Tuesday, November 29, 2016

High-end inequality colloquium at NYU, week 6: Morse & Bertrand, Trickle-Down Consumption

This past Monday, Adair Morse of Berkeley presented her paper (coauthored with Marianne Bertrand of U Chicago), Trickle-Down Consumption.

The paper is a nice, compressed job of empirical research, published recently but of particular interest to us given its relationship to my co-convenor Robert Frank's interest in expenditure cascades, whereby rises in market consumption at the top triggers attempts to keep up via increased consumption just below, then again just below that, and then continuing until it has radiated far down the distributional chain.

The two main explanations for the Frank story are (1) positional externalities, whereby my having a bigger house requires you to get a bigger one, too, just to restore our relative positions to what they were before, and (2) context, whereby my bigger house simply triggers you (without necessarily having competitive motives) to need a larger house in order to feel that yours is big enough.  These two views are closely related and can be hard to tell apart, although (2) is framed in such a way as to sidestep criticisms to the effect that one should not give social weight to "envy" (a criticism that I consider wide of the mark in any event).

In the Bertrand-Frank study, state-level data suggests that, when the consumption of the top 20% in the income distribution in a state increases, those in the bottom 80% start consuming more even if their current and expected future income are flat.  Hence, their savings rates fall and they experience increased rates of bankruptcy and financial distress.

With respect to types of consumption, the effect is not greater for what seem to be positional or status goods than for other types of consumer outlays.  This might tend to rebut viewing the issue as status competition via positional goods, although how it affects the context view is less clear.

Based on meticulously testing various alternative explanations, the authors suggest that the methodology might involve a rise in well-off consumers in a neighborhood triggering an increase in appealing high-end goods and services that the others then start consuming without due regard for their budget constraints.  (But they manage to rule out mere price level changes for the same goods.)

To me, this seems to invite (as a plausible explanation) what I call the chocolate chip cookie or temptation problem.  When there are more nice things around, I tend to buy them, just as I might gobble chocolate chip cookies left on the seminar table. [I actually don't do that these days, but never mind that.]  So I'm inclined to view it as an internalities problem, in which consumers who are tempted by what might actually be nice things (from which they do indeed derive utility) get themselves into worse spots overall.

It's not as obvious why a preference for positional goods or the influence of context on consumer choice would necessarily involve irrationality.  But Bob Frank, in terms that he once nicely explained in an NYT column, invoked the work of James Duesenberry, a Harvard economist whose model for how people make consumption choices dominated the economics field & textbooks, because it nicely explained actual observed behavior, until Milton Friedman's permanent income hypothesis wholly supplanted it.  To paraphrase the old joke, Duesenberry's account worked in practice but not in theory, whereas Friedman's worked in theory but not in practice, so the economics profession unanimously voted for Friedman.

Friedman views people as rationally and farsightedly allocating consumption opportunities across their lifespans in order to equalize its marginal utility in all periods, as judged when one decides, and hence one's total lifetime utility.  (If bequests are added to the picture, a common move is to model the multigenerational household as if it were a single infinite-lived individual.)  But he can't readily or convincingly explain, e.g., why rich people generally save higher percentages of their incomes than poor people.  A smoothing rationale alone, for example, might tend to apply equally to each.

Friedman also had trouble explaining the fact that, as societies grow richer, their savings rates generally don't increase.  But, to quote Frank's NYT column:

"Mr. Duesenberry's explanation of the discrepancy is that poverty is relative. The poor save at lower rates, he argued, because the higher spending of others kindles aspirations they find difficult to meet. This difficulty persists no matter how much national income grows, and hence the failure of national savings rates to rise over time.

"To explain the short-run rigidity of consumption, Mr. Duesenberry argued that families look not only to the living standards of others, but also to their own past experience. The high standard enjoyed by a formerly prosperous family thus constitutes a frame of reference that makes cutbacks difficult, which helps explain why consumption levels change little during recessions.

"Despite Mr. Duesenberry's apparent success, many economists felt uncomfortable with his relative-income hypothesis, which to them seemed more like sociology or psychology than economics. The profession was therefore immediately receptive to alternative theories that sidestepped those disciplines."

Whence the shift by universal acclaim to Friedman's model, even though it was psychologically less realistic and also a worse fit with the key data points noted above.

This phenomenon has much in common with what I say about public economics and optimal income tax theory in my recent U Miami Law Review article on the "mapmaker's dilemma."  I too call there for more sociology, and less exclusive reliance on rational choice-based economic models.

In any event, once one adds the Duesenberry view to the positional and context models, they can join the temptation / chocolate chip cookies model in positing that a rise in high-end consumption may trigger planning failures that we might describe as involving internalities from people in lower income tiers.  This might be viewed either as further ground for deeming high-end inequality injurious to those below, or else simply as providing support for the use of policy instruments that focus on increasing private saving where it appears to be suboptimal, and/or on addressing the misuse of consumer credit.

No comments: