A Better Form of Unemployment Protection

Spring 2021 • Regulation

When it comes to finding creative ways to expand the size of government — especially with old, tired, and problematic programs — legislators are extremely resourceful. Case in point: 2020’s expansion of unemployment insurance (UI) as part of the Coronavirus Aid, Relief, and Economic Security (CARES) Act. Expanding unemployment insurance during a recession is to be expected. What’s unusual, though, was the scale and means of this latest expansion.

Like UI expansion in the past, and like UI benefits in general, this most recent program created major economic distortions and is likely to continue slowing the recovery even if firms are allowed to get back to business as usual. It is time for a serious reform of the program.

UI today/ Unemployment insurance is a joint federal/​state program financed through payroll taxes. Each state operates its own program under federal guidelines. States set their own eligibility requirements, coverage limits, financing methods, and formulas for determining benefits. However, in general, most states provide unemployment benefits that replace about half of a worker’s previous wages for up to 26 weeks.

Take Virginia for example. The state pays UI benefits equal to half of the worker’s prior wages up to a maximum of $378 per week. The 50% replacement of prior wages holds for employees earning up to $39,312 annually, and a smaller and declining percentage for earnings above that amount. For instance, an unemployed worker who earned $100,000 annually receives a 19.6% replacement rate.

During nationwide downturns, the federal government usually provides supplemental funding to increase the UI replacement rate, or to lift the upper limit on benefits, or both. Also available for expansion is the number of weeks unemployed workers are eligible for UI. For instance, during the Great Recession, the federal government allowed UI recipients to receive up to 99 weeks of benefits.

The CARES Act expanded eligibility to many people who would not otherwise qualify for state UI but who nevertheless cannot work because of the ongoing pandemic. This group includes self‐​employed workers, independent contractors, part‐​time employees, and those who quit their jobs for coronavirus‐​related reasons (for example, people who are sick or taking care of a dependent if they do not have paid‐​leave benefits). In addition, the expansion offered workers an additional $600 a week for four months on top of what state unemployment programs pay. The projected cost to taxpayers for all this relief was $260 billion.

Generous unemployment benefits create a disincentive for workers to look for jobs and for families to work at other income sources like secondary earners’ jobs.

With the CARES Act expansion, Congress meant to address the sudden increase in unemployment — millions of workers lost their jobs almost overnight — because of the virus. Unfortunately, this policy was both a benefit and a harm. The benefit is easy to see. Paying people who have lost their job during the pandemic alleviates a lot of pain. But paying generous benefits made some people better off from not working rather than from returning to work; economists at the University of Chicago found that roughly two‐​thirds of the UI beneficiaries received 134% of their past wages under the CARES expansion.

The harm from this policy was predictable and is now well documented: the UI expansion created disincentives to work, prompting many people to drop out of the workforce. While this result might have been the one desired by state officials when they shut down their economies in the hope of keeping people from spreading the virus, it has been terrible for employers that have continued to operate. It became an even bigger problem when the economy started reopening and employers began calling employees back to work.

Examining the effect of expanding this provision of the CARES Act, the Congressional Budget Office noted that the “extension would also weaken incentives to work as people compared the benefits available during unemployment to their potential earnings, and those weakened incentives would in turn tend to decrease output and employment.” The CBO report also put to rest the idea that UI expansion offers much of a boost to the economy from a benefit‐​fueled increase in aggregate demand; the CBO found that the bonus and UI expansion under the CARES Act had a multiplier of just 0.67.

This result is neither surprising nor new. The unintended consequences and moral hazards of UI during normal times and normal recessions are well‐​documented. Generous UI benefits create a disincentive for workers to look for jobs. Families respond to unemployment benefits by working less at other income sources, like secondary earners’ jobs. UI benefits also create disincentives to save by crowding out as much as half of private savings for the typical unemployment spell.

In addition, UI creates financial troubles for the states. During times of high unemployment — and, hence, of higher benefit payments — UI funds get depleted in most states while states’ tax bases are shrinking. As a result, once the economy begins to recover, states are tempted to increase their payroll taxes in order to replenish their trust funds. Depending on the trajectory of the recovery, these raised UI taxes dampen business hiring at a time when unemployment remains high.

This is exactly what happened during the last recession. According to the Federal Reserve Bank of Boston, during the Great Recession states had higher unemployment claims, lower ratios of taxable to total wages, and they faced insolvency of their trust funds. The report notes that “at least 35 states borrowed at some point to maintain fund solvency.” To replenish their reserves, states facing insolvency increased their UI taxes on businesses, thus creating further distortions and delaying the economic recovery.

The time has come for a different approach to unemployment protection.

PISAs/ Personal Unemployment Insurance Accounts (PISAs) were pioneered by Chile in 2002. The accounts are financed through a payroll‐​tax contribution from both the employer and employee and are individually owned by workers. During spells of unemployment, idled workers can make withdrawals to compensate for the loss to their incomes, but when employed they continue to build their balances. At retirement, workers can use the balances in these accounts to bolster their retirement income or transfer the funds to their heirs. The program includes a solidarity fund — a public safety net — financed by employers and the government. Unemployed workers can receive payment from the solidarity fund when their own savings are insufficient to cover their period of employment.

These accounts provide insurance while keeping strong incentives for people to return to work. Several studies have confirmed that under this system, workers are motivated by a desire to keep their own savings for retirement, so they are careful about tapping into this money during their working years. On net, workers seem as well off with PISAs as they were under the old UI system.

Similar plans have been adopted in other Latin American countries as well as in Austria and Jordan. Scholars have recommended similar programs for developing countries, in part for PISAs’ minimal burden on governments’ budgets.

One drawback of the Chilean model is that the solidarity fund creates the same kind of disincentives to work as a traditional UI system. For instance, some workers wait until the solidarity funds are exhausted before looking for a job, just as some workers on traditional UI wait until their benefits are about to end before getting serious about their job hunt. A U.S. PISA program should not include this feature. With this revision, Chile’s program provides a good template for a U.S. program that would replace the current unemployment insurance program.

In a perfect libertarian world, the accounts would be funded solely on a voluntary basis. Realistically, they would probably be funded by mandatory contributions from employers and employees. The contributions would be made to the account of each employee until it reaches a certain level of benefits (for instance, 80% wage replacement for six months).

A U.S. PISA program should also allow employees to voluntarily contribute additional funds to their accounts, thus giving them a chance to build a larger UI benefit and more savings. Ideally, the extra savings, including the interest paid to the account based on the additional contributions, should be available to the employee for withdrawal for any need at any time. Also, withdrawals would be allowed after a separation from an employer regardless of the reason.

Under such a system, there would be no need for state UI trust funds and hence states’ outlays would not be affected by UI during recessions. The program would also free the states from federal pressure to permanently expand their benefits and create future spending, as occurred in the last recession. By eliminating the current system’s complexity, PISAs would also reduce the scale of improper payments.

Finally, once the federal and state governments get out of the UI business, it is likely that innovative private options would emerge. As Chris Edwards and George Leef have documented, before the federal government stepped in to provide unemployment insurance in 1935, labor unions, employers, and private insurance companies offered UI benefits. Traditional UI finances its benefits through implicit taxes on savings. Under a PISA system, we can expect total personal savings to increase as workers prepare for unemployment episodes.

Conclusion/ Incentives matter. UI benefits provided by the government are well known for creating disincentives to work, at great cost to taxpayers. During a recession, these effects are compounded by benefit expansions. As a result, economic recoveries take longer than they should and the economic costs are unnecessarily heavy. The current system also is a serious drain on states’ trust funds.

Now is the time to replace America’s traditional UI with a system that better fits our modern economy. PISAs would provide a far superior way to respond to unemployment with fewer of the moral hazards that are endemic to the current system.


  • “Does Unemployment Insurance Crowd Out Spousal Labor Supply?” by Julie B. Cullen and Jonathan Gruber. Journal of Labor Economics 18(3): 546–572 (2000).
  • “Failures of the Unemployment Insurance System,” by Chris Edwards and George Leef. Downsizing Government (website), Cato Institute, June 1, 2011.
  • “Job Search and Unemployment Insurance: New Evidence from Time Use Data,” by Alan B. Krueger and Andreas Mueller. Journal of Public Economics 94(3–4): 298–307 (2010).
  • “Quasi‐​Experimental Evidence on the Effects of Unemployment Insurance from New York State,” by Bruce Meyer and Wallace K. Mok. National Bureau of Economic Research Working Paper no. 12865, 2007.
  • “The Chilean System of Unemployment Insurance Savings Accounts,” by Kirsten Sehnbruch and Rafael Carranza. University of Chile Department of Economics Working Paper SDT 401, April 2015.
  • “The Folly of Subsidizing Unemployment,” by Robert Barro. Wall Street Journal, August 30, 2010.
  • “The Impact of the Potential Duration of Unemployment Benefits on the Duration of Unemployment,” by Lawrence F. Katz and Bruce D. Meyer. Journal of Public Economics 41(1): 45–72 (1990).
  • “Unemployment Insurance and Precautionary Saving,” by Eric Engen and Jonathan Gruber. National Bureau of Economic Research Working Paper no. 5252, September 1995.
  • “Unemployment Insurance in Chili: A New Model of Income Support for Unemployed Workers,” by German Acevedo, Patricio Eskenazi, and Carmen Pages. World Bank, Social Protection Discussion Paper no. 0612, October 2006.
  • “US Unemployment Insurance Replacement Rates During the Pandemic,” by Peter Ganong, Pascal Noel, and Joseph S. Vavra. University of Chicago, Becker Friedman Institute for Economics Working Paper No. 2020–62, August 24, 2020.
  • “When the Tide Goes Out: Unemployment Insurance Trust Funds and the Great Recession, Lessons for and From New England,” by Jennifer Weiner. Federal Reserve Bank of Boston, New England Public Policy Center Research Report 12–1, April 2012.

Download the Regulation Article

About the Author