The aesthetics of inequality

This post was triggered by a recent Will Wilkinson essay on inequality.  Earlier I linked to a Wilkinson essay where he cast a bemused, skeptical eye on happiness research.  His inequality essay is more serious, as the issue is much more highly charged.  BTW, it’s interesting to consider why “happiness,” the supposed goal of humanity (according to economists) is regarded by many as a fairly frivolous research topic, where as inequality is very serious.  One answer is that we can’t measure happiness very well, but as Wilkinson shows there’s no evidence that our inequality measures are any better.  Indeed I’ll add a few criticisms of my own.

Although I sympathize with Wilkinson, I also think those on the left have a very powerful argument, and the response from those on the right can sometimes seem a bit too convenient.  In the end I am going to agree with Wilkinson, partly because of arguments that he makes, but partly for some very different reasons of my own.

Wilkinson starts off with some comments on Krugman.  Readers of this blog may assume that I am as obsessed with Krugman as Ahab was by the white whale.  But I’ll use a different animal metaphor—for anyone on the right, Krugman is like a big elephant standing in the middle of the road.  He is the most articulate spokesman on the left, and unless you can get past his arguments you won’t get anywhere.

It’s generally agreed that the income distribution in the US has become significantly more unequal since 1970.  Wilkinson points out that if we are interested in living standards, however, the relevant variable is consumption, not income.  Income is distorted by fluctuations in realized capital gains, and all sorts of other factors.  Nominal consumption inequality shows much less little change.  In addition, he provides evidence that the cost of living for low income Americans has probably risen much more slowly than for wealthy Americans.  Low income Americans have benefited from the Walmart effect, whereas richer Americans have been penalized for wanting to live in areas with a rapidly rising cost of living, such as Manhattan and West LA.

[BTW, I recall reading that Bill Gates drives a Lexus LS, whereas the most popular car in America is the Camry.  Both are made by Toyota, and while the LS is undoubtedly a better car, the actual experience of driving the two is not that different.  I’m sure that a few hundred years ago the difference in transport between the richest man in the world (Louis the XIV?) and the average Frenchman was considerably larger.]

Wilkinson cites a study by Stevenson and Wolfers that shows that happiness has become much more equally distributed since the 1970s, exactly the opposite of what you’d expect if you relied on the trendy left-wing narrative on inequality.  They argued that some of this may be due to the increasing openness of American society.  Women and minorities now have more opportunities, and thus may not feel quite so powerless.  Is happiness what really matters?  Perhaps not to cultural conservatives; but aren’t liberal economists supposed to be utilitarians?

Now let’s get to Wilkinson’s most interesting argument, immigration.  He passionately advocates open immigration, arguing that it would be the most effective program for reducing poverty.  His most powerful argument is directed against what he calls “analytical nationalism,” which is roughly making normative judgments about society is if foreigners, and even Americans born elsewhere, are of no consequence.  For example, assume a large number of low income peasants moved to America, and over time saw an enormous increase in their living standards.  For people who worry about inequality this should surely be a big plus, shouldn’t it?  Actually, it’s just the opposite; the immigration would make America’s income less equal, and might even increase the poverty rate, even if every single person now living in America was better off than before the immigration occurred.  If course in the real world some Americans are hurt by immigration, but that is no reason to ignore the income gains of those that are helped, the immigrants themselves.

Those on the left might reasonably reply that immigration wouldn’t fully explain the rise in income inequality.  No, but it might completely explain the much smaller rise of consumption inequality.  Alternatively, given how many immigrants come from countries with very low incomes, it might make the US look much better from a (Rawlsian) maximin perspective—which emphasizes helping those worst off.  So how do left liberals feel about the huge welfare gains from immigration?  Here’s Wilkinson quoting Krugman:

Analytical nationalism has serious real-world consequences.  It leads well-meaning people to countenance, or even support, acts of injustice against fellow members of our transnational society—restrictions on the free movement of persons across political boundaries—in the name of combating the illusory injustice of an uptick in the national Gini coefficient.  These gaffes lead Paul Krugman, for example, to tie his conscience in a liberal knot.  “I’m instinctively, emotionally pro-immigration,” Krugman confesses.  But he is also instinctively, emotionally committed to the moral relevance of national income inequality statistics.  Thus does a modest rule that tells the Census Bureau where to stop counting come to tell Krugman whose welfare really counts.  “We’ll need to reduce the inflow of low-skilled immigrants.  Mainly that means better border controls on illegal immigration, ” Krugman concludes.  After all, “the net benefits to the U.S. economy from immigration, aside from the large gains to the immigrants themselves [emphasis added], are small.”

I’d like to shift the focus from whether immigration is desirable to why left-leaning liberals often oppose immigration.  When Krugman says he is instinctively pro-immigration, I take that to mean he is comfortable with immigration at a cultural level.  Surveys suggest that liberals tend to be relatively open to new experiences, new cultures, diversity, etc.  In contrast conservatives often feel threatened by the mixing of languages, religions, races, cultures, etc.  So let’s assume that Krugman is culturally liberal.  Why then would he favor restrictions on the immigration of unskilled labor, given that such immigration could dramatically improve the living standards for low income Americans who live here after the immigration had occurred ?  (I.e. including the welfare of the immigrants themselves.)

The best way of discovering what Krugman really believes is to read Keynes.  Unfortunately I could not find the relevant quotation (can anyone help me?) but I recall Keynes saying something to the effect that poverty is aesthetically ugly.  It seems to me that analytical nationalism is essentially saying that what really matters the not the well-being of Americans, but rather the well-being of America.  (No, I’m not going to call liberals “fascists,” as conservatives tend to feel this way even more strongly than liberals.)  In other words the country is to be judged from an aesthetic perspective; a “Great Society” is a morally beautiful society.  A peasant village perched on a hillside in a third world country can be aesthetically beautiful, but a shantytown of former peasants on the edge of a large modern urban area (even if the peasants are now better off) is aesthetically ugly.  BTW, this might explain the vastly different appearance of Indian and Chinese cities.

So if we want an attractive society without abject poverty and shocking inequality, then what are the most aesthetically beautiful societies on earth?  Who should we emulate?  Switzerland would make any list.  Singapore is obviously the Switzerland of Asia, and perhaps we could add Sweden.  But I’ll substitute Denmark for Sweden, and then we have my list of the three possible neoliberal models for the 21st century (in a paper a wrote a year ago and briefly discussed on my Great Danes post.)  Hyper-egalitarian neoliberalism (Denmark), hyper-economistic neoliberalism (Singapore) and hyper-democratic neoliberalism (Switzerland.)  These countries all look good from a utilitarian perspective.

Note that what is missing from this group is statism.  Singapore is the second most free market society on earth according to the Heritage Institute, and Denmark is number one if you excluded government spending and taxes, and averaged the other eight categories.  Switzerland is a bit further down the list, but has always been considered a fairly capitalistic country by Europeans standards, and certainly has relatively low income taxes.  So if you want the “Great Society” that LBJ called for, one place to start is with free market policies.  Surprisingly, another example is China.  Yasheng Huang pointed out that when Chinese reformers followed a free market model in the 1980s, inequality was reduced.  When they switched to a state-led approach in the 1990s, inequality widened.

[Just to be clear, when I speak of a society’s “beauty” I mean it in the moral sense employed by Keynes, not in terms of artistic beauty or natural scenery.  Italy would surpass Denmark in artistic beauty, and almost any country would surpass Singapore in natural beauty.]

So this is the liberal’s dilemma.  They may delight in the cultural diversity produced by immigration, and enjoy eating in all the great Thai restaurants.  But if the immigrants are unskilled peasants then they may have to accept that American cities will not be as neat, tidy, and prosperous looking as Zurich, Stockholm and Singapore.  And that thought is apparently too much to bear.

Now let me add a few of my own thoughts on inequality.  Is inequality the issue, or is poverty?  Imagine a society where a third of families make $100,000 a year, another third make $300,000 a year and another third make $900,000 a year.  It is rather unequal, but there is no poverty.  To me, this wouldn’t seem aesthetically ugly.  Indeed, the town I live in is a bit like what I just described; there are middle class families, upper middle class families, and rich families.  (And don’t give me any national income data, I know $100,000 is above average, but in Newton it is just middle class.)  But what about the saying “the poor are always with us?”  Why hasn’t poverty gone away as average living standards have increased enormously?

Consider the consumption bundle of America’s poor.  In many ways Americans below the poverty line do quite well in international comparisons.  I recall reading that the average dwelling of a poor American is bigger than a middle class unit in most European countries.  Statistics on ownership of household appliances, cars, etc, are also relatively good.  Even so, only the most willfully blind would argue that there is no poverty in America.  Public schools, access to health care, crime rates, etc, are often far worse for America’s poor than those in other developed countries.   But this supports Wilkinson’s argument that instead of fixating on the abstraction of “inequality,” we should instead focus on fixing specific problems like inner city schools.

I would also argue that the distribution of income is strongly distorted by age.  Lots of people who are in the bottom quintile are basically middle class, but young and single.  And lots of people in the top quintile are working class; the Boston cop married to a nurse, for example.  And some people in the 20-40% quintile are actually poor, a single mom raising three kids in NYC on $28,000 year.  At various times in my life I have been in all 5 quintiles of family income distribution, and yet I have always felt like I was in the same “class,” and I have never felt like my happiness had anything to do with how much income I was making.  Some of that is due to the fact that I was never a single mom trying to raise kids, but wouldn’t you expect money to have at least some impact on my perceived happiness?

[Perhaps I am atypical.  Michel Houellebecq argues that in a modern society the real inequality is between nerdy guys and cool guys, or beautiful women and ugly ducklings.  I suppose when I was young I felt that way.  But how would I know?  I’ve always been a nerd, and the grass looks greener on the other side of the fence.  And what about inequality associated with mental health, or physical disability?  It seems much more important than income to me, but again perhaps that merely reflects my middle class upbringing.]

Now let’s consider the touchy issue of ethnic diversity.  Is it a “bad thing” that some ethnic minorities such as Japanese-Americans, Jewish-Americans, and (Asian) Indian-Americans now have higher average incomes than us formerly dominant WASPs?  I’ve never heard any reasonable person answer this in the affirmative.  Why?  Because in some sense it is assumed that these minorities have “earned” there success.  Now you might be surprised but I actually think the left liberals have a very powerful argument in favor of progressive taxation.  The argument that successful minorities have earned their success is not the same argument as successful individuals have earned their success.  To simplify things, assume that these minorities did well because they tended to go into high-paying professions that required a lot of education.  Then it would be perfectly OK for a left liberal to argue that lots of Indian-Americans have earned the right to be MDs, and yet also argue that high-paying MDs of all ethnicities should pay high income taxes to help the poor.

So if that isn’t my argument, what is?  Go back and ask why these groups are generally viewed as deserving their success.  If it is deserved by virtue of some aspect of their culture, then isn’t that just another way of saying that in more “diverse” societies you would expect more income inequality?  And indeed not just that you would expect more income inequality but that it would be in some sense morally justified?  I’m not saying that in the long term we shouldn’t strive for ethnic equality as well, certainly cultures can and do change over time.  My point is that cultures change slowly, so we shouldn’t necessarily expect an optimal social welfare policy in America to immediately produce the same results as that policy would produce in a more ethnically homogeneous country.

Again, I think those on the left would have a very strong counterargument—even within various ethnic groups, income in America tends to be more unequal than in many Europeans societies.  So let’s get to the bottom line, why don’t we take some money from the rich and give it to the poor?  This is one question that I don’t think Wilkinson was able to answer in a convincing fashion.  I tried to put myself in the shoes of a supporter of the Nordic model, and imagine how they would react to Wilkinson’s essay.  It seems to me that a left-leaning utilitarian might reasonably conclude that Wilkinson was looking for all sorts of reasons to minimize the problem of inequality.  Isn’t the bottom line that $100 is worth much more (utility) to a poor person than a rich person?  So why talk about whether income or consumption is appropriate, or whether liberals are hypocrites for not supporting immigration?  Aren’t conservatives just looking for excuses to avoid paying taxes to help the poor?

One response would be that Wilkinson is not a traditional conservative; indeed his views on immigration put him far to the left of even most liberals.  But he does seem to pick up some other conservative themes, such as that there is something appropriate about the initial distribution.  Or at least that’s what I think he implies:

Nicole Kidman is fabulously wealthy because millions of individuals have chosen to see a movie with Nicole Kidman in it instead of a non-Kidman movie, or instead of going bowling.  Of course, these myriad choices take place within a framework of political, legal, economic, and social institutions and norms—all of which affect the eventual pattern of incomes.

I just don’t find this type of analysis convincing, although at one time I would have.  I used to think that people deserved what they earned, but no longer.  Now I think rich people should keep what they earn if and only if trying to take it away from them it will do more harm than good.  In other words, I’ve gone from being a dogmatic libertarian who thought the Nordic model was bad, to being a pragmatic libertarian who thinks it’s worth considering.  I lean towards Singapore’s slightly less egalitarian forced saving regime for many different pragmatic reasons.  But if someone can convince me that Denmark’s undeniably successful social welfare system is better, I’d jump ship in a moment.  As far as I am concerned the fact that “millions of individuals” have “chosen” to spend their money on Nicole Kidman films has no more normative implications than if a bag of Federal Reserve notes had fallen from an airplane into her front yard.

I’m not a card-carrying utilitarian, but I think we might as well use utilitarian methods for policy evaluation until we come up with something better.  I don’t know Wilkinson’s views on utilitarianism.  He is an economic philosopher, and must have an opinion, but he refrains from using explicit utilitarian arguments even when they would naturally buttress his policy views (as with immigration.)  So what does this all mean?  I believe that supply side economics is the key issue.  We take Kidman’s money if it doesn’t hurt the supply-side of the economy very much, and we don’t if the costs aren’t worth the benefits.  Right now the Singapore system with low MTRs looks more robust for the long run than Denmark, and less dependent on a high trust culture, so I’m inclined to favor a low tax economy with a basic safety net for pragmatic reasons.

Earlier I argued that many “economistic” views were highly counterintuitive.  One of those is supply-side economics, which suggests that high marginal tax rates can have a very adverse affect on wealth creation.  Perhaps because they instinctively feel it is hard to make the supply-side argument, or perhaps because they don’t buy it themselves, many libertarians rely on the argument that those who work hard to create wealth deserve what they get.  This argument appeals to the politically powerful conservatives, a much larger group than pragmatic libertarians.

But in the end I think libertarian economists will need to confront the supply-side issue head-on.  Take away the disincentive effects of high MTRs and the left liberals have a much stronger case.  Kidman may have worked hard, but she was very lucky not to be born an ugly peasant farmer in some third world country.  Great wealth is always created by teams.  Marginal revenue product is not a moral argument.

Pragmatic libertarians need to look beyond the US borders, and point out that even much more egalitarian societies than the US found a need to sharply slash MTRs in recent decades.   Unfortunately, President Obama managed to find the one University of Chicago economist (Goolsbee) who dismisses supply-side economics.  The proposed health plan in Congress would actually raise the top rate (combined Federal, Medicare, and state MTRs) to Swedish levels.  And yet Singapore is able to provide health insurance for all its citizens despite having MTRs far lower than those in the US or Europe.

For those who don’t find traditional supply side arguments convincing, there are two other interesting pieces of evidence for the US system over the European model, one objective and one subjective.  The objective argument is that immigrants seem to prefer the US.  Most people I talk to from China and India say they would much prefer the US, and they also indicate that their view is fairly widespread back in their home country.  The subjective argument is that in surveys Americans tend to say they are a bit happier than Europeans.  The theoretical argument for income redistribution seems powerful; a dollar probably is worth more to a poor person than a rich person.  But “the people” don’t seem persuaded.

I better stop here, although there is much more to say.  Wilkinson is an engaging writer (and speaker, as those who watch know.)  There are many more good arguments that I didn’t have time to mention; don’t be put off by the one argument I singled out as unconvincing.  I also highly recommended his “In Pursuit of Happiness Research.”  Another excellent essay on a related topic is Brink Lindsey’s “Nostalgianomics.”  Lindsey is another participant with a pleasant personality.  How can people on the left think these Cato guys are part of an evil organization?



54 Responses to “The aesthetics of inequality”

  1. Gravatar of R. Pointer R. Pointer
    18. July 2009 at 19:12

    Jeez, that was a long post. I’m still not finnished! [sic] I mean done.

    Additionally, I find it interesting that you attribute the Danish system to a high trust culture when it could be that institutions have reduced the incentives for free-riding and engendered path-dependent usage of said institutions. The VOC nor Welfare State literature mentions not any cultural background for the development of these institutions. Indeed for some, cross-cutting political compromises at the start of industrialization help facilitate solutions to employment risk.

    Dr. Sumner, how do you come to understand institutions? Are you a Neoclassical or Transactions cost economist? Or is that the wrong kind of question to ask a Macroeconomist?

  2. Gravatar of Will Wilkinson Will Wilkinson
    18. July 2009 at 20:09


    Thanks so much for this terrifically thoughtful post. Lots I could say, but let me just limit it her to saying that I’m some kind of liberal contractarian, and I reject utilitarianism mostly for the usual contractarian reason. See, for example, Scanlon’s classic “Contractualism and Utilitarianism.” Matt Zwolinski has a nice overview of Rawls and Nozick’s different separateness of persons arguments against utilitarianism here:

    About desert and distribution, I’m a Hayekian, and think it’s mostly sort of category error to think of people as deserving their income levels or the difference between their income level and somebody else’s. I can deserve my income in the sense that I have an employment contract that specifies such and such an income from such and such services, and I perform the services. I deserve that income from the other party to the contract. But I nobody deserves the bargaining conditions under which the contract written. A plumber can make a killing if there aren’t other plumbers around. And he deserves to have people pay him what they agreed to pay him. But he doesn’t deserve to be the only plumber in town. If I run the fastest, I deserve a blue ribbon. But I don’t deserve the fact that a faster guy sprained his ankle in warm ups. Etc.

    In the passage you quote, I wasn’t saying Nicole Kidman deserves her income. I was saying that her income wasn’t “distributed” to her by anyone. It is the outcome of a huge number of individual choices made within a context of a huge number of rules and constraints. My view is like Nozick’s and Hayek’s. If Kidman’s income (and the pattern of incomes generally) is the outcome of a bunch of choices made according to basically just/fair rules and constraints, then she’s entitled to it. But it’s nonsense to say she deserves it, or that the pattern generally tracks or reflects desert.

    I’ve blogged quite a bit about desert and distribution. Try some of these, if you’re interested:

    Also, I think Krugman is instinctively pro-immigration mainly because he’s a good trade economist. But it’s also true that he’s a basically liberal-minded guy.

  3. Gravatar of rob rob
    18. July 2009 at 20:12

    relative wealth matters more than absolute wealth in terms of attracting mates, at least for men. i dont know why more liberals dont use that as an argument for why division of the pie in some sense matters more than the size of the pie. probably because it makes the one making the argument seem like a loser. we r all competing for happiness. economists like to ignore that, but posting this from a bar tonight makes me hyperaware of that fact. somehow the chicks arent interested in the guy who cant keep hias nose out of a blackberry.

    i refuse to believe the desire to make money has anything to do with the pursuit of happiness. we r just like bowerbirds building our nests. we dont really know what compells us, but better nests tend to attract better mates on average.

  4. Gravatar of Rob Rob
    18. July 2009 at 20:26

    One thing I always wonder when people compare models and say “maybe we should consider this or that less-free-market model” is… How much does the prosperity/GDP/wealth in those countries result from technological and other innovations that started in the non-model U.S.?

    IF 1) the U.S. is a major net exporter of productivity-enhancing technological or other innovations, and 2) the U.S. would not be as innovative or productivity-increasing if it adopted other less-free economic models, and 3) other countries’ current prosperity/GDP/wealth is increased by imported innovations more than is so in the U.S., THEN the comparison of models is mostly worthless, because built into the other models’ is the benefit of having the U.S. be non-model.

    To consider whether the U.S. should adopt this or that other model, then, one should ask, what technological advances are less likely to have happened if the U.S. had adopted that model 50 years ago; what would the Nordic model look like today in that environment; and how does THAT compare to the U.S. in actuality today.

    It is astonishing that the U.S. is, on average, 1/3 richer than France NOTWITHSTANDING France’s ability to adopt all the technological innovations invented here and elsewhere.

    What would the disparity be if countries were only limited to their own inventions? That gives some clue as to whether the world, and the U.S., would be better or worse off if the U.S. adopted a different model.

    (Yes, there is of course IN-flow into the U.S. of innovation from all over the world, so that the question goes both ways. But that doesn’t mean it all nets out to zero in the end. I think supporters of free markets are willing to stake a claim as to which way the net benefit runs…)

    Put in more simplistic terms, if I figure out how to create a goose that lays golden eggs, and my neighbor who does nothing but drink and beat his wife copies me and reaps the benefits from having a golden-egg-laying goose, he appears to be just as prosperous, but that says nothing about his model of living or of running his household.

    It’s similar to the old joke that someone famous (Vaclav Havel, maybe?) used to tell about how, in the Cold War, even if the U.S. had fallen to the communists, the Soviets would still have allowed, say, New Zealand to stay a free market. Why? Well, because we’d still need to know how much milk should cost…

  5. Gravatar of Rob Rob
    18. July 2009 at 22:23

    “How can people on the left think these Cato guys are part of an evil organization?”

    Because as you point out, they keep making these ridiculous Kidman arguments: that people get what they deserve. Most people only tend to feel that way when they see someone taken away in an ambulance, not a limousine.

  6. Gravatar of david david
    18. July 2009 at 22:44

    As Bryan Caplan @ econlog as previously pointed out, most Singaporeans would prefer a larger social safety net than Singapore currently has. The Singaporean system of a basic net, as you describe it, relies explicitly on an authoritarian system where politicians openly tell their electorate on the front page of their national newspaper that they know better, and are powerful enough to keep suppressing naïve populist movements (which exists, even in Singapore).

    It may be economically more sensible in the long run, but I’m not sure any libertarian would endorse it, however pragmatic. Except Caplan, I guess.

  7. Gravatar of TGGP TGGP
    18. July 2009 at 22:57

    Women and minorities now have more opportunities, and thus may not feel quite so powerless. Is happiness what really matters?
    Minorities have gotten more happy, but women LESS. Happiness surveys used to put women above men. The first paper on Justin Wolfers page is The Paradox of Declining Female Happiness.

    As a Stirnerite egoist, I say the real question is whether a policy makes ME better off. Because foreigners have far less political influence on our government, current citizens may decide to better themselves at the expense of foreigners. The reason I don’t support trade restrictions is because that would actually make me worse off. I do have one reason to be worried about inequality though, which is discussed in Amy Chua’s World on Fire. I wrote about my fears of a two-tiered society here.

    I actually thought Will was a Rawlsian (or “Rawlsekian”, to show his appreciation of the power of markets combined with maximin ethics) rather than a Nozickian. Yet his Kidman example is just a modern version of the Wilt Chamberlain argument from Anarchy, State & Utopia (which I haven’t actually read).

    rob: Some liberals can argue that it is fine to tax positional goods, because it doesn’t actually make us worse off. But that doesn’t serve as an argument for redistribution. Let’s simplify by considering a set of two people: a billionaire and a hobo. If we think absolute wealth is what’s important (though declining in marginal utility) we can take a large cut from the rich man that won’t make him much worse off and give the hobo food, shelter, medical care and whatnot which will make him far better off. But if it’s really position that’s important, than the hobo is still on the bottom and the billionare on top, so we haven’t changed anything. You can’t improve a set of ordinals.

  8. Gravatar of johnleemk johnleemk
    19. July 2009 at 00:52

    “How can people on the left think these Cato guys are part of an evil organization?”

    Don’t forget the dogmatic libertarians who accuse Milton Friedman of being a statist — with the rise of Ron Paul it seems like they’re everywhere on the internet, and most of them don’t have good things to say about Cato. I don’t agree with everything Cato puts out, but what I like about them is that they have a healthy mix of pragmatic and dogmatic libertarians.

  9. Gravatar of Dan Dan
    19. July 2009 at 01:31

    One thing I would say is that “the theoretical argument for income redistribution seems powerful; a dollar probably is worth more to a poor person than a rich person” is actually very less clear cut than you imply. David Schmidtz has a paper called Diminishing Marginal Utility and Egalitarian Redistribution (it’s in the Journal of Value Enquiry, vol 34 and also in his book the Elements of Justice) which absolutely destroys the theoretical argument at least; the upshot is that it ignores the fact that we live in a world of production as well as consumption, and that the differing marginal propensities of the rich and poor to consume have an impact on utility.

  10. Gravatar of StatsGuy StatsGuy
    19. July 2009 at 03:05

    The core ECONOMIC arguments around massive inequality has always been incentives… We know these well, I won’t repeat.

    The core POLITICAL argument around massive inequality is that even if we could generate more happiness by starting with a different “initial wealth distribution” in our models, the political costs in achieving this are destructive because the wealthy classes will devote vast resources to distributional activities (e.g. preserving their wealth). Thus, society comes up with a compromise… (But even in this compromise, some sides are advantaged by the international aspects – trade and capital flows, and in particular the relative mobility and specificity of factors of production. Note how fears of capital flight are constraining monetary policy.)

    The core MORAL reason why we do not redistribute wealth is that rich people DESERVE their money, because they (or their forebears) somehow contributed massively to the welfare of other people. Those other people therefore OWE rich people a debt that rich people claim by spending money. (Rich people, btw, have a massive psychological bias to believe this argument – generally attributing a far lower portion of their success to luck than skill or hard work. Others have always doubted this argument, but watching the financial crisis unfold has further undermined it.)

    But it’s not just rich people who suffer from the bias to believe they DESERVE their wealth. So does every American.

    Imagine what would happen if before every time a person bought a Starbucks coffee a little angel whispered in their ear “that could feed a starving child for a month”.

    Or if before you went out to an expensive dinner, an angel whispered “that could drill a well to provide clean water to a village for a year”.

    Instead, society (through media) tells us that we deserve to spend money on ourselves. Otherwise, commerce would plummet.

    But, fundamentally, the MORAL argument for wealth inequality such as we observe today is sadly weak. The ECONOMIC argument (incentives) justifies some inequality, but probably not the type of extreme inequality we have. It’s really the POLITICAL argument that matters (if we tried to equalize things, it would make the world even worse).

  11. Gravatar of StatsGuy StatsGuy
    19. July 2009 at 03:20

    WHY ECONOMISTS SHOULD fear immigration:

    Chicago economists have a well-known pro-immigration and pro-free-trade bias, which I will argue is fundamentally a political argument. These aspects of internationalization are useful tools to weaken state power. Period.

    Free trade undermines the ability of society to pass laws that compensate for externalities that occur in the production of goods/services, as well as laws that fix other well known market failures. This is because it creates a “race to the bottom”.

    Free trade can be fixed, however – “simply” apply tariffs to incoming goods and services which reflect the additional costs to production of the good or service due to violations of US regulations (e.g. labor code, environmental code). Chicago economists have yet to propose a good counterargument, but many viciously hate this argument because it would undermine precisely the thing they like about free trade (that it punishes the US for trying to pass labor/environmental regulation).

    Immigration has a huge _economic_ problem however. It’s an incentive problem, you see, and it comes down to this: you may think it’s great to “share the wealth”, but doing so creates a disincentive for communities to invest because they will capture only a small portion of that investment.

    The same arguments that apply to Chicago arguments for decentralization of government services _also_ apply to the anti-immigration debate.

    Cato economists like to point out that those liberals who argue for “share the wealth” at the national level argue against it when it comes to immigration. But Chicago economists are no less hypocritical – perhaps more so – in arguing for “free” trade and open immigration, but arguing against internal wealth restribution and centralization of govt services.

    The success of Chicago economics has less to do with its intellectual sophistication, and more with the fact that its conclusions are useful to powerful elements of society.

    Of course, Chicago economists would argue that the DESERVE their success. 🙂

  12. Gravatar of malavel malavel
    19. July 2009 at 04:27

    If we want to maximize happiness we should try to maximize the population. The Maltusian trap for everyone. Doesn’t sound very appealing.

  13. Gravatar of Bill Stepp Bill Stepp
    19. July 2009 at 05:52

    Wilkinson cites a study by Stevenson and Wolfers that shows that happiness has become much more equally distributed since the 1970s, exactly the opposite of what you’d expect if you relied on the trendy left-wing narrative […]

    Are you serious? The goal of economic action is not happiness, it’s what Mises called the removal of felt uneasiness. Happiness has hedonistic or other qualitative implications that have no place in economics, the science of human action.

  14. Gravatar of Will Wilkinson Will Wilkinson
    19. July 2009 at 07:47


    I think policy (the taxes + the transfers, structured in a certain way) that actually helps the least well off is part of a system of fair rules. So when I say, “these myriad choices take place within a framework of political, legal, economic, and social institutions and norms””all of which affect the eventual pattern of incomes,” I’m already including things like fiscal policy. These are the sorts of things that we’re evaluating when deciding if the rules and constraints are just. I assume Nicole Kidman paid a lot in taxes. I think she probably paid too much, but I think she is in fact entitled to what she ended up with. The point in that passage is to say that the pattern of incomes emerges from tons of choices according to certain rules and constraints. The pattern isn’t what we should be evaluating. It’s the rules and constraints. When Hayek says he agrees with Rawls, he’s agreeing with him about THIS, the procedural nature of the subject of justice. When Hayek says he disagrees with Rawls, it’s when Rawls has forgotten this and starts worrying about the properties of the pattern itself — about “justifying inequalities” — in a way that doesn’t make a lot sense given the overall structure of his argument. This is also what Nozick picks up on (and caricatures in a way Hayek doesn’t) in his discussion of patterns in ASU.

  15. Gravatar of Will Wilkinson Will Wilkinson
    19. July 2009 at 07:52


    “The goal of economic action is not happiness, it’s what Mises called the removal of felt uneasiness.”

    Speak for yourself! Anyway, how is “felt uneasiness” any less qualitative than happiness? It isn’t! And how is reducing the experience of a negative emotion somehow categorically different than increasing the experience of a positive one? It isn’t!

  16. Gravatar of Will Wilkinson Will Wilkinson
    19. July 2009 at 08:10


    “Because as you point out, they keep making these ridiculous Kidman arguments: that people get what they deserve. Most people only tend to feel that way when they see someone taken away in an ambulance, not a limousine.”

    And, as I point out, I have argued against desert arguments over and over and over again. I think Hayek is right! The pattern of incomes tracks supply and demand in labor markets, not virtue, merit, deservingness, or whatever else you would like the pattern to track.

    Here’s an example. New technology has disproportionately increased the productivity of highly-educated workers. Workers who produce more in an hour get paid more per hour. Also, demand for these workers is high relative to their supply. This pushes up their wages further. So the wages of highly-educated workers has gone up at a much faster rate than less-educated workers. People surely deserve some credit for working to get an education. They’re responsible for that. But the fact that new technology especially complements their skills is something they’re not responsible for at all and don’t deserve. (This is complicated somewhat by the fact that people often choose skill training based on price signals in labor markets, which signal demand, and so often are in an important way responsible for the match between their skills and technology, but set this aside.) Also, the fact that the U.S. has strict limits on the number of work visas for skilled workers acts as a subsidy for native skilled workers. Their wages would come down if we let more foreigners in. They don’t deserve this subsidy either.

  17. Gravatar of TGGP TGGP
    19. July 2009 at 08:12

    david, I would endorse Singapore’s system or something even more radical. Democracy has NOTHING inherently to do with liberty. The system of maximum liberty (theoretically) would be anarchy, in which there are no politicians to elect or referenda to vote on.

    johnleemk, I resemble that remark! I used to hang out a lot at the Mises blog before they banned me. I think they make a lot of good points, even though Cato & Friedman (no more a statist than Ron Paul, I’d estimate) are/were really alright. When they slip up, it is only right that they be criticized. I think Sam Dolgoff had something to say about that sort of thing.

    StatsGuy, you throw around the word bias, not indicating that people are making inaccurate predictions, but that they don’t share your ethical positions. Couldn’t one just as easily claim that you are biased for not sharing their ethical positions? I don’t see how either side to such a dispute could have any evidence.

    There are some environmental externalities that are global in scope (carbon emissions). But many are about preserving the local environment, and so an importing country has little reason to care about the effects, which apply to the exporting country. Labor codes don’t involve externalities at all.

    I didn’t quite get your point about decentralization/centralization of government services. Could you elaborate?

    The success of Chicago economics has less to do with its intellectual sophistication, and more with the fact that its conclusions are useful to powerful elements of society.
    John Henry made a similar argument (though starting with the Whig classical economists) in The Ideology of the Laissez-Faire Program. Luigi Zingales made that argument when it comes to Keynesianism. John Wood & Greg Mankiw have both argued that academic macroeconomics has had little influence on policy, but rather that academics develop theories to justify existing policies.

    malavel, it depends on whether you’re an average or total utilitarian (assuming you are a utilitarian) different people have different Repugnant Conclusions. I believe Robin Hanson and Bryan Caplan are both favor population expansion. Hanson’s “hardscrapple scenario” is a particularly extreme version of that vision, though one I endorse over Eliezer Yudkowsky’s singleton.

    Bill Stepp, some “economists” are going to be interested in it anyway. Perhaps the field could be called “economics and hedonics”?

  18. Gravatar of ssumner ssumner
    19. July 2009 at 08:59

    Pointer, I don’t have a firmly held view of where either culture or institutions come from. I think institutions obviously reflect culture in some places, and just as they obviously don’t in other places (such as the two Koreas.) I do think that high trust cultures have an easier time building better free market institutions. But I am by no means a determinist in this area. Cultures can change over time, and external pressures (as in Korea) can dramatically affect institutions. Another example is the EU, which seems to have improved the Spanish and Irish economies, perhaps because they were pressured to adopt EU practices.

    One interesting question is why do the Nordic countries have high trust societies? Is it because small bands of settlers were forced to work together for the common good during the long cold winters? Who knows. It’s probably much more complicated.

    I think free markets tend to instill trust, and trust tends to instill free markets. I’m sorry that I can’t really answer your question, as I haven’t read the relevant literature, and the issues are so interconnected that I have trouble making sense of it all.

    Will, Thanks for responding. I had a nagging feeling that I might be slightly misinterpreting the Kidman quotation. But what you say here isn’t actually that different from what I assumed you meant. Rather the problem is that I wasn’t able to clearly convey my ideas very clearly in the post. I knew you didn’t mean “deserve” in some cosmic sense, and I shouldn’t have used that term. But I’m not even comfortable with the term “entitled,” unless it means nothing more than that she isn’t violating any laws. If I understand you right, the idea is that we should set up some sort of simple but fair constitution, which treats people equally, and then let the market determine incomes. But why not write a constitution that is silent on tax progressivity? Don’t get me wrong, I don’t favor a lot of redistribution, but not for any sort of broad philosophical principles. Rather for (supply-side) pragmatic reasons.

    I can’t decide whether I’m a utilitarian because I can’t decide what the term means. I favor the societal set-up that makes the total expected well-being of society highest. But that doesn’t mean I think each public policy issue should be decided on a case by case basis on utilitarian grounds. I think what I favor is called “rules utilitarianism” which in my case ends up leading me to much the same libertarian policy views as most non-utilitarian libertarians.

    Inequality proved too much for me to handle. The post was already too long, but I felt that it would need to be three times longer to really get my points across. And I don’t know the relevant philosophical literature like you do. But I will look at the papers you sent and perhaps add some comments when I am better informed.

    rob, A lot of liberal economists do look at things that way. I believe Robert Frank wrote a book using that idea. But although the view seems very plausible, even to me, its policy implications don’t seem to hold up very well. Wilkinson’s essay on happiness shows that policies of income redistribution don’t seem to boost happiness, even though the evolutionary psych perspective suggests they should.

    Rob, That’s a very good point. The Canadian government likes to buy drugs from US firms at the cost of manufacturing, not necessarily including all the R&D that went into drug research that didn’t pan out. How good would Canadian health care be if they had to rely just on drugs invented in Canada?

    david, That’s a very good question, but I disagree for several reasons. First of all if Singapore is not democratic, then it is almost impossible to know what people really believe. In Eastern Europe polls often show people opposed painful economic reforms that required then to restructure on market lines. But did they really? Then why do polls also show that Eastern Europeans overwhelming wanted their country to join the EU, which meant that they would have to adopt the sort of practices that made Western Europe rich?

    2. Why do so many people want to move to countries like Singapore? I have students tell me that they want an easy “A.” Do they really want to go to a college where everyone gets a “A”, and the college’s reputation goes down the tubes? There is the fallacy of composition to consider. Singaporeans say they want some benefit, but what about higher taxes? Do the poll questions include taxes? In America lots of people say they want more government spending on health care. But if you frame it in terms of higher taxes to pay for that care, the answer changes dramatically. I recall one poll that didn’t tell people how fast Medicare was growing, but offered various rates as options, and only 2% favored growth as fast as what actually happened (around 8% or 10% I think.).

    3. Singapore is not an open democracy. But it does have contested elections and news sources like The Economist say the government probably does actually win about 70% of the vote. Again, I’m not saying the other party has a fair chance, but its not like Stalinist Russia either. The electorate does seem to have a grudging respect for the government’s expertise.

    4. Suppose Singapore became a pure democracy, and the other party had a fair chance. Due you think it would be elected on a policy of much higher tax rates? Especially when Singaporeans can look around their neighborhood and draw the obvious conclusions? Look what trouble the Dems in New Hampshire are having getting an income tax, even though they now control the government. They may get it in the end. But it isn’t easy.

    5. Even if you are right, the purpose of blogs is not simply to reflect opinion, it is to shape opinion. Laffer didn’t give up when people told him Scandinavians like 80% top tax rates, he changed opinions and changed the world. I want people to think about what Singapore has achieved. Why shouldn’t we look at successful economic models and learn from them? If Singaporeans read lots of articles about how they have the most successful economy in the world, it may instill a sense of pride and make them less likely to change things drastically when they do have an open democracy.

    TGGP, I’m not an egoist, but if I was I’d still favor immigration, as it makes me better off.

    You may be right about women, but I was reporting what Wilkinson said, and he read the Wolfers research.

    I may add comments later when I have time to read your paper and Wilkinson’s.

    johnleemk, I agree about Cato. But I would think left liberals would like pragmatic libertarians more than dogmatic libertarians. At least pragmatists like me basically share the same values as left liberals. But to be honest, I think most people on the left don’t even understand the distinction, and simply assume all libertarians are dogmatic libertarians.

    Dan, I once read a book called something like “The uneasy case for progressive taxation” which did many of the things you mention. But I still think that none of those right wing attacks are convincing without supply-side economics backed up by empirical studies. And it is really, really hard to do those studies well, because supply-side effects can take decades to show up. So I’m not sure their arguments demolished redistribution quite as thoroughly as you claim. It’s not easy to do.

    Statsguy, Good points. I believe Americans reached the right view on redistribution for the wrong reason (i.e. the conservative “just deserts” argument.) Perhaps there is a reason why the richest big country in the world has a strong Puritan streak.

    Statsguy, Ricardo demolished the “race to the bottom” view 200 years ago, so Chicago economists don’t need to worry about that issue. If there is a global externality problem like global warming, you might have an argument. But surely most protectionist arguments over the years haven’t been based on global warming.

    Immigration isn’t about “share the wealth,” otherwise the US, Canada, and Australia would not be three of the most prosperous countries in the world. I do think there are reasonable arguments against completely open immigration, or for a balance of countries that better reflects the world population (i.e. Asia), or for a better balance between skilled and unskilled (which would hurt me, BTW.) But crude arguments that immigrants hurt our economy don’t seem well-supported by the facts.

    You said:

    “The success of Chicago economics has less to do with its intellectual sophistication, and more with the fact that its conclusions are useful to powerful elements of society.”

    So let me get this right. Chicago economists have won far more Nobel prizes than economists from any other school not because there have done important work, but because those right-wing Swedes are trying to prop up the capitalist classes? The same Swedes who just gave a Nobel prize to Krugman?

    malavel, Yes, that is a great problem for utilitarianism that I mentioned in an earlier post. And here is a great irony here. The one group that seems to favor a big population policy (the Catholic Church) is among the least utilitarian institutions in the world. Go figure.

    Bill, I don’t even know what happiness is, so I’m not going to argue with you. My point is that mainstream economic theory treats happiness as the goal. So if this is their view, then why aren’t they achieving success according to their own criterion?

    I need to do another post, so will catch up to comments on earlier posts tonight.

  19. Gravatar of ssumner ssumner
    19. July 2009 at 09:14

    Will, After reading your reply to TGGP, I feel that we are closer than I had thought. I agree that in some sense Kidman is entitled to keep what she has earned, after she has paid her taxes. I originally misinterpreted the “entitled” comment as meaning something closer to “entitled not to have to pay tax money which would go to help the poor.”

    Rawls’ thinking struck me as horribly confused. Am I right that the minimax principle is really nothing more that the idea that the welfare of the worst off member of society is all that matters? I know Rawls would deny that, but if we assume that any significant public policy has at least a trivial impact (say one cent) on each person, then minimax would seem to reduce to nothing more that “focus only on the very poorest.” Something that could justify Maoist economics.

    If Rawls thought that we needed a “do over” to redistribute income even after taxes had been paid, than that further diminishes my view of his economics.

  20. Gravatar of StatsGuy StatsGuy
    19. July 2009 at 09:38

    Very quick comments:

    1) By bias, I mean deviation in views of the wealthy from the average view (in so far as to the degree that wealthy people truly deserve their wealth). So either the national average is biased, or the wealthy are biased.

    2) The decentralization argument is at the root of the markets vs. command economy debate – there’s an information component and an incentive component.

    3) I think Ricardo’s rejoinder to the race-to-the-bottom argument is still very much an unsettled debate.

    4) There’s a legitimate argument about labor regulation in the context of shifting the bargaining power between firms and employees, and the externalities associated with investments in human capital. The other arguments are mostly distributional, although distributional arguments (at the extreme) have externality components as well (e.g. crime rates in depressed neighborhoods).

    5) US immigration, since the Great Depression, has been extremely limited and highly strategic, focusing primarily on critical skills and capital. In that context, it has overwhelmingly benefited the US (and most people in the US) by filling in complements to existing factors of production. Note the public response, however, when immigrants begin to act more like substitutes and less like complements. We’ve not had free immigration for decades (indeed, even during the Ellis Island waves, we turned away a vast proportion of people who were viewed as unhealthy or defficient).

    6) Chicago School – please note that this was mostly tongue in cheek (hence the smiley-con). The Chicago Boyz have certainly pulled their weight and more… HOWEVER, let’s also note that Milton Friedman didn’t win his Nobel until 1976, MANY YEARS after his theories became politically popular and useful to certain political constituencies. Nobels tend to follow social popularity/acceptance of theories, not lead them. Thus, it’s hard to disentangle what portion of Chicago’s success comes from usefulness to dominant political factions vs. true intellectual contributions.

    But… it’s not my intent to annoy anyone too much, just to raise some questions.

  21. Gravatar of Taylor Taylor
    19. July 2009 at 09:42


    I’m not a card-carrying utilitarian, but I think we might as well use utilitarian methods for policy evaluation until we come up with something better.

    There are two major problems with this:

    1.) You can’t compare utility because utility is subjective. You have no way to judge whether the wealth taken from a rich man in taxes causes a smaller loss to his utility/happiness than the gain that a poor man receives in having the wealth bequeathed to him. If you can’t measure utility, you can not successfully base policy off of it and expect it to make everyone better off.
    2.) You can find yourself in a horrible situation whereby, say, the Stalinist regime has calculated with their utility functions that if they murder 10million Kulaks, everyone else would be better off. Are you ready to defend such an action on utilitarian grounds? Are you ready to write off the theft of people’s property right in their lives, simply because some utility comparison someone came up with yielded that as the conclusion?

  22. Gravatar of TGGP TGGP
    19. July 2009 at 11:12

    It’s just a short step from evaluating patterns to evaluating rules. I haven’t read Rawls in the original (since I haven’t read Nozick, that makes me fair & balanced!), but his minimax principle seems to say that if the outcome of the least well off can be improved by changing the rules, than we ought to change them. There is no “after tax do-over”, it’s just setting a higher tax rate in the first place. Our current tax system was not the result of a contract made by disembodied somnambulists behind the veil of ignorance. The Rawlsian question for you and Will would be, what OUGHT Kidman pay in taxes? I thought I once heard it said that Rawls really meant the rules of the economy ought to be structured in such a way that before-tax income should go more toward the least well-off, but whatever.

    StatsGuy, since 1965 (or perhaps the 80s) immigration took a very different turn. Some countries have highly selected immigrants, but the largest share of immigrants nowadays are relatively unskilled ones from Mexico. I noted in an earlier comment that there are massive gains from trade for even that sort of immigration, and so I don’t favor merely shutting it down (it might even be expanded). But I think the way we do things now is far from optimal.

    Something that could justify Maoist economics.
    Maoism is a horrible system, including for the very poor. They would be better off in a more free-market system. Note that even Haitians don’t free to Cuba (although I think that regime technically might not be Maoist, but rather some other Stalinist variety).

  23. Gravatar of johnleemk johnleemk
    19. July 2009 at 11:21

    “But to be honest, I think most people on the left don’t even understand the distinction, and simply assume all libertarians are dogmatic libertarians.”

    Scott, I agree — and it doesn’t help that many dogmatic libertarians view pragmatic libertarians as too statist. Simultaneously, I wouldn’t be surprised if there are quite a few “liberals” who are actually pragmatic libertarians in the sense that they desire more liberty than the mainstream Democrats, but don’t identify as such because they think the only option is dogmatic libertarianism.


    I agree with both your points, but neither are particularly original critiques of utilitarianism. John Stuart Mill’s work focused on addressing #2 about a century and a half ago. And as for #1, while this is true for many cases, you can easily come up with many counterexamples. The Prisoner’s Dilemma is a classic one.

    I agree that utilitarianism is not a fully workable philosophy, but I think it represents an ideal that is often at least a good starting point for thinking about policy problems.

  24. Gravatar of Jeremy Goodridge Jeremy Goodridge
    19. July 2009 at 12:07


    There are other utilitarian arguments against progressive taxation besides the ‘incentive’ effects. I think economists too often miss these and put all their eggs into the ‘incentive’ argument. As a result, they end up being vulnerable to Saez and others who do empirical research that suggests a weaker incentive effect, particularly with labor effort.

    Just to be clear before giving those arguments, I should say that I favor a completely flat tax covering ALL taxes not just the personal income tax. It should include business income taxes and social security taxes (which are now actually REGRESSIVE). And even better, this flat tax rule should be in the constitution — so progressive taxation would actually be illegal forcing politicians to raise taxes if govt needs to grow.

    I think the most important of these other arguments is the ‘capital accumulation’ effect. Basically the capital accumulation effect arises because high marginal tax rates block people from accumulating capital as quickly and thus reduces new business formation or at least causes business formation to rely more on external capital markets. Part of the reason economists miss this is that they think capital markets are so effecient. But are they? We certainly have no way of KNOWING that they are. And the recent financial crisis suggests they aren’t that great at finding terrific entrepreneurial opportunities on behalf of investors. My bet is that if people could keep more money, the overall quality of investment would rise. And much of that investment originates with people who have higher incomes and consequently higher marginal rates.

    Related to this is that govt usually redistributes moneys it taxes from people who are more likely to save/invest to people to who are more likely to consume and not invest. So, the overall savings/investment rates are made lower by progressive taxes.

    There are other less critical, but still non-zero, arguments against progressive taxation. One is the ‘complexity’ issue. I think one could reasonably say that if we had an enforced flat tax, the number of deductions and exemptions would be fewer and they would tend to be much more broad based. It’s high tax rates that encourage the development of deductions and credits. And these then cause all kinds of distortions and wasted effort at tax evasion. Another ‘complexity’ issue is that taxes are, for the most part, collected based on a year’s worth of income. With progressive taxation, it’s possible for someone who earns 50,000 one year and 150,000 the next year to be taxed differently from someone who earns 100,000 for both years. Of course, the designers of the code can prevent this, but it takes some effort and ultimately results in a lot more complexity.

    Another argument against progressive taxation is the ‘big government’ argument. I suspect that progressive taxation allows govt to grow bigger than it would ordinarily be. If the middle class can get more health care on the backs of the upper-middle class and wealthy, then they are much more likely to say ‘yes’. I we had a constitutional requirement that all income needed to be taxed at an equal rate, politicians would be forced to either borrow money, or to ask the middle class to make a judgement — “do you want health care or not for this particular price”. It would be much harder to object to a big government with a flat tax — after all it would mean that the avg person basically voted to be taxed more.

    And finally, there are real moral risks to a progressive tax — it gives the majority the power to confiscate the wealth of a few. I have much less of a moral problem with the majority of people voting to tax themselves to help the poor (I would myself vote ‘yes’ for that). But I have a lot bigger of a problem with the majority voting to tax the few to give themselves more benefits. And the main reason is that there is no limit to that — the majority get to decide when too much is too much. Maybe they will be rational, or maybe they won’t.

    To be sure, the basic empirical argument Krugman would give against all this is this (the words are mine not his): “we grew a lot in the 50’s and 60’s, a period with high minimum wages, strong unions, and a highly progressive tax code. Thus those things dont hurt growth, and maybe even help.” Of course, the argument is empirically meaningless because it involves ONE data point. Lots of other things were true too: (1) govt was smaller (state,local, federal together), by 5-10 percentage points of gdp. (2) the percent of federal spending going to defense was also higher — 10% of gdp versus 5% of gdp. I doubt he would point to these issues.


  25. Gravatar of Will Wilkinson Will Wilkinson
    19. July 2009 at 12:38

    You say…

    “Rawls’ thinking struck me as horribly confused. Am I right that the minimax principle is really nothing more that the idea that the welfare of the worst off member of society is all that matters? I know Rawls would deny that, but if we assume that any significant public policy has at least a trivial impact (say one cent) on each person, then minimax would seem to reduce to nothing more that “focus only on the very poorest.” Something that could justify Maoist economics.”

    Rawls is actually talking about the least well-off class. So, basically, the working class, labor. As I understand it, about the time Rawls was writing the stuff that ended up in ToJ, the economics profession was transitioning in a big way from a functional conception of distribution — this much to labor, this much to capital, this much the the landed, etc. — to the now completely dominant idea of individual income distribution. A lot of political theorists writing about inequality seem not to realize there was a massive paradigm shift in thinking about income distribution. Anyway, as far as I can tell, Rawls minimax concerns the least well-off class, not individual. Anyway…

    The difference principle is often misinterpreted by economists in the way you seem to approach it. It is the second part of Rawls’ second principle of justice. The first principle, a version of the principle of equal liberty, has to be satisfied before moving on to the second (it is “lexically prior”). The difference principle comes up after we’ve established a framework of rights and liberties, and is, in part, supposed to insure the value or worth of those liberties to the least well-off class. (So that they aren’t “merely formal” in the way Marxists always complained about.) So we’re picking among schemes that satisfy the first principle. The idea is, within that set, and pick the scheme that leaves the least well-off class best off. That’s the one “primary goods”-maximizing rational agents would pick behind a veil of ignorance. It’s still a crazy principle if applied strictly. (One reason Rawls’ protege Scanlon gave up on this original position stuff and talks about something much more vague, but more realistic: find rules that people of good will and not too much partiality would not “reasonably reject.” It is not unreasonable for wealthy people to reject a scheme that would make them dramatically worse off for only a tiny benefit to the poor. Anyway, even given the exotic Rawlsian analytical machinery, I’ve never been convinced that maximin would be preferred over a scheme that simply maximizes expected primary goods, but with a floor that ensures a minimum level of these goods. This is, in fact, what people tend to prefer when you try to simulate the original position experimentally.

    “If Rawls thought that we needed a “do over” to redistribute income even after taxes had been paid, than that further diminishes my view of his economics.”

    Yeah, he didn’t think that. He stresses the difference between “distribution,” which the the pattern that emerges from the basic rule, and “allocation,” which is the tax and transfer corrective “do over”. So Rawls preferred basic rules that minimized the need for reallocation via fiscal policy. The idea that the difference principal is primarily about tax rates and the generosity of transfers is a very widespread misinterpretation of Rawls. He explicitly rejects what he calls “welfare-state capitalism” in favor of “property-owning democracy.” His claim is that is that welfare-state capitalism (Denmark is almost the ideal type) allows too much inequality in the ownership of the means of production, and this makes the owners of capital too politically powerful to ensure the value of everyone’s rights to equal democratic participation. The idea is to prevent large inequalities in wealth and capital from arising in the first place. The picture he paints is a VERY heavily regulated economy, with limitations on individual capital-ownership, state-sponsored education, publicly funded elections, very strict limits on bequests, etc. When one looks at the actual world, this kind of scheme looks terrible. It does much, much worse materially than Danish style welfare-state capitalism, and not a bit better in terms of the kind of democratic equality Rawls is really concerned about. The two systems he endorses as satifying his principles of justice, property-owning democracy and liberal socialism, are extremely prone to the kind of regulatory capture and concentrations of political power characteristic of dirigiste economies. It seems like he and his disciples wave off the actual social science as dealing with “non-ideal theory,” but there’s no evidence that the kind of system he wants can work. The system that seems to actually do the best relative to his basic standards is re-allocative welfare-state capitalism. Laissez faire on the front end to finance redistributive education and social insurance at the back. If the Singapore model is democratically viable (not clear to me), it might be even better.

  26. Gravatar of Cauchy and Schwarz should be ashamed of themselves!! « This Blog Has No Theme Cauchy and Schwarz should be ashamed of themselves!! « This Blog Has No Theme
    19. July 2009 at 14:01

    […] should be ashamed of themselves!! By darfferrara Another Sunday brings another fantastic post from Scott Sumner.  I started reading his blog around the same time that everyone else did, after […]

  27. Gravatar of Rob Rob
    19. July 2009 at 15:02

    I finally got around to reading the actual paper, and Will makes a lot of very good points. However:

    “Suppose you made a million dollars last year and put all but $50,000 of it in a shoebox. Now imagine you lose the box. What good did that $950,000 do you? Maybe it purchased some temporary peace of mind. It’s certainly reassuring to know that you have resources at your disposal. But it likely did rather less for your well-being than did the $50,000 you spent on housing, food, entertainment, healthcare, transportation, gadgets, toys, and so on.”

    Unless you’ve lost many nights of sleep worrying about your debt and how you’re going to pay it, or worried about how you are going to pay your bills if you lose your job tomorrow, I suppose you may not appreciate the significance of that “peace of mind” the $950,000 in the shoebox gives you. I don’t think it’s far-fetched to say that sleeping well at night can be the difference between happiness and unhappiness. It’s the difference between living in fear and not. One might ask: If peace of mind is worth so much, why do so many consume themselves into debt instead of saving? I’ll argue that people must consume a certain amount in order to keep up with societal norms in order to compete for mates–AND INCOME INEQUALITY LEADS DIRECTLY TO THIS OVER-COMPENSATION. If you forgo the decent car, the HBO, the wining and dining, the expensive vacation, the toys, etc. — you will have a harder time competing with those who don’t. You may not sleep well at night, but peace of mind takes a back seat to evolutionary imperatives. Moreover, that peace of mind is tied to real risks for the future. (We might want to track the divorce rate over the next couple years.)

    I do agree, however, that the attempt at a cure may be far worse than the disease.

    (Sorry for the caps. I don’t know how to bold or italicize here.)

  28. Gravatar of Rob Rob
    19. July 2009 at 15:05

    I meant over-consumption, not over-compensation. Freudian?

  29. Gravatar of happyjuggler0 happyjuggler0
    19. July 2009 at 17:16

    Sorry for the caps. I don’t know how to bold or italicize here

    To make the word “italics” in italics, type [i]italics[/i], except use arrows “” instead of brackets. For bold, do the same thing except use b instead of i.


  30. Gravatar of happyjuggler0 happyjuggler0
    19. July 2009 at 17:17

    Um, ignore the “” in my previous post.

  31. Gravatar of TGGP TGGP
    19. July 2009 at 18:04

    I’ve never been convinced that maximin would be preferred over a scheme that simply maximizes expected primary goods, but with a floor that ensures a minimum level of these goods.
    Yeah, that makes much more sense to me. Though I think a utility function that places great (marginal) weight on the minimum could substitute for the floor.

    Will, would you say that Rawls’ “property-owning democracy” resembles the distributism of ChesterBelloc? I’m not too well informed about either, but they sound similar to me.

  32. Gravatar of ssumner ssumner
    20. July 2009 at 06:39

    Statsguy, Rich Americans voted 52-48 for Obama, so I don’t think their views are much different from other Americans.

    You are right that there are more complex arguments for trade barriers. But I’d make two comments:

    1. Most people don’t understand those arguments; when they talk about a race to the bottom they have in mind a theory that was discredit by Ricardo. Read Krugman’s Pop Internationalism. (Did I just recommend Krugman?)

    2. The empirical importance of these complex models is probably trivial. I am pretty sure that American workers would suffer badly if we became protectionist. It’s not a zero sum game, and it’s pretty hard for a very large group to get a bigger absolute slice when the total size of the pie is falling.

    Friedman’s contributions weren’t even accepted by his profession until much later than when he made them. I was at Wisconsin in the mid-1970s and I recall that Friedman was viewed as a nut.

    Taylor, All public policies are based on estimates of things we cannot measure. We can’t measure the effect on GDP of stopping global warming, but we still might want to have a policy. Instead we make estimates. Society has estimated that the poor benefit from an extra dollar more than the rich, and even people in the private sector (with no government coercion) act on those estimates every time they give to charity. So I do think it is possible to base public policies on guesses rather than hard facts.

    The other point is a common objection to utilitarianism. My response is twofold. First, I find it interesting that every real world objection I have ever seen to utilitarianism is not some public policy that raised utility but was moral offensive, but rather policies that almost certainly reduced total utility. That tells me that total utility might be a pretty reliable guide to public policy, since it is hard to find a counterexample.

    But your basic point of course is slightly different. You argue that a utilitarian procedure might (mistakenly) lead to atrocities that would not occur if we relied on some other procedure such as religion or natural rights. That may be correct. I would continue to judge current public policies on utilitarian grounds, but I would not object to “natural rights” being embedded in a constitution in order to prevent future Stalin’s from making the same mistake.

    [BTW, Stalin was not a utilitarian, I doubt he took into account the loss of utility for all his Kulak victims. But you point remains valid, as he might have been.]

    So where does this leave utilitarianism? In my view it makes it sort of tautological. We have a first amendment rather than decide the utility of each speech on a case by case basis. Does this violate utilitarianism? Who knows? You could argue it is still utilitarianism at a deeper level, as we think society would be happier if the government wasn’t deciding the utility of each utterance on a case by case basis.

    My state has a constitutional amendment against progressive income taxes. It might surprise you to know that I support this provision on utilitarian grounds.

    TGGP, All your points are good. You are right that Mao probably hurt even the very poor. But I was focusing more at what Mao was trying to achieve. I presume he wanted to help the poorest, but took no notice of lower utility for the better off. (Actually he favored punishing the better off even if it didn’t help the poor, so that is another difference from Rawls.)

    Johnleemk, I agree with your points. Brink Lindsey is trying to build bridges between liberals and pragmatic libertarians.

    Jeremy, Very good post. Those are all good points. Krugman has a fondness for the 1950s and 1960s, but Brink Lindsey does a very good job of critiquing that nostalgia. Krugman forget to mention that tax rates rose sharply in the 1930s, and fell in the 1920s, 1960s, and 1980s. And even in the 1990s the tax rate on capital was cut. Krugman also ignores what happened in the rest of the world, and doesn’t seem too curious as to why it happened.

    I completely oppose the income tax, even a flat income tax, but am OK with payroll taxes. I think you are right that a flat rate is better for all the reasons you mention, but I hate the income tax so much that I’d even prefer a progressive payroll tax to our current system. Ordinary workers like me shouldn’t have to spend many hours each year working on our tax forms. The taxes should be withheld by the government, just like FICA. And my other suggestion is that we should go with a forced saving plan to replace social programs. This would allow us to have very low tax rates.

    Will, Thanks for that very informative discussion of Rawls. I guess I am the typical economist who wants to reduce everything interesting in life to mathematical criteria. Even so, I still end up with the same problem with the maximin principle that you and Scanlon apparently have.

    I was vaguely aware of the importance Rawls placed on human rights.

    I have this bizarre view that the “behind the veil” thought experiment treats all of humanity as if it were a single individual. Thus a person doesn’t much care which brain cells in his mind are “happy” as long as he is happy, then if you didn’t know which person you would be, the analogous choice would be to maximize total utility. I think Borges once said that in a sense each person is every person. But this stuff is so far removed from reality as to be no help in public policy formation.

    Rob, That’s a good point. I think that Will would say that consumption is what matters, but a that stable consumption leads to more happiness than the same amount of consumption delivered with great uncertainty. The peace of mind aspect may be very important if consumption is highly unstable, as you suggest.

    TGGP, Rawls views do sound a bit like Chesterton’s. I wonder if anyone has pursued that angle, as one is a famous liberal and the other is a famous conservative.

  33. Gravatar of Current Current
    20. July 2009 at 07:55

    Well, this has been very interesting. I think though that folks are making all of this much more difficult than it need be. There is no question that these are difficult problems, but there are some simple and essential points missing from what folks are saying.

    In this post I’m going to say “Mises was right” about a few things. To be clear, I don’t think Mises was right about everything. My point is that in this area a lot of what he wrote makes sense, although most of it isn’t original.

    Firstly, on individual happiness. None of this is really about economics or Mises. Michel Houellebecq is certainly right that extra-economic individual circumstances are important. The concerns he mentions about sex drive are only the tip of the iceberg. As another poster has mentioned mental health is very important. From my personal experience it seems to me that the sort of person that is always beset by worries is always in a poor position, and that isn’t even a mental illness. Several of my friends are unemployed and have no assets. I think they are much happier though than people who suffer from worries.

    Regarding partners and sex, I find lots of the stuff folks have wrote unconvincing. The three men I know who are most successful with women are all unemployed and broke. I fail to see how income inequality leads to “over consumption” compared to what would occur otherwise. This is a Veblenesque argument and suffers from all the same failing as the rest of his arguments. Status exists independently of wealth. I am seen a middle-class engishman because of my mannerisms, my accent, and because I get all my jokes from “The Onion” and the english version “Newsbiscuit”. Most of all I am middle-class because of the style of my shirts. Income doesn’t play that much of a part. The idea that re-distributing income will make people happier by reducing the significance of status, or that it will reduce wastage due to conspicuous consumption is gambling with social policy on the basis of dubious psychology. Nobody really knows which way the pieces will fall.

    What I have written above is of only tangential relevance to economics. We cannot measure the happiness of a person. However we can ask if a person has the means to remove any of the felt uneasiness he suffers from or if other will impose more upon him. We can then ask what the institutional environment needed to best bring this about the ability to remove felt uneasiness. This is the utilitarian problem phrased in a reasonable way.

    As for contractarian views. These are all well and good, but I think they are a distraction. If a large scale contractarian society were to exist then a large set of the agents within it could choose to form a utilitarian society within it. If there really was a superior utilitarian society available then clearly most people would aim to do that. So, any thorough argument for a contractarian society must really be an argument for why such a society is utilitarian.

    Regarding bias, this discussion is based on a confusion. It doesn’t matter if I’m biased or you’re biased. What matters is: who is right? Rational argument is the only way forward.

    Statsguy, if you have an argument for why Ricardo and the Classical economists were wrong about trade, then what is it? Show us the money.

    Will Wilkinson, regarding “Misean Grassmowing”. Certainly Mises is wrong about Buddists in Human Action. But that is more a misunderstanding of what Buddism teachs than one of philosophy or psychology. I think Mises corrected this in later books.

    So, enough of the tangents, the main question here is “Is there a justification for not redistributing wealth”. The simple answer is “yes” and Jeremy points out the best argument for why: “Related to this is that govt usually redistributes moneys it taxes from people who are more likely to save/invest to people to who are more likely to consume and not invest. So, the overall savings/investment rates are made lower by progressive taxes.”

    Redistribution is intellectualized robbery of the future. Rich people can afford to have low rates of time-preference and to invest, poor people can’t. If I were taxed to fund the spending of my three unemployed friends (which of course I am) then they would spend the money on women, marjuana and beer. There is nothing wrong with that and I spend lots on that sort of thing myself. It is not a moral judgment but a simple acknowledgment of economic reality. Since I already have those consumption desires satisfied I can go onto others desires such as saving.

    It is exactly because this sort of redistribution did not occur in periods of the past that we are able to enjoy so much wealth now.

    Anyone interested in the “Aesthetics” of inequality should read T.B.Macaulay on this, it takes a little time to get going, but it’s great when it does…

    “Here is wisdom. Here are the principles on which nations are to be governed. Rose-bushes and poor-rates, rather than steam-engines and independence. Mortality and cottages with weather-stains, rather than health and long life with edifices which time cannot mellow. We are told that our age has invented atrocities beyond the imagination of our fathers; that society has been brought into a state, compared with which extermination would be a blessing; and all because the dwellings of cotton spinners are naked and rectangular. Mr. Southey has found out a way, he tells us, in which the effects of manufactures and agriculture may be compared. And what is this way? To stand on a hill, to look at a cottage and a factory, and to see which is the prettier. Does Mr. Southey think that the body of the English peasantry live, or ever lived, in substantial or ornamented cottages, with box-hedges, flower-gardens, bee-hives, and orchards? If not, what is his parallel worth? We despise those mock philosophers, who think that they serve the cause of science by depreciating literature and the fine arts. But if anything could excuse their narrowness of mind, it would be such a book as this. It is not strange that, when one enthusiast makes the picturesque the test of political good, another should feel inclined to proscribe altogether the pleasures of taste and imagination. “

  34. Gravatar of Current Current
    20. July 2009 at 08:06

    Scott: “The peace of mind aspect may be very important if consumption is highly unstable, as you suggest.”

    Just for fun, think about what that means for the stronger sorts of efficient financial markets hypothesis.

  35. Gravatar of Sumners One-Sentence Class Autobiography, by Bryan Caplan – Money Management, Financial Markets Sumners One-Sentence Class Autobiography, by Bryan Caplan - Money Management, Financial Markets
    20. July 2009 at 08:49

    […] years ago I asked economists to share their class autobiographies.  Yesterday Scott Sumner boiled his down to one sentence: At various times in my life I have been in all 5 quintiles of family income distribution, and yet […]

  36. Gravatar of StatsGuy StatsGuy
    20. July 2009 at 09:33


    “Statsguy, Rich Americans voted 52-48 for Obama, so I don’t think their views are much different from other Americans.”

    The limited amount of polling data on this topic are pretty clear, even though it does appear that Bush Co’s incompetence managed to alienate even his natural allies.

    For example:

    “In 1997, researchers at Roper Starch polled a national cross-sample of America’s most affluent 1 percent. Everyone surveyed made at least $250,000 in income or held $2.5 million in assets. These wealthy Americans were asked to agree or disagree with a simple statement: “I deserve all my financial success.” Nearly 90 percent agreed, 54 percent “strongly” and 32 percent “mostly.””

    “The public does not begrudge rich people “” a majority of Americans, 54%, say that those who make a lot of money deserve it”

    So 54% of the public believes the rich deserve it. But among rich people, 90% think they deserved their wealth. These are different polls, and the data are not directly comparable, but it’s pretty clear which way the wind blows. I’ve read other polls that I recall showing the same trend.

    The belief that the rich, fundamentally, do not DESERVE what they have undergirds populism. Populism is not a utilitarian movement (aka, “The rich earned it fairly, but we’re going to take it we get more happinness from it than they do at the margin…”) Populism is a moral movement.


    Other interesting point:

    “Just 34% of people in upper-income households [>75k in this poll] agree that money is the root of all evil, compared with 48% of those in middle-income households and 52% of those in lower-income households.”

  37. Gravatar of StatsGuy StatsGuy
    20. July 2009 at 10:01

    An AEI text, somewhat dated (1997) – But overall, same conclusions – a majority of Americans (~60% in many versions of the question) tended to defend current inequalities (unlike in other countries). But this masks some large disagreements across strata. (Americans also severely overestimated their chances of getting rich.)

    Can’t find any 2009 data. Some more recent data (2005, 2008) at Roper Center, for those who still have their academic passwords. I would bet a lot of money that all of these numbers have changed in the past 8 months.

  38. Gravatar of James James
    20. July 2009 at 16:31

    Scott, you write, “Now I think rich people should keep what they earn if and only if trying to take it away from them it will do more harm than good.”

    Do you think courts should allow this as a defense for private persons charged with theft? Or are you proposing a standard to be used for judging the actions of the state, with the unwritten caveat that different and less permissive standards should be used to judge the actions of private persons?

  39. Gravatar of ssumner ssumner
    21. July 2009 at 05:43

    Current, Thanks for the great Macaulay quotation.

    Statsguy, The two polls you mention are comparing apples and oranges. It doesn’t mean their attitudes vary by class. It would be like asking rich people “are you an above average driver” and then asking other people “do you think rich people are above average drivers?” Of course you’d get very different answers. But you get equally different answers if you asked rich people “do you think other rich people drive as well as you do?”

    The point is that everyone tends to overrate their own merit compared to that of other people.

    James, No I don’t view that as a valid defense. Theft is generally not beneficial from a utilitarian perspective (with a few obvious exceptions like a person dying of thirst stealing water from a rancher’s well.) This is because theft causes huge deadweight losses. In addition, I support rules utilitarianism.

    I should mention, however, that I don’t think we should take extra money from rich people, because I do believe the cost exceeds the benefit. And on the other hand if theft didn’t cause any deadweight losses then I would not object to it if transferred wealth to those who would enjoy it more (perhaps the poor.)

  40. Gravatar of James James
    21. July 2009 at 08:42


    I wasn’t asking if you’d accept that defense in every case. Is that defense not even valid if true? Why, then the special treatment for government? (You may believe that the state should not be judged by the norms applied to private actors, but can you get to your conclusion without explicitly assuming it?)

    Rules utilitarianism is equivalent to act utilitarianism unless you impose an arbitrary constraint on how fine grained the rules may be.

  41. Gravatar of Current Current
    22. July 2009 at 01:34

    James: “Rules utilitarianism is equivalent to act utilitarianism unless you impose an arbitrary constraint on how fine grained the rules may be.”

    That’s what rule utilitarianism is, it’s about general rules not arbitrary ones.

    That said, it’s still very incomplete. Ridiculous general rules can be created.

    This is why enormous books on political economy are necessary, to understand the problems properly.

  42. Gravatar of ssumner ssumner
    22. July 2009 at 05:47

    James, This is a complicated question. Let’s separate out several issues:

    1. Does it makes sense to examine anything on a case by case basis? Or should we always rely on broad rules?

    2. What about specific issues such as private crime? Or public decisions over tax rates?

    For crime, I think there are huge advantages to having one overarching rule, and not deciding on a case by case basis how much the thief needed the money vs the owner, plus the costs imposed by thieves as owners take precautions, etc.

    Income redistribution is a bit more complicated. I think there is a reasonable argument for redistribution–letting the government decide tax rates each year, and allowing progressivity. However, because I see a lot of practical problems with progressive income taxes, I favor the sort of constitutional prohibition on progressive income taxes that we have in Massachusetts (a very liberal state, by the way.) So in the end I am opposed to progressive income taxes, but not because it is unfair to the rich, rather because of supply-side arguments, and some of the others mentioned by commenters here. If studies show that the supply-side arguments are wrong (and they might be wrong) they I would change my mind.

    I think I am actually using the same pragmatic criteria in both cases, but you may see it differently.

    Current, I agree. I hope my answer to James gives a sense of where I am coming from, but I agree that politics are almost infinitely complex, and hard to fit into a simple set of rules.

  43. Gravatar of James James
    22. July 2009 at 08:35


    On 1) Max(u) subject to a limit on the number of rules results in a lower u than unconstrained maximization. If you favor the constrained maximization anyway, some considerations other than utility are affecting your analysis.

    On 2) I still don’t understand how you get to the conclusion that the range of acceptable activities is different for people in government than for others. E.g., suppose I am a tax collector. As I understand your view, it’s permissible for me to take your money without your consent so that the state may spend it. Suppose I resign tomorrow and become a thief. I again take your money without your consent and write a check to the treasury for the same amount. You seem to regard the former as good and the latter as bad. Assuming that I’m not misunderstanding you, why is the former good and the latter bad? What assumptions do I need to make to reach your conclusion?

  44. Gravatar of GNZ GNZ
    22. July 2009 at 22:06

    1) But there is a cost to doing this analysis in time and effort.
    2) Concequentialism and somthing other than “all groupings must always be treated exactly the same”.

  45. Gravatar of ssumner ssumner
    23. July 2009 at 05:47

    James, I totally disagree with your assumption in point one. There are all sorts or real world factors that argue for rules, even from a Max(u) perspective.
    1. Adminstration costs (of cost of checking out each case you might regulate in an unconstrained system on utilitarian grounds.)
    2. Predictability so that people can make decisions.
    3. The time inconsistency problem in policy-making.

    These immediately popped into my mind and I am not even an expert in this area. A better economist than me could come up with many more such reasons.

    I believe that if decisions are made democratically they have some legitimacy, although they may be bad rules, and often are. In addition I think democracy also maximizes expected utility. Elections are a not perfect, but at least roughly match social preferences. When an individual steals they are only looking out for their own self-interest, and thus they impose negative externalities on society. I admit that redistribution also has costs, but it is at least a bit more plausible that the benefits might outweigh the costs than in the case of theft. But as I say, I am even dubious of redistribution by the government on cost benefit grounds. But I am much more certain that theft should be banned.

    I see these as empirical questions, not something decided by saying “X is in the same category as Y, so we must impose the same rules on X and Y”. No we don’t, the cost/benefit calculus for government redistribution and theft may be very different.

    GNZ, I agree.

  46. Gravatar of James James
    23. July 2009 at 08:44


    In point one, I assume that a constrained maximum can never exceed an unconstrained maximum. I’m sure that’s not what you disagree with.

    This bit about the number of rules is not a red herring. If the right number of rules is 1, then it might well be the principle of non-aggression. If the number of rules is unbounded, then we can have one rule per scenario and we get the same implications as act utilitarianism. I realize that you might have calculated a marginal rate of substitution between expected social utility and admin costs. Or you could have choosen your preferred end state in advance and used that to back into the right number of rules to get your preferred result. I suspect that this is what most rule utilitarians actually do, but maybe you really did calculate the optimal number of rules.

    Perhaps you did not read my example above carefully, but I deliberately gave a pair of scenarios which are identical in terms of their effects and differ only in whether the person taking the money works for the government. The costs and benefits are identical! Do you honestly believe that the rightness or wrongness of taking money from unwilling persons turns on whether or not the doer is acting on behalf of a government?

  47. Gravatar of GNZ GNZ
    23. July 2009 at 17:44

    In real life I cannot tell the optimum ‘Act Utilitarian’ action from a whole set of other actions, but I can increase the expected utility of my actions, by using information at my disposal such as which rules have worked in the past (which amounts to Rule Utilitarianism). Better for me to maximize Expected U than to just ensure max U is on the table by keeping everything on the table.

    As to number of rules – we know the nature of the limitations to the number of rules in real life scenario (which is why Scott was able to list some) and they imply the optimum is much more than 1 and much less than infinity (given any reasonable definition of a util).

  48. Gravatar of ssumner ssumner
    24. July 2009 at 05:17

    James, I don’t see the two situations you describe as being at all equal.

    1. In once case the decision is assumed to be made by voters. Thus it reflects a more “intelligent view” as to what sort of redistribution we should have than an individual would be able to derive. If it is not democratic, then that makes the two cases somewhat more equal, but still not totally equal.

    2. The other difference is that the state can just withhold money from my paycheck, whereas the thief must breakdown my door to get the money.

    Just to be clear, I am not saying all theft is wrong, I think it is OK for a man dying of thirst to steal water from a well. Rather my point is that the real world considerations in theft and taxes are very different. The sorts of costs and benefits we weigh are very different. I still come down on the side of opposing both theft and progressive income taxes (indeed all income taxes.) But I have no principled objection with an economist who favors redistribution because they think the government doesn’t have to “break down many doors” to collect money from the rich. I am a supply-sider, and think too many doors would be broken down in any attempt to do major redistribution. (I’m excluding enough taxes for a Singapore-style safety net.) So I favor low overall taxes, and no taxes on capital.

    One other point. If the tax collector resigned and became a vigilante, collecting money from rich people who cheated on taxes, and if he did so without “breaking down doors”, then I’d see no difference, and I wouldn’t object.

    GNZ, You express my thinking much more elegantly than I do. I have no training in philosophy.

  49. Gravatar of James James
    24. July 2009 at 17:35


    If you just rely on rules that have worked well in the past, then there is no optimization going on. You aren’t maximizing anything at all, ergo you aren’t even applying the utilitarianism you claim.

    Scott & GNZ:

    Whatever the benefits of utilitarianism may be, you can’t claim those benefits just by mentioning the theoretical advantages of utilitarian reasoning whenever you make moral claims. At best, you can say something like “Utilitarianism seems great. I think that if I were actually to apply it, I’d find that XYZ maximizes utility.”

    At least as a morality-as-principles kind of guy, I can show that my policy views really do follow from the moral theory that I claim as a basis for my views.


    As a tax collector, I send the money to the treasury. As a thief I send the money to the treasury. The resulting expenditures are identical and determined democratically. I deliberatly constructed the example so that the ONLY difference is my employer at the time of the taking.

    Just to be clear, I’m not defending private Robinhoods. I’m (perhaps too coyly) accusing you of moral relativism/particularism. If you want to say that the costs and benefits differ from one doer to the next, so what is right for me may be wrong for you and wronger still for the next guy, that philosophy requires a lot of bullet biting, but that’s your choice. Your last paragraph indicates your willingness to bite at least one of those bullets.

    This will be my last comment on this post.

  50. Gravatar of ssumner ssumner
    25. July 2009 at 06:15

    James, I agree that utilitarianism is a very vague criterion. I see economists as having utilitarianism in the background, and then taking shortcuts. Thus a growth economist says “Let’s raise India’s GDP.” with the implicit assumption that higher GDP usually (not always) raises utility. Or they say “Let’s reduce unemployment in a recession” with the implicit understanding that much of the unemployment is probably involuntary. I think they also feel this way about redistribution. They are making a judgment that some redistribution will help the poor more than it hurt the rich, and then have no trouble reconciling that will keeping theft illegal, because legalizing theft would cause massive problems.

    I meant that the amount of redistribution is determined democratically in one case but not the other.

  51. Gravatar of Why I Love Scott Sumner Why I Love Scott Sumner
    3. August 2009 at 20:24

    […] love his long post on my paper. But much more than that, I love that he’s willing to write a post denying scientific realism […]

  52. Gravatar of The Knack … and How to Forget It: An Inquiry into Consumption Deskilling « Generation Bubble The Knack … and How to Forget It: An Inquiry into Consumption Deskilling « Generation Bubble
    6. August 2009 at 03:28

    […] on the vulgar masses. It’s an aesthetic masquerading as ethics, as Scott Sumner describes here. Skilled and deskilled consumers, as described by Jaffe and Gertler, are easily transformed […]

  53. Gravatar of Taylor Taylor
    10. November 2009 at 18:52


    Here’s a scenario for you:

    Imagine a society of thieves and producers. Imagine the thieves constitute a democratic majority in this society. Imagine that each thief has an amount that he would steal privately from the producers to satisfy his needs. Imagine that the thieves realize the producers will be upset if the thieves individually break down doors to take what they want. Imagine the thieves form the National Thieves Party and elect their representatives, who proceed to enact laws that “redistribute” (as you call it) the specific, individual amounts to each thief that he had planned to steal privately.

    This decision was made democratically and the theft was carried out via government redistribution. Every thief was thinking only of himself when voting for the spoils at the ballot box.

    Any thoughts?

  54. Gravatar of Scott Sumner Scott Sumner
    11. November 2009 at 09:12

    Taylor, Not a good situation. Fortunately the most democratic countries tend to do the opposite, they protect property rights. Switzerland is an example. It is the less democratic countries where you tend to see one group loot another.

Leave a Reply