An Egalitarian Argument for Punishing Poor People More Harshly

Consider the following argument: The same punishment for different people is not in fact the same. Thinking of a criminal fine is the easiest example. A $1000 fine levied on an offender who is a millionaire is simply not as serious of a sanction as a $1000 fine leveled against a poor criminal. The millionaire can pay the fine without noticing it, while the poor criminal may be subjected to considerable economic hardship. The result is that the $1000 fine will not have much deterrent effect against the millionaire. To get his attention we require a much harsher punishment. So far so good.

Now consider prison sentences. Here, it seems to me, that the situation is reversed. Even a modest prison sentence — say of six months to a year — could have a devastating effect on an upper class defendant. The opportunity costs of foregone income will be substantial. Furthermore, there may be far more ancillary consequences in the form of foregone opportunities, and the like. A lawyer, for example, might lose the ability to practice law in the future. In contrast, a poor criminal with few opportunities faces fewer costs from the same length of time in prison. The opportunity costs in terms of foregone income will be much lower, and if the criminal really has few opportunities to begin with then the prison term may have a relatively small impact on one’s future prospects.

If the analysis in the above paragraph is correct it suggests that those in poor communities receive dramatically less protection than those in richer communities. This is not simply because those in richer communities tend to have more resources lavished upon them in the form of protection and law enforcement. It is also that — to the extent that the population of potential criminals who might victimize them are also wealthy — they are protected by larger sanctions against criminals. In contrast, so long as prison sentences are insensitive to the wealth of the criminal, those in poorer neighborhoods are protected by comparatively weaker deterrents for wrong doers. (Again, assuming that the population of criminals likely to victimize poor communities is also poor.)

In order to achieve equal levels of deterrence across communities using incarceration, it seems to me that we must punish poor criminals with longer prison sentences than comparable rich criminals. It is not that poor criminals are more deserving of punishment, although my initial intuition is that that those who prey on the poor are more culpable than those who prey on the rich. Rather, the argument for punishing poor criminals more harshly is to equalize the level of deterrence provided to all citizens. A regime of equal incarceration, in contrast, creates dramatic inequalities in the level of protection provided to citizens of different socio-economic status.

Of course, another solution would be to adopt a method of punishment where the cost imposed on the criminal is less sensitive to the criminal’s wealth. The stocks or publish whippings come to mind…

You may also like...

14 Responses

  1. Orin Kerr says:


    This argument is based a number of assumptions that strike me as dubious. Among the dubious assumptions:

    1) Criminals generally or always victimize those in their socioeconomic class.
    2) The deterrence effect of criminal punishment is based on the financial hit of imprisonment as measured in absolute dollar terms of lost income.
    3) Rich people and poor people are equally likely to be caught and prosecuted for their conduct, and if they are prosecuted, they are likely to receive the same quality of legal representation and receive equal treatment by prosecutors, judges, and juries.
    4) It is possible to fairly, accurately, and justly administer a system that takes this alleged dynamic into account.

    These assumptions strike me as either questionable or just wrong, and my sense is that your argument depends on them.

  2. Jack Chin says:

    Isn’t this already the policy? It seems to me that crimes poor people are more likely to commit tend to be punished more harshly. It also seems to me that those in the criminal justice system are disproportionately poor (as evidenced by the fact that a supermajority have assigned counsel in the jurisdictions I know).

  3. I think another problem with the argument is that “poor” is undefined. If poor means someone who is destitute and has no family to feed, then prison may very well be a step up in quality of life.

    If poor means someone who work but earns very little to barely get by, then a prison sentence can be incredibly costly as a percentage of income. The millionaire may have a large opportunity cost, but the family can eat from investment income. The working poor has no such fallback.

    If you have any kind of diminishing marginal utility of income (see Sarah Lawsky’s posts on this – ), then an equal prison sentence may very well be necessary to inflict equal deterrence in the form of opportunity costs. Indeed, if you are going to measure opportunity cost at all, you might well need a progressive sentencing length to provide equal deterrence (and even then, only if all of the other assumptions pointed out above hold true).

  4. “The working poor has no such fallback.”
    Should be “The working poor defendant has no such fallback.”

  5. Nate Oman says:

    Orin: Assumption 1 strikes me as quite plausible. Assumption 2 is not necessary for my argument. 3 and 4 are, it seems to me, reasonable objections.

  6. Number 2 is implicitly part of your argument because the financial hit is the only basis for comparison that you use. There are other ways in which a poor person may be more greatly injured than a rich person by virtue of a prison sentence. For example, many poor people have their parental rights terminated because they are incarcerated.

  7. Bruce Boyden says:

    Nate, it’s true your 2nd paragraph doesn’t rely solely on lost income, but it does rely on lost income plus foregone opportunities. You might be able to throw in different levels of shame, although even that strikes me as questionable.

    But I doubt any of those are the most significant costs from being in prison. I would guess that the major costs are: being separated from the people you know and love; having your freedom severely restricted in multiple ways; and the risk of violence from fellow prisoners. If there is any difference between rich and poor offenders on those vectors, I suspect they tend to swing heavily against poor offenders, who are probably more likely to be imprisoned at distant locations, in higher-security facilities, among more violent fellow prisoners.

  8. A.J. Sutter says:

    While perhaps Nate intends this post in a Swiftian vein, it is depressing to see how many people respond in terms of the same economistic analysis as he adopts, including the focus on “costs” rather than “harms” (though Bruce is a bit more humane in what he includes). OTOH, the real target of what I hope is Nate’s satire seems to be packaged in the word “egalitarian,” so the question becomes: why?

  9. FWIW, the reason I respond in the same economic terms is to meet the argument head on to show its nuttiness even on its own terms. I don’t view incarceration this way at all. I suspect the others feel the same way.

  10. Nate Oman says:

    A couple of quick points:

    1. I agree that there are lots of inequalities in terms of how rich and poor are treated in prison. My only point is that even if rich and poor were subjected to the same treatment in prison the costs of prison would not be the same.

    2. Even if the costs of prison are largely non-pecuniary (I certainly believe that they are!) the pecuniary costs of prison are likely to vary according to income. (Although I think that the point about how the costs of incarceration to criminals’ families may vary with income is a good one.)

    3. I started thinking about this argument after hearing Dan Markel present a paper about differences in subjective harm from nominally equal punishments. My main point — to the extent that I have one — is that nominally equal treatment of criminals can have unequal effects on communities. While there are obviously huge administrative problems with my proposal and perhaps moral costs as well, it is worth thinking about how equality of punishment may have very inegalitarian impacts on the rates at which different communities are victimized by crime.

  11. Nate Oman says:

    Note: While criminal law is definitely not my area, I am aware of the way in which “poor” crimes are often given harsher sentences than “rich” crimes. This argument doesn’t necessarily justify such disparities, but it does suggest that they may have egalitarian as well as inegalitarian implications.

  12. TJ says:


    I think it is easier to frame your argument if you define “wealth” not so much in monetary terms, but more generally as a person’s accumulated social, educational, and monetary capital. What you are observing is a more formal statement of the old adage that you can’t deter someone who has nothing to lose. And while we can quibble at the edges, the basic intuition seems to me to be unassailable. Whether any good policy can be crafted in response to the observation is another question, and Orin has pointed out many of the problems.

  13. Dan Markel says:

    Not that you’re attributing any position to me in particular, but it’s worth noting two things briefly. First, I take most of the argument advanced by Nate to be structured or motivated by deterrence and egalitarian concerns, not retributive and egalitarian concerns, or retributive and non-egalitarian concerns. It’s interesting to think about the ways egalitarian commitments might shape a purely instrumental approach, rather than an “intrinsic intelligibility” approach that characterizes retributive justice theory. (In case anyone is interested, my paper with Chad Flanders that Nate alludes to is focused on retributive justice considerations, not instrumentalist deterrence.)
    Second, to the extent I’m right about Nate’s post being fashioned in largely instrumental and deterrence terms, it’s vital to think about the ways in which criminal offenders of varying backgrounds do and do not internalize the various signals meant to be emitted for the purpose of deterrence by criminal law and sentencing policy. On this point, and registering deep skepticism toward rules calibrated to achieve instrumental goals, some good work by Robinson and Darley was done in Oxford J LS a few years ago (i think 2004).