Arnold Kling  

Income Distribution Stories

PRINT
Comment of the Week, 2003-06-2... Perspectives on Social Securit...

This week, the topic of income distribution made news. Several news outlets highlighted a report from the IRS on the 400 highest-income taxpayers.


annual samples of individual income tax returns for Tax Years 1992 through 2000 were sorted by AGI, and the 400 returns with the highest AGI in each year identified.

Using 1990 dollars, the income found in the top 400 returns rose from a total of $17.4 billion in 1992 to $52.8 billion in 2000. Capital gains income accounted for much of this rise, going from $6.3 billion in 1992 to $37.9 billion in 2000.

'Jane Galt' pointed to an NBER working paper on the distribution of consumption. This description says,


By one measure, inequality of after tax labor income has increased by 25 percent in the 1972-98 period. Yet consumption inequality has risen less than 2 percent. What's happened, the authors explain, is that higher-income Americans have been saving more of their income.

For Discussion. Consider two households, each with $40,000 in labor income. The households are headed by young adults, age 25. The first household saves nothing. The second household saves 20 percent of its income each year. This difference in saving will affect income and wealth over the next fifty years, and it will result in differences in taxes paid over that span. What tax policy would you consider "fair" between these two households, and how do you think our actual tax policy would compare with this fair policy?


Comments and Sharing


CATEGORIES: Income Distribution



COMMENTS (24 to date)
Bernard Yomtov writes:

Part 2:

Consider two households, each with $40,000 in total income. the gouseholds are headed by young adults, age 25. The first household derives all its income from labor. The second does no labor, and its income consists of dividends on inherited wealth.

What tax policy would you consider "fair" between these two households, and do you think the recent changes in the tax law moved us closer to this fair policy?

David Thomson writes:

“The first household saves nothing. The second household saves 20 percent of its income each year.”

Common sense dictates that we should enact tax policies which encourage a prudent lifestyle. What about those households who, through no fault of their own, are unable to save? These people are the exception and not the norm. There are far too many laws passed premised upon the notion of victimhood. The reality is that most people are responsible for their own predicament. We also dare not inadvertently convey the message that those families saving 20% of their annual income are suckers for doing so.

David Thomson writes:

"The second does no labor, and its income consists of dividends on inherited wealth."

Once again, why should we be focussing on the exceptions to the general rule? I doubt very much if even 5% of Americans are living solely on inherited wealth. Moreover, many of the sons and daughters of the very affluent end up working hard. Few become complete wastrels and self centered jerks.

Bernard Yomtov writes:

"I doubt very much if even 5% of Americans are living solely on inherited wealth."

I share those doubts. I also doubt that very many households with $40,000 in labor income save $8,000, unless they have substantial other sources as well.

So we have a duel of unlikely scenarios. My point is that Arnold's framing of the question was hardly neutral. It was designed to elicit the answer he prefers.

Don Lloyd writes:

The criterion of 'fair' is NOT an appropriate way to choose tax policy.

Tax policy SHOULD be judged on its effectiveness in helping and allowing the creation and broad distribution of NEW wealth, resulting in widespread improvements in a standard of living measured by a real economic freedom to choose from among both quantities and varieties of goods and services to consume.

This would likely mean no taxes on earnings or production or investment of any kind, and no politically motivated monetary supply inflation that prevents the benefits of improved productivity in specific areas from generally diffusing throughout the economy by means of lower prices.

It conceivably could mean a retail sales tax on CURRENT consumption of NEW products, with an effective beginning annual threshold set by fixed pre-funds or refunds or credits.

The CURRENT consumption qualification would mean that the tax on a new durable good of 5 years of projected service would be discounted by 80%.

Regards, Don

Arnold Kling writes:

I'll answer Bernard's question about the household that lives on inherited wealth vs. the household that has to work.

I think that a consumption tax would be the most fair. If I thought that we lived in a society where everyone who does not inherit a fortune lived in poverty, I might think differently. But I don't think that confiscating inherited wealth is a better form of taxation than a consumption tax.

Bernard Yomtov writes:

Arnold,

I'm not suggesting we confiscate inherited wealth. I'm suggesting we recognize its importance, and more bradly the general importance of family wealth, even when it's not massive.

Take your example: two households with $40,000 in labor income. It is not uncommon that one would receive significant assistance from parents. I'm not talking about million-dollar bequests. I'm talking about getting a decent used car free when your parents buy a new one, or some furniture. I'm talkig baoutthe occasional cash gift, the help with a down payment on a house, the contribution to the kid's college fund. I'm talking about having smaller educational loans to pay back.

These are commonplace advantages, and go far towards enabling one household to save and the other not to. Of course this represents admirable generosity on the part of parents, but let's not pretend that the beneficiaries are particularly virtuous for accepting it.

Your proposed consumption tax needs to be much better defined before it can be sensibly discussed. For example, exactly what would you tax, and would you exempt anything? How would you make the transition?
And why would it be simpler than the income tax?

Trudy thinks that with a $40,000 in household income there is no 20% left to save, and little time to study tax law and ponder "should I save the money I don't have or not?"

Tax law doesn't influence behavior at all, other than that of rich people.

Trudy the Monkey

dsquared writes:

Arnold's example; tax labour income @ 25%, income from savings @25% and capital gains @25% when realised.

Bernard's example: The second family would have to work for a living, as I've always supported 100% inheritance taxes (with perhaps exemptions for a family house) and nobody's ever managed to convince me that their reasons for not agreeing aren't self-serving.

Mcwop writes:

D2,
Are you suggesting that politicians don't have self-serving reasons for keeping the estate tax? Most sides have self-serving reasons on the issue of estate taxes.

I look at the estate tax and inheritance this way. What happens to the money? If the government gets it they spend it. If an individual keeps it, they spend it and/or invest it and/or give it away. There are economic/social benefits in all of these situations. The real question is which situation produces the most economic and social benefit. Nobody's ever managed to prove to me which situation wins in the scoring.

Eric Krieg writes:

20% savings rate? That's easy.

Start out by maxing out your 401K.

Next you've got the Roth IRA. Max that sucker out.

A portion of your mortgage is savings, albeit a small portion.

I make much more than $40,000, but I do live in Chicago. I make about 15 grand less than the median income here. My wife doesn't work. I have 2 kids. I own 2 recent vintage vehicles outright.

And I save a little less than 30% of my income.

Saving is a lifestyle, just as is conspicuous consumption. I have NEVER carried over a credit card balance, for example. The credit card gets paid off every month. I drive used Chevies, not new Bimmers. When I do splurge on something, it is generally something for the house, at the Home Depot, because in my mind that is an investment.

We already have tax policy that encourages people to save (401Ks, Roth IRAs, 529, etc.). Yet people still don't save. Is a consumption tax going to make any difference? Do states with high sales taxes have less consumption and more savings, for example?

This is much more a cultural issue than an economic one. We're a consumerist society, much more so than in the past, much more so than other countries (56 % of German GDP is due to consumer spending vs. 70% of US GDP, I think I read that in BusinessWeek recently). It took much more than tax policy to get us here, and it is going to take much more than tax policy to change our bad habits.

JT writes:

Dsqaured: a 100% estate tax would simply create a big industry creating loopholes to evade these taxes. You should know better: extraordinary levels of taxation create extraordinary efforts to evade those taxes.

Mr Hyde writes:

"I've always supported 100% inheritance taxes (with perhaps exemptions for a family house) and nobody's ever managed to convince me that their reasons for not agreeing aren't self-serving."

Gee d-squared any money I save to provide for my widow and orphaned children (money I have already paid tax on) is confiscated – just when they have lost their breadwinner.

That’s great if you want everyone dependent on the government.

One of my prime reasons for working (and for taking out life insurance) is to see that those who are dependent on me will be taken care of after I am gone.

I think destroying the incentive to provide for your family is morally and economically wrong. In effect it abolishes the family as a decision-making unit. It will reduce productivity and undermine the virtues and attitudes that create prosperity and make a free society possible.

Tom writes:

>Consider two households, each with
>$40,000 in total income. the households
>are headed by young adults, age 25. The
>first household derives all its income
>from labor. The second does no labor,
>and its income consists of dividends
>on inherited wealth.
>
>What tax policy would you consider "fair"
>between these two households...?

I'll play. A "fair" tax policy would tax the income of the two families by the same dollar amount. (As an aside, neither would pay Social Security taxes, but both would be subject to the same mandatory retirement savings law. Since the dividends household clearly has much larger savings, the effect of the mandatory retirement savings law on the $40000 of income of each of the two households may be different.) I am sympathetic to the efficiency argument that says the household relying on dividend income should have its tax taken out as corporate income taxes paid by the dividend-paying companies, with no personal income tax on dividends. I am more sympathetic, however, to the transparency argument that says the companies should deduct dividends from their corporate incomes before paying corporate income taxes, and that the dividend household should fill out an identical income tax form as the labor household (with dividend and labor income going into the same line).

Eric Krieg writes:

Those of you in favor of the estate tax should check out this article:

http://www.techcentralstation.com/1051/techwrapper.jsp?PID=1051-250&CID=1051-063003C

The estate tax is probably the tax that is the most larcenous.

Bernard Yomtov writes:

Tom,

Thanks for the answer. Let's complicate the question a bit. Should they still pay the same if there are substantial differences between them?

Say one household consists of a single individual and the other is a family of four. Or suppose one has large medical expenses during the year. What then?

Jim Glass writes:

To have a "fair" tax code people need to agree on a narrow, clear definition of "fairness". If people insist on comparing their different subjective impressions it'll never happen.

Take the simplest of cases, a married couple no children. In principle you can tax them as two single people, multiplying all the tax bracket amounts by two, so they are taxed at bracket rates as two single people with equal income. That at first may seem fair, but it ignores a basic fact about marriage: two people live together more cheaply than two live apart and consequently have more disposable income on the same gross income. The Census estimates a couple on average live together at a cost of about 1.4 single people -- which means marriage adds about 30% of gross income to disposable income per person, which is a lot. So this option violates the most basic precept of the progressive income tax, that it should be imposed by *ability to pay*. That is unfair!

To remedy this we can try an option (b) that adds a couple's income together, divides by 2 and taxes at the rate paid by a single person with 40% more income than that. So a couple with $14x of income pays as much tax as two single people with a combined $20x. But this creates a "marriage penalty". It gives many people an incentive to not marry. Moreover, if the second spouse takes a job or gets a better-paying one, the marginal income gets pushed into a top tax bracket -- and a 'punitive' higher tax rate is incurred for working compared to a single person taking the same job, just because one is married rather than single. That is unfair! For those who overcome the deterrent to marry this is a deterrent to work, and to improve the financial situation of their families. That doesn't seem right, and the solution to that is (a).

There is no resolution to this conflict because "fairness" here is being judged by two different conflicting standards both of which are reasonable. Whichever option one chooses reasonable people can object and say the other is more fair -- and be correct.

So in practice the choice is driven by the politics of winners and losers, and with the growing numbers of voters in two-income families in recent years the US tax code has been gradually moving from (b) to (a).

But if this simplest choice between (a) and (b) can't be logically resolved, nothing can. Should child costs be deductible? Which ones? Medical bills? How much of them? If mortgage interest is, why not rent? Why deductions, why not tax credits? With no consistent logic it all becomes just politics and vote buying. The simplest single thing in the entire US tax code probably is the cents-per-mile deduction for driving your car -- and there are three different deduction rates for business, medical and charitable driving. What's fair about that? ;-)

I've been working with taxes for a long time and the only simple, consistent standard of "fairness" that has ever looked remotely practical to me is, as our host suggests, consumption taxation -- even by "progressive" standards. Tax people by how much they consume.

After all, if one wants to tax the most well off by the most, the relevant measure of welfare is consumption, not income. There are many ways for many people to enjoy lavish consumption without having much (or any) income. But those people can't dodge consumption taxes. And if two people both have the income to live lavish lives -- with yachts and champagne and burning $100 bills to impress floozies -- but only one does so while the other saves and invests to create jobs today and the wealth needed to fund Social Security 30 years from now, is it really "progressive" to tax them both at the same rate because they have the same income? Or is it more progressive to tax the wealth spender more? And "progressiveness" can easily be increased by giving everyone a lump sum refund that is a larger portion of total consumption for the poor than the rich, and/or exempting from tax items that form a larger portion of the budgets of the poor than rich.

Moreover, accurately determining income from businesses and sophisticated investments can be very difficult even when everybody is honest about it, and for those who aren't there are great opportunities to game the system -- both of which present real issues of "unfairness" that aren't present in a consumption tax. And consumption taxation solves the problems of estate and gift taxes.

But all that said, funding government by consumption tax doesn't seem practical. Income taxes exist in all modern states for a reason, governments today are too big to be funded by consumption taxes alone. And adding a general consumption tax to an existing income tax doesn't look like it would necessarily be a beneficial thing.

Eric Krieg writes:

>>Income taxes exist in all modern states for a reason, governments today are too big to be funded by consumption taxes alone.

Jim Glass writes:

"That's always been my reaction to consumption taxes. They'd have to be so high ..."

Well, it's possible to have a 'backdoor' consumption tax by having an income tax that allows a current deduction for all savings, and then taxes all savings as income when they are cashed in. Income - savings = consumption, so the amount of consumption is what gets taxed.

Bush's first team of economic advisors, Hubbard & Co., was clearly pushing things in this direction with their proposal for greatly expanded tax-favored savings accounts as well as the dividend deduction. We'll see what Mankiw & Co. do now.

It certainly wouldn't be easy to get there from where we are now. Among other things you'd want a broad-based income tax with no deductions (good luck!) and there'd still be all the problems with calculating income and gaming the income tax system. But it is at least theoretically conceivable that we could get there from here, which it isn't with conventional consumption taxes.

Bernard Yomtov writes:

Jim,

I agree with much of what you say about "fairness." It's not an easy concept to pin down. (This by the way, is one reason I have little sympathy for arguments that the estate tax is unfair).

That said, I fail to see how a consumption tax, as it would actually work as opposed to sound bite descriptions, would be an improvement. You claim it would be progressive, but I don't see why. If you tax consumption at a fixed rate for everyone then by definition it's regressive on income, except for those who spend more than their income, who will tend to be low-income people.

And you have to define "consumption." It's easy to write equations, but not so easy to decide how to treat the cost of a house, or educational expenses, or even a car. As you point out, to define it as income minus savings requires defining income, the main source of the complexity and gamesmanship in the current system. You would also have to define savings - not trivial. It has to include a negative figure for borrowing. That is, the definition of consumption you want to use is really income plus borrowing minus savings.

The large number of people who now file very simple tax returns would have to file considerably more complex ones, to show net savings and borrowings.

If you try to collect the tax at the retail level you create, as Eric says, enormous incentives to cheat. Even at the sales tax rates that prevail now, we've all seen merchants fail to ring up a sale.

This is why I commented earlier that it is impossible to discuss a proposed consumption tax without a bit more detail as to what is being proposed.

Eric Krieg writes:

>>As you point out, to define it as income minus savings requires defining income, the main source of the complexity and gamesmanship in the current system.

Jim Glass writes:

"If you tax consumption at a fixed rate for everyone then by definition it's regressive on income,..."

But taxing on income is regressive by consumption. So which do you choose?

Consumption is the actual measure of economic welfare. People seek income only so they can consume. If you have wealth some other way you don't need income to consume. So income is only a proxy for welfare, and it can be a misleading one if the goal is to impose the highest taxes on the most well off.

Take a person who consumes $100k a year on lavish spending by tapping capital (obtained by inheritance, gift, lucky find of a treasure trove, savings, whatever) or by borrowing against capital, so he has no income. At the same time another person has $50k of income and saves mightily from it, so he consumes only say $35k. What's progressive about income taxation here, so the second person gets a tax bill while the first is tax exempt?

I would argue on principle that if we want to tax the most well off, consumption not income is what we want to tax. Consumption determines economic welfare and consumption is consumption, there's nothing misleading about it. It's a rational standard that people could agree on (though they won't). And if one wants to create a consumption tax that is more progressive than a flat-rate tax, I've pointed out there are easy ways to do it.

"...except for those who spend more than their income, who will tend to be low-income people."

But not *poor* people, necessarily. Remember the Fed's Survey of Consumer Finances says the bottom 5% by income average in the top 15% by wealth. (Lots of business owners having loss years, etc.) Income is not such a good proxy for "well off" as many assume. That should be kept in mind when people assume the income tax is the source of progressive taxation.

"... I commented earlier that it is impossible to discuss a proposed consumption tax without a bit more detail ..."

And I commented before that all the practical objections you raise are right, and any kind of "pure" national consumption tax of the sort I am discussing in principle is impossible in practice.

But it can pay to keep principles in mind when discussing tax issues generally. After all, an intellectually consistent and uniformly "fair" income tax system is also impossible in practice for many of the same reasons discussed here. ;-)

Eric Krieg writes:

Forget fairness. How about simplicity, ease of implementation, ease of collection, and minimization of unintended consequences.

The flat income tax might meet these conditions the best.

Bernard Yomtov writes:

Jim,

Taxing income is not necessarily regressive by consumption. It is in some cases, but not in general, and is progressive when consumption is positively correlated with income.

Your argument rests on two principles:

1. That “Consumption is the actual measure of economic welfare. People seek income only so they can consume.”

2. That “..consumption is consumption, there's nothing misleading about it. It's a rational standard that people could agree on..”

I don’t think either of these is quite as clearcut as you claim. Take the second first, just for confusion. The definition of consumption is by no means easy to arrive at. Sure, we can just write “C” in our equations and not worry about it, but in reality there are all sorts of problems: durable goods, education, job-related purchases, housing, medical expenses, charity, etc. all raise issues. Despite conventional assumptions, households make real (physical) investments. These look like consumption, but are they? I’m not at all convinced that consumption is in fact any easier to define than income for tax purposes.

Your first principle is also a stretch. It makes a psychological assumption about what people want that is unjustified. Some people want income so they can consume. Others are happier with less consumption and a healthier net worth. Why exactly does the first group have higher economic welfare than the second? The argument that the savers will ultimately spend their savings doesn’t apply, because often it is simply having the savings that generates utility. We might almost say that savings provide “option value” over and above what is shown on the account statement.

And of course there are other ways in which greater consumption does not mean greater “economic welfare.” Jack and Jill both have $50,000 incomes. Jack spends $40,000 and saves $10,000. Jill spends $40,000 in much the same way as Jack, but has $10,000 in medical expenses due to illness. Is she better off than Jack? Are we going to hand her a higher tax bill than we hand him? Fairness may be hard to define, but sometimes unfairness jumps right out at you.

We seek income because we want income, to dispose of as we see fit, and only part of that disposition is what is conventionally, and somewhat artificially, called consumption. In economic terms, our utility functions include all sorts of items. You are trying to segregate out certain ones as defining “economic welfare.” I don’t think that works.

Comments for this entry have been closed
Return to top