Introduction and executive summary
The Great Recession and its aftermath created severe challenges for unemployment insurance (UI) programs in the United States and for jobless workers relying upon them. In this briefing paper, we show that state UI programs are failing their critical goals of income replacement and supporting economic growth. The proportion of jobless workers receiving benefits from state programs, referred to as the UI recipiency rate, fell to 23.1 percent in December 2014—below the pre-Great Recession record low of 25.0 percent in September 1984.
Due to the expiration of federal emergency unemployment benefits at the end of 2013, jobless individuals were solely dependent upon state UI programs for support in 2014. While state UI benefit recipiency overall has declined due to the improving economy, these state programs in many cases failed to assist jobless workers. This brief focuses special attention on those states that have cut their potential available weeks of UI benefits to below the long-accepted norm of 26 weeks. Because state UI programs are mainly designed to address short-term unemployment, we focus our analysis on the short-term recipiency rate, which excludes people who have been unemployed for 27 weeks or more from the proportion of jobless workers receiving benefits from state programs.
The key findings of this brief include the following:
- Since 2011, nine states have cut maximum durations of unemployment benefit recipiency: Arkansas, Florida, Georgia, Illinois, Kansas, Michigan, Missouri, North Carolina, and South Carolina.
- Eight of these states have experienced faster-than-average declines in their short-term recipiency rates. The exception is Illinois, which cut available benefits by only one week for a single year (to 25 weeks for 2012). In four of the states (Florida, Georgia, North Carolina, and South Carolina), short-term recipiency rates declined by between 1.7 and 8.6 times as much as the U.S. average decline.
- By cutting available weeks of benefits, these eight states’ already-low short-term recipiency rates fell even further below the recipiency rates of all other states. In 2014, Florida, Georgia, North Carolina, and South Carolina ranked in the bottom eight states in short-term (less than 26 weeks) recipiency rates.
- In North Carolina, one of the states with the most severe cuts (cutting the duration of benefits from 26 weeks in 2013 to 14 weeks in 2014 as well as cutting the level of weekly benefit amounts), the decline in the short-term recipiency rate was 14.4 percentage points (or 8.6 times) greater than the nation’s average decline since the cuts went into effect in July 2013.
Expanding our analysis to the regular (versus short-term) UI recipiency rate, we find that jobless people exhausting state UI benefits in 2014 had less protection from income loss than any cohort of jobless individuals exhausting state UI benefits over the last few decades.
The current labor market situation
The damage caused by the Great Recession was extensive and prolonged. Further, the recovery that officially started in June 2009 has been characterized by historically tepid growth. Despite a more recent acceleration of employment growth, the labor market is far from a full recovery. As of December 2014, the unemployment rate was 5.6 percent, down from its peak of 10.0 percent in October 2010.
The drop in the official unemployment rate overstates the overall improvements made in the underlying labor market. The United States lost 7.8 million jobs between December 2007 and October 2010 but the working-age population continued to grow over that period. As a result, even with steady job growth in recent years, the current labor market is still short 5.6 million jobs needed to keep up with the growth in potential labor force (see Figure A). We are still far from a healthy labor market.
Payroll employment and the number of jobs needed to keep up with the growth in the potential labor force, 2000–2014
|Year||Payroll employment – actual||Jobs needed to keep up with growth in potential labor force|
Note: The potential labor force is the actual labor force plus the “missing workers,” potential workers who, because of weak job opportunities, are neither employed nor actively seeking a job. How EPI calculates missing workers can be found here: http://www.epi.org/publication/missing-workers/
Source: EPI analysis of Bureau of Labor Statistics’ Current Employment Statistics public data series and Current population Survey public data series
Background on UI programs
Unemployment insurance (UI) is a federal-state program that provides income support for jobless workers in economic downturns.
A recent EPI report provides a succinct overview of how the federal-state system has historically worked:
In the United States, the federally supported but state-administered unemployment insurance (UI) system typically provides someone who has lost a job through no fault of his or her own with unemployment benefits for up to 26 weeks. In states that have experienced a sharp rise in unemployment rates, the extended benefit (EB) program kicks in, providing an additional 13 to 20 weeks of jobless benefits. And in times of severe economic distress, Congress routinely votes to provide extra weeks of aid beyond EB. The most recent Emergency Unemployment Compensation (EUC) program was authorized by Congress in June 2008. … It was allowed to lapse in December 2013. (Bivens, Smith, and Wilson 2014)
The EUC program authorized in 2008 lasted 66 months and provided at its maximum 63 weeks of additional benefits.
While the federal government fulfills a more central role than states in providing benefit extensions during recessions, state governments have responsibility for setting the main parameters of regular state UI programs, including eligibility rules, benefit amounts, and weeks of benefits available. (Detailed descriptions of how UI programs changed during and after the Great Recession are provided by Bradbury (2014) and Isaacs (2012).)
The principal goals of UI programs are providing involuntarily unemployed workers with adequate, temporary income replacement while maintaining consumer spending levels during an economic downturn. Related goals include supporting the job search of unemployed individuals by permitting them to find work that matches their prior experience and skills, as well as enabling employers to retain experienced workers during layoffs (Advisory Council on Unemployment Compensation 1995, 27–30). One important performance indicator for state UI programs is the proportion of unemployed individuals who get UI benefits, or the benefit recipiency rate. Both the income support and economic stimulus objectives of UI programs are better fulfilled with higher levels of benefit recipiency (Vroman 2011).
Viewing UI as an important social insurance program, this brief assumes that falling recipiency rates are a troubling development. Admittedly, many economists concentrate on the moral hazard risks posed by unemployment benefits. These theoretical concerns (that UI benefits encourage unemployment) are undercut by more recent research finding empirical evidence that UI benefits improve job-matching (the fit between a worker’s skills and earnings experience and the pay provided and skills required by the new job) and support work search. Bivens, Smith, and Wilson (2014) provided an overview of recent economic research concerning the relationship between duration of unemployment and UI, casting doubt on traditional perspectives that concentrate largely upon moral hazard. The focus on moral hazard has been criticized for making unrealistic assumptions about how jobless individuals value work, avoid unemployment, and conduct job searches (Altman 2014; Howell and Azizoglu 2011). Significantly, in addition to the works cited in Bivens, Smith, and Wilson, there is new empirical evidence based upon labor market and administrative data that UI benefits and extensions improve job matching and help long-term unemployed individuals remain connected to the labor market.1
State UI benefit recipiency rates reach historic lows in 2014
Figure B presents our calculated benefit recipiency rate for regular state UI programs from 1977 through December 2014. It does not include federal extensions to UI often instituted during recessions.2 It is notable that during the prior lowest point for benefit recipiency in the early 1980s, there were federal benefit extensions available. Given the phase-out of federal benefit extensions at the end of 2013, this figure indicates that UI benefit recipiency is now lower than the previous historical low in 1984. By December 2014, only 23 out of every 100 jobless workers were getting state UI benefits. Because there were federal benefit extensions in place in 1983 and 1984, this means that those exhausting UI benefits in 2014 had less protection from income loss than any cohort of jobless individuals exhausting state UI benefits for several decades.
U.S. unemployment insurance (UI) recipiency rate, 1977–2014
|Date||U.S. recipiency rate (12-month moving average)|
Note: The UI recipiency rate is the share of unemployed workers receiving benefits from regular state programs, and is calculated by dividing the number of weeks compensated by the total number of unemployed persons. It is presented as a 12-month moving average. Shaded bars denote recessions.
Source: EPI analysis of Department of Labor (DOL) administrative data and Current Population Survey basic monthly data
The shaded months of recessions in the figure illustrate the cyclical nature of recipiency in UI programs. Overall, when unemployment peaks around recessions, UI recipiency rises (Vroman, 2011).3 In general, more claims are filed and individuals draw benefits for longer periods as unemployment rises. When unemployment falls, there are fewer workers to file claims and fewer workers with qualifying earnings.4 For these reasons, there is a clear cyclical pattern in benefit recipiency.
Compared with previous recessions, the U.S. labor market in 2014 is at a point in the recovery when we would expect benefit recipiency rates to fall from their peak levels. What stands out about the current low recipiency levels is the degree of their decline and the extremely low levels to which some individual states’ recipiency rates have fallen. We now examine some details regarding these recent declines.
Most short-term unemployed workers do not get state unemployment benefits
Unemployment benefits are paid by trust funds maintained for individual states in the U.S. Treasury and financed with employer payroll taxes (three states also use employee contributions).
Bivens, Smith, and Wilson (2014) explain how the federal-state UI system is designed to work:
The federal-state UI system is designed to enable states to increase the balances of their unemployment trust fund accounts during periods of economic prosperity and low unemployment so that the accounts maintain solvency during economic downturns, when the unemployment rate increases. Conversely, during recessions, when unemployment rises and growth slows, expenditures increase while revenues decrease. In this manner, during economic downturns, the federal and state governments replace a portion of the economic activity that is lost though decreased wages by injecting money in the form of UI benefits into the economy. Moreover, during times of prosperity, states pay less in benefits than they receive in revenues, allowing them to build up account balances … which tend to decline during downturns, when states must pay out more in benefits than they receive in taxes.
But a failure to adequately fund state trust fund accounts during the economic recovery and expansion between 2001 and 2007 set them up for insolvency when the Great Recession hit (Bivens, Smith, and Wilson 2014).
As early as 2011—when the year-round unemployment rate averaged 8.9 percent—some states reacted to record benefit payments from their unemployment trust funds by enacting restrictive legislation designed to cut UI benefit costs. In particular, between 2011 and 2014, nine states abandoned a long-established national pattern of providing at least 26 weeks of potential duration for a state unemployment claim. At the end of 2013, with long-term unemployment still more thandouble what it had been when the most recent federal unemployment extension program was adopted in mid-2008, the federal government let all emergency extensions expire, leaving jobless workers wholly dependent upon state UI programs. As noted, this combination of state cuts and federal inaction meant that at the close of 2014, levels of UI benefit recipiency in the U.S. were the lowest in the program’s history.
The costs of UI benefits are driven by three main factors: the unemployment rate, the benefit recipiency rate, and the share of prior wages replaced by UI benefits. Of these three factors, recipiency rates have shown the biggest variation over the post-World War II life of UI programs (Vroman 2011). Indeed, state UI legislation is often passed during and following recessions and is explicitly designed to cut benefit costs. Not unexpectedly, this legislation then plays a role in falling state UI program recipiency rates (GAO 1993). In this briefing paper, we present clear evidence that recent legislative cuts in duration of unemployment benefits have led to significant declines in benefit recipiency rates in those affected states which exceed the overall declines in recipiency observed in other states.
Many studies have attempted to provide explanations for the observed long-run decline of recipiency during the early 1980s, and these studies provide useful insight into why variations exist between states’ recipiency rates. A report commissioned by the Department of Labor and produced by the Lewin Group (Wittenburg et al. 1999) provided a summary of many such analyses to that time, analyses that primarily focused on explaining the rapid decline of the national recipiency rate from the late 1970s to the 1980s. Among those that studied labor force or unemployment compositions, Blank and Card (1991) found that the decline of unionization explained a substantial portion—25 percent—of the decline between 1977 and 1987, 25 percent. Other labor market variables, such as industry or demographic profiles of the unemployed, had negligible effects according to Burtless and Saks (1984) and Wittenberg’s own analysis. Another substantive explanation was the federal taxation of UI benefits, first enacted in 1978, which significantly reduced the recipiency rates by reducing the incentive to file claims (Anderson and Meyer 1997).
Benefit recipiency rates vary across states. While labor market and other factors contribute to variation in UI recipiency, public policy decisions have played a major role in recipiency declines. The most recent comprehensive assessment of recipiency rates was undertaken by Wayne Vroman in a report for the Department of Labor in 2001. After reviewing prior research, Vroman (2001) combined a regression analysis with site visits to eight states falling on the high and low ends of the range of state recipiency rates. Vroman’s paper found that no single factor explained the differences between low and high recipiency states, but that state laws and administrative practices made “an important contribution” to differences in UI recipiency. In his later 2011 paper, Vroman observed that UI recipiency rates and benefit-wage replacement rates are “strongly influenced by state UI statutes and program administration.” Similarly, Wittenburg et al. (1999) concluded that administrative and policy changes can explain a “large portion” of recipiency declines, but these were difficult to measure as independent factors.
In this paper, we focus on states that have passed restrictive legislative packages that have included cuts in available weeks of state benefits to levels below the customary 26 weeks. We find significant reductions in recipiency that exceed the overall declines in recipiency accompanying falling unemployment rates in other states in recent years.
Given that state UI programs are mainly designed to address short-term unemployment, for this portion of our paper we refine our consideration of recipiency by focusing on benefit receipt by the short-term unemployed (those unemployed 26 weeks or less). These jobless individuals are properly considered the target population for state UI programs. Our calculation of short-term recipiency rates uses a 12-month moving average of weeks compensated divided by a 12-month moving average of the number of short-term unemployed (26 weeks or less) in each state.5 This calculation gives us a short-term recipiency rate reported for each state in Figure C, with those states that cut maximum durations highlighted with red bars. Strikingly, these short-term recipiency levels indicate that regular state UI programs are failing jobless workers who need income replacement and support of work search.
Short-term UI recipiency rate, by state (ordered from highest to lowest), 2014
|State||Short-term recipiency rate, ordered from highest to lowest|
Source: EPI analysis of Department of Labor (DOL) administrative data and Current Population Survey basic monthly data
Many critics of UI programs wrongly assume that the lion’s share of jobless workers get benefits. This is plainly wrong over the history of UI and especially in the more restrictive states.6 The U.S. short-term recipiency rate (black bar) was 34.7 percent in 2014, meaning that over 65 percent of short-term jobless workers did not get state UI benefits. Figure C shows the short-term recipiency rates for 2014 ranged from a low of 14.8 percent in South Carolina to a high of 65.7 percent in New Jersey. In 21 states, 70 percent or more of short-term jobless workers did not get UI benefits in 2014.
Cuts in weeks of available benefits cause significant recipiency declines
From the late 1960s to 2011, all states paid regular benefits for at least 26 weeks. One important potential reason for the recent decline in benefit recipiency is that since 2011 nine states have cut the maximum available number of weeks of regular UI benefit duration: Arkansas, Florida, Georgia, Illinois, Kansas, Michigan, Missouri, North Carolina, and South Carolina. Except Illinois, all these states made other legislative changes to their programs which may have reduced benefit recipiency. Our study does not try to distinguish among the variety of state law changes which accompanied the cuts in weekly benefit duration in these nine states. As a shorthand, we refer only to the cuts in available weeks of benefits and we focus on those states with these cuts in this section of our analysis.
The timing, approach, and severity of the cuts in benefit duration in these nine states have varied.Table 1 shows the new maximum durations in each state and the effective dates of the cuts in weeks. Arkansas and Illinois enacted relatively minor cuts to duration, reducing the maximum from 26 weeks to 25 weeks. Arkansas passed other restrictions along with cutting available weeks. In the case of Illinois, the cut of one week was effective only in calendar year 2012. Michigan, Missouri, and South Carolina all reduced their maximum duration from 26 to 20 weeks. Effective in January 2014, Kansas adopted a sliding scale formula that reduces benefits from 26 weeks to 20 weeks, and finally to 16 weeks as the state’s unemployment rate falls. Kansas paid 20 weeks on claims in 2014 and will pay only 16 weeks in 2015. Finally, the most severe cuts were made in Florida, Georgia, and North Carolina, where the maximum duration depends on a sliding scale formula based upon each state’s unemployment rate, with the variable number of weeks potentially available ranging from 12 to 20 across the three states. With falling state unemployment rates, the impact on available weeks of benefits was significant in these formula-based states. In December 2014, Florida paid a maximum 16 weeks of benefits, Georgia provided 15 weeks, and North Carolina’s limit was 14 weeks. Starting for claims in January 2015, Florida is providing a maximum of 14 weeks, Georgia is providing at most 17 weeks, and North Carolina is providing 15 weeks.
Maximum weeks of unemployment insurance benefits, by states that cut benefits duration in the aftermath of the Great Recession
Note: The number of weeks of benefits available to recipients in states with sliding scales is “determined by the state’s unemployment rate.” In Illinois, the criteria for the duration cut were met in 2012 but not 2013 (meaning 26 weeks were available in 2013). In North Carolina, labor market conditions at the time of implementation were such that the maximum fell to 19 weeks.
Source: EPI analysis of Isaacs (2012) and state laws
Quite predictably, those states that cut available weeks of benefits have experienced faster-than-average declines in UI recipiency rates since the cuts. Figure D compares the short-term UI recipiency rates in eight of the nine states that cut their maximum available weeks (excluding Illinois as its temporary cut did not negatively impact recipiency) with the average UI rate across the 41 states (plus the District of Columbia) that did not cut benefit weeks.
Percentage-point difference between short-term UI recipiency rates in duration-cutting states and states that did not cut duration of benefits
*The figure charts the benefit-cutting states’ short-term UI rates relative to the average short-term UI rate across all other states. Illinois was excluded from the all-other-state average but is not included in this graph because its temporary cut did not negatively impact recipiency.
Source: EPI analysis of Department of Labor (DOL) administrative data and Current Population Survey basic monthly data
Because each state enacted its cuts at a different time, the x-axis in Figure D reflects the months since the cuts became effective in each state. Because we use 12-month averages and the cuts have a delayed effect on new claimants, the effects of these cuts to available weeks typically begin to impact calculated recipiency rates from six to 12 months after the cuts went into effect.7 While these states’ recipiency rates were below the average recipiency rate even before their cuts, all of them now display an even steeper deviation from the rates of all other states. Arkansas, which reduced the duration of UI to 25 weeks, saw a milder drop from 4.7 percentage points below overall rates to 7.7 percentage points below. Kansas’s legislation became effective in January of 2014, leaving only a small sample to observe, but its deviation from the norm dropped from 3.8 percentage points to 4.9 percentage points after the first 12 months. Michigan and Missouri, two states that cut their durations to 20 weeks, which initially kept recipiency rates relatively flat, experienced drops that brought them to 6.6 and 13.2 percentage points below the overall recipiency rate of other states. In January 2015, Kansas’s duration formula dropped its maximum duration to 16 weeks, which will only accelerate its decline in recipiency.
States with more dramatic cuts to their durations also saw rapid declines in their recipiency rates. South Carolina, the fourth state that cut its durations from 26 to 20 weeks, saw one of the biggest declines. South Carolina’s recipiency rate fell from 9.9 to 22.8 percentage points below all other states’ recipiency rates, a decline of 12.9 percentage points. (This likely reflects the fact that South Carolina’s legislative package included several other restrictive measures.) Among the group of states with the most severe cuts in available weeks, both Florida and Georgia already had recipiency rates far below those of all other states. Still, their cuts to maximum durations did correlate with similarly large declines: 8.1 and 5.5 percentage points, respectively. North Carolina, a state whose recipiency rate was only 1.3 percentage points below the average recipiency rate of all other states, saw its rate fall sharply, by 15.7 percentage points so that it now stands 17.1 percentage points below the short-term recipiency rate of all other states that had no cuts to durations. This was the fastest and largest drop of any state.
Further analysis of individual state changes shows that the percentage-point declines in recipiency rates are greatest in those states with the biggest cuts in available weeks of benefits. Table 2compares the changes in the recipiency rate for each of the nine states that cut maximum durations to changes in the the U.S. average recipiency rate over the same timeframes as those cuts took effect in each state. The table breaks down these changes and shows how the magnitude of the rate cuts in nine states with cuts in available weeks depend upon the severity of the duration cuts.
Percentage-point change in short-term UI recipiency rates for states that cut maximum durations and the corresponding changes in the U.S. average
* Effective month defined as the first month in which the cuts are effective for the majority of the month. Unless otherwise indicated, changes extend to the most recent month available, December 2014.
Source: EPI analysis of Isaacs (2012), state laws, Current Population Survey Outgoing Rotation Group microdata, and Department of Labor administrative data
As shown in Table 2, short-term recipiency rates in Arkansas and Illinois, the states that cut maximum durations least, experienced percentage-point declines that were about the same as or below the decline in the average U.S. decline. Illinois actually showed a slower-than-average decline (1.6 percentage points versus 2.7 nationally) over the one year for which it cut maximum durations. Since the enactment of their maximum duration cuts, Kansas, Michigan, Missouri, and South Carolina all experienced higher-than-average declines in their recipiency rates of 1.4, 10.7, 12.2, and 19.8 percentage points, respectively. Compared with the national average, these declines were 1.3, 1.6, 1.2 and 2.3 times the overall national average declines. These results show cuts to available weeks meant a decline in the recipiency rate. Increasingly severe cuts to the duration of benefits also meant more severe declines in the recipiency rate.
Figure E graphically presents these changes as reductions in the share of unemployed workers receiving short-term jobless benefits since the duration cuts went into effect and compares those reductions to the fall in short-term UI recipiency in the U.S. overall over the identical time periods. With the exception of Illinois, all states cutting available weeks of benefits below the norm of 26 have experienced larger reductions in short-term UI recipiency than the national average reduction in recipiency.
Percentage-point change in the short-term recipiency rate in duration-cutting states compared with national average*
* From effective date of legislation in each state
Source: EPI analysis of Department of Labor (2015) and Current Population Survey Outgoing Rotation Group microdata
States cutting weeks of UI have saved some money by reducing the number of weeks of benefits they paid out, but when spread across all covered employees, these savings are minimal, especially when compared with the importance that those benefits have for UI recipients. Bivens, Smith, and Wilson (2014) projected the costs of paying jobless benefits to workers affected by the cuts to duration, excluding Illinois, Kansas, and North Carolina because cuts there had been in effect for a limited time. For the other six states (Arkansas, Florida, Georgia, Michigan, Missouri, and South Carolina) however, the analysis found that the cuts saved $0.37 per covered worker per week on average.8 These savings are small compared with the average weekly benefit amount of $251.61 lost due to these cuts. Bivens, Smith, and Wilson also showed that those states that cut maximum durations did not experience anything resembling a boost in the labor market as some theorize. Rather, both employment growth and employment attachment (as measured by the prime-age employment-to-population ratio) remained about average in those states.
Can we expect states to maintain robust UI programs?
States play a large role in U.S. unemployment insurance programs, especially when it comes to operating regular state UI programs and setting their eligibility rules. Nearly 80 years into the life of UI as a federal-state program, there is considerable evidence that a significant portion of states fall short in terms of benefit recipiency, even when measured by the “low bar” of short-term recipiency rates. While there are economic and demographic differences among the states, from a policy perspective it is hard to defend state autonomy when jobless workers in South Carolina and other stingy states are denied benefits while those in more generous states get UI benefits under identical circumstances. Permitting states to operate restrictive UI programs undercuts the goals of UI and creates competitive pressures on those states with decent programs to engage in a “race to the bottom.” Both these factors undercut the economic and policy rationales for UI.
Other serious problems with state UI programs are easily identified. Although this brief has focused on falling recipiency rates, many states failed to sufficiently build UI trust fund reserves in advance of the Great Recession (Evangelist 2012), requiring them to borrow federal loans and raise payroll taxes while providing a rationale for the restrictive cuts examined here. In particular, larger states have failed to adequately finance their programs, and over time inadequate financing bodes ill for the welfare of UI programs under state control over financing (Vroman 2012).
Adequate wage replacement rates are key to achieving the goals of UI. Some states are unable or unwilling to keep UI programs at reasonable levels. Many states pay low benefits. There were 11 states with maximum weekly benefit levels of $350 or less in 2014 (U.S. Department of Labor 2014), meaning that workers earning more than $700 a week (well below the median weekly earnings) do not get half their prelayoff wages replaced by UI benefits. Average benefits overall were only $315 a week in 2014 (U.S. Department of Labor, 2015a) with average weekly benefits below poverty levels in the poorly performing states.
The recently released FY 2016 federal budget contains one possible federal policy response to the erosion in state UI program performance. Budget details released in early February 2015 by the Labor Department revive the concepts found in the 2009 UI modernization proposal, but reconceive the approach to address many pressing problems with states’ stewardship over UI programs (U.S. Department of Labor, 2015b). Under the first UI modernization program, states were offered federal financial incentives to adopt modest eligibility measures. Over $4 billion in payments were made to state trust funds during the lifetime of the program (March 2009 to August 2012).
Under the proposed UI modernization program, in order to gain access to a portion of the $5 billion in federal incentives available, states would have to agree to five administrative measures. For purposes of this brief, the most critical administrative requirement is that states must pay at least 26 weeks of regular state program benefits. In addition to the required administrative measures, the new version of UI modernization asks states to select two among a range of eligibility options and two options from a set of “work connection” strategies. Regardless of its short-term political viability, the inclusion of the renewed UI modernization concept in the FY 2016 budget proposal offers a potential pathway toward encouraging states to adopt sensible options which will, at a minimum, help states better resist the temptation to adopt cuts like those seen in the worst-performing states. Based upon state responses to the initial version of UI modernization prior to 2012, financial incentives may prove insufficient to improve policy in states that regard UI as a program which is best kept as small and stingy as possible. In that case, a campaign for federal standards for benefits and financing—advocated for decades by supporters of robust state UI programs—should be a central aspect of ongoing UI reform efforts.
While the proportion of jobless workers receiving UI benefits has varied over business cycles and varies between states, in 2014 the overall benefit recipiency rate for state UI programs reached its lowest level in post-WWII U.S. history. Further, this low level of benefit recipiency has come at a time when the U.S. labor market remains unambiguously slack, with significantly more unemployed workers than available job openings. Until 2011 all states provided at least 26 weeks of available benefits in their regular state UI programs. Nine states broke this national pattern beginning in 2011, and our analysis shows that this change had a negative and substantial impact on the UI benefit recipiency rate in eight of those states.
Focusing only on the short-term recipiency rate, we find that (with the exception of Illinois) those states cutting available weeks below 26 weeks had greater reductions in recipiency than the national average decline in recipiency. The states with the most severe cuts (Florida, Georgia, and North Carolina) were also among the four states whose recipiency declines most exceeded the national average. Reducing the duration of benefit receipt is especially detrimental to the unemployed. Clearly, these states ignore the proven advantages of UI programs in providing a countercyclical stimulus, protecting families against income loss, and supporting work searches by jobless individuals.
About the authors
Rick McHugh is an attorney and policy advocate who has worked on unemployment insurance for more than 35 years. He has testified before congressional and state legislative committees, litigated many cases in courts and before administrative agencies, and speaks frequently on UI topics. This is his second publication for EPI. McHugh has a bachelor’s degree from Wabash College and a J.D. degree from the University of Michigan.
Will Kimball joined EPI in 2013. As a research assistant, he supports the research of EPI’s economists on topics such as wages, labor markets, macroeconomics, international trade, and health insurance. Prior to joining EPI, Kimball worked at the Center on Budget and Policy Priorities and the Center for Economic and Policy Research.
1. A recent working paper by Nekoei and Weber (2014) suggests that moderate extensions of UI benefits can improve job quality for those individuals receiving benefits while searching for work. The authors found that increased earnings and revenue from taxing these higher-earning positions more than overcame the increased costs of extended benefits as well as the moral hazard issue arising from any increased duration of unemployment. Bradbury (2014), examining data from 2005 through 2013, finds that the main impact of UI on job transitions is to delay the transition from unemployment to nonparticipation in the labor market, while the transitions from unemployment to employment were “virtually unaffected” by UI program extensions. As a result, the overall effect of extensions in recent years was to raise unemployment rates by keeping jobless individuals from dropping out of the labor market. According to this study, job-finding rates by benefit recipients were distributed across the weeks of unemployment and not closely related to benefit exhaustion. Finally, three authors of papers reviewed in Bivens, Smith, and Wilson (2014), collaborated on a follow-up review of more recent data and found “little or no effect on job-finding but a reduction in labor force exits due to benefit availability” (Farber, Rothstein, and Valletta 2015).
2. Data are as reported by the U.S. Department of Labor. We have calculated benefit recipiency rates using the monthly number of those getting benefits, reported as “weeks compensated,” divided by the monthly number of total unemployed individuals. We opt to use weeks compensated by the alternative weeks claimed because weeks compensated reflect all benefit weeks that are actually paid out, rather than including those claiming benefits who are not paid because they are claiming during a waiting week or other ineligibility period. The monthly figures are smoothed by using 12-month moving averages. While there are other measures of UI recipiency, including the ratio of insured unemployed to the total unemployed, or weeks claimed compared with job losers, these measures show a similar pattern to our measure. Our use of 12-month moving averages shifts the peaks and troughs a few months later in our time series as compared with those reported by others.
3. A large group of nonrecipients of UI programs are those who do not apply for benefits. The reasons for significant non-application rates in state UI programs are not well understood. In this paper, non-applicants looking for work fall into the number of total unemployed in the denominator as with other past analyses of UI recipiency.
4. Within this broad cyclical pattern for recipiency, there are other labor market interactions with UI rules influencing benefit recipiency. Soon after a recession, recipiency increases as laid-off workers with recent wages file UI claims and fewer unemployed workers have quit their jobs. These freshly unemployed workers are more likely to be eligible for benefits compared with those unemployed later in the recovery period. Later in a recovery a higher share of jobless workers may have experienced a prior period of unemployment (reentrants), as well as those entering the labor market (new entrants and nonparticipants).These groups lack the sufficient wage history benefit eligibility as compared with those unemployed earlier in a downturn. And, when unemployment levels are lower, more workers in the ranks of the jobless leave their jobs and some of these job leavers are disqualified for voluntarily quitting.
5. We adopt this short-term recipiency approach from H. Luke Shaefer and Michael Evangelist (2014) who employed it in an analysis of the impact of Michigan’s restrictive 2011 legislation.
6. Vroman (2001, p. 9-10), calculated centered 5-year average recipiency rates using two alternate measures and showed that from 1949 until 1997, recipiency declined from peaks in the early 50s to a trough in the early 80s, then rose modestly in the 90s. Figure B (above) shows that higher recipiency levels in the 90s did not persist, and as this brief has shown, overall recipiency is now below the 80s trough. See also McKenna (2015).
7. Restrictive legislation impacts only new claims beginning on those effective dates. As a result, the effects of the cuts occur only four to six months after the effective date (depending on the severity of the cut in available weeks). Additionally, the use of 12-month averages in the calculation of our recipiency rates has a smoothing effect.
8. A covered worker is an employee on whom the employer is required to pay UI-dedicated state taxes (also known as State Unemployment Tax Acts or SUTA taxes).
Advisory Council on Unemployment Compensation. 1995. Unemployment Insurance in the United States:Benefits, Financing, Coverage. U.S. Department of Labor, Report.
Altman, Morris. 2014. “Insights from behavioral economics on how labor markets work.” School of Economics and Finance, Victoria University of Wellington (NZ), Working Paper.
Anderson, Patricia M., and Bruce D. Meyer. 1997. “Unemployment Insurance Takeup Rates and the After-Tax Value of Benefits.” The Quarterly Journal of Economics, 1997, vol. 112, no. 3. 913–937.
Bivens, Josh, Joshua Smith, and Valerie Wilson. 2014. State Cuts to Jobless Benefits Did Not Help Workers or Taxpayers. Economic Policy Institute, Briefing Paper #380.
Blank, Rebecca, and David Card. 1991. “Recent Trends in Insured and Uninsured Unemployment: Is There an Explanation?” The Quarterly Journal of Economics, 1991, vol. 106, no. 4, 1157–1189.
Bradbury, Katharine. 2014. “Labor Market Transitions and the Availability of Unemployment Insurance.” Federal Reserve Bank of Boston Working Paper No. 14-2.
Burtless, Gary, and Daniel H. Saks. 1984. The Decline in Insured Unemployment During the 1980s. The Brookings Institution. Report prepared for the U.S. Department of Labor, Employment and Training Administration.
Evangelist, Mike. July 2012. Lessons Left Unlearned: Unemployment Insurance Financing After the Great Recession. National Employment Law Project.
Farber, Henry S., Jesse Rothstein, and Robert G. Valletta. January 2015. “The Effect of Extended Unemployment Insurance Benefits: Evidence from the 2012-2013 Phase-Out.” Princeton University Working Paper #586.
Government Accountability Office (GAO). 1993. Unemployment Insurance: Program’s Ability to Meet Objectives Jeopardized. Report.
Howell, David R., and Bert M. Azizoglu. 2011. “Unemployment Benefits and Work Incentives: The U.S. Labor Market in the Great Recession.” Oxford Review of Economic Policy vol. 27, no. 2, 221-240.
Isaacs, Katelin P. August 2012. Unemployment Insurance: Consequences of Changes in State Unemployment Compensation Laws. Congressional Research Service.
McKenna, Claire. February 2015. The Job Ahead: Advancing Opportunity for Unemployed Workers. National Employment Law Project, Report.
Nekoei, Arash, and Andrea Weber. 2014. “Does Extending Unemployment Benefits Improve Job Quality?” Job Market Paper.
Shaefer, H. Luke, and Michael Evangelist. April 2014. The Impact of the 2011 Changes to Michigan’s Unemployment Insurance Program on Unemployed Workers and Their Families. Michigan Unemployment Insurance Project, Report.
U.S. Department of Labor, Employment and Training Administration. 2014. “Significant Provisions of State Unemployment Insurance Laws, July 2014.”
U.S. Department of Labor, Employment and Training Administration. 2015a. “Latest Statistics.”
Vroman, Wayne. 2001. Low Benefit Recipiency in State Unemployment Insurance Programs. The Urban Institute.
Vroman, Wayne. 2011. Unemployment Insurance: Problems and Prospects. National Academy of Social Insurance, Brief #2.
Vroman, Wayne. August 2012. Financing Unemployment Insurance After the Great Recession. Urban Institute Brief # 7.
Wittenburg, David C., Michael Fishman, David Stapleton, Scott Scrivner, and Adam Tucker. 1999. Literature Review and Empirical Analysis of Unemployment Insurance Recipiency Ratios. The Lewin Group, for the Division of Research and Policy of the U.S. Department of Labor Unemployment Insurance Service.