UNCOMMON SENSE 14 © rev. October 2001
#1 of FAIR WORK AND WELFARE: a welfare reform packet
By Philip Harvey, Assoc. Prof. of Law, Rutgers University School of Law (Camden), and Advisory Board, National Jobs for All Coalition
In October 2000 the national unemployment rate stood at 3.9 percent. Ten months later it had risen to 4.9 percent. Thus, even before the September 2001 attack on the World Trade Center, all of the reductions in the nation’s rate of unemployment achieved in the preceding three years had been lost, and there is every reason to expect the situation to deteriorate even more rapidly in the wake of the attack. This trend will put the problem of unemployment back on the public’s political agenda, but it is important to remember that even with unemployment at the 3.9 percent level achieved in the fall of 2000, over 5.5 million active job seekers were entirely without work, over 2.5 million were employed part-time even though they wanted full-time jobs, and over 4 million people reported that they wanted a job even though they were not actively looking for one.
Even in periods of general prosperity, there are not enough jobs to satisfy the needs of everyone who wants to work, and the burdens of this joblessness tend to fall disproportionately on disadvantaged population groups. With unemployment rising, millions more will suffer, and this increase, too, will be born disproportionately by members of disadvantaged population groups. As of September 2001, before the attack on the World Trade Center, official unemployment had risen by 1.5 million compared to a year before and the number of involuntary part-time workers had grown by another million.
While jobless individuals and their dependents endure the most harm from the economy’s failure to provide work for everyone who wants it, they are not they only ones to suffer. We all share in the sacrifice. We are poorer as a society because millions of people who want to work are kept idle instead. Our taxes are higher because there are fewer workers to pay taxes. And government expenses are higher because of the benefits and services which must be provided to deal with the multitude of problems that unemployment either causes or aggravates.
Why doesn’t our government simply create the additional jobs needed to ensure that everyone who wants to work can do so? There is a variety of considerations that keep us from taking that step, and I shall address only one of them here-concern about the cost of the needed job-creation effort.
There is no question that job creation is expensive. What most people fail to take into consideration is that unemployment costs governments money, too. The true cost of creating jobs for everyone who wants to work is the difference between the cost of creating the jobs and the costs of unemployment that governments already bear. When those costs are compared, the surprising conclusion is that creating jobs for all might end up saving taxpayers more money than it would cost them.
The Costs of Joblessness
Is it possible to finance a large-scale initiative without adding to what we
now spend coping with joblessness? The answer may be yes. The key to understanding why lies in an assessment of the costs of leaving people unemployed.
1. Governments lose the taxes that jobless people would pay if they were working. The result is higher taxes on others, lower public expenditures, or larger deficits.
2. Government revenues pay for income-assistance benefits to the jobless, like unemployment insurance or food stamps.
3. We lose the goods and services that the jobless would produce if they were employed. The result is a lower standard of living for society as a whole.
4. We bear a myriad of indirect costs attributable to the social problems that joblessness aggravates–from increased criminal-justice costs to increased health-care expenditures and family breakdown.
5. We suffer deeper recessions because rising unemployment cuts consumer spending, further discouraging business.
The fact that joblessness imposes costs on society and that governments already bear a substantial portion of these costs raises the possibility that government job creation might be at least partially self-financing.
The Cost of Job Creation
I have explored this idea, analyzing both the costs of a large-scale job-creation initiative and the additional public-sector revenues and savings it would generate. Although the numbers have changed somewhat since then, the basic conclusion still applies.
I asked what it would have cost to operate a program capable of providing employment to every job-seeker unable to find work in the regular labor market between 1977 and 1986–a period of unusually high average unemployment. I included not only officially unemployed workers but also involuntary part-time workers, discouraged workers, and able-bodied welfare recipients not already counted in the labor force.
If the program offered employment at market wages (the average wage earned by similarly qualified individuals when employed), program wages would have averaged about $9.25 per hour in 1986, or almost $19,000 per year (both in 1999 dollars). Other assumed program features included the following:
1. Program participants would pay the same Social Security taxes and receive the same Social Security benefits as other workers (with the program paying the employer’s share of the payroll tax);
2. Program participants would be offered the same health insurance benefits as federal government employees and on the same terms;
3. Free, high-quality child care would be provided for all working parents;
4. Paid job training would be offered to anyone who initially could not qualify for a job paying at least a poverty-level income, as well as to others who need training to provide needed services;
5. Individuals who could not function in a normal working environment would be offered employment, either temporarily or permanently, in a sheltered workshop or other appropriate setting.
Not surprisingly, such a program would have cost a lot of money– less than $225 billion a year average (in 1999 dollars) over the ten-year period analyzed. Although very large, it is worth noting that this level of social-welfare spending is not unprecedented. We regularly spend more on Social Security benefits. In 1986, for example, the job program would have cost $143 billion in current dollars compared to $194 billion for Social Security (not including Medicare).
Paying the Bill
Costs would have been substantially higher for this hypothetical jobs program but for an interesting side-effect. The program would eliminate certain expenses that add to the cost of providing jobs in the regular labor market. Consider child care, for example. Child care is an expensive benefit when purchased from third-party providers, but a comprehensive job program could operate child-care centers as one of its job-creation initiatives. Free, high-quality child care could then be offered to program participants without adding a penny to total program job costs. The same is true of the costs of job training. Creating jobs for trainers would reduce the number of other positions the program would have to create. This means program participants could be furnished ample training without increasing program costs.
The program wouldn’t have to limit its hiring to currently unemployed individuals either. If someone were hired away from another job because the program needed a person with particular skills (e.g., a child-care professional), the job that person left behind would become available to someone else. This vacancy would substitute for a job the program otherwise would have to create. Projects also could be started slowly, permitting program participants to gain experience. This would reduce the program’s total output, but it wouldn’t increase its total cost. The latter would depend on the total number of jobs created, not the distribution of participant assignments between training and direct production.
Administrative costs and most other support services in the program could be financed in the same way. Even the cost of purchased goods and services (e.g., child care supplies and building materials) would be partly internalized since jobs created in the private sector to produce these goods and services would substitute for jobs the program otherwise would have to create.
The bottom line is that a comprehensive jobs program could offer employment at a lower cost per job than would be possible in a more narrowly targeted program. Given the persistence of the economy’s job deficit, many of these jobs could be essentially permanent. Others would be temporary, designed to provide work for cyclically unemployed workers.
The results of my job-creation study were a pleasant surprise. $225 billion is not chicken feed. Yet, according to my calculations, these costs would have been offset almost entirely by lower government costs and increased revenues resulting as people moved from unemployment to jobs.
Begin with taxes. I estimate that about 20 percent of program costs would have been immediately recouped through increased income and Social Security tax payments by program participants to all levels of government. Another 60 percent of the program’s cost would have been covered by funds actually spent between 1977 and 1986 on unemployment compensation and means-tested income assistance provided to able-bodied adults of working age and their dependents.
After accounting for additional tax receipts and income-assistance savings, the job program’s remaining funding deficit would have averaged only 20 percent of total program costs, an average of $43 billion per year (in 1999 dollars). That’s not an unreasonable amount to pay for full employment. Even this amount, however, assumes that everything the program produces would be given away for free. If anything at all were charged for the goods and services produced, the program’s funding needs would be reduced, very possibly to less than zero.
In the period I studied, for example, the program would have produced an average of about $225 billion of goods and services a year (expressed in 1999 dollars). This level of output could include child care not only for program participants but for other working parents. It could include low-cost housing for homeless families, additional maintenance and recreational staff to work in our parks, additional classroom and lunchroom aides for our schools, home-care aides for the elderly, conservation workers, and so forth. Those who worry about the “end of work” or a “jobless future” can be reassured. There is no shortage of work that needs to be done. The problem is having the political will to translate real needs into real jobs.
The public might not be willing to pay $225 billion a year for these goods and services, but it would be willing to pay something. If it had paid the $43 billion per year needed to close the program’s remaining budget gap, it would have gotten the program’s output for an average of 20 cents on the dollar.
The Macro-Economic Fine Print
Economists will note that such a program would have increased disposable household income by an average of close to $43 billion a year (the program’s total cost less direct taxes paid by the newly employed and income-assistance benefits they would have received had they remained jobless). This would have provided the necessary income and a rationale (the prevention of excess economic stimulus) for charging fees for some portion of the goods and services the program would have provided.
Account also needs to be taken of the likely counter-cyclic effect of such a jobs program. In the period studied, almost half of the estimated ten-year funding deficit would have been incurred in the recession years of 1982 and 1983. A jobs program that offered employment to all job-seekers unable to find work would be a powerful automatic stabilizer. If it had been in place during the early 1980’s, the deep recession of that period almost surely would have been moderated. This would have reduced the program’s cost, but a bigger payoff for society would have come in the form of higher aggregate income levels. The nation’s gross domestic product fell by $121 billion (1999 dollars) between 1982 and 1983. Compared to the 3 percent expansion we might have had, the loss was $288 billion.
This is not to say the recession would not have occurred if a comprehensive jobs program had been in place. It is just to indicate that any reduction in the severity of recessions attributable to such a program could generate a large social payoff.
Finally, indirect savings likely to flow from reduced joblessness have not been taken into account here. There are countless ways in which joblessness is harmful to both individuals and families. Its negative effects on physical and mental health are well documented, and it is implicated in the full range of social problems associated with poverty. These problems weigh most heavily on the unemployed and their families, but the public shares the cost. Government spending is affected in areas as diverse as health care, education, criminal justice, and social-service delivery. While the indirect savings and benefits society would enjoy from reduced unemployment are hard to quantify, they are undoubtedly large.
Good News and Bad News
When all these factors are considered, it is hard not to conclude that a jobs program capable of providing work for everyone who needs it would not cost us money. It would save us money. Existing policies are more expensive. A comprehensive job-creation program would lighten the burden on taxpayers. As with the provision of universal health insurance, making sure that all Americans have jobs they can live on is not only the humane and just thing to do, it’s just plain smart.
The bad news is that it would be very hard to mobilize the resulting pool of savings. The savings would show up in thousands of budget lines scattered across all levels of government as well as in real income gains by ordinary tax payers. No one level of government would save enough to fund such a program on its own. Gathering the money saved by a large-scale job-creation initiative into a single pot would be a gargantuan political task. In conceiving specific funding devices for job creation, proponents should look for ways these savings could be gathered together.
There also are other challenges to winning public support for job-creation. Given the association of low levels of unemployment with higher rates of inflation, questions would be raised about the inflationary effect of such a program. This is not a simple issue to analyze. The inflationary effects of the direct job-creation described here would be less than those of the more conventional way of creating jobs through an increase in overall spending (possibly financed by increased government budget deficits.) Job creation targeted to unemployed labor and depressed areas can avoid spot shortages that can lead to inflation. However, the bargaining strength of employees would increase in both instances, resulting in wage increases. These, if not offset by productivity increases or a reduction in the share of profits, would trickle through to prices. However, in the absence of total spending growing beyond capacity and/or more expansionary monetary policy, it is unlikely that a wage/price spiral would persist for long.
The net effect of a job-creation initiative such as I have described–following an inflationary adjustment period–might be a redistribution of income in favor of wage earners, particularly at the lower end of the wage scale, without much change in underlying inflationary tendencies. This is worth exploring in more depth. If inflation were a problem, other inflation-controlling techniques are preferable to the socially destructive policy of intentionally increasing joblessness.
Other issues that would have to be addressed in persuading the public that deliberate job creation makes sense include the question of whether there really is a job shortage in the economy, whether government has the capacity to administer a successful job-creation initiative, whether unemployed workers really are “employable,” whether a direct job-creation effort would adversely affect regular government employees, and whether such an initiative would tend to undermine labor productivity and discipline in the private sector. Even if the money could be found to fund a major program, the importance of these other questions should not be underestimated.
Nevertheless, in today’s political climate, cost is a key issue. Proponents of job creation are unlikely to make much headway in promoting deliberate job creation as long as the public believes someone is going to have to take a big tax hit to pay for it. On that score, the public can be reassured. Notes available on request. _______________________________________________________
*A version of this article appeared in Social Policy, Spring 1995.