In a recent essay on Greg Mankiw’s forthcoming paper on income inequality, Noah Smith writes,
The usual economist case for income redistribution is based on utilitarianism; the idea is that $1000 matters more to a poor person than to a rich person.
This doesn’t sit well with me, and not for ideological reasons (although this may have some influence). But, I’ve had some trouble putting my thumb on the reason why. The economy is complex. People earn their money through production processes that are complicated, and some of them earn their money by funding a fund that funds production processes. Further, it’s difficult to make sense of what is a return to entrepreneurship. Entrepreneurial activity is the result of uncertainty and disequilibrium, and profits are solely the outcome of the latter. Entrepreneurs preform the positive — and very important — economic function of assuming risk and helping to coordinate markets, but their main source of income seems to come from “denying” capital and labor their equilibrium returns. But, I think we can get around some of this complexity by just thinking through the issue in the context of gains from trade.
A typical response is that economists cannot make interpersonal utility comparisons, which means we can’t compare the utility of some good between persons ‘A’ and ‘B.’ The simple reason this is true is because we can’t measure utility, and utility is something that’s subjective — the utility of $1 will be different between ‘A’ and ‘B.’ Nevertheless, this approach never seemed very persuasive. Take two people: ‘A’ is very poor ($100 income) and ‘B’ is very well off ($1 million income). We can assume very different utility schedules, but it seems counter-intuitive to claim that ‘B’ will value her last $1,000 more than ‘A’; to put it slightly differently, $1,000 represents 0.1% of ‘B’s’ income, and 1000% of ‘A’s.’ While most economists accept the fact that interpersonal utility comparisons can’t be made, it just doesn’t come off as a strong reason to not make certain utilitarian assumptions regarding welfare.
But, there’s another approach that undermines the case for redistribution, even if it is the case that ‘A’ values $1,000 more than ‘B.’ If this were true, then there’s an opportunity for a gain from trade. That is, taking advantage of asymmetric preferences, it pays ‘A’ to bargain for those $1,000 by offering ‘B’ something that the latter values more (and ‘A’ values less). Maybe the wealthy person will buy some consumer good from another producer, exchanging the money for that consumer good. But, probably more likely, that wealthy person will exchange his money for others’ labor (and other inputs, like capital — but, the more capital abundant an economy, the higher the relative returns to labor [↑K/L, ↓r/w]; that is, higher real wages). In other words, the wealthy are likely to invest “excess cash” (cash holdings beyond their demand for money), precisely because others prefer that cash to the commodities (including labor) they have available to exchange.
Sometimes the complexity of an economy can make this point difficult to see. What if a wealthy person instead invests excess cash in some kind of financial asset? Well, remember that this is essentially an act of lending. The investor is lending some sum of money for the promise of an even greater sum of money at some point in the future (dictated by that person’s time preference; liquidity preference determines the kind of savings instrument), but the borrower is the one using that credit to buy labor and capital (or consumers’ goods, if it’s consumer credit). Even if we add several stages of financial intermediation, ultimately there has to be some production to afford the interest. (There are issues of fraud and non-Pareto optimal transfers of income related to finance, but these are issues specific to finance and shouldn’t factor into the discussion of redistribution — solve those issues, don’t place the wrong sized band-aid on them.) These kind of confusions seem to be behind the belief that the wealthy have some schedule for the provision of savings towards physical investment, and at some point they prefer some alternative (they “sit” on their money, which actually usually means they lend it out).
Returning to the gains from trade point, consider the consequences of redistribution versus those of trade. In the former case, we have a zero-sum game. Following our ‘A, B’ distinction above, ‘A’ gains utility from now owning $1,000 more, but ‘B’ suffers a loss of utility (whatever $1,000 was worth to her). Whether this transfer is Kaldor-Hicks efficient (if ‘A’ can, theoretically, recompense ‘B’ for her loss; that is, if there is a net gain in utility) is an empirical question, but this position would nevertheless be inferior to that which would come out of trade. In the case of trade, the game is a positive-sum. ‘A’ gains utility from owning $1,000, but ‘B’ also gains utility from whatever was exchanged for her money — it’s a Pareto optimal outcome. In the case that the exchange involves a productive asset, redistribution can also lead to the foregoing of the opportunity to add to society’s stock of wealth (especially if ‘A’ uses her new $1,000 to consume). Note, if the trade fails to occur, we either have a market failure — the most plausible, but still unlikely, being a failure of intermediation — or ‘B’ really does value $1,000 more than ‘A,’ despite the disparity in income.
Noah’s point about inherent inequalities between humans (e.g. disparities in the distribution of intelligence) seen in the “gains from trade,” or Ricardian, light doesn’t seem quite right either. The economy is built on the asymmetric distribution of skills, preferences, and endowments. We trade precisely because because these inequalities create the opportunities for gains from trade. The perfect world, of course, is that where there is no longer any need from trade, or where all resources have been allocated to those who value them the most. But, this world doesn’t exist. This leads us to another point on the raison d’etre of markets: markets do a better job of allocation than governments do. It’s amazing that people lose sight of this point when talking about inequality. We prefer trade to government redistribution precisely because the former is likely to do a better job than the latter.
This being said, I do think that there’s inequality that we can’t trace back to the causes of comparative advantage. For example, I believe that the recent recession illuminated one method of wealth extraction from borrowers, where “illusory wealth” induced borrowers to take on debts and interest, only to find their real wealth take a dramatic hit during the recession (e.g. the fall in home values). Where there is fraud there is an added market failure (information asymmetry). Other markets failures may provide a necessary, but insufficient, condition for redistribution. One example is monetary disequilibrium; where financial intermediation fails, maybe the government ought to redistribute. Ultimately, whether collective action should be taken at all is a function of the relationship between the size of the externality and the net costs of intervention. But, I do think there are some legitimate things government can do to reduce inequality, although all, or most, of them have a libertarian twist to them (e.g. free banking).
I agree with Noah that economic theory is not likely to persuade the masses. But, that’s why we’re economists. We have a certain set of information that non-economists don’t. We can analyze the world through a certain lens and we can present what we see in our writing. We would all like a perfect world, but economists know as well as anybody else why certain means of achieving the perfect world will instead lead to greater imperfection.