Best Practices and Lessons Learned From the Administration of Pandemic Related Unemployment Benefits Programs
- Document type
- Memorandum
- Date
- 2022-02-16
Summary
A memorandum from the Pandemic Response Accountability Committee (PRAC) dated February 16, 2022, transmitting a MITRE report, Best Practices and Lessons Learned from the Administration of Pandemic-Related Unemployment Benefits Programs. The memorandum states that the PRAC engaged MITRE in May 2021 and that an estimated $872 billion has been allocated to unemployment insurance benefit programs. MITRE contacted all 54 state and territory workforce agencies and 12 responded, and the report summarizes those responses. It organizes observations into standard operating procedures, eligibility phase controls, public communications and risk management, and lists key considerations such as standardizing payments to claimants with unverified identities and requiring timely financial institution compliance with fraudulent payment recovery. Appendices cover methodology, study sources and acronyms.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Full text
MEMORANDUM
February 16, 2022
MITRE Report:
Best Practices and Lessons Learned from the Administration of
Pandemic-Related Unemployment Benefits Programs
The Pandemic Response Accountability Committee (PRAC) is charged with conducting oversight of
pandemic-related spending to prevent and detect fraud, waste, abuse, and mismanagement. In May
2021, we engaged MITRE, a not-for-profit federally funded research and development center, to
conduct an independent study of lessons learned from the administration of pandemic-related
emergency funding for unemployment insurance (UI) benefit programs in a sample of states. An
estimated $872 billion has been allocated to UI benefit programs to address the economic downturn
created by the COVID-19 pandemic. The objective of this study was to increase understanding of how
states implemented pandemic UI benefit programs and how their different implementation
approaches may have reduced the fraud risk, including identity theft-related fraud. MITRE, in
coordination with the U.S. Department of Labor (DOL), contacted all 54 state and territory workforce
agencies (SWAs) and 12 responded. This report summarizes the responses of those 12 SWAs except
where otherwise indicated.
The use of the term “best practices” recognizes those instances where responding SWAs’ actions
may have reduced fraud risk and these practices are spotlighted so they can be replicated and
expanded elsewhere; use of the term is not intended to suggest that the DOL or the SWAs have or
have not done enough to combat UI fraud and improper payments. The DOL Office of Inspector
General (OIG) has raised significant concerns regarding the DOL and SWAs’ ability to deploy UI
benefits expeditiously and efficiently while ensuring integrity and adequate oversight, particularly
during the pandemic and in response to national emergencies and disasters. As the DOL OIG
reported, improper payment rates in the UI program have historically been among the highest in the
federal government. Moreover, the unprecedented infusion of federal pandemic UI funds provided
individuals and organized criminal groups a high-value target to exploit. Despite DOL’s efforts in
issuing new guidance, distributing additional antifraud funding, and providing technical assistance,
improper payments stemming from fraudulent activity continue to pose a significant threat to the
integrity of the nation’s UI program. The DOL OIG has a large body of audit and investigative work on
this topic, with more oversight projects on-going. This work is posted on the DOL OIG pandemic
response website. Moreover, MITRE’s independent observations and suggestions should be read in
context with the PRAC’s December 2021 report, Key Insights: State Pandemic Unemployment
Insurance Programs, which summarized the findings of state auditors overseeing UI benefits.
BEST PRACTICES AND LESSONS LEARNED
FROM THE ADMINISTRATION OF PANDEMIC
RELATED UNEMPLOYMENT BENEFITS
PROGRAMS
© 2022 THE MITRE CORPORATION. ALL RIGHTS RESERVED
Lessons Learned from Pandemic-Related Unemployment Benefits Programs
Record of Changes
A = Add
Change
Date Reference M = Modify Change Description
Number
D = Delete
1.0 01/31/22 Initial Publication - -
Lessons Learned from Pandemic-Related Unemployment Benefits Programs
EXECUTIVE SUMMARY
The Pandemic Response Accountability Committee, a committee within the Council of the
Inspectors General on Integrity and Efficiency tasked by Congress to promote transparency and
conduct oversight of the pandemic response related to the Coronavirus Aid, Relief, and
Economic Security Act, the Continued Assistance for Unemployed Workers Act, and the
American Rescue Plan Act, seeks to identify best practices and lessons learned for minimizing
fraud risk from the implementation of pandemic unemployment insurance (UI) benefits
programs.
Over the course of the pandemic, state workforce agencies (SWAs) have worked to minimize UI
fraud while providing timely benefits to claimants, employing a range of fraud prevention and
deterrence methods. Produced by MITRE, a not-for-profit organization and operator of federally
funded research and development centers authorized by Federal Acquisition Regulation 35.017,
this report showcases interviewed SWAs’ UI fraud prevention practices and lessons learned and
offers key considerations for potentially overcoming some of the challenges to UI program
administration that emerged during the pandemic. It draws on federal and state UI program
implementation documentation and reporting as well as interviews with federal and state UI
stakeholders. This work is intended to inform stakeholders in Congress, the executive branch,
state and local governments, and the public about the status quo and potential of state UI fraud
prevention strategies and tactics.
MITRE corresponded with all 54 states and territories to elicit their responses. A limited number
of states responded, and this report summarizes their narratives. MITRE recognizes the
Department of Labor (DOL) Office of Inspector General (OIG) has repeatedly reported
significant concerns with DOL and SWAs’ ability to deploy UI program benefits expeditiously
and efficiently while ensuring integrity and adequate oversight, particularly during the pandemic
and in response to national emergencies and disasters. MITRE’s observations and key
considerations are based on the analysis of the SWA responses.
Observations – State Best Practices for Preventing UI Fraud During the Pandemic
Drawing on a fraud prevention framework produced by the Organisation for Economic Co-
operation and Development, MITRE organized the observations gathered from federal and state
UI documents and interviews into the following categories:
• Standard Operating Procedures
• Eligibility Phase Controls
• Public Communications
• Risk Management
Standard Operating Procedures
Interstate and interagency coordination and exchange of information about emerging UI fraud
schemes and technical practices to prevent, detect, and deter fraudsters took on heightened
importance as fraud tactics evolved over the course of the pandemic. The National Association
of State Workforce Agencies UI Integrity Center provided open lines of communication and data
resources for states.
iii
Lessons Learned from Pandemic-Related Unemployment Benefits Programs
Coordination between SWAs, federal and state law enforcement, OIG investigators, and states’
attorneys general was critical for investigating and prosecuting fraud and recovering fraudulent
payments deposited with financial institutions.
SWAs utilized federal and state emergency grant funds to hire and train new staff. Furthermore,
some SWAs invested emergency funds into building out their technology and data analytics
expertise for fraud prevention and fraudulent payment recovery.
Eligibility Phase Controls
SWAs implemented a wide array of eligibility phase controls to identify and freeze fraudulent UI
claims.
Upfront identity verification tools were critical to stopping fraudsters in their tracks. While
identity verification tools are not foolproof, SWAs that implemented them claimed to have seen
a reduction in UI fraud associated with identity theft. Additionally, multifactor authentication
and bot prevention and detection technologies augmented identity verification tools in freezing
fraudulent claims.
SWAs pursued different approaches to conditionally paying claimants with unverified identities.
Each approach is a technique to mitigate the losses from unverified claims continuing to be paid.
Some SWAs iteratively updated the fraud indicators and filters used in their fraud prevention
data analytics and cross-matching. Fraud indicators look for commonalities among data points
across multiple claims and multiple data environments.
Public Communications
Much of the public communication conducted during the pandemic by the interviewed SWAs
was less focused on deterring fraud through forceful messaging than it was on educating the
public about the risks of fraud and identity theft and instructing the public on how to effectively
communicate with the SWA.
Some SWAs increased their phone line capacity to accommodate higher call volumes, stood up
contact centers, and made their websites easier for UI claimants to navigate in order to find key
information.
Risk Management
SWAs did not specifically mention use of risk management frameworks aside from risk-based
scoring matrices used to quantify the relative fraud risk associated with a particular claim.
However, they did note cultural shifts and tactical innovations that could help them identify and
reduce fraud risk.
Tactical innovations for risk management include novel vendor engagement and embrace of
artificial intelligence and machine learning (AI/ML) for fraud discovery. One SWA director was
particularly adamant about the use of AI/ML for fraud discovery, arguing that effective use of
automation is paramount for 21st century UI fraud prevention.
Considerations for Further Exploration
Synthesized from review of the observations above, the following key considerations are
opportunities to couple with existing Department of Labor Employment and Training
iv
Lessons Learned from Pandemic-Related Unemployment Benefits Programs
Administration or Office of Unemployment Insurance Modernization strategic initiatives to
deliver transformational impact:
• Standardize policies and practices for administering payments to claimants self-certifying
unemployment and claimants with unverified identities.
• Establish a recommended baseline for risk tolerance for UI fraud associated with self-
certification and unverified identities, both under normal circumstances and during severe
emergencies.
• Develop and standardize conditional payment options to mitigate fraud losses when
administering emergency UI programs where claimant information and identity
verification is difficult.
• Develop and conduct regular UI demand surge stress tests to prepare for future
emergencies.
• Require timely financial institution compliance with fraudulent payment recovery.
• Support SWAs in determining eligibility and making informed eligibility decisions
through requirements that collect additional detail, such as enhanced wage records.
• Develop UI fraud prevention (pre-award) performance measures.
• Explore requirements for consistent data usage and claim adjudication risk assessment
protocols to prevent fraud.
v
Lessons Learned from Pandemic-Related Unemployment Benefits Programs
Table of Contents
EXECUTIVE SUMMARY .....................................................................................................III
1. BACKGROUND OF UNEMPLOYMENT INSURANCE DURING THE PANDEMIC....1
2. PURPOSE ..............................................................................................................................4
2.1. Scope .........................................................................................................................4
3. OBSERVATIONS – STATE BEST PRACTICES FOR PREVENTING UI FRAUD
DURING THE PANDEMIC ................................................................................................5
3.1. Focus Areas for Fraud Prevention and Deterrence ...................................................5
3.2. Observations by Focus Area .....................................................................................6
4. KEY CONSIDERATIONS FOR FURTHER EXPLORATION.........................................12
4.1. Standardize Policies and Practices for Administering Payments to Claimants
Self-Certifying Unemployment and Claimants with Unverified Identities ...............13
4.2. Require Timely Financial Institution Compliance with Fraudulent Payment
Recovery ....................................................................................................................15
4.3. DOL Needs to Help SWAs Determine Eligibility and Make More Informed
Eligibility Decisions Through Requirements That Collect Additional Detail ..........17
4.4. DOL Could Develop UI Fraud Prevention (Pre-Award) Performance Measures...18
4.5. DOL, in Collaboration with SWAs, Can Explore Requirements for Consistent
Data Usage and Claim Adjudication Risk Assessment Protocols to Prevent Fraud .19
APPENDIX A - METHODOLOGY .......................................................................................21
REFERENCES AND SOURCE DOCUMENTATION ..........................................................21
APPENDIX B – ACRONYMNS.............................................................................................26
List of Figures
Figure 1. DOL ETA UI Claims Data .........................................................................................2
Figure 2 Evaluation Methodology ...........................................................................................21
List of Tables
Table 1. Fraud Prevention Key Focus Areas .............................................................................6
Table 2 Study Sources .............................................................................................................22
vi
1. BACKGROUND OF UNEMPLOYMENT INSURANCE DURING THE PANDEMIC
The COVID-19 pandemic has caused an unprecedented negative impact on the well-being of the
American people. Aside from the catastrophic health impacts and the devastating loss of life in
the United States, the economic effects of the pandemic have been among the most challenging
for the U.S. government and citizens. State lockdown mandates and measures to enforce social
distancing have massively disrupted normal business operations, resulting in an unprecedented
spike in unemployment across the country as employers have permanently or temporarily
reduced their labor force. Newly unemployed or furloughed workers have suffered the brunt of
this painful adjustment, and the surging ranks of unemployed Americans have translated into
surging demand for unemployment insurance (UI). U.S. Department of Labor (DOL) data reflect
the immense scale of this demand (see Figure 1).
Jointly administered by the DOL and state workforce agencies (SWAs), “unemployment
insurance programs provide unemployment [cash] benefits to eligible workers who become
unemployed through no fault of their own and meet certain other eligibility requirements.” 1
SWAs administer their own UI programs under state law while adhering to standard federal
guidelines, and they “establish requirements for eligibility, benefit amounts, and the length of
time that benefits can be paid.” 2 Meanwhile, the DOL Employment and Training Administration
(ETA) provides federal UI program direction and oversight. 3
1
U.S. Department of Labor, “How Do I File for Unemployment Insurance?,” Accessed 11 November 2021,
https://www.dol.gov/general/topic/unemployment-insurance.
2
U.S. Department of Labor Office of Inspector General, “DOL-OIG Oversight of the Unemployment Insurance
Program,” 6 October 2021, https://www.oig.dol.gov/doloiguioversightwork.htm.
3
Ibid.
1
Unemployment Insurance Weekly Claims Data
25000000
20000000
15000000
10000000
5000000
0
1/7/1967 1/7/1971 1/7/1975 1/7/1979 1/7/1983 1/7/1987 1/7/1991 1/7/1995 1/7/1999 1/7/2003 1/7/2007 1/7/2011 1/7/2015 1/7/2019
Initial Claims S.A. 4-Week Continued Claims S.A. 4-Week
Figure 1. DOL ETA UI Claims Data 4
The scale of the pandemic UI demand surge is captured in UI weekly claims data recorded by
ETA (see Figure 1). Between March 14, 2020, and April 18, 2020, the seasonally adjusted
(S.A.) 5 four-week average of initial unemployment claims filed across the country skyrocketed
from 225,500 to a peak of 5,301,250. Between March 14, 2020, and May 16, 2020, the
seasonally adjusted four-week average of continued unemployment claims filed across the
country skyrocketed from 1,730,750 to a peak of 21,199,000. This spike dwarfs all other surges
in UI demand since 1967, the earliest year for which ETA provides UI weekly claims data; as
shown in Figure 1, the UI demand surge during the Great Recession between 2007 and 2009
pales in comparison to the COVID-19 pandemic UI demand surge in 2020. 6
Even before the onset of the COVID-19 pandemic and this unprecedented spike in UI demand,
the efficiency and integrity of UI programs nationwide were of serious concern. The DOL Office
of Inspector General (DOL-OIG) notes that “historically the UI program experienced some of
the highest improper payment rates among federal government benefits programs. The reported
improper payment estimate for the regular UI program has been above 10 percent for 14 of the
last 17 years.” 7 Among the areas of concern related to UI raised by DOL-OIG at the outset of the
4
U.S. Department of Labor Employment & Training Administration, “Unemployment Insurance Weekly Claims
Data,” 10 November 2021, https://oui.doleta.gov/unemploy/claims.asp.
5
Seasonal adjustment is a statistical technique that attempts to measure and remove the influences of predictable
seasonal patterns to reveal how employment and unemployment change from month to month.
6
U.S. Department of Labor Employment & Training Administration, “Unemployment Insurance Weekly Claims
Data,” 10 November 2021, https://oui.doleta.gov/unemploy/claims.asp.
7
U.S. Department of Labor Office of Inspector General, “DOL-OIG Oversight of the Unemployment Insurance
Program,” 6 October 2021, https://www.oig.dol.gov/doloiguioversightwork.htm.
2
pandemic were “state preparedness, initial eligibility determinations, benefit amount, return to
work, improper payment detection and recovery, and program monitoring.” 8
Improper payment detection and recovery stems in part from the federated nature of UI programs
nationwide, which enables fraudsters, both individuals and organized criminal entities, to exploit
enforcement gaps within and across states. The rate and magnitude of the increase in UI claims
filed across the country during the pandemic and the expansion of both UI eligibility and benefits
that were legislated through the Coronavirus Aid, Relief, and Economic Security (CARES) Act,
the Continued Assistance for Unemployed Workers (Continued Assistance) Act, and the
American Rescue Plan (ARP) Act placed unprecedented administrative strain on SWAs and
created openings for heightened fraudulent activity. Considering this, DOL-OIG estimates that
“UI program improper payments, including fraudulent payments, will be higher than 10
percent.” 9 Considering the hundreds of billions of dollars of federal funding allocated to the UI
program, this could translate into tens of billions of dollars of improper, including fraudulent,
payments.
More specifically, the CARES Act established three new UI programs — Pandemic
Unemployment Assistance (PUA), Pandemic Emergency Unemployment Compensation
(PEUC), and Federal Pandemic Unemployment Compensation (FPUC) — whose scale and
implementation structures were exploited by fraudsters in unprecedented ways. PUA “extended
UI benefits to individuals who were not traditionally eligible for UI benefits until December 31,
2020,” including “self-employed workers, independent contractors, those with limited work
history, and others.” 10 PEUC “provided up to an additional 13 weeks of unemployment
compensation to individuals who had exhausted their regular unemployment benefits until
December 31, 2020.” 11 Lastly, FPUC “provided a supplemental payment of $600 per week to
individuals receiving traditional and non-traditional UI benefits until July 31, 2020.” 12 Despite
the intention of these congressional efforts to strengthen the social safety net through wider UI
coverage and more generous benefits, the massive infusion of funds into the UI program and
corresponding ambiguity about the implementation of the new CARES Act programs made the
UI programs extremely lucrative targets for fraudsters.
DOL-OIG has documented the primary sources of improper UI payments, including fraud, in a
series of audit reports since April 2020, identifying the following high-risk areas for UI fraud:
“individuals with social security numbers filed in multiple states,” “individuals with social
security numbers of deceased persons and federal inmates,” and “individuals with social security
8
U.S. Department of Labor Office of Inspector General, Pandemic Response Oversight Plan, 27 April 2021,
https://www.oig.dol.gov/public/oaprojects/DOL_OIG_Updated_Pandemic_Response_Oversight_Plan.pdf.
9
U.S. Department of Labor Office of Inspector General, “DOL-OIG Oversight of the Unemployment Insurance
Program,” 6 October 2021, https://www.oig.dol.gov/doloiguioversightwork.htm.
10
U.S. Department of Labor Office of Inspector General, COVID-19: States Struggled to Implement CARES Act
Unemployment Insurance Programs, 28 May 2021, https://www.oig.dol.gov/public/reports/oa/2021/19-21-004-03-
315.pdf.
11
Ibid.
12
Ibid.
3
numbers used to file for UI claims with suspicious email accounts.” 13 DOL-OIG also found that
“state reliance on self-certifications alone to ensure eligibility for the PUA program” made state
UI programs particularly susceptible to fraud. 14
Delivering unemployment benefits to claimants in a timely manner without compromising
payment integrity is critical. Throughout the pandemic, SWAs have had to prioritize getting
benefits to claimants while minimizing fraud. To accomplish this, they have employed a
collection of techniques and fraud prevention and deterrence methods, including through
collaboration with external partners and organizations and cultivation of new relationships with
public and private entities, to meet their obligations to safeguard UI trust fund dollars while
mitigating the disastrous economic impact of the pandemic for legitimate claimants.
2. PURPOSE
The purpose of this report is to identify best practices and lessons learned for minimizing fraud
risk during the implementation of pandemic UI benefits programs. Because of the scope and
scale of the estimated UI fraud during the pandemic, the Pandemic Response Accountability
Committee (PRAC), a committee within the Council of the Inspectors General on Integrity and
Efficiency tasked by Congress to promote transparency and conduct oversight of the pandemic
response related to the CARES Act, the Continued Assistance Act, and the ARP Act, contracted
with The MITRE Corporation (MITRE) to conduct this work. As a federally funded research and
development center operator authorized by Federal Acquisition Regulation 35.017, MITRE
provides unbiased and conflict-free advice, guidance, and technical subject matter expertise to
government sponsors. To develop this report, MITRE’s analysis included evaluation of federal
and state approaches, policies, technology, and processes for implementing pandemic UI benefits
programs. The analysis was conducted through review of federal and state documentation and
reporting on UI program implementation as well as interviews with federal and state UI
stakeholders.
3. SCOPE
MITRE was tasked to provide technical expertise, assessment, and guidance to support the
PRAC’s strategic goal of improving the transparency of pandemic-related funding, focusing in
this report on UI fraud prevention. This report includes:
A description of states’ approaches to implementing pandemic unemployment programs
•
and their efforts to prevent UI fraud
• Key considerations related to proposed alternatives for program implementation to
minimize fraud risk
With this analysis, the PRAC can summarize lessons learned from implementation of pandemic
UI programs to date and share best state practices for UI fraud prevention. This work is intended
13
U.S. Department of Labor Office of Inspector General, Pandemic Response Oversight Plan, 27 April 2021,
https://www.oig.dol.gov/public/oaprojects/DOL_OIG_Updated_Pandemic_Response_Oversight_Plan.pdf.
14
Ibid.
4
to inform stakeholders in Congress, the executive branch, state and local governments, and the
public about the status quo and potential of state UI fraud prevention strategies and tactics.
4. OBSERVATIONS – STATE BEST PRACTICES FOR PREVENTING UI FRAUD
DURING THE PANDEMIC
MITRE engaged with 12 state UI agencies 15 to determine how they successfully addressed the
unprecedented challenges of UI program administration during the COVID-19 pandemic.
Furthermore, MITRE sought information about the vulnerabilities and gaps exposed during the
pandemic that the SWAs hope to remedy moving forward. MITRE synthesized the observations
below from interviews, responses to MITRE’s request for information, and reports from 12
SWAs. The Appendix A environmental scan references and source documentation list the
specific SWAs engaged for this research.
5. FOCUS AREAS FOR FRAUD PREVENTION AND DETERRENCE
The scope of this report is to showcase successful practices and lessons learned by SWAs to
prevent fraud — the intentional deception to obtain benefits. During the pandemic, most of the
misappropriated distribution of claimant benefits was performed by external entities. This type of
fraud differs from internal fraud committed by staff. The report also does not include
unintentional errors by claimants in applying for, or SWAs in processing, claims for UI benefits.
Furthermore, this report does not address waste — the thoughtless or careless expenditure,
mismanagement, or abuse of resources. 16
MITRE leveraged global industry best practices on fraud detection, including an international
report – Countering Fraud in Social Benefit Programmes: Taking Stock of Current Measures
and Future Directions 17 – produced by the Organisation for Economic Co-operation and
Development (OECD). The OECD established guidance on steps to prevent and deter external
fraud based on lessons learned from the administration of social benefits programs across
numerous nation states.
The observations from interviews with SWA stakeholders are categorized and grouped below
within the fraud prevention key focus areas laid out in the OECD report.
15
This report articulates the summarized narratives expressed by the SWAs that participated in this project. MITRE
corresponded with all 54 states and territories to elicit their responses. A limited number of states responded, and
this report summarizes their narratives. Reference to SWA activities does not necessarily mean all states performed
the activities, but rather a number of states did. MITRE recognizes DOL-OIG has repeatedly reported significant
concerns with DOL and SWAs’ ability to deploy program benefits expeditiously and efficiently while ensuring
integrity and adequate oversight, particularly during the pandemic. MITRE’s observations and key considerations
are based on analysis of the SWA responses.
16
Information regarding payment accuracy and improper payments as well as the effective stewardship of taxpayer
funds is a critical responsibility of the Federal Government. Payment accuracy focuses on the prevention and
recovery of improper payments while ensuring the right individuals and communities benefit from federal funds;
https://www.paymentaccuracy.gov/.
17
OECD, Countering Fraud in Social Benefit Programmes: Taking Stock of Current Measures and Future
Directions, OECD Publishing, Paris, 2020, https://doi.org/10.1787/71df2657-en.
5
Table 1. Fraud Prevention Key Focus Areas 18
Key Fraud Prevention Focus Area Description and Application
Standard operating procedures – This includes public organizations incorporating fraud prevention into
strengthening strategies, goals, and their strategies, objectives, and procedures, and making sure that they
objectives for combating fraud strike the right balance between prevention, detection, and prosecution
measures. Fraud prevention techniques are essentially integrated into
standard operating procedures.
Eligibility phase controls – targeting To minimize the chance of fraudulent tactics succeeding at this high-
prevention measures at the registration risk phase, governments must ensure that they put in place adequate
phase policies, controls, and measures to verify identities and data submitted
during the registration process, in particular during the registration
phase of unemployment benefits administration.
Public communication – tailoring By integrating behavioral perspectives into their prevention approaches,
communication campaigns and public organizations can develop nuanced communication campaigns
messaging to improve fraud that include a range of messages to deter fraud, for example by
deterrence including soft messages and reminders and outlining the penalties for
committing fraud.
Risk management – focusing on the Risk management and assessments can contribute to savings and
highest risks promote efficiency by targeting the application of preventive controls
and identifying areas that are most susceptible to false claimants and
fake registrations; includes risk models and scoring to apply risk
mitigation strategies.
6. OBSERVATIONS BY FOCUS AREA
The OECD recommends a holistic approach to reduce and mitigate external fraud by examining
benefits program strategies within the key focus areas in Table 1. MITRE conducted interviews
with SWAs to glean insights on how states addressed fraud within these key focus areas. The
observations from the SWA fraud prevention narratives describe successful efforts to thwart
fraud as well as opportunities to enhance existing UI resources, processes, and technologies.
6.1.1. Standard operating procedures
MITRE found that interstate and intrastate coordination related to UI programs administration
matured over the course of the pandemic. SWAs from neighboring states or within a particular
region established working groups with regular meetings to exchange information about the
fraudster tactics and techniques they were identifying. The National Association of State
Workforce Agencies (NASWA) UI Integrity Center played a critical role in both connecting and
maintaining open lines of communication for states, and the number of SWAs regularly
interacting with NASWA and utilizing its data resources for UI claim cross-matching increased
considerably over the course of the pandemic. 19 Cross-matching was used to identify fraudulent
Ibid.
18
Employment and Training Administration, Response to the Office of Inspector General Alert Memorandum: The
19
Employment and Training Administration needs to Issue Guidance to Ensure State Workforce Agencies Provide
6
UI claimants or claims leveraging incarceration databases, deceased persons databases, motor
vehicle registration databases, and multi-state cross-match databases.
Within states, coordination between SWAs, federal and state law enforcement and IG
investigators, and states’ attorneys general was critical for investigating and prosecuting fraud
and recovering fraudulent payments from financial institutions. Working with the state attorney
general has been critical for some SWAs in requiring financial institutions to return improper UI
payments to the UI trust fund.
SWAs also utilized federal and state emergency grant funds to hire and train new staff, in
multiple cases hiring hundreds of full-time and temporary employees to fill investigatory and
claims processing roles. Furthermore, much of this hiring, training, and claims processing was
done virtually as SWAs closed their in-person offices to comply with social distancing
guidelines.
For claims processing roles, SWAs transitioned non-UI staff to process claims and experimented
with virtual training programs and training pods, which consisted of a mentor and mentee to
ensure retention of training for new staff. One SWA emphasized how it successfully utilized
virtual training to reduce the time to complete UI training from 6 weeks to 3 weeks. The
successes of virtual hiring and training should be documented and considered for incorporation
into standard training procedures even after the pandemic.
For investigatory roles, SWAs invested emergency funds into building out their technology and
data analytics expertise for fraud prevention and fraudulent payment recovery. One SWA
highlighted how a close relationship between analytics and investigations made for the strongest
fraud detection and prevention regime as new fraud schemes emerged. SWAs hired in-house
technologists and external vendors to fine-tune existing fraud indicator queries and develop new
queries to improve fraud detection. Multiple SWAs established cyber fraud units to complement
the work of traditional benefit payment control units; these cyber fraud units, in addition to
general fraud training units that SWAs stood up, were critical to both executing fraud prevention
and detection activities and providing necessary guidance to wider SWA staff, including claims
processors, as new fraud tactics and events emerged.
To ensure that the technological, investigatory, and analytical expertise that SWAs developed
over the course of the pandemic does not fade, these burgeoning capabilities and the talented
employees who deliver them must be fostered and made permanent to the extent possible.
Multiple SWA leaders expressed concern about their ability to sustain such capabilities due to
the eventual withdrawal of emergency funds (federal and state) and the difficulty of keeping
talented employees (investigators, technologists, and claims processors) on the payroll with
limited budgets and caps on the number of permanent employees that SWAs can hire. Without
more sustainable funding for permanent employees and an increase in the permanent employee
cap, especially during periods of heightened unemployment, SWAs expressed that they would
continue to struggle with persistent staffing shortages.
Requested Unemployment Insurance Data to the Office of Inspector General, 15 June 2021,
https://www.oversight.gov/sites/default/files/oig-reports/DOL/19-21-005-03-315UI-Data-Access-Alert-MemoFinal-
Rpt06162021.pdf.
7
6.1.1.1. IT Modernization
For SWAs that had them, modernized IT systems were critical for both administering pandemic
UI programs efficiently and preventing fraud. In a May 2021 report, DOL-OIG emphasized its
finding that “states with modernized IT systems implemented CARES Act programs
significantly faster than those using antiquated IT systems. The results of our analysis
demonstrate a clear correlation between states’ IT modernization status and the time needed to
implement new PEUC and PUA programs. For example, states that completed IT modernization
started the PEUC program 15 days faster and the PUA program 8 days faster (on average) than
those still planning IT modernization.” 20
SWAs stressed the importance of modernized IT systems in helping them detect and prevent
fraud. One SWA actually launched a fully modernized UI IT system in the middle of the
pandemic (after having received funding for the modernization effort a few years earlier),
providing a unique case study of the impact of IT modernization for pandemic UI fraud
prevention. Before it launched its modernized IT system, this SWA had claims processing staff
manually reviewing and freezing claims suspected to be fraudulent; its legacy IT system could
not freeze suspicious claims in batches through staff direction or automation. The SWA knew
this was a slow process that could not freeze suspicious claims at scale, but it was the best it
could do at the time. With the introduction of its modernized IT system, the SWA froze in bulk
batches all claims identified by its algorithms as suspicious; through this process, 150,000
suspicious claims were frozen on the first night of this modernized system’s operations.
Other SWAs noted that modern IT and advanced analytics enabled them to build multilayer
fraud defense tactics, which will be detailed further in the next section. These were critical for
conducting the claim crossmatches needed to flag duplicative or potentially fraudulent claim
information. These systems contrasted significantly with antiquated SWA IT systems described
in a DOL-OIG audit, which, among other deficiencies, “did not have the mainframe capacity to
perform cross-matches for such a large volume of claims” and “did not include [improper
payment] detection and recovery functionality.” 21
Over the past 6 months, the importance of IT modernization has further been made evident by
the DOL OUIM priority to develop “IT solutions to modernize antiquated state technology by
centrally developing open, modular technology solutions that can be adopted by states as
needed.” 22 Through federal funding and hands-on technology co-development with SWAs, DOL
OUIM “hopes to provide software to support end-to-end administration of UI, including benefit
delivery, employer tools and appeals.” 23
20
U.S. Department of Labor Office of Inspector General, COVID-19: States Struggled to Implement CARES Act
Unemployment Insurance Programs, 28 May 2021, https://www.oig.dol.gov/public/reports/oa/2021/19-21-004-03-
315.pdf.
21
Ibid.
U.S. Department of Labor Employment and Training Administration, Fact Sheet: Unemployment Insurance
22
Modernization – American Rescue Plan Act Funding for Timely, Accurate and Equitable Payment in Unemployment
Compensation Programs, 11 August 2021, https://oui.doleta.gov/unemploy/pdf/FactSheet_UImodernization.pdf.
23
Ibid.
8
6.1.2. Eligibility phase controls
SWAs implemented new techniques and technologies and enhanced existing techniques and
technologies to tighten their eligibility phase controls. Given the sensitivity of SWA tactics and
operations to prevent and detect fraud, there were limits to the tactical information that SWAs
were willing to disclose. Despite this, MITRE was able to glean and can present higher-level
observations on approaches and concepts that SWAs employed to detect, prevent, and deter UI
fraud during the pandemic.
Upfront identity verification tools were lauded by multiple SWAs as critical to stopping
fraudsters in their tracks. SWAs explained that there are many ways for fraudsters to either
generate convincing personally identifiable information (PII) or steal actual PII, which can then
be used to file for UI. Identity verification tools require proof of identity before an individual can
move forward with any type of registration or claim process, thereby ensuring that the individual
presenting the PII is indeed who they say they are. In the case of the PUA program, individual
claimants had to prove their identity before receiving payment.
While identity verification tools are not foolproof, SWAs that implemented them saw a reduction
in UI fraud associated with identity theft. One SWA emphasized that the incorporation of
identity verification tools was associated with a significant decrease in the number of daily PUA
claims and initial UI filings, which may suggest that knowledge of that SWA’s identity
verification requirement deterred would-be fraudsters from submitting initial unemployment
claims. In summary, there are multiple identity verification technologies that SWAs can
incorporate into their fraud prevention and detection tactics. The key for SWAs was to determine
how best to fit identity verification into their existing IT architecture in a way that would enable
them to achieve their desired balance between payment timeliness and integrity.
SWAs pursued different approaches to conditional payouts to claimants with unverified
identities, even as federal and state executive orders pushed them to prioritize payment
timeliness over payment integrity, especially for PUA. One state simply did not pay claimants
whose identities could not be verified. Another state established a minimum payment amount for
claimants with unverified identities until their identities could be confirmed. Yet another state
established a “kill switch” (stop payment) mechanism that halted conditional payouts to
claimants with unverified identities if there was an unexpected spike in UI claims that could
indicate a fraud attack; the purpose of this “kill switch” is to ensure that a fraud event does not
result in a massive continuous payout of conditional pay.
SWAs also implemented a range of other techniques to build a multilayer fraud defense.
Multifactor authentication augmented identity verification tools in weeding out fraudsters filing
claims with fake or stolen PII. Bot prevention and detection technologies were used to establish
web application firewalls that, among other functions, were able to geo-block Internet Protocol
addresses that came from areas known to generate a high degree of fraudulent claim activity.
Additionally, SWAs would establish average claim submission baselines; any significant
deviation above this average would flag to the SWA that it might be facing a potential fraud
event. With the surge of legitimate claims early during the pandemic, significant deviations
above the average did not always indicate fraudulent activity; however, as the pandemic
progressed and SWAs were better able to handle the influx of claims, utilizing such baselines
and deviations was effective for noting abnormal activity.
9
SWAs also iteratively updated the fraud indicators and filters used in their data analytics and
cross-matching; fraud indicators are used to search across multiple claims and multiple data
environments for commonalities (duplicate information) and fraudulent information that would
flag a claim as illegitimate. This enabled them to expand the range of indicators that would mark
a UI claim as suspicious as they encountered new fraud schemes and tactics over the course of
the pandemic. One SWA noted that it started the pandemic with less than 10 standard fraud
indicators and now has incorporated close to 60 indicators into its standard cross-matching
analytics. Finally, some SWAs established risk-based scoring matrices to quantify the relative
fraud risk associated with a particular claim; certain fraud indicators would weigh more heavily
in the scoring scheme, enabling the SWAs to prioritize immediate review of the riskiest claims.
On the other hand, this risk quantification was also used in some states to drive the speedy
release and payment of relatively low risk claims that had been frozen.
Finally, timely and accessible data to query and crossmatch against was critical for SWAs in
iterating on the fraud indicators incorporated into their analytics and using those analytics to
detect and prevent fraud. NASWA’s UI Integrity Center and its data tools and resources,
including the Integrity Data Hub (IDH), played a pivotal role in enabling multi-state cross-
matching efforts. One SWA explained it exchanged data files with NASWA daily for cross-
matching purposes, and another SWA director remarked that UI fraud during the pandemic
would have been much worse without NASWA serving as a hub for interstate data coordination.
Some SWAs also emphasized that positive relationships with employers, who would provide
more real-time worker records or forward fraud alert notices if an individual still employed by
them filed for UI, were influential in getting reliable data that helped SWAs identify fraudulent
claims.
6.1.3. Public communication
Public communication was a fraught topic across the SWAs that MITRE interviewed. All
interviewees mentioned the tradeoff that exists between transparency with the public and the
need to safeguard against the disclosure of counter-fraud tactics to fraudsters. One SWA
highlighted how the public release of one of its counter-fraud tactics prompted bad actors to
adjust their activities and succeed in evading SWA controls and detection through that particular
tactic. Considering this, much of the public communication conducted by SWAs during the
pandemic was less focused on deterring fraud through forceful messaging than it was on
educating the public about the risks of fraud and identity theft and instructing the public on how
to effectively communicate with the SWA. With that said, some SWAs did note that
communication through law enforcement task forces that publicly messaged about UI fraud
arrests and prosecutions could serve as an effective deterrent.
SWAs were forced to improve their communication pathways with the public because pandemic
stay-at-home and social distancing guidelines resulted in the temporary closure of in-person
centers and offices. Thus, SWAs had to transition from an in-person communication presence to
a largely virtual communication presence. While the communication tactics detailed hereafter
may seem simple, they were critical to enabling a reliable virtual presence to which citizens
could turn for information related to UI.
SWAs increased their phone line capacity to accommodate higher call volumes, stood up contact
centers, and utilized “contact us” web forms and automated message services to record customer
10
information without subjecting customers to long telephone wait times. One SWA emphasized
its efforts to increase outreach to traditionally underserved communities, increasing the number
of documents translated into non-English languages so more customers could self-serve and
better understand and access the programs available.
SWAs also sought to make it easier to find critical information on their websites, standing up
distinct UI fraud and identity theft webpages. These distinct pages provided information on how
to identify and report UI fraud and identity theft as well as recommendations on what to do if an
individual believed they were a victim of identity theft. These distinct fraud pages also provided
links to related law enforcement webpages for reporting identity theft, although one SWA
advised victims of identity theft to report directly to the SWA website because local law
enforcement offices were not capable of dealing with identity theft associated with fraudulent
unemployment insurance claims. Lastly, some SWAs sent mailings to claimants or those who
had reported identity theft (employees and employers) as another medium through which to
provide instructions and recommendations for protecting against and reporting identity theft.
Two isolated comments are worthy of note given the unique perspective they provide. First, one
SWA proactively communicated to claimants about the incorporation of identity proofing
technologies into its claims filing process in order to increase public trust in the UI system as
well as expedite the claims submission and adjudication process by alerting claimants ahead of
time that they should be prepared to verify their identities. Second, one SWA director
emphasized the importance of educating the public about solid cyber hygiene and the creation of
usernames and passwords that are not recycled across online accounts. While this latter point is
not exclusive to PII theft associated with UI fraud, legitimate UI claimants would certainly
benefit from better cyber hygiene.
6.1.4. Risk management
Risk is an event that, if it occurs, adversely affects an organization’s ability to achieve its
objectives. Risk management is a formal and disciplined practice for addressing risk and
reducing it to an acceptable level. It includes identifying risks, assessing their probabilities and
consequences, developing management strategies, and monitoring their state to maintain
situational awareness of changes in potential threats. As such, risk management is critical to
increasing the likelihood of successful program outcomes.
Risk management can be practiced on an individual project, within a specific program, or across
the entire enterprise. The Association for Federal Enterprise Risk Management defines
Enterprise Risk Management (ERM) as “a discipline that addresses the full spectrum of an
organization’s risks, including challenges and opportunities, and integrates them into an
enterprise-wide, strategically-aligned portfolio view. ERM contributes to improved decision
making and supports the achievement of an organization’s mission, goals, and objectives.” 24
SWAs did not specifically mention use of risk management frameworks aside from the risk-
based scoring matrices (described in Section 3.2.2) used to quantify the relative fraud risk
G. Milbourn, Dr. S. Brady, G. Ingber, J. Stehle.,” GOVERNMENT-WIDE PAYMENT INTEGRITY: NEW
24
APPROACHES AND SOLUTIONS NEEDED,” The MITRE Corporation, Mclean, VA, April 2016
11
associated with a particular claim. However, they did note cultural shifts and tactical innovations
that could help them identify and reduce fraud risk.
Representatives from multiple SWAs highlighted the tension between payment integrity and
payment timeliness. They admitted that identity proofing and fraud prevention efforts on the
front end can slow payment delivery to claimants. Modern identity verification technologies can
enable payment integrity without creating claim backlogs as severe as those of the past, but
prioritization of payment integrity and fraud prevention in SWA standard operating procedures
naturally inhibits payment timeliness. Thus, to mitigate fraud risk at scale, one SWA stated, “The
cultural change needs to be that everyone who touches UI should view payment integrity as part
of their responsibility and bake anti-fraud into how we do business.” However, the latitude that a
particular SWA has to prioritize and operationalize payment integrity and risk management
depends on state and local political sensitivities to potentially extended payment timelines.
Tactical innovations for risk management include novel vendor engagement and embrace of
artificial intelligence and machine learning (AI/ML) for fraud discovery. One SWA mentioned
its subscription to proprietary threat intelligence feeds to monitor the dark web for fraudster
chatter about tactics and targets for UI fraud. Another SWA described its efforts to acquire
enhanced (i.e., more timely) wage record data from a private vendor to supplement its databases
for cross-matching and querying to flag and freeze suspicious claims. Lastly, multiple SWAs
raised the potential value of AI/ML technologies that search claims for signs of fraud that fall
outside standard rules-based fraud discovery filters. With commonly used rules-based discovery
systems, SWAs are alerted only when hard-coded fraud indicators are identified in a claim; with
AI/ML platforms, SWAs could potentially learn about previously undiscovered fraud schemes
and tactics straight from patterns in claims data. One SWA director was particularly adamant
about the use of AI/ML for fraud discovery, arguing that effective use of automation is
paramount for 21st century UI fraud prevention. To pursue and sustain these tactical innovations
and fraud prevention technologies, SWAs will require federal and state fraud prevention funding
that is greater in both quantity and duration than the fraud prevention funding they were
receiving before the pandemic.
7. KEY CONSIDERATIONS FOR FURTHER EXPLORATION
The administration of a decentralized, federated UI program requires collaboration and
coordination involving multiple stakeholders, including DOL, SWAs, other federal agencies
(e.g., Internal Revenue Service), and public-private entities. Stakeholder engagement and
partnership is essential to foster alignment, agreement, adoption, and action on any suggestions.
The U.S. Department of Labor recently established the OUIM to provide strategic leadership and
work with state workforce agencies and federal partners to modernize and reform the
unemployment insurance system. The following key considerations do not consider current or
future initiatives by ETA or OUIM and have yet to be prioritized or adopted. These
considerations should be considered as opportunities to couple with existing strategic initiatives
to deliver transformational impact.
12
8. STANDARDIZE POLICIES AND PRACTICES FOR ADMINISTERING PAYMENTS
TO CLAIMANTS SELF-CERTIFYING UNEMPLOYMENT AND CLAIMANTS WITH
UNVERIFIED IDENTITIES
Since the passage of the CARES Act and implementation of the PUA program, ambiguous
federal guidance about how to administer UI payments to claimants self-certifying
unemployment and claimants with unverified identities has hampered SWAs’ ability to prevent
UI fraud. Absent clear federal direction, SWAs adopted varied policies and practices to handle
these payments. To better support SWAs in fighting fraud associated with self-certification and
unverified identities, DOL might consider developing requirements and standards for SWA
administration of these unique and challenging UI payments.
DOL and SWAs cannot write off self-certification as a one-time pandemic-era solution that will
not be relevant in future crises. Instead, DOL and SWAs need to reflect on their experience
administering PUA with self-certification and record the lessons learned to build strategies and
tactics to mitigate fraud associated with self-certification and unverified identities in future UI
demand surges. They can take steps to make administration of these payments standard across
states as well as targeted toward a balance between payment timeliness and payment integrity to
which both DOL and SWAs agree. Key considerations related to this include:
• DOL, in collaboration with SWAs, can determine a benchmark for UI fraud
associated with self-certification and unverified identities that they are willing to
tolerate during emergencies – Effectively balancing fraud prevention with timely
payment provision during UI emergencies requires federal and state fraud tolerance
benchmarks, which can help SWAs calibrate their UI operations and tactics.
• DOL, in collaboration with SWAs, can research, develop, and standardize
conditional payment options to mitigate fraud losses when administering emergency
UI programs under which claim information and identity verification is difficult –
Self-certified unemployment was extremely difficult for SWAs to verify during the
pandemic as states tried to minimize fraud up front during the application process and
comply with Federal guidelines to expedite payments, resulting in extreme losses to UI
fraud. Considering this, it could be beneficial to explore conditional payment options in
instances of unverified claimant information or identity that balance the need for timely
payments against the need to reduce fraud risk.
• DOL ETA, in collaboration with SWAs, can develop and conduct regular UI
demand surge stress tests to prepare for future emergencies – Based on stress test
results, DOL ETA and SWAs can determine emergency response resources, guidance,
and standard operating procedures needed to effectively balance fraud mitigation and
payment timeliness.
8.1.1. DOL, in collaboration with SWAs, can determine a benchmark for UI fraud
associated with self-certification and unverified identities that they are willing to tolerate
during emergencies
Due to executive (federal and state) pressure on SWAs to prioritize payment timeliness over
payment integrity during the early stages of the pandemic, especially after the instantiation of
PUA, states were forced to put an unplanned emphasis on recovery of fraud losses associated
13
with payments to claimants self-certifying unemployment and claimants with unverified
identities. To build resilience into UI systems, DOL, with input from SWAs, can explore
regulating how SWAs operationalize the extreme prioritization of timeliness over payment
integrity, which characterized the administration of pandemic UI programs, so as to mitigate
fraud losses. In determining these operational regulations, unemployment self-certification and
other eligibility schemes under which identity verification would be difficult must be factored in
as potentially expedient methods to deliver payments to claimants during a UI demand crisis.
A benchmark (or bounds) for tolerable levels or rates of UI fraud associated with self-
certification and unverified identities would be needed to underpin these operational regulations
and inform how states operationalize the tradeoff between payment integrity and payment
timeliness. DOL is best positioned to set this “tolerable fraud” benchmark at the federal level,
and it could engage SWAs to receive their input about what is desirable and feasible for the
benchmark. SWAs could then tailor and resource their fraud prevention operations and tactics to
meet the federal benchmark, as opposed to trying unrealistically to prevent all UI fraud
associated with self-certification and unverified identities at the expense of timely payment
delivery.
An initial step toward establishing this benchmark would be to isolate pandemic UI fraud
associated with self-certification and unverified identities from pandemic UI fraud rooted to
other sources or gaps in the system. This would provide national and state-level foundational
statistics on UI fraud associated with self-certification and unverified identities, which DOL and
SWAs could use to begin determining reasonable benchmarks for tolerable UI fraud in
emergency scenarios.
8.1.2. DOL, in collaboration with SWAs, can research, develop, and standardize
conditional payment options to mitigate fraud losses when administering emergency UI
programs under which claim information and identity verification is difficult
If it is envisioned that self-certification, or other UI eligibility schemes that make claim
information and identity verification difficult for SWAs, will be instituted to improve payment
timeliness in future UI crises, DOL, in collaboration with SWAs, might consider researching,
developing, and standardizing conditional payment approaches that minimize fraud losses in
instances of unverified claimant information or identity.
SWAs pursued different approaches to conditionally paying claimants with unverified identities
or information, citing their approaches as critical to tightening post-delivery payment controls
and preventing continued fraud through already-released claims with unverified information.
Some SWAs made conditional payments to claimants with unverified PII up until a deadline, at
which point identity verification was required for payments to continue. Incorporating identity
verification tools into the claim submission process can reduce the number of claimants with
unverified information. However, identity verification tools are not foolproof, so there is still a
need for established conditional payment protocols to deal with instances of unverified claimant
information. Another SWA maintains a “kill switch” (stop payment) mechanism that halts
conditional payments to claimants with unverified PII if there is an unexpected spike in UI
claims that may indicate fraudulent activity; this “kill switch” ensures that a fraud event does not
result in a massive improper conditional payout. Yet another SWA explained that it only paid a
minimum amount per conditional payment until the claimant’s identity was verified.
14
Considering these distinct state practices, DOL, in collaboration with SWAs, can research and
establish a standard set of effective conditional payment options for SWAs to employ if forced to
administer UI programs under which payments are made to claimants with unverified PII. These
options could enforce ceilings for the amount of money and length of time that a claimant with
unverified information will be paid, and they could be applied consistently across states to close
off opportunities for fraudsters to exploit differences in state conditional payment protocols and
practices. Lastly, DOL and SWAs could refine these conditional payment approaches over time
as they identify new fraud schemes and tactics.
8.1.3. DOL ETA, in collaboration with SWAs, can develop and conduct regular UI demand
surge stress tests to prepare for future emergencies
Tactical, operational, and strategic actions need to be taken to prepare for future UI demand
surges during which the prioritization of payment timeliness over payment integrity exacerbates
the potential for severe UI fraud losses across the country. These actions could be determined
and refined through regular stress tests administered by DOL ETA with SWAs.
Stress tests, whether through simulations or exercises, probe systems to determine how robust or
brittle they might be in the face of extreme circumstances. For example, in the aftermath of the
2007-2009 financial crisis, Congress passed legislation—the Dodd-Frank Act—requiring the
Federal Reserve to conduct capital planning stress tests to ensure that the nation’s largest
financial institutions had sufficient capital to weather severe economic downturns without
significantly disrupting financial markets. 25 Stress test methodologies establish the requirements
and performance metrics against which the organization being tested is evaluated.
DOL and Congress could consider regularly stress testing state UI systems to improve fraud
prevention capabilities before the next unemployment crisis. Federal and state legislation
requiring UI system stress testing might be needed to ensure DOL and SWAs adequately
resource and prioritize it. Through lessons learned and weaknesses identified from regular stress
test exercises, modeling, and simulation, DOL and SWAs would have a consistent feedback
mechanism with which to develop fraud prevention resource requirements and standard
operating procedures that help keep fraud within the organization’s risk tolerance level while
maintaining the prioritization of payment timeliness over payment integrity during
unemployment crises.
DOL ETA, in collaboration with SWAs, could begin making UI system stress testing a reality by
establishing UI stress test methodologies, which would evaluate how SWAs’ people, processes,
and technology perform under extreme UI surge circumstances with elevated fraud risk.
9. REQUIRE TIMELY FINANCIAL INSTITUTION COMPLIANCE WITH
FRAUDULENT PAYMENT RECOVERY
Financial institutions are not currently required to return UI funds identified by SWAs or other
state authorities as fraudulent within an established timeline; this challenges the ability of SWAs
to recover those funds. The regulatory authority to establish a required timeline for expedient
25
Board of Governors of the Federal Reserve System, “Stress Tests and Capital Planning,” 5 August 2021,
https://www.federalreserve.gov/supervisionreg/stress-tests-capital-planning.htm.
15
financial institution compliance with fraudulent payment recovery may be outside DOL’s
purview and require federal legislation.
Multiple SWAs raised the recovery of identified fraudulent payments from financial institutions
as a major challenge to increasing improper payment recovery rates during the pandemic. This
challenge is rooted to the fact that there is no federal regulation compelling banks and financial
institutions to expeditiously return identified fraudulent UI payments to state UI trust funds
before those fraudulent payments are withdrawn and no longer recoverable.
Washington state has emerged as a leader in recovering stolen UI funds through its attorney
general’s use of the state’s “asset forfeiture power to recover stolen funds”; Washington is the
first state to utilize this authority to recover stolen funds not yet withdrawn from financial
institutions. 26 The Washington Employment Security Department shared its UI fraud data with
the Attorney General’s Office, supporting the attorney general with case development to
subpoena “35 banks across the country to identify accounts with balances of $1,000 or more that
bore…indicators of fraud.” 27 The Washington Attorney General intends to continue utilizing the
state’s asset forfeiture authority to recover identified stolen UI funds from financial institutions.
Despite Washington’s success, the lack of urgency by financial institutions in returning stolen UI
funds held in their accounts to state UI trust funds poses a significant obstacle to SWAs,
warranting federal regulation that helps SWAs expediently recover stolen UI funds from
financial institutions before they can be withdrawn. DOL did issue Unemployment Insurance
Program Letter (UIPL) No. 19-21 in May 2021 providing “guidance to states on the proportional
distribution methodology for recovering [comingled] federally funded UC benefits, which are
held by banks and financial institutions as a result of suspicious and/or potentially fraudulent
activity.” 28 However, much of the language in that UIPL places the onus of outreach for
fraudulent payment recovery on the SWAs rather than the financial institutions. To improve
fraudulent payment recovery rates, financial institutions need to be required to be more proactive
in returning identified fraudulent UI payments sitting in their accounts before they are withdrawn
by fraudsters.
Representatives from one SWA further explained that they have not received any guidance from
the U.S. Department of Labor during the pandemic about subpoenaing financial institutions and
compelling their compliance with improper payment recovery. Considering this, the Department
of Labor, Department of Justice, and other federal agencies involved in financial institution
oversight could consult with Washington and any other states beginning to use asset forfeiture
26
Washington State Office of the Attorney General, “Attorney General Ferguson uses innovative approach to
recover $495,000 in stolen unemployment benefits,” 25 October 2021, https://www.atg.wa.gov/news/news-
releases/attorney-general-ferguson-uses-innovative-approach-recover-495000-stolen.
27
Ibid.
28
U.S. Department of Labor Employment and Training Administration, Advisory: Unemployment Insurance
Program Letter No. 19-21; Subject: Benefits Held by Banks and Financial Institutions as a Result of Suspicious
and/or Potentially Fraudulent Activity and the Proportional Distribution Methodology Required for
Recovering/Returning Federally Funded Unemployment Compensation (UC) Program Funds, 4 May 2021,
https://wdr.doleta.gov/directives/attach/UIPL/UIPL_19-21.pdf.
16
powers to recover stolen UI funds about the regulatory framework needed to compel timely
financial institution compliance with fraudulent payment recovery.
A working group could be identified to research and define the responsibilities of financial
institutions and SWAs in meeting timely fraudulent payment recovery requirements, the
timelines by which those responsibilities must be fulfilled, and the associated reporting
requirements.
10. DOL NEEDS TO HELP SWAS DETERMINE ELIGIBILITY AND MAKE MORE
INFORMED ELIGIBILITY DECISIONS THROUGH REQUIREMENTS THAT
COLLECT ADDITIONAL DETAIL
DOL has previously considered using enhanced wage records (EWR) for UI program
administration to gain more insight into employment outcomes related to training and education
and state and local area information on employment by occupation, hours of work, and wages,
which can inform employer location and recruitment decisions and enable the production of
occupational wage trends that cannot currently be produced. 29 Benefits of EWR would also help
to address the number and types of fraudulent claims related to claimant employment history. In
addition to challenges with the verification of identities, validating specific employment
attributes—status of employment, occupational codes, hours worked, etc.—is not attainable with
the current data attributes captured by most states; 30 the information is also provided quarterly as
opposed to monthly or even bi-weekly. One SWA explained that access to EWR could have
greatly reduced the number of fraudulent claims because the department would have been able to
confirm claimant employment histories with more recent information. The additional specificity
into a claimant’s work history could be matched against the initial UI claim to accurately
determine currency of employment and type of employment.
As identified in other wage information reports, DOL should consider implementing the
following activities to institute the usage of EWR.
10.1.1. Pilot EWR
DOL ETA has been working with the Workforce Information Advisory Council to identify the
challenges, benefits, and pathways to establish EWR. 31 Recent council recommendations include
advocating for the adoption of EWR by including information on the occupational job title(s),
hours worked, and job site location; engaging the private sector for successful adoption; and
appointing a leader from within the Secretary’s Office or Deputy Secretary’s Office to oversee
this initiative, pulling in the appropriate individuals in the public and private sectors to achieve
29
Workforce Information Advisory Council, Recommendations for the Improvement of the Workforce and Labor
Market Information System, 2018, https://www.dol.gov/sites/dolgov/files/ETA/wioa/pdfs/WIAC-Recommendations-
Report-DRAFT-for-Secretary-v2.pdf.
30
Increased Reporting Frequency: Illinois now requires some employers to report monthly, which, if expanded to
other states, could broaden the benefits, and uses of the wage records.
31
Workforce Information Advisory Council August 2021 summary meeting notes outline several recommendations
for the Secretary of Labor, August 2021,
https://www.dol.gov/sites/dolgov/files/ETA/wioa/pdfs/Minutes_8.31.21.pdf.
17
the recommendations. The changes will require changes to policy, business processes, and
technical workstreams.
The establishment of an EWR pilot could provide a means to test the changes and the impacts
with key stakeholders and identify any additional business process or technology changes
required to provide the level of specificity SWAs require to accurately determine eligibility. A
pilot program to improve the capture of individual-level data to help establish currency of
employment records 32 (i.e., when an employee was employed and duration as well as status,
hours worked, and occupational codes using existing enhancements in information records)
could demonstrate the utility of EWR for deterring fraudulent claims during the registration
phase. The pilot could include the ability to determine federal and state reporting requirements to
be addressed, scenarios to crossmatch employment records history to claimants’ UI registration
information, development of comprehensive status reports, and establishment of a universal data
dictionary. DOL can formalize the requirements to collect the data with more specificity by
reviewing existing regulatory authorities that define data definitions and schemas, obtaining
Paperwork Reduction Act approval as part of digital transformation, and satisfying any other
legislative requirements.
10.1.2. Deploy policy, business process, and technical guidance based on the pilot to
instantiate EWR
Based on the lessons learned from piloting EWR, DOL could determine the implementation
strategy including formalizing the policy recommendations, the business processes, and technical
requirements, including data and technical integration to deploy across the landscape of UI
programs. DOL could also consider establishing a working group to monitor and review the
implementation, identify suggested future EWR enhancements, and incorporate metrics to reveal
EWR’s effectiveness in preventing UI fraud.
11. DOL COULD DEVELOP UI FRAUD PREVENTION (PRE-AWARD)
PERFORMANCE MEASURES
Current DOL performance measures incentivize SWA counter-fraud efforts after claim payout
(i.e., “right of check”). In its agency strategic plans and annual performance reports, DOL tracks
three performance measures related to its strategic objective to “support states’ timely and
accurate benefit payments for unemployed workers”—payment timeliness, detection of
recoverable overpayments, and improper payment rate. As the DOL Fiscal Year 2020 Annual
Performance Report makes clear, decreasing UI improper payments was an agency priority goal
under the Trump administration; more specifically, the goal was for the UI improper payment
rate to be 9% by September 30, 2021. 33 While these measures are important, federal performance
measures that incentivize fraud prevention investments and activities that take place prior to
32
Transformational Pathway Recommendation as specified in https://www.nga.org/futureworkforce/policies/data-
and-
assessment/#:~:text=Indiana%2C%20Illinois%20and%20Missouri%3A%20The%20Coleridge%20Initiative%20Pil
ot,%E2%80%93%20as%20long%20as%20federal%20minimums%20are%20met.
33
U.S. Department of Labor, FY 2020 Annual Performance Report,
https://www.dol.gov/sites/dolgov/files/general/budget/2022/CBJ-2022-V1-01.pdf.
18
claim payout (i.e., “left of check” or pre-award) can be explored. Furthermore, SWAs could be
recognized and rewarded for fraud prevention investments and activities that take place prior to
claim payout.
As MITRE heard from multiple SWAs, SWAs need to undergo a cultural shift from prioritizing
detection and recovery of fraudulent payments to prioritizing prevention of fraudulent payments.
However, this cultural shift cannot happen unless DOL advances agency performance measures
that quantify and reward pre-payment fraud prevention activities. If DOL incorporates fraud
prevention into its evaluation of SWAs’ performance, it may incentivize SWAs to dedicate
staffing and resourcing to fraud prevention like they do with improper payment detection and
recovery. Such federal performance measures for fraud prevention may spur SWAs to
proactively and robustly staff and resource UI fraud prevention investments and activities.
SWAs could work with DOL to categorize and quantify their current fraud prevention
investments and activities and propose achievable, or aspirational, fraud prevention measures
that could be standardized across states. These measures could encourage SWAs to invest in staff
and IT to prevent fraud, and they could capture SWA strides and success in making more
accurate eligibility and payment determinations; additionally, they could lend themselves to
straightforward reporting so that SWAs could simply document the success of their “left of
check” fraud prevention investments and activities.
An increased focus on fraud prevention could reduce payment timeliness by lengthening the
claim review process (although ID proofing technologies may mitigate this). However, this focus
deserves consideration given the extreme levels of fraud during the pandemic. This cultural shift
trickling down to SWAs may be best achieved through codification at the federal level in the
form of DOL UI fraud prevention performance measures.
12. DOL, IN COLLABORATION WITH SWAS, CAN EXPLORE REQUIREMENTS
FOR CONSISTENT DATA USAGE AND CLAIM ADJUDICATION RISK
ASSESSMENT PROTOCOLS TO PREVENT FRAUD
In its audits of UI program administration throughout the pandemic, DOL-OIG has consistently
cited insufficient SWA data transparency and sharing as an impediment to more effective fraud
prevention. DOL has sought to facilitate cross-state information and data sharing by funding the
UI Integrity Center IDH operated by NASWA. Despite this, there is no mechanism ensuring
consistent use of data and claim adjudication protocols across states to prevent fraud.
Considering this, DOL, in collaboration with SWAs, can explore data usage and claim
adjudication risk assessment protocol requirements as a condition for states to receive federal
funding for UI program administration. In doing so, DOL and SWAs would move past
discussion of data and information sharing into concrete and potentially binding steps to align
their counter-fraud technologies and processes.
More specifically, DOL and SWAs could jointly determine the UI datasets required for fraud
checks as well as the standard fraud indicators that SWA investigators and algorithms should be
scanning for in those datasets. With this said, SWAs should obviously have the flexibility to scan
for fraud indicators outside the standard set agreed to with DOL, as they are on the front lines of
fraud prevention and consequently best positioned to identify emerging fraud indicators.
19
However, mandated usage of agreed-upon UI datasets and fraud indicators could be beneficial in
overcoming some of the downsides of our federated UI system.
Finally, DOL, with input from SWAs, can establish standard protocols or guidance for how
SWAs should make a UI claim adjudication decision based on the identification of a particular
fraud indicator or set of fraud indicators; in this way, DOL could begin to bring consistency to
fraud risk scoring and assessments, which are currently done individually by SWAs based on
their internal evaluation of the severity of particular fraud indicators.
Establishing and executing claim adjudication risk assessment protocols are currently the
purview of SWAs, but DOL has already taken steps during the pandemic to bring greater
uniformity to state-administered UI programs. For example, DOL is already incentivizing
standardization of SWA IT and fraud prevention systems by making “identification verification
services available to states to purchase.” 34 Furthermore, to overcome some of the obstacles to
data integration at scale that have enabled fraudsters to exploit cross-state UI enforcement gaps
during the pandemic, DOL has conditioned federal fraud prevention grants on SWA compliance
with federal data and information sharing requirements.
In the Employment and Training Administration’s August 2021 UI Program Letter No. 22-21
notifying SWAs “regarding the availability of up to $140 million to support states with fraud
detection and prevention, including identity verification and overpayment recovery activities, in
all UC (Unemployment Compensation) programs,” grant conditionality is used as a mechanism
to nudge data and information sharing between SWAs and DOL. 35 Specifically, “as a condition
of receiving funding…[states] must agree to provide all confidential UC information to DOL-
OIG for purposes of both investigating fraud and performing audits through weeks of
unemployment ending before December 31, 2023.” 36 Consideration of standardized UI fraud
prevention datasets, indicators, and risk assessment protocols would align with this already
existing federal effort to make SWA fraud prevention investments and activities more consistent.
U.S. Department of Labor Employment and Training Administration, Fact Sheet: Unemployment Insurance
34
Modernization - American Rescue Plan Act Funding for Timely, Accurate and Equitable Payment in Unemployment
Compensation Programs, 11 August 2021, https://oui.doleta.gov/unemploy/pdf/FactSheet_UImodernization.pdf.
35
U.S. Department of Labor Employment and Training Administration, Advisory: Unemployment Insurance
Program Letter No. 22-21; Subject: Grant Opportunity to Support States with Fraud Detection and Prevention,
Including Identity Verification and Overpayment Recovery Activities, in All Unemployment Compensation (UC)
Programs, 11 August 2021, https://wdr.doleta.gov/directives/attach/UIPL/UIPL_22-21.pdf.
36
Ibid.
20
APPENDIX A - METHODOLOGY
Environmental Scan
Overview
This assessment report provides valuable information to key stakeholders, including the public,
Congress, executive branch agencies, and SWAs on best practices and lessons learned from the
implementation of pandemic UI programs. MITRE’s environmental scan leveraged a variety of
information sources to characterize the current landscape, highlight challenges and obstacles, and
identify areas for improvement.
MITRE used a qualitative data-gathering protocol to capture the factors that affected the
execution of state UI programs during the pandemic and the processes, governance, and
technology involved in UI program execution. The protocol provided a systematic approach to
discover and document the challenges, successes, and opportunities raised during MITRE’s
interviews and document review. MITRE conducted the analysis in four steps as shown in the
figure below.
Figure 2 Evaluation Methodology
The analysis process synthesizes information gathered from the stakeholder interviews and
document reviews. Initially, observations are gathered from these data sources. Observations are
specific facts drawn from the document review or interview comments; they are directly
attributable to a particular data source. The team categorizes and collates the observations into
broader findings and themes and uses these to highlight best practices, lessons learned, and
optimal approaches for implementing pandemic UI programs. Findings are key inferences drawn
from the observations. Themes are high-level concepts that recur across multiple findings;
themes group findings to communicate the broadest analytical takeaways.
References and Source Documentation
21
Table 2 Study Sources
Title/Name Author Date Description
Advisory Report – CARES Act: Initial Areas of Concern Department of Labor Apr-20 Report
Regarding Implementation of Unemployment Insurance Office of Inspector
Provisions General (DOL-OIG)
Alert Memorandum: The Pandemic Unemployment DOL-OIG May- Memorandum
Assistance Program Needs Proactive Measures to Detect 20
and Prevent Improper Payments and Fraud
Unemployment Insurance Program Letter No. 23-20: Department of Labor May- Memorandum
Program Integrity for the Unemployment Insurance (UI) Employment and 20
Program and the UI Programs Authorized by the Training
Coronavirus Aid, Relief, and Economic Security Administration (DOL
(CARES) Act of 2020 – Federal Pandemic ETA)
Unemployment Compensation (FPUC), Pandemic
Unemployment Assistance (PUA), and Pandemic
Emergency Unemployment Compensation (PEUC)
Programs
Response to the Office of Inspector General’s (OIG) DOL ETA Jun-20 Memorandum
Alert Memorandum: The Pandemic Unemployment
Assistance Program Needs Proactive Measures to Detect
and Prevent Improper Payments and Fraud
Top Pandemic Challenges Facing the U.S. Department of DOL-OIG Jun-20 Report
Labor
Unemployment Insurance Program Letter No. 25-20: DOL ETA Jun-20 Memorandum
Benefit Accuracy Measurement (BAM) Program
Operations in Response to the Coronavirus Disease of
2019 (COVID-19) Pandemic
Countering Fraud in Social Benefit Programmes: Taking OECD Jul-20 Report
Stock of Current Measures and Future Directions
COVID-19: More Can Be Done to Mitigate Risk to DOL-OIG Aug-20 Report
Unemployment Compensation Under the CARES Act
Unemployment Insurance Program Letter No. 28-20: DOL ETA Aug-20 Memorandum
Addressing Fraud in the Unemployment Insurance (UI)
System and Providing States with Funding to Assist with
Efforts to Prevent and Detect Fraud and Identity Theft
and Recover Fraud Overpayments in the Pandemic
Unemployment Assistance (PUA) and Pandemic
Emergency Unemployment Compensation (PEUC)
Programs
COVID-19: States Cite Vulnerabilities in Detecting DOL-OIG Oct-20 Report
Fraud While Complying with the CARES Act UI
Program Self-Certification Requirement
Unemployment Insurance Program Letter No. 9-21: DOL ETA Dec-20 Memorandum
Continued Assistance for Unemployed Workers Act of
2020 (Continued Assistance Act) – Summary of Key
Unemployment Insurance (UI) Provisions
22
Title/Name Author Date Description
Alert Memorandum: The Employment and Training DOL-OIG Feb-21 Memorandum
Administration (ETA) Needs to Ensure State Workforce
Agencies (SWAs) Implement Effective Unemployment
Insurance Program Fraud Controls for High Risk Areas
Update: Top Challenges in Pandemic Relief and Pandemic Response Feb-21 Report
Response Accountability
Committee (PRAC)
Response to the Office of Inspector General’s Alert DOL ETA Mar-21 Memorandum
Memorandum: Employment and Training Administration
Needs to Ensure State Workforce Agencies Implement
Effective Unemployment Insurance Program Fraud
Controls for High Risk Areas
2021 State of the Workforce Report: Responding to the National Association Mar-21 Report
Pandemic of State Workforce
Agencies (NASWA)
Unemployment Insurance Program Letter No. 14-21: DOL ETA Mar-21 Memorandum
American Rescue Plan Act of 2021 (ARPA) – Key
Unemployment Insurance (UI) Provisions
Pandemic Response Oversight Plan DOL-OIG Apr-21 Report
Unemployment Insurance Program Letter No. 16-21: DOL ETA Apr-21 Memorandum
Identity Verification for Unemployment Insurance (UI)
Claims
COVID-19: States Struggled to Implement CARES Act DOL-OIG May- Report
Unemployment Insurance Programs 21
Unemployment Insurance Program Letter No. 20-21: DOL ETA May- Memorandum
State Instructions for Assessing Fraud Penalties and 21
Processing Overpayment Waivers under the Coronavirus
Aid, Relief, and Economic Security (CARES) Act, as
Amended
Unemployment Insurance Program Letter No. 19-21: DOL ETA May- Memorandum
Benefits Held by Banks and Financial Institutions as a 21
Result of Suspicious and/or Potentially Fraudulent
Activity and the Proportional Distribution Methodology
Required for Recovering/Returning Federally Funded
Unemployment Compensation (UC) Program Funds
Alert Memorandum: The Employment and Training DOL-OIG Jun-21 Memorandum
Administration Needs to Issue Guidance to Ensure State
Workforce Agencies Provide Requested Unemployment
Insurance Data to the Office of Inspector General
Alert Memorandum: The Employment and Training DOL-OIG Jul-21 Memorandum
Administration Does Not Require the National
Association of State Workforce Agencies to Report
Suspected Unemployment Insurance Fraud Data to the
Office of Inspector General or the Employment and
Training Administration
23
Title/Name Author Date Description
Unemployment Insurance Program Letter No. 23-21: DOL ETA Aug-21 Memorandum
Grant Opportunity for Promoting Equitable Access to
Unemployment Compensation (UC) Programs
Unemployment Insurance Program Letter No. 22-21: DOL ETA Aug-21 Memorandum
Grant Opportunity to Support States with Fraud
Detection and Prevention, Including Identity Verification
and Overpayment Recovery Activities, in All
Unemployment Compensation (UC) Programs
Lessons Learned in Oversight of Pandemic Relief Funds Pandemic Response Aug-21 Report
Accountability
Committee (PRAC)
Fact Sheet: Unemployment Insurance Modernization – DOL ETA Aug-21 Memorandum
American Rescue Plan Act Funding for Timely, Accurate
and Equitable Payment in Unemployment Compensation
Programs
Top Management and Performance Challenges Facing DOL-OIG Nov-21 Report
the U.S. Department of Labor
Pandemic Response Accountability Committee (PRAC) - - Interview
Department of Labor Employment and Training - - Interview
Administration (DOL ETA)
Department of Labor Office of Unemployment Insurance - - Interview
Modernization (DOL OUIM)
Department of Labor Office of Inspector General (DOL- - - Interview
OIG)
National Association of State Workforce Agencies - - Interview
(NASWA) UI Integrity Center
Identity Theft Resource Center (ITRC) - - Interview
Nevada Department of Employment, Training and - - Response to
Rehabilitation (DETR) Request for
Information
Oregon Employment Department (OED) - - Response to
Request for
Information
Iowa Workforce Development (IWD) - - Response to
Request for
Information
Georgia Department of Labor - - Interview
Connecticut Department of Labor - - Interview
Washington State Employment Security Department - - Interview
(ESD)
Idaho Department of Labor - - Interview
24
Title/Name Author Date Description
New Jersey Department of Labor and Workforce - - Interview
Development
Michigan Department of Labor and Economic - - Response to
Opportunity (LEO) Request for
Information
Arizona Department of Economic Security - - Report
Colorado Department of Labor and Employment - - Interview
Virgin Islands Department of Labor - - Interview
25
APPENDIX B – ACRONYMNS
Acronym Definition
AI Artificial Intelligence
ARP American Rescue Plan
CARES Coronavirus Aid, Relief, and Economic Security Act of 2020
DOL Department of Labor
ERM Enterprise Risk Management
ETA Employment and Training Administration
EWR Enhanced Wage Records
IDH Integrity Data Hub
MITRE The MITRE Corporation
ML Machine Learning
NASWA National Association of State Workforce Agencies
OECD Organisation for Economic Co-operation and Development
OIG Office of Inspector General
OUIM Office of Unemployment Insurance Modernization
PEUC Pandemic Emergency Unemployment Compensation
PII Personally Identifiable Information
PRAC Pandemic Response Accountability Committee
PUA Pandemic Unemployment Assistance
SWA State Workforce Agency
Treasury U.S. Department of the Treasury
UC Unemployment Compensation
UI Unemployment Insurance
UIPL Unemployment Insurance Program Letter
26
File and source
- File
- 2022-02-16-best-practices-lessons-learned-pandemic-unemployment.pdf
- Size
- 1,470,358 bytes
- SHA-256
- 938430c1f487bc3e274fcce6ba295015b77fb97df1ca420ddabfa0d34d0afb68
- Original
- www.oversight.gov