Commentary | Economic Growth

Addressing the unemployment problem | EPI Viewpoints

Opinion pieces and speeches by EPI staff and associates.



ddressing the unemployment problem

by Larry Mishel

Mr. Chairman and Members of the Committee:

Thank you for the opportunity to discuss our nation’s current employment and unemployment. As you know, unemployment has been rising fairly steadily since October 2000, rising from 3.9 percent to its current level of 5.7 percent, with 8.1 million workers unemployed. Although the economy has been growing in 2002, the growth has not been strong enough to create jobs and keep unemployment from rising. We now have 1.7 percent fewer private sector jobs than in October 2000. Moreover, current slow growth will generate few jobs and will lead unemployment to grow further – perhaps to 6.5 percent or more – and to remain near or above 6 percent for at least another year.

The nation, thus, has a serious, persistent jobs and unemployment problem that will not go away on its own and should be addressed in two ways. First, we need to cushion the impact of this recession on those directly affected. Second, we need to generate growth by stimulating demand – that is, by creating more customers for business. The fastest, most effective way to do this is by boosting government spending quickly and for the next year of so. Extending unemployment insurance and improving the unemployment insurance system will both assist those in need and stimulate demand quickly.

Today’s job situation
The Bureau of Labor Statistics just released its survey on employment and unemployment for August 2002. Unemployment stands at 8.1 million, 5.7 percent of the labor force. Although 39,000 jobs were created last month, the number of private sector jobs did not expand and is now 1.8 million jobs, or 1.7 percent, lower than in October 2000. I believe, with many other economists, that last month’s 0.2 percent drop in unemployment is a temporary dip that will be followed by a ratcheting up of unemployment over the near term.

It is true that today’s unemployment rate is relatively low compared to the unemployment reached in the recessions over the last thirty years (although it is sizeable relative to those of the 1950-73 period). There are four reasons, however, not to be complacent: unemployment will rise further; our target for unemployment should be 4 percent; the current level of umemployment has significant adverse consequences; and, when average unemployment is at its current levels some segments of the workforce experience very high unemployment.

The important point about today’s unemployment is that it is 1.8 percent higher than the 3.9 percent rate we had just two years ago, an increase equivalent to 2.6 million more unemployed workers. Although an unemployment rate of 5.5-6 percent used to be considered acceptable, and any lower rate considered potentially inflationary, we now know that this conventional wisdom was and is wrong. The late 1990s have reestablished that 4 percent is the proper benchmark target for unemployment, one that facilitates broad-based noninflationary growth. The costs to families of having unemployment higher than 4 percent are substantial, as discussed below. Unfortunately, despite last month’s dip in unemployment, I expect unemployment to rise over the next six months and to stay near or above 6.0 percent, throughout 2003.

The explanation is simple. In order for the unemployment rate to drop, the economy must grow faster than the sum of productivity plus labor force growth. This is because productivity growth allows us to generate more goods and services without expanding employment and the economy needs to continuously absorb new workers (recent high school or college graduates and other “new entrants” or “reentrants”). Since the long-term productivity trend is from 2.0 to 2.5 percent and the labor force trend is 1.0 percent, the economy must grow at least 3.0 to 3.5 percent just to keep unemployment from rising. It takes a growth rate of 4.0 to 4.5 percent for unemployment to drop 0.5 percent.

Even though the economy began growing over the last three quarters, growth has been so slow (0.1 percent, 1.4 percent, 2.1 percent) that very few jobs have been created and unemployment has risen giving us the ingredients of a “jobless recovery.” The unemployment rate would have grown further if the labor force had expanded at its recent historical rate. We estimate that the current recession in the labor market has kept roughly 2.2 million people from seeking employment – a group not captured in the unemployment rate. Had just half of the workers in this “hidden unemployment” been in the labor force, the current unemployment rate would be close to 6.5 percent.

There are several reasons why I expect unemployment to grow. First, the impact of the recent slow growth has not yet been fully felt, since unemployment “lags” behind growth. Second, growth is forecast at the 2.5 percent to 3.0 percent rate through the end of this year and at about 3.5 percent for next year, a growth rate not sufficient to achieve a 4 percent unemployment rate anytime in the next few years. This is the conclusion of the both the CBO and the Blue Chip consensus. In fact, The Blue Chip survey finds that two-thirds of the forecasters do not believe that unemployment has peaked. Moreover, forecasters expect unemployment to peak at 6.2 percent, 0.5 percent more than today’s rate. Third, unemployment will be higher than is typically forecast. That’s because if and when growth accelerates, we can expect the labor force will resume its more rapid expansion, making up for the very slow labor force growth (especially in the share of the adult population in the labor force) over the last two years. For instance, from July 2001 to July 2002 the labor force grew by just 0.5 percent, in contrast to the expected growth of 1%. My conclusion, I am sorry to say, is that the unemployment situation will grow worse before it gets better and unemployment will stay high through 2003.

It is also important to note that the national unemployment rate is an average that conceals low unemployment among some groups but very high unemployment among others. Therefore, complacency in the face of roughly 6.0 percent unemployment is inappropriate. For instance, the unemployment rate has averaged 5.8 percent over the last three months. This includes groups with unemployment rates that are low: professionals and managers, 3.1 percent; college graduates, 2.8 percent; and government workers, 2.4 percent. Other groups, however, have very high unemployment rates such as construction workers (9.6 percent), those with less than a high school degree (8.4 percent); and blue-collar operatives and laborers (8.7 percent). Especially, notable is the higher unemployment rates of minorities, with blacks at 10.1 percent and Hispanics at 7.5 percent. Most worrisome is that some disadvantaged groups, such as young (ages 16-25) black men and women with a high school degree have unemployment rates of roughly 24 percent. This is especially relevant to concerns about the success of welfare policy in the current environment, since high unemployment and underemployment in the low wage labor market obviously impedes former welfare recipients from getting and maintaining employment.

Higher unemployment has had, and will continue to have an adverse impact on working families. The impact of the recession obviously falls hardest on those who become unemployed. Less well appreciated is that the unemployment rate does not adequately reflect the number of workers who become unemployed at some ti
me during the year, a number two to three times as large as the number who are unemployed at any one point in time. As unemployment rises so does underemployment, including, for example, workers in part-time jobs who want full-time jobs, workers in temporary jobs who want permanent, regular jobs and those working at jobs for which they are overqualified.

Higher unemployment also affects those who remain employed by leading to slower wage growth. Wages are still growing faster than inflation but at only half the pace of a few years ago (at roughly 1 percent rather than 2 percent real growth). Wage growth has decelerated the most for low-wage workers whose wages are now barely keeping up with inflation. The wage growth among middle-wage men has also been substantially reduced. Consequently, we are now seeing a renewal of the trend of a widening of the wage gap between low and middle-wage workers, as well as the continued growth in the wage gap between high-wage and middle-wage workers. In other words, the recession is reestablishing the trend toward an across-the-board growth in wage inequalities, something we haven’t seen since the mid-1990s. The bottom line is that higher unemployment is lowering family incomes — by $1,800 this year for a middle-income family — and leading to wider inequalities.

Another way to understand the costs of high unemployment is by examining the benefits of the period of low unemployment that prevailed from 1995 to 2000. We saw broad-based real wage growth for the first time in two decades. Job quality improved as health insurance and pension coverage improved, part-time work diminished, self-employment shrank, and the need to work two or more jobs declined. Family incomes rose across the board and the income and wage gaps by race shrank for the first time in years. All of this is to say, we certainly miss those days of low unemployment and we need to get back to them as quickly as possible.

Characteristics of the current recession
Several characteristics of the current recession are noteworthy. In particular, compared to earlier recessions: the private sector job loss has been more severe; the rise in unemployment has been broadly shared by gender and education; and, college graduate unemployment has risen more. Plus, there has been an usually large increase in the duration of unemployment – how long people stay unemployed – as witnessed by the historically large share of the unemployed who are officially designated as “long-term” unemployed – meaning they have been out of work for at least 26 weeks.

Some analysts have labeled this recession as shallow because of the relatively small decline in economic output and the relatively low increase in unemployment. What has not been noticed, however, is that there has been an even more severe loss of private sector jobs in this downturn than in previous ones. The decline in private employment — 1.7 percent, or 1,833,000 jobs — is larger at this point in the cycle than in the recessions of the early 1980s and 1990s. This reflects the severe loss of manufacturing jobs and the stagnant growth of private sector service jobs. Increased employment by government is a singular bright spot. The private sector job loss is the downside of the relatively fast productivity growth we have enjoyed.

The increase in unemployment has been more uniform in this recession, affecting a broader array of the workforce. The unemployment rate has grown comparably among those with “some college'” a high school degree and those without a high school degree. In contrast, unemployment in prior recessions rose much more among those with less education. Unemployment among women has risen almost as much as among men, a stark contrast with earlier recessions, which more strongly affected men. This gender pattern reflects the heavy job losses in the female-intensive service industries. It is interesting to note, in this regard, that the unemployment rate among college graduates has risen more than in the early 1990s and is on par with the much steeper increase during the early 1980s recession.

There is one way that this recession’s impact is more narrowly focused – those who became unemployed tend to stay unemployed longer. In August there were 1,474,000 unemployed who had been out of work for more than twenty-six weeks, comprising 18.1 percent of the unemployed. This is a larger share of the unemployed than at this point in the business cycle in the 1970s and 1990s but less so than in the early 1980s recession (Figure A). This longer-term unemployment suggests a greater mismatch between the types of jobs available and the types of workers that are unemployed. Obviously, an increase in long-term unemployment suggests the continued need for extended benefits in the unemployment insurance system.

Policy targets: stocks and employment
It is understandable that there has been much media discussion of the fall of stock prices over the last few months, with the Dow falling 10.4 percent, the NASDAQ falling 14.5 percent and the S&P 500 falling 10.4 percent since June. Yet, economic policy should not target getting stock prices back up as a goal, even though the losses are painful to some. Besides the fact that stock prices were probably inflated and unsustainable, the government should not be in the business of providing a safety net to stockholders.

There are many reasons that economic and budget policy should not focus on the stock market. First, we do not have a basis for establishing what the correct level of stock prices should be – the target. Second, we do not understand well what moves the stock market. Third, the government does not have effective tools to affect the stock market. Fourth, there is no clear connection between a rising stock market and the economic well-being of the vast majority, as a rising stock market can be consistent with improving or declining family incomes and real wages. Last, contrary to popular belief, the stock market has not been a source of investment funds – companies have used more cash to buy stock than they have received from selling stock. Consequently, the government should structure and regulate the stock market so as to generate accurate financial information and transparency but leave the value of the stock market to supply and demand forces.

On the other hand, at a time of high and rising unemployment it is not only appropriate, it is also necessary for government economic policy to focus on job generation and putting the economy on a track to achieve 4 percent unemployment. It is easy to establish the size of the existing “job gap” – the difference between the number of job openings and the number of unemployed – using the new JOLTS data provided by the Bureau of Labor Statistics, as shown in Figure B. In December 2000, the earliest data, there were roughly 5.2 million unemployed and 4.0 million job openings – a gap of 1.2 million jobs. By August 2001 – before the terrorist attack – the job gap grew to 3.1 million as job openings slipped by 0.2 million and the unemployed grew by 2.0 million. The latest data, for June 2002, show a further 0.6 million decline in job openings and a further 1.2 million growth of unemployed workers, leaving a total job gap of 5.4 million jobs. As of June 2002, there were 2.67 unemployed workers for every job opening.

What’s needed-more customers
The Administration is correct in one important sense in saying that the economy’s fundamentals are good. That is, we are enjoying a relatively fast growth in productivity even during this recession, a continuation of a trend that began in 1996. The long-term productivity growth rate is now estimated to be 2.0 to 2.5 percent, which is 0.5 to 1.0 percent greater than the roughly 1.5 percent trend over the 1973-95. Thus, one problem we do not seem to be having is improving our efficiency.

The problem we face is that businesses do not have enough customers – they need more sales. As sales rise, so will profits and so will the number of workers hired. Investment in
plants and equipment will follow. In economic terms, we have excess supply – plenty of available workers and productive capacity (industry is now working at 76.1 percent capacity, down from 81.8 percent in 2000) – but insufficient demand. Therefore, to achieve a stronger cyclical recovery we need mechanisms to quickly increase expenditures by government consumers (both foreign and domestic). That is, we need a short-term (over the next two years) boost to demand.

From 1996 to mid-2000, demand was growing at a 3.5 to 4.5 percent pace but fell to a 1.5 percent pace over the last five quarters because consumption growth slowed, investment fell and net exports declined. In order to lower unemployment by 1.5 percent over the next two years – a modest goal that gets us close to the 4 percent unemployment rate target – we will need to increase overall growth 1.5 percent each year for the next two years – 3 percent more growth overall.

There are two means to do this. One would be to gradually lower the value of the dollar through a policy coordinated with other advanced countries and using lower interest rates. This would lead to greater exports and fewer imports, thereby boosting the manufacturing sector. This is an important policy but not one that directly involves congressional legislation.

The other mechanism to boost demand is to increase government expenditures without raising taxes or having offsetting cuts in other programs. This will require extra spending of $100 billion in each of the next two years. To be clear, this extra spending must be associated with an increase in the fiscal deficit over the next two years in order for these new expenditures to add to demand. However, the increase in expenditures and the accompanying deficits need not, and should not, create long-term fiscal imbalances. Thus, the characteristic of a short-term spending stimulus is that the expenditures decline and then disappear as unemployment falls. It is possible to implement an ongoing spending plan if it is paid for by bringing in more revenues after the first two years. Preventing the future execution of the planned tax cuts and retaining the estate tax can help offset the costs of higher spending implemented now, allowing policymakers to achieve long-term fiscal solvency.

It is also possible to increase short-term demand through consumer-oriented tax cuts. The tax cuts being discussed by the Administration, however, are permanent and therefore raise deficits in both the short- and the long-term. They are also not consumer-oriented. In fact, many of the tax cuts are oriented toward increasing saving, which would decrease rather than increase demand. The proposed tax cuts are thus wrongly timed (permanent rather than short term) and incorrectly targeted (at increasing saving rather than consumer spending).

Another advantage of spending over tax cuts is that with tax cuts there is leakage to savings and imports, which do not increase domestic demand. That is, part of tax cuts received by households will be saved rather than spent and any consumer spending generated by tax cuts is more likely to go to imports than will government spending on infrastructure, school repair or government services.

Unemployment insurance: softening the blows, stimulating demand
Unemployment insurance plays two critical roles in the U.S. economy. It forms the first line of defense against income lost during periods of unemployment and it provides an automatic stimulus during periods of economic decline by sustaining consumption and therefore demand. By extending benefits for the long-term unemployed and improving eligibility to incorporate more workers into the system, policymakers can both cushion the impact of a recession on individuals and provide a counter-cyclical stimulus that will moderate the recession. Thus, a package of unemployment reforms should be part of any stimulus plan.

Softening the blow of unemployment
The current unemployment insurance system is an important but weak safety net. From the worker’s perspective there are few if any protections from lay-off. While some may be eligible for generous severance packages, most workers rely on income from unemployment insurance once they lose a job. The unemployment insurance program has weaknesses, but overall it has worked remarkably well since its creation in 1935. Research indicates that in the absence of unemployment insurance benefits many families would quickly descend into poverty. For moderate income households, unemployment insurance benefits lifted 20 percent of these households out of poverty (Danziger and Gottschalk, 1990). Nearly one-third of workers are unable to replace even ten percent of their lost income from savings, while those who receive unemployment insurance benefits draw down their savings and assets more slowly (Gruber 2002).

While unemployment insurance benefits are a much-needed palliative measure for those unfortunate enough to lose a job, they do not provide enough income for families to make ends meet. On average a single working parent with two children falls $1,317 short each month of the amount of money needed to maintain a minimal, no-frills living standard. A two-parent household with one parent unemployed and the other employed part-time at the median wage falls short of their family budget by $334 per month (Boushey & Wenger 2001). While unemployment insurance is very helpful it is not overly generous.

Currently, only 48 percent of workers who become unemployed will receive unemployment insurance. Those who receive UI can expect to remain unemployed for 14.7 weeks and receive $254 per week in benefits.

Unemployment insurance as stimulus
The second benefit that comes from unemployment insurance is its effect on the macro-economy. Unemployment insurance acts as an automatic stabilizer – injecting money into the economy as the labor market wanes and unemployment increases. During periods of low unemployment and economic expansion unemployment insurance benefits naturally decline and trust fund balances are increased. Thus, the program injects spending into the economy when the economy is weak and takes out spending when the economy is strong. Improvement in the system will not create long-term fiscal imbalances.

The stabilizing role of unemployment insurance is considerable; research by Chimerine et al. (1999:6) indicates that the unemployment insurance program “mitigated the loss in real GDP by about 15 percent over all the quarters in each recession.” Moreover, and importantly, the authors indicate that the effect of unemployment insurance in the 1990 recession was more robust than in the 1980’s recession. This was likely due in large part to the extended benefits program passed by the 102nd Congress in 1992.

Other research has concluded that unemployment insurance plays an important role in stabilizing the economy. Von Furtenburg (1976), Dunson et al. (1990), Uri, et al. (1989), and Vroman (1998) all find that increases in unemployment are met with commensurate increases in UI expenditures and that these expenditures result in significant economic stimulus during periods of economic decline.

The importance of the unemployment insurance system on macro-economic performance has been well documented. However, with the exception of the 1990-1991 recession, the policy and demographic changes have reduced the availability and generosity of unemployment insurance benefits – weakening its effect as a stabilizer. In particular, recipiency rates have declined from the late 1970’s. During 1975, the year of peak unemployment in the 1970s, the recipiency rate averaged 75 percent1. During 1992, at the end of the last recession, when unemployment peaked, the UI recipiency rate was 52 percent. Currently the percent age of unemployed workers receiving UI benefits is 48 percent.

Low levels of unemployment insurance eligibility are problematic for particular groups of workers. First and fo
remost, the lowest earners have a difficult time qualifying for UI benefits. These workers often earn low wages and have periodic interruptions in employment. As a consequence, many of them fail to earn sufficient income or work enough hours to qualify for unemployment insurance benefits. This means that many new entrants to the labor force, in particular women who previously received welfare, will be unable to qualify for unemployment insurance benefits.

The second group of workers who have limited access to unemployment insurance are workers employed on a part-time basis. These workers, most of whom are women, make up 20 percent of the US labor force as of July 2002. The earnings from these jobs comprise nearly 30 percent of household income – a significant loss of income should a part-time worker lose a job. Yet most part-time workers are ineligible to receive unemployment insurance benefits. Research indicates that only 17 states have rules allowing part-time workers to qualify for unemployment insurance benefits. (Wenger, McHugh & Segal 2002)

Apart from the problem of access are ongoing problems of recipients who exhaust benefits prior to finding a job. During first quarter of 2002, 36.9 percent of UI benefit recipients ran out of regular benefits, making them eligible for federal extended benefits (Tier 1 TEUC – Temporary Extended Unemployment Compensation). TEUC allowed many workers to receive as many as 13 weeks of additional benefits. It is estimated that 60 percent of those on extended benefits will exhaust their benefits before finding work. Currently, over one million people have exhausted all of their UI benefits.

The UI system of the United States should be significantly expanded to better serve all of the unemployed. Not only would this serve the interests of the families who have experienced job loss, but it would serve as a well-targeted and highly effective economic stimulus.

Offsetting the state fiscal crisis
There has been a larger shortfall in state budgets in this recession than in earlier recessions, according to the National Governors Association. States are drawing down their reserves and cutting spending and raising taxes in order to balance their budgets. Unfortunately, when states raise taxes and reduce services they reduce the overall demand for goods and services, thereby exacerbating the recession. The layoffs and spending reductions also curtail many needed services. A very useful way to restore demand in the economy would be to provide fiscal relief to state (and local) governments. This is administratively easy to do- it would take a staff of twenty-five to distribute $50 billion of fiscal relief to the states. This would also have an immediate impact, as plans to cut programs, raise taxes, or execute layoffs would be

School renovation
There is wide agreement that there’s a need to improve and expand school infrastructure. The GAO estimated the need for school repairs at $112 billion dollars in 1995. The problem has only gotten worse. There are existing programs that can be used to quickly expand school repair activity. One mechanism is to expand the repair fund program of the Individuals with Disabilities Education Act (IDEA). Given that many school districts are currently curtailing planned repairs, any new funding for this purpose can be implemented quickly. Another mechanism is to expand federal funding of the Qualified Zone Academy Bond (QZAB) program, which provides tax credits toward the interest on school construction bonds. We could expand school repair by $10 to $20 billion annually through these programs and not only increase spending and accelerate the recovery but also address a much needed investment shortfall in our schools.

The jobs and unemployment situation will only worsen as slow growth leads to higher unemployment, driving us further from achieving a much-needed goal of returning to 4 percent unemployment. The cost of the higher unemployment is lost wages and incomes to workers and their families, a further widening of inequalities, an inhospitable environment for welfare reform and the social costs of greater crime and worsened health. It is critical that policymakers address the unemployment problem by seeking a measured decline in the value of the dollar and through various temporary spending measures of about $100 billion each year for the next two years.

1Data from the Department of Labor, Employment and Training Administration and the Green Book, published by the Ways and Means Committee of the United States House of Representatives.


Boushey, H. and J.B. Wenger (2001). Coming Up Short Issue Brief 169. Washington DC: 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. Washington DC. US Department of Labor.

Danziger, Sheldon & Peter Gottschalk. (1990). Unemployment Insurance and the Safety Net for the Unemployed. in W.L. Hansen & J.F. Byers (eds.) Unemployment Insurance: The Second Half Century. Madison, WI: University of Wisconsin Press

Dunson, B. S.C. Maurice, and G. Dwyer. (1991). The Cyclical Effects of the Unemployment (UI) Program. Unemployment Insurance Occasional Paper 91-3, Washington D.C.: US Department of Labor.

Gruber, J. (2001). The Wealth of the Unemployed,” Industrial and Labor Relations Review, 55(1): 79-94
Uri, N.D., J.W. Mixon, B.L. Kyer. (1989). Automatic Stabilizers Reconsidered. Public Finance. 44(3):476-91
von Furstenberg, G.M. (1976). Stabilization Characteristics of Unemployment Insurance. Industrial and Labor Relations Review 29:363-76

Vroman, W. (1991). The Decline in Unemployment Insurance Claims Activity in the 1980s. Unemployment Insurance Occasional Paper 91-21, Washington DC: US Department of Labor.

Vroman, W. (1998). Labor Market Changes an Unemployment Insurance Benefit Availabilty. Report prepared for the U.S. Department of Labor under Contract No. F-5532-5-00-80-30.

Wandner, S.A. & A. Stettner. (2000). Whare are many jobless workers not applying for benefits? Monthly Labor Review. 123(6): 21-32

Wenger, J.B, R. McHugh and N. Segal. (2002) Part-Time Work and Unemployment Insurance. Indicators: The Journal of Social Health. 1(4):1-13


See related work on Economic Growth