Why Do We Tax?

As the US presidential election bears down for November, it's prime time to ask how the income tax system could be improved. Assistant Professor Matthew C. Weinzierl suggests how.
by Martha Lagace

There is a mismatch between what many scholars assume is the purpose of taxes and what most people believe for themselves. That mismatch means that the advice experts offer to policymakers and citizens debating American tax policy can sometimes seem to miss the point. We could all benefit from fixing this gap.

For 40 years, economists have drawn from the well of Utilitarian theory—which has the goal of maximizing overall well-being in society—to help design tax systems in the United States and around the world. Although the theory works brilliantly in the laboratory of ideas, in reality its implications are a mixed bag. While taxpayers and policymakers endorse some aspects, others are widely rejected.

“My paper is about revising economists' dominant approach to evaluating taxes so that it reflects this messier reality.”

For example, should we tax the tall more than the short? Standard theory recommends substantial "tagging" or tailoring of taxes to personal characteristics (such as height) that are, on average, linked to wages. In the real world, this is seen as a terrible idea, except for some select traits, such as blindness, for which the United States has a tax exemption.

Lawmakers, following public opinion rather than scholars' theories, have put in place very little tagging. Does this mean it's time to bury the Utilitarian approach?

Not quite, says economist Matthew C. Weinzierl. The Harvard Business School assistant professor has a better plan for research and policy: expand the theory to include the diverse criteria for taxes that most people find compelling.

Weinzierl's research shows that a theory with a more realistic perspective on what people believe makes for good tax policy, can make up for the limited use of tagging in the United States. The key is that this policy be designed not just to meet Utilitarian goals but also to align with the principle of Equal Sacrifice.

The principle of Equal Sacrifice declares that the goal of tax policy is to share the costs of funding public goods evenly across all people. This goal was a staple of tax theory through the 1960s. Combining Equal Sacrifice with Utilitarianism better represents the preferences that people hold in the population at large, according to Weinzierl, and makes the theory's recommendations align much better with the reality of tax policy.

Weinzierl lays out the case for this revision in a new paper, Why Do We Redistribute So Much but Tag So Little? Normative Diversity, Equal Sacrifice and Optimal Taxation.

The fundamental challenge, he writes, is that "different people find different criteria compelling, and most people find multiple criteria partially compelling. Specifically, not all people are Utilitarians, and perhaps more important, most people are not all Utilitarian."

We asked him to break it down for overtaxed managers.

Martha Lagace: You want theory and policy to better agree. What is the basic problem keeping them apart?

Matthew Weinzierl: To get at the basic problem, it's easiest to consider this question: Why do we tax? Different people give different answers, and most of us tend to think of a number of answers ourselves.

Economists who study taxation, however, have assumed—with few exceptions—a single answer: to maximize overall well-being in society. We can call that goal the "Utilitarian criterion" after the philosophical framework that supports it. Since pioneering work in the 1970s by Nobel laureate James Mirrlees, the Utilitarian criterion has dominated tax research. It has had enormous impact on policymaking around the world.

The Utilitarian criterion is a very reasonable, perhaps the most reasonable, goal. But it is not the only one that matters to people.

To prove this point, all we need do is examine one aspect of the tax policy we would use if we were fully convinced Utilitarians: people should pay different taxes depending on their gender, height, race, and other personal characteristics. Economists call these characteristics "tags."

The Utilitarian logic for taxing tags is that these traits are both hard to change and systematically related to individuals' innate talents and abilities. This combination is the holy grail of Utilitarian taxes. Why? To a Utilitarian, people with greater innate ability find it easier to earn income, but all people equally enjoy having money to spend. To maximize overall well-being, we should require those with high ability to work harder than those with low ability and then, to fully redistribute the income so that everyone has the same money to spend. Of course, this would eliminate the incentives for anyone to work, so we can't achieve this "ideal" policy.

But tags can help. We know that tall people are, for example, higher earners than short people on average. It's hard to change your height, so we can redistribute income without hurting incentives by using tags. My research with [Harvard economist] Greg Mankiw, which has also been profiled in a Working Knowledge Q&A, develops this point in detail.

Q: Your new paper notes that just about everyone rejects the idea of tagging, except in special circumstances. What alternative do you propose?

A: Tagging is such a central implication of the standard model that, as tax scholars, we really must be able to explain its general rejection in real-world tax systems. If we can't explain why we don't tag, we would have to explain why our entire approach shouldn't be jettisoned. My revision to the standard theory allows us to reject tagging without taking this drastic step.

The key idea is that the Utilitarian criterion is not the only one that people find relevant to what makes for good tax policy. I argue that one alternative criterion, Equal Sacrifice, is a plausible candidate for a criterion that people use in combination with the Utilitarian criterion to judge tax policy. Incorporating a role for Equal Sacrifice can explain why tagging is used in only limited ways, but it can do so without rejecting the progressive taxation that otherwise might seem inconsistent with a rejection of tags. It also has a number of additional implications for policy, each of which improves the match between theory and reality.

Q: As you've said, tagging means the tailoring of taxes to personal characteristics. To what extent does the US tax code tag?

A: We do some tagging, but tagging is severely limited in reality relative to what the standard theory would say. Theory suggests we should tag height, gender, race, facial symmetry, place in birth order, native language, parental traits, macroeconomic conditions at age 18, and so on—anything that is hard to change and correlated to income-earning ability. We use none of these.

Tagging is not totally absent, however. In the United States, taxes and transfers depend on old age, the number of young children in a household, disability, and blindness. All four of these characteristics have something in common: they are very strong predictors of lower ability to earn income, at least per person in the household. This fact is clear for the last three traits.

But even old age, which may seem less clear, proves the point. When age-dependent programs such as Social Security or Medicare were initiated, they kicked in at an age when work had become difficult for most people. They were intended, in fact, to prevent poverty among the elderly. These days, 65 doesn't seem nearly so old. Productive work at that age is now quite possible for many people. Consistent with that change, we are now debating whether we should raise the age at which these transfers kick in.

We can learn a lot from the fact that the tags we use in our tax code are so closely related to having lower income-earning ability. When we give a tax break to blind people, we can be pretty sure that they have trouble earning income. Similarly with the elderly, children, and the disabled. In contrast, giving a tax break to short people because they have lower ability on average than tall people would give tax breaks to many highly able short people and force higher taxes on many less able tall people.

To the extent that those "errors" bother us—and they do bother us if we care about the principle of Equal Sacrifice—we would want to tag only those characteristics closely related to income-earning ability.

Q: Could the tax code tag other personal characteristics?

A: Tags have been suggested for age, gender, height, and race. The most provocative, however, is for genetics. Once we know and understand genetic information better, we could, in principle, do a very good job of inferring individuals' innate abilities from the genetic code. Of course, most of your readers would probably, and understandably, balk at taxing genetic information. I think it's very unlikely that any real-world tax code will ever tax personal traits to a large extent.

Q: You argue that research should include a role for the principle of Equal Sacrifice. Why?

A: Most people, whether from the right or the left of the political spectrum, are likely to see at least some intuitive appeal to the principle of Equal Sacrifice. It says that the goal of tax policy is to spread the costs of public goods so that each individual's "sacrifice" from paying taxes, which depends on his or her ability to earn income, is the same. Equal Sacrifice got its start with the famous philosopher John Stuart Mill 150 years ago. It held enormous sway for nearly a century among leading economists and tax scholars.

The basic appeal of Equal Sacrifice is that it rejects the assumption implicit in the Utilitarian approach that, as summarized in a critique by [Harvard economist] Martin Feldstein, "all property and individual abilities should be regarded as society's common resource." Instead, Equal Sacrifice assumes individuals have the first claim to their output, and that they voluntarily agree to form societies that collect taxes in order to purchase public goods.

Equal Sacrifice has many implications for tax policy. For instance, it says that everyone should pay some taxes, so it makes no room for redistributing income from one person to another. But because having higher ability makes it easier to earn the money to pay taxes, Equal Sacrifice is generally consistent with progressive taxation.

“"Equal Sacrifice is a reasonable way to capture one component of how many if not most of us judge tax policy.”

For our purposes, an important implication of Equal Sacrifice is that tagging is a very bad idea. Taxing a tall person more than a short person with the same ability will cause the tall person to bear more sacrifice than the short. A Utilitarian doesn't care about this unequal sacrifice, but it strikes most people as unfair. If we include a role for Equal Sacrifice in how we judge tax policy, then we can explain why tagging is limited in reality.

I want to be clear that I am not arguing that Equal Sacrifice is a better criterion than Utilitarianism. Instead, I think that Equal Sacrifice is a reasonable way to capture one component of how many if not most of us judge tax policy. A huge body of research, mostly by political scientists, has shown that people are ambivalent about how to judge policies such as taxes. Even people who generally support progressive economic policy usually see some appeal in principles like Equal Sacrifice.

My current research is about revising economists' dominant approach to evaluating taxes so that it reflects this messier reality. As I summarize it in the paper:

"This paper argues that the puzzle of limited tagging is a symptom of a more fundamental problem: conventional optimal tax theory evaluates policy based on a criterion that is unrealistically narrow. That criterion, Utilitarianism, is powerful and compelling to many. But extensive evidence has shown that no single criterion, however appealing, can claim to be the criterion of the tax design problem as society perceives it. Different people find different criteria compelling, and most people find multiple criteria partially compelling. Specifically, not all people are Utilitarians, and perhaps more important, most people are not all Utilitarian."

Q: The word "justice" occurs often in the paper's bibliography. Do notions of justice affect US tax policy?

A: The clearest example of ideas of justice mattering for current tax policy is the "Buffett Rule" debate. The Buffett Rule would ensure that people with substantial incomes from investments paid average tax rates at least as high as middle-class workers.

Both sides of this debate think they have the upper hand in terms of justice. How can that be?

President Obama has supported the Buffett Rule because he knows that it seems "fair" to most people. Individuals living off of investment income are surely not more deserving than workers, the argument goes, and they should therefore have to pay at least as much in taxes. Especially at a time when wealth and incomes are increasingly concentrated among very few, it is unfair to let the richest pay so little.

Republicans, to put it mildly, disagree. They argue that taxing those with large capital incomes is discriminating against people who chose to save their incomes rather than spend them. If society wants to use progressive taxation—in which average tax rates rise with income—they believe it should do so on income earned from work. Taxing investment income is taxing that same income again, but only on those who chose to save it.

Issues of justice such as these are at the heart of tax policy. Of course, these issues have been faced by human society forever, and resolving them with a definitive answer is impossible.

My research, rather than trying to decide which idea of justice is "right," seeks to capture the diversity of individual beliefs about justice. Only by doing so will we be able to truly understand the policy we have and how we might improve it.

Q: What are you working on now?

A: To add evidence to my argument about people using a variety of criteria to judge tax policy, I am running surveys in which people are given choices among a range of tax policies. Depending on what they choose, I will learn how much importance they assign to different principles of tax design. I hope this research, by making more relevant the theory that we tax scholars use to understand and advise on tax policy, will lead to a better tax system for all of us.

About the Author

Martha Lagace is senior editor of Working Knowledge.

    • David Bley
    • Unemployed, Not my strong suit
    I commend your solution of equal sacrifice. But isn't the concept of equal benefit implied also?

    As far as the Buffett Rule goes, there is nothing that prevents Mr. Buffett from making a voluntary contribution to the US Government in any amount he feels would right the inequity.
    • Kapil Kumar Sopory
    • Company Secretary, SMEC(India) Private Limited
    Tax is the key source of revenue for the government which needs money to provide various facilities/benefits to the people. There are very many types of direct and indirect taxes in India and these seem to be a great burden to the common man when he observes misuse of tax monies as the services expected to be provided are conspicuously absent. Cost overruns due to delays and negligence, corruption, etc. are covered by designing new taxes and raising rates of taxation making the have-notes poorer day by day.
    People would be willing to pay taxes if there is no mismatch between the amounts shelled out and the quantity and quality of benefits provided.
    Tagging is necessary to even the gulf between those with better earning and the less privileged. Such a sacrifice is in the overall interests of the society at large.
    • Anonymous
    On April 25, 1938, the Supreme Court overturned the standing precedents of the prior 150 years concerning "COMMON LAW" in the federal government.


    Since the Erie Decision, no cases are allowed to be cited that are prior to 1938. There can be no mixing of the old law with the new law. This is no longer taught for obvious reasons.

    The Sixteenth Amendment gave Congress no power to tax, full stop.

    It (Congress) had that power before. All the Amendment did was to remove the need for apportionment among the several States.

    Where are all the CONLAW people? Read what it says,don't spin it into something it isn't. It was written by men with grade school educations, it is not rocket science.
    • Anonymous
    If you are going to only look at the "income tax" I think you may miss the mark, of why we have tax. Why not look also at the idea of a consumption tax as well. In my opinion it is the most fair form of taxation there is.
    • Anonymous
    You had me for a while when you were talking about shared sacrifice, it seems like a good argument for a flat tax. Why you feel it is a good basis for a progressive tax escapes me.
    • Anonymous
    Any discussion of the Buffet rule for individuals is flawed if it does not look at double taxation and also account for taxes already paid on the same income. If you make the argument for dividends by adding together the taxes the company paid on the dividend income and then the tax the individual pays then you may have a point, if you only talk about the rate the individual pays then you are missing the point.
    • Jim Peers MBA '51
    • retired
    Am really disappointed in this article. Read "Trickle Down theory and Tax Cuts for the rich" - posted by the Hoover Institute this week
    • Ted Gutelius
    Instead of making deductible contributions to Bill Gates, one of the richest men, and taking a write off for that transfer so that he can lower his taxes, why doesn't Warren just give it all to the government? Is he concerned the government might be a bit wasteful?
    • Eric
    • Attorney, The Marks Law Firm
    The trouble (or one of the troubles) with something like the Buffet Rule is that, in addition to double-taxing monies previously earned (as you discussed), it also disincentives investment. Rules like this one discourage people from investing because they're potential reward is being limited while their risk, largely, is not. What would this do to the economy? How many successful business, large and small, got their start thanks to the investment of others? I find this equally compelling as the double taxation argument, and probably a more practical concern. The double taxation argument only directly impacts the investor. The reduction of investment altogether directly impacts many others, in addition to the investor.
    • Bob M
    • AVP, Not Important
    Well, I thought this article had promise but Weinzierl quickly lost me for good here:

    "To maximize overall well-being, we should require those with high ability to work harder than those with low ability and then, to fully redistribute the income so that everyone has the same money to spend. Of course, this would eliminate the incentives for anyone to work, so we can't achieve this "ideal" policy. But tags can help. We know that tall people are, for example, higher earners than short people on average. It's hard to change your height, so we can redistribute income without hurting incentives by using tags."

    Someone, please pull the plug on Weinzierl's research project before he completes his policy work and creates the potential that someone might actually listen to him.
    • Chris
    RE Equal Benefit

    When considering some social services and government run institutions (such as roads, water/sewer, military) the concept of equal benefit can easily lead one to think of equal taxation in terms of dollars.

    Personally, I think you have to extend the concept of equal benefit to consider other aspects of society such as education, living in a stable society etc. Most of those that 'have' don't seem to understand that that structure of society has been a key factor in allowing them to 'have'.

    I am reminded of the expression "If you can make it in NY you can make it anywhere". Really? Try 'making it' in Kabul or Guadalajara.
    • Clark Phippen
    • Retiree, HBS '62
    Yikes! The simple question: "Why Do We Tax?" has an equally simple answer - to run the government. We should NEVER attach other reasons like income redistribution, but, of course, we do. The debate ought to be strictly on what the government needs the money for, and debates on that appropriately abound. Tax policy should be (1) to collect monies equal to needed governmental expenditures; (2) in an open, SIMPLE, transparent manner (as a retiree with modest savings I should not need to file 40 to 50 pages each year); in a way that (3) is fair and as close to a straightforward constant or graduated percentage of income as possible. "Equal sacrifice" is a frightening term; I would suggest "equal relative share".
    • Hugh Quick
    • home, none
    "All people equally enjoy having money to spend". That is obviously untrue. All people are different and specifically are very definitely different in their enjoyment of spending money.
    • Edward Stern
    • Researcher
    The author says that Republicans ...argue that taxing those with large capital incomes discriminates against people who chose to save their incomes rather than spend them. But not everyone with large investment income earned their capital. Many inherited it. What would be a fair tax rate on an income from a $20 million inheritance? And what portion of investment incomes over $25,000 is from inheritances? That would be a good study.
    • Anonymous
    This is a very interesting article. However, no matter how you look at tax, if the public is not confident that the tax monies are wisely and honestly used and that fraud is closely monitored and government manpower is minimal, then the first question should be "How can we use our tax funds more efficiently?".
    • Anonymous
    I commend you on taking on such a political topic during an election year. I appreciate the thoughtful approach in thinking through tax policy. You are certainly hitting on a key aspect of how policy is formed - what real people think is fair. More than any time before, I believe this will be a leading factor as Congress debates and legislates tax reform in the coming months. Look forward to the next piece on this topic.
    • Anonymous
    Interesting views on the bases of taxation and the division such bases create between the ideal and the real.

    I would have thought that taxation is all about the projection of power and what constitutes that power. Any recourse to utilitarianism or an equal sacrifice addendum is all about masking what taxation is really all about to make it palatable.

    Of course, over time, the signficant decline in the use of force for the purpose of collecting taxes has narrowed the gap between the "face of reality" and its mask. Unfortunately, that such a gap exists, remains.

    I would be interested in what your research would show if, at an extreme, taxation stopped. No doubt, a pathological case unlikely to ever see the light of day, but neverthless instructive I suspect.
    • Anonymous
    It seems amazing that one can build an article on such flawed premises. The author sets up the 'straw man' of 'utilitarianism and tagging' as a basis for discussion, and on which to reject progressive taxation.
    The reason people reject tagging may have nothing to do with its fairness or otherwise. It is just a bad policy to tax a correlate of income, when these correlates are extremely weak predictors [on an individual level], and taxing actual REALISED income is much more direct.
    It is simply flawed logic to suggest people reject progressive taxation because they would reject the notion of tagging.
    • Anonymous
    Taxing means nothing if spending is not on the horizon. The author's normative framework is therefore incomplete, making publication / interview quite premature. He would profit from following his own advice (footnote 3 to the Working Paper), and actually reading Rawls and other normative ethics scholars' take on tax policy. Their discussions on "why do we tax?" shed much more light on tax policy goals than the author's weak correlations, lack of focus, and half-baked attempt to combine Utilitarianism with egalitarian prescriptions.
    • Cass Apple
    • Retired
    Thinking of "tagging" a bit differently, consider that some taxes are levied to pay for specific purposes. Consider the Medicare Payroll Tax and the Medicare Part B Tax or Fee.
    If the tag were an individual's annually-determined BMI, such that the Tax or Fee rates were adjusted downward for those of normal BMI (in no more than expected value of the government's cost savings from better health), then the tag, which is under an individual's control, would reinforce actions to lower the government's total costs for all health care.
    I do not propose that this specific suggestion has wider application than to health care related taxes (or insurance policies) but that considerations of the effect of taxes on behavior might be included in a broad theory of taxation.