Shattered Consensus. James Piereson

the overwhelming burden of the federal income tax, while those below the median income have been taken out of the income tax system altogether.

      For several generations, progressive reformers have looked to the income tax as the instrument through which they might take resources from the rich and deliver them to the poor. In reality, the income tax—in the United States, at least—is not a sufficiently large revenue source for the national government to do the job that the redistributionists want it to do. In 2010, when the federal government raised $2.144 trillion in taxes, just 42 percent of that sum came from the individual income tax, while 40 percent came from payroll taxes, 9 percent from corporate taxes, and the rest from a mix of estate and excise taxes. Since the early 1950s, the national government has consistently relied upon the income tax for between 40 and 50 percent of its revenues, with precise proportions varying from year to year as economic conditions change.

      Payroll taxes account for nearly as large a share of federal revenue, and these taxes fall more heavily on the wages and salaries of the working and middle classes than on the incomes of the wealthy, whose salaries far exceed the maximum earnings subject to those taxes and who derive income disproportionately from capital gains. In 2010, the wealthiest 1 percent paid just 4 percent of payroll taxes (and 39 percent of income taxes). The top quintile of earners paid 45 percent of payroll taxes (and 93 percent of income taxes), while the middle quintile paid 15 percent of all payroll taxes (and just 3 percent of income taxes). The more widely shared burdens of the payroll tax tend to counteract the redistributive effects of the income tax.

      An increase in the top marginal tax rate in 2010 from 39.6 to, say, 50 percent might have yielded around $100 billion in additional revenue, assuming no corresponding changes in tax and income strategies on the part of wealthy households and no negative effects on investment and economic growth (all risky assumptions). That is real money, to be sure, but it would represent only about 0.5 percent of GDP (using 2010 figures), or less than 3 percent of total federal spending—not enough to permit much in the way of redistribution to the roughly 60 million households in the bottom half of the income scale.

      * * *

      Turning to the spending side of fiscal policy, we encounter a murkier situation due to the sheer number and complexity of federal spending programs. The House of Representatives Budget Committee estimated in 2012 that the federal government spent nearly $800 billion on ninety-two separate antipoverty programs that provided cash assistance, medical care, housing assistance, food stamps, and tax credits to the poor and near poor. The number of people drawing benefits from antipoverty programs has more than doubled since the 1980s, from 42 million in 1983 to 108 million in 2011. The redistributive effects of these programs are limited, however, because most of the funds are spent on services to assist the poor, while only a small fraction is distributed in the form of cash or income. This means that most of the money goes to providers of services, who are generally in the middle class or upper middle class, and not to poor or near-poor households. Senator Daniel Patrick Moynihan once tartly described this dynamic as “feeding the horses to feed the sparrows.”

      The American welfare state evolved in the direction of services rather than income in part because the American people have long viewed poverty as a condition to be overcome rather than one to be subsidized with cash. Many also believe that the poor would squander or misspend cash payments and so are better off receiving services and in-kind benefits like food stamps, health care, and tuition assistance. Regarding aid to the poor, then, Americans have built a social service state but not a redistribution state.

      Social Security is the only substantial federal program that transfers money income from one group to another, in this case from workers and employers to retirees. It is by far the largest of all federal programs, claiming $850 billion or 24 percent of the federal budget in 2014. It is funded by a payroll tax split equally between employees and employers. As of 2014, about 59 million Americans were collecting benefits under Social Security, with an average benefit of $1,260 per month. Social Security has a progressive benefit formula and it contains a feature (Supplemental Security Income) that provides cash benefits to elderly, blind, or disabled persons with incomes below the poverty line. Nevertheless, it was designed and it still functions to provide income for retirees, not to redistribute income from the wealthy to the poor.

      The National Bureau of Economic Research, in a series of studies on the redistributive aspects of Social Security, concluded that the program transfers income in various complex ways but does not transfer it from the rich to the poor. One NBER study by Julia Lynn Coronado, Don Fullerton, and Thomas Glass bluntly concluded that “Social Security does not redistribute from people who are rich over their lifetime to those who are poor. In fact, it may even be slightly regressive.” This is partly because wealthier recipients tend to live longer and partly because they are more likely to have nonworking spouses also eligible to collect benefits.

      Medicare and Medicaid, two other expensive programs that together claim nearly 25 percent of the federal budget, provide important health-care services to the elderly and the poor, but no actual income. The flow of money through these health-care programs, more than $850 billion in federal funds in 2014 (plus another $180 billion in state funds for Medicaid), goes mainly to hospitals, nursing homes, pharmaceutical companies, doctors, insurance companies, and health maintenance organizations. Both programs have been plagued by fraud and corruption since their origins in 1965 because some doctors, nursing home entrepreneurs, and other providers have sought to game the system for financial advantage, and in many cases have succeeded all too well. No one has ever attempted a study of the redistributive aspects of the flow of funds from Medicare and Medicaid, but one surmises from the nature of these payments that most of the money goes to those in the upper reaches of the income distribution.

      The federal government does provide cash assistance to the poor and near poor through two programs: (1) Temporary Assistance to Needy Families (TANF, popularly known as welfare), which currently provides cash benefits to about 4.5 million households at a cost of $17 billion per year to the federal government and about $14 billion (in 2014) to various state governments; (2) Supplemental Security Income (noted above), which provides cash benefits to the disabled poor, aiding about 8.5 million households at a cost of around $50 billion per year to the federal government. These numbers work out to approximately $7,000 on average per year per household under TANF and $6,000 per year per household under SSI, in each case around half of the average benefit under Social Security.

      Lower-income working families are also eligible to receive rebates on payroll taxes through the Earned Income Tax Credit (which, again, the CBO counts as a negative tax rather than income). The House Budget Committee estimated that 28 million taxpayers took advantage of this program in 2011, at an estimated cost of $60 billion to the federal government; the average family with children claiming this benefit received $2,900 in tax rebates.

      Of the $800 billion spent on antipoverty programs in 2012, less than $150 billion, or about 18 percent of the total, was distributed in cash income, if one includes the tax rebate under the EITC. The rest of the funds were spent on services and in-kind benefits, with the money paid to providers of various kinds, most of whom have incomes well above the poverty line.

      With respect to the recipients of federal transfers, the CBO study reveals a surprising fact: households in the bottom quintile of the income distribution receive less in federal payments than those in the higher income quintiles. According to that study, households in the lowest quintile received on average $8,600 in cash and in-kind transfers in 2012; households in the middle quintile about $16,000 in such transfers; and households in the highest quintile about $11,000. Even households in the top 1 percent of the distribution received more in dollar transfers than those in the bottom quintile. The federal transfer system may move income around and through the economy, but does not redistribute it from the rich to the poor.

      * * *

      It is well known in Washington that the people and groups that lobby for federal programs are generally not those who receive the services, but those who get salaries and income for providing the services. They, as Senator Moynihan observed decades ago, are the direct beneficiaries of most of these programs, and they have the strongest interest in keeping them in place. The nation’s capital is home to countless trade associations, corporate lobbyists, hospital and medical associations,