In the Tuesday Wall Street Journal, Professor Alan Blinder wrote of his puzzlement at the very slow growth of productivity in the last three years. There is really no mystery. The rate of growth of investment in...
- Frequently Asked Questions about the Tax Foundation'...
Frequently Asked Questions about the Tax Foundation's Fiscal Incidence Study
Why did you publish this study?
The goal of this study is to show lawmakers and the public that there are two sides of fiscal policy: taxes and government spending. Currently, most tax debates in Congress only focus on tax distributional tables. But while taxes are important, how those tax dollars are spent is also important. By focusing only on taxes, we're missing half the fiscal picture.
With this study we hope to encourage official agencies like the Congressional Budget Office and the U.S. Treasury to start producing official estimates of the distribution of government spending as well as taxes. Government statistical agencies in the U.K., Australia and elsewhere have done so for years, and the United States government should follow their lead.
Has this type of study ever been done before?
Yes. Our working paper shows there are many academic studies on this topic dating to the 1940s. Official agencies such as the World Bank and the International Monetary Fund have conducted similar studies throughout the world. Official agencies in the United Kingdom and Australia have long produced similar studies. Our study follows conventional assumptions in the previous literature, both on the tax and spending sides.
Is your study relevant to tax and spending debates?
Yes. Lawmakers have to decide how progressive they want the fiscal system to be. Ideally, they would set a goal for how much they want to assist lower-income households and then achieve that goal in the most efficient way possible. But by only measuring tax progressivity and ignoring spending, lawmakers are missing half the fiscal picture.
For example, say lawmakers want to give additional aid to poor households. One way is to boost the Earned Income Tax Credit (EITC), which sends cash payments to low-income households in the form of a negative tax. Since the CBO and others count EITC in tax distribution tables, such a policy would show a large tax cut for the poor, and thus an increase in progressivity.
However, if that same goal were achieved by boosting Temporary Assistance to Needy Families (TANF) spending, tax distributions alone would show zero increase in progressivity. Clearly this is incorrect, as the two ways of helping low-income households are very similar. But today, lawmakers are not even counting the increased progressivity in the fiscal system that comes from these direct transfers.
By combining taxes and spending, our study shows the supposed tradeoff between "efficiency" and "fairness" in tax policy is often illusory. In many cases, we can achieve both a more efficient tax code and whatever level of aid to low-income households lawmakers desire.
Does your study really tell us who "benefits" from government spending?
No. We only measure the amount of money that governments spend on households. We don't measure how much those households value that spending or how much they economically "benefit" from it. Similarly, estimates of tax burdens from the CBO and others don't tell us who "benefits" or is "hurt" from taxes; they only measure which households are likely to pay dollars of tax collections.
For example, CPAs likely benefit highly from various provisions in the tax code, which is not counted in tax distributions. Similarly, tax distributions attribute the value of the home mortgage interest deduction to homeowners, ignoring the fact that realtors and home builders benefit substantially from those tax preferences. Tax distributions only answer the very limited question of which households likely fund government operations. They do not answer the question of who "benefits" or "loses" from taxes. Our study follows this conventional approach as well, and simply applies the logic of tax distribution studies to the spending side of government operations.
No current studies of either tax or spending distributions answer the question of which households "benefit" from government taxes and spending. Just as one dollar of government spending may benefit Karen more than Lisa, one dollar of taxes may hurt Joe more than Bob. To do so would require detailed knowledge of what economists call the "utility functions" for each household, which are not available to researchers. As a result, our study only measures dollars of tax burdens and dollars supplied back to households in spending.
Are those government spending totals really money that households can spend?
Yes and no. There are different kinds of government spending. Some spending can be easily spent or turned into cash, and some cannot be spent. In our study about 42 percent of government spending is counted as transfer payments. This includes transfers of pure cash to households like Social Security payments, as well as in-kind spending such as Medicare and Medicaid payments that are economically very similar to transfers as they represent government payments for items that would otherwise have to be purchased out of households' own resources. In this sense, households can "spend" the amount of government spending they receive in the form of transfer payments.
However, many other kinds of government spending cannot readily be spent by households. For example, a household's share of government spending on national defense, roads or police protection can't be easily traded away for an equivalent amount of money. For these types of non-transfer spending, it is not appropriate to consider these items as "cash income" to households. In our study, these only represent dollars spent on households by the state and do not represent actual additions of income to households. As a result, it is important to make a distinction between what our study refers to as "fiscal redistribution" and actual cash income redistribution in the sense of dollars being transferred directly between households by government. Our study only measures the former.
Throughout the study, we are clear to point out the distinction between these different types of government spending: transfers, private goods, quasi-private goods and public goods. While only some of them can be considered "cash in-hand" to households, we believe that it is important to count all government spending in studies like ours. Ignoring government spending on non-cash items can lead to a significant understatement of the amount government spends on households each year. Additionally, if our study excluded spending on things like defense, roads and police protection and only counted transfer payments, the resulting distribution of government spending would be dramatically more progressive than we find in our study, giving a misleading impression of how lawmakers spend tax dollars throughout society.
Much of what we identify as fiscal redistribution in our study cannot in fact be spent by households. Instead, the phrase "fiscal redistribution" is simply used to represent an inequality between a household's tax burden and the amount government spent on them in return. It is important to be cautious when interpreting the figures from our study as representing pure redistribution of income. Our study only measures the more limited phenomenon of a household being supplied with more government spending than it pays in taxes. It does not measure actual changes in the distribution of income due to government fiscal policy.
Aren't your findings just an illusion caused by Medicare and Social Security that go to low-income seniors?
No. Social Security and Medicare have an impact in our study's findings, but the impact is not large enough to change our basic conclusions when those items are removed from the study.
Transfer programs like Social Security and Medicare result in large tax payments in youth and large government spending in old age. In our study they are treated as taxes and spending in the current year only. However, for the sake of transparency we show how these programs affect our results in two ways in Appendix A.
First, we show our results excluding Social Security, Medicare and payroll taxes. This has a small effect on our results. Second, we control for age and show taxes and government spending for income groups within each age group. We find that even as households move through the life cycle from youth to old age there is substantial fiscal redistribution between income groups at every age.
Do government deficits and surpluses affect your numbers?
Yes. Our study doesn't require that government spending exactly equal taxes in any year. For example, in 2004 the federal government ran a large deficit while in 2000 it ran a surplus. During deficit years, it's possible for everyone to receive more spending than they paid in taxes. In surplus years everyone's taxes can exceed the amount of government spending received. In Appendix A, we show how our results change if we force government spending and taxes into balance instead.
How do you treat "public goods" like national defense in the study?
Some kinds of government spending are supplied equally to all households. That is, once government provides things like national defense they can't easily stop others from using them, and one person's use of them doesn't reduce how much is left for others. Economists call these things "public goods." In our study they include national defense, environmental protection, public health spending and a few others.
In our study we distribute public goods equally to households since this is the way they're supplied by governments. Some argue wealthier households "benefit" more from these services. However, while that may or may not be true, our study does not attempt to answer the question of who "benefits" from government spending as noted above. It only tells us how much government spends on our behalf, which is a substantially different question.
In the interest of transparency, in Appendix A we show what happens to our results if public goods are allocated mostly to wealthy households instead of equally to all. We find that doing so doesn't change the basic conclusions of our study.
Doesn't government just waste all our money?
Some argue it is unrealistic to say that government provides anything valuable to people. That is, some argue that nearly all government spending is simply wasted or is supplied only to political favorites, shareholders of government contractors, politicians and government employees.
Our study does not assume this. While government waste exists, the only way a study like ours can proceed is to assume households pay taxes for a reason. We assume the basic role of government in society is to use valuable tax dollars to supply useful things back to taxpayers. Although this is sometimes not the case, it is the approach adopted by nearly every previous study in the fiscal incidence literature.
Join the Tax Foundation's fight for sound tax policy Go
Tax Policy Blog
The official weblog of the Tax Foundation.
Tax By State
For information on your state, select it from the drop-down menu.
Ask a Tax Expert
Contact information for Tax Foundation policy staff Ask