Issue Brief on Unemployment Insurance. Wayne Vroman* January 2002

Transcription

1 Issue Brief on Unemployment Insurance Wayne Vroman* January 2002 The Congress is likely to enact an economic stimulus package in 2002 that includes significant increases in Unemployment Insurance (UI) benefits. Details of bills passed by the House and the Senate during 2001 differ widely. Great uncertainty surrounds the provisions of the bill that may be enacted sometime in the current year. This issue brief provides background information on the UI program, discusses key aspects of benefit recipiency, and describes the legislative developments to date. It provides a general discussion of UI before examining the federal legislation of The starting point is the current labor market and evidence of a recession. The Recession Throughout 2001 the U.S. economy demonstrated consistently weak performance. Real output (GDP) growth which fell below one percent in each of the first two quarters, was negative in the third quarter and probably was negative in the fourth quarter as well. In late November, the Business Cycle Dating Committee of National Bureau of Economic Research (NBER) formally designated that a recession started in March Their decision was made even before preliminary estimates of real GDP growth for the fourth quarter had been released by the Commerce Department. 1 During past recessions the economy has contracted rapidly in short time periods. Negative quarterly real GDP change of 6 percent or more occurred in one quarter of the years 1958, 1975 and In comparison, the changes during the first three quarters of 2001 were all smaller than one percent in absolute value. To this point in time the available evidence suggests the economy is experiencing a mild recession, perhaps akin to the pattern of modest decreases observed during the contraction of the early 1990s. In the labor market unemployment has increased, but the pattern in the first six months of 2001 was remarkable only in comparison to the strong labor demand observed 1 A discussion of the committee's decision is given in the Fall 2001 issue of the NBER Reporter, available at

2 in 2000 and earlier years. Dramatic increases in unemployment as measured in the household labor force survey (LFS) did take place in the second half of A series of increases moved the overall unemployment rate from about 4.5 percent in July to 5.8 percent in December. While the labor market report for December pointed to some positive developments such as increases in hourly earnings and employment gains in services, unemployment rate developments continue to receive primary attention. Claims for UI benefits provide a second clear indication of a deterioration in the labor market. In recent months UI claims have been more than 50 percent higher than one year previously, and the year-over-year changes have exceeded 0.5 million since April 2001 and 1.0 million since August Chart 1 summarizes developments with monthly claims data from 1999, 2000 and Every month in 2001 has had higher claims than in 2000 and during October-December the twelve month increase ranged from 1.3 to 1.5 million. The increase in UI claims dates from September Previously, monthly claims in 2000 had been lower than in 1999 but after September claims were higher than in The Chart 1 claims data suggest that the labor market deteriorated gradually during There was no dramatic movement in any single month of 2001, although the slope for September and October (indicating the monthly rate of increase) was higher than for earlier and later months. The decline in December 2001 is probably linked to an increased rate of benefit exhaustions since new claims continued to exceed new claims of the previous year. Thus the economy has entered a recession. But to this point in time (January 2002) it seems to be more akin to a growth recession rather than a sharp contraction as sometimes observed in the past with quarterly real GDP falling by 6 percent or more. Fourth quarter real GDP data will help in assessing the severity of the current downturn. Overview of Unemployment Insurance Unemployment Insurance (UI) pays cash benefits to eligible unemployed workers who have lost jobs through no fault of their own. Benefits provide partial replacement of lost wages for temporary periods. Weekly benefits are determined by formulas that typically replace 50 to 60 percent of previous wages subject to a maximum (typically 2

3 from 50 to 70 percent of statewide average wages). Maximum potential benefit duration is finite, most typically 26 weeks for workers in the so called "regular" UI program, and those who reach this limit are said to exhaust their benefits. Because benefits have a finite duration, the only way a worker can collect following benefit exhaustion is to reenter employment and earn sufficient amounts to establish a future entitlement. During past recessions, the maximum period of entitlement has been extended. The UI statutes that govern benefits in the regular UI program, e.g., entry eligibility requirements, weekly benefit levels and maximum potential benefit duration, are determined by the states. States undertake most UI administrative activities related to both paying benefits and collecting from employers the payroll taxes that support the program. The Federal government is also an important partner in UI program administration, requiring the states to make prompt and accurate administrative decisions over eligibility, to satisfy certain statutory tax provisions (a maximum tax rate of at least 5.4 percent and an annual tax base per worker of at least $7,000 2 ) and to practice efficient financial management of contributions and benefit payments. UI Benefit Generosity and Recipiency Because UI involves state statutes and administration, the programs in the individual states differ widely, especially in the share of the unemployed who receive UI payments. Summing across all state programs, it is also possible to characterize the program with national aggregates and averages. Chart 2 summarizes national data on generosity and recipiency for the years 1949 to 1999 using centered three year averages to smooth out noise in data from individual years. Note that the last data point is 1999 since that refers to averages for Benefit generosity is summarized with the ratio of weekly benefits to weekly wages. This measure, commonly termed the replacement rate, shows average payments relative to average weekly earnings. Over the years spanned by Chart 2, this measure of generosity ranged from a low of (1952) and a high of (1975 and 1976). Note the smooth evolution of benefit generosity with lowest ratios occurring in the 1950s. For 2 Strictly speaking, having a $7,000 tax base is an inducement rather than a requirement. A state can have a tax base below $7,000 per worker, but if it does all employers will be taxed at a uniform statutory rate of 5.4 percent rather than at current rates which have averaged close to 1.7 percent in recent years 3

4 most years between 1971 and 1983 the ratios exceeded and between 1984 and 1994 the ratios averaged almost Benefit generosity for those receiving payments has been quite stable for a long period. There was a modest decrease in this measure of generosity during the 1990s. By 1999 the three year average ratio stood at 0.320, 2.0 percentage points below the average for the full 51 years (0.340) and 2.9 percentage points below the three year ratio for 1992 (0.349), the highest ratio for the 1990s. The explanation for the decrease in the replacement rate, however, is more related to developments in average weekly wages than developments in UI benefits. The denominators of the Chart 2 replacement rates refer to the weekly wages of all covered workers. Overall average wages increased more rapidly during the 1990s than the average wages of those receiving UI benefits. 3 The different trends in wage growth probably reflect the faster rate of increase of wages for high-wage workers (with below-average unemployment) than for low-wage workers. Thus overall benefit generosity has been quite stable, even during the past decade. The evolution of benefit recipiency over these same years is also depicted in Chart 2. Two recipiency measures are shown. Recipiency 1 is the ratio of all beneficiaries to unemployment. This includes recipients under the regular UI program plus recipients of Federal-State Extended Benefits (EB) and recipients of various temporary federal programs operative just during recessions. Recipiency 2 is the ratio of just regular UI program beneficiaries to unemployment. The vertical gap between the two recipiency measures shows that EB and the temporary programs were important in past recessions. Three aspects of these measures are noteworthy. 1) There has been a strong longrun decline in recipiency under the regular UI program. Recipiency 1 averaged 0.40 and above in most years up to 1961 while it averaged below 0.30 in most years after Since 1982 the Recipiency 1 has averaged ) While Recipiency 1 increased after the recession of the early 1990s, the highest ratio observed in the 1990s was only (1999). The ratio in 1999 was only 1.6 percentage points higher than in 1995 despite the strong economic growth of the period. 3) The role of the federal government in providing 3 Data from the Benefit Accuracy Measurement (BAM) system show weekly benefits and weekly wages of UI beneficiaries. These data, available only since 1988, show higher replacement rates than in the usual series reflected in Chart 2 (so called Handbook data). The ten year averages of replacement rates from 4

5 benefits to the long term unemployed during recessions has grown. There were no federal (or EB) benefits during the recessions of 1949 and 1954 and only a small federal caseload during the recessions of 1958 and The importance of long-term UI benefits was greatest during the period, but there were also relatively large payments during , and This importance is indicated by the vertical distances between the squares and diamonds in Chart 2 for a given year. During the downturn of nearly all UI payments for long-term unemployment were from the federal program (Emergency Unemployment Compensation or EUC) as few states paid any EB benefits. Over the period covered by Chart 2, the federal role in UI benefit provision during recessions has become firmly established. Further, although not obvious in the chart, the relative mix of financing longterm benefits has shifted sharply towards the federal partner, partly due to changes in the early 1980s in EB program triggers (the unemployment rates that activate the payment of EB). The EB program accounted for roughly half of long-term benefit payments during but less than two percent during Unlike initial claims for unemployment benefits which tend to lead the economy during downturns, long-term unemployment lags behind economic performance. As a consequence, payments for long-term unemployment also tend to lag during economic recoveries. If a federal long-term benefits program is enacted, this pattern can be anticipated in the current recession. Over the past five decades, the average duration of unemployment has been gradually increasing. The increase is apparent in data from the monthly household labor force survey (LFS) and in data from the regular UI program. Mean duration in LFS data averaged 11.3 weeks during the 1950s but 15.7 weeks during the 1990s. Potential benefit duration in the regular UI program has not changed much in more than three decades. Most who become unemployed and file initial claims are entitled to a maximum of 26 weeks during the next 52 weeks. The result of increasing average duration coupled with stable potential duration has been an increased rate of benefit exhaustions. In the 1950s the average exhaustion rate was 25 percent, but it was 36 percent in the 1990s. Even in 1991 to 2000 were in Handbook data and in BAM data. The three-year averages for 1992 and 1999 in BAM data were and respectively. 5

6 2000, a year of very low unemployment, 30 percent of UI recipients exhausted their benefits. The long-term unemployed are known to experience high poverty rates. Given the high prevalence of UI exhaustions and predictable increases as the downturn unfolds, provision of long-term benefits will be especially important in preventing hardship among families and individuals with long-term unemployment. Substantial hardship can be anticipated even in the early stages of the ensuing recovery. Workers with low wages, part-time workers and those with irregular patterns of earnings typically do not collect UI benefits despite having above-average unemployment rates. A recent study by the General Accounting Office documented both their high unemployment and low receipt of benefits. 4 While many factors contribute to this situation, UI eligibility rules are important in causing low recipiency. Three aspects of UI statutes and administrative procedures that affect eligibility deserve particular attention: the designation of the base period, the treatment of those who quit for personal reasons and the issue of job availability for part-time workers. The combined effects of the three frequently cause ineligibility even when claimants' past earnings exceed UI monetary eligibility requirements. Brief discussions of each situation follow. State UI programs make decisions about eligibility after considering both monetary and nonmonetary factors. In most states, monetary eligibility is decided by comparing the worker's annual earnings and earnings in the high quarter of the base period with statutory requirements. The base period is usually the earliest four of the past five fully completed calendar quarters ending prior to filing a claim, e.g., July 2000 to June 2001 for a claim filed in December Depending on when the claim is filed, three to six months will elapse between the end of the base period and the filing date. In the UI monetary eligibility determination, the claimant's earnings from these recent months will go unrecognized unless the state also has an "alternative base period" (ABP). Several states have ABP provisions that are used for those initially found ineligible under the regular base period. The most common ABP uses the last four 4 U.S. General Accounting Office, "Unemployment Insurance, Role as Safety Net for Low-Wage Workers is Limited," Report GAO , (Washington, D.C.: U.S. General Accounting Office, December 2000). 6

7 completed quarters (October 2000 to September 2001 in the preceding example) in making a second monetary determination. Statutory provisions vary across the 12 states with an ABP. A summary of the added costs from the ABP would be roughly as follows, a 5 to 8 percent increase in beneficiaries and a 4 to 6 percent increase in benefit payouts compared to a situation with no ABP. Because the ABP disproportionately helps lowwage workers, it increases UI recipiency more than it increases costs. Eight of twelve states with an ABP have had it for at least five years. Hence there is good information on both the added benefit payouts and the added administrative costs due to the ABP. People who quit jobs typically are not eligible for UI benefits. In determining eligibility, some states draw a distinction between job-related and personal reasons for quitting. Someone who quits due to uncontrollable personal reasons but leaves in such a way as to retain the job connection, e.g., requesting and being granted a leave of absence, may later collect UI when they subsequently try to return to work but the job is no longer available. For a quit to be compensable in most states, however, the cause must be related to the job, e.g., a transfer to another shift for someone unable to arrange child care. Good personal reasons for quitting are simply not recognized in the majority of UI programs. As a condition for UI eligibility most states require that the claimant be available for full-time work regardless of the past pattern of work. Those who lose jobs but previously worked part-time (less than 35 hours per week) could only collect UI benefits if they would accept full-time work. Only a minority of states routinely permit those seeking part-time work to collect UI benefits. The circumstances described in the preceding paragraphs undoubtedly have become increasingly prevalent as the composition of the labor force and the nature of work have evolved during the past five decades. Ineligibility among unemployed persons in these situations has contributed to the long-term decline in recipiency depicted in Chart 2. Note that most of the affected workers have demonstrated a substantial commitment to the world of work and will continue to seek paid employment in the future. During their unemployment spells, however, they do not collect UI benefits. Eligibility rules vary across states, but eligibility is enhanced in states where there is an ABP, good personal reasons for quitting are recognized and availability for part-time employment is allowed. 7

8 Considerations of access to UI benefits for low-wage and part-time workers are of special concern in the reformed welfare system. Many former TANF recipients have jobs with these characteristics and many would experience unemployment spells, especially during recessions. Thus improving general access to UI will help these particular workers avoid a return to welfare dependency. Because states play a central role in UI, it is important to note that recipiency varies widely from one state to the next. During the three years workers in the ten states with highest recipiency (beneficiaries as a proportion of unemployment) were nearly three times as likely to collect UI benefits compared to workers in the ten states with the lowest recipiency, averages of 0.50 and 0.18 respectively. Map 1 summarizes recipiency rates during with the states (plus the District of Columbia) arranged into 3 groups of 17. All states in the High group had recipiency rates that averaged at least 0.35 while all in the Low group had recipiency rates below There is an obvious geographic pattern to high and low recipiency. States with high recipiency are located in the Northeast, the upper Midwest and along the West Coast. Low recipiency is concentrated in the deep South, the Southwest and along the Rocky Mountains. The patterns depicted in Map 1 have been persistent since the late 1960s, i.e., for as long as state-level measurements have been available. Low recipiency is not fully understood but some aspects are known. 5 In states with low recipiency, application rates for benefits are low. Nationwide, about half of new onsets of unemployment are followed by an application for UI benefits. In some states with low recipiency, the application rate is half the national average, i.e., one fourth of new onsets. In most of these states the rate of disputes over eligibility is above-average. Disputes over misconduct seem to be especially prevalent and high rates of misconduct proceedings are associated not only with low initial application rates but also low rates of awards to claimants. In a few states, the UI agency rules on a misconduct issue in one fifth of all applications. These so-called separation issues arise much less often in states with high benefit recipiency. It is much easier for claimants to collect UI in some states 5 This discussion draws from a recent project report to the U.S. Department of Labor: Wayne Vroman, "Low Benefit Recipiency in State Unemployment Insurance Programs," The Urban Institute (June 2001). This report is available at online at 8

9 than in others. 6 If recipiency were increased in low-recipiency states, it would also increase the response of total UI benefit payments to economic downturns and improve the performance of UI as an automatic stabilizer of the macro economy. The combined effects of the preceding factors yields low overall UI recipiency. Unless changes are made, we can anticipate continued low receipt of benefits among the unemployed with low wages, part-time jobs, workers with irregular work patterns and especially among workers in the deep South, the Southwest and Rocky Mountain states. Those who exhaust regular UI benefits will also need extended benefits, and the need for extended benefits will be present throughout The deliberations over an economic stimulus package may yield a federal vehicle to address these needs. UI Financing Payroll taxes paid by employers support UI. There are both federal and state UI payroll taxes. Contributions are deposited into trust fund accounts maintained at the U.S. Treasury. Fifty three state accounts 7 are the source of monies for paying regular UI benefits and the state share of EB. The state accounts can be used only to pay UI benefits. A federal account finances the administrative costs of UI and the Employment Service in the states. The federal share of EB is paid from a second federal account and a third federal account makes loans to states with insolvent trust funds. At the end of September 2001, the combined balances in the state and federal trust fund accounts exceeded $90 billion, $51.6 billion in the state accounts and $38.6 billion in the federal accounts. The federal UI accounts are overflowing. All three are currently at required statutory maximums. Under current law, excess monies in the federal trust funds are slated for return to the states (so called Reed Act distributions). Excess federal trust fund balances are to be distributed according to each state's share of federal taxable UI wages. Most state accounts are also large relative to potential recession-related needs. While the method for assessing needs has uncertainties (mainly related to questions of the depth and duration of the current economic downturn), the standard measures (termed 6 Short duration of UI benefits could also contribute to low recipiency, but while there is wide variation in benefit duration, it is less consistently associated with low recipiency than low initial application rates and low ratios of first payments to new claims. 7 There are programs in fifty states plus the District of Columbia, Puerto Rico and the Virgin Islands. 9

10 reserve ratio multiples or high cost multiples) 8 indicate that just three states have very low balances. If payouts were to occur at a rate equal to the average of the three highest payout years (of the past twenty), the September balances for New York, North Dakota and Texas would represent less than four months of reserves for benefit payments. While several other states have balances that represent from six to eight months of benefits, these states all have time to consider solvency legislation during their 2002 legislative sessions before becoming insolvent. Technical insolvency does not pose an immediate threat to the continuation of benefit payments since there is a well established procedure for borrowing from the federal trust fund intended for this express purpose. During the recession of the early 1990s seven states needed such loans, but only two (Connecticut and Massachusetts) needed large loans. 9 Thus the states, with a few notable exceptions, have trust fund reserves that are fully or nearly adequate by the common actuarial standard used to assess solvency. Most states are very concerned about the sufficiency of funds allocated from the Employment Security Administration Account (ESAA), the federal account that finances both UI program administration and Employment Service (ES) administration in the states. Supported by a federal payroll tax, monies returned to the states for UI and ES administration have systematically fallen short of amounts paid by employers to support these administrative activities. In fiscal year 1999, for example, ESAA payroll taxes totaled $5.4 billion while administrative allocations totaled $3.2 billion. Low funding levels have caused states to make several adjustments in UI and ES administration. Payroll taxes and general revenues are being used in several states to supplement monies from the ESAA allocations. In fiscal years 1999 and 2000 these state supplements totaled roughly $250 million and $300 respectively. 8 The reserve ratio multiple or high cost multiple is the ratio of two ratios. The denominator ratio is the three year average high cost rate (payouts expressed as a percentage of covered payrolls). The numerator is the ratio of the fund balance to last year's covered payroll (also expressed as a percentage). It is recommended that states have multiples of 1.0, i.e., balances that represent twelve months of payouts at the three year high payout rate. At the end of 2000 the national multiple was 0.91 and 30 states had multiples of 1.0 or larger. New York, North Dakota and Texas had ratio between 0.26 and Four states had ratios between 0.48 and 0.59 (Illinois, Minnesota, Missouri and West Virginia), four had ratios between 0.60 and 0.69 (Alabama, Arkansas, Ohio, Pennsylvania) and twelve had ratios between 0.75 and Total loans to these two states in the early 1990s exceeded 1 percent of (1990) their covered payrolls. 10

11 Federal UI taxes have systematically exceeded monies returned to the states throughout the past decade. As a consequence, the federal trust funds have grown and now exceed $38.0 billion. After reaching statutory thresholds, excess monies in the federal trust funds are to be returned to the states as Reed Act distributions. The current fiscal year is the first when substantial sums are due to be returned to the states. Proposed Federal Legislation in 2001 Several legislative proposals intended to stimulate the economy were offered in Two bills received majority votes but they differ widely, and the Congress recessed in late December without finding an acceptable stimulus package. This section discusses the parts of the proposals that deal with UI benefits, monies for UI-ES administration and health insurance benefits for the unemployed. The President proposed a stimulus package in September that included extensions of UI benefits and improved health insurance coverage for the unemployed. In the House this proposal was modified substantially. The House bill that narrowly passed in October had two main provisions to help the unemployed and bolster UI-ES administration. 1) There would be an accelerated distribution of Reed Act monies to the states. States could use these monies for an unspecified combination of three activities: i) UI benefit increases, ii) increased resources for UI-ES administration and support of other worker adjustment programs, and iii) reduced employer payroll taxes that support the regular UI program in the states. The states could enact temporary benefit provisions, to be operative through March 2003, with costs financed through the Reed Act distribution. 2) Assistance with health insurance costs would be provided to the unemployed through larger block grants to the states. The Congressional Budget Office has estimated the costs of these provisions at $9.3 and $3.0 billion respectively. Most of the budgetary impact of the House bill would come from other provisions, mainly reduced taxes for corporations and individuals. The total first year impact of all provisions was about $101 billion See Congressional Budget Office (CBO), "Congressional Budget Office Cost Estimate, H.R Economic Security and Recovery Act of 2001," (Washington, D.C.: CBO, October 16, 2001). 11

12 A Senate bill originating in the Finance Committee with substantially larger help to unemployed workers was approved by a vote on November 9. Because this bill did not receive the supermajority of 60 votes required for formal passage under its budget rules, it cannot be described as enacted by the Senate. However, since the bill's UI provisions (to be effective through December 2002) were the basis for the bargaining position maintained by the Senate Democrats in discussions with Senate Republicans, the House and the Bush Administration in December 2001, its provisions will be discussed. The Senate bill has five main provisions affecting the unemployed and UI benefits. These are the following (with CBO cost estimates in parentheses): 11 1) expanded UI eligibility through the ABP and part-time worker provisions ($1.01 billion), 2) 13 weeks of extended UI benefits for those who exhaust regular UI after September 11 ($11.02 billion), 3) increases in weekly benefits by (the larger of) 15 percent or $25 ($6.35 billion), 4) partial funding of health insurance costs of the unemployed ($15.3 billion) and 5) reimbursement to the states for added administrative expenses ($0.50 billion). Health insurance cost reimbursement would be for three-fourths of the costs of maintaining policies for those previously insured and full reimbursement for the uninsured with low income through Medicaid. The total cost of all five provisions (including interactions and effects of increased entitlements on actual benefit duration) was estimated by CBO to total about $35 billion in fiscal years 2002 and 2003 combined. Health insurance and Medicaid constituted $15.3 billion while the UI provisions (including administration) totaled $19.6 billion. The total Senate package (with tax reductions and other new spending) increases the deficit by $79 billion over these two years with the bulk of the effects concentrated in Under what CBO would probably consider more realistic economic assumptions the increase in UI benefits would total $26.5 billion while health insurance increases would be $19.7 billion. Thus the Senate measure would have about three times the amount of spending related to UI and health insurance compared to the House bill. More generally, the House bill relies mostly on tax reductions to provide economic stimulus 11 See Congressional Budget Office (CBO), "Congressional Budget Office Cost Estimate, H.R Economic Security and Recovery Act of 2001," (Washington, D.C.: CBO, November 15, 2001). 12

13 while the Senate bill places greater weight on spending increases, mainly cash benefits and health benefits for the unemployed. As this issue brief is being completed, the House and Senate are slated to return and take up negotiations on a compromise package to be enacted early in If a package can be crafted, it is likely to have smaller benefit increases than the Senate measure but larger than the House measure. Considering both measures, it is interesting that several important provisions can be described as first time initiatives. In the past, extensions of benefits for the long-term unemployed have constituted the centerpiece of recession-related federal UI legislation. The proposed bills of 2001 had three new elements: 1) help with costs of health insurance for the unemployed, 2) increases in weekly benefits and 3) expanded eligibility for those who do not presently qualify for regular UI. 12 A fourth new element is reliance on a onetime Reed Act distribution to state trust fund accounts, giving the states broad latitude in decisions regarding the economic stimulus monies to be made available. As noted, the biggest one-time grant under consideration is the Reed Act disbursement of $9.3 billion contemplated in the House bill. To help understand the consequences of this potential disbursement, the National Association of State Workforce Agencies (NASWA) circulated a questionnaire with ten questions to all state employment security agencies. Responses were obtained from 38 of the 53 programs. 13 While these responses were qualitative rather than quantitative, they provide valuable information. NASWA was careful to note that the responses summarize the advice the agency directors would provide to their Governors and legislators. Ultimately, the actual use of the monies would be governed by decisions made in the political processes of each state. At issue here is the weight of the advice given by UI administrators to political actors regarding the uses to be made of a distribution of monies from the U.S. Treasury. Key responses to the questions were the following. 1) States would do more than simply add the monies to the state UI trust funds. 2) Many states viewed the monies as 12 The one exception was the Supplemental Unemployment Assistance program enacted in the 1975 recession. This temporary program made payments to claimants who did not qualify for regular UI, typically because of inadequate earnings or earnings in sectors not (then) covered by UI, e.g., many employees of state and local governments. 13 See National Association of State Workforce Agencies, "How Would Your State Use An Accelerated Reed Act Distribution" A Summary of A State Survey," November

14 providing added insurance against having to borrow from the Treasury if the recession proved to be serious. 3) About one-third indicated they would avoid UI tax increases or benefit reductions because of these added monies. 4) Uses besides benefit enhancements would figure predominantly in state actions. Only about one third indicated they would consider state-level benefit extensions targeted on exhaustees. Only five would consider enacting an ABP and just three were interested in expanding access to part-time workers. Finally, using monies to increase weekly benefits was not a common response. 5) Most states indicated some of the monies would be used to enhance UI administration and provision of adjustment services to the unemployed. Those who indicated to the contrary typically said it was because monies were needed to ensure trust fund adequacy. The states have ended a ten year period where economic performance should have automatically raised trust funds by substantial amounts. However, tax cuts during these prosperous years offset much of the natural increase in state accounts that would otherwise have occurred. To use the Reed Act monies to increase solvency during a recession seems contrary to the intent of the UI program. Considering likely political developments in the states, it seems that little of these added monies would go to enhance eligibility or to provide benefits to exhaustees of regular UI benefits. While adding monies to program administration is probably necessary for the effectiveness of UI administration in the long run, it would seem that adding monies to the income of the unemployed is a more pressing short-run concern, especially during a recession. Two final observations about a potential Reed Act distribution are important. 1) The Congress could specify how some or all of the Reed Act monies are to be used by the states. 2) The provisions of a stimulus package (should one be enacted) could affect the size of the Reed Act distribution. If, for example, there were to be a temporary extension of UI benefits to the long term unemployed, the added payouts could offset Reed Act monies. At present, roughly $4 billion is slated for return to the states. This could be totally eliminated under an extension of long-term benefits. While the health insurance provisions and increases in benefit levels also represent new developments, note that all these monies will go directly to families and individuals with unemployment. This should provide stimulus through the same reasoning that led to the original enactment of unemployment insurance in the 1930s. 14

15 These monies would provide enhanced replacement of lost wages caused by the recession and help to maintain spending during spells of unemployment. There may be a question of how to "turn-off" the higher benefit levels once the temporary (or emergency) period has ended. But the targeting of the increases is appropriate, i.e., the unemployed. Two other observations about the legislative proposals can also be offered. First, expanding recipiency is clearly a long-run need of the state UI programs. The ABP and part-time provisions of the Senate measure are important benefit enhancements, but singly and even combined they have modest effects in terms of both numbers of recipients and added payouts. The recipients of these enhancements are generally lowwage persons. Their hardships would be helped but the societal gains are mainly in the areas of improved equity and reduced hardship than in enhanced macro stabilization. Second, none of the measures address questions of interstate variation in recipiency. As noted, recipiency is very low in several states. This situation contributes at least as much to low overall recipiency in national data as the specific situations addressed by ABP and part-time worker enhancements. Further, if recipiency were increased in low-recipiency states, there would be a larger effect on spending per newly entitled person since their wages (hence benefits) would be higher (compared to enhancements through ABP and part-time provisions). This type of eligibility enhancement would be more effective in restoring the performance of UI as an automatic stabilizer of the macro economy. * The author is an economist at the Urban Institute. Helpful comments on a earlier draft were provided by Richard Hobbie of the National Association of State Workforce Agencies and Ralph Smith of the Congressional Budget Office. Any errors of fact or interpretation are the sole responsibility of the author. This document does not represent the official views of the Urban Institute or its sponsors. 15

16 Chart 1. Continued Claims for UI Benefits, 1999, 2000, and 2001 Thousands Jan. April July Oct

17 Chart 2.Unemployment Insurance, Long Run Trends in Recipiency and Generosity, 1949 to Proportions Recipiency 1 Recipiency 2 Generosity Recipiency 1 = (all beneficiaries(regular UI, emergency and EB)/unemployment), Recipiency 2 = (regular UI beneficiaries/unemployment) Generosity = (average weekly benefits/average weekly wages). All data centered three year averages.

18 Recipiency Rates by State, Below to 0.35 Above 0.35