POLICY ROUNDTABLE: The Challenge of Job Creation

Policy Paper
Oct. 18, 2009

What the Government Can Do to Better Promote Job Creation
by Timothy J. Bartik

Jobs: What Can We Do?
by James K. Galbraith

The Time Has Come for Direct Job Creation
by L. Randall Wray

What the Government Can Do to Better Promote Job Creation

Timothy J. Bartik, Senior Economist, W.E. Upjohn Institute for Employment Research

The federal government could significantly increase U.S. employment rates over the next two years using targeted job creation policies. Such policies could more quickly turn a jobless recovery into a recovery for workers.

Why is job creation needed now? First, the U.S. economy is far short of employment rates that are normal or desirable. Since the December 2007 start of the recession, U.S. employment to population ratios have declined from 62.7% to 59.2% (August 2009). This 3.5% reduction is equivalent to 8.3 million jobs.[1]

Second, for many disadvantaged Americans, employment conditions have been declining for 30 years. From 1979-2006, among prime-age (age 25-54) white men, the employment to population ratio dropped by 13 percentage points among high school dropouts and by 5 percentage points for those with only a high school degree. From 1979-2006, employment rates for prime age black men dropped by 21 percentage points for high school dropouts and 10 percentage points for those with a high school degree.[2]

Third, even though the U.S. economy may be beginning to recover, it will take several years for employment rates to significantly recover. In the last two recoveries, employment rates were still low for years after the recovery began. The 1990-91 recession ended in March 1991. The employment rate did not increase by more than 0.2% from its March 1991 level until January 1994. The 2001 recession ended in November 2001. The employment rate did not increase by more than 0.2% from its November 2001 level until October 2006. If today's economy follows these precedents, the U.S. employment rate may not significantly recover from current levels until 2012 or later.

The jobs problem is being addressed by the $787 billion fiscal stimulus package passed in February of 2009. But the stimulus is insufficient. According to the Council of Economic Advisors (CEA), the stimulus package created 1.0 million jobs by the third quarter of 2009.[3] But the 8.3 million job shortfall figure given above is calculated after those 1 million jobs are included. The CEA predicts that the stimulus will create an additional 2.5 million jobs by the fourth quarter of 2010.[4] This is far less than the 8.3 million jobs needed.

Another limitation on the job creation ability of the Recovery Act is its high cost per job created. Through 2012, it will cost about $112,000 to create one "job-year".[5]

Additional jobs could be created more cheaply through policies that target job creation. Two options are revised versions of the New Jobs Tax Credit (NJTC) used by the federal government in 1977-78, and the MEED (Minnesota Emergency Employment Development) Program used by the state of Minnesota from 1983 to 1989.

The NJTC provided wage subsidies via tax credits to private employers who expanded their employment. The wage subsidy was equivalent, in 2008 dollars, to a little over $7,000 per worker per year. The program subsidized the hiring of 2.1 million workers.

Research evidence suggests that the NJTC increased U.S. employment by about 700,000 workers, at a cost per new job created of about $20,000 in today's dollars.[6] These research findings are based in part on comparisons of job creation by businesses that knew about the NJTC versus businesses that did not. In addition, the design of the NJTC favored particular industries and small businesses, and these industries and businesses did relatively better during 1977-78. Finally, the timing of job growth in 1977-78 is consistent with NJTC having a significant effect.[7]

I have estimated that a revised NJTC, expanded to also subsidize non-profit employers, would increase U.S. employment by 1.3 million jobs. The estimated gross cost is $26 billion per year.[8] John Bishop has estimated that a differently revised NJTC, expanded to be more useful to large businesses, would create 1.5 million jobs at a cost of $55 billion per year.[9] Bishop's estimates are more pessimistic because he assumes more employers take up the subsidy offer, and he bases his estimates on general economics research on labor demand. My estimates rely on prior experience with the NJTC, which suggests somewhat higher responses and lower take-up rates.

Suppose these directly created jobs have an employment multiplier of 1.5. Then my estimates imply that a NJTC would create 1.95 million jobs, at a cost per job of $13,000. Bishop's estimates imply that a NJTC would create 2.25 million jobs, at a cost per job of $24,000.

Surveys indicate that more than 55% of participating employers would not have expanded without the MEED program.[10] This statistic is likely to be valid, because the program did not require that the jobs be created solely by the subsidy. This estimate that about half of all subsidized jobs were created due to the program is consistent with similar programs.[11]

Suppose a national MEED program was run at a scale of 1 million job slots. Then the program's estimated annual wage subsidy cost would be approximately $20 billion (=$10 per hour times 2000 annual work hours time 1 million job slots). Add in another 50% to pay for other program costs, including job developers who would help recruit and match employers and workers. Total program costs would then be $30 billion. Based on the Minnesota experience and other research, direct jobs created would be 550,000. With a multiplier of 1.5, total jobs created would be 825,000. Government costs per job created would be $36,000.

MEED is more expensive per job created than a NJTC. MEED requires more overhead, and its wage subsidy is bigger.[12] MEED's scale is more limited because of the needed administrative capacity and its clientele. But MEED's costs are more predictable because it is a spending program. MEED also is more targeted to create jobs for the low-skilled or disadvantaged.

Together, a revised NJTC and national MEED program could create around 3 million jobs. This would significantly augment the 2.5 million additional jobs that the stimulus may create by the fourth quarter of 2010. This would move us close to the 8.3 million jobs needed to restore December 2007 employment conditions.

Proposals for a revived NJTC have faced some skeptics. For example, Howard Gleckman of the Tax Policy Center of the Urban Institute and Brookings has argued that the evidence for success of the NJTC is "very limited".[13]

However, the evidence for the NJTC is similar to the evidence for any macro policy. The evidence for the NJTC includes survey results, correlations, and economic principles. Businesses that knew about the NJTC or had large benefits were more likely to expand in 1977-78. Significant effects of a NJTC are consistent with theories of labor demand.

Skeptics of wage subsidies such as MEED, which target the disadvantaged, worry about "stigma effects". The worry is that employers may avoid hiring workers who identify themselves as being disadvantaged. However, this criticism mainly applies to wage subsidy programs in which unscreened disadvantaged workers are sent out on their own to unscreened private employers. In the MEED program, job developers recruit employers who do not stigmatize disadvantaged workers, and work with employers and job seekers to identify good matches.[14]

There is good reason to think that job creation over the next two years will be well short of what is needed to restore pre-recession employment conditions, even with the fiscal stimulus. Targeted job creation policies, which subsidize employers to create jobs, could help. These programs could create millions of jobs at a cost per job created of less than $40,000. The question is whether the compelling need for job creation will force political leaders to consider policy solutions that do not easily fit into conventional ideologies.

Jobs: What Can We Do?

James K. Galbraith, Professor, LBJ School of Public Affairs, University of Texas at Austin

"From a technical perspective, the recession is ... over." So spake Bernanke, a few weeks back. Technically I do not disagree. Budget deficits, if they are big enough, work wonders. From the perspective of last year, it's remarkable that the deficit was allowed to get as big as it did.

Few realize how this happened. Obama didn't do it. The deficit doesn't owe that much to the stimulus bill, which is getting seriously started only now. Most of the deficit--and therefore most of the impetus to recovery--comes from unemployment itself.

Over six million jobs disappeared. The people who held them stopped paying taxes. They signed up for UI, disability, food stamps, Social Security and welfare. Production dropped, investment collapsed, inventories ran down--but consumer spending only declined a bit. And so when inventories run out, orders started up again--and the economy is jumping forward. It's ugly--extremely ugly--but it works.

Let me amend that. It works up to a point. Yes, economic growth resumes. But, as we shall shortly see, most jobs don't reappear, and many never will. Firms find it far easier to give existing staff more hours, or to find ways to produce more with less labor. In manufacturing, there is the rich potential of out-sourcing. It's no surprise that China's factories are rehiring already and ours are not.

Even so, the fate of manufacturing is not the real story. The manufacturing sector is small--about eight percent of employment. It will shrink even more, no matter what happens. Yes, the occasional tire tariff can keep some jobs in place, but that number is in the thousands, not the millions. Practically all the net new jobs of the past fifty years have been in other sectors.

So what about services? Here there's more bad news: three big job creators of recent times, in retail trades, finance and homebuilding, are not likely to bounce back soon. The retail boom was partly an artifact of the suburban explosion, and as housing imploded, so did the malls. And so have the banks.

So: what to do about unemployment?

First, there is no excuse for a single layoff in the state and local public sector. The people want--the country needs--the work of teachers, police, firemen, librarians, street sweepers and park rangers. The fact that federal tax revenues collapsed did not force the government to fire Members of Congress. The fact that local tax revenues have collapsed should not lead to cuts in music and arts in public schools. Enough already: the federal government should support state and local budgets, in full, until the crisis is over.

Second, expand higher education. American colleges are a tremendous buffer for the young, keeping them out of formal employment--and formal unemployment--for years into adulthood. Community colleges, public universities and graduate schools are also an excellent mid-life experience for many displaced workers. There is nothing wrong with higher education as an alternative to the dole--and it beats both the army and the prison system by a lot.

Third, let people retire sooner. Let's have no more talk about increasing the Social Security retirement age. The very last thing you want, in this situation, is to force the near-elderly to hold on to their jobs. As an emergency measure, cut the Medicare eligibility age to 55, clearing a good many near-elderly into early retirement. This would ease unemployment, open up chances for the young--and be a better health care reform than we're likely to get.

So much for starters. Obviously, we also need more jobs. How can we get them?

First, let's focus on our biggest problem--energy and global warming. There's work to do there. Much of it is labor-intensive, like weatherizing houses. Some of it is industrial, like converting the car fleet. Some of it is research, design and engineering, like building a smart electric grid and efficient public transportation. Some of it would go into the research, education and training required for these missions. All of it is necessary. Doing it right could give millions of people good jobs for decades.

Second, we have an aging population. Old people require care. Much of what they get, these days, especially in nursing homes, is pretty lousy. And when elderly parents have to rely on grown children--well, that can be quite hard on the children. How about a Home Care Corps? You could employ a lot of nurses, and even more nurses' aides, in such a program--and make millions of lives easier than they are.

Third, we have a foreclosure crisis. Millions of homes are being emptied out, prey to decay and to vandals. New slums are the very predictable result. How about a Neighborhood Conservation Corps? Such an entity could buy foreclosed houses at auction, restore, weatherize and maintain them, and rent them out, with an option to buy--in the first instance to their former owners. In the Depression, the Home Owners Loan Corporation employed 20,000 people to manage a million mortgages. Today we could create ten times that many decent jobs at least.

Fourth, about eight percent of all American jobs are in non-profits. (This has been a huge part of our jobs advantage over Europe.) All of this is under severe threat now, as collapsed asset values lead to budget cuts. Federalizing the non-profits would be crazy--but we've got an emergency. So how about plugging some of their funding gaps, with a Federal Foundation Funding Facility? The function would be similar to that of the Reconstruction Finance Corporation of yore: keeping hospitals, universities, and museums in business at (say) the average funding level of the previous three years.

Employers like to hire the already-employed. A federal jobs pool would give them a good place to look, where they could find workers with track records and supervisors able to give recommendations. An employed buffer stock of labor would be a great improvement over the present, vast mass of job-seeking unemployed. And when private employers finally decide they need labor, they would hire from the pool--and spending on the public program would automatically decline.

Would all of this add to the deficit? Yes. Providing jobs is more expensive, in budget terms, than keeping people on the dole. But that's not a bad thing right now.

And in a larger economic sense, it would be much cheaper. You'd save the cost of the dole. And you'd get three things out of it--economic goods that the entire country could enjoy, solutions to some of our most pressing problems, and a working population that would be working, acquiring skills, getting on with life--and no doubt happier, into the bargain.

The Time Has Come for Direct Job Creation

L. Randall Wray, Professor of Economics, University of Missouri-Kansas City

According to an ILO report[15] issued before the global economic crisis hit, even though more people were working than ever before, the number of unemployed was also at an all time high of nearly 200 million. Further, "strong economic growth of the last half decade has only had a slight impact on the reduction of workers who live with their families in poverty...", in part because the growth was fueling productivity growth (up 26% in the past decade) but was not creating many new jobs (up only 16.6%). The report concluded: "Every region has to face major labour market challenges" and that "young people have more difficulties in labour markets than adults; women do not get the same opportunities as men, the lack of decent work is still high; and the potential a population has to offer is not always used because of a lack of human capital development or a mismatch between the supply and the demand side in labour markets." All of these statements applied equally well to the United States even at the peak of our business cycle in early 2008.

Now, of course, our labor market is in dire straits--having lost more than 6 million jobs, with official unemployment approaching 10%, and with millions more workers facing reduced hours and even reduced hourly pay. According to a New America Foundation report[16] released late last spring, if we add "marginally attached" workers, those forced to work part-time, and those who would like to work but have given up looking, the effective unemployment total is over 30 million. Add to that another 2 million incarcerated individuals--many of whom might have avoided a life of crime if they had enjoyed better economic opportunities, and it is likely that a more accurate measure of the unemployment rate would be about 20%.

These numbers are similar to those I obtained for the Clinton boom when I estimated how many potential workers remained jobless even when the economy was supposedly at full employment.[17] Labor force participation rates--the percent of working age population that is employed or unemployed--vary considerably by educational level; high school dropouts have very low participation rates, and correspondingly high incarceration rates. I calculated that as many as 26 million more people would be working if we brought labor force participation rates of all adults up to the levels enjoyed by college graduates. That number would be higher now because of lackluster job creation during the Bush years and due to the economic crisis. Thus, we can safely conclude that whether the US economy is booming or busting, it is chronically tens of millions of jobs short.

Comparing such numbers with President Obama's promise that his policies will create, or at least preserve, three or four million jobs demonstrates that current policy is not up to the task of dealing with our labor market problems. To be sure, there is no single labor market policy that can deal with the scope of our problems. We certainly need to resolve the financial crisis and to restore economic growth. But as experience demonstrates, even relatively robust growth does not automatically create jobs.

We also have severe structural problems: some sectors, such as manufacturing, will create far too few jobs relative to the supply of workers with appropriate skills, while others, such as the FIRE sector--finance, insurance and real estate--likely should be downsized, and still others, such as nursing and trained childcare, face a chronic shortage. Finally, it could be argued that we face another kind of structural problem identified a half century ago by John Kenneth Galbraith: a relatively impoverished public sector and a bloated for-profit sector. Thus, while recognizing the multi-faceted nature of our problem, I believe that direct job creation by government would go a long way toward resolving a large part of--and probably the worst of--our unemployment problem even as it could put people to work to provide needed public sector services.

Direct job creation programs have been common in the US and around the world. Americans immediately think of the various New Deal programs such as the Works Progress Administration (which employed about 8 million), the Civilian Conservation Corps (2.75 million employed), and the National Youth Administration (over 2 million part-time jobs for students). Indeed, there have been calls for revival of jobs programs like VISTA and CETA to help provide employment of new high school and college graduates now facing unemployment due to the crisis.[18]

But what I am advocating is something both broader and permanent: a universal jobs program available through the thick and thin of the business cycle. The federal government would ensure a job offer to anyone ready and willing to work, at the established program compensation level, including wages and benefits package. To make matters simple, the program wage could be set at the current minimum wage level, and then adjusted periodically as the minimum wage is raised. The usual benefits would be provided, including vacation and sick leave, and contributions to Social Security.

Note that the program compensation package would set the minimum standard that other (private and public) employers would have to meet. In this way, public policy would effectively establish the basic wage and benefits permitted in our nation--with benefits enhanced as our capacity to provide them increases. I do not imagine that determining the level of compensation will be easy; however, a public debate that brings into the open matters concerning the minimum living standard our nation should provide to its workers is not only necessary but also would be healthy.

The federal government would not have to micromanage such a program. It would provide the funding for direct job creation, but most of the jobs could be created by state and local government and by not-for-profit organizations. There are several reasons for this, but the most important is that local communities have a better understanding of needs. The New Deal was more centralized, but many of the projects were designed to bring development to rural America: electrification, irrigation, and large construction projects. To be sure, we need infrastructure spending today, but much of that can be undertaken by state and local governments. This program would provide at least some of the labor for these projects, with wages and some materials costs paid by the federal government.

More importantly, today we face a severe shortage of public services that could be substantially relieved through employment at all levels of government plus not-for-profit community service providers. Examples include elder care and childcare, playground supervision, non-hazardous environmental clean-up and caring for public space, and low-tech improvement of energy efficiency of low-income residences. Decentralization promotes targeting of projects to meet community needs--both in terms of the kinds of programs created but also in terms of matching new jobs to the skills of unemployed people in those communities. Also note that by creating millions of decentralized public service jobs, we avoid one of the major criticisms of the stimulus package: because there were not enough "on the shelf" infrastructure-type projects, it is taking a long time to create jobs. Instead, we should allow every community service organization to add paid jobs so that they can quickly expand current operations.

As the economy begins to recover, the private sector (as well as the public sector) will begin to hire again; this will draw workers out of the program. That is a good thing; indeed, one of the major purposes of this program is to keep people working so that a pool of employable labor will be available when a downturn comes to an end. Further, the program should do what it can to upgrade the skills and training of participants, and it will provide a work history for each participant to use to obtain better and higher paying work. Experience and on-the-job training is especially important for those who tend to be left behind no matter how well the economy is doing. The program can provide an alternative path to employment for those who do not go to college and cannot get into private sector apprenticeship programs.

There are some recent real world examples of programs that are similar to the one I am proposing. When Argentina faced a severe financial, economic, and social crisis early this decade, it created the "Jefes" program in which the federal government provided funding for labor and a portion of materials costs for highly decentralized projects, most of which created community service jobs.[19] The program was targeted to poor families with children, allowing each to choose one "head of household" to participate in paid work. The program was up-and-running in a matter of four months, creating jobs for 14% of the labor force--a remarkable achievement. More recently, India has enacted the National Rural Employment Guarantee, which ensures 100 days of paid work to rural adults. While the program is limited, it does make an advance over the Jefes program: access to a job becomes a recognized human right, with the government held responsible for ensuring that right.

Indeed, the United Nations Universal Declaration of Human Rights includes the right to work, not only because it is important in its own right, but also because many of the other economic and social entitlements proclaimed to be human rights cannot be secured without paying jobs. And both history and theory strongly indicate that the only way to secure a right to work is through direct job creation by government. This is not, and should not be, a responsibility of the private sector, which employs workers only on the expectation of selling output at a profit. Even if we could somehow manage economic policy to produce a permanent state of boom, we know that will still leave tens of millions of potential workers unemployed or in part-time and underpaid work. Hence, a direct government job creation program is a necessary component of any strategy of ensuring achievement of many of the internationally recognized human rights.

[1] All calculations of employment rates (employment to population ratios) for the aggregate U.S. economy use seasonably adjusted monthly employment rates for ages 16 and over from the U.S. Bureau of Labor Statistics.

[2] P. 5 of Bartik, Timothy and Susan Houseman (2008), "Introduction and Overview", in A Future of Good Jobs? (edited by Timothy Bartik and Susan Houseman), Kalamazoo: Upjohn Institute.

[3] Council of Economic Advisors, "The Economic Impact of the American Recovery and Reinvestment Act of 2009, First Quarterly Report", September 10, 2009.

[4] Council of Economic Advisors, "Estimates of Job Creation From the American Recovery and Reinvestment Act of 2009", May 2009.

[5] This combines data from the CEA and CBO. In its May 2009 report, the CEA estimates 6.8 million job-years created by the end of 2012. The March 2,2009 CBO analysis implies about $762 billion in fiscal stimulus will be used by the end of 2012. (This adds one-quarter of CBO's estimates for FY 2013 to the estimates for previous fiscal years.) Congressional Budget Office, "Estimated Macroeconomic Impacts of the American Recovery and Reinvestment Act of 2009", Letter to Senator Charles Grassley, March 2, 2009.

[6] See review of evidence on pp. 226-228 of Bartik, Timothy (2001), Jobs for the Poor, Kalamazoo: Upjohn Institute. Among the relevant studies are Perloff, Jeffrey, and Michael Wachter (1979), "The New Jobs Tax Credit: An Evaluation of the 1977-78 Wage Subsidy Program", American Economic Review 69 (2) : 173-179; Bishop, John (1981), "Employment in Construction and Distribution Industries: The Impact of the New Jobs Tax Credit". In Studies in Labor Markets (edited by Sherwin Rosen), Chicago: University of Chicago Press.

[7] Evidence of these effects of the NJTC is reviewed in Bishop, John (2009), "How to Restore Animal Spirits and Reduce Unemployment", working paper, School of Industrial and Labor Relations, Cornell University.

[8] Bartik, Timothy (2008), "The U.S. Economic Crisis and a Revised New Jobs Tax Credit". Upjohn Institute memo, October 16, 2008.

[9] Bishop, John (2009), "How to Restore Animal Spirits and Reduce Unemployment".

[10] Rode, Peter (1988), "MEED Means More Business", Minneapolis: Jobs Now Coalition.

[11] Pp. 235-241 in Bartik, Timothy (2001), Jobs for the Poor.

[12] MEED provides up to a $10 wage subsidy for up to 6 months, which will total $10,000. The MEED jobs paid only 25% more than the subsidy, so the subsidy averaged 75% of wages. The revised NJTC that I examined is a $7000 wage subsidy, but paid out over one year, and the jobs are likely to have average pay closer to economy wide average wages of $19.39 (data from U.S. Bureau of Labor Statistics for average private sector wages, 2009II). The average subsidy percentage is 18% (=7000/(19.39 times 2000). Bishop considers a 15% wage subsidy.

[13] January 9, 2009 entry on Tax Vox blog by Howard Gleckman, at http://taxvox.taxpolicycenter.org/blog/_archives/2009/1/9/4051427.html.

[14] Stigma effects are discussed on pp. 220-224 of Bartik, Jobs for the Poor. The best known study arguing for stigma effects is Burtless, Gary (1985), "Are Targeted Wage Subsidies Harmful?", Industrial and Labor Relations Review39(1): 105-114.

[15] Global Employment Trends Brief 2007, International Labour Office; results summarized in "Global Unemployment Remains at Historic High Despite Strong Economic Growth", ILO 25 January 2007, Geneva. See also The Employer of Last Resort Programme: Could it work for developing countries?, L. Randall Wray, Economic and Labour Market Papers, International Labour Office, Geneva, August 2007, No. 2007/5.

[16] Not Out of the Woods: A Report on the Jobless Recovery Underway, New American Contract, New America Foundation, 2009, www.newamericancontract.net.

[17] Can a Rising Tide Raise All Boats? Evidence from the Kennedy-Johnson and Clinton-era expansions, L. Randall Wray, in Jonathan M. Harris and Neva R. Goodwin (editors), New Thinking in Macroeconomics: Social, Institutional and Environmental Perspectives, Northampton, Mass: Edward Elgar, pp. 150-181.

[18] See Not Out of the Woods, referenced above.

[19] See Gender and the job guarantee: The impact of Argentina's Jefes program on female heads of households, Pavlina Tcherneva and L. Randall Wray, CFEPS Working Paper No. 50, 2005.