Two myths about the distribution of income comforted liberals during the 1950s. They believed that incomes were becoming steadily more equal, and that government spending programs uniformly accelerated that trend. Frederick Lewis Allen spoke for a self-satisfied New Deal generation when he said that “we had brought about a virtually automatic redistribution of income from the well-to-do to the less well-to-do”1 through the mechanism of the welfare state.

As the Fifties ended the first myth began to give way to evidence. Income distribution, it appeared, had barely changed since the war. The changes that did occur were at the expense of the poorest as well as the richest parts of the population, yielding a slight bulge around the middle classes. In 1947, the poorest 20 percent of the nation’s families shared only 5.1 percent of the nation’s income; by 1960, they had to make do with 4.9 percent.2 Between 1947 and 1960, the percentage of families earning less than one-half the median income—a good relative measure of poverty—increased from 18.9 percent to 20.3 percent.3

As liberals began to acknowledge the persistence of inequality in the private economy, they redoubled their faith in spending by “the public sector.” John Kenneth Galbraith’s 1958 book, The Affluent Society, summed up the postwar liberal strategy. Galbraith argued that inequality could not be attacked directly through a stiffer income tax. Most Americans, he observed were too prosperous to care very much about soaking the rich. Thus equality had faded as a political issue. It could be approached only indirectly, through a general expansion of public spending. The poor, Galbraith argued, “would be among the first beneficiaries of [government spending on] education, health, housing and other services….” Admittedly, an enlarged public sector would support itself in major part through regressive taxation, for example by hiking sales taxes. But that was no objection:

It will be argued that some people are still very poor. The sales tax, unlike the income tax, weighs heavily on the small consumption of such individuals. But if the income tax is unavailable or in service to other ends, the only alternative is to sacrifice social balance [between public and private spending]…. The modern liberal rallies to protect the poor from the taxes which, in the next generation, would help eliminate poverty.4

Galbraith’s ideology made virtuous what liberal politicians had already found necessary: a program which did not rely principally on the defunct Roosevelt coalition based on labor and the ethnic minorities. The defeat of national health insurance and the emasculation of public housing had by 1952 ended hope for a second New Deal. Liberals could succeed only with new measures and new partners. Within a few years the transition was accomplished. Labor and the academic liberals joined with downtown business interests in sponsoring urban renewal; with auto manufacturers and highway contractors in supporting vast road projects; with suburban legislators in promoting state-supported higher education.

But the result was not a redistribution of income. Galbraith had assumed that a basically middle-class electorate could be hoodwinked into supporting a measure of equality through public spending that it would never endorse directly. If anything, the outcome was the reverse. The new public programs built homes and roads for the middle class, provided college education for the upper middle class, and lavished private aviation facilities on the rich. The poor were placed at the wrong end of a bulldozer. Only in the mid-Sixties; with the proclamation of the War on Poverty, was any new measure of redistribution even attempted. But its scope has been minor and, in significant part, rescinded. In the meantime, regressive sales and social security taxes have been growing; the income tax during the Fifties was “progressive”—i.e., it imposed larger proportionate burdens as income rose—only to a minute degree, and this has now been largely eroded.5

Precise statistics have become available on only some of these questions. They confirm the durability of inequality in the private economy and the primarily passive role of the public sector, except for compulsory education and subsistence welfare for the very poor. H. P. Miller’s book, Rich Man, Poor Man, shows that the distribution of income among classes was even less mobile from 1957 to 1968 than in the previous decade. There was a shift along racial lines—the relative income of non-whites increased from 51 percent of the white median in 1958 to 63 percent in 1968. But much of this change was due to black migration from the low-wage South and from rural to urban areas, not to waning discrimination or better education. The real gains from such higher income in wretched city ghettos is debatable to say the least. A more detailed study by J. Gwartney has found, moreover, that, within particular regions, non-white males have made no relative progress during the entire 1947 to 1969 period, although there were substantial gains for non-white females.6


Gwartney shows that changes in relative education have actually worked against non-whites since World War II. The jobs requiring the most education show the greatest racial disparities in income; thus, as both races become better educated, the comparative situation of blacks tends to deteriorate. This effect has been offset somewhat by faster educational progress for nonwhites. But even if blacks were suddenly to receive the same amount of schooling as whites, gross income differentials would persist. H. P. Miller quotes 1960 census figures showing that “the average non-white with four years of college could expect to earn less over a lifetime than the white who did not get beyond the eighth grade.”

Taxes have hardly any effect on the relative distribution of income. The progressive rates of the income tax are offset by well-known loopholes, more respectable but equally regressive benefits (e.g., the deductibility of mortgage payments), and regressive social security taxes. The over-all burden of federal taxes is roughly proportional to income for all but the affluent.7 State and local taxes are sharply regressive for the very poor and roughly proportional for everyone else.8 The net impact is a small gain for the middle classes at the expense of the extremes. The tax system is not, and has never been, a vehicle for significant redistribution.

Government spending does accomplish some redistribution, though in a far more limited way than liberals have imagined. The basic distinction is between straight payments which transfer money directly from the government—such as welfare and social security—and government programs designed to provide specific goods and services, such as schools, housing, and transportation. Transfer payments are indispensable aids for the very poor—families earning under $2,000 a year in 1965 received more from government transfers than from all other sources. But transfers barely affect the distribution of income for the highest-earning 85 percent of the population.

For other government programs the story is more complex. The largest item is primary and secondary education. If one makes the assumption that schools are worth what they cost, the poor gain substantially from stateprovided education.9 But studies such as the Coleman report—an extensive statistical survey which found that the level of school expenditures made no difference on achievement test scores—suggest a more skeptical appraisal.10 Similarly, one must be skeptical about the stated cost of programs like Medicaid and public housing which provide transfers of services and buildings to the poor. Segregated and substandard facilities and a lack of other options make these programs less valuable to their beneficiaries than an equivalent amount of cash.

Even more striking evidence concerns the programs central to the post-New Deal liberal strategy, such as subsidies for higher education and tax relief for homeownership. In 1966, the most recent year studied, federal tax benefits for private housing amounted to $7 billion, dwarfing the $500 million subsidy for public housing. These tax provisions were strongly regressive, conferring twenty times as much benefit, in absolute terms, on a family earning $50,000 as on one earning $5,000.11

State-supported higher education, although legally available to all, is of use primarily to the more affluent. University of Wisconsin economists W. Lee Hansen and Burton Weisbrod have studied California’s supposedly egalitarian system of colleges and universities.12 They found that students from poor families were the least likely to be eligible for state college or the University of California, and, if eligible, the least likely to attend. Proportionately, six times as many high-school graduates from families earning over $25,000 planned to attend U.C. as did those whose parents earned less than $4,000.

To be sure, all high-school graduates in California are eligible for two-year junior colleges. But the state subsidy for a student who completes state college or university is between four and five times that for a junior college graduate. The net effect of free higher education, even in California, has been to reserve the highest scholarships for the richest students.

For many other public programs the results are similar. Federal transportation subsidies overwhelmingly aid the automobile, not the subway; farm programs sustain the squire, not the serf; urban renewal calms the middle classes and dispossesses the poor. In general, government enterprise has proven little more effective than taxation in rearranging the American distribution of income.

For Great Britain the evidence is remarkably parallel. During the 1950s, writers on both the left and the right assumed that wartime taxation and postwar Labour rule had accomplished a major leveling of incomes.13 Anthony Crosland’s revisionist manifesto, The Future of Socialism, argued that financial equality had been achieved to such a degree, and that the perquisites of ownership had become so diminished, that Britain was no longer a capitalist state.

By the Sixties, however, work by Richard Titmuss and his associates at the London School of Economics Department of Social Administration began to discredit the egalitarian myth.14 The new study by Webb and Sieve confirms and extends Titmuss’s argument. British welfare state benefits, they observe, were financed largely by regressive taxation. The net impact of government taxes and benefits has been simply proportional to income, and thus not redistributive. Nor has the private economy become more evenhanded. Webb and Sieve conclude that “the estimated inequality of final income remained constant over a period of twenty years (1939-1959) which saw the establishment and growth to some stability of the ‘welfare state’” (emphasis in original). The recent Labour government plugged some income tax loopholes but also escalated the taxes paid on consumer purchases; a respectable case can be made that the poor grew relatively poorer under Labour.15


What explains the failure of egalitarian policy? Much of the blame lies with the liberal planners and their overambitious multiplicity of goals. Liberals have believed not merely in redistribution, but in the inherent desirability of a large public sector, in free and universal access to the social services, and in the use of tax exemptions for social purposes. The result has been to multiply the burden on the income tax while contracting its base. “We have traveled down the vicious path of decadent tax systems,” observed the Labour economist Nicholas Kaldor, “the path of charging more and more on less and less.”

As the income tax became overloaded and unpopular, liberal politicians turned to regressive payroll and consumption taxes to pay for the welfare state. Britain’s pension system—financed largely by a highly regressive head tax—provides an illustration. In America, adoption of proposals for universal health insurance will compound the burden of the regressive payroll tax and increase pressure for a national sales tax (probably in the form of the European tax “on value added,” a scheme already receiving friendly scrutiny in the Administration).

To be sure, much of what has been done was thought to be politically unavoidable. Labour and the Democrats needed middle-class votes; the price they paid was middle-class programs. Tax benefits for homeownership, and Social Security pensions that vary inversely with need, have helped to ensure the permanence of New Deal legislation and of Democratic Congressional majorities. In this explanation, however, lies a refutation of Galbraith. The public sector, rather than serving to camouflage an unpopular program of equality, has simply hidden from liberals evidence of their own willingness to ally themselves with the more prosperous and to abandon the poor.

In the Fifties, liberal enthusiasm for public spending probably did not thwart redistribution, since major political forces were in fact unwilling to bring pressure for it. Galbraith accurately summarized the political stalemate on equality. But in recent years, the War on Poverty, the collapse of the welfare system, and the growing militancy of the poor themselves have given legitimacy to proposals for a negative income tax. Some form of explicit redistribution—probably the Nixon Administration’s diminutive Family Assistance Plan—seems sure to be adopted.

The constraint at present is more financial than ideological. Even a halfhearted negative income tax would have an enormous cost. Providing a minimum $3,600 income for a family of four would add $25 billion to the federal budget; a $5,500 plan would cost $71 billion.16

These sums are not about to be voted by today’s Congress. But the amount that will be available depends,in large part, on other claims on the federal budget. Theoretically, room could be found for an adequate negative income tax by lopping off the most wasteful and destructive parts of the budget, such as excess defense spending and road building. The United States could defend itself on $20 billion less per year. Additional funds—Joseph Pechman estimates $25 billion—could be raised through federal tax reform. But in practice, liberals have never had the votes to squeeze $10 billion from these sources, much less thirty or sixty billion.

Funds of this magnitude will be raised, if at all, only at the expense of traditional liberal enthusiasms. Conversely, a budget that provided free universal health insurance, generous urban aid, no-tuition university education, and huge housing subsidies would not contain funds to end poverty as well. Income redistribution and growth in government spending programs are, even more than in the past, rival attractions.

This Issue

August 12, 1971