oversight

FSIS Rulemaking Process for the Proposed Rule: Modernization of Swine Slaughter Inspection

Published by the Department of Agriculture, Office of Inspector General on 2020-06-25.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

     United States Department of Agriculture




FSIS Rulemaking Process for the
Proposed Rule: Modernization of
Swine Slaughter Inspection




Inspection Report 24801-0001-41
June 2020
                                               OFFICE OF INSPECTOR GENERAL
FSIS Rulemaking Process for the Proposed
Rule on the Modernization of Swine Slaughter
Inspection
Inspection Report 24801-0001-41
OIG reviewed the proposed Modernization of Swine Slaughter Inspection rule
to determine whether FSIS’ worker safety analysis complied with applicable
rulemaking requirements.


OBJECTIVE                          WHAT OIG FOUND
                                   The Department of Agriculture’s (USDA) Food Safety
We determined whether FSIS,
                                   and Inspection Service (FSIS) enhances public health
with respect to its worker
safety analysis section of the     and well-being by protecting the public from foodborne
proposed rule: (1) complied        illnesses and ensuring that the Nation’s meat, poultry,
with public transparency           and egg products are safe, wholesome, and correctly
requirements under E.O. 13563;     packaged. On February 1, 2018, FSIS published its
(2) made information about its     proposed rule, the Modernization of Swine Slaughter
preliminary analysis on worker     Inspection (“the proposed rule”). As part of this rule,
safety clearly accessible to the   FSIS proposed to revoke maximum swine slaughter line
public during the comment          speeds for participating establishments and authorize
period; (3) adhered to the
                                   them to set their own line speeds based on their ability
Guidelines in developing the
                                   to maintain quality and performance measures. In the
proposed rule; (4) came to a
reasonable determination about     proposed rule, FSIS compared worker safety data from
the reliability of the OSHA        the Occupational Safety and Health Administration
injury data it used for the        (OSHA) for large establishments with different allowed
proposed rule; and (5) consulted   line speeds (“worker safety analysis”).
with OSHA and NIOSH about
the impact of the proposed rule    On March 26, 2019, 16 members of Congress sent a
on workplace safety and health.    formal request to USDA’s Office of Inspector General
                                   (OIG) to review USDA’s rulemaking process related to
                                   the proposed rules’ worker safety analysis. Based on our
REVIEWED                           inspection, we concluded that FSIS generally complied
We interviewed FSIS officials      with the public participation requirements under
and reviewed the proposed and      Executive Order (E.O.) 13563 and, to the extent required,
final rules; relevant E.O.s and    communicated to OSHA and the National Institute for
the Guidelines; worker safety      Occupational Safety and Health (NIOSH) about the
analysis documentation; and        impact of the proposed rule. However, we found that
communications during the
                                   FSIS did not fully disclose its data sources in its worker
rulemaking process.
                                   safety analysis. Additionally, we concluded that it
                                   did not fully adhere to the USDA Information Quality
                                   Activities Guidelines (“Guidelines”) data presentation
RECOMMENDS                         and transparency requirements in the worker safety
We recommend that FSIS update      analysis section in the proposed rule. Finally, we
its internal procedures for the    concluded that FSIS did not take adequate steps to
rulemaking process, determine      determine whether the worker safety data it used for the
the impact of the omissions        proposed rule were reliable.
from the proposed rule, and
communicate to the public the      FSIS provided its response to our findings and
actual review period and known     recommendations, and we accepted management decision
limitations of the OSHA data.      on two of the four recommendations.
                              United States Department of Agriculture
                                      Office of Inspector General
                                        Washington, D.C. 20250


DATE:            June 23, 2020

INSPECTION
NUMBER: 24801-0001-41

TO:              Paul Kiecker
                 Administrator
                 Food Safety and Inspection Service

ATTN:            Cara LeConte
                 Chief Financial Officer
                 Office of the Chief Financial Officer

FROM:            Gil H. Harden
                 Assistant Inspector General for Audit

SUBJECT:         FSIS Rulemaking Process for the Proposed Rule: Modernization of Swine
                 Slaughter Inspection


This report presents the results of the subject review. Your written response to the official draft
is included in its entirety at the end of the report. We have incorporated excerpts from your
response, and the Office of Inspector General’s (OIG) position, into the relevant sections of the
report. Based on your written response, we are accepting management decision for
Recommendations 1 and 4. Please follow your internal agency procedures in forwarding final
action correspondence to the Office of the Chief Financial Officer (OCFO).

We initiated this inspection in order to answer five specific questions raised by members of
Congress. These questions—and therefore OIG’s inspection work—pertained solely to the
worker safety analysis section of the proposed rule.

FSIS has laid out two primary concerns in its response to our report regarding our findings.
First, FSIS stated that OIG’s inspection of compliance with certain criteria was misguided
because it did not believe the criteria applied to the worker safety analysis section of the
proposed rule. 1 Although FSIS stated that the worker safety analysis was not a basis for the
rulemaking, FSIS also stated that it “recognizes that evaluation of the effects of line speed on


1
 These criteria are Executive Order 13563 and the USDA Information Quality Activities Guidelines. Both are
applicable analyses used in support of rulemaking activities.
Paul Kiecker                                                                                                        2


food safety should include the effects of line speed on establishment employee safety.” 2
Because FSIS chose to include the worker safety analysis in the rule, it follows that the analysis
was prepared and presented in support of the rulemaking activities. Therefore, we feel the
criteria is applicable. 3

Additionally, FSIS believes that OIG has placed distorted emphasis on minor errors in FSIS’
proposal text, which they described as an omission of a website citation and a typo in a table
header. 4 While we acknowledge that these issues may have been inadvertent, the scope of our
inspection addressed specific, Congressional questions, including whether FSIS made
information about its worker safety analysis clearly accessible and whether it complied with
specific USDA guidelines. We therefore believe that omitting the citation and inaccurate
presentation of the worker safety analysis’ scope period fall within the context of our inspection
scope. We do not believe that there is disproportionate or distorted emphasis on these errors.
(See Objectives 2 and 3 of this report.)

Based on your written response, management decision has not been reached on
Recommendations 2 and 3. The information needed to reach management decision on the
recommendations is set forth in the OIG Position section following each recommendation. In
accordance with Departmental Regulation 1720-1, please furnish a reply within 60 days
describing the corrective actions taken or planned, and timeframes for implementing the
recommendations for which management decisions have not been reached. Please note that the
regulation requires management decision to be reached on all recommendations within 6 months
from report issuance, and final action to be taken within 1 year of each management decision to
prevent being listed in the Department’s annual Agency Financial Report.

We appreciate the courtesies and cooperation extended to us by members of your staff during our
inspection fieldwork and subsequent discussions. This report contains publicly available
information and will be posted in its entirety to our website (http://www.usda.gov/oig) in the
near future.




2
  Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
3
  The general requirements of the Guidelines apply to all types of information disseminated by USDA agencies and
offices. The regulatory supplement of the Guidelines states, “These supplementary guidelines pertain to information
disseminated to the public by USDA agencies and offices in conjunction with their regulatory activities, rulemaking
activities, and program implementation activities that are subject to notice and comment procedures. Such
information includes economic, cost/benefit, scientific, environmental, risk assessment, reporting and record
keeping, and other pertinent analyses prepared or presented by agencies in support of those activities.”
4
  For clarification, the misstated scope of analysis discussed in Objective 3 was not within a table header but in the
proposed rule’s narrative discussion of FSIS’ worker safety analysis.
Table of Contents

Background and Objectives .........................................................................1
Section 1: Transparency and Accessibility of Worker Safety Information in
FSIS’ Proposed Rule—Modernization of Swine Slaughter Inspection ............5
Objective 1: Did FSIS, with respect to its worker safety analysis section of the proposed rule,
comply with public transparency requirements under E.O. 13563? ......................................... 5
Objective 2: Did FSIS, with respect to its worker safety analysis section of the proposed rule,
make information about its preliminary analysis on worker safety clearly accessible to the
public during the comment period?......................................................................................... 7
         Recommendation 1.............................................................................9
Section 2: Quality of Worker Safety Information used in FSIS’ Proposed
Rule—Modernization of Swine Slaughter Inspection .................................. 10
Objective 3: Did FSIS, with respect to its worker safety analysis section, adhere to the USDA
Information Quality Activities Guidelines in developing the proposed rule?.......................... 10
         Recommendation 2........................................................................... 12
         Recommendation 3........................................................................... 13
         Recommendation 4........................................................................... 14
Objective 4: Did FSIS come to a reasonable determination about the reliability of the OSHA
injury data it used for the proposed rule?.............................................................................. 16
Objective 5: Did FSIS consult with OSHA and NIOSH about the impact of the proposed rule
on workplace safety and health? ........................................................................................... 19
Scope and Methodology............................................................................. 20
Abbreviations ........................................................................................... 22
FSIS’ Response ......................................................................................... 23
Background and Objectives
Background
The Department of Agriculture’s (USDA) Food Safety and Inspection Service (FSIS) enhances
public health and well-being by protecting the public from foodborne illnesses and ensuring that
the Nation’s meat, poultry, and egg products are safe, wholesome, and correctly packaged. FSIS
operates under multiple food safety statutes,1 including the Federal Meat Inspection Act (FMIA).
Since 1906, FMIA has authorized FSIS to inspect all meat products sold through interstate
commerce and imported products to ensure that they meet U.S. food safety standards. For fiscal
year 2018, FSIS was appropriated $1.06 billion to carry out services authorized by FMIA, the
Poultry Products Inspection Act, and the Egg Products Inspection Act.

FSIS aims to prevent foodborne illness by performing food safety inspection activities at more
than 6,000 establishments nationwide and ensuring that its inspections align with existing and
emerging risks. FSIS deploys over 7,500 inspection program personnel to these establishments
to ensure compliance with Hazard Analysis and Critical Control Point Systems (HACCP) plans,
policies, and regulations. 2 FSIS’ Office of Policy and Program Development develops
regulations and associated instructions for inspectors to follow in order to prevent foodborne
illnesses and protect public health. The office’s mission is to leverage science, use statutes, and
interpret data to develop policies and instructions for the field that further promote the safety of
meat, poultry, and processed egg products and lead to improvements in FSIS inspections.

Modernization of Swine Slaughter Inspection Proposed Rule

On February 1, 2018, FSIS published a proposed rule in the Federal Register titled
Modernization of Swine Slaughter Inspection (“the proposed rule”). 3 This rule proposed to
establish a new optional inspection system for market hog slaughter establishments called the
New Swine Slaughter Inspection System (NSIS). Under the proposed rule, NSIS would require
swine slaughter establishment employees to sort and remove unfit animals before FSIS’
antemortem inspection and to identify defects on carcasses and parts before FSIS’ postmortem
inspection. The rule was intended to shift FSIS resources to enable inspectors to conduct more
offline inspection activities and to reduce the number of online inspectors to a maximum of three




1
  FSIS operates under the following statutes: Poultry Products Inspection Act of 1957, 21 U.S.C. §§ 451-472; Egg
Products Inspection Act of 1970, 21 U.S.C. §§ 1031-1056; Agricultural Marketing Act of 1946, 7 U.S.C.
§§ 1621-1639s (select sections); Humane Methods of Slaughter Act of 1958, 7 U.S.C. §§ 1901-1907; and Federal
Meat Inspection Act of 1906, 21 U.S.C. §§ 601-695.
2
  HACCP is a science-based process control system for food safety which identifies hazards and measures for their
control. HACCP’s goal is to assess hazards and establish control systems that focus on prevention rather than
relying mainly on end-product testing.
3
  Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,780 (Feb. 1, 2018).

                                                                INSPECTION REPORT 24801-0001-41                 1
per line per shift. 4, 5 Additionally, the proposed rule states that a key element of the proposed
NSIS is “revoking maximum [swine slaughter] line speeds [for NSIS establishments] and
authorizing establishments to determine their own line speeds based on their ability to maintain
process control for preventing fecal contamination and meeting microbial performance measures
during the slaughter operation.” 6, 7

In the proposed rule, FSIS acknowledged that evaluation of the effects of line speed on food
safety should include the effects of line speed on establishment employee safety. FSIS also
included a summary of its analysis comparing worker safety data for establishments with
different allowed line speeds (“worker safety analysis”). To perform its analysis, FSIS used
publicly available data from the Occupational Safety and Health Administration (OSHA). The
OSHA data included 56 establishments that voluntarily submitted work-related injury and illness
data to OSHA within the scope of FSIS’ analysis. FSIS compared in-establishment injury and
illness rates from calendar years (CYs) 2002 to 2011 between 5 establishments participating in
an FSIS pilot program with higher line speeds (HACCP-Based Inspection Models Project
(HIMP) establishments) and 24 traditional establishments. 8, 9 The results of FSIS’ worker safety
analysis found that the HIMP establishments had a lower mean number of reported injuries for
three OSHA injury and illness rates: (1) total case rate; (2) days away, restricted, and transfer
case rate; and (3) days away from work case rate. 10 However, in its discussion of these findings
in the proposed rule, FSIS noted that factors other than line speed may affect injury rates (for
example, automation and number of sorters per line). 11

Laws and Orders Governing the Federal Rulemaking Process

Congress enacted various laws to protect the integrity of the Federal rulemaking process. One of
these laws, the Administrative Procedure Act, requires Federal agencies to publish notices of
proposed rulemaking in the Federal Register so the public may participate in the rulemaking
process. 12 Further, Executive Order (E.O.) 12866 requires that each agency should, where
appropriate, seek the involvement of those who are intended to benefit from and those expected

4
  Offline inspectors verify that establishments have adopted controls in their food safety system that demonstrate
effectiveness in reducing the occurrence of pathogens, including the controls that prevent contamination of carcasses
and carcass parts.
5
  As part of the postmortem inspection, online inspectors in livestock establishments inspect each carcass to ensure it
and its attached parts are free of contamination.
6
  Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,781 (Feb. 1, 2018).
7
  The current maximum swine slaughter line speed is dependent on the size of the slaughter establishment, the
number of FSIS inspectors on the slaughter line, and the type of hog being slaughtered. The maximum line speed
for establishments similar to those that would be eligible to implement NSIS is 1,106 hogs per hour. Comparable
establishments slaughter market hogs and have seven online FSIS inspectors.
8
  FSIS initiated its HIMP pilot program in October 1999. HACCP refers to a science-based process control system
for food safety. There are five swine slaughter establishments nationwide that participate in HIMP.
9
  The other 27 traditional swine slaughter establishments were excluded due to their low slaughter volume. FSIS
believed the 24 traditional establishments with high slaughter volumes would provide a better comparison group
because all 5 HIMP establishments have high slaughter volumes.
10
   The days away from work case rate is the number of calendar days away from work as a result of work-related
injuries and illnesses. The days away, restricted, and transfer case rate includes the days away from work case rate
as well as injury and illness cases with job transfer or restricted work, or both.
11
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
12
   5 U.S.C. § 553.

2     INSPECTION REPORT 24801-0001-41
to be burdened by any regulation (e.g., State, local, and Tribal officials) before issuing a notice
of proposed rulemaking. In addition, agencies are to provide the public a meaningful
opportunity to comment on any proposed regulation via a public comment period. 13
Furthermore, for both proposed and final rules, to the extent feasible and permitted by law,
agencies must provide timely online access to the rulemaking docket on regulations.gov,
including relevant scientific and technical findings, in an open format that can be easily searched
and downloaded. 14 For proposed rules, such access must include an opportunity for public
comment on all pertinent parts of the rulemaking docket, including relevant scientific and
technical findings. 15

Congressional Interest in the Proposed Rule

On March 26, 2019, a formal request, signed by 16 members of Congress, was sent to USDA’s
Office of Inspector General (OIG) regarding the findings and methods FSIS used during the
rulemaking process for the proposed rule. 16 In the letter, Congressional members expressed
concern about reports that FSIS used “flawed worker safety data” in developing the proposed
rule, and that the agency’s actions raised serious concerns about a lack of required
transparency. 17 The letter requested that OIG review specific areas of FSIS’ proposed
rulemaking process related to its worker safety analysis.

In a letter dated June 21, 2019, the Inspector General responded to Congressional members’
concerns and stated that OIG would conduct an inspection to assess FSIS’ proposed rulemaking
process based on our determined objectives. 18

Final Rule Published During Our Fieldwork

On October 1, 2019, FSIS published the final rule for the Modernization of Swine Slaughter
Inspection. In the final rule, FSIS included the results of its worker safety analysis as they were
presented in the proposed rule and included additional statements related to its worker safety
analysis. Specifically, FSIS stated that “[a]lthough FSIS conducted an analysis of injury rates
during the development of the proposed rule, FSIS did not use the analysis to draw conclusions
on worker safety in HIMP or non-HIMP establishments.” Further, FSIS emphasized that while it
“recognizes that working conditions in swine slaughter establishments is an important issue, the
[a]gency does not have the authority to regulate issues related to establishment worker safety.
OSHA is the Federal agency with statutory and regulatory authority to promote workplace safety
and health.” 19




13
   E.O. 12866 (Oct. 4, 1993).
14
   E.O. 13563, § 2(b) (Jan. 18, 2011).
15
   Ibid.
16
   Durbin, The Honorable Richard, et al., Letter to The Honorable Phyllis K. Fong (Mar. 26, 2019).
17
   Ibid.
18
   Fong, The Honorable Phyllis, Letter to The Honorable Richard J. Durbin, et al. (June 21, 2019).
19
   Modernization of Swine Slaughter Inspection, 84 Fed. Reg. 52,300, 52,305 (Oct. 1, 2019) (discussing and
responding to comments received during the rulemaking process).

                                                                INSPECTION REPORT 24801-0001-41              3
Objectives
Our objectives were to determine whether FSIS, with respect to its worker safety analysis section
of the proposed rule: (1) complied with public transparency requirements under E.O. 13563; (2)
made information about its preliminary analysis on worker safety clearly accessible to the public
during the comment period; (3) adhered to the USDA Information Quality Activities Guidelines
in developing the proposed rule; (4) came to a reasonable determination about the reliability of
the OSHA injury data it used for the proposed rule; and (5) consulted with OSHA and the
National Institute for Occupational Safety and Health (NIOSH) about the impact of the proposed
rule on workplace safety and health.

While addressed in this inspection report, we did not make any formal recommendations related
to Objectives 1, 4, and 5.

As further described in the Scope and Methodology section of this report, we conducted this
inspection in accordance with the Council of the Inspectors General on Integrity and Efficiency’s
(CIGIE) Quality Standards for Inspection and Evaluation.20




20
     CIGIE, Quality Standards for Inspection and Evaluation (Jan. 2012).

4        INSPECTION REPORT 24801-0001-41
Section 1: Transparency and Accessibility of Worker Safety
Information in FSIS’ Proposed Rule—Modernization of Swine
Slaughter Inspection
Objective 1: Did FSIS, with respect to its worker safety analysis section of the
proposed rule, comply with public transparency requirements under E.O. 13563?
We concluded that FSIS’ discussion of the scientific and technical findings of its worker safety
analysis in the proposed rule generally abided by the requirements of E.O. 13563. 21

E.O. 13563, Improving Regulation and Regulatory Review, requires agencies to provide for both
proposed and final rules, to the extent feasible and permitted by law, timely online access to the
rulemaking docket on regulations.gov, including relevant scientific and technical findings. 22, 23
E.O. 13563 also requires agencies to ensure the objectivity of any scientific and technological
information and processes used to support the agency’s regulatory actions, consistent with the
President’s March 9, 2009, memorandum regarding scientific integrity and the Office of Science
and Technology Policy’s (OSTP) Scientific Integrity guidance memorandum. 24 The OSTP
Scientific Integrity Memo states, “[a]gencies should communicate scientific and technological
findings by including a clear explication of underlying assumptions; accurate contextualization
of uncertainties; and a description of the probabilities associated with both optimistic and
pessimistic projections, including best-case and worst-case scenarios where appropriate.” 25

In the proposed rule, FSIS stated that it recognized that evaluation of the effects of line speed on
food safety should include the effects of line speed on establishment employee safety, and it
compared in-establishment injury rates between HIMP and traditional establishments. 26, 27 FSIS
included a brief description of this analysis and a summary of its findings in the proposed rule.
Specifically, the proposed rule stated, “[t]he preliminary analysis shows that HIMP
establishments had lower mean injury rates than non-HIMP establishments.” 28 It further
specified, “[t]he results showed HIMP [establishments] had a lower mean number of injuries
using three OSHA injury rate measures.” 29

However, the agency did not provide enough detail to be consistent with the guidance in the
OSTP Scientific Integrity Memo. For example, FSIS performed statistical tests on the OSHA

21
   For the purpose of this report, our use of “public transparency requirements” noted in Objective 1 refers to the
public participation and science requirements under E.O. 13563.
22
   For the purposes of this engagement, we reviewed FSIS’ scientific analysis and findings based on OSHA
collected in-establishment injury and illness rates in HIMP and traditional swine slaughter facilities.
23
   E.O. 13563, § 2 (Jan. 18, 2011).
24
   E.O. 13563, § 5 (Jan. 18, 2011).
25
   OSTP, Scientific Integrity, Memorandum for the Heads of Executive Departments and Agencies § I.4. (Dec. 17,
2010) (“OSTP Scientific Integrity Memo”).
26
   HIMP establishments operate at higher line speeds than traditional (non-HIMP) establishments.
27
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
28
   Ibid.
29
   The three OSHA injury and illness rate measures were: (1) total case rate; (2) days away, restricted, and transfer;
and (3) days away from work.

                                                                  INSPECTION REPORT 24801-0001-41                   5
data that concluded HIMP establishments had lower mean injury rates than traditional
establishments, but did not disclose either which tests it performed to reach its conclusion or the
assumptions involved in those tests. 30 Therefore, we determined that FSIS did not include a
clear explanation of underlying assumptions, as recommended by the OSTP Scientific Integrity
Memo.

Further, FSIS disclosed in the proposed rule that factors other than line speed may affect injury
rates, such as automation and number of sorters per line. However, the agency did not disclose
other known limitations of the data it analyzed, thus hindering the accurate contextualization of
uncertainties as recommended in the OSTP Scientific Integrity Memo. 31

As neither E.O. 13563 nor the OSTP Scientific Integrity Memo explicitly require agencies to
include a clear explanation of underlying assumptions of data used or an accurate
contextualization of the uncertainties of findings in their proposed rules, we concluded that FSIS’
discussion of its scientific and technical findings of the worker safety analysis in the proposed
rule abides by the requirements of E.O. 13563. Therefore, we do not make any formal
recommendations; however, FSIS may want to ensure it clearly communicates all elements
outlined in the OSTP Scientific Integrity Memo when discussing scientific and technical findings
in public communication. Full disclosure of this type of information aids the public’s ability to
make informed decisions and promotes transparency and confidence in the rulemaking process.




30
   FSIS performed two statistical tests using OSHA data: a t-test and a Levene’s test. According to the agency, a
Levene’s test was performed to test the assumption of equality of variances between data sets and a t-test for
equality of means was performed to test for statistical significance.
31
   See Objective 3 for further details.

6     INSPECTION REPORT 24801-0001-41
Objective 2: Did FSIS, with respect to its worker safety analysis section of the
proposed rule, make information about its preliminary analysis on worker safety
clearly accessible to the public during the comment period?

We concluded that, while FSIS provided a summary of its preliminary worker safety analysis in
the language of the proposed rule, it did not fully disclose its data sources. This occurred
because FSIS considered the discussion of the worker safety data analysis to be non-essential,
supplementary information for the proposed rule. As a result, clear and complete information
about the data sources used for FSIS’ preliminary worker safety analysis was not easily
accessible to the public during the proposed rule’s comment period.

USDA’s Information Quality Activities Guidelines (Guidelines) stipulate that an agency will
“ensure transparency of the analysis, to the extent possible, consistent with confidentiality
protections, by . . . providing transparent documentation of data sources.” 32

In the proposed rule, we found that FSIS did not fully disclose the data source for the OSHA data
it analyzed. In the proposed rule, FSIS stated that “[t]he analysis uses injury rate data by
occupational injury estimates that are derived from the [Bureau of Labor Statistics’] annual
Survey of Occupational Injuries and Illnesses (http://www.bls.gov/iif/data.htm). The survey
captures data from [OSHA] logs of workplace injuries and illnesses maintained by employers.” 33
This language could be interpreted to indicate that FSIS gathered the information for its worker
safety analysis from the listed website. However, FSIS did not download the information from
the website noted in the proposed rule; the agency acknowledged that the proposed rule did not
include the web address for the OSHA data it analyzed. As such, FSIS should have provided
more transparent documentation of the data source for its worker safety analysis in the proposed
rule.

However, FSIS included this acknowledgement and the web address for the OSHA data it used
in the final rule, published in the Federal Register on October 1, 2019. 34 Because the agency
ultimately provided transparent documentation for this specific data source to the public through
the final rule, we do not make a formal recommendation relating to FSIS providing this specific
documentation to the public. However, to ensure FSIS includes transparent documentation of
data sources in future proposed rules, the agency should update its internal procedures for the
rulemaking process to include a review of proposed rules to ensure compliance with the
Guidelines, including the Guidelines’ data source transparency requirements.

Secondly, we found that FSIS did not clearly identify the 24 traditional establishments it used for
worker safety analysis in the proposed rule. During our fieldwork, FSIS explained that, “while
FSIS did not post the exact data that the [a]gency pulled from its Public Health Information
System (PHIS) to select swine slaughter establishments present in the OSHA data set, the same
information can be found in other formats on FSIS’s [sic] website. Establishment level

32
   USDA Office of the Chief Information Officer (OCIO), Regulatory (in effect since Oct. 1, 2002; last accessed
July 24, 2019), https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-
information/regulatory.
33
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
34
   Modernization of Swine Slaughter Inspection, 84 Fed. Reg. 52,305 (Oct. 1, 2019).

                                                                INSPECTION REPORT 24801-0001-41                   7
production volume information is available at
https://www.fsis.usda.gov/wps/portal/fsis/topics/data-collection-and-reports/data. This [sic] data
would allow interested parties to identify the high-volume establishments.” 35 However, based on
our review of this information, this data set did not explicitly state slaughter volumes. Rather, it
included a “slaughter volume category” that used the number designations 1-5 and the
“establishment size” category. Subsequent to our fieldwork, FSIS updated its data dictionary
defining these fields. However, based on this spreadsheet alone, we could not determine which
24 traditional establishments FSIS included in its analysis. 36 Therefore, we concluded that FSIS
did not ensure the transparency of the analysis by providing transparent documentation of the
data source for its comparison group of traditional establishments.

In the final rule, FSIS stated that it included a list of establishments used in its worker safety
analysis in its Freedom of Information Act (FOIA) Electronic Reading Room. 37 As such, the
agency has now provided transparent documentation for this specific data source and we,
therefore, do not make a formal recommendation related to FSIS communicating the names of
the 24 traditional establishments to the public. However, as noted previously, to ensure FSIS
includes transparent documentation of data sources in future proposed rules, the agency should
update its internal procedures for the rulemaking process to include a review of proposed rules to
ensure compliance with the Guidelines, including review of data source transparency
requirements.

We concluded that FSIS did not fully comply with the Guidelines related to providing
transparent data sources when it published its proposed rule. 38 Specifically, FSIS’ information
about the sources of the data used for its worker safety analysis was not easily accessible to the
public during the comment period. 39 Although FSIS disclosed the web address for the OSHA
data it analyzed in the final rule published on October 1, 2019, and updated its FOIA Electronic
Reading Room to include a list of the establishments used in the worker safety analysis, FSIS
should also update its internal procedures for the rulemaking process to include a review of
proposed rules in order to ensure compliance with data source transparency requirements.



35
   PHIS is a web-based FSIS system that contains establishment names, addresses, and general information
regarding the establishments.
36
   Using the web address FSIS provided during fieldwork, we downloaded FSIS “Establishment Demographic
Data,” available as of July 23, 2019. We filtered the spreadsheet for pork slaughter establishments based on
“slaughter volume category.” Per the data dictionary on FSIS’ website, establishments in this category labeled as 4s
and 5s have slaughter volumes of 100,000 or greater. This information narrowed the universe to 64 establishments.
We recognize that establishment slaughter volumes may have changed since FSIS issued the proposed rule in
February 2018; however, we noted that the 24 establishments could not be easily identified using the information
FSIS provided.
37
   FSIS’ FOIA Electronic Reading Room is a webpage included on the agency’s site that contains documents in the
following four categories: (1) final opinions and orders made in the adjudication of cases; (2) policy statements and
adopted interpretation not published in the Federal Register; (3) administrative staff manuals and instructions that
affect the public; and (4) records that are frequently requested/of interest. USDA FSIS, FOIA Electronic Reading
Room (last accessed, Nov. 5, 2019), https://www.fsis.usda.gov/wps/portal/footer/policies-and-links/freedom-of-
information-act/fsis-electronic-reading-room.
38
   For more information about FSIS’ compliance with the USDA Information Quality Activities Guidelines in
publishing the proposed rule, see Section 2, Objective 3.
39
   The public comment period for this proposed rule was open from Feb. 1, 2018, through May 2, 2018.

8     INSPECTION REPORT 24801-0001-41
Recommendation 1

Update FSIS’ internal procedures for the rulemaking process to include a review of proposed
rules to ensure compliance with USDA’s Information Quality Activities Guidelines, including
data source transparency requirements.

FSIS Response
In its May 21, 2020, response, FSIS stated:

As explained above, FSIS maintains that it complied with Departmental guidance because the
preliminary worker injury analysis was not used as a foundation for the rulemaking.
Nevertheless, FSIS intends to update FSIS Directive 1232.4, Regulations Development and
Clearance, to include additional instructions for FSIS employees who review Federal Register
documents before publication. FSIS will include key points from the Department’s Information
Quality Activities Guidelines.

FSIS provided an estimated completion date of October 30, 2020, for this action.

OIG Position
We accept FSIS’ management decision on this recommendation.




                                                      INSPECTION REPORT 24801-0001-41          9
Section 2: Quality of Worker Safety Information used in FSIS’
Proposed Rule—Modernization of Swine Slaughter Inspection
Objective 3: Did FSIS, with respect to its worker safety analysis section, adhere
to the USDA Information Quality Activities Guidelines in developing the
proposed rule?

We concluded that FSIS did not fully adhere to the Guidelines in developing the proposed rule
and presenting it during the comment period. This occurred because FSIS did not fully satisfy
the data presentation and transparency requirements of the Guidelines. As a result, the public
may not have fully understood the limitations of the data used for the worker safety analysis in
the proposed rule, during the public comment period.

The Guidelines require agencies to “ensure that the information they disseminate is substantively
accurate, reliable, and unbiased and presented in an accurate, clear, complete, and unbiased
manner.” 40 Additionally, the Guidelines stipulate that agencies must evaluate the data and
ensure their transparency. 41

In the proposed rule, FSIS stated that the new rule would revoke maximum line speeds at
participating swine slaughter establishments. FSIS also recognized that evaluation of the effects
of line speed on food safety should include the effects of line speed on establishment employee
safety. 42 In this context, FSIS performed a worker safety analysis using injury and illness data
obtained from the OSHA website. 43 FSIS stated that it compared in-establishment injury
incidence rates between 5 HIMP and 24 traditional establishments, from CYs 2002 to 2010, in
the proposed rule. Based on its analysis of the data, FSIS stated that HIMP establishments had
lower mean injury rates than traditional establishments. 44

When we reviewed FSIS’ information related to its worker safety analysis, we found that FSIS
did not fully comply with the data presentation and data transparency requirements of the
Guidelines, as described below:

        Data Presentation Requirements

        The Guidelines require agencies to “evaluate data quality and, where practicable, validate
        the data against other information when using or combining data from different



40
   USDA OCIO, Information Quality Activities, General Requirements (in effect since Oct. 1, 2002; last accessed
May 23, 2019), https://www.ocio.usda.gov/policy-directives-records-forms/information-quality-activities.
41
   USDA OCIO, Regulatory (in effect since Oct. 1, 2002; last accessed July 24, 2019),
https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-information/regulatory.
42
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
43
   FSIS downloaded the OSHA injury and illness data from the following website: OSHA, Establishment Specific
Injury & Illness Data (OSHA Data Initiative) (last accessed July 23, 2019),
https://www.osha.gov/pls/odi/establishment_search.html.
44
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).

10      INSPECTION REPORT 24801-0001-41
        sources.” 45 Additionally, the Guidelines state that agencies must ensure that the data are
        “presented in an accurate, clear, complete, and unbiased manner.” 46 We noted that FSIS
        was not able to verify the injury and illness rates for accuracy and reliability against other
        information because, according to an FSIS official, the agency used the only reasonably
        obtainable information concerning the consequences of the intended regulation related to
        worker safety. However, FSIS checked the OSHA data for completeness and
        applicability of the included establishments by verifying OSHA establishment
        information against independent data from PHIS.

        The proposed rule stated, “FSIS compared in-establishment injury rates between HIMP
        and traditional establishments from [CYs] 2002 to 2010.” 47 However, when we reviewed
        the universe of data FSIS used in its worker safety analysis, we found that it additionally
        included CY 2011 results for 5 of 24 traditional establishments, which were outside of its
        stated scope of CYs 2002 to 2010. 48 FSIS informed us that the statement in the proposed
        rule related to the time period reviewed was a typographical error. Therefore, while FSIS
        took steps to verify OSHA’s data for the establishments, we determined that FSIS did not
        ensure the data period in the proposed rule was presented in an accurate manner.
        Because of this error, we concluded that FSIS did not fully adhere to the data presentation
        requirements of the Guidelines.

        FSIS’ final rule was published in the Federal Register on October 1, 2019. 49 However,
        FSIS did not disclose the actual period reviewed for its worker safety analysis. For
        transparency, FSIS should communicate to the public the actual review period for its
        analysis.

        Data Transparency Requirements

        The Guidelines require agencies to, “[w]hen using the best available data obtained from
        or provided by third parties, ensure transparency in its dissemination by identifying
        known sources of error and limitations in the data.” 50

        We found that FSIS disclosed the following limitations in the data: (1) the data were
        voluntarily submitted; (2) only 9 percent of all market hog slaughter establishments
        submitted illness and injury rate data to OSHA; and (3) factors other than line speed may



45
   USDA OCIO, Regulatory (in effect since Oct. 1, 2002; last accessed July 24, 2019),
https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-information/regulatory.
46
   USDA OCIO, Information Quality Activities, General Requirements (in effect since Oct. 1, 2002; last accessed
May 23, 2019), https://www.ocio.usda.gov/policy-directives-records-forms/information-quality-activities.
47
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
48
   FSIS documentation did not include evidence of CY 2011 data for the other 19 establishments. As discussed
under this Objective’s Data Transparency Requirements subsection, the OSHA data FSIS used for all
29 establishments assessed in its worker safety analysis were incomplete, and it did not include annual injury and
illness rates for all establishments for the 10 years between CYs 2002 and 2011.
49
   Modernization of Swine Slaughter Inspection, 84 Fed. Reg. 52,300 (Oct. 1, 2019).
50
   USDA OCIO, Regulatory (in effect since Oct. 1, 2002; last accessed July 24, 2019),
https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-information/regulatory.

                                                               INSPECTION REPORT 24801-0001-41                  11
         affect injury rates (for example, automation and number of sorters per line). 51 However,
         FSIS did not disclose other known limitations. For example, FSIS did not disclose that
         OSHA’s website advised against using the injury and illness data to make any general
         conclusions. 52, 53

         In another example, FSIS did not disclose that the OSHA data used for the
         29 establishments assessed in its worker safety analysis were incomplete, and it did not
         include annual injury and illness rates for all establishments for the 10 years between
         2002 and 2011. For instance, for 3 of the 24 traditional establishments, injury and illness
         rates were available for only 1 of the 10 years. 54 Additionally, three of the five HIMP
         establishments had injury and illness rates for only 5 of the 10 years. Only 1 of the
         29 total establishments had data for all 10 years.

         During our discussions with FSIS officials, one official stated that worker safety was
         considered supplementary to the proposed rule and did not relate to the foundation of the
         proposed rule as worker safety is outside of FSIS’ jurisdiction. In addition, FSIS officials
         stated that the agency did not use the analysis to draw conclusions on worker safety in
         either HIMP or traditional establishments.

We determined that FSIS did not fully adhere to the Guidelines in developing the proposed rule.
Specifically, FSIS neither ensured that the data in the proposed rule were presented in an
accurate manner nor disclosed all known limitations of the data. FSIS should: (1) communicate
to the public the actual review period and known limitations of the OSHA data associated with
FSIS’ worker safety analysis; and (2) determine the impact of the omissions from the proposed
rule. FSIS should also update its internal procedures for the rulemaking process to include a
review of proposed rules in order to ensure compliance with the Guidelines’ data presentation
and data transparency requirements. However, we believe Recommendation 1 in this report
sufficiently addresses this specific recommendation and, therefore, we did not make a related
recommendation under this objective.


Recommendation 2

Communicate to the public the actual review period associated with FSIS’ analysis.




51
   Modernization of Swine Slaughter Inspection, 83 Fed. Reg. 4,796 (Feb. 1, 2018).
52
   U.S. Department of Labor, Establishment Specific Injury and Illness Data (i.e., OSHA Data Initiative) (last
accessed July 23, 2019), https://www.osha.gov/pls/odi/establishment_search.html.
53
   Specifically, the website states, “[f]or each data collection cycle, OSHA only collects data from a small portion of
all private sector establishments in the United States (80,000 out of 7.5 million total establishments). Therefore,
these data are not representative of all businesses and general conclusions pertaining to all US business should not
be drawn.”
54
   For one of the three establishments, the OSHA data showed 0 percent for the covered period.

12      INSPECTION REPORT 24801-0001-41
FSIS Response
In its May 21, 2020, response, FSIS stated:

OIG takes issue with the sentence in the proposed rule that states, “FSIS compared in-
establishment injury rates between HIMP and traditional establishments from 2002 to 2010” (83
FR 4780, 4796), because the Agency also looked at data from 2011. FSIS has explained several
times that the “2010” is a minor typographical error and did not affect the conclusions of the
analysis. Further, even with the minor typographical error, the sentence in the proposed rule is
factually correct. Regardless of what time span is utilized – 2002-2010 or 2002-2011 – both
show that HIMP establishments had lower mean injury rates than non-HIMP establishments.
Again, the typographical error did not affect the conclusions of the analysis, so while the dates
did contain an error, it had no bearing on the outcome, discussion or understanding of the
document.

As FSIS has explained to OIG multiple times, this recommendation has already been addressed
with the publication of the final rule “Modernization of Swine Slaughter Inspection” (84 FR
52300). In the final rule (84 FR 52300, 52305), FSIS included a link to its Electronic Freedom of
Information Act (FOIA) Reading Room, which contains documents that show FSIS’ full analysis
of worker injury data. FSIS believes that this response is sufficient because the preliminary
worker injury analysis was not used as a foundation for the rulemaking.

OIG Position
We do not accept FSIS’ management decision for this recommendation. While the linked
documents in FSIS’ Electronic Freedom of Information Act (FOIA) Reading Room contain 2011
data for some individual establishments, the documents also contain the erroneous statement that
the scope of FSIS’ worker safety analysis was from 2002 to 2010. 55 To reach management
decision, FSIS needs to communicate to the public the actual review period associated with its
analysis.


Recommendation 3

Communicate to the public the known limitations of the OSHA data used for FSIS’ analysis.

FSIS Response
In its May 21, 2020, response, FSIS stated:

Similar to the response above, FSIS believes that this recommendation has already been
addressed with the publication of the final rule “Modernization of Swine Slaughter Inspection”

55
  USDA FSIS, FOIA Electronic Reading Room: Records Frequently Requested/Of Interest (last accessed, May 27,
2020), https://www.fsis.usda.gov/wps/portal/footer/policies-and-links/freedom-of-information-act/fsis-electronic-
reading-room/ct_index3. See linked documents 18-194 (p 40); 18-277 (p 30); and 18-213-c (p 3).

                                                              INSPECTION REPORT 24801-0001-41                 13
(84 FR 52300). In the final rule (84 FR 52300, 52305), FSIS included a link to its Electronic
FOIA Reading Room, which contains documents that show FSIS’ full analysis of worker injury
data. FSIS believes that this response is sufficient because the preliminary worker injury analysis
was not used as a foundation for the rulemaking.

Additionally, FSIS explained to OIG that when the USDA guideline recommends verifying
third-party data, it is not referring to data from other Federal agencies. Federal agencies generally
accept data from other Federal agencies. FSIS does not have the authority to contact
establishments to independently verify OSHA’s worker injury data and doing so would place an
unnecessary information collection burden on industry.

OIG Position
We do not accept FSIS’ management decision for this recommendation. The linked documents
in FSIS’ Electronic Freedom of Information Act (FOIA) Reading Room disclose additional
known limitations that were not included in the language of the proposed or final rule. 56
However, we determined that FSIS was aware of other limitations but did not disclose them. To
reach management decision, FSIS needs to communicate all known limitations to the public.
Specifically, FSIS still needs to communicate: (1) the data the agency used in its analysis of the
29 establishments did not include injury and illness rates for all establishments for each of the
10 years, and (2) the data used did not differentiate whether injuries/illnesses occurred on the
swine slaughter line or elsewhere within the establishment.


Recommendation 4

Determine the impact of: (1) publishing an inaccurate review period related to the worker safety
analysis in the proposed rule; and (2) not disclosing all known limitations related to the data used
for the worker safety analysis in the proposed rule.

FSIS Response
In its May 21, 2020, response, FSIS stated:

There is no impact related to the preliminary analysis of worker injury data because, as FSIS has
explained multiple times, the preliminary analysis was not used to support the proposed rule. The
preliminary analysis was only included as part of a larger request for comments. Many
commenters suggested that FSIS should not use the data to inform decisions on worker safety,
and FSIS clarified in the final rule that it did not use the data as a foundation for the rulemaking
(see 84 FR 52300, 52305).

FSIS completed this action on October 1, 2019.

56
  USDA FSIS, FOIA Electronic Reading Room: Records Frequently Requested/Of Interest (last accessed, May 27,
2020), https://www.fsis.usda.gov/wps/portal/footer/policies-and-links/freedom-of-information-act/fsis-electronic-
reading-room/ct_index3. See linked documents 18-194 (p 41); 18-277 (p 31); and 18-213-c (p 4).

14      INSPECTION REPORT 24801-0001-41
OIG Position
We accept FSIS’ management decision on this recommendation.




                                                 INSPECTION REPORT 24801-0001-41   15
Objective 4: Did FSIS come to a reasonable determination about the reliability
of the OSHA injury data it used for the proposed rule?
While FSIS’ rationale for using OSHA’s data to perform a worker safety analysis of HIMP and
traditional establishments appeared to be reasonable, we determined that the steps FSIS took to
determine the reliability of OSHA’s data prior to use were not reasonable. FSIS:
(1) underestimated the ultimate importance of these data; (2) did not assess the strength or
weakness of any corroborating evidence to verify their reliability; and (3) underestimated the
anticipated level of risk in using these data. As a result, FSIS’ may have used data that were not
suitable for its worker safety analysis.

The Guidelines require agencies to ensure they disseminate objective information in conjunction
with their rulemaking activities and state that agencies and offices will “ [u]se reasonably reliable
and reasonably timely data and information.” 57 For our purposes, data reliability refers to the
accuracy and completeness of OSHA’s computer-processed data, given their use in FSIS’
proposed rulemaking. 58 In order to address this objective, we reviewed FSIS’ actions to assess
the reliability of OSHA’s data. The goal of a data reliability assessment is to determine whether
the data can be used to answer the research questions. The extent of an agency’s data reliability
assessment can depend on three factors: (1) the expected importance of the data; (2) the strength
or weakness of any corroborating evidence; and (3) the anticipated level of risk in using the
data. 59 Based on these three factors, we determined that FSIS did not come to a reasonable
determination about the reliability of the OSHA injury data.

         Expected Importance of the Data

         The extent to which an agency is to assess the reliability of data it uses depends, in part,
         on its perception of the expected importance of the data. 60 FSIS considered the
         discussion of the worker safety data analysis to be non-essential, supplementary
         information for the proposed rule. However, the Regulatory section of the Guidelines
         states, “[w]ith respect to influential scientific information disseminated by USDA
         regarding analysis of risks to human health, safety, and the environment, USDA agencies
         and offices will ensure, to the extent practicable, the objectivity of this information.” 61, 62
         Since FSIS’ worker safety analysis dealt with risks to human safety, we believe that it

57
   USDA OCIO, Regulatory (in effect since Oct. 1, 2002; last accessed July 24, 2019),
https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-information/regulatory.
58
   Government Accountability Office (GAO), Assessing the Reliability of Computer Processed Data, GAO-09-
680G, at 4 (July 2009). While the Guidelines require agencies to use reasonably reliable data, they do not explicitly
define how to determine data are reasonably reliable. We chose to use this GAO document as further criteria as we
believe it describes a practical approach to data reliability assessments.
59
   Ibid. at 9.
60
   Ibid.
61
   USDA OCIO, Regulatory (in effect since Oct. 1, 2002; last accessed July 24, 2019),
https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-information/regulatory.
62
   The Guidelines define objectivity as “substantively accurate, reliable, and unbiased and presented in an accurate,
clear, complete, and unbiased manner.” USDA OCIO, Information Quality Activities, General Requirements (in
effect since Oct. 1, 2002; last accessed May 23, 2019), https://www.ocio.usda.gov/policy-directives-records-
forms/information-quality-activities.

16      INSPECTION REPORT 24801-0001-41
        qualifies as influential scientific information. As such, we believe the expected
        importance of the data goes beyond non-essential, supplementary information.

        Strength or Weakness of Any Corroborating Evidence

        The extent to which an agency assesses the reliability of data it uses depends, in part, on
        its determination of the strength or weakness of corroborating evidence. When we asked
        why FSIS selected OSHA’s data to perform its worker safety analysis and how the
        agency verified the accuracy of these data, an FSIS official stated that OSHA’s data were
        from another Federal agency and were the only data available related to worker’s safety.
        This same official also stated that, although FSIS did not perform any tests to verify the
        accuracy of OSHA’s data, FSIS matched the establishments’ information in OSHA’s data
        to establishment information in PHIS in order to verify that OSHA’s data could be used
        for the analysis. 63 FSIS officials further explained that the agency did not use its analysis
        to draw conclusions on worker safety in HIMP or non-HIMP establishments, and that
        establishment worker safety was outside of FSIS’ jurisdiction. Based on our review of
        documentation and discussions with officials, we determined that FSIS did not compare
        the OSHA data to any corroborating evidence to verify the reliability of these data used.

        Anticipated Level of Risk in Using the Data

        The extent to which an agency assesses the reliability of data it uses depends, in part, on
        its anticipated level of risk in using the data. Agencies conduct risk assessments to weigh
        how the use of such data could have substantial negative consequences on decisions by
        policymakers and others. FSIS was aware of OSHA’s noted data limitations and
        statement that its data should not be used to draw general conclusions, but the agency
        ultimately chose to use the data for its worker safety analysis. As noted before, FSIS
        considered the discussion of the worker safety data analysis to be non-essential,
        supplementary information for the proposed rule. However, when doing a risk
        assessment, one must consider whether the data are relevant to a sensitive or
        controversial subject. 64 FSIS chose to include this analysis because a similar proposed
        rule related to poultry slaughter received worker safety-related comments and requests. 65
        This decision indicates FSIS was aware that this was a sensitive or controversial subject
        and, therefore, could anticipate that use of the data might involve increased risk.

When we discussed FSIS’ determination of data reliability with its officials, we found that FSIS
decided to use OSHA’s occupational injury and illness data to conduct a worker safety analysis
and report its results in the proposed rule because the agency anticipated comments about the
proposed rule’s impact on worker safety. According to an FSIS official, the agency included the
results of its worker safety analysis in the proposed rule because a previous, similar proposed

63
   We reviewed documentation related to FSIS’ data verification activities and found that FSIS compared the names
and addresses of establishments in OSHA’s data with the agency’s data in order to select the relevant swine
slaughter establishments for its analysis.
64
   GAO, Assessing the Reliability of Computer Processed Data, GAO-09-680G (July 2009).
65
   FSIS referred to its Modernization of Poultry Slaughter Inspection final rule, published Aug. 21, 2014. 79 Fed.
Reg. 49,565.

                                                              INSPECTION REPORT 24801-0001-41                 17
rule related to poultry slaughter received worker safety-related comments and requests. 66 The
official maintained that the analysis was intended to stimulate public comment, but also
maintained that the changes to the swine regulations were not contingent on the results of the
worker safety analysis.

In summary, we acknowledge FSIS’ position that OSHA’s data were the only available data at
that time and that FSIS does not have the authority to regulate issues related to establishment
worker safety. However, we determined that FSIS did not adequately assess the reliability of
OSHA’s data as FSIS: (1) underestimated the ultimate importance of these data; (2) did not
verify the reliability of the data with corroborating evidence; and (3) underestimated the
anticipated level of risk in using these data. We believe Recommendation 1 in this report
sufficiently address any recommendations we would make regarding this issue; therefore, we did
not make any recommendations under this objective.




66
     Ibid.

18           INSPECTION REPORT 24801-0001-41
Objective 5: Did FSIS consult with OSHA and NIOSH about the impact of the
proposed rule on workplace safety and health?
We concluded that FSIS, to the extent required, communicated to OSHA and NIOSH about the
impact of the proposed rule on workplace safety and health.

In a March 26, 2019, letter to the USDA’s Inspector General, Congressional members referenced
E.O. 12866, Regulatory Planning and Review, and asked OIG to determine whether the agency:
(1) consulted with OSHA and NIOSH about the impact of the proposed rule on workplace safety
and health; (2) provided OSHA or NIOSH with its preliminary worker safety analysis; and
(3) addressed any concerns raised by OSHA or NIOSH related to its worker safety information. 67

Based on our review of E.O. 12866, we identified no provisions that explicitly require FSIS to
consult with other Federal agencies regarding FSIS’ proposed regulations. 68 However, we
determined that at least three of the E.O.’s provisions and other guidance strongly imply that
FSIS should consult with other Federal agencies. 69 For example, the Office of Management and
Budget (OMB) Circular A-4 states, “[a]s you design, execute, and write your regulatory analysis,
you should seek out the opinions of those who will be affected by the regulation as well as the
views of those individuals and organizations who may not be affected but have special
knowledge or insight into the regulatory issues. Consultation can be useful in ensuring that your
analysis addresses all of the relevant issues and that you have access to all pertinent data.” 70

When we asked FSIS officials about their communications with OSHA and NIOSH, we found
that FSIS held a meeting with OSHA representatives to discuss the proposed rule before its
publication. 71 However, FSIS did not provide OSHA or NIOSH with details of its preliminary
worker safety analysis. We also found that FSIS provided these agencies an advance copy of the
proposed rule for commenting. Both agencies provided feedback, but neither agency had
questions specific to the results FSIS’ worker safety analysis. 72 FSIS considered the comments
from the agencies and incorporated changes into the proposed rule before its publication.

We concluded that FSIS’ communication with OSHA and NIOSH aligned with the requirements
of E.O. 12866, which does not explicitly require consultation with Federal agencies. Therefore,
we did not make a formal recommendation. However, FSIS may want to consider sharing
relevant analysis with organizations (including Federal agencies) that have specific knowledge or
insight regarding future proposed regulations.


67
   In relation to E.O. 12866, the Congressional members’ letter to USDA OIG stated, “[E.O.] 12866 requires
agencies to consult with other agencies that have significant interests in proposed regulations.”
68
   E.O. 12866.
69
   E.O. 12866, § 1(b)(2), (10), (11); see also OMB, Regulatory Analysis, Circular A-4, at 3 (Sep. 17, 2003)
(recommending consultation with organizations that have special knowledge or insight into the regulatory issues).
70
   OMB Circular A-4, at 3.
71
   FSIS did not meet with NIOSH prior to publication.
72
   For example, OSHA requested that FSIS include requests for comments, best practices, and other measures that
establishments can take to protect workers throughout the establishment. NIOSH suggested that, before FSIS would
allow implementation of a line speed waiver, the agency and involved establishments consider results and
recommendations from an evaluation of the effects of line speed waivers on employee health.

                                                             INSPECTION REPORT 24801-0001-41                 19
Scope and Methodology
We conducted an inspection of the FSIS rulemaking process for the proposed rule on the
modernization of swine slaughter inspection. We initiated this inspection in response to a formal
request from 16 members of Congress. In their letter to OIG, Congressional members requested
that OIG review the proposed rulemaking process “to better understand whether worker safety
data was accurately analyzed and appropriately shared with the public.” 73 Our inspection scope
primarily covered FSIS’ actions related to its worker safety analysis and the Modernization of
Swine Slaughter Inspection proposed rule published in the Federal Register on
February 1, 2018. 74 We conducted our fieldwork from June 2019 through October 2019.

To accomplish our inspection objectives, we:

     •   reviewed the Modernization of Swine Slaughter Inspection proposed and final rules; 75
     •   reviewed E.O.s 12866 and 13563, USDA Information Quality Activities Guidelines, and
         other related guidance; 76
     •   interviewed FSIS officials to discuss the compilation, verification, and analysis of the
         OSHA worker safety data; the process of drafting the proposed rule; the communications
         between FSIS and OSHA and NIOSH; the public responses to the proposed rule; and the
         potential changes to the final rule;
     •   consulted OIG’s Office of Data Sciences regarding FSIS’ use of statistical terminology
         and methods to evaluate the OSHA injury and illness data cited in the worker safety
         analysis section of the proposed rule;
     •   reviewed FSIS-provided supporting documentation for its worker safety analysis,
         including combined injury and illness data obtained from the OSHA website;
     •   reviewed communications between FSIS staff members engaged in the drafting of the
         proposed rule;
     •   reviewed preliminary summaries of the worker safety analysis;
     •   reviewed comments provided by OSHA and NIOSH staff about the worker safety
         analysis; and
     •   reviewed a summary of the meeting between FSIS and OSHA representatives.

We conducted this inspection in accordance with CIGIE’s Quality Standards for Inspection and
Evaluation. These standards require that we plan and perform the inspection to obtain sufficient,
appropriate evidence to provide a reasonable basis for our conclusions and recommendations


73
   Durbin, The Honorable Richard et. al., Letter to The Honorable Phyllis K. Fong (Mar. 26, 2019).
74
   We primarily reviewed documents related to FSIS’ worker safety analysis for the proposed rule that were dated
between June 2015 and May 2018, the month the comment period closed.
75
   Our inspection was initiated to review the worker safety analysis section of the proposed rule. However, during
the inspection, on October 1, 2019, FSIS published the final rule. While addressing our objectives as related to the
proposed rule, we performed a limited review of the final rule in formulating our recommendations.
76
   E.O. 12866; E.O. 13563; President’s 2009 memo on scientific integrity; OSTP Scientific Integrity Memo; OMB
Circular A-4; GAO, Assessing the Reliability of Computer-Processed Data; and USDA OCIO Information Quality
Activities (last accessed July 24, 2019), https://www.ocio.usda.gov/policy-directives-records-forms/information-
quality-activities, https://www.ocio.usda.gov/policy-directives-records-forms/guidelines-quality-
information/regulatory.

20       INSPECTION REPORT 24801-0001-41
based on our review objectives. We believe that the evidence obtained provides a reasonable
basis for our conclusions and recommendations based on our review.




                                                    INSPECTION REPORT 24801-0001-41           21
Abbreviations
CIGIE .................................. Council of the Inspectors General on Integrity and Efficiency
CY ....................................... calendar year
E.O. ..................................... Executive Order
FMIA ................................... Federal Meat Inspection Act
FOIA ................................... Freedom of Information Act
FSIS ..................................... Food Safety and Inspection Service
GAO .................................... Government Accountability Office
HACCP ............................... hazard analysis and critical control point
HIMP................................... HACCP-Based Inspection Models Project
NIOSH................................. National Institute for Occupational Safety and Health
NSIS .................................... New Swine Slaughter Inspection System
OCIO ................................... Office of the Chief Information Officer
OIG...................................... Office of Inspector General
OMB.................................... Office of Management and Budget
OSHA .................................. Occupational Safety and Health Administration
OSTP ................................... Office of Science and Technology Policy
PHIS .................................... Public Health Information System
USDA .................................. Department of Agriculture




22      INSPECTION REPORT 24801-0001-41
FSIS’ Response




                 FSIS’
    RESPONSE TO INSPECTION REPORT




                    INSPECTION REPORT 24801-0001-41   23
Food Safety and
Inspection Service
                     TO:             Gil H. Harden
                                     Assistant Inspector General
1400 Independence                    Office of Inspector General
Avenue, SW,
Washington, D.C.     FROM:           Paul Kiecker             / s / 5 / 21 / 2020
20250                                Administrator
                                     Food Safety and Inspection Service

                     SUBJECT:        OIG Official Draft Report, FSIS Rulemaking Process for the
                                     Proposed Rule: Modernization of Swine Slaughter Inspection,
                                     Inspection Number 24801-0001-41

                     FSIS appreciates the opportunity to review and comment on this Official Draft report
                     concerning the New Swine Slaughter Inspection System (NSIS) rulemaking. The Food
                     Safety and Inspection Service (FSIS) reviewed the Official Draft report and has general
                     comments below followed by a response to each recommendation. FSIS believes that
                     the report’s findings and recommendations are derived from the misapplication of E.O.
                     13563 and the USDA Information Quality Activities Guidelines to an analysis in the
                     NSIS proposal, as well as from a distorted emphasis placed by the auditors on minor
                     errors made in the proposal text. In summary:

                         •   The Occupational Safety and Health Administration (OSHA) and the National
                             Institute for Occupational Safety and Health (NIOSH) are the Federal agencies
                             with jurisdiction over meat and poultry establishment worker safety.
                         •   FSIS presented a preliminary analysis comparison of injury rates between
                             HACCP Inspection Models Project (HIMP) and traditional slaughter
                             establishments to solicit comments for use by OSHA and NIOSH and not as a
                             basis for the NSIS rulemaking.
                         •   E.O. 13563 and the USDA Information Quality Activities Guidelines do not
                             apply to the preliminary analysis; therefore, the audit of FSIS compliance with
                             these documents in developing the preliminary analysis was misguided.
                         •   OIG auditors placed too much weight on minor errors made by FSIS in the
                             NSIS proposal (a typographical error in a table header and the failure to cite a
                             specific website) as a basis for their findings.

                     FSIS General Comments

                     Members of Congress specifically requested that OIG examine whether FSIS complied
                     with E.O. 13563 and the USDA Information Quality Activities Guidelines
                     (“Guidelines”) when it developed the preliminary analysis comparison of injury rates
                     between HIMP and traditional establishments, which was published in the NSIS
                     proposed rule (83 FR 4796). E.O. 13563 and the Guidelines apply to situations where
                     scientific data and analysis is used as support for regulatory action. However, since the
                     preliminary analysis was not a basis for the NSIS rulemaking, E.O. 13563 and the
                     Guidelines do not come into play. OIG incorrectly accepted the assumption that E.O.
                     13563 and the Guidelines apply and did not engage in its own analysis to determine
whether those standards were triggered by the NSIS rulemaking. The fact is that they were not.

As FSIS explained to OIG auditors, the preliminary analysis was not used as a basis for the NSIS
rulemaking. Likewise, it was not used to draw conclusions on worker safety in HIMP or non-
HIMP establishments, nor was it used to determine whether there is an associated impact on food
safety. FSIS published the preliminary analysis to solicit comments about line speed and worker
safety, specifically for use by the OSHA and NIOSH, the Federal agencies with jurisdiction over
worker safety. Notably, immediately following the discussion of the preliminary analysis in the
preamble to the proposed rule, FSIS stated at 83 FR 4796:

   FSIS is requesting comments on the effects of faster line speeds on worker safety.
   Specifically, FSIS is requesting comments on whether line speeds for the
   NSIS should be set at the current regulatory limit of 1,106 hph or some
   other number. The Agency is also interested in comments on the
   availability of records or studies that contain data that OSHA or the
   National Institute for Occupational Safety and Health (NIOSH) may be
   able to use in analyzing the effects of increased line speed on the safety
   and health of employees throughout the establishment, including effects
   prior to and following the evisceration line.

E.O. 13563 and the Guidelines do not apply to the preliminary analysis because the analysis was
published to solicit comments for OSHA and NIOSH and not as a basis for the NSIS rulemaking.
As such, OIG findings that FSIS did not comply with these documents are false because they are
premised on a mistaken assumption about the purpose of the preliminary analysis. FSIS
maintains that it complied with all applicable Executive Orders and Departmental guidance in
developing the NSIS regulations.

Additionally, OIG gave too much weight to two inadvertent and insignificant errors committed by
FSIS. The first error was a typographical one in a table and did not affect the overall outcome of
data therein. The second error was failing to list the website address for OSHA data. OIG cites
these missteps as support for a finding that FSIS was not adequately transparent in presenting the
data sources for the preliminary analysis. As indicated earlier, the preliminary analysis was
published not as a basis for the NSIS rulemaking, but to solicit comments on line speed and
worker safety for use by OSHA and NIOSH. Furthermore, FSIS has publicly clarified the source
of the OSHA data and posted documents that show the full scope of the Agency’s analysis. FSIS
aims to be as transparent as possible in presenting data and data analysis. Any implication that
these minor errors misrepresented the basis for the NSIS rulemaking is false.

Objective 2: Did FSIS, with respect to its worker safety analysis section of the
proposed rule, make information about its preliminary analysis on worker safety
clearly accessible to the public during the comment period?

Recommendation 1
Update FSIS’ internal procedures for the rulemaking process to include a review of proposed
rules to ensure compliance with USDA’s Information Quality Activities Guidelines, including
data source transparency requirements.




                                                2
FSIS Response
As explained above, FSIS maintains that it complied with Departmental guidance because the
preliminary worker injury analysis was not used as a foundation for the rulemaking. Nevertheless,
FSIS intends to update FSIS Directive 1232.4, Regulations Development and Clearance, to
include additional instructions for FSIS employees who review Federal Register documents
before publication. FSIS will include key points from the Department’s Information Quality
Activities Guidelines.

Estimated Completion Date: FSIS estimates that FSIS Directive 1232.4, Regulations
Development and Clearance, will issue by October 30, 2020.

Objective 3: Did FSIS, with respect to its worker safety analysis section, adhere to the
USDA Information Quality Activities Guidelines in developing the proposed rule?

Recommendation 2
Communicate to the public the actual review period associated with FSIS’ analysis.

FSIS Response
OIG takes issue with the sentence in the proposed rule that states, “FSIS compared in-
establishment injury rates between HIMP and traditional establishments from 2002 to 2010” (83
FR 4780, 4796), because the Agency also looked at data from 2011. FSIS has explained several
times that the “2010” is a minor typographical error and did not affect the conclusions of the
analysis. Further, even with the minor typographical error, the sentence in the proposed rule is
factually correct. Regardless of what time span is utilized – 2002-2010 or 2002-2011 – both
show that HIMP establishments had lower mean injury rates than non-HIMP establishments.
Again, the typographical error did not affect the conclusions of the analysis, so while the dates
did contain an error, it had no bearing on the outcome, discussion or understanding of the
document.

As FSIS has explained to OIG multiple times, this recommendation has already been addressed
with the publication of the final rule “Modernization of Swine Slaughter Inspection” (84 FR
52300). In the final rule (84 FR 52300, 52305), FSIS included a link to its Electronic Freedom of
Information Act (FOIA) Reading Room, which contains documents that show FSIS’ full analysis
of worker injury data. FSIS believes that this response is sufficient because the preliminary
worker injury analysis was not used as a foundation for the rulemaking.

Estimated Completion Date: Completed October 1, 2019.

Recommendation 3
Communicate to the public the known limitations of the OSHA data used for FSIS’ analysis.

FSIS Response:
Similar to the response above, FSIS believes that this recommendation has already been
addressed with the publication of the final rule “Modernization of Swine Slaughter Inspection”
(84 FR 52300). In the final rule (84 FR 52300, 52305), FSIS included a link to its Electronic
FOIA Reading Room, which contains documents that show FSIS’ full analysis of worker injury
data. FSIS believes that this response is sufficient because the preliminary worker injury analysis
was not used as a foundation for the rulemaking.




                                                 3
Additionally, FSIS explained to OIG that when the USDA guideline recommends verifying third-
party data, it is not referring to data from other Federal agencies. Federal agencies generally
accept data from other Federal agencies. FSIS does not have the authority to contact
establishments to independently verify OSHA’s worker injury data and doing so would place an
unnecessary information collection burden on industry.

Estimated Completion Date: Completed October 1, 2019.

Recommendation 4
Determine the impact of: (1) publishing an inaccurate review period related to the worker safety
analysis in the proposed rule; and (2) not disclosing all known limitations related to the data used
for the worker safety analysis in the proposed rule.

FSIS Response:
There is no impact related to the preliminary analysis of worker injury data because, as FSIS has
explained multiple times, the preliminary analysis was not used to support the proposed rule. The
preliminary analysis was only included as part of a larger request for comments. Many
commenters suggested that FSIS should not use the data to inform decisions on worker safety,
and FSIS clarified in the final rule that it did not use the data as a foundation for the rulemaking
(see 84 FR 52300, 52305).

Estimated Completion Date: Completed October 1, 2019.




                                                 4
Learn more about USDA OIG
Visit our website: www.usda.gov/oig
Follow us on Twitter: @OIGUSDA

Report Suspected Wrongdoing in USDA Programs

OIG Hotline: www.usda.gov/oig/hotline.htm

Local / Washington, D.C. (202) 690-1622
Outside D.C. (800) 424-9121
TTY (Call Collect) (202) 690-1202

Bribery / Assault
(202) 720-7257 (24 hours)




In accordance with Federal civil rights law and U.S. Department of Agriculture             Relay Service at (800) 877-8339. Additionally, program information may be made
(USDA) civil rights regulations and policies, USDA, its Agencies, offices,                 available in languages other than English.
employees, and institutions participating in or administering USDA programs
are prohibited from discriminating based on race, color, national origin, religion,        To file a program discrimination complaint, complete the USDA Program
sex, gender identity (including gender expression), sexual orientation, disability,        Discrimination Complaint Form, AD-3027, found online at How to File a Program
age, marital status, family/parental status, income derived from a public                  Discrimination Complaint and at any USDA office or write a letter addressed to
assistance program, political beliefs, or reprisal or retaliation for prior civil rights   USDA and provide in the letter all of the information requested in the form. To
activity, in any program or activity conducted or funded by USDA (not all bases            request a copy of the complaint form, call (866) 632-9992. Submit your completed
apply to all programs). Remedies and complaint filing deadlines vary by                    form or letter to USDA by: (1) mail: U.S. Department of Agriculture, Office of the
program or incident.                                                                       Assistant Secretary for Civil Rights, 1400 Independence Avenue, SW, Washington,
                                                                                           D.C. 20250-9410; (2) fax: (202) 690-7442; or (3) email: program.intake@usda.gov.
Persons with disabilities who require alternative means of communication for
program information (e.g., braille, large print, audiotape, American Sign                  USDA is an equal opportunity provider, employer, and lender.
Language, etc.) should contact the responsible Agency or USDA’s TARGET
Center at (202) 720-2600 (voice and TTY) or contact USDA through the Federal               All photographs on the front and back covers are from USDA’s Flickr site and are in
                                                                                           the public domain. They do not depict any particular audit or investigation.