This blog got started in response to the insistence by many on the right that evidence from the real world is no longer relevant for public policy decisions. But the left and right have been talking past one another since way before the appearance of the reality deniers. When the left says that redistributive taxation is required in the name of social justice, for example, the right counters that forced income transfers are morally illegitimate. There the conversation typically ends.
There seems little hope that the right will start trying to see things from the left’s point of view. But progress may be more likely if the left is willing to confront the right’s arguments in their own terms. As a case in point, I call your attention to a potentially fruitful dialog provoked by this strategy.
According to many libertarians, an ideal world would be one in which well informed people could exchange freely with one another with no significant market frictions. They concede that when transaction costs stand in the way of this ideal, collective action can sometimes improve matters. But the guiding principle behind any such action, they argue, should be to mimic the kinds of arrangements people would have negotiated on their own, had transaction costs not prevented them from doing so.
With this framework as a starting point, I argued in my most recent New York Times column that redistributive taxation is justified because it closely mimics the pay schemes we observe in every private firm. In those schemes, the most productive workers in any group are paid substantially less than the value of what they contribute to the employer’s bottom line while the least productive are paid substantially more.
Close scrutiny suggests that this pattern reflects the workings of an implicit market for high-ranked positions in every work group. If, as in the United States, people cannot be forced to work for an employer against their wishes, no one can occupy a high-ranked position in any work group unless others agree to occupy low-ranked positions in that same group. The logical implication is that if people prefer high-rank to low-rank, no stable group of mixed productivity can survive in which all workers are paid the value of what they produce. Those who find low-ranked positions aversive could always quit such a group in favor of a new one composed of others like themselves.
But markets make a better outcome possible for everyone. Those who value high rank can transfer some of their pay to less productive colleagues to compensate them for the burdens associated with low rank. Along the resulting pay schedules, the distribution of wages is substantially compressed relative to the corresponding distribution of productivity, just as we observe in practice.
Under these pay schedules, the best option available to someone who doesn’t care about local rank would be to join a group in which he was the least productive worker. For in that position, he would receive a substantial pay premium relative to the value of what he produces.
Rank matters not just within work groups but also in society at large. And irrespective of whether someone cares about rank per se, social rank matters for instrumental reasons. People at the bottom of society’s income distribution, for example, find it much harder to send their children to good schools, since a good school is an inherently relative concept. To send your children to a good school, you must buy a house in a neighborhood surrounding such a school, and those houses are expensive for precisely that reason.
If new societies could form and dissolve as easily as private work groups can, people might be able to sort this kind of distributional externality out on their own. But transaction costs make it impractical to form new societies. Because we’re more or less stuck in the societies we’re born into, social institutions cannot suit every preference exactly. We must choose those that work best on average.
What the labor market tells us is that when people are free to form groups of their own choosing, they always agree to an implicit tax that transfers income from the most productive members to the least productive ones. So if libertarians believe the best public institutions are those that mimic the arrangements people would have chosen for themselves if transaction costs had not been an obstacle, they cannot consistently insist that redistributive taxation is morally illegitimate.
That, at any rate, is my challenge to libertarians. David Friedman, son of the late Milton Friedman and a leading libertarian thinker, has taken up my challenge. His attempt to rebut my argument is here, and he has graciously posted my response to his rebuttal here.
He’ll have more to say, I am sure. We may not bridge the communication gap between us, but I hope it will be clear to readers of our exchange that we’re each seriously trying to engage the other’s position in its own terms.