppliers.
The agency concedes that it did not include as importers processors
that purchase products directly from foreign suppliers. The agency
does not have exact data on the number of these entities but is
estimating that there are approximately 1,000 total importers. The
agency estimates that the aggregate cost to all importers is
$500,000 annually.
Cost of Processing Authorities
Several commenters questioned the ability of firms to hire
processing authorities for $1,000. These commenters provided no
information as to the cost of processing authorities. In the
absence of more information, the agency will retain the initial
estimate of $1,000.
Cost of Changing Suppliers
One commenter asserted that the cost of changing suppliers would
range from $15,000 to $50,000 instead of the $1,000 that the FDA
used in its preliminary cost estimate. In the absence of more
supporting documentation for these estimates, the agency is not
convinced that such a large expenditure would be required to find
a different supplier. The agency also believes that the need to
change suppliers will diminish as mandatory HACCP takes effect and
suppliers themselves begin to practice HACCP.
Cost of Preventing Viral Contamination of Shellfish Harvesting
Areas
The agency believes that some portion of viral infections from raw
molluscan shellfish will be prevented by eliminating the fecal
contamination of molluscan shellfish growing waters caused by the
harvesting vessels operating in those waters. One of the most
significant ways that viral diseases (e.g., illnesses related to
Norwalk and hepatitis A viruses) could be reduced is if molluscan
shellfish processors only receive raw product from harvesters that
have a means of containing human waste so that it is not discharged
overboard. If there are 10,000 harvesting vessels operating
nationally in these waters and 50% of them do not have some means
of containing waste, then these 5,000 vessels could remedy this
situation by purchasing a 5 gallon sealable bucket (approximately
$5), as proposed by the FDA to the ISSC. FDA believes that this
is an acceptable option for handling waste on harvesting vessels
with no other means of controlling waste. The total cost of adding
such facilities is negligible relative to the overall estimates of
the costs of this regulation.
Cost of Controlling Vibrio vulnificus
As the preamble to the proposed regulations pointed out, FDA does
not expect processors to install impractical controls for hazards
or to have controls that are beyond known technology. The agency
does expect processors to be aware of new developments regarding
controls and to integrate them into their HACCP systems where
practical. Moreover, where definitive controls do not yet exist
for a hazard, FDA expects that processors will take any reasonable,
prudent measures available that will reduce the risk of a hazard.
For V. vulnificus in raw molluscan shellfish, the preamble noted
that time and temperature controls constituted such prudent
measures. As better controls become available, the preamble
advised, processors of these products should incorporate them.
Better control strategies must reduce V. vulnificus infections to
achieve the benefits claimed for this regulation later in this
analysis.
Foodborne cases of infections with V. vulnificus and deaths are
associated with the consumption of raw molluscan shellfish taken
from the Gulf of Mexico between the months of April and November.
The current state of knowledge regarding factors leading to V.
vulnificus infection is incomplete. However, epidemiologic and
ecologic studies suggest that risk of infection increases with
increasing bacterial cell numbers in oysters. While data do not
support establishment of an infectious dose, reduced exposure to
the pathogen through reduction in cell counts in meats showed a
reduction in the risk of infection. It is a common practice in
many food processing plants, as well as clinical settings, to
attempt to reduce risks from bacterial pathogens by taking actions
which either will prevent proliferation of the bacteria, or which
will actually decrease the number of bacteria.
Interstate molluscan shellfish shippers could insist that
harvesters rapidly chill oysters at the harvesting location rather
than dockside. Although normal cooling will not reduce the numbers
of V. vulnificus cells, reduction would occur if raw molluscan
shellfish are cooled to 13ÓC or below. Another possible control
step is for shippers to insist that vessels without cooling
capabilities relay their harvest to refrigerated vessels or spend
a short amount of time in the harvesting area, so that the harvest
is quickly delivered to dockside refrigeration. Both control steps
could be costly to molluscan shellfish harvesters. In places where
the harvesting areas are located far from shore, shortening the
time spent at the harvesting area may not be effective since,
regardless of the time spent in the harvesting area, the shellstock
would still be out of the water for the full duration of the trip
to the dock. In the calculation of the benefits for this
regulation, FDA has assumed that 50% of all molluscan shellfish
shippers will employ time and temperature controls for the first
two years that this regulation is in effect in an effort to control
V. vulnificus.
If 50% of the shippers require, as a matter of purchasing
specifications, rapid chilling of the harvest, it may affect 50% of
the estimated 2,500 harvesting vessels operating in the Gulf of
Mexico. A few of the 1,250 oyster dredging vessels operating in
the Gulf currently use on-board refrigeration units for rapidly
cooling shellstock. These units cost approximately $5,000 each.
However, because of the large size of the units relative to the
limited deck area of many dredges, perhaps only 300 vessel
operators will install on-board refrigeration. The majority are
expected to relay to appropriate vessels or to land their product
in a shorter time after it has been harvested. For the 1,250
tonging vessels that operate relatively close to the landing sites,
shorter harvesting periods are the most feasible alternatives. FDA
estimates the cost of such an operational change to be $1,000
because of additional fuel and smaller output per unit of labor and
capital. If 1,000 of the tonging vessels shorten their times in
the harvesting areas, then the total cost for both dredges and
tongers is estimated to be $2.5 million for the first year. In
subsequent years, the cost is $1 million per year for tonging
vessels (without including the extra fuel costs for powering the
refrigeration units on dredges).
However, in order to reduce the number of cases of V. vulnificus to
between 20% and 50% of existing cases, FDA assumes that
manufacturers will employ more extensive processing changes. Some
firms may build new facilities capable of individually quick-
freezing (IQF) oysters. Such expenditures for plant and equipment
would amount to approximately $150,000 per firm. With an average
yearly Gulf Coast output of 18 million pounds of oyster meat, and
considering that approximately half of this product is be harvested
during the warm weather months, if half of the warm weather product
is subjected to an IQF process, then approximately 4.5 million
pounds of oyster meat would need to be quick-frozen. If there were
no transportation problems this quantity of product could be
processed by only a few IQF plants. Several such plants currently
exist along the Gulf Coast, with under-utilized capacity in the
warm weather season. However, almost all of the existing plants
are freezing other species. Thus, it is likely that for logistical
reasons, as many as eighteen IQF processing plants (one for each
major Gulf Coast oyster harvesting area) would be needed for a
total cost of approximately $3 million (excluding the costs of
energy and refrigerants). This analysis does not account for the
additional costs of processing oysters through IQF, for which FDA
has no information.
Another control that interstate molluscan shellfish processors
could employ is to cease marketing raw shellstock harvested from
the Gulf of Mexico from April through October. This option would
have a significant impact on those areas of the Gulf region where
the harvesting, processing, shipping and sale of molluscan
shellfish play an important part in the local economies.
The agency is currently in the process of studying the costs of
controlling V. vulnificus using actions such as the ones mentioned
above. (Contract No. 223-91-2238, Task Order No. 4, "Cost Analysis
of Options Affecting Seasonal Gulf Oyster Marketing")
Cost Effects of the Regulation on Quality
A few commenters expressed concern that compliance with the draft
Guide for tuna would require that products be frozen in order to
prevent histamine growth, and that for swordfish the mercury
testing expectations would reduce freshness due to the time needed
to perform the test. The agency is in the process of revising the
draft Guide and will take these comments into account. In summary,
the first edition of the Guide will make it clear that these
species do not need to be frozen. In addition, the agency does not
believe that there will be deterioration in product quality as a
result of any product testing that may be necessary for methyl
mercury because product need not be held pending analysis.
Cost to the Federal Government
Several commenters were dissatisfied with the agency's statement
that all Federal government costs that result from this regulation
will be covered by existing appropriations. These commenters
stated that the relevant issue is the government resources that
will be devoted to seafood inspection as a result of this
regulation that otherwise would have been devoted to the inspection
of other FDA regulated products. FDA agrees that this description
relates to the opportunity cost of this regulation and that it is
relevant. The question of agency resources that may be spent on
seafood inspections is more fully discussed in the preamble to the
final regulations. However in summary, the agency believes that
there is enough flexibility in a HACCP-based inspection system to
accommodate whatever resource situation exists in the agency at the
time of implementing HACCP.
Cost of Refrigeration
Many commenters stated that, in order to meet the requirements in
the Guide that refrigerated products be held at or below 40ÓF, they
would need to purchase new or additional refrigeration units.
Commenters generally agreed that the cost of new refrigeration
units is $50,000 per unit. FDA's preliminary cost analysis did not
include estimates of costs of increased refrigeration, but
specifically requested information from commenters. To estimate
the number of plants that will need new refrigeration units, FDA
has used data obtained from its High Risk Survey. That survey of
442 processors of temperature sensitive shellfish and specialty
products found that 6% of the storage refrigerators were operating
above 45ÓF and 14% were operating at or above 41ÓF and at or less
than 45ÓF.
For the purposes of this analysis, FDA assumes that one-tenth of
the refrigerating units operating above 45ÓF will need to be
replaced, as it is possible that they either cannot achieve 40ÓF or
the units would be so inefficient at achieving that temperature
that replacement would be the lowest cost option. This suggests
that 0.5% of processors of cooked, ready-to-eat and smoked fish
products (7 plants) will need to purchase new refrigeration units.
Assuming each plant needs only a single new refrigeration unit, the
total cost for new refrigeration units is $353,000.
The agency assumes that the estimated 14% of processors of cooked,
ready-to-eat and smoked fish products that have coolers operating
at or above 41ÓF, but at or less than 45ÓF and 90% of the 6% of the
units that are operating above 45ÓF, could achieve the 40ÓF setting
by one or more of the following: lowering the temperature at which
the current refrigeration units are set to operate, changing
operating procedures (e.g., keeping cooler doors closed or
prechilling product with forced air or ice), or making repairs to
the units. In some cases this would require more electricity,
parts replacement, or other repairs. Because the amount of
electricity needed to attain a given temperature is determined by
so many variables, each with a wide variance, the agency is not
attempting to estimate the extra cost of electricity and thus, this
may be a source of underestimation. However, this underestimate
may be partially offset by the possibility that the new
refrigeration units purchased by 0.5% of the processors of cooked
ready-to-eat and smoked fish products will be more energy efficient
than the older units they replace.
Attribution of Costs to Other Laws and Regulations
Several commenters disagreed with the agency's decision to
attribute all the costs of HACCP implementation in Alaskan plants
to an Alaskan State law that requires all processors to operate
under HACCP plans. The commenters objected that the Alaskan law
has only been applied to Alaskan seafood canning operations, though
it was enacted several years ago. Because this law has been
enforced only on canneries since its enactment, only Alaskan
canneries have incurred any HACCP costs related to the law. The
FDA agrees with these commenters. For determining both costs and
benefits, the agency uses as its baseline the best estimate of the
practices, activities, expenditures and cases of illness that exist
prior to the regulation or are predicted to change because of
market behavior. Just as the agency has not decreased its estimate
of the benefits of this regulation in anticipation of the Alaskan
law, the agency will also not decrease its estimate of the cost of
this regulation because of the possible implementation of the law.
The inclusion of costs that will be incurred by Alaskan firms in
the analysis adds 497 to the number of domestic plants affected by
the rule.
Cost of Controls on Vessels Harvesting Scombrotoxin Forming Species
Several commenters stated that the cost of temperature controls on
vessels that harvest scombrotoxic species would be approximately
$10,000 per year. Several commenters reported that the "cost of
installing a deck chill tank on albacore tuna trollers was not
included in the cost estimate. Recording time and temperature in
such a deck tank will probably be the only way for fishermen to
document that they have properly handled their fish. An insulated
ice chill tank made of stainless steel might cost $5,000." A few
commenters also reported that costs of temperature recorders and
labor for monitoring account for an additional $5,000 for a total
of $10,000. One commenter suggested that one half of the boats
might need to make this expenditure, and another commenter reported
that 8,000 vessels were involved in harvesting scombrotoxin forming
species.
FDA believes that the vast majority of scombrotoxic fish species
harvesting vessels probably already are taking proper precautions
to avoid this problem. Because scombrotoxin is at least partially
correlated with decomposition, it is in each harvester's best
interest to control this hazard. Illnesses are not always from
well established commercial fishing vessels, but can also be
associated with poor handling practices on sport fishing vessels
and by irregular suppliers of finfish into interstate commerce.
FDA expects that, under a system of HACCP controls, processors will
require that all vessels that supply them with raw materials keep
records of properly handled seafood before selling their catch.
FDA has no good data on the numbers of vessels that annually supply
scombrotoxic fish species into interstate commerce but, for the
purposes of this analysis, will assume that 100 small vessels will
need to make additional investments. It is assumed that $500 will
cover the annual cost of ice supplies, containers to hold the ice,
thermometers and catch/temperature log books to meet purchasers
specifications or approximately $50,000 per year (for the
industry).
If the foreign fleet is of the same size and condition as the
domestic fleet, the total cost for scombroid fish harvesting
vessels is $100,000 per year. However, consistent with assumptions
made elsewhere in this document (that only 25% of foreign costs are
passed on to US consumers), FDA assumes that only $12,500 of the
foreign vessel costs will be passed on to U.S. consumers.
Unintended Cost Consequences of Holding Processors Responsible for
Transporter Handling of Product
Two comments from Hawaii expressed concern that if processors need
to place time and temperature controls on airlines that transport
products to the US mainland, the airlines would likely refuse to
carry seafood. The commenters stated that air carriers find it
difficult to handle such cargos under current circumstances and
that air carriers do not want to take cargo that needs special
attention. In the absence of data to support these concerns, the
agency is unable to verify this cost.
Cost of Administrative Overhead
Two commenters questioned why FDA used rates for administrative
overhead that are significantly below the rates that the agency
uses for estimates of its own financial costs. In the PRIA, the
agency used a 20% administrative overhead factor for first year
processor costs and a 10% administrative overhead factor for
processor costs in succeeding years. The commenters noted that,
for its own activities, the agency uses an annual factor ranging
from 70% to 110%. In general, the commenters are close to the
overhead factors used by the agency in its own activities.
However, in the absence of data supplied by commenters, FDA is not
persuaded to change the overhead rates used in the PRIA for seafood
processors. Administrative costs of government and of seafood
processors are unlikely to be similar. In addition, some of the
activities for which costs are estimated could be reasonably
construed to include administrative activities whereas FDA's
administrative activities are more narrowly defined.
Furthermore, the agency will be providing processors with
administrative assistance in the form of the Guide. FDA has held
several meetings around the country to explain the proposed
regulation, as well as sending speakers to numerous meetings with
trade associations. The agency expects to engage in extensive
outreach soon after the final regulations are issued. In all of
these ways the agency is assisting the managerial implementation of
the rule, which will reduce the administrative overhead costs.
Nevertheless, the agency acknowledges that these estimates must not
be considered to be precise estimates of the total administrative
cost. The agency concludes that the estimates used in the PRIA,
20% of all other costs in the first year and 10% thereafter, are
essentially correct and will use these estimates in the analysis.
Cost of Smoked, Cured & Dried Fish Guidelines
A few commenters wrote to state that under this regulation they
would have to reformulate their smoked fish products due to the
recommendation that brine be refrigerated. These comments did not
provide any information as to the number of plants that would need
to reformulate or the cost of reformulation. The agency
acknowledges that some products may need to be reformulated and
that reformulation is not without cost. However, without any data
on the cost and number of products to be reformulated, FDA will not
make an estimate of the reformulation cost of smoked fish products.
The final regulations contain a performance standard for these
products that require them to be free of botulism toxin through the
shelf life of the product. Any reformulation that will be required
of processors will depend on whether they can meet this standard
with their existing formulations.
FDA has included in the RIA the cost of labeling smoked fish
products with a statement containing the information that the
product is to be kept refrigerated at 38ÓF or below. The costs are
$168,000, the lower bound of the $168,000 to $2.5 million estimated
in the PRIA. Based on FDA's inspection experience the agency
believes that 75% of the products already bear such a statement.
The labeling change is estimated to result in a one-time cost of
$1,000 per small plant.
Foreign Plant Costs
FDA has estimated that, on average, foreign plant costs will be
half that of domestic plant costs, due to generally lower labor
costs in foreign countries. Based on information from US Customs
Service databases on imports into the US, the agency estimates that
6,000 foreign firms ship seafood products to the US. The agency
believes that this is a generally accurate estimate of the number
of foreign plants that export to the US, since one processing plant
may supply several exporters and one exporter may purchase product
from several processing plants.
One commenter challenged the agency's assertion in the PRIA that
additional costs that will be incurred by plants now exporting to
the EU should not be attributable to this regulation. The
commenter stated that none of the EU member nations have yet
enacted HACCP requirements. But in fact, an EU directive has been
enacted to require member nations to enact HACCP requirements for
seafood within a short period of time. The agency is counting
compliance costs with this regulation by foreign processors and
vessels for the first two years. This attribution reflects some
uncertainty as to these estimates. The first uncertainty is that
the agency is not able to accurately estimate how much of the
foreign cost of this rule will be passed on to US consumers. In
this analysis the agency has assumed 25% of the foreign costs would
be passed on. The second uncertainty is that the agency does know
how many of the foreign plants that export to the US are not
exporters to the EU. The third uncertainty is that the agency does
not know to what extent the exact HACCP requirements that EU
nations place on importers will match the exact requirements of
this rule.
Because it is highly unlikely that the entire output of the plants
that export to the US is devoted to supplying the US market, FDA
attributes only a portion of the costs incurred by foreign
processors and vessels to this regulation. The agency believes
that foreign plants that make expenditures to comply with this
regulation will spread the costs of compliance over all of their
output rather than only over the portion of output that is shipped
to the US. FDA has assumed that 25% of the costs of compliance
will be passed on to US consumers. Based on these assumptions, the
agency finds that the relevant foreign costs of this regulation are
$18 million in the first year and $10 million in the second year.
Cost Estimates Based on the FDA Model
In addition to the cost estimate based on the NMFS modeling
(described above), FDA is presenting a second cost estimate for
these regulations. The uncertainties associated with the choices
that will be made by seafood processors for controlling hazards
justify providing a range of potential costs based on more than one
model.
In examples created by seafood experts within FDA, the cost of
compliance with these regulations was estimated for two small
seafood processors that the agency believes to be representative of
a significant portion of the seafood industry. One of the plants
is assumed to be in substantial compliance with existing GMP
requirements. Therefore, the costs experienced by that plant are
attributable exclusively to the establishment and maintenance of a
HACCP system. The other plant has some Current Good Manufacturing
Practice (CGMP) deficiencies that the agency believes are typical
of those displayed by seafood processors. This plant is identical
to the first plant except for the CGMP deviations. The costs
calculated for this second plant represent the costs associated
with the establishment and maintenance of HACCP as well as costs
associated with the correction and monitoring of CGMP deficiencies,
i.e., poor sanitation conditions.
The models concern two plants that cut and package tuna which is
received frozen and that also distribute orange roughy fillets.
The complexity of the processing operations, and the nature and
number of hazards, are assumed to be equivalent to those of the
other types of operations. FDA recognizes the difficulty in fully
validating this assumption. Nonetheless, the results of this
modeling demonstrate that processors may have costs that are
significantly below the averages estimated with adjustments to the
NMFS report. Moreover, as discussed later, data received about
firms that have implemented HACCP are generally supportive of the
results of this modeling.
Small Plant Cost Example 1
This is the example of a plant that is a processor of frozen tuna
steaks and distributor of imported orange roughy fillets who
receives all fish frozen. This plant is located in a major seafood
processing region, so there is no need for plant personnel to
travel to other cities to receive training. It would be available
locally. This processor operates 280 days per year. The plant
manager is paid $15 per hour and production workers are paid $8.50
per hour. No food safety hazards are reasonably likely to occur in
orange roughy, so a written hazard analysis shows hazards for tuna
only.
This processor has no need to make GMP improvements so the plant
costs are limited to the following:
1. Training: ($760) This is calculated as follows: $400 tuition
plus the opportunity cost of training time (24 hours x $15 per
hour). The processor is expected to do most of the hazard
analysis during the class.
2. HACCP Plan Refinement: ($240) This is calculated as taking 16
hours billed at $15 per hour using the Guide.
3. Plant Sanitation Audit: ($0) This is done 3 times daily for 20
minutes each time. However, because the firm is modeled as
being in compliance with existing GMPs, it is assumed that
these audits are already being done. It is assumed that there
is a negligible cost for recordkeeping.
Critical Control Points
4. Receiving CCP (histamine): ($3,200) This processor gets a
freezing log from the tuna harvester and makes a visual check
of the fish to see that they are frozen. The processor keeps
a copy of the freezing log and makes a note of the visual
check. The fish are then transferred to a plant freezer. The
monitoring takes 15 minutes per lot for 4 lots per day.
Similar monitoring is already occurring and the marginal cost
for the recordkeeping is negligible.
The processor drills a representative sample of each lot and
performs an organoleptic examination for decomposition of the
tuna. It is assumed that this monitoring is not being done
previous to this regulation and takes 20 minutes per lot for
4 lots per day. Monitoring is billed at $8.50 per hour.
Also, there is a cost for a new drill ($50), and it is assumed
that recordkeeping costs are negligible.
5. Cutting CCP (metal fragments): ($0) A worker checks the saw
blade at every break to look for broken saw teeth and keeps a
log of checking on the teeth. Monitoring takes a few minutes
per break. It is assumed that there is a negligible marginal
cost for the monitoring and recordkeeping. Fish is weighed,
packed, labeled and returned to the freezer.
Corrective Actions
6. Problems with incoming product: ($0) It is expected that
product rejects in the first year would be higher but they
would return to current levels in the second year as
harvesters became aware of the processor's new requirements.
The total cost for the industry is $1 million for the first
year and zero in the following years. Because harvesters and
not processors bear the cost of rejected raw product, this
cost is included in table 5 as a separate line item and not in
table 4 which includes only costs borne by processors.
7. A saw tooth breaks every two years: ($20) A worker needs to
examine potentially affected product every two years. This is
expected to take 4 hours billed at $8.50 to check two hours
worth of cutting.
Verification
8. Record review: ($400) This involves a review of 5 sanitation
records, 5 receiving records, and a log book for the cutting
operation. These are expected to be very simple (e.g.,
records with check marks). Consequently, this review is
expected to take 30 minutes per week billed at $15 per hour.
9. Review hazard analysis & HACCP plan: ($60) This is expected to
take 4 hours per year at $15 per hour.
10. Administrative changes - 20% of all of the other costs in the
first year and 10% in the second year.
Small Plant Cost Example 2
The categories of costs that are different from Example 1 are
explained below. Beyond these differences the plants are assumed
to be identical.
1. Plant Sanitation Audit: ($2,800) This will need to be done 3
times daily taking approximately 20 minutes for each audit. It is
assumed that some minimal sanitation assessment is already being
done once per day, but an additional 40 minutes would be required
to perform three adequate audits. Again, it is assumed that there
is a negligible cost for recordkeeping.
2. Extra Equipment Cleaning and Sanitizing: ($2,480) This is
assumed to take 1 hour per day billed at $8.50 per hour.
Additional water, and cleaning and sanitizing materials are assumed
to cost $100.
3. Eliminate Fly Infestation: ($330) Torn screens need to be
repaired taking 2 hours billed at $8.50 per hour. Screening
materials assumed to cost $15 are needed. An exterminator to apply
pesticides costs $300.
Table 5 represents the models described above in tabular form.
Table 5
FDA Models of Small Plants
Small
Plant 1
(No GMP
Costs)
Small
Plant 1
(No GMP
Costs)
Small
Plant 2
(GMP
Costs)
Small
Plant 2
(GMP
Costs)
CATEGORY
Year 1
Year 2 -
Year 1
Year 2 -
Training
760
0
760
0
HACCP Plan
Refinement
240
0
240
0
Sanitation
Audit
0
0
2800
2800
Receiving CCP
3200
3200
3200
3200
Cutting CCP
0
0
0
0
Sawtooth
Monitoring
20
20
20
20
Record Review
400
400
400
400
HACCP Plan
Review
60
60
60
60
Equipment
Cleaning
0
0
2500
2500
Eliminate
Pests
0
0
330
0
Administration
940
370
2100
900
PER PLANT
COSTS
$5,600
$4,000
$12,400
$9,900
Extrapolation to Total Industry Cost
In order to estimate an average plant cost from these FDA model
plants, FDA assumed that, based on the results of the agency's
1990/1991 survey of the seafood industry, 20% of the small firms
are similar to the model plant that requires some CGMP improvements
(Small Plant 2) and that 80% of the small firms are similar to the
model plant that is in compliance with CGMPs (Small Plant 1).
These estimates (20% and 80%) come from FDA's survey of the seafood
industry.
The agency has also assumed that the cost of compliance for large
firms is the same as that of small firms. There are offsetting
considerations that have led the agency to make this assumption in
this model. For example, agency experience suggests that it is
likely that small firms will, on average, have larger sanitation
costs and thus incur greater expenses to rectify existing CGMP
deviations. Large firms, on the other hand, are more likely to
have a greater number of products and processing lines, resulting
in greater costs of plan development and monitoring. However, the
agency believes that large firms are more likely to already have
preventive controls, formalized sanitation programs, and
recordkeeping systems in place than small firms. Additionally,
large firms are more likely to take on new monitoring regimes with
their existing quality control and production staffs than are small
firms. The agency believes that these considerations counteract
each other and should result in equal costs for large and small
firms.
To complete the FDA model, FDA assumed that exporters (one half of
the 1,000 large firms) would only need to spend $1,000 in order to
comply with this rule. Combining the two plant total costs as
reported in Table 5 and weighting the proportion of the industry
they are assumed to represent, average plant costs are estimated to
be $6,400 in the first year and $4,800 in subsequent years.
The foreign processor costs associated with this rule and passed on
to US consumers are estimated to be 13% of the average domestic
plant costs. The total cost of this regulation using this method
of cost modeling is $69 million in the first year and $38 million
in the fourth year and beyond.
Total costs for compliance with these regulations using the FDA
model are shown in table 6. Table 6
Disaggregated Costs from FDA Model
First Year Costs
Domestic manufacturers and importers: $ 32 million
Major plant repair and renovation: $ 13 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
Harvesters for rejected raw product: $ 1 million
Shellfish vessels: $ 3 million
Foreign processors: $ 5 million
First Year TOTAL: $ 69 million
Second Year Costs
Domestic manufacturers: $ 23 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
Shellfish vessels: $ 1 million
Foreign processors: $ 3 million
Second Year TOTAL: $ 42 million
Third Year Costs
Domestic manufacturers: $ 23 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
IQF Shellfish plants: $ 3 million
Third Year TOTAL: $ 41 million
Fourth Year (and subsequent years) Costs
Domestic manufacturers: $ 23 million
Sea Grant expertise: $ 1 million
Repackers and warehouses: $ 14 million
Fourth Year TOTAL: $ 38 million
Total Discounted Costs
Beyond the fourth year and discounted at 6%, the costs are $677
million. Comparison of Costs of the NMFS and FDA Models
Table 7 provides a summary of the combined estimates of average
domestic seafood manufacturer plant costs. As the table makes
clear, there is still some uncertainty surrounding the cost
estimates. There is no single source of data that FDA has found to
be entirely satisfactory for estimating the costs for the
regulations. FDA has provided estimates based on two models, as
described above, that provide a range of possibilities.
In Table 7 the first column reports plant costs developed by making
adjustments to the data in the NMFS reports. The second column
reports the model plant average presented by FDA.
Table 7
Per Plant Cost Estimates for Domestic Manufacturers
Year
Adjusted NMFS
Model
FDA Model
Plants
First Year
$23,000
$6,400
Subsequent years
$13,000
$4,800
There are a number of explanations that would account for the
uncertainty between the FDA and NMFS models. Virtually all of the
difference can be explained by the two different estimates of what
it would take to come into compliance with 21 CFR Part 110 (FDA's
CGMP regulations). In the case of the NMFS study, the contractors
estimated the cost of coming into full compliance with all CGMPs.
Using this methodology, they found that approximately 80% of the
plants were out of compliance. On the other hand, the FDA model
uses the results of FDA's own survey of the industry, which only
listed plants as being out of compliance if the CGMP violations
were related to potential contamination of the seafood product. In
this case, FDA found that only about 20% of the firms were out of
compliance. In addition, the FDA cost model assumes the simplest,
least expensive corrective action to solve the CGMP violation. The
NMFS model did not use the same approach in all cases.
Other Cost Reports
Reports received by FDA on the cost of implementing HACCP discussed
below appear generally to support the results of the FDA modeling
across the seafood industry. While the modeling was limited to
certain types of small operations, the firms for which FDA has
information on reported actual costs represent a cross section of
processing operation types, including canned, fresh, frozen,
smoked/salted, molluscan shellfish, and cooked, ready-to-eat
products as well as warehouses and repacking operations. It should
be noted that these costs are reported only as an additional source
of information. They were not used to generate FDA's model plants.
The cost information obtained from industry includes responses to
a 1991 evaluation questionnaire from 4 firms that participated in
the FDA/NOAA seafood HACCP pilot in 1990-1991. It also includes
information provided to FDA from 7 firms through the assistance of
the National Food Processors Association. (These seven firms
operate a total of 44 processing plants.) It further includes
information from two seafood trade associations, the National
Fisheries Institute and the New England Fisheries Development
Association (NEFDA), which provided FDA with summary information
about member firms that had implemented HACCP systems. The two
seafood trade associations provided information on 16 firms. NEFDA
operated a HACCP pilot with member firms through a Federal grant.
All of this information was received by FDA before the publication
of the proposed regulations and was reported in the PRIA. After
the publication of the proposal, FDA received information from a
large processor-exporter on its HACCP start-up costs. This
processor reported start up costs of $1,000 per plant. In total,
FDA has information on 86 plants.
Many of these firms have implemented HACCP as participants in pilot
programs, the NOAA fee-for-service program, or the State of Alaska
program, and therefore their HACCP systems have been subject to
some form of third party verification. Virtually all of these
plants have developed HACCP plans, many of which included CCPs for
quality or economic fraud or both in addition to safety. In this
respect, many firms implemented a more extensive form of HACCP than
is being mandated by FDA.
More complete information on start-up costs received from 22 firms
who have implemented HACCP is summarized in table 8. Some of these
costs are for multi-plant firms and some for firms operating only
one plant. Table 8
Start-up Costs
# Firms Start-up costs
4
< $1,000
15
$1,000-$5,000
1
$5,0000-$10,000
1
$10,000-$15,000
1
> $20,000
FDA notes that there are several uncertainties with these data.
The agency does not have sufficient information to extrapolate the
costs observed by these firms to the entire industry. FDA also
does not know the extent of previous HACCP-type activities in these
firms so that they may have different incremental costs than the
industry average. Additionally, for costs in subsequent years,
some of the firms reported costs that exceeded the start-up costs
shown in Table 8 although some were below, and it is not clear if
costs that might be incurred in order to comply with CGMPs are
represented. Nevertheless, the range of reported start-up costs
are consistent with the FDA model for a processing operation that
is in full compliance with CGMPs.
Notably, the estimates developed for NMFS of the costs of operating
HACCP systems for small businesses are consistent with the FDA
model and with the costs reported by firms actually operating HACCP
systems as discussed above. Thus, three independent sources of
information suggest that average annual HACCP costs, at least for
small businesses, are within a range of $3,000 to $6,000 per plant
if sanitation costs are not included. Although the HACCP cost
estimates made for NMFS did not include certain aspects of a HACCP
system such as HACCP plan development, plan verification, and
taking corrective actions, the estimates did include the costs of
operating HACCP systems for quality and economic adulteration in
addition to safety. The FDA HACCP system involves safety only and
is therefore less expensive in that respect.
It is also worth noting that three independent sources (FDA's own
inspection experience, NMFS' inspection experience with plants
purchasing its voluntary inspection services, and the contractor's
report for NMFS) confirm the existence of sanitation deficiencies
in some seafood plants. Since FDA holds that these conditions must
be corrected under existing requirements, then the costs associated
with these corrections will be borne by processors regardless of
whether sanitation provisions are included in the seafood HACCP
regulations or somewhere else. Sanitation controls for processors
may address a number of enteric pathogens discussed elsewhere in
this analysis including Salmonella, Shigella, hepatitis A, Listeria
monocytogenes, campylobacter, and Clostridium botulinum.
Contamination may come from either the raw product or from poor
hygiene practices such as insufficient control of vermin (flies and
rodents) or insanitary water. In addition, poor sanitation may
cause contamination of the product with pesticides, lubricants,
cleaning compounds, or other toxic substances because of improper
labeling, storage or use. The sanitation control system in the
seafood HACCP regulations is based on the monitoring of sanitation
conditions by processors. FDA is not aware of any method for
processors to take control of the sanitation conditions within
their plants other than by a method that involves routine
monitoring. Recording the results of these observations, as
required by the regulations, need involve only minimal additional
cost.
Seafood Prices
A number of commenters referred to the effect that the regulation
will have on the price of seafood that consumers experience at the
retail level. The PRIA presumed that most of the cost of
compliance of the proposed regulations would be passed on to
consumers. It calculated that if the domestic industry passed on
to consumers all of the costs estimated in the PRIA, prices for
domestically produced seafood would increase by less than 1% in the
first year and less than 0.5% in succeeding years. The PRIA noted
that price changes of such magnitude are unlikely to have a
significant impact on general seafood purchases.
Some commenters claimed that all of the cost of the regulation
would be born by processors, and that none of the increase in cost
would be passed on to consumers. These commenters explained that
seafood is currently at a disadvantage compared to other flesh
foods for consumers' food dollars because seafood has a higher
price per pound. If the relative price of seafood were to increase
further, consumers would eat less seafood. These commenters also
explained that domestic seafood processors are at a competitive
disadvantage compared with seafood that can be imported at low
cost. If domestic processors were to raise their prices, seafood
imports would take an even larger part of the seafood market away
from domestically produced seafood.
Other commenters said that processors will pass on all of the cost
of the regulation, and that the regulation will cause the consumer
price of seafood to rise. Some said that the price increase would
be large enough to cause a decrease in seafood consumption.
Both theories have some merit, although neither is completely
correct. The agency agrees that, all other things remaining the
same, an increase in the price of seafood will decrease seafood
consumption and increase the consumption of other flesh foods.
However, the decision of a consumer to purchase a product depends
on a number of factors.
Seafood includes many species which vary significantly across types
of products and over time in price per pound. Such diversity of
price makes it clear that there is not perfect substitution among
the flesh foods. Nevertheless, databases on food consumption are
equally clear at showing that as people have increased their
consumption of various seafood products, they have reduced their
consumption of meat and poultry.
There are other non-price factors in the consumption decision. A
consumer survey found that taste, quality, and freshness were rated
above price in decisions to order seafood in a restaurant or to
purchase for preparation at home. In a survey of retailers'
experiences, consumers ranked quality ahead of price in making
seafood selections and rated the need for information on cooking as
a concern equal to price.
Another relevant consideration is the fact that a disproportionate
percentage of seafood is consumed in restaurants as a luxury item
where the cost of the raw material plays a smaller role in
influencing the purchasing decisions of consumers.
All of this information is consistent with other data in this
analysis that suggests that a 1% change in price results in less
than a 0.5% change in seafood consumption.
Another major factor that lessens any competitive cost advantage
that meat and poultry products might gain from an increase in the
price of seafood is that USDA is proposing similar HACCP
regulations for meat and poultry. USDA's proposal, if finalized
for meat and poultry products, suggests that all segments of the
flesh food market may face cost increases in the near future. It
is entirely possible that the price of seafood relative to meat and
poultry will not change. The agency agrees that some seafood
imports have a cost advantage over domestically produced seafood,
primarily due to lower labor and capital costs of production.
However, because the regulation applies to imports as well as
domestic products and because exporters from and importers to EU
member nations will soon be under HACCP requirements and
experiencing increased costs, it is reasonable to assume that the
price of imported seafood relative to domestic seafood will not
change.
In the short run, the ability of producers to pass on cost
increases is largely determined by the elasticity of demand (the
degree to which consumers reduce their consumption of a good in
response to a given increase in price) and the elasticity of supply
(the degree to which producers increase their production of a good
in response to a given increase in price). The elasticity of
demand is determined in turn by, among other things, the presence
or absence of close substitutes. Thus, for example, if there are
close substitutes and the price of a good goes up, consumers will
not continue to consume the higher priced good but switch to one of
the substitutes.
If manufacturers know that consumers will not switch to a
substitute when there is a price increase, then they are free to
pass along all of the increased costs (from complying with the
regulation) in the form of price increases. However, where there
are substitutes for seafood, such as other flesh foods, consumers
respond to price increases by reducing their consumption of the
higher priced good. Rather than attempting to pass on all of the
costs of the regulation in the form of higher prices, producers
must accept reduced profits and bear some, if not all, of the
burden of the cost increase.
In very competitive markets, such as the market for flesh food,
where meat, fish, and poultry are substitutes, producers bear the
entire burden of any increases in fixed costs. Fixed costs are
costs that do not change, despite the size of the firm and changes
in the level of output. Examples of fixed costs are costs of
plant, equipment, and management; much of these costs are expected
to be borne by processors. Because large firms spread fixed costs
over larger output, they are more able to pass on these costs than
are smaller firms.
In addition, also in the short run, producers may bear some portion
of the variable costs that cannot be profitably passed on to
consumers. Variable costs are costs that vary with changes in the
amount of output. Examples of variable costs are costs of raw
materials and hourly labor. However, it is likely that much of the
variable costs will be passed on to consumers.
When firms in a competitive market cannot pass on all of a cost
increase in the short run, profits decline. Beyond some point
profits become either so low or negative that the firm is forced to
close (discussed more fully in the Regulatory Flexibility Analysis
below). In the long run, the exit of these marginal firms reduces
the industry supply of seafood and permits the remaining firms to
raise prices to cover the full costs of production, both variable
and fixed costs. Thus, in the long run, seafood prices will rise
by the full cost of the regulation.
A few commenters requested a better analysis of price changes.
These commenters criticized the approach used to estimate price
increases in the Executive Summary of the PRIA. Rather than
dividing the estimated domestic cost of the regulation by the total
domestic production, the commenters suggested estimating price
changes for each market segment. The advantages of this approach
are that different types of seafood are treated separately (the
change in the price of frozen tuna might be very different from the
change in the price of ready-to-eat shrimp cocktail) and that
different sized firms are treated separately (small firms may be
forced to raise prices more than large firms).
FDA agrees that this method of determining price changes is more
appropriate than the method employed in the PRIA. However, FDA did
not receive any information from commenters that would enable the
agency to calculate prices in this manner. It is worth noting,
however, that the NMFS reports did take product type into account
when estimating price increases. That contractor estimated a range
of price increases from negligible to 1.3%, depending on the
product. Again, it is important to note that that study included
costs for the control of types of hazards not covered by this final
regulation.
Finally, while the methodology used in the PRIA might not produce
accurate price changes, it suggests that overall price increases
due to this regulation could well have a negligible effect on
demand.
Unquantified Costs
Several commenters suggested that costs of litigation of issues are
unquantified costs of this regulation. The agency agrees that
these are real costs and that it would be difficult to quantify
them. Other costs that FDA has either not quantified or quantified
in the PRIA but were later challenged as wrong by commenters
without supplying sufficient data to change, include:
o Record storage costs which were not fully quantified.
o The cost of the repair of current refrigerators, and the
cost of operating current refrigeration units at lower
temperatures.
o Time and temperature controls on airlines.
o The cost and number of smoked, cured and dried fish that
must be reformulated.
o The cost of operating IQF molluscan shellfish plants.
BENEFITS
Benefits that will be realized as a part of this rule include both
those which FDA has quantified and those which have remained
unquantified. All benefits are shown in the table 9:
TABLE 9
TYPE OF BENEFIT
LOWER BOUND
UPPERBOUND
Safety Benefits - (decreased
illness and death)
$45 million
(4th year)
$116 million
(4th year)
Export Benefits (elimination
of duplicative effort)
$20 million
$20 million
Reduced Enforcement Costs
$22 million
$22 million
Increased Consumer Confidence
in Seafood
Unquantified
Better Process Control
(Reduced processing costs)
Unquantified
Improved Employee Morale
Unquantified
Reduced Rent Seeking
Unquantified
For the four unquantified benefits, FDA is more convinced that
positive benefit will obtain from increased consumer confidence and
better process control, as opposed to the latter two, improved
employee morale and reduced rent seeking, which FDA regards as
being more speculative.
In the PRIA, FDA estimated that the proposed option, which is being
adopted in this final rule, would, (1) reduce the amount of seafood
related illness, (2) generate nutrition benefits as a result of the
increased consumption of seafood (brought about by a decrease in
consumer anxiety about seafood) with a concomitant decrease in the
consumption of meat and poultry, (3) reduce the amount of rent
seeking (rent seeking is a term economists have applied to
activities which do not contribute to societal welfare but only
seek to transfer resources from one party to another), and (4)
generate export benefits to firms by allowing U.S. exporters to
continue to export to countries that require Federal oversight and
certification of HACCP programs.
In addition to the benefits cited in the PRIA, the agency is
claiming benefits derived from reduced enforcement costs, and is
discussing other unquantified benefits of finalizing this rule.
The agency has fully considered all of the comments on benefits.
What follows is FDA's conclusion as to how these benefits should be
valued.
(1) Safety Benefits (reduced seafood-related illness)
A number of commenters did not view HACCP as having large, or in
fact any benefits in terms of either improved safety or consumer
confidence. As discussed in the preamble to the final regulations,
FDA reasserts the benefit of HACCP with respect to illness
reduction. Part of the commenters concern may be the result of a
misconception on the part of the commenters of how mandatory HACCP
is intended to work: dividing responsibility between the
government and industry. For example, it is likely that new
technology or information will become available in the future that
will allow industry to more readily control hazards at CCPs than is
currently possible. Because HACCP systems of preventive controls
are operated by the processors themselves, they will need to avail
themselves of new technologies and information as they become
available. No additional regulation will be necessary to ensure
the adoption of new technologies. The role of the government will
be to confirm that hazards are controlled to the maximum extent
practicable with the prevailing knowledge and technology.
For example, at present there is no quick, inexpensive test to
determine if finfish are contaminated with ciguatera. If one is
developed in the near future, it would be advantageous to
processors for whom control of ciguatera is necessary to avail
themselves of that test to reduce the probability of a hazard being
present in their product. FDA's role would be to verify that
processors are, in fact, controlling the hazard. This verification
will take into account whether the best practical technology is
being used where it would produce superior assurances of safety
than other forms of control. In the interim, processors should
refrain from purchasing certain species of fish that have been
caught on reefs that are known to be ciguatoxic at the time of
harvest.
The introduction of superior technologies that can be used to
control ciguatera under the HACCP framework should eventually
reduce or eliminate ciguatera intoxications. However, as with any
innovation, the timing of development of such technologies is
difficult or impossible to predict. Therefore, it is also
difficult to predict costs and benefits of the adoption of future
innovations. For the most part, FDA has not attempted to speculate
on these innovations, and costs and benefits are estimated only for
known control technologies. However, because some information is
known about the control of Vibrio vulnificus, and because
innovative techniques are being explored now, FDA has made a
preliminary attempt in this analysis to explore costs and benefits
of future actions which may occur to control this hazard.
Comments Pertaining to the Baseline Number of Cases
One commenter said that it is improper to lump all seafood species
together in the baseline figure as many species may not be
responsible for any illness. FDA agrees that it would be desirable
to link illness data to a seafood species or a finished product
type to the extent possible. This is addressed in table 11.
One commenter said that the number of illnesses reported reflect an
individual's perception that the illness resulted from the last
food consumed.
FDA's method of calculating the baseline illnesses should have
minimized this problem. The baseline illnesses were calculated
from the data reported to the Centers for Disease Control and
Prevention (CDC). To the extent this bias exists in CDC's data,
then this criticism is correct. However, the CDC data, upon which
the baseline illnesses are based, are contributed by State
epidemiologists and other public health professionals. Although
there is some difficulty with diagnostic tests that are performed
to identify the cause of illness so that some illnesses are
misreported, sufficient screening by qualified health professionals
before data entry minimizes the risk of improperly attributing the
source of the illness reported to CDC.
One comment suggested that FDA should include in its benefits
tables other hazards addressed by the regulation including
"chemicals, drugs and decomposition." Another comment mentioned
that mercury, PCBs, amnesic shellfish poisoning (ASP) and
environmental contamination should be included. Other commenters
also addressed hazards covered in the Guide but not in the economic
impact assessment, again including hazards such as mercury and
other chronic chemical hazards.
FDA has not included as part of the benefit analysis any chemicals
(such as mercury), pesticides, animal drugs, food or color
additives or decomposition. Because the manner in which risk
assessment is performed for exposure to chemicals, additives and
drugs, it is difficult to predict a baseline number of illnesses
from exposure to these hazards. If, however, this regulation
prevents chronic illnesses from chemicals, pesticides and drugs
which may result in death, the benefits would be large. FDA and
CDC estimate that there are approximately 20 cases per year of
Neurotoxic Shellfish Poisoning, Amnesiac Shellfish Poisoning and
Demoic Acid poisoning from shellfish each year. Table 9 lists all
of these illnesses under "Other Marine Toxins."
Decomposition not involving safety is outside the scope of this
rulemaking. FDA notes however, as a secondary effect, these
regulations could reduce the amount of low quality seafood due to
decomposition in the marketplace. For example, the controls for
histamine that will be put in place to meet the requirements of
these regulations could improve the quality of histamine-producing
species as well as their safety. A reduction in decomposition
could similarly be expected in such products as cooked, ready-to-eat seafood, mo
lluscan shellfish and smoked fish as a result of
time-temperature controls intended to prevent pathogen
proliferation. Although FDA has not quantified this benefit, the
agency believes that it may be substantial.
One commenter noted that no illnesses are associated with L.
monocytogenes. FDA is also unaware of any domestic occurrence of
listeriosis associated with the consumption of seafood. However,
listeriosis has been attributed to the consumption of smoked
mussels in New Zealand, a product also distributed in the U.S. It
has been difficult to study attributable risks for listeriosis
because of the lengthy incubation period (3 weeks) and because the
milder gastrointestinal syndrome has only recently been identified.
Therefore, it is likely that listeriosis may be considerably
underreported. FDA believes that current measures to control L.
monocytogenes will control potential cases of listeriosis but has
not attempted to ascribe any numbers of illnesses or benefits at
this time.
Control for this pathogen will be necessary in some seafood
products. FDA has not included as a benefit to consumers a
reduction in the numbers of cases of listeriosis, but has included
benefits to the industry, shown in tables 16, 17, and 18 (and
ultimately consumers) through reduced regulatory action relative to
products contaminated with L. monocytogenes. The reduction in
regulatory actions and contamination from L. monocytogenes can be
attributed to the identification of hazards and improved control
measures mandated by the regulation, especially with respect to
sanitation.
One commenter stated that "statistics related to foodborne illness
caused by recreationally harvested seafood should be separated from
illnesses caused by commercial products." The commenter's point
was to ensure that consumer perceptions of the safety of commercial
seafood are based on accurate data. Other commenters also
suggested that seafood illnesses traced to both recreational
harvest and from closed or unauthorized areas be separated.
Finally one commenter cited an article in the Journal of Food
Technology (February, 1994. p. 117) that states that "only three
percent of reported seafood illnesses are due to commercial
processing. The remainder is due to home handling preparation or
other reasons." Other commenters pointed to abuse by food service
operations as the other main source of seafood illness.
FDA agrees that this sort of exercise, separating illnesses by the
various types of cause, would be useful but does not believe that
this would be possible given the fragmentary nature of the reported
illness data. For example, for some types of hazards, such as
Neurotoxic Shellfish Poisoning (NSP) and Paralytic Shellfish
Poisoning (PSP), virtually all illnesses may be attributable to
recreational harvesting. For other hazards, such as Vibrios, there
is unlikely to be any difference in the numbers of illness per
pound of product harvested whether it is from recreational or
commercial harvest, because vibrios can be ubiquitous and cannot
now be well detected before harvesting. Regarding the point the
commenter made as to harvest from closed or unauthorized areas, FDA
is convinced from experience that much of this harvest finds its
way into commercial channels illegally, and although it is not now
possible to determine the actual amount of this harvest relative to
that which is legally harvested. Much of the catch from
recreational harvest also ends up in commercial channels. It is
clear that although some qualitative information concerning the
causes of these illnesses is known, it is difficult to precisely
determine the ultimate cause. Moreover, the numbers of cases of
illness averted as reported in this RIA take into account, to the
fullest extent possible, the likely contribution of recreational
harvests. That is, the estimates of the illnesses averted did not
include those caused by recreational harve