Bryan Caplan  

Poverty: The Stages of Blame

Media Bias: RT May Have Less T... An update on Mr. Renzi...
I've repeatedly argued that there's a connection between (a) how deserving the poor are, and (b) how the poor ought to be treated.  Unfortunately, as soon as I make this deliberately vague claim, many readers rush to ascribe specific, silly views to me.  To preempt future misinterpretations, I now sketch my view in greater detail.

1. Claims about desert and poverty are meaningful.  Asking, "Does he deserve to be poor?" can be rude, but that doesn't mean the answer is "No."

2. A person deserves his problem if there are reasonable steps the he could have taken to avoid the problem.  Poverty is a problem, so a person deserves his poverty if there are reasonable steps he could have taken to avoid his poverty.

3. Common sense can usually resolve whether reasonable steps to avoid poverty were available to a particular person.  A good rule of thumb: If you wouldn't accept an excuse from a friend, you shouldn't accept it from anyone.

4. The fact that a person deserves his poverty does not imply that it is morally wrong to help him.

5. However, the fact that a person deserves his poverty is (a) a strong moral reason to give him low priority when weighing how to allocate help, and (b) a strong moral reason not to force a stranger to help him.

6. The fact that a person does not deserve his poverty does not imply that it is morally wrong not to help him.

7. However, the fact that a person does not deserve his poverty is (a) a strong moral reason to give him high priority when weighing how to allocate help, (b) an extra moral reason for individuals morally responsible for his poverty to cease and remedy their wrongful behavior, (c) a moral reason to force these morally responsible individuals to cease and remedy their wrongful behavior, and (d) a plausible though not totally convincing moral reason to force strangers to help the deserving person if the benefits heavily outweigh the costs.

Coming soon: What these claims imply about government policy and personal behavior.

HT: Bill Dickens for spurring me to clarify my position.

Comments and Sharing

COMMENTS (36 to date)
Nicholas Weininger writes:

"A person deserves his problem if there are reasonable steps he could have taken to avoid the problem."

Do you really believe this as a general principle? Being a victim of violent assault is surely a problem, and often a bad one. Does it follow that, if someone is walking through an unsafe neighborhood alone at night and is assaulted, and they could reasonably have taken a safer route, they therefore _deserved_ to be assaulted?

Sudhir Jain writes:

Who gets to define "reasonable" steps? The neo-liberal or the socialist or the "objective" academic (if such a creature exists!)?

DJ writes:

Re: Nicholas's comment - I think the word "deserve" carries too much emotional baggage.

Turn the knobs on your hypothetical a bit:
(1) If someone is walking through an unsafe neighborhood at night wearing a "Please assault me!" sign and they are assaulted, did they "deserve" to be assaulted if they could have reasonably not worn the sign?
(2) If someone is walking through an unsafe neighborhood at night and then they trip and fall, did they "deserve" to be injured if they could have reasonably taken a safer route?
(3) If someone is walking through unsafe woods at night and they are attacked by a bear, did they "deserve" to be attacked if they could have reasonably taken a safer route?

These all have the same essence as your scenario, but with these I feel less indignant when I consider a "yes (to some extent)" answer.

Becky Hargrove writes:

Don't forget that poverty always exists in a larger context that goes beyond the individual. For instance, say someone was gainfully self employed, when circumstances changed and bankruptcy ensued, hence (possibly) poverty as well. Do we blame the individual for not getting out in timely fashion, to do something supposedly "failproof"?

Daublin writes:

@Nicholas, you are hovering around in the gray area. Consider a more extreme version of your example. For example, suppose someone actively seeks out the roughest neighboorhoods, so much so that they walk an additional half hour home each night instead of taking the direct route. Everyone they see they heckle and insult, just begging them to lash out. You've talked to this person a dozen times about how dangerous this is, and they keep doing it anyway.

Such a person deserves what they get when someone inevitably attacks them. That doesn't mean the attacker is in the right, and it doesn't mean anybody is happy about the situation. It just means you wouldn't go "poor Johnny, he never did anything to deserve being beat up in an alley".

Steve Roth writes:

Addressing it from a bigger-picture perspective.

Assume an economy in which an unchanging 10% of people live in poverty.

Could all of them "take reasonable steps to avoid the problem?" If they all did so, would poverty be alleviated?

Or are they inevitably subject to larger forces? Do they happen to be the ones unlucky enough to land there? (And luck might include being born stupid, ugly, or lacking in motivation, or just...poor to begin with...)

If they could all overcome their failings through force of will, would poverty go away? It certainly never has.

Scott Freelander writes:

How can it fail to be morally wrong to refuse to help someone who deserves help?

The sort of libertarianism that is hinted at here strikes me as the same that Milton Friedman espoused when he favored a negative income tax. As he once explained to William F. Buckley, he only favored it because it was superior to the then current welfare programs. He said he did not think that people should be forced to pay for any programs to help the poor against their will, and he hoped inflation would gradually eat away at such a program's benefits.

Well, that's a nice cop out, isn't it? It's like the pacifists who oppose the draft, even when the country they live in actually comes under dire external military threat. Instead of staying and fighting, they flee the country, if the enemy starts closing in.

People like Einstein were practical enough to give up pacifism in the face of the threat of fascism in World War II. Even Gandhi favored letting the Allies use Indian bases to conduct military operations during that war.

But, wide-eyed, holier-than-thou, crazily idealistic libertarians are locked into an all too convenient ideology, which allows them to hide from the consequences of the human condition by denying the nature of that condition and of humanity itself. Some of them even take it as far as anarcho-capitalism, which has all of the realism of Leninism and Stalinism, with just as much morality.

In the real world, adults sometimes have to make decisions and choose winners and losers. Such choices will never be perfect or fair, and no system is without ludicrous injustices.

All of politics is a contest of force and will, with the force of violence as the ultimate arbiter. You can hate this and you can deny it, but you cannot escape it. There is not a government in the world that will not ultimately enforce its laws with the threat of violence, including its tax laws. And I cannot imagine how it can ever be different. It certainly never has historically.

The only question is, toward which ends should force be applied?

It's far worse to stand aside, pretending to be superior and trying to make judgments about who deserves help, than to pitch in, get your hands dirty, and be willing to break a relatively small number of cheapskates who don't want to contribute to universal entitlements or national security.

Some of us are adult enough to understand that freedom is always limited and that we must fight to make sure our preferred freedoms and limits prevail. We certainly need more libertarianism in a number of ways, but the idea of trying to decide who deserves help, and even then not to force tax money into the hands of those who need it most, is disgusting.

philemon writes:


No, it doesn't follow. In general, it doesn't follow from the mere fact that

(1) X happened to S


(2) There are reasonable steps S could have taken to avoid X

But Bryan made no such inference. Whether or not (2) is the case has to be established separately from (1). In fact, if he has made such an inference, he couldn't have made the distinction between deserved and undeserved poverty--the distinction presupposes that there are some cases of:

(3) X happened to S, and there ARE reasonable steps S could have taken to avoid X (X-deserved)

And there are also cases of:

(4) X happened to S, and there AREN'T reasonable steps S could have taken to avoid X (X-deserved)

@Sudhir Jain

When faced with a particular case of (1), whether the corresponding (2) is also true is an EMPIRICAL matter involving judgments regarding the agent's capacities and the objective circumstances surrounding the happening. Incidentally, as far as I know, we make such judgments all the time.

For instance, we often excuse someone for having done something to us, that would otherwise have been blameworthy, because we recognized that there weren't reasonable steps that he could have taken to avoid doing it. Say someone crashed into me, knocking off the plate of hot soup I was holding, causing me to be hurt. Not nice. But it makes a difference whether he did it deliberately, or whether it happened because the floor was slippery, but not obviously so (objective circumstances), and he--like everyone else who is not a high level kungfu expert--has next to no chance to steady himself once he stepped on to the slippery spot, nor does he--like everyone else who is not superman--have super vision to have noticed the slippery spot, etc., etc. The perfectly fit guy with normal vision who stepped in despite there being a bright yellow sign saying "slippery, do not step", on the other hand...


Insufficient data to compute.

I hope that your "Assume an economy in which an unchanging 10% of people live in poverty" is not just another way of saying "Assume an economy in which 10% of people are such that there aren't any reasonable steps for which they could not have taken to escape from poverty." Or it would be simply question begging. So I'll take it that you are not making such an assumption.

In other words, all we know is that there is always 10% of the people who are poor. (And by that, I take it that it's not because of a rigged definition of poverty such that 10% will always be poor no matter what happens, e.g., the bottom 10% is by definition 'poor'.)

But the question will remain: Why are they poor? What are the objective circumstances? What are their capacities like? What are the options available to them? Again, these are empirical questions. And presumably not easily answered too.

While I'm willing to go with Bryan Caplan on the distinction he is drawing and its moral significance, agreeing on this point is compatible with also believing that, as a matter of fact, all or almost all of the people living in a particular society at a particular span of time are such that there aren't reasonable steps they could have taken to escape poverty. That's not an impossible outcome for me. Bryan has not established that such an outcome is impossible, nor does he say or imply that he has. I take it that the post is only meant to put forward the moral distinction. What the moral distinction entails by way of our actual judgment of actual people in an actual society and thus what we do with them will depend on how the empirics work out.

Not that I have a lot of hope that getting the empirics will be easy or uncontentious.

Pajser writes:

How about this line of thinking:

(1) If Poor John risked his health to save someones child, and Poor John was seriously injured in attempt, he didn't deserved his poverty, although he had reasonable chance to avoid his own poverty by simply - not risking his life.

(2) If Poor John is poor because he was injured risking his health to save your child, and Poor Ahmed is poor because he is injured risking his life to save my child, you have somehow stronger moral obligations to help Poor John than to help Poor Ahmed.

(3) If Poor John risked his health to save your children, he wasn't injured, but he is poor because he is lazy, you have stronger moral obligation to help Poor John than to help Poor X who is just lazy.

(4) If you're citizen of USA, then every mentally healthy adult USA citizen risks his health to save your children. His very decision to be US citizen means that he accept the possibility that he might be recruited to fight against aggressor. Admittedly, it is not much of risk and not much for your children, but yes, it is some risk and to some degree, for your children.

Kevin Driscoll writes:

Since we're talking ethics:

When you say "reasonable steps he could have taken to avoid the problem," do you mean could have done otherwise in the (philosophical) libertarian free will sense? Is it enough that he could have done otherwise, assuming some minimum of circumstances were different (a more soft determinist sense)?

I'm just trying to point out that you run into a lot of problems with libertarian free will/hard determinism/soft determinism when you start considering counter-factuals and possible worlds.

john hare writes:

I think the do not deserve people are more the following than the examples mostly given above.

A person that refuses to look for a job.

An addict (drugs, alcohol, gambling, etc) that won't try to get clean.

Someone that will not even try to learn to cope with current society when given numerous opportunities.

One that is a criminal by choice.

And so on as there are numerous types of people that we should never be forced to help/enable.

Al writes:

There are downsides to focusing on statistical descriptions without a historical backdrop. For example, one might say that black Americans deserve lower income, worse housing, and stagnant upward mobility because the group tends to show lower academic achievment and higher crime rates. Obviously, Caplan did not say anything close to this, and it ignores point 7., but this argument is a vulgarization of points 1 - 6. The easy availability of certain data points can lead to underestimation of the variability of human outcomes.

What if you took a child, and regularly impaired his ego by making him feel worthless and incapable? He may indeed grow up to be an incapable man. Individually, you might see him as being undeserving of support. But as a group, his unimpressive behavior might indicate some underlying processes that should be interrupted.

Curtis L. writes:

Interesting. I'm left-leaning but I didn't find anything offensive. A few reactions, though.

If we are going to attach desert to material wellbeing, that opens up a whole can of worms. If we shouldn't offer support to the poor who deserves it, what does that imply we do to the wealthy who does not deserve it? I've got no clue.

There is one demographic of the poor that never deserves it and that is children. I believe in helping children first and foremost. You made a good case for that. Thanks!

Interesting that you always use the masculine pronoun. Never feminine. That's usually a bad thing. In this case not so much. I think if we were to identify the "deserving poor" it would be majority male. So again, let's focus on the moms and kids. I've got no problem with that.

McMonster writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment and your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Bostonian writes:

The poor are on average less intelligent, disciplined, and hard-working than the non-poor, but we know that IQ is highly heritable, and there is some evidence that character is as well. The Stanford Marshmallow Experiment showed that self-control of preschoolers was positively correlated to later life outcomes.

To what extent the poor deserve to be poor is arguable, but giving them money because they are poor discourages them from becoming self-sufficient. Programs such as the Earned Income Tax Credit, which makes large payments only to poor people with children, may encourage them to have more children, who will inherit the same genes increasing the likelihood of poverty. These considerations should be weighed against the declining marginal utility of income. Previous blog posts have noted that happiness is more highly correlated to log(income) than income.

Bostonian writes:

My observation about the declining marginal utility of income in the prior post is an argument for welfare spending, since the poor "need" the transferred money more than the non-poor. The trade-off between need and the disincentive effects of welfare is tricky.

Jeff writes:

Scott said:

How can it fail to be morally wrong to refuse to help someone who deserves help?

I think it depends on what kind of help we're talking about here. If I see someone drowning and I fail to throw them the life raft I've got handy, I think everyone would characterize that as wrong. On the other hand, suppose I don't have a life raft, the seas are really rough, and I'm not a great swimmer to begin with. Am I obliged to jump in the water and try to save the person, at great risk to my own life? I think most people would say no, even though they'd agree that virtually anyone who's drowning is "deserving" of help. In the same sense, I don't think people are morally obligated to make big sacrifices of their own property or wealth or time or what have you to help others, even if those people are deserving of help. What constitutes a "big" sacrifice, though, is going to mean different things to different people.

RPLong writes:

I'd like to try to express Nicholas' comment in a way that I think will resonate a little better with some of the hardcore rationalists out there:

Caplan's point #2 seems to carry a strong risk of hindsight bias. The level of this kind of bias could potentially be enormous.

Finch writes:

This is related to what Bostonian said:

How can desert be dependent on taking reasonable steps to avoid problems when being able to identify and take those steps depends heavily on things like IQ and conscientiousness which are highly heritable, and therefore essentially randomly endowed? You didn't "deserve" your parents in any meaningful sense.

Doesn't this render "deserving" meaningless? At least as a moral tool. It may still make sense to act as if it's real, punish cheaters, execute murderers, and reward inventors, etc, because that gets the most out of imperfect people.

But I'd bet we get better results by dropping the sloppy thinking and emotional baggage of desert as a moral concept. This answers all the difficult questions you dodge in your post about what happens when someone else deserves something more than you. You have to pick and choose when to apply your reasoning and give up what you have for someone else, whereas I just get to say "there is no such thing as deserving."

Randy writes:

Its a fun exercise, and well thought out, but of course most of the poor do deserve their poverty, which is precisely why Social Security, Medicare, and now the ACA, are not designed as welfare programs. The political class wants the ability to manage people without having to answer such sticky questions.

And if the result of such programs is the creation of some free riders, well... so what? After all, the whole point of a political system is to create free riders. It only becomes a problem if and when it hinders the ability of the members of the political class to be free riders.

Chris H writes:

@Finch and Bostonian,

Wouldn't your points on IQ and its relations to heritability and poverty be something that is part of capacity defining rather than an attack on the idea of "moral desert?" While IQ and poverty are correlated that does not mean all people who are poor are low IQ, or that all of them did not recognize actions they could take to avoid those circumstances. Your point is thus an empirical one on the extent of deserving vs. undeserving poor, not a point on the existence of that distinction as you seem to be making by my reading.


I just find it interesting that you decry libertarians for being crazy idealists, when the libertarian in question is crafting a nuanced point on a moral question, while you seem to want to take the more black and white stance of those who don't deserve to be poor do deserve to be helped. I'm not taking a stance as to whether those are the same thing, but rather simply pointing out that that also sounds idealistic (and so does the "people have to go out and act" line). Perhaps it's more cynicism that you're complaining about?

TMC writes:

"Interesting that you always use the masculine pronoun. Never feminine. That's usually a bad thing."

No, it's common English usage.

Finch writes:

Chris H, you make a reasonable point.

Perhaps one must have an awful lot of faith in the concept of free will to believe in desert? I might lack that faith - I believe in physics, so I view what I do as largely a consequence of the state of the world an instant prior. I can imagine other people thinking differently.

Chris H writes:


The free will question is an interesting one. I think overall I actually lean more to your side than Caplan's which does make the "desert" question a bit moot insofar as it assumes free will. At that point I think we just reach either a utilitarian position of trying to determine which levers cause the least of stuff we don't like and the most of stuff we do (maybe deserving-ness can act as a heuristic there?), or a stance of moral nihilism at which point who cares. The former is a bit more appealing to me I suppose.

Al writes:

Chris H, why would you go from a "utilitarian" position to who cares? Counting utils would be meaningless without describing and delineating values. It's a process of everyone figuring what they care about, which, if you believe it, is simply a de facto description of how the world works anyway. There would be no noticeable difference between a utilitarian approach, and a values approach, unless the values had changed.

MingoV writes:

The line of reasoning used by Dr. Caplan for poverty also is used in assigning priorities for organ transplants. For example, a chronic alcoholic with liver failure gets a lower priority than an adult with primary biliary cirrhosis (a disease not affected by earlier behavior). The former "deserved" his liver failure, the latter did not.

Some people in the transplant world believe that the above policy is morally wrong. I believe their morals arise from misguided beliefs.

Matt_S writes:


If suffering is greatly out of proportion to a person’s mistake, how does that modify our sympathies?

For instance: Your friend becomes poor because he lost his temper, told off his boss, and was fired. His fault. In Actual America, he goes on unemployment or uses the social safety net to skate by until he figures things out. Your friend still likely grows old and enjoys modest comforts.

Now how about Dystopian America, where everyone who becomes poor dies of starvation within one month?

My question is: does this modify how sympathetic we would behave toward our friend, who is admittedly at fault? In other words, is it purely a question of fault, or does desert also depend on how much suffering results?

This is not me trying to puncture the argument, because Third World poor are probably both more deserving and have greater levels of suffering. But, conversely, if relatively small mistakes can lead to relatively severe suffering, it becomes at some point improper to say a person “deserves” their suffering. Consider: a policeman makes the mistake of forgetting his bullet proof vest, fails to wear it per departmental policy, and is shot and killed.

Did the policeman deserve to die?

No. His mistake was small and the price he paid was total. Even if there are other people who had even less a chance of survival and made even smaller mistakes, I do not think that policeman deserved to die.

Desert is not solely determined by our actions, but also by the consequences. I don't know if I've seen you address this before, but I would like to.

vikingvista writes:

How about we stop assuming poverty is an affliction? Disease is an affliction. Starvation is an affliction. Dependency may very well be an affliction. But not poverty per se. Some people deserve poverty not as a punishment, but as part of the life they knowingly and gladly chose.

Believe it or not, there really are people who live on incomes well below the poverty line, are independent, reasonably happy, and would be offended at the condescension of sympaths.

Seth writes:

"Being a victim of violent assault is surely a problem, and often a bad one. Does it follow that, if someone is walking through an unsafe neighborhood alone at night and is assaulted, and they could reasonably have taken a safer route, they therefore _deserved_ to be assaulted?"

Who's the assailant with poverty?

Matt_S writes:


There are two problems brought up by the assault hypothetical and you rightly knock out one of them. 1) Desert in instances of victimization. 2) Desert and problems of proportionality.

You're right to point out that desert is not much of a factor when there is an assailant. Our society awards nearly all blame to assailants during any type of an attack, and poverty is not an "attack." My own hypothetical involving the shot policeman, above, falls into the same trap, so it's a good criticism and I feel compelled to address it.

But there is a second problem, and that is when the cost of your mistakes is greatly out of proportion to the mistake.

Therefore, consider an alternative hypothetical that controls for the assailant problem, but not for the proportionality issue: A teenager chooses not to wear his seat belt, crashes, and dies as a result. It seems likely that wearing his seat belt would have saved his life.

Did the teenager deserve to die?

Keep in mind we are not asking, "Should the teenager have worn his seat belt?" Only whether he deserved death for not wearing his seat belt.

Brad writes:

“…that it is morally wrong to help him.” Who is doing the “helping” here? If the government takes resources from one and gives to another, is this not morally questionable, no matter the circumstances? It undoubtedly has the potential to create a moral hazard problem. How do we avoid this conundrum? Let private actors do the helping.

Charities, churches, and individuals are much better-equipped to dish out “help” to the poor. Let’s not forget that churches throughout history have done “more” for the poor than any coercive government.

Chris Thomas writes:


I'm sure you don't intend it to, but doesn't 2. imply that people engaged in civil disobedience deserve to be arrested? Being arrested is a problem, and there are certainly reasonable steps a person can take in this case to avoid being arrested, namely, don't engage in civil disobedience.

Devon Herrick writes:

Your point is well thought out and I believe supported my most people. If you look at poverty policy in the United States, it reflects your view to some degree. For instance, Medicaid has always been more generous to children; half of births are paid for by Medicaid. Medicaid eligibility is more generous to poor children than to their parents. Obviously, children have no earning capacity and society takes pity on them since being born into poverty is beyond their control. In addition, universal education provides for the fact that children should not be disadvantaged because their parents cannot afford to pay for schooling them. Likewise, seniors (who are mostly too old to work and suffer health problems) are provided for -- most notably through Medicare.

If you look at our safety net, it follows to some degree the notion of deserving poor versus those who could avoid it with more effort. Children and seniors are the ones we (as a society) view as most deserving of assistance. Single adults without dependents, not so much.

Floccina writes:

Some low income Americans are unwilling to do the things that I did. I am ok with that, I think they have valid values but I am not so motivated to give them money on the other hand some poor people work hard, are fugal and mind their own business and I like to help them and they generally do not need much help.

Sebastian H writes:

"Am I obliged to jump in the water and try to save the person, at great risk to my own life? I think most people would say no, even though they'd agree that virtually anyone who's drowning is "deserving" of help."

Part of the problem here is that our intuitions in the US are formed under the framing of equality or at least not dramatic inequality, but the reality is of very pronounced inequality. The rich are not being asked to jump in the water at great risk to their own life, they have three life rafts on their yacht right now.

Professor Hulk Hogan writes:

[Comment removed for supplying false email addresses. Email the to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Comments for this entry have been closed
Return to top