Friday, May 09, 2008

Equality Times Two

"West of the mountains nobody owes allegiance to anyone."
(from my Harald)

There are two senses in which people may believe in equality, one having to do with attitudes to people, one having to do with the distribution of income and other good things. The first can, I think, be illustrated with three examples: H.L. Mencken, my father, and the protagonist of my one published novel. All three are egalitarian in the first sense, none in the second.

In his published diary, Mencken mentions going to visit a friend who lived some distance from the railroad station. The friend sent his chauffeur to pick Mencken up. Mencken found that the (black) chauffeur, although uneducated, was an intelligent man, and had a very interesting conversation with him, finding out how the world looked from his point of view. On a second visit Mencken was looking forward to another conversation with the chauffeur. To his disappointment, there was a second guest, a white woman, and in her presence the chauffeur remained silent. Mencken was a famous, influential, and comfortably well off man. There is no hint in his account of the incident that he felt as though that made him superior to the chauffeur. What mattered was not race, income, education, or status but that the chauffeur was an intelligent person with something interesting to say.

In interactions with my father when I was growing up it was always clear that what mattered was who was right, who had the better argument, not who was older—status was simply irrelevant. Many years later I was shocked to hear an intelligent elderly man tell a child not to contradict his elders. From the point of view I had grown up in, the statement was not merely wrong, it was close to obscene.

The protagonist of my novel is an able, successful, famous man—from a society whose status hierarchy is very flat. He takes everyone he meets on the same level, treats a king, the grandson of an Emperor, and a farm boy more or less in the same way, allowing for the difference in their ages and a certain amount of prudence in dealing with the king.

All of this is egalitarianism—of a sort. But it says nothing at all about whether any of the people in question think incomes should be more equal, gender roles less well defined, or anything of the sort. It is an attitude towards other people, not a political or economic philosophy.

The connection with the other sort of egalitarianism comes, I think, from the belief that people don't see the world in the way I have described—a belief that is, of course, often correct. One of the motives for wanting a world that produces more nearly equal outcomes is the belief that your being richer than me makes you better than me, or at least that you will think so, hence that if incomes are not reasonably equal many people will be treated as inferiors.

My point is not that the motive is wrong. Inequalities of income can indeed lead to inequalities of status—although I suspect that people would find lots of other grounds for considering themselves better than other people even in a world where all incomes were equal. My point is rather that there is no necessary connection between being richer, or more famous, or better educated than someone else and thinking yourself his superior in any more fundamental sense. One can be a democrat in the first sense without being a Democrat in the second.

Or, of course, the other way around.


Clayton said...

I have been thinking about the focus on the income gap and how this actually contributes to inequality of the first sense. If the public discourse presupposes that the income gap is a measure of inequality of the first variety, then it becomes a self-fulfilling prophecy. By treating the income gap itself as a social ill, we are implicitly asserting that income is - or ought to be - an objective measure of social rank.

Your past blog post on social status was mind-blowing to me - while social status is not completely subjective, it's a good deal more subjective in free society than in societies with explicit class or rank. But even in free society, the more strongly we measure rank in terms of income or material possessions or beauty, etc., the more we (socially) impoverish ourselves by voluntarily restricting ourselves from measuring our own social rank in alternative terms.

jbm said...

Income distribution & égalité have always been an interesting issue, as someone who grew up in a contemporary liberal family and only started to challenge those views after studying economics (and particularly after reading Capitalism & Freedom, for that matter.)

I may be mistaken, but don't those who argue in favor of income distribution ultimately have to rely on subjective sense of morality to show that it's a problem? I've often wondered, after finding out that the "Immiseration of the Masses" was patently false, what was inherently wrong with uneven distribution of income. Oftentimes it seems that the only argument in favor of redistribution is that income inequality doesn't jibe with many people's concept of "justice" or "morality."

[Cue the cynicism]. It's tough to shake the feeling that, in trying to achieve income equality, many advocates of redistribution seem more interested in hurting the rich, regardless of whether their actions actually help the poor.

Anonymous said...

This is something I've been struggling with a bit lately. Hard life experiences have led me to the belief that much of talent and intelligence is genetic, which surely also means it's unearned. Therefore I'm fairly sure meritocracy, in its simplest form, does have an element of elitism in it. This realization troubles me a bit.

I'm not saying simple meritocracy is really equivalent to stupid status hierarchies. But neither can meritocracy claim to be all about desert. Tomorrow, you could get in a car accident, and it might take away your faculties, change you from beautiful to ugly, or change you from able to lame. Does that mean you suddenly deserve less?

David Friedman said...

"Does that mean you suddenly deserve less?"


I think we have two sets of moral intuitions, one about desert and one about (in Nozick's terms) entitlement. For a simple example, suppose we agree to bet a dollar on the flip of a coin and you win.

You don't deserve the dollar, since you didn't deserve to win the coin flip. But you are entitled to the dollar, since you acquired it in a legitimate way--a voluntary transaction with me.

One can generalize this distinction. I think of the "desert" approach as a "God's eye" view of the world--a view from above by an imaginary person capable of judging what everyone else deserves. I think of the "entitlement" approach as a way of judging things from the human level, where the judge is no more competent to judge what people deserve than anyone else is--but we can all, to a considerable degree, see what people are entitled to.

Anonymous said...


Those two perspective pretty nicely track with Thomas Sowell's distinction between the normal kind of justice (the same rules are applied to everyone) and cosmic justice (in which we try to make life outcomes fair).

One of the most important things we can do to encourage pursuit of the first kind of justice over the second, IMO, is to make it clear that your worth as a person, and your rights and responsibilities as a citizen, aren't based on your social or economic position, but on being a human and a citizen.

It also seems to me that a lot of the top of our society imposes a certain kind of hierarchy on people based on academic, financial, or career achievements. And this seems pretty unhealthy, and leads pretty directly to justifications for stuff like affirmative action programs in education and employment.

Aaron Krowne said...

Nice post.

It seems to me that an income/money-centric notion of classism/elitism is really quite stupid for a variety of basic reasons.

For one, money is just one sort of thing one can be compensated with. It is not actual "useful stuff," nor is it many other monies that could potentially be disbursed. The euro is stronger than the dollar, so is everyone in Europe superior to me because they earn Euros? The pound is stronger than both, so are Brits superior to all of us?

And if I managed to be paid in gold, would I be a veritable god among men?

In a way, that line of rhetorical questioning would immediately strike most people as silly. But that is part of my intent -- you can't "spend" euros, pounds, and gold at all in the United States (taking that as a point of reference). But these other currencies and monies do have a higher exchange power than the dollar, and arguably preserve their spending power much better.

I for one would be impressed with someone who was being paid in gold (especially if it were in fixed quantity). Most people wouldn't know what to think.

The upshot: in a sense, these dollar-substitutes are better than what we take for granted, and in a sense, they are not. So even money is relative and its value, and ability to confer "status" subjective.

Even if we hold fixed the kind of money all are paid in, who is to say a lot of it is such a great thing for everyone? Isn't the man who earns less but is happy and content much "wealthier" than the rich miser, in a very important sense? How about the man who works only for himself, just scraping by, versus the man getting paid handsomely to toil under a tyrant?

It is very likely in the real world the two would actually envy each other.

Finally, addressing the point of anonymous about innate (genetic) ability, I don't consider this a significant theoretical problem for the income distribution issue. People are born with all kinds of innate abilities and disabilities -- I will guarantee to you that everyone has both! It is a matter of many factors as to whether an individual turns their abilities into the level of income they desire; not just intelligence, but "EQ", mental health, experience, willingness to take risks, dedication, and so forth.

Earning a lot of money is really a value decision on the part of the earner, to a much greater degree than is typically acknowledged. Most people are happy to pine wistfully for great riches (and support policies that pilfer them from others), but not to really do anything it takes to attain them. This might be because that lifestyle wouldn't actually be to their liking.

Joel Davis said...

I agree with the sentiment behind the former category more than the latter even though I do consider myself a market socialist. That is actually how I reconcile the two, with the knowledge that (economically speaking) even Donald Trump isn't Donald Trump.

But I do get the feeling that right-libertarians do tend to (on the whole) downplay the role of the worker as a source of creativity as well as labor. With comments like with Ellis Wyatt "I am leaving it as I found it. Take over. It's yours" it seems as if the character is attributing all the success totally to himself and none to the people "underneath" him who probably made millions of creative decisions he wasn't the least aware of. You see it in real life with Trump et al as well. He makes a big deal about being "broke" and coming back, filling his ego as being some super-productive individual. When you hear him (or anyone in his immediate family for that matter) talk about it you can hear a certain tone that seems to imply (when they're polite about it) that he was able to start with less than others had and made more than others usually do. But that ignores the aforementioned worker creativity and the factors of knowledge and reputation that were skewed in his favor. I also don't doubt he believes his own propaganda. That's part of the reason I consider myself a socialist, because I see an individual's privilege pushing them ahead of those who have as much or more potential.

But like I said, the "we're all the same basic kind of talking apes" sentiment is why I believe a free market will (in the long term) level the income considerably.

So I guess that's a long-winded way of saying "the two categories aren't necessarily exclusive of each other unless you take the latter to the most extreme point."

jimbino said...

Ï'm fascinated by the effect that language has on the perception of intelligence and status.

I consider myself a brilliant graduate of the U of Chicago and, in the States, I judge the intelligence, education and status of people by the words they use.

But here in Brazil, most kids can speak the language better than I can, and I have to judge people's intelligence and social status by other cues. And most of them probably think I'm really dumb, since I can't say lots of simple things properly.

The result is that I do get to know and appreciate lots of kids, bartenders, bricklayers, mechanics, hookers and others of a social status I would consider "beneath" my social rank in the States.

All Amerikans need to get out more.

Micha Ghertner said...

David initially argues that under the first sense of egalitarianism, "what mattered was who was right, who had the better argument, not who was older—status was simply irrelevant." But then David proceeds to tell us that this same sense of egalitarianism says nothing at all about whether gender roles should be less well defined.

So age shouldn't matter but gender should? Is gender not also a form of social status, very much irrelevant to questions like who has the better argument?

David Friedman said...

Micha writes:

"But then David proceeds to tell us that this same sense of egalitarianism says nothing at all about whether gender roles should be less well defined."

You can have well defined gender roles in the sense of the expectation (for instance) that most women will be housewives, and still judge what women say on the basis of whether they have good arguments, not whether the arguments are made by women.

Similarly, when I was growing up, there were well defined age roles--I wasn't expected to work to support the family, was expected to go to school. That didn't imply that arguments I made got discounted because I was not an adult.

Charles T. Wolverton said...

Just out of curiosity, from whence comes the recurring assumption in libertarian circles that there are a significant number of people who want to "equalize incomes"? I don't believe I've ever heard anyone express that preference, even among those who would be labeled "liberal" and presumably typically vote D. (And I think I would remember since such irrationality would be quite jarring.) Surely it isn't an inference from support for graduated tax rates, which as far as I know have neither the objective nor - according to some studies I've seen referenced - the effect of after-tax income equalization.

Also, what does being a "democrat" - a matter of political philosophy - have to do with either first or second sense egalitarianism, the former seemingly moral philosophy (essentially the golden rule with a dash of rational self-interest), the latter a mix of moral philosophy and economics?

- Charles

Trey said...

Does income disparity grow as a function of a society growing richer?

The lowest incomes cannot get lower than zero, while the highest incomes have no limit.

Anonymous said...


aren't there many examples of people favoring increased income equality and government intervention to make it happen?

One argument is that a poor person gets more utility (happiness) out of a dollar than a rich one, so transferring a dollar from the rich guy to the poor leads to increased happiness. The ideal state, then, that maximizes utility (happiness), would be one where we all get the same amount of money.

I believe this "argument from utility" to be part of the conventional wisdom. You'll hear it from Michael Kinsley.

Trey said...

Am I the only one who constantly hears about the problems of "the growing income disparity" in the country? Or how the rich are not paying "their fair share" of taxes. I mean, there is a Democratic party primary going on right now!

That is only true if you consider the distribution of income as not fair.

Do the rich deserve to be rich? I would say 'yes' to Bill Gates and 'no' to Paris Hilton. But are they both entitled to be rich? Yes!

My impression is that most folks would say Hilton is not entitled (along with rich ball players, CEOs, and most actors), and therefore the government has to “fix” it with income distribution.

Trey said...

One of my favorite parable deals with the difference between ‘deserved’ and ‘entitled to’.

From Matthew 20:1-16

Basically, several cohorts of workers put in different hours in a vineyard; some starting in the morning and some late in the day. At the end of the day, the owner of the vineyard gave them each the same amount of pay. When those who started earliest complained, the owner said “Friend, I am doing you no wrong; did you not agree with me for a denarius?”
So, we might say they deserved different payments, but they were only entitled to what was agreed to.

Anonymous said...

My own egalitarian hero (in the first sense) is also my father. When I was a kid in Sweden, we would always walk to the grocery store and after getting our groceries, we sat down on a bench outside. A group of winos hung out there too, and we befriended them. They were all friendly and often interesting. One was from Norway and like my father well read in Norwegian literature. He had been part of the resistance during the war.

Dad was always like that. We had many friends that were politically more socialist than my father. Many of them were comically preoccupied with status and money and shit like that. Their demands for 'justice' and for subsidies to things they wanted often seemed motivated by very petty anxieties.

Charles T. Wolverton said...


"aren't there many examples of people favoring increased income equality and government intervention to make it happen?"

There are obviously people who favor graduated tax rates, but to call that "favoring increased income equality" is misleading at best (not to mention logically flawed - incomes are equal or they aren't under any reasonable definition of "equal"). Suppose you pay $200K in taxes and your after-tax income is $600K, I pay $2K and my after-tax income is $18K. If the top marginal tax rate is raised and your after-tax income is reduced to $550K, can you honestly say that the resulting situation is fairly described by "our incomes are now more equal"? If so, I see a promising career in political speech writing.

I don't mean to be merely quibbling. My point is that a situation can be described in different ways depending on the intended message. "Some are concerned about the societal effects of growing income disparity", " the rich are not paying their fair share of taxes", and "they think incomes should be more equal" all address more-or-less the same issue, but the first is a polically neutral statement of fact, the second is D campaign.rhetoric, the last is R campaign rhetoric. On a quality blog like this one, I expect the first way. Partisan hackery I can get anywhere.

I see graduated tax rates as driven not by moral philosophy but by Willie Sutton's philosophy - "that's where the money is" (banks for Willie, the wealthier for the tax man). If some want to attack spending and then flatten tax rates, fine. But they should do it honestly and responsibly, not like the current admin. And they shouldn't assign manufactured motives to those opposed.

I'm not sure how you're defining "conventional wisdom", but if Michael Kinsley - one of the country's leading "public intellectuals"- is an example, you are using "conventional" in a way foreign to me. (And an admittedly half-hearted try with google unearthed no attributions of that concept to Kinsley). Anyway, my original statement re my experiences vi-a-vis income equalization stands.

- Charles

StefanoC said...

I suspect that on average a society that is more equal in the second (income) sense would be also more equal in the first (status) sense.

Compare North European countries with most of South America.

On the other hand, I can't think of any example of a place or time where there were very rich and very poor people, and yet that difference was irrelevant.

I admit that perhaps the causation goes in the reverse direction (greater equality in status causes greater equality in income).

William Newman said...

charles writes "I don't mean to be merely quibbling. My point is that a situation can be described in different ways depending on the intended message. 'Some are concerned about the societal effects of growing income disparity', 'the rich are not paying their fair share of taxes', and 'they think incomes should be more equal' all address more-or-less the same issue, but the first is a polically neutral statement of fact, the second is D campaign.rhetoric, the last is R campaign rhetoric. On a quality blog like this one, I expect the first way."

A minute's Googling "krugman income inequality" gave me . "clinton income inequality" looked like mostly arguments about statistics during Bill Clinton's administration, but "hillary income inequality" gives . It don't see an "they think incomes should be more equal" direct quote. But try "income inequality has risen to the highest levels since 1929, and wages have stagnated," "lowering taxes for middle class families," "empower our workers and ensure that all Americans contribute their fair share," and "restore the basic bargain. Hillary will restore the basic bargain that if Americans work hard and take responsibility, government will do its part to make sure they have the tools to get ahead."

I don't see how you can support the position that "some are concerned about the societal effects of growing income disparity" is a fairer paraphrase of this kind of thing than something like "most Democrats, and indeed a large fraction of non-Democrats, consistently argue that incomes [for people who inherited their citizenship, and similar 'well, duh' cynical side conditions] should be more equal."

You concluded with "Partisan hackery I can get anywhere." It seems to me that you have a pretty bizarre standard for what constitutes 'partisan hackery'." (What do you tell people who say things like "many Republicans still favor the invasion of Iraq" instead of, oh, "some are concerned about nuclear proliferation and terrorism?")

Clayton said...

@Charles, 10:14

"There are obviously people who favor graduated tax rates, but to call that "favoring increased income equality" is misleading at best..."

Well, a half-hearted attempt to Google it ("income equality") turned up a Wiki page devoted to it and an entire website with no other topic, among 407,000 other related results. :-P

But Google aside, I think you're right, the idea that liberals care about income equality is just a libertarian myth created to further their evil plot undermine the social order and wreak havoc on all humanity.

Charles T. Wolverton said...

simon made a very specific statement re Kinsley and the "argument from utility". Searching on that combination of words (entered with and without quotes), I didn't find an attribution of that concept to him. To respond that searching on various other sets of words resulted in lots of hits is a non sequitur.

"I don't see how you can support the position that ..."

I made an argument (my numerical example) supporting my contention - maybe neither sophisticated nor convincing, but an argument nonetheless. You responded by merely expressing skepticism. Try making an actual counterargument; maybe I'll be swayed and concede.

- Charles

Will said...

Charles: you are using equality in a different and stricter sense than most people. There is a spectrum from perfectly equal (your sense) to perfectly unequal (one person has literally everything). The mathematical expression of this is the Gini coefficient. In more colloquial terms, we tend to say "equal" to mean "more towards the perfect equality end of the spectrum." I do not believe any serious politician in the United States has called for perfectly equal income.

My response to the argument from utility is that people learn and form rational expectations. If you perfectly redistribute wealth one time, then given assumptions of risk aversion you have by definition raised social utility. Period.

However, if this begins to happen regularly, people will realize that they will receive the same income no matter what they do in the meantime, and will simply stop responding to any incentives. Rational expectations will always destroy the relationship you are trying to manipulate: just look at when governments tried to use the Phillip's Curve. Failure to account for incentives devastates so much public policy, and I think both parties are guilty in different areas. When it comes to income redistribution, I think the Democrats happen to have this one wrong.

Charles T. Wolverton said...


Thanks for a substantive response.

I had heard of the Gini coefficient, but never bothered to check it out. I obviously agree that no one is talking about achieving GC=0, but I still consider calling a concern about "extreme income inequality" (GC very high) a desire to achieve "more nearly equal outcomes" (GC~0, in the case of "outcomes" meaning "incomes") is an abuse of language. Suppose I suggest that in playing Russian roulette you put bullets in only five rather than six chambers to make your probability of survival "more nearly equal to one". Would you find that literally true statement comforting?

In your discussion of utility, you revert to perfect redistribution of wealth. I thought we agreed that that is neither rational nor being proposed by anyone, so I don't see the relevance.

I find arguments against modest increases in marginal tax rates that are based on disincentive unconvincing (in realistic scenarios, eg, marginal tax rates in their recent range). First, a claim that someone will work less or less hard if the top marginal rate is (X+5)% instead of X% (again, for X in a relevant range) just seems counterintuitive for what seem to me to be rather obvious reasons: the income inelasticity of many jobs, (95-X)% of a lot of money is still a lot of money, etc. But I could be persuaded by credible evidence. More important is that there apparently is no disincentive: as I alluded to earlier, there are studies that show the overall tax rate (percentage of total income paid in taxes) to be essentially flat for the non-poor. Ie, people with lots of potential income find tax avoidance schemes (although not particularly wealthy, even I do). In one study the rate for the highest income percentile actually went down.

As for Ds and income redistribution, I stated my position on that earlier with my Willie Sutton analogy: the rich(er) by definition "is where the money is". So if one wants to fund programs for the poor(er), the funding has to come from them.

If D candidates use "soak the rich" language, that's campaign rhetoric. All I'm suggesting is that if one wants to seriously discuss the merits or demerits of "help the poor" programs specifically or in general, one should use serious language.

- Charles

Anonymous said...

Charles, people don't necessarily work less if taxes are higher, but some will find tax avoidances (and evasions) that they might not otherwise have bothered with. Others will relocate to a different country (if not themselves, their assets).
And of course that will be the wealthiest doing that as they can afford the alternatives and have much more money to save by doing so, leaving the bill to, not the rich, but the middle income earners.

Clayton said...

@Charles 6:46

"I find arguments against modest increases in marginal tax rates that are based on disincentive unconvincing ..."

But the most important argument against taxes is not disincentive, it is deadweight cost. Yes, $1 has less marginal utility to me than it does to someone poorer than me. Yes, $1 collected in taxes has less utility to each individual from whom it is taken than the $150M (assume 150M taxpayers for argument's sake) has to whatever government bureaucracy is the beneficiary of those taxes. But for each $1 taken from taxpayers, the cost of living increases by $1. Or, stated another way, for each $150M taken from taxpayers, the aggregate cost of living increases by $150M.

So, whatever theoretical argument can be made about increases in marginal utility resulting from this or that rearrangement of wealth, the cost of living really increases dollar for dollar by the exact amount by which government taxes. Which is really nothing more than the old adage, "Money doesn't grow on trees." :)

Note further that deadweight cost is always regressive, it affects the poor more by percentage than it affects the wealthy.

"... people with lots of potential income find tax avoidance schemes (although not particularly wealthy, even I do). In one study the rate for the highest income percentile actually went down."

So what's the point? If the wealthy can afford the army of tax lawyers required to dodge whatever regulations Congress tries to throw at them (and they can afford them), the people who really get hurt are those who have too much money to be considered poor, but not enough to afford an army of tax lawyers. That is, the middle class, of whom the Ds are supposed to be the appointed champions.

Will said...


A quick reply about my utility argument: I did use the words "perfect redistribution" simply because it's an easy case to work with, and I was acknowledging that this necessarily results in an increase in total utility. However, my argument itself does not rely on perfect redistribution, as I think there is a spectrum of disincentives. Which leads me into the next part of my reply, against your assertion that incentives do not affect income.

First, an empirical point: income does indeed tend to be very inelastic for most people at most tax rates. However, it is very well documented that second-earners in a household have a much higher elasticity, and will frequently cut back working in response to increases in the marginal tax rate. A quick Google or JSTOR search should turn something up, but if not, I am happy to supply a link.

My second point is theoretical, in response to your skepticism towards incentives affecting income earning. Consider one scenario where you keep everything you earn, and consider another scenario where you lose everything you earn. In the second scenario, I think we can agree that the worker will never work, that is, they are completely disincentivized. However, if you accept that conclusion, then incentives necessarily must decrease from scenario one to scenario two.

Either this decrease follows some kind of smooth curve, or there is a discontinuous break somewhere. If you believe that it follows a smooth curve, then incentives must be monotonically decreasing with decreasing income. This conclusion is in contradiction with yours, and so you have to assume a discontinuous break in incentives somewhere along the marginal tax rate... but where would this occur and why? A break at 100% tax makes sense, given that it's now entirely pointless to work, but do you really think people would still be working at a 99% tax rate? What about a 98% tax rate?

Perhaps I am misunderstanding your argument, and for that I apologize. I think what you are actually trying to say is that income tends to be highly inelastic at sufficiently low marginal tax rates, which is where most taxation we have observed occurs. However, it is a logical fallacy to then conclude that income does not respond to incentives at all. I highly doubt you believe that a tax increase from 20% to 80% will not cut back work at all.

The reason I am harping on this point so hard is because I think ignoring incentives is very dangerous. Incentives will always exist in every situation, even if they're not immediately obvious or powerful enough to notice.

Will said...


There is no deadweight loss to a tax if supply or demand is perfectly inelastic, which is Charles's claim. The argument of elasticity is logically prior to the argument of deadweight loss; they are not independent of each other. Until Charles believes income is not perfectly inelastic, he has no reason to be convinced by deadweight loss.

Charles T. Wolverton said...

First, let me say that the last few comments are represenatative of the kind of discussion of these issues I would like to see. In the face of such a discussion by others, I would be inclined to retire to a read-and-learn mode since I would have little or nothing to add.


OK, I buy your point about second incomes - been there, done that.

As for the "theoretical" point, you are of course quite right that disincentive is monotonically increasing with increased marginal rate. But I see the important parameter as being the amount of disincentive for a proposed increment in marginal tax rate. Which is why in my example I incremented the marginal tax rate by five percent, not 60-70%. Yes, there is disincentive; but I suspect not very much.

No one is proposing a return to marginal tax rates of 90%, so what motivates people like my conservative neighbor who thinks a D in the White House will result in "our taxes going through the roof"? (Answer: partisan hackery.) Besides, we now seem to have bi-partisan support for unbounded deficit spending, so why worry about tax rates at all? No need to increase them if we don't intend to pay for our profligate spending.


Although "deadweight loss" is a new concept for me, I think I get the point. But implicit in the concept (and the perjorative label) seems to be the assumption that taxes which pay the salaries of government bureaucrats are a total loss. I am reluctant to point out the obvious, that most (all in a zero saving society) of the money ultimately returns to the economy through consumption, so in that sense the loss is not total.

I assume the rebuttal is that the government bureaucracy is relatively low productivity, so there is nonetheless some marginal loss relative to a higher productivity job in the private sector. But doesn't this assume that there is always a more productive alternative job in the private sector? And doesn't the very nature of the high consumption US economy belie that assumption? Eg, is being on the production staff of American Idol really in any meaningful sense more "productive" than a job in the government bureaucracy?

Anyway, we've come a long way from "Ds want to make incomes more equal". And IMO, that is progress.

- Charles

Jonathan said...

No-one seems to have brought up the early communist slogan, "From each according to his ability, to each according to his need."

It's interesting to note that this is not a prescription for income equality. If the slogan were implemented, someone who needed a lot of expensive medical treatment would have, in effect, a much higher income than a healthy person.

Will said...

I wish to make one final post, just to clarify what deadweight loss from taxes is, and where it comes from.

Suppose you have a good, the usual supply and demand curves apply, etc. Assume the price of the good is $2. Every consumer who values the good more than that will buy and every firm who is capable of supplying that good for less cost will sell. Every economically efficient transaction will occur in this situation.

Now say the government imposes a 5% tax, and the price of the good is $2.10. The result is that every consumer who valued this good between $2.00 and $2.09 will no longer be willing to buy it (and subsequently firms supply less of the good). These are economically efficient transactions that would occur under a free state of the world, and no longer will. Thus, the surplus utility that would have been created by those trades is now gone. That is deadweight loss.

Clayton said...


Although "deadweight loss" is a new concept for me, I think I get the point. But implicit in the concept (and the perjorative label)

Note that deadweight loss is applied to any kind of monopolistic pricing, not just government taxes.

seems to be the assumption that taxes which pay the salaries of government bureaucrats are a total loss.

As William explains above, the money taken in taxes itself is not the deadweight loss (on re-reading it, my original comment seemed to imply that this is the case) - deadweight cost is a result of the difference between the "free market value" and the actual price charged for something as a result of taxes or monopolistic pricing. Wikipedia has a good price graph depicting it (just go to the Wiki article titled "Deadweight loss").

I am reluctant to point out the obvious, that most (all in a zero saving society) of the money ultimately returns to the economy through consumption, so in that sense the loss is not total.

The loss is in consumption (and, symmetrically, production) that would have happened but did not due to a contraction in total real demand as a consequence of the artificially higher price on the taxed good (or goods).

Income taxes are not exempt from this because they can be thought of as a tax on all consumer goods.

Eg, is being on the production staff of American Idol really in any meaningful sense more "productive" than a job in the government bureaucracy?

Of course. American Idol operates profitably because they (and all other entertainment content providers) reduce the cost of doing business by connecting willing buyers with willing sellers through ... advertising! :)

As with any other monopoly, there is no way to tell whether a particular government bureaucracy is a net benefit or cost to the economy, but the odds are that it is a net cost.

But in any case, as William noted, the deadweight loss is in the lost consumption which results from the artificially inflated price of goods and services which results from taxes, not the taxes themselves as my first comment implies.

Clayton -

Charles T. Wolverton said...

w & c:

OK, got it. However, I still see an implicit assumption in your description of deadweight loss due to tax - maybe right, maybe wrong, but there either way.

Not being receptive to the anarcho-capitalist view, I assume government does provide some useful functions. These must be paid for by taxes. Therefore, some fraction f of a tax may be deadweight loss but not all (the implicit assumption that I think is present in your description is that f= 1, with which I disagree). Ie, I view 1-f as a cost for the product or service in question.

So, maybe we have an objective way of defining political leaning: anarcho-capitalists think f=1, liberal Ds think f < 0 (taxes always too low for the wonderful benefits of government), Rs don't care what f is but recognize that "deadweight loss" has great potential for use in a sound-bite.

Oh, and thanks for your patience.

- Charles

David Friedman said...

"Although "deadweight loss" is a new concept for me, I think I get the point. But implicit in the concept (and the perjorative label) seems to be the assumption that taxes which pay the salaries of government bureaucrats are a total loss."


That assumption might be true. It might even be optimistic--perhaps government bureaucrats do net damage. But it does not have anything to do with deadweight loss as normally defined by economists.

A tax collects $1000 to pay the salary of government employees. Collecting it imposes costs on the taxpayers--not only the loss of $1000, but also changes in their behavior as a result of the tax that make them worse off without providing any money to the government. Thus the total cost to the taxpayers is $1000+X. X is the deadweight loss.

Another poster writes:

"the deadweight loss is in the lost consumption which results from the artificially inflated price of goods and services which results from taxes,"

Also wrong. The deadweight loss is the lost welfare due to the altered pattern of consumption and production due to the effects of the tax. It's a reduction in consumer (and producer) surplus, not in consumption.

If either of you would like a more detailed explanation, you might want to take a look at my webbed Price Theory. It's at:

The idea of consumer surplus is introduced in Chapter 4. The welfare implications of taxes, including deadweight loss, are in Chapter 7. Figures 7-9 and 7-10 show graphical representations of deadweight loss. As you will see, the question of what the government uses the money for doesn't arise at all.

Will said...


First of all, David Friedman's links should really help clear things up. I can't put it better than he can. I encourage you to read it, and then reread our comments in light of that.

Second of all, I would like to point out a utility accounting error in your conception of government action. Before a tax is introduced, the supply and demand split the total surplus utility. Now consider the tax raising that good from $2 to $2.10. Some people stopped consuming (which is the deadweight loss), but the ones who continued to consume lost $0.10 worth of utility. This is the impact of the tax on the consumer. Thus the supply and demand get part of the surplus, some of it is lost, but the remainder goes to the government in taxes.

I agree that the government spends money, attempting (for the most part) to improve people's welfare. Your point is that the government needs to prove it can provide more benefit than the deadweight loss. However, this view fails to take into account the fact that the very people they are trying to help, the consumers, are all worse off by exactly the amount of the tax. Thus the relevant benchmark that the government has to meet is that it provides more utility than both the deadweight loss AND the amount they collected in taxes, or else it is on net damaging to social welfare. This is a very important distinction.

One could argue that this benchmark is easy to meet by government action. However, consider this point: what would the individuals have spent their money doing had it not been collected in taxes? If you believe that everyone is pursuing their own best interests, then every single person spending that money are putting it to exactly what would make them the best off. This sets an incredibly high benchmark for government: they have to justify to their citizens that the government knows better than every single one of them do about what is best for them, in fact, that they are able to produce AT LEAST more additional social utility than the deadweight loss. This is an extremely tough standard to beat. I think the obvious political ramification is that the burden of proof MUST be on the government: the benefit of any action must be extraordinarily evident, and vast majority of people support it.

(Note that this is an argument about government spending. If you're talking about the absolutely simplest income redistribution, where you literally take money from the rich and give it directly to the poor, it's a slightly different case. In that situation, the government needs to prove that the additional utility of the redistribution is larger than the deadweight loss. I think this argument is a little less intuitive, as it involves the marginal choices of the low-income versus high-income groups, essentially you're trading investment for consumption. I think the government needs to be extremely clear why it should favor consumption, particularly why it thinks this will produce even more utility than the deadweight loss the tax produces.)

Charles T. Wolverton said...

Well, I've skimmed the various references provided (thanks) and need to read them more carefully, but unfortunately am out of time (we're leaving on vacation Sat AM). So, I have to make one last try to convey my perspective and hopefully get a response on why I'm wrong.

The scenario in each write-up and in the various comments has been that there is an a priori situation where the tax is absent and the equilibrium price has been established. Then, out of the blue a tax is imposed. And yes, I see that this has various bad effects including the deadweight loss, and to be justifiable has to meet an arguably insurmountable benefit-to-cost ratio.

But that isn't the scenario I tried (ineptly) to convey in my last comment. I was trying to argue that in the economic system we actually have (ie, not some hypothetical anarcho-capitalist system), there are some government services that are (or at least are viewed by most as being) essentially indispensable. Those services have to be funded, and once a tax has been levied to provide the funding, that tax becomes a cost of "doing business", whether the "business" is producing a good or service or is the consumption of a good or service. Ie, it's included in setting the equilibrium price.

Now, if at some point an additional tax is levied to fund an arguably dispensable government service, I see how the deadweight loss and other ill effects enter the picture. But not prior to that new tax.

In reviewing the comments, I see that the flow was from the disincentive effect of higher marginal tax rates to the deadweight loss effect, specifically on income redistribution. I will just note that it was not I who introduced that specific taxation objective, which might well be a "dispensable" service and therefore an appropriate tax to which to apply the deadweight loss argument. But my "Willie Sutton" argument was that to fund any tax, even one to supply arguably minimal "indispensable services", you have to go where the money is, ie, to those who have it. Hence, graduated tax rates. Call that "soak the rich" or common sense, I don't see how to get around it.

- Charles

Will said...


The reason everyone else in this thread is assuming some a priori taxless state and using that as an alternative case is because of the concept of opportunity cost.

When evaluating ANY possible option, you have to consider the benefit of the next-best option as well. The reason for this is because of scarcity - we cannot simultaneously pursue every available option. Instead, we can only choose one of them. However, in passing up the alternative, we are incurring a cost equal to the benefits we would have received from the next-best option. This is the opportunity cost.

If you fail to factor opportunity cost into your cost-benefit calculations, then it appears you would want to take every single action with a net positive utility. Since these actions are mutually exclusive to some extent (time, money, whatever), you have no guide as to which one you would choose. Once you incorporate opportunity cost, by subtracting out the benefits of the alternative use of your time, you now have a single best possible option.

I understand your point that we currently have a government, and that it provides certain "indispensable services" and requires funding for them. You call the effects of this tax the "cost of doing business" and simply take it as given now. You also admit that if the government proposes a new tax, that it might have an insurmountable hurdle, and so the action should not be undertaken. However, these two views are incompatible unless you fundamentally assume that government actions are irreversible.

Suppose you're looking at a proposed government program. It will impose a tax of 10mm dollars, and create 2mm of deadweight loss. It will use these funds to create a social program generating 11mm worth of value. This results in a destruction of 1mm worth of value, and thus the action should not be taken.

Now consider a currently-active government program. It has created 11mm worth of value from 10mm in taxes, thus creating a social gain of 1mm. However, economists have estimated that if the taxes were not in place, there would be an additional 2mm of value created. From this perspective, which is exactly the same as the previous one, you should want the government to end the program for the social good.

The bottom line is that if a government program is not on net beneficial, regardless of whether it is being proposed or already in place, it should be eliminated. You are attempting to say that some programs are "indispensable," but if they don't result in a social benefit why should they exist? If they do in fact result in social benefit, even after clearing the arguably insurmountable hurdle, then simply say so, and then we can discuss the merits of that claim. Simply asserting a program as "indispensable" is not an argument, it's an assumption.

Your original point was simply that in order to raise revenue you have to "go to where the money is" and thus tax the rich. First you try to separate this from income redistribution, but that necessarily follows from a graduated tax rate (the rich hold proportionally less wealth than they would under a flat tax rate) and thus should be factored in the discussion. Second of all, I don't think one follows from the other, since there are lots of potentially taxable things with money. You could levy a tariff, since there is money in international trade, but I don't think many economists would argue this is the best approach to raising revenue. Corporations have lots of money, more than most "rich" individuals, but that's not the ideal thing to tax either. Discussing the most efficient taxation strategy is itself a very interesting activity, and I am happy to engage in the relative merits of each tax, but it requires a lot more analysis than simply "follow the money."

Anonymous said...

Very cool blog...
I loved it, big-time... And I've linked to you at my blogroll, too.

My blog is written in Arabic, however we share the same Right Libertarian views, exactly...

Nice to meet you, sir.

Ahmed Zidan, Cairo - Egypt.

Charles T. Wolverton said...


"You are attempting to say that some programs are "indispensable, ..."

No, I said "some government services ... are (or at least are viewed by most as being) essentially indispensable". The parenthetical was intended to deflect precisely this response. Whether they are or aren't indispensable is irrelevant - for all practical purposes, they are permanent, and the funding for them is a fixed cost to taxable entities. Maybe it's not clear, but in "government services" I am including congress, the federal judiciary, the agencies, DHS, INS, etc. If you are assuming I am referring to so-called "welfare programs", maybe that's why we don't seem to be communicating.

"... but if they don't result in a social benefit why should they exist?"

I don't know where you're going with this question, since I don't know how you define a "social benefit". Does a B1 bomber provide a "social benefit"? The FDA? Yellowstone Park? Do you mean any benefit or the optimum benefit in accordance with an opportunity cost analysis. If you think no government activity provides a "social benefit", we're just spinning our wheels.

"you try to separate [taxing the rich] from income redistribution ..."

No. I said "it was not I who introduced that specific taxation objective [ie, income redistribution]", which was intended to point out that a consequence is not necessarily an objective. The objective I am addressing is funding some government expenditures via the individual income tax. "Going where the money is" may have graduated tax rates as a consequence, but graduated tax rates are not the objective. This is similar to my objection earlier of ascribing the motive of "making incomes more equal" to support for graduated tax rates.

"Discussing the most efficient taxation strategy ... requires a lot more analysis than simply 'follow the money.'"

Which can be generalized to "discussing anything seriously requires a lot more than making unwarranted assumptions of motive and throwing around sound bites" - the essence of my initial comment.

- Charles