The Persistence of Poverty
Why the Economics of the Well-off Can't Help the Poor
by Charles Karelis
Yale, 208 pp., $30
Charles Karelis is not a philosopher-king, but he can fairly be described as a philosopher-public servant.
Formerly the longtime director of the Fund for the Improvement of Postsecondary Education (in the Department of Education), he has also been president of Colgate. He is currently research professor of philosophy at George Washington, and in this short but ambitious book, Karelis attempts the difficult task of saying something new about poverty--a subject that has been exhaustively studied for decades. Karelis's effort to stake out an original position is laudable but doesn't finally succeed.
He begins by making what is usually a conservative argument: Poverty results from the self-destructive behavior of the poor. He mentions five behaviors in particular. The poor are poor because they "are not working much for pay," are "not getting much education," are "not saving for a rainy day," are "abusing alcohol," and are "taking risks with the law."
Karelis's list of the problematic behaviors of the poor is useful, although in two respects it is idiosyncratic. His contention that drinking to excess is an important cause of poverty is refreshingly retro; that claim was commonplace in the 19th century but is practically never voiced today. More surprising is a cause that Karelis omits: The propensity of the poor to produce illegitimate children. Karelis acknowledges that "having children early and out of wedlock is ... doubtless a big factor in poverty in the United States today," but does not consider it "as global or as perennial as the other factors on our list."
What, then, causes the self-destructive behaviors of the poor? Karelis offers six explanations, which fall into two categories. Three posit the irrationality and "dysfunction" of the poor, who are thought to be characterized by "apathy ... fragmentation of the self, which leads to short time horizons, and weakness of the will." The other three understand the behavior of the poor as a rational response to their situation and deny the dysfunctionality of the poor. Instead, their behavior is explained as a response to "opportunities [that] are unduly limited" or "perverse incentives created by public policy," or as a result of "atypical preferences."
But Karelis goes on to reject all of these arguments. The poor are not, for the most part, dysfunctional. Nor is it the case that they lack opportunity, that their preferences differ significantly from those of the nonpoor, or that public policy encourages their counterproductive behavior. Karelis argues instead that, given their economic situation, "poverty-linked conduct is efficient," and he elaborates on this claim: "Poor people engage disproportionately in the poverty-prolonging and poverty-worsening behaviors because they are poor--and rational. For this conduct is exactly the conduct that makes sense for them."
To make this case, Karelis takes aim at a fundamental postulate of economics: the law of diminishing marginal utility. That law posits that the first dollar of income has the most utility, the second dollar slightly less, the third still less, and so on. Its core idea is that "resources mean most to those who have least. On this basis it is natural to conclude that poor people stand to benefit especially from working for pay."
Karelis contends, though, that the marginalist view is inapplicable to the situation of the poor--and that it underlies all six of the theories of poverty that he finds wanting: "Conventional theories of poverty are divided into those that assert that the conduct in question really contravenes marginalist prescriptions [i.e., the dysfunction theories] and those that contend the contravening is a mirage--but neither side questions the validity of marginalism itself."
Karelis claims that "marginalism is mistaken," but in his view it is mistaken only in part. For those who are well off, he agrees, the value of each additional dollar does, indeed, diminish. That rule does not, however, apply to the poor: "Marginalist economics was an economics of more-than-enough that mistook itself for a general theory, applicable to both surplus and deficit."
Of course, the crucial question is why marginalism does not apply to the poor. Why isn't it worth the while of poor people to earn that first (or next) dollar? To answer that question, Karelis, adopting a familiar philosophical practice, argues by analogy. A poor person is like someone who has suffered multiple bee stings. For someone stung once, salve to relieve the sting would be very valuable. But for someone stung seven times, salve that would relieve only one sting would not be particularly valuable because the pain from the other six stings would remain. He explains:
__The person with seven bee stings ... would not sacrifice much to relieve the sting on his hand, seeing that the pain of it was nearly drowned out by the pain of the six stings on his body. This would seem to be the position of very poor people, for whom work, schoolwork, and (in a much different way) moderation in alcohol use constitute sacrifices that would buy them too little felt relief to be worth making, so many are their troubles.__
Let's stipulate that Karelis has accurately stated the problem of someone who's been stung multiple times. But his book is entitled The Persistence of Poverty, not The Persistence of Bee-Sting Pain. He therefore needs to show, and not just to assert, that the poor--to be fair to Karelis, the "very poor"--can aptly be compared to people who have been stung seven times. To simply assert the equivalence a priori is to be guilty of excessive abstraction.
It's reasonable to suppose that the poor can be placed on a continuum: Some of them would correspond to people with only one sting, others to people with two, etc. To the extent that significant numbers of poor people resemble people with only one or two stings (or, to look at it differently, people who already command the resources to pay for five or six dabs of salve), Karelis's analogy--and, more broadly, his critique of the relevance of marginalist economics for the poor--becomes less tenable.
To determine the resources of the poor and their capacity for self-advancement would, of course, require empirical study, not philosophical speculation. At least with respect to the American poor, data certainly indicate that many have fewer stings or more salve than a reader of Karelis might suppose. For example, Robert Rector of the Heritage Foundation has recently shown, using federal government statistics, that 43 percent of poor American households own their homes. (A typical poor person's home is a three-bedroom house, with one-and-a-half baths, a garage, and a porch or patio.) The typical poor American has more living space than the average (not the average poor) inhabitant of cities like Paris and London.
It seems unlikely that poor Americans like these are the metaphorical equivalent of people with seven bee stings, for whom the effort to work to buy one dab of salve is not worthwhile.
In addition to the problem of determining the number of stings from which the poor actually suffer, there's also the problem of how many they suppose they have, and how many they should suppose they have. Here Karelis contradicts himself. In one place he asserts that a poor person's assessment of his condition can't be disputed. In effect, the poor person (like Cindy Sheehan, according to Maureen Dowd) has absolute moral authority.
__To put it in terms of stings and salve, a poor Asian immigrant may regard an annual income of twenty thousand dollars as verging on sufficient--as leaving just a few stings unsalved--while a poor African American may regard that same income as very insufficient--as leaving so many stings uncured that it is not worth much effort to get another dab or two of salve. ... Behavioral differences [between Asian immigrants and African Americans] are to be explained as equally rational, benefit-maximizing responses to the same economic facts, seen and felt differently.__
Elsewhere, however, Karelis takes a different view, arguing that poor people harm themselves when they exaggerate their plight and minimize what they might do to relieve it. Thus he contends that civil rights leaders (think, for example, of those who disparaged the desirability of "dead end" jobs) harmed their constituents when their rhetoric reduced "the marginal relief to be expected from a small improvement in objective circumstances. It might even be argued that this rhetoric worsened the poverty problem it was meant to help relieve."
In effect, the hypothetical Asian immigrant's reaction ("that $20,000 job is a step in the right direction") appears objectively preferable to--and more prudent than--the hypothetical African American's reaction ("that $20,000 job offers nothing but chump change"). Despite Karelis's earlier claim that the actions of the poor are rational, here he suggests that the failure to take a job is irrational. In short, he seems to question the basic premise of his argument.
When it comes to offering solutions to poverty, Karelis mostly advocates the expansion of the course correctly adopted by current American social policy: that is, the policy known as "making work pay," which seeks to make "work a more attractive option for low-income people through transfers and other provisions ... whose benefits depend on the recipient's working."
But Karelis also supports no-strings assistance to the poor: that is, assistance that is not conditional on any efforts by the poor to help themselves. After all, if the poor are given funds to buy four dabs of salve, they're more likely to work to be able to purchase three additional dabs themselves. To make this case, though, Karelis must refute the familiar contention that no-strings assistance discourages work: Why work, if you'll be paid even if you don't work?
Karelis does not deal adequately with that obvious objection. He limits himself to a paragraph in an endnote, in which he alludes to "the famous income maintenance or negative income tax experiments conducted by the federal government between 1968 and 1982" which "have often been taken to show that [income] transfers reduce work effort." That conventional view may be wrong, Karelis asserts, because the beneficiaries of the negative income tax may have underreported their earnings.
Is there evidence to support this hypothesis? If so, he doesn't cite it.
In the final analysis, Karelis advances a novel argument on behalf of a very familiar position: Giving money to the poor will solve their problems. But in the past few decades most Americans (and more, though not yet most, American poverty experts) have increasingly come to believe that simply giving money to the poor encourages the behaviors--in particular, not working and not marrying--that make and keep them poor.
It is unlikely that Karelis's ingenious philosophical argument will cause many people to change their minds.