Rewarding Work: Feasible Antipoverty Policy

Virtually all economists who have studied the changing income distribution have confirmed what nearly everyone else knows. For most Americans, living standards are stagnating and becoming more unstable. For the bottom half, income is falling. And the prime culprit is not shifts in family values or the work ethic, or even changes in taxes and social benefits. Most of the problem is the erosion of wage and salary income.

The causes are multiple [see Bluestone, "The Inequality Express," TAP, Winter 1995], but there are proven methods for cushioning the impact by raising incomes at the bottom and providing "wage insurance" for those at risk. Chief among these are the minimum wage and the earned income tax credit (EITC). The two fit together like jigsaw puzzle pieces. Each has an inherent weakness—but the weakness in one is precisely the strength of the other. The problem is that the real value of the minimum wage is shrinking and the EITC is under attack.


Between the end of World War II and 1973, inflation-adjusted average weekly wages for non-supervisory employees rose by 60 percent, while real median family income, boosted by the growth in female labor force participation, literally doubled. Those in the bottom quintile of the income distribution saw their incomes rise by no less than those in the top. Such a benign climate ended in 1973. Since then, average weekly wages have fallen by nearly 20 percent and median family income has not improved beyond the level achieved more than two decades ago.

According to Gary Burtless of the Brookings Institution, between 1973 and 1993 the average income available to the bottom quintile fell by nearly 23 percent—from $17,601 to $13,596 per year for a family of three (in 1993 dollars). A small portion of this decline was due to cutbacks in means-tested transfer programs such as welfare. A bit more was due to reductions in the value of private pensions. Yet the largest source of decline, amounting to 60 percent of the total income loss, was a reduction in the earnings of either the head of household or other household members.

As incomes of the bottom fifth of the population have fallen, poverty has naturally increased. As a nation, we made sizable inroads against poverty during the 1960s and early 1970s due to strong sustained growth in the economy, low unemployment and high labor bargaining power, increases in transfer programs, and the billions of dollars spent as part of the War on Poverty. Between 1960 and 1973, the proportion of individuals living below the official poverty line fell by half, from 22.2 percent to 11.1 percent. But then the war stopped. By 1993, the poverty rate was back up to 15.1 percent.

The newly poor, however, were not the same as the old poor. In 1970, 40 percent of the poor were children and 19 percent were elderly. Thanks to the expansion in Social Security and Medicare, only 10 percent of the poor today are age 65 or older. The proportion of the poor who are children has also declined slightly, to 38 percent, because of declining birth rates. As a result, nonelderly adults comprise an absolute majority (52 percent) of all poor persons in the nation. This may be the first time since the era of Charles Dickens that the majority of the poor are not the young, the old, or the infirm, but prime, working-age adults. While conservatives may target a breakdown in family values as poverty's chief culprit, the data suggest otherwise. The main cause is falling wages and diminished employment opportunity.

These trends will only intensify if "welfare reform" compels single mothers to seek paid employment. Thirty years ago, single-parent families headed by women made up less than one-fifth of all impoverished households in the country. Today they make up nearly one-half (47 percent) and therefore the welfare-to-work strategy, in either its more gentle White House guise or its harsher congressional form, will impact millions of the lowest-income families in the nation.

If we push welfare mothers into paid employment without raising wages and benefits, the main effect will be to increase the already large proportion of the poor who currently work for their poverty. In 1989, three out of four poverty households had one or more workers in them while nearly 70 percent of all poor prime-age adults—those between the ages of 25 and 54—did some paid work during the year. Three out of every ten of them worked year-round, full time—giving the lie to Thomas Carlyle's remark a century ago that "work is the grand cure of all the maladies and miseries that ever beset mankind."

Families permanently trapped on the lowest rungs of the income ladder are not the only ones in economic trouble. Job and income instability is becoming "democratized" as a direct consequence of ongoing corporate restructuring that targets professional and white-collar workers for layoffs along with the traditional blue-collar rank and file. Economist Stephen Rose, currently at the U.S. Department of Labor, estimates for the 1970s that 67 percent of men had a "strong attachment" to their firm. They made at most one change in employer during the decade. Such strong company affiliation dropped to only 52 percent in the 1980s. Weak attachment—changing employers at least four times in a decade—doubled from 12 percent to 24 percent. Rose finds that such high employment turnover—presumably dominated by layoffs rather than quits—leads to disrupted career paths, in turn resulting in lower average earnings. For a not insignificant number of families, the loss of a job plunges a family at least temporarily into the ranks of the poor or near poor.

All of these statistics suggest that we have built poverty into the very fabric of the American labor market. In such an economic environment, there is a critical role for public policy designed to boost wages and improve income security. Here is where raising the minimum wage and protecting the EITC come into play.


The federal minimum wage dates to the Fair Labor Standards Act of 1938. In his second inaugural address in 1937 preceding the introduction of this legislation, Franklin Roosevelt called on Congress to help the one-third of Americans who were "ill-housed, ill-clad, and ill-nourished." The original minimum wage of 25 cents per hour helped, but it was only sufficient to raise a worker's earnings to the point where it would, using today's standards, support a family of three at 46 percent of the poverty line.

After World War II, as real gross domestic product (GDP) increased, the federal government improved the minimum wage regularly to make sure the poorest workers would share in the general prosperity. Between 1950 and 1991, Congress raised the wage floor 14 times. At its peak in 1968, the $1.40 per-hour minimum represented—on a full-year, full-time basis—118 percent of the poverty wage for a family of three.

In the 1970s, however, inflation began to outrun the minimum wage. Despite periodic increases in the statutory rate, none was large enough to maintain the minimum's real value. By 1995, a full-time worker on a minimum-wage job could earn only 72 percent of the income needs of a three-person family living at the poverty line. [See "A Minimal Wage," below.] According to New York University economist Edward Wolff, if the minimum wage had kept up with inflation since 1965, it would exceed the current earnings of fully 30 percent of American workers. Nearly one out of three American workers is presently toiling for an inflation-adjusted wage that would have violated the Fair Labor Standards Act 30 years ago.

Subscribe to The American Prospect

Real wages have fallen so far during the past 20 years that, according to the Economic Policy Institute (EPI), raising the minimum wage to just $5.15 an hour from the present $4.25—as President Clinton has recommended to Congress—would directly affect more than 12 million workers who now earn between the current minimum and the proposed one. In addition, because firms try to maintain wage differentials between groups of workers, workers just above the minimum typically benefit as well. EPI estimates that nearly nine million additional workers currently earning between $5.15 and $6.14 an hour will see their wages rise by an average of 10 percent. Together, then, more than 21 million workers—one out of six in the U.S. workforce—would see their wages improve over the next two years if the Clinton wage floor were adopted.

Contrary to propaganda by employer groups, almost three-fourths of these workers are adults; only 25.6 percent are teenagers. Nearly three-fifths are women, whose families are disproportionately in poverty. Over half of all workers who would benefit from raising the minimum wage to $5.15 an hour are found in the poorest 20 percent of all families. Well over two-thirds work in retail trade and services; only 10 percent work in manufacturing where the threat of low-wage international competition may increase if the minimum is raised.

As good public policy, the minimum wage has at least four things going for it. First and foremost, it is a way to increase workers' earnings without placing any burden on the taxpayer. It does not add a penny to the federal deficit. If anything, it decreases the deficit by boosting income tax revenue and reducing welfare payments. Second, it provides increased income to workers who do not qualify for government transfer programs or tax credits. Third, it is an incentive to work in the "above ground" economy rather than in the "underground" economy where wages are often higher than the federal minimum. Fourth, and by no means least, an increased minimum wage may well lead to higher productivity in the economy. At current wage levels, there is little incentive for low-wage employers to introduce new technology or find other ways to boost the output of their workforce. Required to pay a higher wage, firms would have an incentive to find ways to use their workers more effectively.

Despite these advantages, economists' support for mandated minimums has been tepid at best. The minimum wage is viewed as a double-edged sword, boosting wages on the one hand, but forcing employers to reduce employment on the other. Teenagers presumably bear the brunt of minimum-wage-induced job loss, being the least skilled and most expendable in the labor pool. Until recently, empirical research appeared to confirm the economists' concern. A battery of studies during the 1970s and early 1980s concluded that a 10 percent increase in the federal wage floor typically leads to a 1 to 3 percent cut in teenage employment. Among adults, the effect was found to be substantially weaker, but statistically significant, in the range of 0.3 to 0.7 percent.

David Card and Alan Krueger, both economics professors at Princeton University, have carried out meticulous "micro" studies that suggest that in the real world of fast food restaurants and other low-wage employers, modest increases in minimum wages have no negative impact on employment levels. Their results have been attacked by some traditional economists, but none of Card and Krueger's detractors have fundamentally undermined the findings. [See John Schmitt, "Cooked to Order," page 82.] Evidently, there is sufficient slack in the relationship between wages and employment levels—especially in industries not subject to international low-wage competition—that employers either absorb the higher minimum wage in slightly lower profits or find productivity improvements to justify the higher wages, rather than compensate by laying workers off.

Yet even if the new Princeton studies are ignored and we fully accept the critics' estimates of an unemployment effect, it would not constitute much of a case for rejecting a higher minimum wage. Virtually every low-wage worker benefits from a higher minimum wage even if there is an aggregate labor displacement effect. Because of the high job turnover rate among teenagers in particular, and low-wage workers in general, no single individual bears the full burden of unemployment. A recent study of working welfare mothers by the Institute for Women's Policy Research, for example, shows that they change jobs on average every 14 months. Those who benefit from a boost in the federal wage floor share both the higher wages and a portion of any induced unemployment. In the case of the Clinton proposal, the typical working teenager will see his wage go up by 21 percent and—if there is job displacement as severe as early estimates make it out to be—a 6 percent decline in annual hours worked. One presumes few teenage beneficiaries would vote against a wage-job displacement deal that includes a 15 percent annual income increase in return for 6 percent less work! For adults affected by such an increase in the minimum wage, the 21 percent wage increase is offset by no more than a loss of 1.5 percent in annual hours worked—hardly a dent in their improved economic condition. This should vindicate the minimum wage, even without the Card and Krueger results.

But there is another criticism that makes the minimum wage less than ideal as an antipoverty remedy. It concerns what economists call "target efficiency." Only a small proportion of the poor would directly benefit from increasing the wage floor, despite the fact that nearly 75 percent of poor households have someone who works. According to estimates by Richard Burkhauser and Kenneth Couch of Syracuse University and their colleague, Andrew Glenn of Vanderbilt, only 16.9 percent of the workers in poor households in 1991 were in jobs paying below the proposed boost in the statutory minimum to $5.15 an hour. Except for the possible indirect benefits noted above via maintenance of wage differentials, the other 83 percent of working-poor households would not be helped since their working members already earn wages above this level. Even so, according to the EPI study mentioned above, 40 percent of the gains from the proposed Clinton wage hike would go to the poorest 20 percent of working families—those with annual incomes less than $22,000 in 1993. These are clearly deserving families, even if only about half this number are below the official poverty line.

However, the federal wage floor would have to be $6.06 per hour for a full-time worker to earn enough to raise a family of three above the poverty line. Hence, most workers in poor families fall into a "dead man's zone." The Clinton minimum wage is below what they are currently making, but their pay is not sufficient even on a full-time, full-year basis to catapult their families out of poverty.


To assist the most disadvantaged of the working poor, an even better-targeted program is needed—the earned income tax credit. First enacted in 1975, the EITC is a refundable tax credit aimed directly at helping working-poor families with children. Its original intent was to offset a portion of the payroll tax liability of low-income families in order to reduce the regressivity of federal taxes. Being refundable, it has aspects of a "negative income tax." If the credit due a family is greater than its federal tax liability, the IRS remits the balance to the family.

The way the EITC operates is quite simple. For example, a working family with one earner and two or more children receives a 40 cent credit for every dollar earned up to $8,900. Hence, the EITC provides a maximum benefit of $3,560 (0.4 times $8,900). When that family files its Form 1040 in April, total tax liability is reduced by this amount. A check will be sent to the family if the credit exceeds what it owes. A family is eligible for the maximum credit until its total earnings from work reach $11,620. After that, the credit declines by 21 cents per dollar of earnings, and vanishes altogether only when a family's earnings exceed $28,524. For working families with just one child, the maximum credit is $2,156 and the credit "vanishes" at $25,120.

Since the EITC rewards work and not welfare, does not impose mandates on employers, and involves little red tape, even President Reagan was a great fan. In expanding it in 1985, he called it "the best anti-poverty, the best pro-family, the best job creation measure to come out of the Congress." At the time, Senator Bob Dole and Representative Dick Armey chimed in with warm endorsements.

The EITC's greatest advantage from the perspective of battling poverty is precisely the minimum wage's weakness—its target efficiency. According to Burkhauser and his colleagues, more than 46 percent of the program's total tax credit goes to families who are living under the official poverty line while 63 percent goes to families with incomes no more than 1.5 times this low income standard. Only about 15 percent goes to families with incomes 3 times the poverty line or greater—families who have relatively higher incomes but receive only a portion of their income from wages. The U.S. Treasury estimates that in 1996 more than two-thirds of the credit will go to families with income under $20,000. Hence it affects millions of families in the "dead man's zone"—those not helped by the minimum wage. For a family of four with one earner making $7 an hour and working 1,500 hours a year, the EITC fills nearly 80 percent of its "poverty gap," the difference between after-tax earnings and the poverty line.

The EITC has still another great advantage often overlooked by both its supporters and its detractors. It is a form of "family wage insurance" in an era of job instability and earnings insecurity. Stephen Rose has estimated that fully 39 percent of families would be eligible for the EITC at least one year in ten. In any one year, about one in six families is eligible for the tax credit, but over a decade, nearly two out of five families with children will have a year or more in which their wage income declines sufficiently for them to be eligible for the EITC. The EITC proves particularly useful for younger families, those hardest hit by the nation's falling median-wage rate. By the end of any given decade, more than half of families headed by individuals who are no more than 35 years old benefit from the credit. The EITC therefore serves two important functions. It is life support for the permanently low-wage worker; it is earnings insurance for the middle class.

Building a Living Wage

How raising the minimum wage and keeping the EITC intact helps working families put food on the table:


  Gross After FICA EITC Total
Family 1
(one adult earner, two children)
Earnings Witholding Value
Current minimum wage ($4.25)/no EITC $6,375 $5,887 $0 $5,887
Current minimum wage ($4.25)/EITC $6,375 $5,887 $2,550 $8,437
Clinton minimum wage ($5.15)/EITC $7,609 $7,027 $3,044 $10,071
Family 2
(two adult earners, two children)
Current minimum wage ($4.25)/no EITC $12,750 $11,775 $0 $11,775
Current minimum wage ($4.25)/EITC $12,750 $11,775 $3,323 $15,098
Clinton minimum wage ($5.15)/EITC $15,218 $14,054 $2,805 $16,859

Earnings are for persons working at minimum wage 30 hours per week, 50 weeks per year. These simulations assume a 1.5 percent reduction in hours worked as a result of raising the minimum wage from $4.25 to $5.15 per hour.

As impressive as these benefits may be, the EITC is no cure-all and has a number of serious weaknesses. For one thing, it is of no help to a large segment of the poor—unrelated individuals—and its aid to childless couples is nearly inconsequential, amounting to a maximum of $324 per year.

An even greater disadvantage, at least politically, is that the EITC is expensive from the perspective of the taxpayer. This year, the EITC will cost the federal treasury more than $25 billion and its annual cost is projected to rise to $30 billion by the year 2000 in order to keep up with inflation and population growth. It is one of the reasons why, in their zeal to balance the budget by 2002, the Republicans have targeted EITC for significant cuts despite their previous enthusiastic support for it.

A third problem is that the EITC is considered subject to abuse. Until administrative changes were made in the program last year, the so-called "error rate" for the credit was found to be extremely high relative to other IRS provisions. Some families were receiving the credit when they were ineligible; others were receiving more credit than they were legally permitted; still others were receiving less. Since there is benefit to be gained from overstating one's income at very low earnings in order to move up the EITC schedule and benefit from understating income at higher earnings levels to retain the maximum tax credit, there will always be some enforcement issue needing attention—even if it presents the IRS with nowhere near the headache of policing tax compliance among the rich.

A more serious problem is related to what might be called the "Speenhamland" effect. The EITC not only subsidizes workers, it subsidizes employers. It permits employers to keep wages low while relying on the federal government to help workers make up the difference between substandard earnings and something approaching a living wage. The British learned this lesson 200 years ago when they imposed their equivalent of the EITC in the form of the infamous Speenhamland provisions. Introduced during the first decades of the industrial revolution, Speenhamland subsidized factory wages to keep workers from starving to death. Quickly, employers realized they could drive down wages and let the government pick up the tab for their employees. By the early 1800s, Speenhamland was bankrupting local treasuries and the laws were repealed. There is a lesson here. Certainly in an era of "privatization," one would think that "privatizing" wages should be high on the agenda. The EITC socializes them.

For this reason, the credit also has a potential adverse effect on productivity. A wage subsidy tends to reduce the incentive for investment in new technology and capital. If an employer can obtain labor for 80 cents on the dollar, why invest in labor-saving technology?


The conclusion should be obvious. An anti-poverty program aimed at low earnings must include both an increase in the minimum wage and retention (if not continued expansion) of the EITC. The disadvantages of the one are offset by the advantages of the other.

The minimum wage has poor target efficiency; the EITC's efficiency is much superior. The minimum wage does not generate jobs and may displace some; the tax credit has no job displacement effect and might encourage some firms to expand employment. On the other hand, the minimum wage is much superior to the EITC when it comes to taxpayer cost, aiding unrelated individuals and childless families, and countering the Speen hamland and adverse productivity effects. Indeed, periodic increases in the minimum wage will over the long term actually reduce taxpayer liability under the EITC. Whenever the federally mandated wage floor boosts a family's earnings above $11,620, the credit is reduced by 21 cents on the dollar. Combined, then, the minimum wage and the EITC give us the best of both worlds—target efficiency for attacking wage poverty at reasonable public cost.

Clearly, the combination of minimum-wage regulations and the EITC makes for good antipoverty policy, especially in an era when the majority of poor people are working-age adults and job insecurity is on the rise. The chart ["Building a Living Wage," page 45] provides estimates of how much an increase in the minimum wage to $5.15 an hour in tandem with the current EITC would actually help low-income families. Take "Family 1" with two children and one adult earner working 1,500 hours per year at the current $4.25 minimum wage. At the current wage floor, this family would have a total annual income of just $5,887 (after subtracting its share of Social Security taxes) excluding its EITC. The EITC alone boosts this family's income to $8,437. Lifting the minimum wage to $5.15 an hour will raise after-tax earnings to $7,027 and total income to $10,071—even if we take into account the highest possible job displacement effect due to the boost in the minimum wage. Overall, this family's income is raised by 71 percent as a result of the higher minimum and the tax credit.

Depending on the wage rates currently earned, the number of annual hours worked, and the number of earners in the family, the gain from the minimum wage-EITC combo varies substantially. But the examples provided in the chart indicate that few antipoverty programs could be more effective.

That Republicans refuse to increase the minimum wage—and a good number have suggested jettisoning it altogether—shows neither compassion for the working poor nor good economic sense. That a number of the currently fashionable "flat tax" proposals eliminate the EITC adds substantially to the enormous regressivity of these measures.

This is not to say, by a longshot, that reliance on federal wage standards and tax credits is the ultimate answer to low wages and poverty. The social policy jigsaw puzzle has many more pieces that must be fitted together to boost employment opportunity, improve living standards, and reduce income insecurity. Fashioning a full-employment macro policy is part of the overall puzzle, as are education and training programs, community economic development strategies, increased unionization, and fair trade. But in this political era of trying to hold on to some of the gains we have made in the past, the battle for improved minimum-wage regulation and maintenance of the EITC are well worth joining with a lot more gusto than we have seen from the White House this past year.


  • David Card and Alan Krueger, Myth and Measurement: The New Economics of the Nimimum Wage (Princeton University Press, 1995).
  • Lawrence Mishel, Jared Bernstein and Edith Raselle, "Who Gains with a Higher Minimum Wage," Economic Policy Institute, 1995.
  • Stephen J. Rose, "Long Term Eligibility for the Earned Income Tax Credit," Research Report #95-05, National Commission for Employment Policy, 1995.

You may also like