Eduardo Porter, a reporter for the New York Times, has an excellent recent column that summarizes some of the evidence for how government can effectively have large effects in reducing income inequality. His column builds on the research of sociologist Lane Kenworthy, and economist Peter Lindert.

The argument is that developed countries that use government to significantly reduce income inequality, such as some countries in Western Europe, do not do so primarily through having a more progressive tax system, in which higher income households pay much higher percentage tax rates than other households.

In fact, the U.S. overall probably has a more progressive tax system than is true for most countries in Western Europe. Many Western European countries rely heavily on value-added taxes, which are similar to a national sales tax. Because these taxes are based on consumption, the percentage of income paid in value-added taxes tends to decline somewhat for upper-income households.

Instead, Western European countries that redistribute income more than the United States primarily accomplish redistribution on the basis of having higher average tax rates. These higher average tax rates allow more government spending on transfer payments (transfer payments are cash or near-cash payments to individuals or households, such as Social Security, welfare, or food stamps) and on public services such as education.

This higher government spending on transfers and public services is the key mechanism by which Western European countries accomplish greater income redistribution.  The various levels of government in the U.S. are less effective in redistributing income not because of defects in the progressivity of our tax system, but because we have lower overall tax rates.

Government spending is more effective than progressive taxation in redistributing income for several reasons. First, there are political and economic limits to how much progressive taxes can do to redistribute income. Concern over economic incentives limits government willingness to have higher tax rates on the wealthy and on business. Perhaps more importantly, the political influence of higher income groups makes it quite difficult to have highly progressive tax systems. Politics tends over time to create various deductions and loopholes that limit the effective progressivity of tax systems.

Second, even universal services tend to be far more progressive in their influence on the income distribution than is true of even the most progressive tax systems that are politically and economically feasible. If a public service is provided to all households on a universal basis, the value of this public service is usually a far larger percentage of income for lower-income households than for upper-income households. For public services and transfers whose benefits are targeted based on family income, for example welfare benefits or subsidized housing that only goes to lower-income households, the distribution of benefits is even more progressive.

Third, public services that are universal in their benefits, or that at least provide benefits to a broad majority of the public, are far more politically sustainable than is true for progressive tax systems.  Voters are more inclined to pay attention and vote on the basis of preserving benefits they receive rather than on the basis of raising taxes on someone else.

Fourth, for public services that are “productive”, that is that provide benefits whose value is greater than their cost, a relatively large amount of income redistribution can be accomplished for a lower dollar cost. If a particular public service provides $3 in benefits per dollar of costs, then we can benefit lower-income households and middle-income households quite a bit even if we raise their taxes at the same time. In contrast, if increasing the progressivity of the tax system leads to tax evasion strategies, or to real changes in investment, then raising an extra dollar of taxes from a progressive tax system may actually have economic costs of considerably greater than a dollar.

Universal preschool is a good example of a productive public service that can accomplish a great deal of income redistribution even if its benefits are universal and even if the tax system is not very progressive. For example, in my book Investing in Kids, I simulated the effects of universal pre-K financed by a regressive tax system, which is a tax system that takes a higher percentage of income from lower-income households. The regressivity I assumed for the tax system is the typical regressivity of state and local tax systems in the United States.

Even under these assumptions about universal preschool’s financing, I found that the lowest income quintile of households, that is households whose income was in the lowest fifth of the income redistribution, received benefits that were 25 times the extra taxes they paid. In contrast, benefits for middle-income households (the middle-income quintile) were about 3 times the extra taxes they paid. Upper income households (the top-income quintile) received benefits of about 32 cents per every extra dollar they paid in taxes. (For more details, see Table 8.2 of chapter 8 of my book. This chapter is one of those that is available for free online, here.)

Even with the typical regressive financing of state and local tax systems, universal preschool produces clear benefits to a majority of all income groups. But the benefits are highly skewed towards benefitting lower income households the most, followed by middle-income households.

If we want less income inequality, a more progressive tax system is a worthy goal. But even more important is the amount of productive public spending the government is able to undertake.   A government with smart spending on productive public services can do a great deal to affect income redistribution at affordable costs. If these public services are truly productive, the benefits in increased economic output from the services will exceed the economic costs of the taxes levied.