Failing the unemployed
A state-by-state evaluation of unemployment insurance systems
by Maurice Emsellem, Jessica Goldberg, Rick McHugh, Wendell Primus, Rebecca Smith, and Jeffrey Wenger
The U.S. unemployment insurance system, the primary safety net for workers in times of economic recession, is in need of significant repair. The current system, a state-by-state patchwork of policies and provisions, is rife with shortcomings and inequities. Perhaps the most important of these involves the difficulty many workers face in even qualifying for benefits. Unfortunately, those who are eligible to receive benefits sometimes find that the maximum benefit amount does not keep a family from falling into poverty. To make matters worse, unemployed workers and their families certainly aren’t helped by the fact that benefits often run out long before firms begin to re-hire workers. Of course, states could protect workers by extending the benefit duration, but many states have not adopted the provisions necessary to weather an economic downturn like the one the economy is now experiencing.
While these failings of the UI system are troubling during periods of prosperity, they are disastrous in times of economic distress. According to the Bureau of Labor Statistics, from October 2000 to January 2002 unemployment increased 1.7 percentage points (from 3.9% to 5.6%), or 44%. The level of unemployment increased from 5,528,000 unemployed workers in October 2000 to 7,922,000 in January 2002, resulting in 2,394,000 more workers without jobs.
These national rates of unemployment mask considerable differences between regions. Oregon and Washington state have unemployment levels1 above 7.0%, and another 11 states – Alabama, Alaska, California, Florida, Illinois, Kentucky, Louisiana, Mississippi, Nevada, North Carolina, and South Carolina – and the District of Columbia have unemployment rates above 6.0%. Economic forecasts indicate that the national unemployment rate will peak at 6.25% sometime in late 2002 (Greenspan 2002). If the national rate climbs to these heights, then some states may see unemployment rise to nearly 8% before improving.
For a variety of reasons, not all unemployed workers actually receive UI benefits. Nationally, only 43.3% of unemployed workers received unemployment insurance benefits in 2001,2 and women were less likely to receive these benefits than men (40.0% and 45.9%, respectively). In some states the gender gap in UI recipiency is over three times as large as the national average; women in Illinois, Michigan, Ohio, and Washington are as much as 20% less likely to receive benefits than men. The states themselves are responsible for many of the barriers that make unemployed workers ineligible for benefits. In the worst states, fewer than one-third of unemployed workers receive UI benefits, while in the best states more than half do (see Table 1). States with barriers to eligibility, and consequently low recipiency rates, diminish the beneficial effects of unemployment insurance. Denying eligibility diminishes the effect of UI as an income support program and reduces its effects as an economic stimulus. If a state does not have adequate eligibility provisions, then that state’s UI system fails the workers it is supposed to serve.
In addition to being undermined by eligibility problems, the UI system is further hobbled by structural flaws. First and foremost among these concerns is that many state-level extended benefits programs fail to come on line automatically (due to their “triggers” being set too high). Consequently, many workers run out of benefits prior to finding new employment. This problem is especially acute during a recession. More than two million unemployed workers are likely to exhaust their regular weeks of UI benefits in the first six months of 2002, with about one million exhaustions occurring in each of the first and second quarters. On average, 11,000 workers are exhausting their benefits each day, or about 80,000 each week.3 The estimate that two million workers will exhaust their benefits in the first half of 2002 is comparable to estimates made by the Congressional Budget Office, and it represents a sharp increase over the number of workers who exhausted their benefits in the first and second quarters of 2001 (see Table 2).
Exhaustion of benefits
About 1.4 million workers exhausted their UI benefits between September 11, 2001 and the end of January 2002 (see Appendix Table B1). Although many of these workers may have found new employment since exhausting their benefits, a significant portion are no doubt still seeking work and thus would be eligible for additional weeks of benefits if Congress would provide them. As noted above, more than two million unemployed workers are projected to exhaust their benefits during the first six months of 2002 and would potentially be eligible for additional benefits. (For a detailed description of the methodology, see the appendix.)
UI benefits are typically provided for 26 weeks while unemployed workers search for new jobs. In each of the last seven recessions, the federal government has funded additional weeks of benefits for workers facing the end of their period of eligibility. As of early March 2002, the federal government had failed to help unemployed workers in this way. Even adjusting for growth in the labor force since 1973, the number of exhaustees who do not receive any additional weeks of benefits is expected to be larger in the first quarter of 2002 than in the first quarter of any other year since the early 1970s.4
The number of workers who exhaust their benefits is expected to be more than 750,000 higher during the first half of 2002 than it was during the first half of 2001. The increases between the corresponding periods in 2001 and 2002 in the number of workers who exhaust their regular unemployment benefits are expected to be larger than the increases experienced during the recession of the early 1990s, even though the unemployment rate was much higher then.
Although the states cannot be held accountable for the large increases in exhaustions (just as they are not responsible for unemployment), many states were poorly prepared for a recession-induced wave of unemployment. In particular Arizona, Colorado, Connecticut, Hawaii, Massachusetts, Maine, Nevada, North Carolina, Texas, and Vermont are all projected to have exhaustions increase by more than 100% in the fist half of 2002 when compared to the first half of 2001. Of these states, only Hawaii has provided extended or additional weeks of benefits to those who exhaust their regular benefits, and only Connecticut has adopted the optional total unemployment rate trigger for the extended benefits program.
Structural problems with unemployment insurance
Workers are losing both coming and going – many are denied benefits while others see their benefits run out long before the job market rebounds. There are even problems for those who actually qualify for benefits. Most middle-class earners, who receive their state’s maximum unemployment insurance benefit, will struggle to eke out a poverty-level existence from UI. For many this means dipping into savings, using money earmarked for retirement, or increasing debt. For those without any of these resources,
welfare may be their only recourse. Recent research by Jonathan Gruber (2002) indicates that nearly one-third of U.S. families will be unable to replace even 10% of their lost earnings from their savings during a spell of unemployment. For many of these families UI benefits represent the difference between stifling debt and financial security.
Grading unemployment insurance programs state by state
The deficiencies in the state unemployment insurance system result from its highly decentralized structure. The current arrangement allows states to act autonomously in setting eligibility rules, benefit levels and extensions, adequate financing, and taxes. To truly understand the deficiencies of the system, a state-by-state analysis is required. We have chosen critical qualities of the unemployment insurance system – eligibility, benefits, employer taxes, funding adequacy, and recession preparedness – and evaluated them according to each state’s policies. Based on these findings, we have issued a passing or failing grade to every state in each category and overall. (See the methodology section for a detailed analysis of how each category was evaluated.)
The unemployment insurance program is a federal-state partnership, with eligibility for benefits determined at the state level. To qualify for benefits, unemployed workers must meet monetary and non-monetary requirements that vary by state. In simplified terms, the criteria that workers must satisfy are:
- sufficient wages in the past year,
- involuntary separation from employment, and
- availability for work.
Although the principles embodied in these criteria are fair and appropriate, too often these tests result in the denial of benefits to two groups of unemployed workers: part-time workers and workers who have only recently joined the labor force.
Earnings requirements. Eligibility can hinge on a state’s minimum earnings requirements in either the base period or the quarter with the highest earnings from the one-year base period. Base period wage requirements for minimum benefits range from $565 to $3,400, and high quarter wage requirements range from $150 to $2,266,5 though not all states have both base period and high quarter requirements.6
In addition to requiring varying levels of earnings, states also set requirements about when those earnings must occur. In most states, the base period for determining UI eligibility and benefit levels is the first four of the five most recently completed quarters. Under this system, wages earned in both the current calendar quarter (the quarter in which the layoff occurred) and the previous calendar quarter are ignored in determining whether the worker earned enough to qualify for benefits. For example, someone laid off in late December 2001 and who began work in late February 2001 would not qualify for benefits in most states. Ten months of substantial wages does not immediately qualify a recent entrant to the labor force for unemployment insurance benefits in a state that uses the typical base period. Some states use a so-called “alternate base period” that incorporates the most recently completed quarter’s wages.
Non-monetary requirements. In addition to varying earnings requirements, all states require that workers have lost their jobs involuntarily and through no fault of their own. States also require that workers be actively engaged in job search activities and that they be available for work. But states vary in their definitions of involuntary job separation and availability for work. For example, some states would deny a working mother UI benefits if she lost her job because the unavailability of child care prevented her from being able to change her work schedule from first shift to third. Some states also require workers to be available for full-time work, even if the job they lost was part time.
Selected criteria. The criteria in Table 3 were selected to demonstrate how states treat low-wage workers with regard to eligibility requirements for unemployment benefits. To receive a passing grade on eligibility, a state must meet all three criteria as follows:
- Alternate base period. A state is said to have an alternate base period if it considers the most recently completed quarter of wages when determining eligibility and benefit levels for workers who do not qualify under the regular base period. Thirty-eight states and Washington, D.C. do not have alternate base periods, meaning that they ignore up to a half a year’s worth of earnings when determining eligibility for benefits. Recent entrants to the labor market, such as former welfare recipients who leave welfare for the workforce but are the first to be laid off during economic downturns, are disproportionately harmed by the lack of an alternate base period. The standard base period of the first four of the last five completed quarters is a relic of the time when when technology wasn’t sophisticated enough to include the most recent wages; modern computers and information systems now make the alternate base period viable for all states.
- Half-time minimum wage worker. The eligibility of half-time minimum wage workers is based upon 20 hours of work per week at the federal or state minimum wage level (whichever is higher). In eight states, monetary eligibility requirements are set so high that individuals working for 20 hours a week, year-round, at the legal minimum wage do not qualify for benefits.
- Part-time work. The third column lists the states in which people seeking part-time work are considered eligible for UI benefits. Thirty-one states do not provide UI benefits to workers who are not available for full-time employment, even if their previous jobs were part time and paid sufficient wages to meet earnings requirements.
Results. Table 3 paints a dismal picture of the adequacy of the unemployment insurance system in terms of eligibility, with only two states – Rhode Island and Vermont – receiving passing grades. The other 48 states and Washington, D.C. fail to provide basic protections for unemployed workers who may have substantial earnings and work histories. Of the 48 states that receive failing grades on their UI program eligibility criteria, New Mexico, North Dakota, and Utah scored particularly poorly – in fact, none of these states met any of the three criteria.
Failing to meet these eligibility measures means that policies such as extending benefit duration and raising benefit amounts will have virtually no impact on the many workers who can’t even make it beyond this initial hurdle. It is important to note that failing to adopt reasonable eligibility measures has a disparate impact on different groups of laid-off workers. Not counting a worker’s most recent earnings reduces the likelihood that low-income workers will be eligible for benefits. This problem is compounded by those states that fail to allow minimum wage workers working 20 hours per week to qualify for benefits. Of the eight states denying benefits to half-time minimum wage workers, four of them – Florida, New Mexico, North Dakota, and Utah – put workers in a double bind by also failing to count their most recent earnings.
Excluding workers who search for part-time work has a disparate impact on women workers. Since more than 70% of part-time workers are women, states that fail to accommodate workers with part-time hours have effectively adopted a provision that excludes many women from UI coverage despite the fact that taxes are paid on their behalf even when they are not eligible. For workers constrained by family and care-giving responsibilities, this exclusion seems particularly arbitrary.
ibility is the single most important component of the unemployment insurance system, benefit levels are a close second. Paying adequate benefits can mean the difference between moderate hardship and privation. Benefits serve a dual purpose in the unemployment insurance system. First, they provide families the income assistance they need during a period of job loss. Without these benefits poverty rates among the jobless would be considerably higher (Danziger and Gottschalk 1990). Secondly, the money put into the economy by the unemployment insurance system acts as a significant economic stimulus. Estimates indicate that, in the absence of UI benefits, recessions (as measured by a real decline in gross domestic product) would have been 15% deeper (Chimerine et al. 1999)
While the importance of UI benefits is clear, benefit adequacy, especially for those with low earnings, is ambiguous. Over time, little has changed in the way state systems calculate benefits, while much has changed within the U.S. labor market, especially in terms of U.S. poverty policy. This change in policy, initiated by Congress in 1996, requires the poor to work in the paid labor market. Since many of these workers may no longer be able to rely on welfare in times of economic distress, it is incumbent on the unemployment insurance system to cover the holes in the safety net.
Yet replacing nearly half of a poor worker’s lost income is very different than replacing half of a middle-income worker’s earnings. For those hovering on the brink of poverty while working, replacing half of their lost income means certain poverty. With more welfare recipients and low-income workers filing for benefits, a minimum benefit that replaces two-thirds of their lost wages makes more sense. Making benefit payments progressive in this way will help these workers pay for adequate food, clothing, and shelter.
Additionally, there are many states that base the duration of benefits (not just the amount of weekly benefits) on previous earnings. Many low-wage workers with lower incomes will receive far fewer weeks’ worth of benefits. The extent of this “week reduction” is considerable: in 2001, 43.1% of workers who exhausted their UI benefits did so before receiving 26 weeks’ worth of benefits. This means that only 56.9% of UI recipients who exhausted their benefits initially received 26 or more weeks’ worth of benefits. This implies that many of those who lost their jobs after September 11 have now run out of UI benefits.
Selected criteria. In the United States unemployment insurance replaces approximately 47% of an unemployed worker’s lost wages.7 These national numbers mask considerable state-to-state variation. In analyzing the adequacy of unemployment insurance benefits, we examined four components of UI benefit generosity:
- Indexed maximum benefit amount. A state indexes its maximum state benefit if it has a formula that automatically adjusts the maximum UI benefit based on average earnings within the state,
- Benefits exceed poverty level. A state has benefits that exceed the poverty level when the maximum weekly benefit amount is sufficient to prevent a one-parent, two-child family from living in poverty ($274.40/week).
- Wage replacement rates for minimum wage workers exceed 50%. If weekly benefits for a full-time, full-year worker earning the minimum wage do not replace 50% of that worker’s lost income, then the state fails to meet this criteria. In cases where the state’s minimum wage is higher than the federal minimum wage, the state wage is used as the measure.
- Wage replacement rates for median wage workers exceed 50%. If weekly benefits for a full-time, full-year worker earning the median wage do not replace 50% of that worker’s lost wages, then the state is said to fail this criteria.
Table 4 shows how each state measures up according to the above criteria. States that fail to index their maximum benefit find that, over time, the percentage of lost income replaced by UI benefits steadily declines. The maximum benefit is “stuck” at one level while inflation and productivity increases are raising wage levels. Until recently California had not raised its maximum weekly benefit from $230 per week in 11 years. By 2001, California was replacing an average of just over 25% of workers’ lost income.
Poverty-level benefits are problematic for many states. We examine the maximum weekly benefit that an unemployed worker receives in each state to assess whether maximum benefit recipients can keep themselves and a two-child family out of poverty. With more low-wage workers unable to rely on other forms of cash assistance, the poverty-fighting effects of unemployment insurance will become increasingly important.
Finally, we examine whether or not minimum wage and median wage workers have more than 50% of their pre-unemployment wages replaced by UI benefits.
Results. Overall, nine states fail in terms of benefit adequacy. To receive a failing grade a state must have received a “no” in either of the replacement rate categories and a “no” in either of the other categories (poverty benefit or indexing). In this way we prevent double-counting of the replacement rate factor. A sizable portion of the U.S. workforce lives in these nine states, which include Arizona, California, Maryland, and Missouri.
In addition, most of the states that failed on benefit adequacy have adequate trust funds. Five of these nine failing states have trust funds with average high cost multiples above 0.75. Arizona in particular stands out as a state that, after 12 months of recession, still had a large surplus of money in its UI accounts but provides poverty-level UI benefits even to workers who receive the maximum benefit amount.
Many states fail to provide adequate UI benefits for any worker (regardless of prior earnings). In these state maximum weekly benefits are insufficient to keep families out of poverty. Eight states have maximum benefit amounts that result in a poverty-level standard of living – Alabama, Arizona, Louisiana, Mississippi, Missouri, Nebraska, South Carolina, and South Dakota. There are an additional six states – Florida, Georgia, Maryland, New Mexico, Tennessee, and Wyoming – where maximum benefits lift a family of three just $10 per week above the poverty line. In all, 14 states leave working families near poverty regardless of their previous earnings.
What is the UI tax burden borne by employers, and how have the revenues of the UI system responded to increases in wages and inflation? These questions help evaluate whether states properly balance revenue requirements (taxes) and expansion of UI benefits. Table 5 shows that the UI tax burden on employers is, in fact, modest. In addition, it’s clear that state tax policies have failed to keep pace with wages, with most states lagging behind appropriate measures of wage adjustment. Finally, and perhaps most significantly, there is a striking relationship between those state UI systems that are the most restrictive toward workers and those that impose the least tax burden on employers.
The UI system is funded by two separate payroll taxes, one federal and one state. The federal payroll tax on most employers (called FUTA, for the Federal Unemployment Tax Act) is 0.8% on the first $7,000 that each worker earns, or a maximum of $56 per worker. Unlike Social Security and other payroll taxes, the amount of wages that are taxed for federal UI purposes is not periodically adjusted to account for inflation, and the $7,000 minimum has
not been increased by Congress since 1983. The FUTA revenues are deposited in a federal trust, now totaling $45 billion, that pays for the administration of the state UI programs, for federal extensions of unemployment benefits, and for loans to the states.
The state UI tax pays for the costs of the benefits provided to workers. The state funds totaled about $ 51.6 billion at the end of 2001. The rate of the UI tax is determined by the state, as is the amount of each worker’s wages that are taxed, known as the “taxable wage base.” The rate of the state tax also varies for each employer. The rate can increase up to a designated point as the employer lays off more workers, a practice that is known as “experience rating.” According to federal law, the states must tax at least the first $7,000 of each individual’s wages. Beyond that, there is almost no federal role in the system of taxing unemployment benefits. While the business community often argues that UI taxes cut into business profits, the empirical research indicates that most of the UI tax is passed on to workers in the form of lower wages (Anderson and Meyer 1994).
Selected criteria. The criteria used to evaluate how well states have managed the tax structure of their UI programs are:
- State taxable wage base. The amount of earnings subject to taxation is known as the state taxable wage base. The lower the taxable wage base the larger the percentage of total tax burden falling on lower-income workers and their employers.
- Wage base indexed to state wage. As nominal wages increase so too should the taxable wage base. If the wage base is not increased, benefit payments will increase while tax revenues remain stagnant, paving the way for trust fund insolvency. Additionally, a higher proportion of the tax burden is shifted onto lower-income workers and their employers.
- Tax rates. We evaluate the percentage of the total tax burden, that is, the tax rate as a percentage of total wages. While this masks the fact that most employees pay only on the wage base, it allows us to examine the effective tax rates for most employees.
As Table 5 illustrates, the amount of earnings taxed can vary significantly from state to state, with 11 states taxing at only the federal minimum base of $7,000. As of 2002, 21 states are below the average state taxable wage of $10,342. The state system of taxation is thus highly regressive toward smaller employers because these employers typically pay less in wages. The system then favors large employers that end up paying much less in UI taxes as a share of their total payroll.
In order to evaluate whether measures are in place for UI taxes to keep pace with wage gains, as with the Social Security tax system, we look to whether a state requires the taxable wage base to be indexed as a matter of law. In fact, most states (33) do not, thus no increase in the taxable wage base occurs unless it is legislated in a given year. Such legislation is difficult politically given the lobbying influence of the business community.
Any proposed increase in the taxable wage base at the state level is typically met with a counter-proposal by employers to restrict unemployment benefits, thus accounting for the resistance on the part of the business community to index the taxable wage base. Not surprisingly, when states index their taxable wage base, the amount of earnings taxed is much more in line with average earnings in the states. States that index their taxable wage base levy taxes on an average of $19,400 of earnings, while states that do not index their wage base levy taxes on an average of $8,600.
Finally, Table 5 compares the average tax rate in each state for 2001, that is the average tax rate paid by employers after taking into account experience rating. The average state tax rate for 2001 was 0.5% of total wages, which is lower than the rate has been in any year since the data collection on this series began in 1950 (Baldwin 2001). In nine states (Arizona, Colorado, Georgia, Nebraska, New Hampshire, Oklahoma, South Dakota, Utah, and Virginia) and the District of Columbia, employers paid 0.2% or less of their total wages in unemployment taxes.
Results. This is another area in which there are widespread problems in the way that the states have managed the tax structure of their UI programs. Nineteen states have failed on two of the three measures described above. Even more alarming is that eight of these states (Arizona, Indiana, Mississippi, Missouri, Nebraska, South Carolina, South Dakota, and Tennessee) fail on all three measures.
Finally, 15 out of the 19 states that earn failing marks on employer taxation also rank below average in terms of the proportion of the unemployed collecting unemployment benefits. Indeed, the average recipiency rate in these states was just 34.4% (compared to 43.3% nationally), illustrating the inequity between the treatment of employers and the treatment of the unemployed workers in these states.
State trust fund solvency
Another important characteristic in determining the strength of a state’s UI program involves the advance financing of unemployment insurance through accumulated trust fund reserves. Measuring the solvency of a state’s UI program is shaped by a combination of objective factors, risk evaluation, and value judgments. While somewhat obscure and technical, solvency is important in determining the overall health of UI programs. Less solvent states have demonstrated an unwillingness to adopt more generous benefits and less restrictive UI program eligibility. When faced with financial challenges during a recession, less solvent states are more likely to be tempted to restrict their UI programs in conjunction with any tax increases they are forced to impose on their employers (Vroman 1998, 5-23). For these reasons, adequate UI trust fund solvency is a significant issue for protecting the interests of unemployed workers and the health of the economy.
All state UI programs impose payroll taxes on employers to finance their UI benefits.8 State UI contributions are deposited in trust fund accounts in the federal treasury, and they are drawn down by the states solely to pay UI benefits. States that have inadequate trust funds to pay UI benefits during an economic downturn must borrow funds and repay those debts with higher taxes and/or benefit reductions or restrictions.
Social insurance programs like UI were designed for accumulating trust funds in advance of the payment of benefits. Since at least the 1950s, there has been some controversy about how much money states should keep in their UI trust funds (Haber and Murray 1966, 379-96). Two trends have been particularly important. First, state UI trust funds have declined in magnitude in comparison with the size of the workforce and the growth in wages since the late 1980s. Second, the amount of reserves generally considered prudent has diminished over the years (Vroman 1998, 10-12, 14-18). Despite these trends, states overall have maintained sufficient UI solvency to weather the current recession and its aftermath. As of December 31, 2001, the 51 states (including D.C.) had $51.57 billion in their UI trust funds (U.S. Department of Labor 2002). As a result, the great majority of states have adequate funds to provide benefits through the current economic downturn, although a fair number need to improve solvency over the longer term.
Selected criteria. We use two measures of UI trust fund solvency:
- Average high cost multiple (AHCM). The AHCM measures a trust fund’s reserves. A reserve large enough to pay benefits without additional revenue for a year, during an average recession, would be equal to one. We use an AHCM of 0.75 (or nine months) as our pass/fail cutoff, with states having AHCMs above 0.75 getting a passing grade and states less than this level receiving a failing grade. Additionally, any state with an AHCM of less than 0.
5 is automatically given a failing grade for overall solvency.
- Decline in UI tax rates. For this criteria we consider states that lowered tax rates between 1994 and 2000 faster than the national average. This analysis uses taxes as a percentage of total payroll as the tax measure. This measure gives a complete picture of the total tax burden.
While most states below 0.75 AHCM will not borrow money as a result of this recession, barely avoiding bankruptcy cannot be an acceptable standard of solvency. Since we are grading the states’ UI solvency in terms of pass/fail, a 0.75 AHCM after nine months of recession (measures are calculated through the end of 2001) is not an unreasonably high standard. Thirty-three states pass this measure of solvency.
In the 1990s, state policy makers in a majority of states chose to reduce UI taxes rather than to build up UI trust fund solvency. States went about reducing UI taxes in the 1990s in two ways. Some states simply let UI taxes fall as a consequence of the good economy, which automatically lowered experience-rated payroll taxes as UI claims fell.9 Other states were more aggressive and passed outright UI tax cuts during the 1990s.10
In this second key solvency measure, we identify states that have reduced payroll taxes significantly following the early 1990s recession. Under this measure, states fail to make the grade if payroll taxes were reduced faster than the national average rate of state UI tax reductions from 1994 through 2000. Nineteen states fail by this measure of solvency (note that many of the states that received a passing mark also had UI tax reductions in the 1990s, though not big enough ones to exceed the national average).
Results. We scored states’ overall UI solvency by giving a failing grade to states that:
- either had an AHCM less than or equal to 0.50, or
- had an AHCM below 0.75 and cut UI taxes more than the national average.
In this fashion, we score states by combining their trust fund balances with a rough reckoning of their UI tax efforts (Table 6). Using this method, 12 states failed the grade on solvency, and 38 states and the District of Columbia passed. This result indicates that, while overall state UI solvency is sufficient to avoid borrowing from the federal trust fund, some states should improve UI solvency in the coming years.
Individual states vary considerably in terms of UI trust fund solvency. Eighteen states have AHCMs below our measure of 0.75 (nine months). Texas and New York have already advised the U.S. Department of Labor that they will borrow from the federal UI trust fund in the near future to ensure the continued payment of UI benefits in those states. Depending on the unemployment situation in the coming months, an additional three or four states may be forced to borrow in the next year. Illinois, Minnesota, Missouri, and North Dakota are likely candidates for trust fund borrowing in the foreseeable future.
A larger number of states are on the opposite end of the solvency scale, with 33 states having AHCMs above the 0.75 measure and 23 states with AHCMs over 1.0. States with considerably higher AHCMs include some with restrictive UI eligibility rules and/or less-than-adequate benefit levels despite having more-than-adequate balances in their trust funds. Arizona, Florida, Georgia, Louisiana, Mississippi, New Hampshire, New Mexico, Oklahoma, Utah, Virginia, and Wyoming have well-above-average UI trust fund solvency and failing grades on most other aspects of their UI programs. To a large degree, trust fund balances in these 11 states are higher than average because their UI programs are less generous than average. Unemployed workers and their advocates in these states have an especially strong argument for expansion of UI eligibility and increases in UI benefit levels.
Eighteen states have failing AHCM grades while 19 states dropped their UI payroll taxes in the 1990s more than the national average. Of these 19 tax-cutting states, seven (Arkansas, Connecticut, Michigan, Minnesota, Missouri, New York, and Ohio) combined larger-than-average UI tax cuts with AHCMs below 0.75. Over the longer term, this is not a recipe for UI solvency.
As a matter of UI policy, we need to address solvency in new ways. Throughout the ongoing debate about UI solvency measures, states have opposed setting a federal solvency standard. Indeed, imposing a solvency standard, by itself, is potentially harmful because states can reach solvency by restricting their UI programs. This is already the case in New Mexico and New Hampshire, both among the nation’s most solvent UI trust funds and more restrictive UI programs.
The Advisory Council on Unemployment Compensation, a bipartisan federal panel convened under former President Bush in the early 1990s, recommended that the federal government adopt an AHCM of 1.0 as a goal for states’ trust funds and that the federal government use enhanced interest payments on state UI trust funds to reward states exceeding this goal. In addition, the council advised that states forced to borrow to maintain UI benefit payments get preferential interest treatment on loans if they were making progress toward meeting the federal solvency goal (Advisory Council on Unemployment Compensation 1995, 8-12).
Given the importance of UI solvency to ensuring that UI programs protect the economic security of unemployed workers and stimulate the economy during recessions, these modest steps should be adopted. All stakeholders should be able to agree on measures that will increase incentives for states to act responsibly and decrease the pressures on states to ignore UI solvency for perceived short-term political and economic development advantages.
Recession preparedness is, of course, the centerpiece of the unemployment insurance system. Making unemployment benefits available to workers during economic hardship as well as providing the economy with billions of additional dollars of stimulus are two of the main functions of the program. Thus, when only a limited number of workers collect UI in times of recession, and when those who collect receive less in lost wages, then the core functions of the program have been compromised.
During recessions, when work is much harder to find, more workers than usual collect benefits for longer periods of time and thus more workers also exhaust their UI benefits. Without access to extended UI, these families have to deplete their savings and, in many cases, rely on public assistance funded by the state and federal governments. According to the Congressional Budget Office (1990), when extended UI benefits are available as many as one-quarter fewer workers have to rely on public assistance.
One of the fundamental lessons learned in previous recessions is that states cannot expect the federal government to enact UI extensions in a timely fashion. One year into the current recession – and six months since September 11 – the vagaries of the political process left an estimated 1.8 million workers and their families without access to benefits beyond the standard 26 weeks.
To their credit, since September 11, some states have taken responsibility themselves and passed temporary extensions (as in the case of Wisconsin and Hawaii). However, many other states failed to enact similar legislation, due to competing demands on revenue during a recession. In contrast, some states enacted legislation years ago to proactively plan for a recession. As a result, these better-prepared states were able to make extended benefits available when needed and without delay.
Selected criteria. To analyze how well each state’s UI system functions during a recession, we examined three facets of each state’s UI pro
gram. Each of these criteria measures the extent to which a particular state’s UI system has “automatic” provisions in place that extend benefits to workers or improve access to benefits.
- State adoption of a total unemployment rate (TUR) trigger. States that adopt this optional trigger automatically receive federally funded UI benefits when the total unemployment rate exceeds 6.5%.
- State-funded extended benefits. Some states have special, solely state-funded programs that extend benefits during periods of recession. Some of these programs are automatic in that workers from particular industries or those participating in training receive additional weeks of benefits.
- Eliminated or modified waiting week provision. Most states have a provision in place that prevents workers from receiving UI benefits during their first week of unemployment.
Overall, we count as passing those state that have adopted the TUR trigger for extended benefits or have a state-funded provision for benefit extensions.
While Congress and the president are called upon in every recession to extend unemployment benefits, the automatic extended benefits (EB) program has been in effect since 1970. However, the EB program is fundamentally flawed because most states never experience the increase in unemployment necessary to trigger this program. As a result, only three states during the current recession (Alaska, Oregon, and Washington) now qualify to provide the standard 13 weeks of additional UI under the EB program. Oregon and Washington qualify because they have adopted the optional TUR trigger. This situation prompted the Advisory Council on Unemployment Compensation to conclude that the “Council is unanimous in the view that there is a pressing need to reform the extended benefits program” (1994, 10-11).
There is, however, one key provision of the EB law that allows states to adopt an optional trigger that would make the program more available in high unemployment states. Rather than rely on the more restrictive measure of unemployment that most states fail to reach even during recessions, states may adopt a less restrictive trigger formula. Specifically, under the less restrictive formula, the EB program is available when the total unemployment rate (TUR) reaches 6.5% in the state and the rate has increased by at least 110% over either of the past two years.11 While 6.5% unemployment is still a high threshold (and there are proposals in Congress to reduce the rate), as of December 2001, 14 states had unemployment rates at or over 6.0%.
Results. As indicated in Table 7, only eight states (Alaska, Connecticut, Kansas, New Hampshire, Oregon, Rhode Island, Vermont, and Washington) have adopted the optional TUR trigger formula for extending UI benefits. In early January of this year, two of these states (Oregon and Washington) reached the required 6.5% threshold, making an additional 13 weeks of UI available to thousands of workers in the state. Under the more restrictive EB formula, Oregon and Washington would not have qualified for an extension when they did.12 Louisiana, Mississippi, Nevada, and North Caolina are examples of states where, if the TUR trigger were adopted, workers would now be collecting extended benefits.
While the reasons may vary depending on the state, most observers agree that the optional EB formula has not been adopted in more states because the funding for EB benefits is shared equally with the states. In contrast, most of the federal extensions that have passed have been 100% federally funded.
Another option available to the states is to simply fund their own extension of unemployment benefits without relying on the federal program. As described above, Hawaii and Wisconsin recently enacted temporary extensions in response to the events of September 11, and other states are considering similar bills. However, a small number of states (seven) have adopted permanent features of their UI programs that provide additional unemployment benefits in certain circumstances. Most of these programs provide 13 to 26 weeks of additional UI to workers who have been displaced from work in selected industries and are participating in approved training.
Finally, some states continue delayed payment of UI for the first week for workers who qualify for UI. This effectively reduces benefits by one week for all workers (except those who exhaust their benefits – they merely experience a delay). For all unemployed workers, it is a hardship not to collect their unemployment for each week that they qualify, but it is especially unfair for workers seeking to find work when jobs are scarce. As Table 7 shows, only 17 states have eliminated or modified the first-week waiting period.13
This analysis illustrates the disturbing reality that very little has been done at the state level to specifically accommodate workers who are long-term unemployed as a result of a recession. Table 7 shows that 34 states fail to provide even one of the limited forms of recession relief described above. These states include: Alabama, Arizona, Arkansas, Colorado, Delaware, Florida, Georgia, Idaho, Illinois, Indiana, Iowa, Kentucky, Louisiana, Maryland, Minnesota, Mississippi, Missouri, Montana, Nebraska, Nevada, New Mexico, North Carolina, North Dakota, Ohio, Oklahoma, Pennsylvania, South Carolina, South Dakota, Tennessee, Texas, Utah, Virginia, West Virginia, and Wyoming.
Reading the report card
Overall, we find many states to be lacking many of the basic protections that were sought by those who developed unemployment insurance. To a large extent the most significant problem with unemployment insurance is accessibility. Even during non-recessionary periods access to UI benefits remains problematic. During a recession the problem of access is coupled with general lack of preparedness. Many of the unemployed are caught in this bind, with nearly 58% unable to receive benefits. Those that do collect benefits find that they are inadequate to make ends meet and that they run out before most workers find a job.
For a better sense of how well the states are doing in terms of protecting unemployed workers, we have given each state an overall pass/fail grade that takes into consideration all of the criteria discussed above:
Overall failing grade
While most states fail on at least one measure of UI adequacy it is interesting to note the number of states that fail on three or more of these measures. In large part these measures capture the extent to which state UI systems have systematically low performers. Column 1 of Table 8 indicates that 23 states fail on at least three out of five of the major categories (eligibility, benefits, taxation, solvency, or recession preparedness
Overall failing grade
While most states fail on at least one measure of UI adequacy it is interesting to note the number of states that fail on three or more of these measures. In large part these measures capture the extent to which state UI systems have systematically low performers. Column 1 of Table 8 indicates that 23 states fail on at least three out of five of the major categories (eligibility, benefits, taxation, solvency, or recession preparedness).
The most common measures that states failed were eligibility (48 states and Washington, D.C.) and recession preparedness (34). In light of the low recipiency rates and the number of workers exhausting benefits, the high numbers of failures in these categories is troubling. Additionally, as noted before, 43.1% of workers
run out of benefits prior to the standard 26 weeks. Other measures did not inspire confidence in the state’s unemployment insurance system; 19 states fail on adequate tax measures, while nine states fail to provide adequate benefits.
Some states restrict eligibility but have highly solvent UI trust funds. These state UI systems get ranked as inadequate because they have the means but lack the policies that would make their systems more accessible.
We find nine states and the District of Columbia have inadequate UI systems, including Arizona, Florida, Louisiana, New Hampshire, New Mexico, Olahoma, Utah, Virginia, and Wyoming. Sadly, these states have the means to alleviate part of the income loss and privation that some families will experience during this recession but fail to enact the policies necessary to avert these problems.
Many states consistently provide tax incentives to businesses while providing negligible benefit expansions to workers. We classify these states as inequitable if fewer than one-third of the state’s unemployed workers receive benefits and the state failed the employer tax provisions measure.
We find that eight states and Washington, D.C. fail on the measure of equity. These states include Alabama, Arizona, Florida, Maryland, New Hampshire, South Dakota, Texas, and Virginia. In an era of fiscal responsibility the mere specter of inadequate funding is enough to eliminate the possibility of expanding UI benefits. We also find that three states were on the brink of failure — Colorado, Georgia, and Nebraska — failing all of the employer tax provisions while less than 35% of these state’s unemployed received UI benefits.
In this report we analyze five critical components of the unemployment insurance system: eligibility, benefits, taxation, funding, and recession preparedness. In each major category, we examine three or four measures. These measures were chosen based on three criteria: (1) the measure directly relates to the major category, (2) the measure has a significant impact on workers and their experience with the UI system, and (3) information was readily available from a reliable source. These measures were chosen as the minimal thresholds for an adequate, fair, and equitable system of unemployment insurance.
In all of the charts, a “no” indicates that the state in question does not provide some provision of unemployment insurance that we deem important. Additionally, many states receive a “pass” in our measure that are only marginally better than those receiving a “fail.” We encourage the reader to interpret states with passing grades with caution. In many cases these states pass but are nevertheless deficient; if we were using a standard grading scale many states would pass with a “D.”
Being eligible for benefits is the lynchpin of this analysis. For those unemployed workers who remain ineligible for benefits, improvements in benefit adequacy, benefit extensions, solvency, and taxation are moot. Consequently, we have developed a “bright line” standard for systems adequacy. Adequate UI systems should provide all the following: use of a worker’s most recent earnings in determining eligibility; eligibility for minimum wage workers who work for full year at 20 hours per week; and eligibility for workers who seek part-time work.
In determining the adequacy of UI benefits, we considered four factors. First, we determine whether a state indexes its maximum benefit to the state wage. In this way, we determine if states automatically raise benefits as wages grow. Secondly, we examine whether a one-parent, two-child family that receives its state’s maximum UI benefit will live in poverty. Finally, we determine if benefits replace 50% of lost earnings for full-time, full-year workers earning their state’s minimum wage and their state’s median wage.
We consider three measures of business taxation for UI.14 While UI taxes are never popular, they are unavoidable. And given the economic evidence that part of the tax burden is shared by employees, the equity of the tax system matters to lower-wage workers. Systems that fail to collect sufficient revenue during periods of economic expansion will not be able to expand benefit coverage during periods of recession. Too often fiscal restraints are used as a reason not to adequately reform the system. We focus on three measures of tax adequacy. First, how do states compare in their overall tax burden? Second, we examine those states that have moved beyond the federal minimum of $7,000 taxable wage base. Third, we assess whether or not states have indexed their taxable wage base to state wages.
Eligibility, benefits, and taxation come together around the issue of adequate UI funding. UI programs with inadequate tax systems but that nevertheless have considerable money in trust are likely maintaining trusts by limiting worker eligibility to unemployment insurance or paying only minimal benefits. To measure funding adequacy we use the average high cost multiple (AHCM) provided by the U.S. Department of Labor. This summary measure indicates how long the money in the state’s UI trust fund can pay benefits during an average recession. In non-recessionary years, the Advisory Council on Unemployment Compensation recommended that states hold one year’s worth of UI benefits in trust. With this recession well under way, we expect states to have approximately nine months of benefits left in trust.
We focus on how well prepared a state is to deal with a recession. During a recession, many workers exhaust their UI benefits prior to finding employment. Due to congressional changes in the automatic provision of the extended benefits program, many states have come to rely on special legislation passed by Congress during recessions to extend benefits. In examining the recession preparedness of the states’ UI systems, we examine those states that have developed optional state-level triggers for extended benefits. We also examine which states limit or have no “waiting” week prior to receiving unemployment benefits. Finally, we examine those states that have developed state-level benefit extensions or supplements.
While we have tried to provide a comprehensive overview of the states’ unemployment insurance systems, there were areas that we could not analyze due to lack of resources and available data. Most importantly, we did not analyze the extent to which states have adopted policies that make UI more accessible for workers. In particular we did not analyze the extent to which states have adopted UI policies that make it easier for workers who leave work as a result of domestic violence, relocation of spouse, or a change in work hours. In many states workers who leave work for these reasons are ineligible for benefits. Additionally, we were unable to analyze a number of administrative issues, such as states that have large backlogs of contested claims. For an excellent recent analysis of some of these administrative issues, see Vroman (2001).
Appendix A: Technical notes
Recipiency rates are calculated by dividing the total claims made against the UI system by the total unemployment rate for each state. This calculation overestimates the percentage of claimants receiving UI benefits; between 10% to 15% of claimants will be denied benefits. Claims data for calendar year 2001 are provided by the U.S. Department of Labor (report ETA-203). Total unemployment data are from Current Population Surveys January-December 2001. New Jersey did not provide claims data by sex.
A recent analysis by the Center on Budget and Policy Priorities predicted that more than two million unemployed workers will exhaust their regular unemployment insurance benefits during the first half of 2002, a figure that was confirmed by the Congressional Budget Office. This further analysis calculates expected exhaustions in each state.
The earlier analysis used the national number of people who first received UI benefits during the third and fourth quarters of 2001 and predicted exhaustion rates to determine the number of people expected to exhaust their benefits in the first half of 2002. Using the same formula and the number of people in each state who first received UI benefits during the third and fourth quarters of 2001, one can estimate the number of workers in each state who will exhaust benefits in the first half of 2002.
The earlier analysis predicting the national number of exhaustions assumed that the national exhaustion rate in the first quarter of 2002 would equal the national exhaustion rate in December 2001, and that the national exhaustion rate in the second quarter of 2002 would equal the national exhaustion rate in the fourth quarter of 2001. These assumptions followed recent seasonal patterns. One method we used to predict state exhaustions was to make the same assumptions about the state exhaustion rates as were made about the national rate; that is, we assume that in each state the first quarter 2002 exhaustion rate will equal the state’s December 2001 exhaustion rate, and the second quarter 2002 exhaustion rate will equal the state’s fourth quarter 2001 exhaustion rate.
But because of variations in local labor markets, the seasonal pattern of the exhaustion rate is not the same in every state as it is nationally. For example, in some states the second quarter exhaustion rate is typically higher, not lower, than the first quarter exhaustion rate. So another method we used to predict state exhaustion rates was to assume that the recent increase in exhaustion rates will be maintained over the next six months. Using that method, the exhaustion rates for the first and second quarters of 2002 in each state are predicted assuming that the percentage increase between that state’s fourth quarter 2000 and fourth quarter 2001 exhaustion rates will also be true for the states’ first and second quarter 2002 exhaustion rates. This is probably a conservative assumption because exhaustion rates typically continue to accelerate during a recession.
Both methods of estimating state exhaustion rates yield state exhaustion levels that, when summed up to the national level, are consistent with the prediction of two million exhaustions nationally during the first half of 2002. Together, these two methods provide a range of exhaustions for each state. We then chose the midpoint of that range, as shown in Table 2. Data for three states (Maine, New Hampshire, and North Dakota) are not available because of data irregularities. In New Hampshire, the UI system functions on a uniform benefit calendar, so the vast majority of exhaustions come at a single point in the year.
• As shown in column four of Table 2, nine states (Arizona, Colorado, Connecticut, Hawaii, Massachusetts, Nevada, North Carolina, Texas, and Vermont) would have more than twice as many exhaustions in the first half of 2002 as they had in the first half of 2001.
• Another eight states (Delaware, Florida, Georgia, Illinois, New York, Oklahoma, South Carolina, and Virginia), as well as the District of Columbia, would have increases of 85% and 100% in the number of exhaustions between the first half of 2001 and the first half of 2002.
Alternate base period: Data for alternate base periods are from Comparison of State Unemployment Insurance Laws (2001, table 300) and EPI/NELP analysis.
Minimum wage worker eligible: EPI/CBPP analysis based on state-level eligibility rules. In determining whether minimum wage workers are eligible for UI, three factors are important: (1) total base period earnings, 2) high quarter earnings, and 3) state-level minimum wages. Base period and high quarter earnings data are from U.S. Department of Labor Summary of State Unemployment Insurance Laws (table 301) and EPI analysis of state laws. Minimum wages available from U.S. Department of Labor. (http://www.dol.gov/dol/esa/public/minwage/america.htm).
Part-time worker eligible: NELP analysis of state UI laws. See McHugh, Segal, and Wenger (2002).
Maximum benefit indexed to state wage: EPI analysis of state unemployment insurance law.
State maximum weekly benefit greater than poverty: EPI analysis of official federal poverty thresholds and state UI benefit maximums. Poverty thresholds from U.S. Census (Poverty Thresholds for 2001, by Size of Family and Number of Related Children Under 18 Years). Maximum UI benefit from U.S. Department of Labor (Significant Provisions of State Unemployment Law, January 2002). Some states provide additional benefits based on family structure. These supplements, known as dependent allowances, are used in calculating whether the state’s maximum benefits raise a family above poverty. States with dependent allowances include Alaska, Connecticut, Illinois, Iowa, Maine, Maryland, Massachusetts, Michigan, New Jersey, Ohio, Pennsylvania, and Rhode Island.
Benefit amount for a minimum wage worker: EPI analysis of state unemployment insurance benefits. Minimum wages available from U.S. Department of Labor (http://www.dol.gov/dol/esa/public/minwage/america.htm). Note that Virginia and Washington, D.C. have temporarily raised benefits; this change was set to expire on March 9, 2002.
Benefit amount for a median wage worker: EPI analysis of state unemployment insurance benefits. Median wages from Current Population Surveys MORG files 2001.
Maximum UI Benefits: Maximum UI benefit from U.S. Department of Labor (Significant Provisions of State Unemployment Law, January 2002). Note that Virginia and Washington D.C. have temporarily raised their maximum weekly benefits. These provisions were set to expire on March 9, 2002.
Tax rates above national average: Wage and Tax Rate Information by State for CYQ: 2001:3 (http://workforcesecurity.doleta.gov/unemploy/content/data_stats/datasum01/3rdqtr/sum.asp#wag).
Wage Base Above National Average: Wage and Tax Rate Information by State for CYQ: 2001.3 (http://workforcesecurity.doleta.gov/unemploy/content/data_stats/datasum01/3rdqtr/sum.asp#wag).
Wage base indexed to state wage: Comparison of State Unemployment Insurance Laws (2001, table 201).
Average High Cost Multiple (AHCM): Data through December 31, 2001 provided by U.S. Department of Labor.
UI Tax Cuts: Data from Marc Baldwin, Beyond Boom and Bust: Financing Unemployment Insurance in a Changing Economy, April 2001 (Appendix 4). Data are taxes on total wages 1994-2000.
In this report, we adopted the AHCM as our main solvency measure, but we employed an AHCM of 0.75 (or nine months) as our pass/fail cutoff. Since we are using only a pass/fail grading system, us
We captured the impact of UI tax reductions by identifying the 30 states that had larger-than-average falls in payroll tax rates on total wages between 1994 and 2000. UI payroll taxes fell because of outright cuts in taxes by state legislatures and the impact on experience rates of lower UI claims and moderately higher trust fund balances in the later years of the decade. In addition, UI taxes as a percent of total wages fell in many states because taxes are imposed only on a small percentage of wages (called the taxable wage base). States with fixed tax bases effectively reduce payroll taxes on most of their employers simply by not adjusting their taxable wage base
s. We captured all these reasons for falling UI taxes by looking at states that had reduced taxes as a percentage of total wages.
Optional extended benefits trigger: U.S. Department of Labor, Trigger Notice, State Extended Benefit (EB) Indicators (updated weekly). Note Hawaii and Wisconsin provide extended or additional weeks of benefits to those who exhaust their regular benefits.
Limited or no waiting week: National Employment Law Project analysis of state law. Note that some states have adopted temporary provisions to eliminate the waiting week. These states include California, New York, New Jersey, Virginia, and Washington, D.C.
State funded benefit supplement: Comparison of State Unemployment Insurance Laws 2001 (Table 309 fn. 3).
1. State unemployment rates are for December 2002. Data on national unemployment rates are for January 2002. State and regional data for February are currently unavailable.
2. About 43% of all unemployed workers received UI benefits in 2001. This measure, UI claims divided by total unemployment, is the standard measure of UI recipiency. However, total unemployment as measured by the Current Population Survey — the monthly survey that produces the unemployment rate — includes several groups of unemployed workers, some of whom UI was not designed to cover. In particular, UI was not designed for new entrants to the labor force, who accounted for about 6.8% of the total unemployed in the third quarter of 2001 (see Wandner and Stengle 1997 for a complete discussion of UI recipiency rates).
3. See the notes about exhaustions in Appendix A for a discussion of the methodology.
4. Currently, about 6% of exhaustees reside in the four states in which additional weeks of benefits have been or soon will be provided. The other 94% of the more than two million anticipated exhaustees are not receiving additional unemployment assistance. That is, of the approximately 80,000 workers currently estimated to be exhausting their regular UI benefits each week, more than 75,000 are not receiving additional weeks of assistance.
5. See Department of Labor’s Comparison of State UI Laws, January 2001, Table 301.
6. Washington has no earnings requirement but instead requires 680 hours in the base year.
7. Rather than measure benefits in dollar amounts, most analysts consider how much lost income is replaced by UI benefits. A worker who previously earned $500 per week and now receives $250 per week from UI would have a benefit replacement rate of 50%. Data are for 2000.
8. Employees make a very small UI contribution through payroll tax deductions in Alaska (0.5%) and New Jersey (0.2%). Other states, including Pennsylvania, use employee contributions only as a backstop when solvency is low.
9. Experience rating links an employer’s UI payroll tax rate to levels of UI claims by its laid-off employees. A general explanation of how experience rating of UI payroll taxes works is found in the Advisory Council on Unemployment Compensation (1995).
10. Statement of Maurice Emsellem, Subcommittee on Human Resources, Committee on Ways and Means, U.S. House of Representatives (March 9, 2000).
11. Under the more restrictive formula that applies to all the states, states are required to have an “insured unemployment rate” (IUR) of at least 5% combined with an increase in the rate of at least 20% for each of the past two calendar years. The IUR is, in essence, a measure of the output of the UI system itself—not the true labor market conditions in the state—because it is based on the number of regular claims in the state divided by the number of workers covered by UI in the state. As of the week of February 24, only four states were above the required 5% IUR.
12. At the time that Oregon and Washington triggered the EB program, their IUR’s were 4.04% and 3.69%, respectively.
13. A number of states have temporarily lifted the waiting week in response to the recent recession, including California, the District of Columbia, New York, New Jersey, and Virginia.
Advisory Council on Unemployment Compensation. 1994. Report and Recommendations. Washington, D.C.
Advisory Council on Unemployment Compensation. 1995. Unemployment Insurance in the United States: Benefits, Financing, Coverage. Washington, D.C.
Anderson, P. and B. Meyer. 1994. The Incidence of the Unemployment Payroll Tax. Advisory Council on Unemployment Compensation: Background Papers. Volume II. July 1995.
Baldwin, M. 2001. Beyond Boom and Bust: Financing Unemployment Insurance in a Changing Economy. National Employment Law Project. April 2001.
Boushey, H., and J. Wenger. 2001. Coming Up Short: Current Unemployment Insurance Benefits Fail to Meet Basic Family Budgets. Washington, D.C.: Economic Policy Institute.
Chimerine, L., T. S. Black, and L. Coffey. 1999. Unemployment Insurance as an Automatic Stabilizer: Evidence of Effectiveness Over Three Decades. Unemployment Insurance Occasional Paper. 99-8. U.S. Department of Labor, Employment and Training Administration.
Danziger, S. and P. Gottschalk. 1990. “Unemployment Insurance and the Safety Net for the Unemployed.” in W. Lee Hansen and James F. Byers, eds., Unemployment Insurance: The Second Half-Century. Madison, Wisc.: The University of Wisconsin Press.
Greenspan, Alan. 2002. Federal Reserve Board’s semiannual monetary policy report to the Congress befor Committee on Financial Services, U.S. House of Representatives, February 27.
Gruber, J. 2002. The Wealth of the Unemployed. Industrial and Labor Relations Review. Vol. 55, No. 1, pp. 79-94.
Haber, W., and M.G. Murray. 1966. Unemployment Insurance in the American Economy Homewood, Ill.: Richard D. Irwin, Inc.
McHugh, R., N. Stengle, and J. Wenger. 2002. Laid Off and Left Out. Working Paper. Washington, D.C.: American University.
Primus, W. & J. Goldberg. 2002. Two Million Workers Will Likely Exhaust Their Regular Unemployment Insurance Benefits in the First Half of 2002. Washington, D.C.: Center on Budget and Policy Priorities.
U.S. Department of Labor. 2002. UI Outlook. Washington, D.C.: DOL.
Vroman, W. 1998. Topics in Unemployment Insurance Financing. Kalamazoo, Mich.: Upjohn Institute.
Vroman, W. 2001. Low Benefit Recipiency in State Unemployment Insurance Programs. Washington, D.C.: The Urban Institute.
Wandner, S., and T. Stengle. 1997. Unemployment Insurance: Measuring Who Receives It. Monthly Labor Review, Vol. 120, No. 7, pp. 15-24.