The federal government spent around $4 trillion ($30,000 per household) on entitlements in 2025, up from $1 trillion in 2000. This spending is supposed to prevent poverty, but the bulk of it goes to the middle class.
Left-leaning political thinkers around the world argue that expanding publicly funded social-welfare benefits to all citizens is the best way to help the poor. Is this true?
Finally, a reason to check your email.
Sign up for our free newsletter today.
The evidence from the United States and Europe suggests not. As I argue in my new book, The American Way of Welfare, while expanding benefits to the middle-class may help left-wing politicians broaden their electoral support, it does little to benefit the poor.
Limiting the provision of public assistance to those who really need it seems like common sense. It ensures that welfare expenditures alleviate unavoidable hardships, rather than burdening taxpayers in order to support those who could provide for themselves.
Yet middle-class entitlements now dominate the welfare state. In 2025, federal spending on Social Security and Medicare (8.5 percent of GDP) substantially exceeded that on Medicaid and other means-tested benefits (4.9 percent of GDP). Only 8 percent of Social Security benefits go to the poorest 20 percent of seniors in any particular year. The richest quintile receives benefits for nine more years than the poorest does.
It’s been clear for decades that entitlement programs are inadequately focused on filling unmet needs. In a 1971 televised debate on Social Security, economist Milton Freedman said, “in short, viewed as a way to distribute government assistance to the needy, the structure of benefits is not defensible,” as it combines a highly regressive tax with largely indiscriminate benefits. Wilbur Cohen, Lyndon Johnson’s Secretary of Health, Education, and Welfare, responded that this problem owed to deliberate political considerations: “[A] program that is only for the poor—one that has nothing in it for the middle income and the upper income—is, in the long run, a program the American public won’t support.”
Cohen’s opinion remains highly influential. Many on the Left believe that public aid for the poor is stingy and unreliable. They call for expanding entitlement eligibility to all, regardless of need, fueling the push for a federal Universal Basic Income program, Medicare For All, and Universal Free College.
The level of public assistance has long been designed to avoid inducing dependence, regardless of the scope of eligibility for benefits. In 1834, John Stuart Mill argued that “no person who is able to work is entitled to be maintained in idleness; or to be put into a better condition, at the expense of the public, than those who contrive to support themselves by their unaided exertions.” In fact, he deemed this “the principle upon which all good government, and all justly constituted society rest.”
Over the subsequent two centuries, the magnitude of American welfare benefits has been increased in steady proportion to entry-level wages. But welfare benefits for the poor have not increased beyond that level. This is deliberate, and popular: a study of survey data from 1962 to 1992 concluded that “voters wish to maintain a target benefit-wage ratio, either for reasons of equity with the non-welfare working poor or to control work disincentive.”
Fiscal constraints traditionally prevented the broadening of eligibility for benefits beyond the needy poor. Revenue was hard to raise while the bulk of the population were subsistence farmers, and broad-based tax increases remained unpopular even after the nation industrialized. But World War II saw an enormous increase in the share of workers paying income taxes, from 7 percent in 1939 to 72 percent in 1945. Federal revenues surged to 20 percent of GDP—a level around which they have since remained.
As defense spending subsequently declined, abundant revenues were largely diverted into the expansion of Social Security and Medicare benefits. Though President Johnson declared War on Poverty, the bulk of his social-welfare expenditures went to middle-class entitlements and grants to state and local governments. So much of the funding for overtly anti-poverty programs was absorbed by a class of lawyers and professional social workers that Daniel Patrick Moynihan called it “a policy of feeding the sparrows by feeding the horses.”
Eventually, fiscal constraints began to bite, as economic growth slowed and defense spending could not be cut further. When federal court orders expanded welfare eligibility, states managed their budgets by trimming benefit levels. By 1970, government transfers increased child poverty by 4 percent, because the burden of payroll taxes paying for Social Security and Medicare exceeded the magnitude of benefits for poor working families.
The expansion of entitlements to the middle-class was even more substantial in Europe, where World War II had more thoroughly disrupted private employment, savings, and insurance. But the expansion of benefits to the middle class did not benefit the poor. In Britain and France, benefits for the working class, net of taxes, declined relative to their postwar level. After the United Kingdom established its universal National Health Service in 1940s, mortality increased among the working class, as scarce public resources shifted toward the middle class.
Today, the greater cost of Europe’s welfare states relative to America’s is largely due to the extent of middle-class entitlements—particularly pensions for retirees. To finance these, Europe must impose substantial sales and payroll taxes, which fall disproportionately on low earners. Calculating the net effect, French economists in 2022 determined that “after accounting for indirect taxes and in-kind transfers, the US redistributes a greater share of national income to low-income groups than any European country.”
Proposals to expand the welfare state further suffer from the same malady. Medicare For All proposals would raise taxes on the poor, elderly, and disabled, who are currently eligible for federal entitlements, in order to pay for benefits that are currently financed privately. A 2021 study estimated that a Universal Basic Income of $12,000 would require a 27-percentage-point across-the-board tax increase and lead workers to reduce their earnings by $6,000.
Aid for the poor is not inherently unpopular. Last year, only 13 percent of Americans told pollsters that the government spends too much on assistance for the poor, while 62 percent said that it spends too little.
Programs (whether universal or targeted at the poor) similarly struggle politically when they deter employment or are viewed as ineffective. In July 2021, 52 percent of Americans surveyed wanted a $300 per week Pandemic Unemployment Compensation eliminated immediately, while only 16 percent wanted it to continue indefinitely.
A recent study by Robert Greenstein of the Brookings Institution found that budgets for some means-tested programs (such as Medicaid or Food Stamps) have grown faster than those for universal benefits (such as Medicare or Unemployment Insurance). He notes that targeted programs tend to fare better in congressional budgetary negotiations, because they do more with every dollar to fill unmet needs.
In fact, extending benefits to the middle class often makes those benefits less generous. For instance, as means tests have been eliminated for free school lunches, budgets have not been proportionately increased—and so the quality of food has deteriorated.
Government spending tends to stick where it hits. Once established, all benefits are hard to cut, regardless of who receives them. When government distributes public spending to the middle-class, it generally stays there without trickling down to the poor.