Since my last post on the subject, Bryan Caplan has put up two very interesting posts over at EconLog about the distinction between the deserving and undeserving poor.  The issue is this: is there a genuine distinction to be made between the poor who deserve our help and the poor who do not?  If so, should our private philanthropic activities attempt to reflect that distinction?  Should our government's public policies?

The discussion began with Bryan noting that libertarians are more willing than most to make a distinction between the deserving and the undeserving poor, and that this might be one factor that leads people (falsely) to believe that libertarians are "callous."  I responded by saying that there are good reasons to believe that the distinction is a valid one, but that even so it might not be one that we want our public policies to reflect.  

Bryan thinks that distinguishing the deserving from the undeserving poor is relatively easy, or at least "no harder than a thousand other moral distinctions we routinely make."  The deserving poor, he goes onto say, are those who are poor through no fault of their own, and they include "children whose parents can't or won't take care of them" and "severely handicapped adults."  The undeserving poor  - those who are poor by their own fault – are those who could find work but choose not to because it is unpleasant, those who lose their jobs for "tardiness, absenteeism, or insubordination," and those who abuse alcohol and drugs. 

Furthermore, he says,  even if the distinction between the deserving and undeserving poor is difficult to draw, this counts against "forced charity" i.e. tax-funded redistribution.  Taxation is coercive, and there is a strong moral presumption against coercion.  Bryan writes: "'Any morally reasonable person would agree that I'm forcing you to help the deserving poor'  at least arguably overcomes this presumption.  'Who knows whether the people I'm forcing you to help are deserving?' does not."

I'll address the question of force later.  For now, let's take Bryan's point about the ease of making the distinction.  It's important to distinguish between two kinds of problem we might have in making it.  The first is a problem in specifying the criteria that are to serve as the basis of the distinction.  The second is a problem in determining when some particular individual meets those criteria.  Bryan's claim that it's relatively easy to make the distinction is based on his belief that the first of these problems isn't all that serious, but he doesn't address the second at all.  Even if we agree that someone is deserving if they had bad parents, and undeserving if they're lazy, it's not easy at all to tell which of these to is the dominant cause of a particular person's poverty.  Not without a lot of background information on the person's life, and even then weighing the significance of the numerous causal factors and determining their interrelation can be difficult.  And, of course, even the first problem is not as simple as Bryan makes it out to be.  What are we to say, for instance, of someone whose absenteeism is due largely to genetic factors and to factors relating to their early childhood environment?  There are difficult questions at the intersection of philosophy and psychology here about what we ought to count as being within or outside a person's control.

Another area of dispute between Bryan and I concerns the relative moral significance of false positives vs. false negatives in identifying the deserving poor.  To explain: in trying to identify the deserving poor, there are two kinds of errors we might make.  We could identify someone as deserving who is not (false positive), or we could identify someone as undeserving who in fact is deserving (false negative).  I claimed that attempts to reduce false positives will likely increase the incidence of false negatives, and that most plausible moral views will hold that false negatives are moral morally troubling than false positives.  In other words, it's worse for someone who deserves your help to not get it, than it is for someone who doesn't deserve your help to get it.  Think of the criminal justice system as an analogy.  Most of us think that finding someone guilty who isn't is much more troubling than finding someone innocent who is guilty.  And our institutions reflect this judgment by incorporating a variety of factors (such as high burdens of proof on the prosecution) that make false positives less likely, even at the cost of increasing the likelihood of false negatives.

Bryan Caplan is not convinced.  Bryan thinks that (1) most of us do not apply this standard to our own behavior when, say, deciding whether to give a beggar money, (2) it is reasonable for us not to apply this standard to our own behavior, given the costs involved in gaining the information necessary to more consistently avoid false negatives , (3) there is no reason for government to be held to a different standard than individuals in this respect, and (4) even if false negatives are worse than false positives, it's still hard to defend forcing taxpayers to pay more in order to avoid this evil.

In response to (1), Bryan might be right.  Reflection on the Wason Selection Task suggests that we have evolved to care a lot and be very good at detecting cheaters – people who try to take a piece of the social pie without doing the necessary work.  We don't care so much about people who contribute to social goods but don't get the benefit they deserve.  People like that aren't a threat to us, even if their situation is unfortunate in some sense.  So false positives probably bother us a lot more than false negatives in a wide range of social contexts.  Even still, I think that there are a lot of distracting factors that weigh in to our decision of whether or not to help a street beggar.  There are a lot of such persons asking for our help, there are many others who could provide the help, we might have certain background beliefs of the people who seek help in that way rather than by, say, going to a church, and so on.  So it's not clear that it's the false positive / false negative distinction that's really doing the work in Bryan's example.

As for (2), Bryan's response seems to assume that the only way to avoid false negatives is to gather more information in order to more accurately determine desert.  But the analogy with the criminal law above shows that this isn't the case.  Another way to reduce false negatives is to be more tolerant of false positives.  Suppose you have 10 people asking you for a nickel, 5 of whom are undeserving and 5 of whom are deserving.  Sure, you could launch an inquiry to determine who's who.  But if each false positive only costs you 5 cents, maybe the most cost-effective and least intrusive way of reducing false-negatives is simply to give a nickel to everyone.  This, I take it, is part of the argument for a basic income, as I've discussed elsewhere.

(3) doesn't sound right either.  One reason we might not care much about false negatives as individuals is that we believe there is a safety net in place that will catch anybody we miss.  But if government is that safety net, then government needs to be much more cautious than we do.  Additionally, government might have reasons relating to political legitimacy for ensuring that its rules provide everyone with the opportunity to live a good or decent life.  This would give governments extra reason to avoid false negatives that would not apply to individuals.

Finally, Bryan is certainly right to claim that there is a presumption against the use of force.  But what it takes to overcome that presumption depends on your moral priors.  If you believe very confidently that people have a strong moral claim to their current holdings, then it will take a correspondingly strong consideration to override the presumption against taking their income by force and redistributing it to others.  If, on the other hand, you believe that people's current holdings are significantly illegitimate (perhaps because they arose from past acts of injustice) then the burden will be lower.  Similarly, if you believe that it is a moral obligation for the government to serve as the safety net of last resort for the deserving poor, you might indeed think that the mere possibility that someone might be deserving will be sufficient to overcome the justificatory burden.  Finally, even if it's clearly true that there's a presumption against the use of force, it seems equally clear that the strength of this presumption depends on the severity and kind of force that's being used.  As much as libertarians like to say that "taxation is theft," most of us think (as Tyler Cowen has recently pointed out) that there is a difference between a burglar breaking into your house and taking your TV, and the government increasing the marginal tax rate by 1%.  And there's probably a difference between the government's increasing the marginal tax rate by 1% and its increasing it by 10%.  Context matters.

This post is probably too long already.  But I can't in good conscience submit a post on welfare-related issues to this blog without mentioning Daniel Shapiro's excellent book on this subject.  He knows much more about this subject than I, and probably disagrees with some of what I've said here.  Perhaps if we all ask him nicely he'll chime in with a post of his own on the issue!