Tuesday, October 25, 2016

Colloquium on high-end inequality

Yesterday was the first day of the seven-week Colloquium on High-End Inequality that I am co-leading with Robert Frank.  Afternoon sessions, from 4;10 to 6 pm in Vanderbilt 202, are open to the public.

Our first session was devoted to several of Bob's short pieces, including one on the importance of luck, another on expenditure cascades, and a third on winner-take-all markets.  The following is a fleshed-out version of my main comments:

2 discussion topics today: the relevance of luck, what if anything is wrong with high-end inequality.

1. Luck
Obviously, luck matters in life. Everyone knows that.  So why, in relation to thinking about high-end inequality, is it particularly worth mentioning?

Because of the alternative view – which, here, is not that everything was predetermined or inevitable, but that all success is just flat-out deserved, period.

3 examples from the papers: Bob survived a dangerous heart attack, Mike Edwards of ELO was buried by a hay bale, Michael Lewis happened to sit next to a Salomon Brothers spouse, who gave him the idea to write Liars Poker, and launch his amazingly successful career.

The first two counterfactuals, but for the luck, are indisputable.  Bob wouldn’t have lived, Mike Edwards wouldn’t have died.  And there’s no particular policy payoff to them, apart perhaps from promoting emergency treatment and road safety.

But for Michael Lewis, some might say: C’mon, he was Michael Lewis. He had extraordinary talents. Surely he was bound to succeed anyway.

Now, that may or may not be true.  We don’t get to run a thousand simulations in which we see how many times he becomes “Michael Lewis.”  But suppose that he wouldn’t have comparably succeeded in many or most of those simulations, even if he never got a heart attack or was buried by a hay bale. What would we learn?

Maybe it’s just attitudinal. Even if you’re highly successful, you should retain proper humility and compassion.

My own answer is that the role of luck, which Bob highlights in his papers, doesn’t matter all that much substantively to how we should think about high-end inequality.  But it might matter not just attitudinally in people’s private thoughts, but also rhetorically in public policy debate.

I think it matters rhetorically due to what I’d call the ideology around meritocracy.  We live in a society where everyone who’s successful claims to have earned it. It’s not just Donald Trump pretending that he didn’t get $14 million from his dad. It’s all kinds of people who were born on third base and believe they hit a triple.

Meritocracy becomes toxic when it divides the world into deserving “winners” and stupid, pathetic “losers,” and that’s where I think our culture often is these days.

Correcting attitudes become all the more important when we have winner-take-all markets.  One person makes a billion & the second makes nothing. Even if the winner was a hair “better,” not just a hair luckier or five seconds earlier, the discontinuity between relative merit and relative reward is well worth keeping in mind.

But suppose all market outcomes were fixed given people’s genetic endowments.  It would still be brute luck what genetic endowment you had.  And it would still be brute luck how your particular talents happened to fit your particular environment.  For example, how much would Michael Lewis’ particular endowments help him in modern Somalia, or 12th century England?

If you think of Shaquille O’Neal as having been lucky, not just to be over 7 feet tall but to live in a society with millions of basketball fans, you’ll have a point in mind that applies far more generally.

So for me, realizing that successful people were often lucky, in the sense of Michael Lewis and the dinner conversation with the Salomon spouse, doesn’t do that much work, for three reasons.

First, I already knew it was true.

Second, I define luck broadly enough that it had to be true.

Third, I don’t subscribe to a theory of distributive desert that’s based on inherently deserving what you’ve earned.

Instead, for me the rationale for property rights and keeping tax rates at reasonable levels is just about incentives & their effects on behavior.  But admittedly we (including I) have intuitions about desert that are not entirely reducible to this.

2. What (if anything) is wrong with high-end inequality?
In the public economics literature, often the only reason for mitigating high-end inequality is the declining marginal utility.  You’d never reduce a rich person’s wealth by a dollar unless at least a penny, or some fraction of it, was successfully transferred to someone else.  This follows from the standard assumption that people only derive utility from own consumption.

Bob challenges this approach by emphasizing the importance of status and relative position, which are affected by relative consumption.  Plus, he posits socially costly expenditure cascades from the top on down.

I myself tend to agree with these points, but I want to note a few possible objections.

First, it’s often argued that all this is just “envy” and should be discounted.  This combines a descriptive claim about people who care about relative position, with a normative claim about envy’s unworthiness to be counted.

Second, recall the old phrase “keeping up with the Joneses.” Suppose that relative position matters more laterally, between peers, than it does vertically, or from the super-rich on down.  Then we might want to tax positional goods relative to non-positional goods, but with no particular reference to rich versus poor.

One definition of positional goods might be market consumption generally, as distinct from leisure. So we might want a high rate of consumption tax, as a kind of pollution tax on negative externalities, but it wouldn’t necessarily be progressive.

Third, what if high-end inequality has positive externalities as well as negative ones? An example might be, the market for luxury goods, including high-end healthcare, leads to technological advances that then become cheap to provide for everyone.  So why limit the analysis to negative externalities from high-end inequality?

No comments: