Thursday, June 24, 2010


Many economists build a case against policies that are aimed at reducing inequality in income and wealth. Their argument rests on two premises:

  1. Societies should seek to be maximally productive, because this is the best way to provide for the needs of the members of the society, and
  2. There is a trade-off between equality and efficiency - policies that promote equality tend to reduce productivity.
I'm not convinced. 

The first premise invokes Coase: well-defined property rights ensure that any redistribution (of wealth, property, or rights) that will increase societal welfare will happen through the mechanism of the market without the need for the intervention of the government - provided that transaction costs are negligible. 

Transaction costs are rarely negligible, but even if we set that aside there is still a problem. Arnold Kling gives an example that illustrates the problem with Coase. Prof. Kling considers the case of a biker who needs to use a bike path that crosses private land. Here's my retelling: The biker is willing to pay a heavy toll (a large percentage of his total wealth) in order to be allowed to use the path, because he wants to reach the hospital where his father is dying to see his father one last time. If he doesn't use the path then he has to take a much longer route and will not reach the hospital in time. The landowner wants to exclude the biker from using the path because the landowner doesn't like to have strangers on his land. Let's assume essentially zero transaction costs - the biker carries a transponder that automatically computes and pays his toll (with his agreement), according to the rate the landowner has set. The landowner sets the toll at a level that compensates him for the unpleasantness he experiences at having strangers cross his land. 

The problem is that the biker is very poor, and the landowner is very wealthy, and as a result, the price the landowner sets is much higher than what the biker can pay, even though the biker places a very high value on using the path. Under Coase, as long as the biker gets more value out of using the path than the landowner loses when the biker uses the path, then they should be able to agree on a price that compensates the landowner. Why doesn't that work in this case? Clearly, the biker places a very high intrinsic value on using the path - equal to a large percentage of his total wealth!

It doesn't work because the landowner and the biker value money, dollars, differently. Essentially the biker and the landowner are not using a common unit of exchange. You could say that the landowner sets the price in apples, but that the biker is paying in oranges. Or to highlight the difference in value, the landowner is setting the price in coal, but the biker must in diamonds. 

There is a further, even more radical implication to all of this, and that is that when there are differences in wealth among the members of a society, transfers from the wealthy to the poor INCREASE net societal welfare. This is because when a dollar is taken from a wealthy man and given to a poor man, the loss of intrinsic value experienced by the rich man is less than the gain in intrinsic value experienced by the poor man. Prices don't clear the market because prices are not established in units of intrinsic value.

As far as premise number two, I haven't seen any good measures of the magnitude of that trade-off. Is it significant? Is it significant at some degrees of intervention, but not significant at others? If you know where i can see data that describe this relationship I would be very interested.

PostScript: None of this addresses the libertarian arguments against policies that are aimed at reducing inequality. Nor does it address the question of whether governments are needed to effect redistribution (when it is desirable), or whether non-coercive institutions and norms could be a more optimal solution than government.


  1. Great post, Robert. I've had related exchanges with Justin about money (or for that matter, material wealth/comforts) being used as the standard for what is valued by people. I'm glad to see the discussion being extended and expanded. Thank you.

  2. This comment has been removed by a blog administrator.