Redistribution: In Cash or In Kind?

Should government anti-poverty programs transfer income in cash or in kind? 

Transfers-in-cash means programs like welfare or the Earned Income Tax Credit; these provide unrestricted income that recipients can spend as they see fit.  Transfers-in-kind means programs like food stamps, subsidized housing , Medicaid, or public schools; these force recipients to spend their income transfers on the particular goods the government dictates.

The argument for cash transfers is that the needs of recipients vary, so recipients can benefit from choosing how best to spend the income they receive.  The argument for in-kind transfers is that some recipients might spend cash transfers unwisely.

My view is that cash transfers are better overall.  Cash is unambiguously better for households that make good use of the flexibility.  And the households tempted to misuse this flexibility are not readily manipulated by government restrictions.  For example, such households can choose bad neighborhoods, or fail to encourage their kids in school, or avoid using government-provided health care.  So, policy should focus on those most eager to help themselves.

Beyond these considerations, government paternalism (whether towards the poor or otherwise) opens a Pandora's box that permits ill-intentioned as well as benign uses of government power.

Comments

[Peter Meyer]

Agree and disagree. I agree with the danger of government paternalism. The apparatus required for government to provide in-kind transfers is likely inefficient and produces a larger public sector. However, in-kind transfers also provide an incentive for recipients to work because they will want cash for discretionary consumption beyond food, shelter, and health care. Best is a no transfer scenario.

[Mike Huben]

It amazes me how this argument blithely assumes away important economic facts. I guess there's no other way to fit the square peg of reality into the round hole of ideology.

Yes, cash does make sense for basic welfare. BUT, real world evidence shows that other significant problems persist.

First, much aid goes to people who have problems with rational behavior: who might prefer to buy liquor than food, housing, or medical care for themselves. Yes, with effort they could convert the value of the aid to liquor. But if they have trouble with rational behavior, that will be difficult. I suspect more aid is delivered correctly in these cases, and that would be easy to measure in changes in health and homelessness.

Second, there are serious market failures for some goods because of economic irrationalitiy: vaccines for example. It is difficult for individuals to estimate their benefits from diseases evaded due to vaccines. So if you are offered $10 to get a shot, versus offered free shots, many will not get the shots even though probabilistically the payoff for the shot is often 50 times the cost.

Third, there can be significant externalities for some goods. Vaccines, once again: when a high enough proportion of the population is vaccinated, the remainder is protected because the disease cannot spread. Education as well: educating children reduces crime.

Fourth, there is the agent problem due to family structures. Society expects heads of households to act responsably for the other dependent members of the household: children, spouses, aged, sick, etc. But of course that's not always going to be realistic: addictions, for example. Yes, with effort they could convert the value of the food stamps to liquor, but I suspect more aid becomes food in these cases, and that would be easy to measure in changes in childhood (and adult) malnutrition, which has gone down dramatically since the issuance of food stamps.

Fifth, cash presents a greater moral hazard, ie. people lining up for cash who might not otherwise line up for food stamps. If we want to target scarce dollars to particular benefits, this moral hazard must be avoided.

Sixth, don't you think the cash option has been tried? This is a world of second best solutions, because simple best solutions don't exist for all problems.

I could probably list many more obvious problems with Miron's position, but how much time is it worth?

So much real economics to ignore, and so little time to do it! The demands of his ideology are harsh.

[Chris]

Mike, If you looked into the issue at all, you would know that it is very easy to convert food stamps into cash which makes your entire argument moot. The biggest difference is that the targets of the aid only get a fraction of the total aid that is intended for them (something like .90 on the dollar).

Add in the fact that food stamps create additional overhead that cash payments do not and the number of retailers that accept food stamps is less than the number of retailers that accept cash I think your argument falls apart.

As to why you start talking about vaccines, I have no idea.

One thing that I find peculariar is that many liberals cannot tolerate the concept of vouchers when talking about education, but vouchers for food are OK.

[Chris]

I was wrong, typically food stamps only sell for 50 cents on the dollar nationwide. My experience was in central IL which (I guess) had better exchange rates.

[Mike Huben]

Chris, I appreciate your self-correction.

But don't you think that receiving only 1/2 the value of a food stamp is a deterrent to converting it to cash?

And my argument is not mooted by the fact that some conversion occurs: the benefits in terms of improved nutrition are measured.

I brought up vaccines because they are frequently given for free to the poor: a means of redistribution in kind that is very important.

And of course food stamps have overhead: so do credit cards. You have a problem with that? Some retailers don't accept credit cards either: is that a problem too?

I've been examining libertarian arguments for over 30 years now: please don't presume I haven't looked into these issues.

[Chris]

So you discount evidence that contradicts your opinion, I guess that is a clear way to convince yourself that you are right.
The conversion rate for food stamps is low because there is a large number of people that are looking to convert. In one article I found (anecdotal to be sure) the writer reported that thousands of people sell food stamps every day in New Orleans. The low exchange rate doesn't seem to be much of a deterent.

You credit increased nutrition to food stamps which seems tenuous at best. Most of the increased nutrition comes form the fact that we are wealthier, as a nation, than we were 50 years ago, I can think of other causes as well. Unless you can site some study that controls for food stamps you are speculating at best.

You cannot refute my concern that food stamps are not efficient by providing examples of voluntary transactions that have nothing to do with helping the unfortunate. Miron's point is thatfood stamps are an inefficient form of assistance. I provided some additional examples as to why this is true - instead of addressing the primary point of the argument you attack a straw man.

[Mike Huben]

Chris, frequent violation of a law doesn't mean the law has no deterrent effect. There are roughly 40,000 murders per year inthe US, despite the laws against them. Unless you experiement with higher exchange rates, you won't be able to say the low rate has no deterrent effect. But standard microeconomics suggests strongly that it would.

And of course, studies were conducted immediately about whether food stamps helped. Not 50 years later, as you suggest. So food price changes have not been such a large variable.

And finally, you may want to be careful with the term "efficiency": Miron didn't use it, probably because in economics it has a special meaning, referring to several different measures of improvements (Pareto and Kaldor-Hicks, for example.)

[Chris]

I never claimed that it had no deterent effect so I fail to see how a comparison to murder advances the discussion at all.

I fail to see how a study that is conducted "immediately" after food stamps are implemented can possibly show anythin. Changes to human health do not occur overnight. I would have to seriously question a study that attempted to show otherwise.

[soft cialis]

Nutrition must be be the priority of our health care system don't you think ?

[ffxiv gil]

And finally, you may want to be careful with the term "efficiency": Miron didn't use it, probably because in economics it has a special meaning, referring to several different measures of improvements (Pareto and Kaldor-Hicks, for example.)