Where Have the Jobs Gone?

1.

Not so many years ago, “full employment” was regarded as a realistic goal. If certain policies were pursued, all who so desired could have steady work at decent pay. Needless to say, there were disagreements on how to achieve that aim. Conservatives asked that the economy be freed to flourish on its own; given such a climate, private businesses and investors would create sufficient jobs. Liberals and those further to the left were less sanguine on this score. They believed that the government must intervene in the economy: not merely when times were bad, but to plan for full employment as a goal for the future.

The talk we hear today has a less assured ring. We hope to put some people back to work; but no one has clear ideas about what jobs, or how many, can be created for them. Indeed, there is a widespread suspicion that our economy will not expand much beyond its present bounds, and that the parts that will grow will be the ones that depend more on high technology than on labor. If this apprehension is correct, a high level of unemployment may be with us for the rest of this century. One result will be keener competition for such jobs as exist. Who gets them will be a central issue, raising questions of equity and need.

Two recent books consider what might be called the distribution of employment. Victor R. Fuchs, an economist at Stanford University, describes his How We Live as “an economic perspective on Americans from birth to death,” relying largely on statistical studies from governmental sources. He is especially interested in the entry of women into the labor force and their experience there. Melvin R. Levin’s Ending Unemployment is chiefly concerned with the “marginal” men and women who have never gained a foothold in the labor market. However he concludes that they and laid-off workers will have more and more in common if we cannot bring about an industrial revival. Hence his proposal for a permanent program of public works, to include workers of varying experience and skills.

Before turning to these books, however, we might well review the basic facts of employment and its lack. Our most comprehensive source is an annual study issued by the Census Bureau, entitled Money Income of Households, Families, and Persons in the United States. Its most recent edition, released in March of this year, tells us which Americans were employed, how much work they did, and what their earnings were in 1981. While it is true that unemployment averaged 7.5 percent in 1981, as against 10.1 percent for April of 1983, the distributions of earnings and employment are substantially the same. The Census findings come from a survey of 60,000 households, whose members were asked to specify the sources of their income, which the bureau then adjusted to accord with data from other agencies. This sampling is the most reliable we have; in fact it is the same one on which the monthly unemployment rates are based.

During 1981, 117.2 million persons held paying jobs at some point during the year. This is a fairly impressive figure. If we take the ages between eighteen and sixty-five as the approximate employment span, it would appear that almost 85 percent of us had jobs at one time or another. However, appearances can be deceiving. In effect, we have a two-tier employment force as Table 1 shows.

TABLE 1

The lower segment largely consists of the 52 million persons who either held part-time jobs or worked only part of the year. Their wages averaged $5,795, which suggests that in most cases they were largely supplemental: women comprised a majority of the group. Most of the unemployed also fell in this stratum. Their number averaged 8.3 million in 1981, but the total of those who were out of work during particular months was actually higher. Thus there were full-time breadwinners who lost their jobs during the year and settled later for part-time work. In most cases, then, the second tier consists of people who fill in when extra hands are needed, taking available work at modest pay.

This leaves the 65.1 million jobholders who worked steadily throughout the year, and who averaged $18,941 in earnings. Of this group, 64.2 percent were men, who made $22,196 as against $13,112 for women. All in all, the men did rather well: 3.6 million earned between $35,000 and $50,000, while another 1.97 million made over $50,000. This top tier displays the best the country can do in the way of employment. Indeed, the 65.1 million regular jobs command 80.3 percent of all wage, salary, and self-employment income earned by Americans in 1981, leaving only 19.7 percent for the 52 million jobholders in the second tier.

While most Americans still live in families, a large and growing group consists of people the Census classifies as “unrelated individuals.” In 1981, it found 17.6 million employed persons who either lived alone or with someone not related to them. By and large these people had fairly comfortable incomes. The 9.6 million men earned an average of $16,916, while the 8 million women made $12,441. Moreover, 59.6 percent of this group had full-time year-round jobs, where the men’s average income was $21,085 and the women’s $15,553. If two such persons decide to share quarters, as an increasing number do, a typical combined income came to $36,638.

Most of these “unrelated” earners are young and single, or if they are divorced they do not have children living with them. (Divorced parents with resident children are assigned to the “family” columns.) Most single people who want to work have little trouble finding jobs at a living wage. Moreover, if they lose one position—and they have a high rate of doing so—they can more easily take another at a lower salary. The need for steady employment looms largest for families, especially those with children.

A “family,” according to the Census, consists of any two or more related people who currently share quarters. In 1981, the bureau counted 60.3 million families, of which 52.5 million had at least one employed member. (Most of the other 7.8 million were older couples receiving Social Security benefits, or women raising children on the welfare rolls.) Table 1 also shows the distribution of family earnings in two different ways: first, the number of households in various income brackets; and then the proportion of all earnings of each income range. Thus the top 27 percent of families with incomes over $35,000 accounted for exactly 50 percent of aggregate family earnings. And the 23.7 percent making under $15,000 received only 6.8 percent of the payroll total. Those who call for more equality in income distribution can draw on strong evidence.

Insofar as some families fare better than others, the determinant is less the major breadwinner’s wages than how many earners the household contains. Of the 51.8 million families analyzed, only 18.5 million had a single earner. Among the others, 24.9 million had two; and another 8.4 million had three or more. Unfortunately, the Census does not tell us how much each member brought in; some may have been teenagers with a paper route. Still, the average income for those families with multiple employment came to $31,919, a substantial 49.7 percent above the $21,320 for households with a sole jobholder.

It is not sheer need that sends additional members to work so much as a desire for a higher living standard. Among the families in the $35,000 to $50,000 range, 83.1 percent had more than one earner, as did 83.6 percent of those in the $50,000 to $75,000 bracket. However, at incomes of $75,000 and above, the proportion with multiple employment dipped to 72.4 percent. Once that level is attained, wives may feel less impelled to work; or their husbands may prefer that they do not. Many working-class husbands also prefer that women stay at home, a sentiment which continues even when the children are grown. Thus among women aged fifty-five to sixty-four, only 25.5 percent had year-round full-time jobs.

The Census has an overlapping analysis of earnings for the nation’s 40.9 million married couples where the husband was employed. In 26.2 million of these marriages, or 64 percent, the wife also worked; while in the other 14.7 million she remained at home. Where the husband was the sole provider, his earnings averaged $22,095. In marriages where both spouses worked, the husbands’ average earnings came to $20,050, while the typical wife made $8,586. This made a total of $28,636 for such a household, to which the wife contributed 30 percent. Clearly these women did not embark on employment to make up the $2,045 difference between their husbands’ wages and those of men with nonworking wives. They took jobs to raise their households even higher, to the $28,636 level.

Still, the wives’ earnings just cited suggest mainly partial employment. Indeed, in only 9.5 million of the two-job marriages—36.2 percent of the total—did both partners have full-time year-round work. However, the women in those marriages made 55.9 percent of the earnings total for all wives. In these favored marriages the husband averaged $21,539 and the wife $13,250, totaling $34,789, of which her contribution rose to 38.1 percent. Moreover, the majority of these households are ones where the couple have no children or where all of the children are grown.

What the state of the economy seems to suggest is that Americans can earn a decent living so long as they remain unmarried or put off having children; and a family can do so if it can add a second member to the employment force. Members of a household also have an even chance of a decent living with a single earner who happens to be a man. Table 2 gives the breakdown by domestic composition of the 61 million families in 1981. (In this analysis, the Census recorded family income from all sources.)

TABLE 2

If we subtract the marriages where the husband is sixty-five or over, there remain 14.9 million childless couples whose average income is now $33,298, which is discernibly higher than the $28,734 for their counterparts with children.^*

For women raising children without husbands or other men to help, average family income was well under half—40.5 percent—of that for couples who had children; and 84.7 percent of those women had incomes under $20,000. This was so even though a 1981 Bureau of Labor Statistics study found that 55.6 percent of single mothers were engaged in some kind of employment, or at least were at the time the survey was conducted. The report noted that 70.7 percent of divorced mothers worked, compared with 51.8 percent of those who were separated and 36.2 percent of mothers who were never married. The nonworking 44.4 percent were almost all on welfare, since the numbers supported solely by alimony and child support are statistically negligible. Still, if all single mothers were made to work, only a small fraction would find their economic status visibly improved.

2.

While everyone realizes that more women are working, it remains to ask how this affects the total employment picture. Between 1960 and 1981, out of every 1,000 women, the proportion working or seeking work rose from 337 to 514. Looked at another way, during this period women rose from being 35.3 percent of all jobholders to 44.3 percent. And among year-round full-time workers, the women’s ratio went from 25 percent to 35.8 percent. In How We Live, Victor Fuchs offers an economist’s explanation for why women have gone to work in increasing numbers. Single women have usually held jobs, while older wives tend not to, so “most of the increase in women working has occurred among married women with husbands and children present in the home.”

Fuchs attempts to explain this shift as a matter of “opportunity costs.” Wives have not gone out to work because extra earnings are “needed to make ends meet,” although that is the reason they generally give. Nor is it because they have fewer children and more labor-saving devices. Rather, he argues, “the probability of a wife’s working is positively related to her potential wage rate.” Putting aside the variety of other reasons women may give for working, he assumes that women simply compute the cost of remaining at home. Beginning with the 1950s and the explosion in office and service work, the wages women could command began to “cross a certain threshold level,” leading them to conclude that domestic duties were an uneconomic use of their time.

This analysis, though narrow, makes a kind of sense, although Fuchs also says it explains why “the higher the husband’s income, the less likely it is that the wife will work for pay:” As we saw earlier, the number of two-earner families begins to decrease at the $75,000-income level. Still, there are more compelling reasons why the wives of many high-income husbands are likely to stay at home. One is that their high-powered spouses lead hectic lives and expect a full-time homemaker to keep the nest in order. In addition, the wife of a successful man may feel she cannot take just any job, but only one commensurate with her station. Unfortunately few women in that bracket have the kind of employment record that would qualify them for high-status positions. In fact, many may want to work, to broaden their lives, and to have a checkbook of their own. But they are caught in a velvet trap.

Despite their rise in numbers, women are still marking time when it comes to earnings. In 1960, fully employed women made $606 for every $1,000 made by men. By 1981, their share had actually declined to $591. Fuchs attributes the sexual discrepancy to the lack of a “continuous commitment to work” on the part of most women. While it is true they cluster in jobs that traditionally get lower pay, How We Live suggests that women would move ahead if they took work more seriously. For example, they have a higher dropout rate. A recent Bureau of Labor Statistics study found that while the average man had been employed at his current place for four years, the typical woman’s tenure was two and a half. Altogether, 28.7 percent of the men had held their jobs for at least a decade, as against 17 percent of the women. A further matter worth recording is that men seem willing to make an extra effort to get a better-paying job. One index of this is that the typical man travels 10.5 miles to work, whereas the average woman’s trip is 7.5 miles.

The main way women depreciate themselves in the market is by becoming mothers and allowing that condition to dominate their lives. They remove themselves from work for several years or take part-time jobs that let them hurry home. While we hear a lot about working mothers, in fact among married women with youngsters under six, only 15 percent work the entire year in fulltime jobs. To the degree, then, that “women feel more responsible for children after birth, they will be disadvantaged relative to men,” Fuchs argues. So long as women remain fearful of being—or accused of being—uncaring mothers, their earnings will be less than men’s. While How We Live acknowledges that in theory fathers could pitch in more, its author sees “little evidence that men are prepared to take up the slack.” They do not consider the experience of European countries where the provisions for long-term maternity leave and subsidized child care are much more generous than in the US.

That women come closer to parity with men if they stick to continuous careers is evident in the fact that among fully employed single people (“never married” in the Census phrasing) women make $815 for each $1,000 made by men. Even among teen-agers, admittedly not heavy earners, girls receive $918 for each $1,000 going to boys. If those young women continue to work steadily, they may retain that ratio or even improve upon it.

Indeed there are signs of a move away from motherhood perhaps for these very reasons. Each year the Census conducts a special “fertility survey,” to ascertain how many children women have already had, plus the number they “expect” to have in the future. In June 1981, the most recent report, 20.5 percent of the single women interviewed said that they “expected” they would never have a baby. Of course these forecasts may express flights of fantasy as well as considered judgments. Still, they offer some insights into attitudes toward childbearing. And among women between the ages of twenty-five and twenty-nine, no fewer than 37.2 percent were still childless; and of those between thirty and thirty-four, a surprising 20.6 percent had yet to have a child. Some of these women may have babies in their late thirties, although those who do tend to receive more attention than their numbers warrant. Moreover, late mothers are more likely to stop at one child. They realize—as men always have—that serious employment does not allow distractions.

How We Live notes that “service sector employment…has provided nearly all of the additional job opportunities in the US economy since the end of World War II.” Given the willingness of employers to place women in these jobs, the consequence will be fewer openings for men. Table 3 shows several fields that were dominantly male a dozen years ago and where women are now being hired in increasing numbers.

TABLE 3

It seems plain from Table 3 that women are displacing men in many fields, not least because men can no longer claim that only they can handle the demands of the position. In the economy as a whole, the full-time year-round work force has fallen from being 75 percent male in 1960 to 64.2 percent in 1981. And whereas 66.6 percent of all adult men were working in 1960, by 1981 that figure was down to 57.6 percent. At this point it is unclear how far one can contend that men cannot find jobs because those they might have held have been filled by women. But if there are such cases, they are a price that must be paid for gender-blind hiring.

As it happens, much male unemployment is being concealed in earlier retirements. Back in 1961—the first year that workers had such an option—only 4.7 percent of men began receiving Social Security before they reached age sixty-five. By 1981, no fewer than 57.1 percent of them began pensions before sixty-five. In this period, also, it became easier to receive disability benefits, another form of early pension. Between 1960 and 1981, the number of men with this source of support rose from 356 thousand to 1.9 million. A third explanation is that in this two-decade period, a substantial number of men dropped out of the labor force into the underclass. During 1982 the nation’s state and federal prisons averaged 412,303 inmates in residence, over 95 percent of them men. Incarceration is yet another way to reduce the official count of unemployed men.

Finally, because “equal opportunity” has been available for women in more fields of employment, more young men are being rebuffed when they try to get the jobs they want. An already noticeable number can be seen taking positions lower down on the pole, such as flight attendants and telephone operators.

3.

Unemployment that causes hardship—and most obviously does—deserves to be deplored. However, as matters currently, stand, we have not got much beyond deploring. No one has any clear ideas for creating several million jobs, including some, for example, for forty-eight-year-old steel workers whose plants have permanently closed. In fact, the problem involves the economy as a whole, so it would be well to examine who is unemployed and why. Full figures are now available for 1982, in the latest Bureau of Labor Statistics report Employment and Earnings. The official unemployment rate averaged 9.7 percent in 1982, rising from 8.6 percent in January to 10.8 percent at the end of the year, the highest since 1940. In an average month, 10,678,000 people were recorded as unemployed, and this figure forms the base for the various subtexts in the bureau’s analysis.

First, among the 10.7 million unemployed, 1.7 million—or 15.7 percent—had either lost or only wanted part-time work. It could be argued that these people were not fully unemployed and that counting them equally with the others inflates the final rate. However, not listed with the unemployed were 6.2 million men and women who held part-time jobs but said that they “usually work full-time” and would prefer such work. Insofar as they need full incomes, and most of them probably do, they should be counted at least as partially out of work. The government’s announcements about “total unemployment” are therefore seriously misleading.

In addition, the Census—which conducts the survey—located 1.6 million “discouraged” people who wanted work but had given up looking because jobs seemed unavailable. They, too, were not included among the 10.7 million unemployed. Most commentators talk as if this “discouraged” group consists mainly of middle-aged men and inner-city youths. In fact, women accounted for 62.6 percent of those who professed to being discouraged, and a considerable number had only hoped for part-time jobs. All the same, among those counted as unemployed there were 395,000 men aged fifty-five or over who want full-time work. It seems likely that many of them will never work again; and since sixty-two is the earliest age they can start on Social Security, their next several years will be rocky ones indeed.

Much is made of the unemployment rate for black teen-agers, which has run as high as 53 percent in recent months and averaged 48 percent in 1982. Still, it would be well to view that figure in perspective. Altogether, there are 2.25 million young black people between sixteen and nineteen, of whom 63.4 percent are still in school or heading families, and are not looking for employment. So the “labor force” statistics we are given concern only 824,000 black youths, of whom 428,000 had jobs and 396,000 said they wanted work but could not find any. The latter figure is large enough to cause concern—it comprises 17.6 percent of all black teen-agers. It would be interesting to know more about the 428,000 who found work. To what extent did they benefit from favorable local conditions? What personal factors were involved?

The Census analyzes the reasons for unemployment, breaking them into four categories. First, there are “job losers”: laid-off workers who hope to be called back, plus others whose jobs came to a permanent end because a business closed down or work was reorganized. This group comprised 58.7 percent of the unemployed. In addition, there were the 7.9 percent who left their last job voluntarily; plus 11.1 percent who were frustrated “new entrants,” mainly high school and college graduates; and then 22.3 percent called “re-entrants,” mainly married women, many of whom would like part-time work. In other words, a third of the “unemployed” are people who have not been working recently and cannot be said to have “lost” jobs.

A further breakdown is equally revealing. Table 4 gives the sex and domestic status of the 10.7 million unemployed.

TABLE 4

It should be added that of the 2.5 million unemployed husbands, 49.3 percent had wives who were working. And 75.7 percent of the 1.8 million jobless wives had husbands who were still employed. In all, then, 23.2 percent of the unemployed were married men with children or single heads of households of either sex and this is the group whose situation should cause the greatest concern. But the largest single contingent, more than half the total, were unmarried people without dependents. One reason we have a high unemployment rate, therefore, is the postponement of marriage. Only a few decades ago many more young women were married and at home with children, and would not have been counted as unemployed. During the 1958 recession, for example, women numbered 35.2 percent of the unemployed as against 42.1 percent last year. In addition, married men are more likely to look for steady jobs. Their 1982 unemployment rate was 6.5 percent, as against 17.7 percent for men who were still single.

Moreover, in the past fewer men left their families so there were fewer single mothers to enter the labor force. Unmarried people obviously need paychecks, and most wives who want to work mean it seriously. At the same time, changes in domestic arrangements have clearly put more people in the competition for limited employment.

Too many people competing for too few jobs was the classical conception of unemployment. Its corollary was that lines outside the hiring window would beat wages down. This is happening in hard-pressed industries, where wages are being reduced, and where demands that workers “give back” certain benefits are acceded to, and long-term prospects for employment are often being diminished by new technology. But the classical explanation no longer applies to many other industries: a greater proportion of the available jobs are now not only more highly paid but are also insulated from market conditions. This is especially the case with professional and administrative occupations.

During 1982, the unemployment rate for professionals averaged only 3.3 percent while for administrators the figure was 3.5 percent. Clerical occupations had a 7 percent rate while among blue-collar workers it was 14.2 percent. Factory unemployment totaled 17.7 percent, with construction laborers at 29.3 percent. If manufacturing and building workers are laid off when demand declines, there seems an unstated agreement that less exacting criteria will be applied to professionals and managers.

This rule applies to salaries as well. In 1981, a total of 30.5 million men and women filled professional and managerial positions, accounting for 26 percent of everyone who did some work during that year. However, their salaries added up to 40.9 percent of the year’s total payroll. Moreover, 73.1 percent of these people held their posts on a fulltime, year-round basis, as opposed to 49.4 percent of those in other occupations.

The most vivid example is among law firms which pay $40,000 and over for entering associates. Given the large supply of graduates, of whom many are highly qualified, it can hardly be argued that such salaries are needed to attract scarce talent. In other cases, a seniority system prevails, enforced as much by custom as by an explicit contract. At Harvard University, for example, compensation for full professors averaged $59,200 last year, of which $48,500 was salary and the rest fringe benefits. This sum is for about eight months’ work. It strains credulity to believe that all or even most of Harvard’s 529 full professors have such attraction on the market that this range of compensation is required lest they be lured away. However, this largely middle-aged group of 505 men and 24 women was allowed to absorb 71.9 percent of the university’s academic payroll, which meant that only 430 positions could be budgeted for faculty of lesser ranks.

A striking feature of recessions—indeed of depressions as well—is how many people hold on to well-paying jobs. Even if those receiving paychecks cut down somewhat on their spending, members of the middle class can still be seen crowding restaurants, going off on holidays, and purchasing the latest in appliances and clothes. Many pay five-figure tuition bills, move into condominiums, and have something remaining for IRAs and money-market accounts. Hence also the spectacle of parents secure in well-paid positions, while their children are having difficulty landing a first job. Such parents are often well-enough off to subsidize their underemployed offspring, more than a few of whom are moving back home. These in-family transfers may ease the economic strain, but they do nothing to create employment for the coming generation.

It is not really possible to give a figure for how many jobs we would need for optimal employment. For one thing, not everyone wants to work: witness the growing number who are accepting early retirement. Of course, what they are retiring from makes a great difference. Middle-level executives who are not going any higher may prefer golf to a job on the sidelines. On the other hand, college professors protest if they can’t continue into their seventies. Table 5 displays the potential dimensions of a full-employment picture.

TABLE 5

The “labor force” column includes all those who were working or looking for jobs in 1982: the employed, the unemployed, plus those with part-time positions who would have preferred full-time work. “Not in the labor force” lists people who said they did not wish to work or weren’t seeking a position, plus the reasons they gave.

Clearly, we would like jobs for the 10.7 million unemployed and the 1.6 million who are discouraged, along with full-time work for the 6.2 million who prefer it. As for current nonworkers, how many would enter employment under favorable conditions is anybody’s guess. Some of the 12.3 million who are retired might wish to go back to work. (Right now, only 3 million members of the labor force are people over sixty-five; and over half of those had part-time employment.) The same holds for a portion of the disabled if suitable jobs could be arranged. The key group are the 29.9 million who report they are needed at home, of whom 21.3 million are wives and 2.9 million are women heads of households. The outlook of the emerging generation suggests that many or even most would probably choose employment if they could have day-care facilities, congenial jobs, and willing husbands. Moreover, a growing number will want full-time positions, which are more expensive to create.

In 1981, the total sum available for hiring Americans and paying for their services added up to $1,549 billion—$1,439 billion for wages and salaries and $110 billion from self-employment earnings. As I indicated at the outset, most of this figure—80.3 percent—provided 65.1 million full-time year-round jobs at fairly decent rates of pay. This $1,549 billion sum, or its counterpart in current and future dollars, had best be regarded as a constant factor. The odds of increasing real national wealth, relative to population, are not very promising, and increasing automation will eliminate more and more industrial jobs. So if we wish to open employment up for more people, whatever incomes they receive will probably have to come from the existing pool of earnings. To cite a homely example, if Harvard’s 529 full professors agreed to reduce their current compensation of $59,200 to, say, $45,000, they could create 376 new assistant professorships each paying $20,000.

Ending Unemployment, its author tells us, “focuses on the condition and prospects of the marginal worker in America, on society’s chronic losers.” These are able-bodied men and women who work sporadically or not at all, upward of 15 million people many of whom are not officially counted as unemployed. Melvin Levin examines the impact on unemployment of the minimum wage, race and sex discrimination, job training programs, regional migrations, private-sector subsidies, reducing the welfare rolls, or doing nothing at all. Unfortunately, sweeping solutions are not really attainable under current conditions; and most programs that have been tried have had disappointing results. The public employment programs of the 1930s worked partly because the nation was desperate, and also because those taken on were already seasoned workers. All those schools, dams, and bridges were not built by people from the bottom of the barrel.

Even so, Levin proposes that we institutionalize “public-sector employment” on a permanent basis. “One of the anomalies of the economic system is the willingness, albeit reluctant acquiescence, to provide a modest subsistence for the jobless either through unemployment insurance or through public welfare.” What we have at present is “subsidized idleness.” Instead of keeping people on the dole, Levin would assign those who have lost or never had jobs to “salvage, replace, and maintain a deteriorating public infrastructure,” ranging from modernizing reservoirs to repairing local jails. Levin’s list is lengthy, including the rebuilding of dams, bridges, ports, sewers, rail lines, along with “the already deteriorating Interstate Highway System.” Even the chronic losers could be absorbed, because “painting; cleaning, and simple street, park, and building maintenance are skills that can be quickly acquired.” On the whole he would prefer that these projects be contracted out to private firms, so as to enlist the support of private business. “Marginal workers would report,” he asserts in a somewhat military tone; and he heads one of his sections “Everyone Works.”

It would be easy to pick holes in Ending Unemployment, and Levin acknowledges that his ideas may be visionary. But it seems never to have occurred to him that if the economy is to be more productive this will probably mean that sophisticated machines will eliminate more jobs. We may have to adapt our values to a future where subsidized leisure will be widespread. Still, such a change seems far off. Levin is appalled that so many million man-years of potential work now remain unused while so much cries out to be done. He calls his a “modest proposal,” although it isn’t that at all; and he says the costs “would not be massive,” which is equally untrue.

To the objection that welfare money will still have to go to mothers with children, Levin would probably reply that his program would reduce the number of fathers who desert their families. He would also argue that with men from the underclass working, the drop in crime would be a social saving. In many ways, the most radical part of his proposal is to abolish unemployment compensation, currently consisting of tax-free payments to people between jobs. In fact, a lot of Americans use at least some of these weeks as a period of relaxation, one they believe they have earned. To consign them to work, at pay less than they had been getting, would be protested by many as involuntary servitude. In fact, Levin would want to mingle all classes in his public-works projects, which is what happened in the Thirties, although it is hard to visualize the middle and upper classes taking part in them now. His is really a prescription for a deep depression rather than for ebb-and-flow recessions. Still, it is worth attention. We may need something like it sooner than we think.