With his new book, Poverty and Discrimination, Kevin Lang becomes my go-to guy on poverty. I disagree with him on some important and emotional points, and I'll have more to say about that at a later date. But I want to use this post to say why this is such a terrific book, even for non-economists or for economists whose interests lie elsewhere.
On p. 9
Inevitably, on occasion I am forced to say "most researchers believe that"...In general, however, I try to resist the temptation to be the ultimate judge of a body of literature rather than giving you the tools and information you need to evaluate it.
As you might guess from what I've been saying recently, I prefer to be able to evaluate the evidence myself, so I really appreciate this promise, which Lang keeps. What's more, the quality of the presentation of data is simply outstanding. He does not waste my time with tables of coefficients from multivariate regressions, which I would not believe. Instead, he plots data or puts it into tabular format in ways that display clear, reliable relationships.
On p. 35,
Thus we can say, "An individual is poor if he or she lacks sufficient financial resources to obtain adequate food, clothing, shelter, and medical care and to participate in society.
The problem of defining poverty is difficult. Lang uses the phrase "participate in society" because he thinks that better fits our intuition about poverty. A person could have adequate food, clothing, and shelter but still not feel accepted on economic terms in most American neighborhoods. Taking this elastic approach to defining poverty helps to rationalize the official poverty line, which is about $20,000 for a family of four, and perhaps justify a higher figure, even though $20,000 would be considered affluent in much of the world.
On p. 39-40,
Most food stamp recipients spend more on food than they receive in food stamps...Because most food stamp recipients want all their food stamps...food stamps are probably worth close to their nominal value...
At the other extreme, the evidence suggests that people would not purchase insurance with the value of their Medicaid health insurance if it were not provided in its in-kind form.
Lang is discussing the valuation of in-kind transfers in measures of poverty. But this is potentially a fairly damning quote with regard to health insurance mandates and universal coverage, if you think about it.
On p. 89,
If per capita GDP increased but the poverty rate did not fall [which the data for the 1975-2004 period show], there are only two mechanisms...The first is that households become smaller...the second explanation is that income inequality increased.
I think there is a third possible mechanism--immigration. Lang talks about immigration as if it changes the mix of workers, adding to the unskilled and lowering unskilled workers' wages. As he points out, this effect is unlikely to have a large impact on poverty rates. What he omits, as far as I can tell, is the arithmetic effect on the poverty rate of pouring poor immigrant families into the system. I'll have more to say on that some other time--I think it's worth an entire essay.
On p. 100,
a 1 percentage point increase in the poverty rate due to an increase in the number of female-headed families cannot explain why the relation between GDP per capita and poverty was not maintained after 1973. Indeed, the trend toward more female-headed households was present over the 1959-1973 period but did not obscure the decline in poverty associated with economic growth.
On p. 382,
We have seen that the minimum wage is a blunt tool for addressing poverty. Nevertheless, the minimum wage is very low by historical standards...A reasonable goal is to raise the minimum wage to a level comparable to that of the 1970s
It's only a slight exaggeration to accuse Lang of saying that raising the minimum wage won't really do anything to alleviate poverty, but we should do it to show that we care. (Sort of like Robin Hanson's theory of wasteful health care spending.)
In fact, I think that one could fairly conclude from Lang's work that many anti-poverty policies are "blunt tools," with relatively weak or even uncertain impacts on poverty. Personally, I would be comfortable with that conclusion.
Notwithstanding my quarrels with Lang on some key issues, Poverty and Discrimination is social science at its best. The issues are interesting, the analysis is first-rate, the organization is excellent, and as I noted earlier the presentation of data is exemplary.
The book is intended to be used as a course textbook, and it would serve that purpose very well. But I think that many economists who do not teach courses on poverty would still value Poverty and Discrimination as a reference. And I think that non-economists who want to get introduced to some real economics (as opposed to Freakonomics) would find it rewarding.
Finally, I recall my oldest daughter's comment that in sociology courses you only learn that "There's poverty and America sucks." If American poverty is indeed the obsession of sociologists, then Poverty and Discrimination ought to be read by every member of that profession.