-- About CFEPS --
-- Contact Us --








Request a copy of this publication
Public Service Employment: Full Employment Without Inflation
Working Paper No. 3
Jan 2000
L. Randall Wray (info)

The Conservative belief that there is some law of nature which prevents men from being employed, that it is 'rash' to employ men, and that it is financially 'sound' to maintain a tenth of the population in idleness for an indefinite period, is crazily improbable -- the sort of thing which no man could believe who had not had his head fuddled with nonsense for years and years.... Our main task, therefore, will be to confirm the reader's instinct that what seems sensible is sensible, and what seems nonsense is nonsense. We shall try to show him that the conclusion, that if new forms of employment are offered more men will be employed, is as obvious as it sounds and contains no hidden snags; that to set unemployed men to work on useful tasks does what it appears to do, namely, increases the national wealth; and that the notion, that we shall, for intricate reasons, ruin ourselves financially if we use this means to increase our well-being, is what it looks like -- a bogy. (Keynes 1972, pp. 90-92)
Since WWII, it has been the stated policy of the US government to simultaneously pursue high employment and stable prices. Paradoxically, neither accepted economic theory nor practical experience appears to indicate that high or full employment is even possible with stable prices. As a result, for at least the past two or three decades, monetary policy generally has been geared toward raising the unemployment rate as a means to achieving stable prices; unemployment is perceived as the inevitable cost of price stability. Many, perhaps most, economists doubt that it is even possible to achieve anything close to a 3% unemployment rate without at the same time inducing accelerating inflation.

We will argue that stable prices and truly full employment are indeed possible. Indeed, the government can guarantee a zero unemployment rate, defined as all who are ready, willing, and able to work at a socially-established basic wage will be able to find a job. Only those unwilling (or unable) to work at the going wage would be left without work (which are not normally counted as unemployed). At the same time, by setting this basic wage, the government will provide a price anchor that will impart some price stability to the system. We do not claim that this policy would cause any particular price index to remain constant over time, thus, inflation could certainly co-exist with our proposed full employment policy, but would not be caused by the policy. Rather, we will show that a true full employment policy is not, in itself, "inflationary" and indeed could reduce inflationary pressures under some conditions. Further, the full employment policy would help to reduce economic fluctuations (the "business cycle") through a powerful built-in automatic stabilizer.

Before proceeding, it is necessary to admit that our proposed policy could lead to an increase of government spending; indeed, a persistent government deficit could result. However, we do not view this result with horror--as would many economists. While we do not have the space to go into this in detail, we take the position that there is nothing inherently wrong with big deficits; these do not necessarily cause "crowding out", they do not "burden" future generations, and they cannot lead to "financial ruin" of the government. In our view, fear of deficit spending is irrational and should never be allowed to stand in the way of the spending that may be required to generate full employment. This is not to say that deficits cannot be too large. Once an economy is operating beyond full employment, any increase of aggregate demand (whether by government or by the private sector) might be inflationary. This has (but only rarely) been the case in the past. However, conditions today make it easier to pursue a policy of full employment. Most importantly, the universal abandonment of the gold standard by all of the large economies has virtually eliminated all rational barriers to deficit spending as a means to hire all of the unemployed as governments no longer need to worry about gold-backing of currency (and possible loss of gold reserves).

One further caveat. Estimates of cost, number of participants, effects on other social spending, and so on, are all based on US data. However, we believe the analysis can be applied to any developed economy without much difficulty. There is one area in which the analysis here is seriously deficient, however. The US is a large and open economy, but exports and imports are still a small percent of GDP; in addition, the dollar is the international reserve currency. It is possible that implementation of a program such as that outlined here might have an impact on exchange rates and/or on trade position. For the case of the US, we do not believe the impact would be significant. However, for a country like Australia, the impact could be larger. We do not attempt to analyze such impacts in this chapter. See Mitchell (1997), Mitchell and Watts (1997), and Mosler (this volume) for analyses that do attempt to examine the case of small open economies, or that explicitly examine effects on exchange rates.

In the next two sections, we turn to the two primary components of the proposal: the government would
  1. act as employer of last resort, and
  2. exogenously set the "marginal" price of labor.
In the final section we examine possible objections to the proposal and outline the types of activities that might be undertaken by those employed.


The first component of the proposal is relatively simple: the government acts as the employer of last resort, hiring all the labor that cannot find private sector employment. As Minsky said
The policy problem is to develop a strategy for full employment that does not lead to instability, inflation, and unemployment. The main instrument of such a policy is the creation of an infinitely elastic demand for labor at a floor or minimum wage that does not depend upon long- and short-run profit expectations of business. Since only government can divorce the offering of employment from the profitability of hiring workers, the infinitely elastic demand for labor must be created by government. (Minsky 1986, p. 308)
Deviating somewhat from Minsky, we will call this the Public Service Employment (PSE) program, rather than an employer of last resort program, which might bring with it negative connotations. As will be discussed in the next section, the government simply announces the wage at which it will hire anyone who wants to work in the public sector, and then hires all who seek employment at that wage. We will call this the basic public sector employment (PSE) at the basic public sector wage (BPSW). Of course, there will still remain many (non-PSE) jobs in the public sector that are not a component of the PSE and that could pay wages above the BPSW. It is also important to emphasize that PSE policy is not meant to substitute for current public sector employment (PSE workers should not displace current public employees).

The implications for wages and prices, in general, will be explored below. Here we only discuss the implications for employment and the government's budget. For the sake of our discussion in this section, we will assume that the government's announced wage (BPSW) is $6.25 per hour or $12,500 per year for full-time (PSE) employment. We will also assume that this is a "living" wage, and that it is the legal minimum wage that exists at the time the PSE program is implemented. As we briefly discuss below, careful analysis should be undertaken before establishing the BPSW. There is no reason why some individuals might not be allowed to work part-time. However, we will assume throughout that employment is full-time to simplify calculations.

This policy will as a matter of logic eliminate all unemployment, defined as workers willing to work at the BPSW but unable to find a job even after looking. We define this as a state of full employment, or zero unemployment. One implication of PSE is that much social spending that is currently targeted to the unemployed might be reduced or eliminated. For example, unemployment compensation currently provides some income replacement for those who are unemployed. The program has only partial coverage (most of the unemployed are not covered), limited benefits (determined in part by income earned while employed), and time limits, and pays people for not working (generating obvious incentive problems). If instead, unemployment compensation were replaced with PSE, all of the disadvantages of unemployment compensation would be eliminated.

A less extreme change would allow newly unemployed workers the option of engaging in full-time job search in the PSE program for a specific period, say, for six weeks. If a job was not found within this time frame, the individual would undergo counseling and assessment to determine whether continued full time search was warranted; alternatively, retraining or education might be indicated (for example, if the individual's skills did not match job opportunities). In this case, the individual might be placed into a full-time PSE job to obtain on-the-job training; or, the individual might be enrolled in a part-time or full-time educational program. Again, there could be time limits for such programs; at some point the individual would be placed into an appropriate PSE job. As the primary goal of PSE is to prepare workers for employment in non-PSE jobs (whether public or private employment), all PSE jobs should contain at least some training. Thus, PSE could provide something similar to "unemployment compensation", but would differ from the current program in three significant ways. First, coverage could be universal (eg: all newly unemployed would qualify, regardless of the reason for unemployment); second, the job search would be more closely monitored and assisted (for example, each PSE worker in the job search program would be expected to devote a full eight hours each workday to job search--phoning for interviews, developing a cv, completing applications, and attending interviews); and third, the "unemployment compensation" would be equalized (each would receive $6.25 per hour). Clearly, some newly unemployed workers would "opt out" of the PSE program, either because they have negotiated sufficient privately-supplied unemployment benefits (or severance pay), or because they have amassed sufficient savings to enable them to pursue full-time job search. We are concerned only with those who would voluntarily choose to participate in the PSE program.

In addition, at least some spending on other types of social programs could be reduced, such as "welfare", broadly defined. Obviously, the PSE policy is not a substitute for these programs--many individuals currently receiving such assistance are not (and probably could not be) in the labor force. Exactly who would be forced out of these current programs and into the PSE program is a subject of social policy but is beyond the scope of this study. We emphasize that our concern is with those who are ready, willing, and able to work, but who are not able to find a job.

The PSE will also eliminate the need for a statutory minimum wage, as the BPSW will become an effective minimum wage. Indeed, it will have complete coverage, unlike the current minimum wage law, as any worker can always choose to accept PSE. (As Hyman Minsky always argued, if there is any unemployment, the effective minimum wage is zero.) The implication of PSE for the private sector wage is the subject of the next section. Note that a PSE program could also provide a path to de facto universal health care coverage and to provision of universal child care. If the PSE compensation included health and child care benefits, then private sector jobs would also have to provide them (or a salary sufficient to induce workers to forego such coverage).

If we assume that in the current economic environment, 8 million unemployed workers (not all of whom would be officially counted as unemployed) would be willing to accept the BPSW in PSE jobs, the total wage cost to the government would be $100 billion. In addition, as discussed above, PSE might include health care costs or other benefits (child care, transportation to work); this would add to costs (health care benefits could nearly double program costs -- but would also substantially reduce health care costs of current programs, such as Medicare and Medicaid). There would also be administrative costs of the program, which would have to be created, monitored, and evaluated as well as capital costs of equipping PSE workers with necessary equipment and supplies. On the other hand, PSE would lead to savings in a variety of government programs, from unemployment insurance to "welfare" (cash assistance and foodstamps to needy families). We have elsewhere estimated that the net cost of PSE to the government would fall between $25 billion and $50 billion (total expenses in excess of $100 billion, with savings in excess of $50 billion).

Note that we are not including a variety of possible social and private benefits associated with lowering unemployment rates. For example, it is widely recognized that long term unemployment contributes to crime, child abuse, divorce, loss of human capital, and other social and private degradation (including insecurity even of the employed) that may be hard to economically value. Thus, we expect that our estimate of a net cost of $50 billion probably overstates costs, but nothing of substance would change in our analysis even if costs were two or three times greater (or half as much)--economically it would not matter, although it might matter politically.

Obviously, the budgetary effects of the PSE are quite small, relative to the size of the Federal budget, to the size of the Reagan or Bush deficit, and to the size of GDP. We won't provide a detailed rejoinder to the "deficit-busting" arguments of those who advocate balanced budgets. An important question, however, concerns the impact this program would have on aggregate demand: is full employment going to increase aggregate demand sufficiently that accelerating demand-pull inflation would follow? If in the absence of PSE, public plus private sector spending provides a level of employment that leaves 8 million workers involuntarily unemployed, this must be evidence that the deficit is too small. This means that the government can safely increase its deficit spending, lowering involuntary unemployment, to increase aggregate demand. Of course, the PSE program is designed to ensure that the deficit will rise only to the point that all involuntary employment is eliminated; once there are no workers willing to accept PSE at the BPSW, the deficit will not be increased further. Thus, the design of the PSE guarantees that the deficit will not become "excessive".

It might be objected that as the government implements PSE and begins employing some of the 8 million unemployed, this will raise aggregate demand and thus increase private sector employment. This, of course, is true and is desired as it will ultimately reduce the amount of PSE required. By stimulating demand (through the "spending multiplier"), PSE may find that only 4 million workers will eventually accept PSE. Still, PSE automatically operates to ensure that the deficit spending attributable to PSE is at the correct level since every private sector job created automatically reduces PSE by approximately one job and the deficit by at least the cost of an PSE job (and probably more as tax revenues rise and government spending falls).

This should eliminate the fear that a full employment policy must necessarily generate excessive demand-pull inflation. Of course, it can still be objected that full employment and the BPSW will generate cost-push inflation by placing pressure on wages and thus costs and prices. In the next section we will examine the second part of the proposal: exogenous wage setting by the government.


The size of the deficit spending necessitated by the PSE intervention will be "market determined" by the gap between actual and full employment aggregate demand. However, the price paid by the government for PSE is exogenously set--for the purposes of our exposition, at $12,500 per year per worker. Thus, while the quantity "floats", the price is fixed. What are the implications for prices and wages?

Clearly, with a fixed price, the government's BPSW is perfectly stable and sets a benchmark price for labor. Some jobs might still pay a wage below the BPSW if they are particularly desirable (for example, because the work is pleasurable, or where large wage increases are possible for a lucky few--as in sports or the arts). However, low wage jobs which pay at or below the BPSW before the PSE is implemented will experience a one-time increase of wages (or will disappear altogether). Employers will then be forced to cover these higher costs through a combination of higher product prices, greater labor productivity, and lower realized profits. Thus, some product prices should also experience a one-time jump as the PSE program is implemented. If the BPSW is set at the statutory minimum wage, and if this minimum wage had universal coverage before PSE, then low wage private sector jobs will experience only minimal impacts--private wages need rise only sufficiently to make private sector employment preferable to PSE. In short, at the low end of the wage scale, implementation of PSE might cause wages and the prices of products produced by these workers to experience a one-time increase. This one-time jump--no matter how large it is--however, is not inflation nor can it be accelerating inflation as these terms are normally defined by economists.

Still, it can be argued that other wages are likely to also rise because by achieving full employment of labor, the threat of unemployment is removed, emboldening workers to demand higher wages--this is essentially the old "reserve army of the unemployed" argument. Workers who might have previously earned $13,000 per year now demand $13,500, knowing that in the worst case, they might be fired if they are too obstinate--leading to a PSE job at a loss of only $500. By extrapolation, all workers might harden positions, causing wages to jump upward. Prices would move upward to the extent that higher labor productivity and lower profits could not absorb the entire increase of wages. However, again, this is a one-time jump that is not defined as inflation, unless it can be argued that all workers above the $12,500 threshold continuously raise wage demands over time (generating a wage-price spiral). This makes little sense. The marginal $13,000 a year worker who decides to demand $13,500 per year on the calculation that this is worth the risk of losing her job and $500 per year pay (to take the PSE job) will not face the same decision once she is a $13,500 per year worker demanding $14,000--for now the loss is $1000 per year in the worst case. It is hard to see how the guaranteed $12,500 per year job will cause any individual worker to continually increase her wage demand through time, because as she gets further from the $12,500 benchmark, her potential loss due to obstinacy rises.

Of course, it is possible that the aforementioned $13,500 per year worker might calculate that if her wage demands are not met, she will fall back to a $13,000 per year job rather than the PSE--displacing some $13,000 per year worker to the PSE--in which case the expected loss is again only $500 per year. In this way, it might be supposed that continuous wage pressure is applied as workers move up the wage ladder, expecting to fall back only one rung rather than all the way to the BPSW. However, if we can assume that wages and jobs can be loosely sorted by labor productivity, then this is not likely. Essentially, the government's BPSW determines the wage for the lowest productivity group--the pool of unskilled and semi-skilled workers during periods of normal demand. Those workers whose productivity is substantially above $12,500 per year will find jobs in the private sector; those with lower productivity will find PSE. When private demand is below normal, the government will find the average productivity of its PSE pool rising as workers are laid off in the private sector; when private demand is above normal, workers whose productivity was formerly too low to induce private hiring will leave the PSE pool, lowering average productivity of this pool. At normal levels of private demand, then, workers in the private sector have a productivity that is above that warranted by a salary of $12,500.

Given that the relation between wages and productivity is loose, some ratcheting upward of individual wages after the PSE policy is adopted is possible. However, just as workers have the alternative of PSE, so do employers have the opportunity of hiring from the PSE pool. This is the primary "price stabilization" feature of the PSE program. If the wage demands of workers in the private sector exceed by too great a margin the employer's calculations of their productivity, the alternative is to obtain PSE workers at a mark-up over the BPSW. This will help to offset the wage pressures caused by elimination of the fear of unemployment. The PSE pool will operate as a "buffer stock", and just as a buffer stock of any commodity can be used to stabilize its price, the government's labor "buffer stock" will help to stabilize the price of non-PSE labor to the extent that workers in the PSE pool are substitutes for non-PSE labor.

It must be remembered that the PSE workers are not "lost" as a reserve army of potential employees; rather, they can always be obtained at a mark-up over $12,500 per year. In the absence of PSE, these workers can be obtained at a mark-up over the value of the package of social spending and private income obtained when unemployed (unemployment compensation, food stamps, under-the-table work, handouts, etc); this mark-up, however, is likely to be higher than the markup over $12,500 since it must be sufficient to make employment preferable over idleness. Further, recent work has tended to place a high rate of "depreciation" on idle human capital; the productivity of workers falls quickly when they are unemployed, and beyond some point, they probably become unemployable (due, for example, to loss of the "work habit"). With a PSE policy, however, those who are not employed in the private sector continue to work, thus, will not depreciate so quickly. Indeed, social policy could actually be geared toward enhancing human capital of the PSE pool. This would reduce the productivity-adjusted cost of hiring PSE workers relative to unemployed workers, and thereby diminish inflationary pressures.

Indeed, it is hard to imagine that true full employment with a PSE program would be more inflationary than the current system. The current system relies on unemployed labor and excess capacity to try to dampen wage and price increases; however, it pays unemployed labor for not working, and allows that labor to depreciate and in some cases to develop behaviors that act as barriers to private sector employment. Social spending on the unemployed prevents aggregate demand from falling excessively, but little is done to promote aggregate supply (or, growth of potential output). With PSE in place, however, labor is paid for working, which can lead to production of publicly supplied goods and services, can promote efficiency of the private sector (if, for example, PSE generates productivity-enhancing public infrastructure) and reduce private sector costs (for example, by reducing crime), and can increase the education and skills of PSE workers (compared with education and skill levels of the unemployed). Thus, PSE might increase aggregate supply (or potential output) and thereby place downward pressure on prices, rather than causing inflation.

The buffer stock aspects of PSE generate "loose" labor markets even as they ensure full employment. This stands in stark contrast with "Keynesian" demand management policies that were designed to "prime the pump" to increase private demand sufficiently to lower unemployment to the "full employment" level. The danger was that this would lead to such "tight" labor markets that inflation would be generated long before full employment could be reached. Indeed, most economists today believe that Keynesian policy proved to be a "failure" precisely because the tight labor markets did generate unacceptable levels of inflation. PSE is not subject to the same critique, for it allows loose labor markets even at full employment. If the PSE pool shrinks too much in an expansion so that it cannot act as a buffer stock, the government can either raise taxes or reduce non-PSE spending to replenish the buffer stock. Thus, aggregate "fine tuning" would operate through increases or decreases of the buffer stock, rather than by causing unemployment.

We make no claim that this PSE policy will stabilize the overall price level, thus, it is not a close substitute for an "incomes policy" or more formal wage and price controls. However, the BPSW will serve as a benchmark for a more-or-less homogenous "standard" labor input. So long as the government keeps the BPSW at $12,500, employers can always obtain workers from this pool at that price. As discussed above, this is the private sector alternative to hiring workers of greater skill at "market determined" wages. When the "market determined" wage rises to a level that exceeds the productivity-adjusted value of labor employed, there is an incentive to substitute workers from the PSE pool, which serves as the alternate "reserve army" to help dampen wage demands.

From time-to-time, there will be pressure for an upward revision of the BPSW. As the overall price level will not be held constant, and as there are substantial forces in modern capitalist economies that generate trend increases of the price level, the "real" (inflation-adjusted) BPSW will fall over time--generating a need for an adjustment. In addition, there will be obvious pressures by labor to raise the BPSW--just as there are pressures currently to increase the minimum wage. When the government raises the BPSW, this in effect devalues the currency. For example, an increase of the hourly wage from $6.25 per hour to $7.50 per hour reflects a 20% devaluation of the currency. Rather than "causing inflation", the devaluation will merely take account of inflation that results from factors that have little to do with the PSE policy.

In conclusion, the PSE will achieve what most economists would call zero unemployment (well beyond what they would call full employment) without inflationary pressures. We believe that the PSE policy would probably result in greater price stability than is currently the case--but that is not a primary claim of our argument. We need only show that truly full employment can be achieved without generating more inflation pressures than exist under the current system.


The main issues examined here concern the desirability and feasibility of a PSE program. The PSE program is desired because a) a more-or-less free market system does not, and perhaps cannot, continuously generate true full employment; b) no civilized, and wealthy, society can allow a portion of its population to go without adequate food, clothing and shelter; and c) our society places a high value on work as the means through which most individuals should obtain a livelihood. PSE policy cannot resolve all social problems; it cannot even replace all transfer spending. Some individuals will not be able to work in even a PSE program. Some individuals will not be willing to work. However, PSE will ensure that all of those willing and able to work at the BPSW will be able to obtain a job by selling their time to the government at the BPSW. Indeed, "ableness" should be defined as broadly as possible to include virtually all those who are willing to work. There is no reason to impose a narrow "efficiency" standard to ensure "productivity" above the BPSW. Any production will normally be better than no production; if one begins with the belief that even the unproductive must be supported, then government will have to provide income whether or not one works. Generally, it will be better to have someone working. In many cases, the "net product" may well be negative from a narrow economic standpoint because supervision, capital investments, and personal services required to put some people to work (for example, to employ severely disabled) could greatly exceed the economic value of output. However, a rich society can afford inefficiencies, and the noneconomic benefits of work can offset at least some of the economic costs.

PSE intervention is feasible. The modern government does not face the narrow "financial constraints" under which households and firms must operate. As Presidents Reagan and Bush showed, large government deficits do not have the supposed inflationary and crowding-out effects about which orthodox economists have long warned. In any case, however, we have argued that the increase of the deficit associated with operation of a PSE program will be relatively small--and PSE could even "pay for itself" in society-wide savings. While PSE may have associated administrative, capital, child care, health care, monitoring, and transportation costs, some of these can be reduced by using PSE workers (for example, to provide child care for other PSE workers) and "piggy-backing" on existing programs (using local governmental and nongovernmental agencies to administer the program, for example).

Once the primary issues have been resolved, there remain many issues, problems, objections, and extensions that must be analyzed. We shall merely list a few objections that immediately come to mind, and will provide a sentence or two to indicate the direction that might be taken to resolve the problems.

1. It will be impossible to administer the program due to incompetence, corruption, racism, and opposition. Clearly, this is a significant problem; as Minsky used to wonder, are there administrators today as capable as those who administered the New Deal? We can suggest several methods to ease administrative problems. First, the existing unemployment benefits program administration might be used to administer a PSE program. Alternatively, administration could "devolve" to the state and local government level and to not-for-profits. The Federal government would simply provide as much funding as necessary to let every state and local government hire as many new employees as they desired, with only two constraints: these jobs could not replace current employment, and they could pay only the BPSW. Finally, a similar offer could be made to qualifying non-governmental non-profit organizations, such as Americorps, VISTA, the Student Community Service Program, the National Senior Service Corps, the Peace Corps, the National Health Service Corps, school districts, and Meals on Wheels.

2. PSE employment will consist of nothing but "make-work" jobs, like the WPA before it. As we move farther from the 1930s, people seem to have forgotten the contributions made by Works Progress Administration (of the New Deal). WPA workers
not only built or reconstructed 617,000 miles of roads, 124,000 bridges and viaducts, and 120,000 public buildings; they also left the nation with thousands of new parks, playgrounds, and athletic fields. Moreover, they drained malarial swamps, exterminated rats in slums, organized nursery schools, and taught illiterate adults to read and write. Unemployed actors set up theaters throughout the land, often performing in remote towns and backwoods areas. WPA orchestras gave 6,000 live concerts. WPA artists produced murals, sculptures, and paintings that still adorn our public buildings. (Ginsburg 1983, p. 11)
We do not believe it requires much imagination to come up with a list of useful tasks for PSE workers. Even in the worst case, PSE workers must at least "sell" their time in exchange for dollars, which many Americans might find preferable to "money for nothing". Possible PSE jobs include:
  • Companion for the elderly, bed-ridden, orphans, mentally and physically disabled
  • Public school classroom assistant
  • Safety monitor for public school grounds, areas surrounding schools, playgrounds, subway stations, street intersections, and shopping centers
  • Neighborhood cleanup/Hiway cleanup engineers
  • Low income housing restoration engineers
  • Day care assistants for children of PSE workers
  • Environmental safety monitors
  • PSE artist or musician
  • Community or cultural historian
Obviously, this list is not meant to be definitive, but is only to suggest that there are many jobs that could be done by PSE workers. We have not listed the more "obvious" jobs, such as restoration of public infrastructure (patching holes in city streets, repairing dangerous bridges), provision of new infrastructure (hiway construction, new sewage treatment plants), and expansion of public services (new recycling programs) that should be carefully considered because they might reduce private costs and increase private profitability. These are types of social spending that should be done even without an PSE program, and that might be better accomplished by non-PSE (including unionized) workers. However, it should be noted that WPA employees did indeed engage in this sort of work.

3. What can be done with belligerent/anti-social/lazy PSE workers? PSE will require that one show-up for work more-or-less on time; beyond that, requirements would have to be made almost on a case-by-case arrangement. Discipline would be maintained primarily by the promise of promotion to more desirable PSE jobs, and, eventually, to private sector employment. In the worst case, some workers might be so irresponsible that their employment would be day-by-day, or even hour-by-hour with a cash payment for a specified amount of time spent on the job. PSE workers could be fired from their jobs for just cause; there could be conditions placed on re-hiring (for example, the fired worker might have to wait for 3 days--without pay--before re-hiring; the penalty could be increased for subsequent firings). In extreme cases, some individuals may not be allowed to work in a PSE job; PSE cannot provide income for all the needy.

4. What effect will PSE have on unions? On one hand, PSE removes the fear or threat of unemployment, which is often said to be an important disciplinary method used by firms against workers. It also establishes a true, universal minimum wage--below which wages will not fall. It still permits unions to negotiate benefits with employers--such as unemployment compensation (so that although there might not be any federal unemployment compensation, workers could still negotiate privately-supplied benefits). PSE could include a package of benefits, including health care. This would then set the lowest standard (and could, for example, lead to universal health care). On the other hand, the PSE pool will also dampen wage (and benefit) demands of non-PSE workers as employers will have the alternative of hiring from the PSE pool. Thus, it is not clear that PSE is biased in favor of workers or employers.

5. Won't participation in PSE lead to stigmatization? If PSE takes only those workers the private sector "doesn't want", won't participation in PSE be seen as a negative indication of character, education, or skill level, much as participation in "welfare" stigmatizes a person? This danger can be reduced through creative action. For example, PSE can be promoted as a universal "Americorps" service, open to all who would like to perform community service (unlike the current Americorp program, which limits the number of participants). We could institute a national service requirement, much as many countries require military service or national service. Alternatively, we can rely on persuasion: universities could favor applications from prospective students who have served for a year in an PSE position; not only would this provide students with savings for tuition, but it would also enable them to gain skills, training, and maturity before beginning college. Alternatively, colleges could offer "junior year programs" in PSE as an alternative to "junior year abroad" programs. Corporations could allow leaves of absence to professionals and executives to work in the PSE program. Retired executives, professionals, and politicians could serve in the PSE program (much as they now serve with President Carter in Habitat for Humanity). PSE might even provide for some part-time positions (perhaps even unpaid) for volunteers who would like to perform community service without giving up other employment. It is possible that PSE service could come to be seen as an advantage on the resume, rather than as a stigma.

6. What if the Fed or financial markets react negatively? Implementation of a PSE program might cause a reaction by financial markets because they come to expect that the deficit will crowd-out investment and cause inflation, or, more likely, because they expect the Fed to react by raising interest rates. Note that if the Fed did raise interest rates and if this slowed the private sector, this would only increase PSE employment. In other words, the Fed would no longer be able to fight fiscal policy by causing unemployment, but would only be able to reduce private sector employment and raise public sector employment. In response, the appropriate fiscal policy would be to increase non-PSE spending or to reduce taxes. While it would be far preferable to coordinate monetary and fiscal policy, at least with PSE in place, the Fed could not raise unemployment. It would be hoped that the private sector would place pressure on the Fed to relax policy because it would be obvious that the tight monetary policy only hurts the private sector and increases the size of government.

7. Why worry now, when unemployment is lower than it has been for a generation? Many pundits have proclaimed that we have entered a "new age" with the "new economy"; it is claimed that things "have never been better". If true, this means that the best that can be expected is a situation in which six and a half million are unemployed and millions more work fewer hours than desired or are forced to patch together several jobs. It also means that "welfare-to-work" programs are doomed to fail because the best that can be done is to redistribute jobs, still leaving millions unemployed. Finally, it means that price stability can only be obtained at the cost of millions of unemployed. Now, more than ever, it should be clear that "free markets" cannot be relied upon. If our society values work, full employment, and stable prices, then PSE is preferable to the current arrangement.


Cavanaugh, Francis X., The Truth about the National Debt: Five Myths and One Reality, Boston: Harvard Business School Press, 1996.

Epstein, Gene, "A Free-Market Guru's Immodest Proposal: Billions in Tax Breaks to Help the Poor", Barron's, October 27, 1997, p. 65.

Ginsburg, Helen, Full Employment and Public Policy: The United States and Sweden, Lexington Books, 1983.

Gordon, Wendell, "Job Assurance--the Job Guarantee Revisited", Journal of Economic Issues, vol 31, September 1997, pp. 826-834.

Harvey, Philip, Securing the Right to Employment, Princeton: Princeton Univ. Press, 1989.

Heilbroner, Robert, and Bernstein, Peter, The Debt and the Deficit, WW Norton, 1989.

Keynes, John Maynard. The Collected Writings of John Maynard Keynes, Vol. IX, Essays in Persuasion, edited by Donald Moggridge, London and Basingstoke, Macmillan, 1972.

Lerner, Abba P. "Functional Finance and the Federal Debt", Social Research, vol. 10, 1943, pp. 38-51.

Mitchell, William F. "Unemployment and inflation: a Demand side focus", paper presented on the PKT seminar, January 1997.

----- and Martin J. Watts, "The path to full employment", manuscript, University of Newcastle, 1997.

Mosler, Warren, Soft Currency Economics, third edition, 1995.

Minsky, Hyman P., Stabilizing an Unstable Economy, Yale University Press, 1986.

Phelps, Edmund S., Rewarding Work, Harvard University Press 1997.

Wray, L. Randall. Understanding Modern Money: The Key to Full Employment and Price Stability, Edward Elgar: Cheltenham, 1998.

-----. "Government As Employer of Last Resort", Levy Institute Working Paper 213, 1997.


Request a copy of this publication