STATE OF MICHIGAN

GRETCHEN WHITMER DEPARTMENT OF LABOR AND ECONOMIC OPPORTUNITY SUSAN CORBIN

GOVERNOR UNEMPLOYMENT INSURANCE AGENCY DIRECTOR

December 29, 2021

The Honorable Gretchen Whitmer

Governor, State of Michigan
111 S. Capitol Avenue
Lansing, MI 48933

The Honorable Ed McBroom

Chair, Michigan Senate Oversight Committee
201 Townsend Street
Suite #7200
Lansing, MI 48933

The Honorable Steven Johnson

Chair, Michigan House Oversight Committee
House Office Building
Room 1289
Lansing, MI 48933

Dear Governor Whitmer, Senator McBroom, and Representative Johnson:

I am writing to share with you the results of an investigation conducted by Deloitte and Touche
LLP (“Deloitte”) into imposter fraud and intentional misrepresentation payments made by the
Unemployment Insurance Agency (“UIA”). The UIA maintains a consummate commitment to
transparency and engaged Deloitte, an independent qualified third-party accounting firm, in
furtherance of this core value. In releasing the results of this investigation to you and the public,
Michigan sets itself apart across the nation as a leader in both transparency and accountability.
The report distinguishes Michigan from its peers in that it demonstrates how the UIA has
successfully implemented many reforms that have effectively stopped the ongoing efforts to
defraud the UIA and unemployment systems across the country.

According to the analysis conducted by Deloitte and Touche LLP, less than 1% of claims paid
since October 2020 were estimated to be a result of imposter fraud or intentional
LEO is an equal opportunity employer/program.
Auxiliary aids, services and other reasonable accommodations are available upon request to individuals with disabilities.
105 WEST ALLEGAN STREET, LANSING, MICHIGAN 48933  www.michigan.gov/leo  517-241-6712
December 29, 2021
The Honorable Gretchen Whitmer
The Honorable Ed McBroom
The Honorable Steven Johnson
Page 2

misrepresentation. Not only is this below the commonly accepted average, but it is also an
improvement upon UIA’s pre-pandemic performance. UIAs success in fraud detection and
deterrence is the result of a distinguished, multi-tiered approach incorporating a number of
policy, technological, and operations reforms implemented by the agency.

Background
The COVID-19 pandemic placed a tremendous toll on Michigan workers and the economy. Even
the state’s primary economic safety net was impacted by the challenges of COVID-19. The UIA
was faced with a historic number of unemployment benefits claims; at its peak the agency
received 77 times more claims that it did in an average week before the pandemic. In the spring
of 2020, the volume peaked with a high of over 388,000 claims in a single week, compared with
just 5,000 claims before the pandemic and a previous all-time weekly high of 77,000 during the
Great Recession.

Michigan workers found themselves low on cash, unable to pay bills, and siphoning off their
savings to make ends meet and cover the cost of everyday items such as prescriptions, car
repairs, food, and utility bills. Recognizing the financial need, the federal government passed
two legislative remedies—the Coronavirus Aid, Relief, and Economic Security (CARES) Act and
the American Rescue Plan Act, both of which provided extended unemployment benefits,
including payments for workers who traditionally would not have qualified for regular state
unemployment benefits. This legislation was the first of its kind in establishing a financial safety
net for part-time employees, independent contractors, gig workers, and those with irregular
work history.

These congressionally created programs required states to allow claimants to self-attest that
they qualified for benefits.1 While regular state unemployment claims are paid out of a state
trust fund, the Pandemic Unemployment Assistance (PUA), Federal Pandemic Unemployment
Compensation (FPUC), Pandemic Emergency Unemployment Compensation (PEUC), Lost Wages
Assistance (LWA), and other federal programs were financed completely through federal
funding.

In Michigan, 3 million of the 5.5 million claims filed since the start of the pandemic were filed
for PUA or other new federal benefits programs and $32.9 billion of the $39.1 billion paid out in
benefits since March 2020 (the start of the pandemic) were paid for with federal funds.

Amid a massive influx of claims, the UIA was forced to stand up new federal programs on short
notice while at the same time interpreting complex and often shifting federal guidance
resulting in huge challenges for UIA. To assist agency staff, private-sector contractors were
quickly brought on to assist in processing the millions of claims.

1
This was a departure from a typical state unemployment claim, where an employer is generally contacted to
verify that the employee qualifies.
December 29, 2021
The Honorable Gretchen Whitmer
The Honorable Ed McBroom
The Honorable Steven Johnson
Page 3

The complexities surrounding the launch of the new programs provided ample opportunity
for criminals and their networks to exploit the weaknesses in fraud detection. This
exploitation resulted in the theft of billions of dollars from state UI programs across the
country. Sophisticated criminal enterprises were merciless, targeting each state, stealing
money and taking advantage of state UI and federal pandemic programs. Emerging reports
from across the country reveal billions of dollars in fraudulent claims were lost to criminal
activity.

Michigan was no exception. Beginning in March 2020 and continuing to this day, UIA has
detected and prevented millions of fraud attempts, both by criminals using stolen identities to
file fraudulent claims and through intentional misrepresentation, whereby an individual
knowingly misrepresented their qualifications for jobless benefits. While fraudulent claims are
often filed by criminals from other states or overseas, claims associated with intentional
misrepresentation were filed by Michiganders who saw an opportunity during a chaotic time to
be paid more in benefits than they were entitled to or be paid when they wouldn’t normally
qualify for any money.

To date, over 50 Michiganders have been charged with UI fraud by either state or federal
authorities – in some cases netting millions of dollars – with 37 cases pending. Nine people
have pleaded guilty or been convicted and three have been sentenced. Those accused of
facilitating fraud include five UIA employees or contract workers.

UIA invested significant resources in the detection, prevention, and prosecution of criminal
activity. An exact accounting of the fraud losses requires an individual examination of each
claim, infeasible given the sheer number of claims filed since the beginning of the pandemic.
Previous efforts to determine an accurate estimation of fraud losses surmised a total at least in
the hundreds of millions; the exact amount paid out for fraudulent or intentional
misrepresentation claims was previously unknown.

In July 2020, to better understand the scope and develop a more nuanced estimate of
fraudulent claims, the State of Michigan enlisted Deloitte to develop an estimate of the
potential fraud in the Michigan unemployment insurance (UI) program during the economic
crisis brought on by the novel coronavirus, COVID-19. A November 2020 Deloitte report
documented steps UIA had taken to enhance its fraud risk management capabilities to address
identified vulnerabilities in the unemployment system. A copy of the previous report can be
found here.

The November 2020 report presents the factors that led UIA to be vulnerable to fraud and
intentional misrepresentation. It also documents the steps that UIA took in 2020 to resolve
those issues. As described below, the December 2021 report demonstrates that steps outlined
in the November 2020 report were successful in addressing fraud and intentional
misrepresentation, as well as provides a more detailed estimate of potential losses at UIA.
December 29, 2021
The Honorable Gretchen Whitmer
The Honorable Ed McBroom
The Honorable Steven Johnson
Page 4

Report Findings
Deloitte’s findings today provide an overall estimate of the number of claims filed by bad
actors, the amount of benefits paid to these individuals, and the amount of fraud UIA
successfully stopped. There are two types of fraud claims described in these findings:
likely imposter fraud and likely intentional misrepresentation. As previously described,
imposter fraud claims are often filed by criminals from other states or overseas who use stolen
identifies to file multiple false claims, while claims associated with intentional
misrepresentation are generally filed by individuals using their own identity but who overstate
their qualification for UI benefits or misrepresent aspects of their case to qualify for benefits
that they aren’t entitled to.2

Using commonly accepted statistical techniques, the Deloitte Report makes clear that UIA’s
current efforts at fraud detection and deterrence are highly effective. The report found that
from October 3, 2020, to September 30, 2021, UIA paid only 0.57% of both imposter fraud and
intentional misrepresentation cases. For comparison, the DOL shows Michigan’s average pre-
pandemic fraud rate was 2.01% from July 2017 through July 2020, and the federal agency has
estimated that the fraud rate during the pandemic could be much higher than the 3% national
average.

Deloitte also examined how much potential fraud was prevented by UIA during the periods
studied. Deloitte estimated that UIA’s anti-fraud tools and resources successfully thwarted
attempts to steal nearly $43.7 billion from March 1, 2020, to September 30, 2021. During the
same period, an estimated $52.2 billion in fraudulent claims were received. Thus, an estimated
$2.7 -$2.8 billion was paid to claims involving likely imposter fraud and between $5.6 - $5.7
billion was paid to claims involving likely intentional misrepresentation.

The Deloitte report found that less than 3 percent of the total funds paid for imposter fraud and
intentional misrepresentation came from the state UI trust fund. While that had an impact on
the trust fund balance, throughout the pandemic, Michigan has been able to maintain trust
fund solvency, a sharp contrast with the economic downturn in 2008, when the state had to

2
The total amount of intentional misrepresentation estimated by Deloitte also includes
Michiganders who filed a claim under the PUA program using one of four qualification criteria
provided by UIA at the onset of the pandemic, but which were later found by DOL to be
incorrect because they did not require workforce attachment. These individuals were given an
opportunity to requalify using one of the seven statutory criteria established by Congress when
it created the PUA program. Those who were unable to requalify under those criteria were
found to have been provided with an overpayment but can seek a hardship waiver that would
mean they are not obligated to repay the amount received. UIA has already waived $3.7 billion
in overpayments for this population of claimants.
December 29, 2021
The Honorable Gretchen Whitmer
The Honorable Ed McBroom
The Honorable Steven Johnson
Page 5

borrow $3.2 billion from the federal government. Those bonds were paid off in 2019, providing
relief for Michigan employers whose payments fund the trust fund.

What UIA is Doing to Combat Fraud

Due to significant changes in our policy, procedures, and operations, the UIA has successfully
prevented more than 99% of attempted fraud and identified the appearance of intentional
misrepresentation. UIA now employs a holistic approach to fraud prevention and will
continue aggressive action to limit fraud by leveraging tools, capabilities, and partnerships to
build on our near-term success, including:
 Modernize its current IT system. A new claims processing system that is agile, robust,
and secure will allow for quicker response to economic changes and provide more
internal control over touchpoints with our customers.
 Launch an aggressive staff training regimen to address knowledge and skills gaps to
better allow the agency to pivot in times of crisis and move resources fluidly to address
growing issues.
 Identify opportunities for procedural changes or creating new procedures that will
increase Agency efficiency not only in current processes, but in recognizing emerging
trends that need to be addressed.
 Focus on a human-centered, plain language approach with customers by making sure
correspondence isn’t confusing, requests by the Agency are clear, and missteps can be
avoided when dealing with customers.
 Marshalling resources, processes, and stakeholder partnerships to identify potential
fraud cases and continue to deploy new technologies to supplement existing identity
proofing tools.
 Planning and preparing for UIA’s participation in U.S. Department of Labor’s (DOL) Tiger
Teams anti-fraud initiative, tapping multi-discipline experts including fraud specialists. 
 Integrating the National Association of State Workforce Agencies’ (NASWA) Integrity
Data Hub, which identifies high risk claim indicators, into UI fraud detection and
prevention processes.
 Creating partnerships with community organizations to educate potential UI claimants
on eligibility requirements and improve accessibility for disadvantaged communities.
 Collaborating with DOL and NASWA to track trends and emerging fraud schemes
perpetrated by multi-state and multi-national criminal enterprises.

The holistic approach described above serves to enhance the UIA’s fraud risk management
capabilities to address identified vulnerabilities in the unemployment system. UIA:

 Detects anomalies to identify questionable claims for additional review using UIA’s
Fraud Manager software. Fraud Manager analyzes claims at filing and certification,
flagging irregularities or other suspicious patterns.
December 29, 2021
The Honorable Gretchen Whitmer
The Honorable Ed McBroom
The Honorable Steven Johnson
Page 6

 Continues to utilize tools provided by the Integrity Data Hub to identify foreign IP
addresses, suspicious email domains, multi-state claims and other tip-offs to
fraud. (Michigan was one of the first states to join the initiative.)
 Reviews all claim activities and establishing procedures daily to resolve matters for
victims of identity theft who need to file a new claim.
 Expanded its collaboration with state and federal law enforcement to increase
efficiencies, share findings, and investigate fraud cases.
 Utilizes network techniques to enhance claim analysis and prioritize reviews of key
attributes to quickly spot aberrations in data.
 Participates in Attorney General Dana Nessel’s Michigan Unemployment Insurance
Fraud Task Force, along with partners at state, local, and federal law enforcement
agencies.
 Appointed Jeffrey Frost, a retired Special Agent in Charge for the U.S. Secret Service, to
provide expertise from June to November 2020 in the UIA’s efforts to counter the
criminal attacks being experienced by unemployment systems around the country,
analyze fraudulent unemployment activity, and clear legitimate accounts.
 Contracted with consultants to identify improvements in fraud detection, streamline
workflows to minimize delays in case processing, and develop operation metrics to track
successes.
 Contracted with Deloitte to support the UIA in continued identification and remediation
of new, or previously unknown schemes, as well as the analytic support of ongoing
investigations.
 Contracted with Accenture to support victims of identity theft through near immediate
resolution of all fraudulent claim filings attempted by criminals, and one-on-one help
where appropriate.
 Hired approximately 70 limited-term staff into the Investigations Division.
 Participates in the NASWA sponsored Regional Integrity State Workgroup and the bi-
weekly Integrity Data Hub Workgroup to meet with peers across the country, to discuss
current issues, exchange best practices, contribute, and discuss efforts taken on by the
NASWA Integrity Center and Integrity Data Hub.
 Provides resources on UIA’s What is Fraud webpage to explain fraud types, how UIA is
combating fraud, and help Michiganders identify and report fraud or identity theft.

In Conclusion
Effectively combating fraud is necessary to preserve the integrity of the UIA program as well
public confidence. There is no acceptable level of fraud, and the loss of federal taxpayer dollars
to criminal activity means fewer dollars in the pockets of those deserving of UI benefits. UIA
takes seriously its duty to steward taxpayer funds, whether state or federal. We have
implemented significant proactive measures aimed at identifying and stopping fraud work. Our
diligence is a powerful deterrent, but the fight is never over: the criminals targeting UI funds
nationwide include sophisticated, international fraud rings, always looking for new
December 29, 2021
The Honorable Gretchen Whitmer
The Honorable Ed McBroom
The Honorable Steven Johnson
Page 7

opportunities to steal money that should rightfully go to workers and those who find
themselves without a job and need money to keep a roof over their heads, buy food, clothe
their children, and fill prescriptions until they return to work.

The UIA is committed to working collaboratively with our partners and stakeholders on ever
evolving opportunities to improve our fraud detection and deterrence capabilities and will
continue to work closely with law enforcement to prevent bad actors from defrauding the
system.

Sincerely,

Julia Dale
Director
Michigan Unemployment Insurance Agency

Enclosure

cc: Senator Jim Ananich

Representative Donna Lasinski
Senator Mike Shirkey, Senate Majority Leader
Representative Jason Wentworth, Speaker of the House
Susan Corbin, Director, Department of Labor and Economic Opportunity
 CONFIDENTIAL

Deloitte & Touche LLP

200 Renaissance Center, Suite 3900
Detroit, MI 48243
USA

Tel: +1 313 396 3000

www.deloitte.com
December 23, 2021

To: State of Michigan Unemployment Insurance Agency (UIA)

Re: Fraud Measurement Estimation

Pursuant to the Financial Accounting, Integrity Oversight and Auditing change notice dated July 1, 2020,
between the State of Michigan (“SoM,” “Michigan,” or the “State”) and Deloitte & Touche LLP
(“Deloitte”), please find the results of the fraud measurement estimation below. The report documents
multiple dimensions of the estimated potential fraud in the Michigan unemployment insurance (UI)
program during the the economic crisis brought on by the novel coronavirus, COVID-19 (“COVID-19” or
the “Pandemic”).

The report includes an estimate of the identified and potential fraud in the UI program that occurred
between March 2020 and September 2021, including the estimated amount potentially paid to fraudulent
claims and the estimate of fraud payments that the State avoided, but otherwise potentially would have
been paid out. As described herein, fraud reviews were conducted by the State for sampled claims and
calculations were performed on data provided by the State. The methodology described herein leverages
commonly accepted statistical techniques and was agreed upon with the State.

This confidential report is intended solely for the information and use of the State and is not intended to
be and should not be used by anyone else for any other purpose. With the exception of complying with
applicable Freedom of Information Act laws, this confidential report should not be disclosed, quoted,
copied, published, or used in whole or in part without the express written consent of Deloitte. This
confidential report must be read in its entirety, and Deloitte is not responsible for any portion of this
report that is selectively quoted or otherwise used in isolation or any summary or paraphrasing of the
report that is prepared by others.

Page 1 of 5
 CONFIDENTIAL

I. Background on Unemployment Fraud During the Pandemic

The increase in available Unemployment Insurance (UI) benefits by federal programs in response to the
COVID-19 pandemic resulted in a significant national increase in UI claims, both of a legitimate and
potentially fraudulent nature. For example:

“The [USDOL] Office of the Inspector General (OIG) reported a 40-fold increase in fraudulent
UI investigations since the beginning of the pandemic and the implementation of the Pandemic
Unemployment Assistance program.”1

In June 2020, Scott S. Dahl, the Inspector General of the Department of Labor, indicated that “the
fraud rate could range from at least 3% to a much higher percentage” during a briefing to
Congress.2

A combination of historically unprecedented claim volume that many states were unprepared to manage,
unclear federal guidance and eligibility policy for new or expanded federal UI benefits, and the apparent
attraction to suspicious actors of significant increases of federal benefits, resulted in challenges for many
states – including the challenge of detecting and preventing heightened fraud in UI programs. The State
requested assistance from Deloitte in understanding the extent of this issue in Michigan.

II. Methodology for Estimating Potential Fraud

The State faced the greatest risk of fraud in its UI program between March 2020 and September 2021, the
period for which heightened federal funding for UI benefits was available. The potential fraud risk (i.e.,
the total estimated fraud that the State could have paid to claimants) to the State’s program can be
separated into the estimated fraud likely avoided as a result of its fraud detection measures and the
estimated fraud that the State did not avoid (i.e., the estimated fraudulent claims paid to individuals).
Deloitte assisted the State in conducting estimation approaches to measure the potential fraud avoided and
the potential fraud paid. These estimation approaches are described below.

A. Estimating Fraud Avoided

The amount of fraud avoided by the State is estimated by identifying claims prevented due to fraud-
related applications (e.g., Fraud Manager – Michigan UIA’s fraud detection tool) and calculating the
unrealized benefits that could have been paid had the State not detected and stopped payments on these
claims. This approach proceeded in four steps.

1. Identify the claims flagged: From the MiDAS system (Michigan Integrated Data Automation
System – internal database for accessing data on UI claims), identify the claims that were stopped
for fraud-related reasons by Michigan’s Fraud Manager or had an investigation opened due to
another source (e.g., staff investigation, tips and leads, etc.) between March 2020 and September
2021.
2. Estimate the Maximum Benefit Amount: For the claims identified in Step 1, extract the
maximum amount that a claim could have been paid based on the total weeks allowed and weekly
benefit amount (as calculated by MiDAS), including supplemental benefits such as increased or
extended benefits.
3. Identify the amount paid: Sum the benefit amount that these flagged claims were paid between
March 2020 and September 2021.
4. Estimate fraud avoided: Estimate fraud avoided as the difference between the maximum benefit
amount (Step 2) and the amount paid (Step 3) for each claim that was flagged.

B. Estimating Fraud Paid

To estimate the amount of benefits that were paid to potentially fraudulent claims, as agreed to by the
State, Deloitte and the State used a stratified random sampling methodology.3 This methodology included
the use of a series of mutually exclusive, exhaustive, and complete strata to identify potential risk of
fraud. For each of these strata, a statistical power calculation4 was performed to identify the number of
samples to randomly draw within each strata. Claims from each strata were randomly selected and
reviewed by State UIA staff to determine if the claims involved potential fraud. Results from these
reviews were compiled and the estimated amount of fraud paid was extrapolated from these samples. This

1
https://www.ncsl.org/research/labor-and-employment/unemployment-insurance-improper-payments-and-fraud.aspx
2 Subcommittee Briefing with DOL IG Highlights Key Oversight Priorities in Response to Coronavirus Crisis | House Committee on Oversight
and Reform
3
Aoyama, Hirojiro. "A study of stratified random sampling." Ann. Inst. Stat. Math 6 (1954): 1-36.
4
Gelman, Andrew, and Jennifer Hill. Data analysis using regression and multilevel/hierarchical models.Cambridge university press, 2006.

Page 2 of 5
 CONFIDENTIAL

stratified sampling approach was conducted separately on claims filed during the period March 1, 2020,
and October 2, 2020, and again on claims filed between October 3, 2020, and September 30, 2021.

In total, thirty-eight strata were identified based on risk indicators for potential fraud. These indicators
included the rules in the State’s Fraud Manager and other indicators of potential fraud identified in
coordination with the State UIA staff. In addition to strata based on the risk of fraud, two additional strata
were identified including claims that were not included in the previously identified strata. One of these
additional strata included claims that were not included in previous strata but were assigned a non-
monetary issue by the MiDAS system, while the other strata included claims that were not included in the
previously identified strata and did not have a non-monetary issue assigned by the MiDAS system. In
aggregate, these strata covered the entire population of claims filed between March 1, 2020, and
September 30, 2021.

A sample was drawn at random from each of these strata. The size of the sample was determined by
applying a statistical power calculation to identify the minimum sample size to derive a five percent (5%)
error rate within a ninety-five percent (95%) confidence interval (Gelman and Hill 2006). A finite
population correction was applied to the sample size for each strata.5 Additionally, those strata which
were deemed especially important to review by the State were judgmentally allocated larger sample sizes
than suggested by the power calculations. The specified number of claims were then drawn using a
random number generator from each of the strata. The sample sizes in the strata ranged from 22 to 384. In
total, 7,741 samples were drawn for the period March 1, 2020, to October 2, 2020, and 7,096 samples
were drawn for the period October 3, 2020, to September 30, 2021.

State personnel reviewed each of the sampled claims to reasonably infer if the claim was potentially
fraudulent. Each claim was assigned one of five outcomes:
1. Likely imposter fraud: Claims filed by an apparent bad actor in the name of another person to
fraudulently extract funds. Example characteristics include: claimants with suspicious attributes
(e.g., suspicious bank accounts), claimants with suspicious addresses (out of state with no
connection to MI, filing from nursing homes, etc.), claimants who appear on the death or prisoner
lists.
2. Likely intentional misrepresentation fraud: Claims filed by apparently legitimate claimants who
appeared to be misrepresenting their eligibility for benefits. Examples include when individuals
submit apparently fabricated income verification documentation or knowingly fail to report
information which would make them ineligible to receive benefits.
3. Likely non-fraud improper payment: Claims filed by legitimate claimants exhibiting other issues
indicating potential improper payments,6 including potential misapplication of UI policy, not
otherwise determined to be likely imposter fraud or likely intentional misrepresentation.
4. Inconclusive
5. No anomalies (i.e., likely an eligible claim)

The State implemented a protocol where trained State personnel reviewed the sampled claims for risk of
fraud. The personnel reviewed each claim and determined that the claim likely involved one of the
outcomes listed above. A lead reviewer then assessed each claim and confirmed the determined outcome
as a second-level review.

The findings for each strata were compiled and statistical calculations were performed to estimate the
amount of fraud and a 95% confidence interval. The estimated fraud in each time period was extrapolated
from the claims determined to likely involve imposter fraud (Outcome 1 from above) or intentional
misrepresentation fraud (Outcome 2 from above). The estimate was calculated using weighted
extrapolation from each strata.7 The 95% confidence interval for continuous measures was derived from
the normal distribution (Gelman and Hill, 2006, page 18).

III. Findings

The estimated potential fraud, including estimated fraud avoided and estimated fraud paid for the periods
described above are included below in Table 1. These estimates reflect the 95% confidence interval for
the fraud paid, calculated as described above. For claims filed between March 1, 2020, and October 2,
2020, the State avoided an estimated $28.7 billion in fraud while paying out an estimated $8.36-$8.51
billion on potentially fraudulent claims during this time period. By totaling these two figures, this equates
to the State having received an estimated total of $37.1-$37.2 billion in potentially fraudulent claims over
this timeframe. For claims filed between October 3 2020, and September 30, 2021, the State avoided an
estimated $15.0 billion in fraud while paying an estimated $34.2-$35.7 million in potentially fraudulent

5
Burstein, Herman. "Finite population correction for binomial confidence limits." Journal of the American Statistical Association 70.349 (1975):
67-69.
6 Payment Integrity Information Act (2019) https://www.congress.gov/116/plaws/publ117/PLAW-116publ117.pdf
7
Journal of the American Statistical Association Vol. 55, No. 292 (Dec., 1960), pp. 708-713

Page 3 of 5
 CONFIDENTIAL

claims. By totaling these two figures, this equates to the State having received an estimated total of $15.0
- $15.1 billion in potentially fraudulent claims over this timeframe. In total, for claims filed for the
duration of March 1, 2020, to September 30, 2021, the State avoided an estimated $43.7 billion in fraud
while paying an estimated $8.4-$8.5 billion to potentially fraudulent claims. In total, this equates to the
State having received an estimated total of $52.1-$52.2 billion in potentially fraudulent claims over this
timeframe.

Table 1: Potential Fraud Estimates

Time Period for Claims Filed Estimated Estimatd Fraud Estimated Total
Fraud Paid Fraudulent Claims
Avoided Received
March 1, 2020 – October 2, 2020 $28.7B $8.36B - $8.51B $37.1B - $37.2B
October 3, 2020 – September 30, 2021 $15.0B $34.2M - $35.7M $15.0B - $15.1B

The estimated fraud paid can be further separated into claims that involved likely imposter fraud and
those that involved likely intentional misrepresentation fraud. These results are displayed in Table 2
below. For claims filed during the period March 1, 2020, to October 2, 2020, an estimated 9.7% of
benefits paid involved likely imposter fraud while an estimated 20.1% of benefits paid involved likely
intentional misrepresentation fraud. For claims filed between October 3, 2020, and September 30, 2021,
an estimated 0.46% of benefits paid involved likely imposter fraud and an estimated 0.11% of benefits
paid involved likely intentional misrepresentation fraud. In total, for claims filed between March 1, 2020,
and September 30, 2021, an estimated 8.07% of benefits paid involved likely imposter fraud while
16.51% of benefits paid involved likely intentional misrepresentation fraud. For the period March 1,
2020, to September 30, 2021, an estimated $2.7-$2.8 billion was paid to claims involving likely imposter
fraud and an estimated $5.6-$5.7 billion was paid to claims involving likely intentional misrepresentation
fraud.

Table 2: Estimated Percentage of Benefits Paid by Fraud Type

Time Period for Claims Filed Estimated Percentage of Estimated Percentage of

Benefits Paid Involving Benefits Paid Involving Likely
Likely Imposter Fraud Intentional Misrepresentation
Fraud
March 1, 2020 – October 2, 2020 9.7% 20.1%
October 3, 2020 – September 30, 2021 0.46% 0.11%

These amounts of estimated fraud paid can be further separated by funding source (i.e., state vs. federal).
To estimate these amounts, the payments made to sampled claims determined to involve likely imposter
fraud or likely intentional misrepresentation were separated by program type and extrapolated based on
data available in MiDAS (refer to Appendix 1). Examples of unemployment insurance program types
include standard UI, Pandemic Unemployment Assistance (PUA), Federal Pandemic Emergency
Compensation (FPUC), and Lost Wages Assistance (LWA). Table 3 below shows the summary of
estimated amounts paid to likely imposter or international misrepresentation by funding source (i.e., state
vs. federal).

Table 3: Estimated amount paid to likely imposter fraud and misrepresentation fraud for claims filed between March 2020
and September 2021 by funding source

Funding Source Estimated Estimated Benefits Paid to

Percentage Likely Imposter Fraud or
of Total Misrepresentation Fraud
Estimated
Fraud Paid
State Funded (UI) 2.9% $242M - $249M
Federal Funded (FPUC, PUA, PEUC, 97.1% $8.15B - $8.25B
LWA, Other Federal)

Page 4 of 5
 CONFIDENTIAL

Appendix 1: Michigan UI Claims Information

Table A-1 below shows the volume of Michigan UI claims and benefits filed and paid during the period
of March 1, 2020, to September 30, 2021. The data was summed from the MiDas system as of 12/1/2021.

Table A-1: Benefits paid for Michigan UI claims filed between March 2020 and September 2021

Time Period for Claims Filed Total Claims Total Benefits Paid
Paid
March 1, 2020 – September 30, 2021 3,535,495 $34,474,279,879
March 1, 2020 – October 2, 2020 2,527,983 $28,313,914,893
October 3, 2020 – September 30, 2021 1,007,512 $6,160,364,986

Table A-2 below shows the Michigan UI benefits filed and paid between March 1, 2020, and September
30, 2021, by program type. The data was summed from the MiDas system as of 12/1/2021.

Table A-2: Benefits paid for Michigan UI claims filed between March 2020 and September 2021 by program type

Program Type Percentage of Benefits Paid

Total
Unemployment Insurance (UI) 14.97% $5.2B
Pandemic Unemployment Assistance (PUA) 16.60% $5.7B
Federal Pandemic Unemployment 53.58% $18.5B
Compensation (FPUC)
Pandemic Emergency Unemployment 7.44% $2.6B
Compensation (PEUC)
Lost Wages Assistance (LWA) 4.42% $1.5B
Other Federal Programs8 3.00% $1.0B
Other <0.01% $250K
Total 100% $34.5B

8
Other Federal Programs include Extended Benefits (EB), Mixed Earnings Unemployment Compensation (MEUC) and other smaller
adjustments.

Page 5 of 5