Relatively Deprived at The New Yorker

Winterspeak speaks about a recent New Yorker article by John Cassidy on different methods of measuring US poverty. Both links are well worth following. I'll concentrate on the latter here.

You might expect a New Yorker article to be a textbook case in advocacy of "progressive" nostrums. And you wouldn't be wrong in this case. But before he runs into that particular ditch, Cassidy actually does a pretty good job of reporting how the current official measure of poverty was invented. He does a great job of demonstrating what's wrong with it: (a) it's an income measure, not a consumption measure; (b) it doesn't take into account regional variations in the cost of living; (c) it's a pre-tax measure, so doesn't take into account post-tax adjustments such as the Earned Income Credit, food stamps, Medicaid, etc.; (d) it doesn't take into account changing expenditure patterns since the measure was derived over forty years ago; (e) it doesn't distinguish between people poor due to a temporary downturn in their financial fortunes and those truly stuck in chronic hand-to-mouth existence.

Cassidy also quotes conservative/libertarian critics (Nicholas Eberstadt; Cox & Alm) of poverty calculations briefly, which is good. Unfortunately, he unfairly tars them with a bad-faith brush:

Conservatives would prefer a measure that reduces the number of poor people.
Winterspeak demonstrates that two can play that game:
… the Left wants "poor" to be defined in such a way that maximizes its size, and so will bolster the case for government transfers …
That's fine, guys. Probably even par for the course. And it may, in fact, be true that relatively few people are interested in an actual statistical measure of economic hardship, except as a crutch to support their particular political causes. Too bad.

Anyway, back to Cassidy's article. Since he's implied that conservatives are wont to jigger the poverty numbers to support their policy positions, he apparently feels that gives him the green light to do the same. He proposes that poverty not be defined by the current "absolute" measure that's (as everyone admits, poorly) based on a calculation of subsistence; instead, we should base it on "relative deprivation", e.g., "classify a household as impoverished if its pre-tax income was, say, less than half the median income".

Why? Well, because Cassidy's big bugaboo is dat old debbil, Inequality. And by flawless logic, his redefinition allows a repurposing of the forces of the "War on Poverty" into the "War on Inequality". This isn't simply killing two birds with one stone; it's arguing that the two birds are actually just one big bird anyway.

To support this thesis, Cassidy quotes a number of studies that purport to show inequality causes social ills all by itself. The "relatively deprived" are unhappy, no matter how good their situation might be in an "absolute" sense. Their health is poorer, and they die sooner. Even if they have color TVs and dishwashers, they "may lack skills—such as how to surf the Web for help-wanted ads—that could enhance their prospects in the job market."

Now this thesis is alarmingly easy to lampoon, and Winterspeak does so. Let's take it slightly more seriously. Probably the thorniest allegation is the alleged link between inequality and poor health; frankly, in comparison, the other issues seem more like petty whining.

Even this allegation tends to fall apart when looked at skeptically, however. Probably the best refutation is by the aforementioned Nick Eberstadt and the new-kidneyed Sally Satel in a short booklet entitled "Health and the Income Inequality Hypothesis: A Doctrine in Search of Data." (PDF available for free here). [Via commenter 'Tex' in the Asymmetrical Information comment thread.] Nick and Sally review the evidence, demolish the thesis, and conclude:

How did it happen that a notion with such questionable empirical documentation and such a limited relationship to the testable proposition has come to acquire so much respect within the academy and so much authority in policy circles?

As we have seen, the phenomenon surely cannot be explained simply in terms of the quantitative persuasiveness of studies of the inequality hypothesis. To the contrary, the ambitious intellectual claims of this school of thinking have been undergirded by research that has all too often relied upon limited or unrepresentative data sets, hazily expounded causality, elementary econometric fallacies, and results that cannot be replicated.

So Cassidy's strongest argument in support of his poverty-as-inequality thesis probably isn't true. Can anything be salvaged from his article?

Not much. It's almost certainly true that the past few decades worth of expensive anti-poverty measures haven't been all that effective in decreasing the "absolute" poverty rate. Is it possible that the definition change Cassidy advocates might (somehow) be more effective?

This isn't promising, for example:

Therefore, the way to reduce relative poverty is to reduce income inequality—perhaps by increasing the minimum wage and raising taxes on the rich.
Ah, "increasing the minimum wage and raising taxes on the rich." Is there no social problem that these solutions cannot cure? For minimum wage issues, see Jane and Prof Bainbridge. As for "raising taxes on the rich": nothing in Cassidy's article supports the notion that reducing inequality by chopping away at the "top end", decreasing the after-tax incomes of the rich, will alleviate any social ills of the poor. Knocking people off the top rungs of a ladder doesn't make the ladder any easier to climb by those at the bottom.

And then we have handwaving:

Raising public awareness about relative deprivation could help to change attitudes toward the poor, by showing how those at the bottom of the social hierarchy continue to face obstacles even as they, along with the rest of the society, become more prosperous. The Times recently reported that more than half of black men in inner cities fail to finish high school, and that, nationwide, almost three-quarters of black male high-school dropouts in their twenties are unemployed. "It doesn't do a poor person any good to say 'You are better off than you would have been thirty years ago,' " [American economist Victor] Fuchs said. "The pathologies we associate with poverty—crime, drug use, family disintegration—we haven't eliminated them at all."
… and, of course, nothing in the article supports the notion that changing the poverty definition will do anything to alleviate, let alone eliminate, the named social pathologies either. Nor is it clear why changing the poverty definition will raise "public awareness", or why "public awareness" can't be raised without changing the poverty definition.

Cassidy does make the valid point that the relatively-poor tend to be "socially and geographically isolated" from other Americans. Since I've read Mickey Kaus, I'm convinced that's a serious problem. Again, however, Cassidy makes no convincing argument that the problem is specifically due to relative deprivation, nor that the social engineering he proposes is likely to solve or assuage it. He should take this Kaus quotation to heart:

We're Americans--we don't mind people getting rich. We do mind richer people lording it over less rich people, or even thinking they're better than less rich people.
This would involve abandoning the obsession with inequality and economic resentment, which is a cherished position by some on the Left. On the other hand, it might actually help.