Book Summary: The Persistence of Poverty

To treat the issue of the poor we have a variety of perspectives from which to begin. The central starting point for our purposes is, of course, the church. Along the way, though, in order to be the church we also have to know the positions that are not of the church. That means digesting a wide ranging diet of thought on the poor. My first contribution to this diversity is a summary of a recent book (2007) by Charles Karelis entitled The Persistence of Poverty.



Karelis's book seeks to address the rationality that underlies the persistence of poverty, or why poor people seem to do things repeatedly that will logically prolong their poverty (e.g., neglecting to save a portion of their small income). When the homeless spend a day's panhandling returns on a glut of alcohol rather than, say, a bus ticket to look for a job, or perhaps saving it toward a rent deposit on an apartment, their actions seem to thwart their own best interests. It just seems irrational. This, in essence, is one of the central puzzles of poverty. Are poor people generally less rational people as their behavior seems to suggest? If so, do irrational people wind up poor, or do poor people wind up irrational?

The standard theoretical accounts of this behavior fall into two categories, according to Karelis. There are those theories which explicitly label the behavior irrational, arising from psychological dysfunction, and there are those theories that claim that the behavior is rational by citing external opportunity constraints (there just aren't any available jobs), atypical preferences (the poor aren't dysfunctional, just atypical), or perverse consequences of public policy. In contrast, Karelis argues that all of these theories rely on a common mistake accepted from classical economic theory: the law of diminishing marginal utility.

The law of diminishing marginal utility states that as consumption of a good increases by constant amounts, the consumer's positive experience that the consumer receives from the consumption increases at a slower rate than the consumption. That is to say that if you sit down to eat a delicious looking chocolate cake, the first bite of cake gives you more satisfaction than the last bite, say the one that finishes off the entire cake at that one sitting. Or, for perhaps a more intuitive example, the luck of finding $10 will be significant to a homeless man (who has no dollars) and perhaps not worth the trouble of bending over to pick up the bill off the street for a millionaire.

By this same logic, the increase in utility afforded by earning the first few dollars by employment should be so great for the impoverished that very little could deter their seeking a job. You can see how the theory suggests that poor people are irrational. That minimum wage job could so drastically alter their basic quality of life, and yet they can't be bothered to look for a job, or even perhaps show up on time once they have one. Strange.

Karelis answer to this conundrum is three-part. First, he explicitly defines poverty in a subjective manner, "The essence of poverty is lacking the material resources to meet basic needs." The "basic needs" may be relativized based on culture, and even perhaps class. That is, someone attempting to live among upper-class circles may feel that they lack basic needs in order to be a part of the group if they do not have access to certain goods that would be seen as luxury items by those of a lower class. In this way Karelis definition of poverty is strongly psychological, and it is debatable whether it is specific enough to be useful in the end.

Having defined poverty based on the perceived lack of resources to meet basic needs, the second part of his argument is that the theory of diminishing marginal utility represents an equivocation on "utility". Karelis makes the point that nearly all economic theorists have been members of the well-to-do classes. The result is that they failed to distinguish between goods that are pleasers and goods that are relievers. For Karelis it matters whether you are seeking relief from a lack of resources or whether you are consuming to increase pleasure. His central image is that of having multiple painful bee stings on one's arm and desiring the use of a powerful salve that relieves the pain of individual stings. The salve comes in units of "a dab" which is enough to relieve exactly one sting. Karelis points out that the psychological benefit of having one dab applied when one has, say, 10 stings on the arm is tiny since one will still be in a great deal of pain. Contrast this with the psychological effect (for one not in any pain) of eating one bite of a 10-bite piece of cake. The utility of the first unit of salve for the man with 10 bee stings is clearly less than the utility gained by the first bite of cake.

Karelis uses this image to argue that for those people perceiving a lack of resources to meet basic needs (enough salve to stop the pain), the marginal utility actually increases with consumption. Returning to the salve example, while the first dab of salve produces very little relief, the relief offered by the 10th dab of salve (the one that relieves the final sting) is tremendous by comparison. Thus, from the first dab to the 10th the marginal utility actually increases. From that point on the marginal utility of additional dabs must decrease rapidly.

So, for the math savvy out there, Karelis argues that the utility curve has an inflection point that is determined by the level of consumption at which the agent perceives that he/she is meeting basic needs. Below this point marginal utility increases with consumption, while above this point marginal utility decreases with consumption.

What does this mean for behavior? Remember that we are still operating within conventional economic theory. We have just tweaked the theory a bit. So, rational behavior should still be that which maximizes utility. First, consider the case of pleasing goods, i.e., the case of a relatively wealthy consumer. Since marginal utility decreases with consumption for the wealthy it makes sense to try to keep consumption as smooth as possible. Why is this? The fundamental insight is that if marginal utility decreases with consumption, then the utility you gained from spending the next to last dollar was more satisfying than spending the last dollar. The net result is that it is more painful to lose 50% of your consumptive potential than it is satisfying to gain 50% more. This is essentially the rationale behind saving. It allows one to consume at a consistent level without, say, being quite poor every once in a while. The decreases in utility caused by being quite poor (say in a transition between jobs) would outweigh the utility gained from a larger consumption during the well-off periods. Balancing these out results in the optimal saving strategy.

Now consider the case of a poor person who, on Karelis's theory, is not meeting basic needs and thus experiences increasing marginal utility with consumption. The result is that smooth consumption does not maximize utility. Because marginal utility increases, one has every incentive to splurge one day and be very poor the next. To again use the salve analogy. If I am "earning" 7 dabs of salve on odd days and 3 dabs on even days which are used to treat my 10 stings, it is more rational to save dabs every odd day and splurge by using all 10 dabs available on the even days so that half the time I am in no pain and half the time I have 10 painful stings. The idea is that the pain caused by the 10th sting relative to the other nine is very small, and similarly that the psychological difference in pain between 5 stings and 10 stings is less than the difference between five stings and no stings. The result is that it is irrational to consume a constant rate of dabs, say 5 dabs per day, since that results in a relatively low utility since I am in pain every day.

It is hopefully becoming clear how this starts to make sense of the behavior of the poor. In the case of job hunting, if the difficulties of job hunting, or even having a job, are likely to smooth out the pain of being poor, so that I am not ever destitute but am not meeting basic needs either, then the rational thing to do is to break the smoothness of the consumption to produce high highs and low lows. The unfortunate result is rather erratic behavior, often involving narcotics, failure to show up to work on time, etc. In short, the uneven consumption that maximizes utility also makes it difficult to keep a job, save for the future, etc.

This is Karelis argument in significantly abbreviated form. He goes into much more detail to show how this theory fits the actual data concerning the behavior of the poor, as well as avoiding some rather silly pathologies of other accounts (such as suggesting that all poor people are irrational).

In the last two chapters he applies this theory to specific policies addressing poverty such as the Earned Income Tax Credit. The EITC has been a success despite what appear to be competing incentives put in place by the policy. The two counteracting effects are called the Income Effect and the Substitution Effect. The Income Effect states that if you offer a higher hourly wage, the worker will have less incentive to work an additional hour since his needs are being met by fewer hours. The result is as incentive to work fewer hours and take more leisure time. The substitution effect states that if you offer a higher hourly wage the cost of leisure time then also increases (because each leisure hour is an hour you could have been working for this higher wage), and one may decide that it costs too much to take leisure time and thus decide to work additional hours. On the conventional theory, these two effects work against each other. However, on Karelis account the increasing marginal utility reverses the income effect since the marginal utility of those extra hours is actually larger than the marginal utility of the previous hours worked (at least up to the point that the worker is meeting basic needs). The result is that what are competing economic incentives for the well-off are actually reinforcing effects for the impoverished. Karelis therefore endorses the EITC as a doubly good policy for combating poverty.

That is enough to get across the premise of Karelis book. The question is whether it has any implications that are of use for us. It likely at least helps understand the behavior of the poor by putting it into a framework that may show us what sorts of responses to expect from the poor. In addition, I think that it suggests that large gifts to the poor are in order. That is, while small gifts of money here and there are likely to be used for alcoholic escapism, to voluntarily raise the standard of living of a poor person to a level that meets basic needs through a simple gift might be enough to get that person "back on their feet" by altering the incentives (as diminishing marginal utility takes effect and encourages consumptive smoothing). In effect, the poor person has more to lose by engaging in erratic consumption than before.

So, my first suggestion is that Karelis argument results in a justification for much more radical generosity to the poor if one is to discourage self-destructive behavior.

Secondly, Karelis makes the off-handed remark that the perception of basic needs being met may also include non-material needs, such as inclusion in society. For obvious reasons he does not dwell on this remark. For our purposes, however, it is significant. It suggests, in the same way as the first point above, that a radical inclusion into a community may be integral to breaking the persistence of poverty.

Finally, Karelis does not explore in depth the communal aspect of utility maximization. There is an argument embedded in his picture for maximizing the collective utility, whereby clearly if I am meeting my basic needs then a marginal increase in my own consumption would produce far less community utility than would my gift to an impoverished person to use that marginal resource (since their marginal utility is increasing with consumption). The result is that in a resource-scarce environment it is inefficient/irrational for any one to use more than is necessary to meet basic needs. It also means that it is more efficient for nearly everyone to be provided basic needs with a few who are truly destitute than for everyone to be just short of meeting basic needs.

In summary, Karelis gives us a psychological framework based on conventional microeconomic theorizing for understanding what behavior to expect from the poor and how it is rational. His aim is to understand the rationale for the behaviors of the impoverished that seem to promote the persistence of poverty. His recasting of the conventional framework also suggests, on my reading, that in order to combat the self-destructive tendencies of poverty our gift giving should be more generous rather than less --such that we should provide for the basic needs of our neighbors (this is not, apparently, to "enable" as, ironically, the giving of lesser gifts would be)-- and that the gifts we give should include membership in a community where social needs may be provided. These conclusions are not explicitly the ones that he was looking for (he is more concerned with evaluating government policy), but it suggests that a Christian anthropology, and the economics it requires, is rational even by social scientific standards (as one would expect since the behaviors of real people are the data).

JR