By Nearly five years after federal pandemic unemployment programs expired, they remain in the news. Federal investigators continue to uncover billions of dollars in fraudulent pandemic unemployment claims. States are still repaying debts incurred during the COVID-19 economic downturn. Despite the massive federal assistance, fewer state unemployment funds are prepared to weather a recession today than they were before the pandemic.
My paper, A Safety Net Full of Holes, explores unemployment insurance (UI) before, during, and after 2020. Long after the emergency has subsided, many of the system’s financial and administrative problems remain. That persistence raises the paper’s central question: If every state experienced the same national recession and had access to the same federal emergency programs, why did some unemployment insurance systems recover while others remained vulnerable?
Before exploring that question, it is important to understand how UI is designed. It was created in 1935 to accomplish two goals: providing temporary, partial wage replacement while unemployed workers search for work (thereby lowering the cost of job searching) and acting as an automatic stabilizer by maintaining household spending during recessions.
UI is a joint federal-state program. Although federal law establishes broad rules for the program, states administer their own systems. State policy largely determines benefit formulas, eligibility standards, payroll tax schedules, and how regular benefits are financed.
Each state has its own “trust fund” at the U.S. Treasury that acts as a reserve account. States levy payroll taxes on employers to fund regular benefits and, generally, half of the cost of extended benefits. Meanwhile, federal unemployment taxes cover the administrative maintenance of state accounts, the federal share of extended benefits, federal accounts that make loans to insolvent state funds, and several federal unemployment accounts (such as two UI accounts related to the Railroad Retirement Board).
During periods of economic expansion, states are expected to accumulate reserves in these accounts. When unemployment rises during a recession, those reserves are drawn down to pay benefits. If a state’s account is exhausted, it may borrow from the federal government until sufficient tax revenue is available to repay the loans.
This financing system creates an important incentive structure. States largely determine how much they tax employers before a recession, how generous regular benefits are during one, and how quickly trust funds are replenished afterward. Employer tax rates are also “experience rated,” meaning firms with greater histories of layoffs generally pay higher unemployment tax rates than firms with more stable employment.
In principle, this structure should encourage stable employment and adequate reserve accumulation. In practice, states entered 2020 in very different positions. These starting positions strongly predicted how states emerged from the COVID-19 economic downturn.
Resilience begins before a crisis. States that accumulated adequate reserves before 2020 were generally better positioned to absorb the surge in unemployment claims without exhausting their funds. Conversely, states with weaker reserves were more likely to exhaust their accounts, borrow, and remain vulnerable after the downturn.
In response to the pandemic, Congress created several emergency UI programs in 2020 and extended them in 2021. These programs sent additional federal dollars through state systems by expanding eligibility, extending benefit duration, and supplementing weekly payments. Participating states entered agreements to administer these benefits under federal rules.
As economists Casey Mulligan, Stephen Moore, and E.J. Antoni argued, the supplemental benefits weakened work incentives because, for some recipients, total benefits exceeded prior earnings or the pay available from returning to work. In response, twenty-six states announced plans to terminate their participation before the programs expired in September 2021. Among the twenty-six, four states (Arkansas, Indiana, Maryland, and Oklahoma) were prevented from fully withdrawing due to legal or administrative challenges. The paper classifies these states as “partial withdrawal states.”
Ultimately, states that withdrew early from these programs experienced stronger post-2020 trust fund solvency than states that remained in the program until it expired. Partial-withdrawal states also experienced stronger outcomes than full-duration states, although the relationship was weaker than it was for full-withdrawal states. It is important to note that these findings identify statistical associations rather than proving causation. They nevertheless suggest that the timing and execution of policy decisions deserve closer attention.
Borrowing also offered short-term relief but prolonged financial challenges. Loans allowed states to continue paying benefits, but repayment obligations extended well beyond 2021.
California still carries a large federal loan balance, while Massachusetts employers continue paying a recovery assessment associated with bonds issued to retire the state’s federal advances.
Administrative capacity also mattered. States with higher improper payment rates (which include fraud as well as payments that should not have been made or were made in the incorrect amount) generally experienced weaker trust fund outcomes, although these relationships were less consistent than those associated with pre-pandemic solvency or program withdrawal. While fraud alone did not determine a state program’s durability, weak administration can make recovery more difficult.
These findings help explain why unemployment insurance continues to generate headlines years after the pandemic. Recent reports of ongoing fraud investigations and federal efforts to recover improperly paid benefits are not merely historical cleanup. They reflect institutional weaknesses that existed before COVID-19 and became impossible to ignore once the system faced extraordinary stress. Likewise, the continued weakening in state trust fund solvency is evidence that many unemployment insurance systems entered the crisis underprepared and have struggled to rebuild.
The lesson extends beyond the pandemic. Future recessions are inevitable, and states must prepare for them. Improving program integrity, encouraging adequate reserve accumulation, and designing financing systems that better align incentives would strengthen the current framework. Policymakers should also consider more fundamental reforms. Worker-owned savings mechanisms, such as personal unemployment insurance accounts or more flexible universal savings accounts, deserve renewed attention, especially as Trump Accounts have reopened the debate over how the tax code should treat household saving.
The pandemic exposed weaknesses that had accumulated over decades. The latest headlines show those weaknesses have not disappeared; they have merely become harder to ignore.
Read the full paper: A Safety Net Full of Holes: Unemployment Insurance in the Wake of 2020 and the Path to Reform