Redistributive Taxation: just for the hell of it?

Great Debates
1

http://slate.msn.com/?id=2072196

In this Landsburg article, he relates an interesting experiment in which one person is given various strategies for passing money to another person, anonymous to them.

What Landsburg thinks the experiment demonstrates (not conclusively, of course, but enough to suggest disturbing implications) is that while people are generally unwilling to give money to others they are oddly willing to pay money to force other people to give yet other people money, all parties being anonymous.

That sounds confusing. It IS confusing. Why are people doing such incredibly strange things like this?

Personally, I think his second to last conclusion is the most likely: people just somehow forget that there are no free lunches: that if anonymous person B is to get money, it has to come from SOMEWHERE, and that somewhere could be just about anyone. Even if they think it comes right out of the experimenter’s or university’s pocket, they still have absolutely no reason to think that person B is any more deserving than the experimenter.

I think that boils down to people being more liable to empathize with the person getting the money, which seems more immediate, concrete and most importantly: an act in which they can claim a generous agency. People are much less likely to see themselves as being involved in the taking of money from someone else in a way that exactly undoes their previous story of generosity, especially when that action takes place in the past, or is less concrete because it involves many individuals.

It’s certainly disturbing to suppose that people might just LIKE moving money around, utterly irregardless of who gets and who loses.

Ultimately, I think the most important elements here are what Landsburg points out about anonymity.

Now, as I see it, the strongest case (whether or not I buy it is a different thread) for a resdistributive tax is Rawlsian. The problem is, Rawlsian justification would demand that we care VERY MUCH about who we are “taxing”: not just about their income, but all of their resources and capacities with which we are endowed at birth.

The problem is, our tax system seems to display very little of that. If I, a well educated young guy with lots of energy and the standard allocation of 24hrs a day choose to sit on my ass all day, I get taxed very little. If I choose to spend my time working, I get taxed very heavily. That makes no sense: I’m the same person no matter what I choose to do. If the government has the right and the obligation to tax my income, then consistency demands that it also has the right and obligation to tax my leisure. I mean, to me, they are both goods that I can trade off for each other, if I choose. Why should my decision to consume one good or another change my social or moral obligation to provide for others? Why should my choice of how to spend my time affect what other people can legitimately demand of me?

As I see it, I might as well be anonymous to it: that the government takes things from my no more capable but, by his own choice, more hard-working friend and not from me, is pretty arbitrary and random of it. It really is as if it doesn’t much care who anyone is, who is really more deserving or capable of fulfilling social obligations. It doesn’t see me: it sees money that it can shuffle around (in my benefit, I should note).

The other side of the coin is the arbitrariness of the distribution. In the context of the experiment, Landsburg points out that it makes far more sense, even from an altruistic point-of-view, to keep ALL the money, and then donate whatever portion you were going to donate to an actual charity: someplace you KNOW for sure needs the money more than you do, as opposed to handing it off to some anonymous stranger, who, for all you know, might be better off than you, or a real asshole you’d never give a dime to.

The parallel here is this: in a lot of ways, being taxed by the government for redistributive programs is like giving to the United Way. That is, even the most thoughtless person can surely find some charities they think are more pressing and deserving than the others in the United Way’s roster: giving to the UW instead of those charities directly, is just pure laziness (or coercion from your employer sometimes, it seems)

Here’s where that line of thought takes me: if you are going to make it mandatory to redistribute money via a portion of my taxes, shouldn’t people at least get to choose WHAT their money gets spent on, so they retain at least SOME agency over it? I mean, consider homelessness. I think some organizations that the government funds with my tax dollars do great work: but some do work that I think is not only counter-productive to solving the problem itself, but actually works DIRECTLY against the aims of the other organizations! That’s nuts (it’s like subsidizing tobacco growth and then taxing tabacco purchase, as well as trying to get the public to stop smoking altogether) If I had a choice, I’d target my contributions at those agencies I’m pretty sure would do the most good. Sure, maybe you can argue that most lazy people are too uninformed to make such choices, but okay: let THEM be lazy: I’M not, and I DO want that degree of agency over where my money goes.

We don’t even have to get that complex: I certainly don’t want my money distributed to corporations so they can advertise in Brazil or wherever. I certainly don’t want it propping up farmers who no longer serve a productive purpose. If I’m forced to give my money away for fulfilling social obligations, at the very least I shouldn’t be forced to donate it to causes and organizations I think are insane or immoral, or work AGAINST the very social causes that are supposed to justify the appropriation of my money.

I’ve said a mouthful, and I don’t even have time to edit things down to size. Have at my musings. But be sure to read the article so you can see what I’m reacting to.

2

I think there are a lot of problems with trying to apply this experiment’s results directly to taxation.

Where to begin …

Let’s start with this part of the experiment:

This immediately removes it from the realm of taxation. A better experiment (that would never fly, of course) would be to require each subject to bring $10 of their own money to the test. If I find $10 on the street, might I not be more inclined to give some of it away than if I had spent an hour in a coal mine earning that cash? It doesn’t indicate whether or not that $10 is part of the subjects’ compensation for the experiment or not, but I think then we’d have to get into the whole issue of why the subject is participating in the first place, if they’re doing it just for the money, just out of curiosity, boredom, extra class credit in Psych 101, etc.

This is where it gets really tenuous. I don’t think a single test subject is concerned/aware that taxpayers fund this research. And, quite frankly, I disagree that it’s even necessarily true that the resarch is funded by taxpayers–most of this sort of stuff goes on at universities, where, in addition to NSF grants, etc., the researchers also get money from tuition, from private and corporate donors, etc. Besides which, why should the subject be concerned about how the researcher spends his grant? How is forcing the researcher to give away $2 any different than a test that consumes $2 worth of test supplies–let’s say a diet study, in which the subjects eats as much as they want in a one-hour period at the food sciences lab… should the subjects feel bad that they’re costing taxpayers more money if they eat more food? Hasn’t that $2 loss already been anticipated, budgeted, and approved anyway? Will the researchers somehow learn less from you if you make them give money away?

Then there’s the whole way that this is being couched (by the writer)–the test subject is “forcing” the taxpayer/researcher to give money to someone else. If the researchers had told the subject, “now, I really can’t afford this, but…” or “I don’t want to give this money away, but if you force me to, I’ll have no choice,” etc., then it might have some validity. But it could just as easily have been, “If you tell me to, I’ll go right ahead and give $2 to that person next door.” That sounds awfully nice of them! How charitable! Sure!

Really, though, you need to put yourself in the subject’s chair: You’re in a totally artificial situation, being handed play (unearned) money, and being offered bizarro choices of what to do with it. It’s a game, pure and simple.

All in all, this experiment is poorly constructed to draw any conclusions about people’s attitudes toward redistributive taxation (and as the writer noted, the researcher disagreed with the writer’s conclusions). I am curious re: what the researcher genuinely intended to learn.ˇ

3

Well, you’re being a little silly in treating this exactly how I said: as if it was anything other than a study in some aspects of human nature, which only afterwards can be extrapolated into taxation. As Landsburg notes, not even the experimenter agrees with his take, because no one even considered the experiment from that perspective (in a way being as guilty of the full reality of the complete system as the subjects themselves).

—This immediately removes it from the realm of taxation. A better experiment (that would never fly, of course) would be to require each subject to bring $10 of their own money to the test.—

Uh, no. Money in their pocket is money in their pocket. From the point they are given it, it’s theirs. If you are desperate to see it as income, then it’s income for spending their time in a lab.

—I don’t think a single test subject is concerned/aware that taxpayers fund this research.—

That’s EXACTLY the problem. If this is true, the question is: why do people FAIL to recognize this? Why do they think it comes from nowhere?

The other key factor is that no matter where they think the money is coming from, they know nothing at all about the person who is getting it. They are acting as if, by donating more, they are actually doing something positive: somehow making money. But all they are doing is moving it around: but why would they prefer one distribution to another? And why would the prospect of making a totally random redistribution of money actually be WORTH giving up their own money to them?

4

Landsburg’s article made me pretty angry. He takes the work of a Nobel prize winner and distort it to make a cheap ideological point. He’s amused me in the past (I liked his micro textbook and his The Armchair Economist), but this garbage pissed me off.

An example of a gift that has the “tripling” qualities Vernon Smith uses: donating clean blood. Another is the whole Pareto Optimal Redistribution literature started by Hochman and Rodgers and effectively ended by Brennan. If you care about other people and would like to see them better off, and there are many people like you, then contributions are a public good and potential donors have an incentive to free ride. A willingness to donate provides benefits to others beyond one’s own contribution. Another example: donations to public radio. Again a public good, beset by seemingly intractable free ride problems. But people give and provide benefits beyond their own contribution (ie the consumer surplus of pseudo-demands of all listeners).

This is well known. The experiments performed by Smith are known as the Voluntary Contributions to Public Goods experiments. They are NOT about voting. They are NOT about coercive redisribution. Quite the opposite - taken seriously they question the need for government provision of public goods even in large numbers because people do not behave in a Nash manner. Landsburg’s “interpretation” is nothing of the kind. It is at best a lazy misrepresentation. At worst it’s a sleazy piece of propaganda from someone who is not content to show that selfishness may not be a disaster, but has to insist that it’s somehow a virtue.

5

I disagree. In fact, I think he raises a rather troubling point about Smith’s experiment, in that it DOESN’T take into account the reality that the extra money has to come from somewhere. It acts as if they source of the money isn’t really an element of the experiment.

—At worst it’s a sleazy piece of propaganda from someone who is not content to show that selfishness may not be a disaster, but has to insist that it’s somehow a virtue.—

That’s pretty sleazy of you, considering he doesn’t make any such point.

The problem is the fact that you don’t know anything about the person who you are giving money too. You have absolutely no idea whether your donations make the world better off at all. That’s one reason why Vernon’s experiment doesn’t show what it purports to. If subject A could be assured that their money would do more GOOD than it does subject A, that would be one thing, and might be in line with the idea that certain voluntary transfers make the world better off. But they aren’t: they are simply assured that the anonymous person in the next room will get more MONEY.

As Landsburg points out, if you were REALLY a generous person, you could turn that uncertainty into a certainty by taking the money and donating it to a charity, a public good cause, a homeless person: ANYONE but an utterly random person who might as well already have more money than even you.

Indeed, if you agree with the Voluntary Contributions to Public Goods model, then that would itself be much better than giving money to an anonymous person.

But that’s not what people are doing here. They are acting as if someone else being given more money by someone else is worth doing (indeed, worth paying for), irregardless of the prescence of any such effects.

6

Point number one:

This is a pretty cool thread.

Point number two:

Regarding your suggestion that we, the taxpayers, be given the freedom to designate where our dollars go, I think that’s not a bad idea. We probably couldn’t allow taxpayers to designate every dollar, because taxpayers in general don’t have the foresight to know which government funded operations are essential (military, roads, etc), and which are fluff, and moreover, they don’t know how much each operation should get. I can imagine a nightmare scenario in which the public at large decides to give 90% of our money to Save the Whales, and 2% of it to our military. The whales propser, but our nation is invaded and conquered by New Guinea.

However, I’m envisioning a system where X% of our tax dollars are used to fund the essentials, and are divvied up by government employees who know what they’re doing. The remaining (100-X)% is allocated on a personal basis. (Okay, libertarian Republican that I am, I’d prefer we just cut taxes to the point where we’re only paying X% in the first place, but that’s another thread. :slight_smile: ) If you choose not to earmark your tax dollars, or earmark them towards ineligible programs (though how to define “inelligible” would be a whole 'nother tangent), then they go into a general fund, that could either be delegated by the government, or could go into a government trust fund that could be invested and used in the future, thus possibly allowing for lower taxes later on. I’d much prefer this, than watching my money go to fund programs that are ineffective, or actively against my beliefs.

Unfortunately, while I think this system could be made to work, and would be vastly superior to what we’re currently doing, there’s not a chance in hell of us ever seeing something like this. I think the little box that says “check here if you want $1 of your taxes to go to the general campaign contribution fund” is as close as we’re going to get.
Jeff

7

Well, I indeed was talking about only that portion of our taxes which is earmarked for directly redistributive purposes.

8

Landsburg’s spin on this study seems exceedingly thin. I can imagine several threats to external validity, especially Hawthorne effects (which I had to mention with a nod to hawthorne).
People will do all kinds of strange things in the presence of an experimenter. You want disturbing, check out the Milgram experiments.

More likely they think it goes nowhere, being a highly artificial environment (they may think A doesn’t really get the money, whatever the experimenter tellls him/her). They may think the money, already allocated to the experiment, will simply be wasted if it stays with the experimenters. In other words, regardless of whether it came from taxpayers or not, it is not likely to go back to the taxpayers.

Anyway, more information is needed about the actual experimental set up.

9

It is commonly supposed that the “perfect stranger” aspect of this experiment is what contributed to the attitudes towards charity in the Old South. Charity was given personally, not anonymously. You wanted to feel good about what you’d done, see a human face in the person you helped, and have that person know that he has you to thank. This retarded the development of large-scale charity operations (such as soup kitchens) in the South (though you can argue the need for such institutions was lower due to population dispersal, the attitude did exist.)

I wonder what the result would have been in this experiment had the subject been told that the other person needed the money, for whatever reason. Would the fact that the donee is a stranger prevent the donor’s generosity, even if the donor had reason (by present cultural standards) to donate?

10

I would be very surprised if the donations did not go up when the subjects are told the other person needs the money.

—Landsburg’s spin on this study seems exceedingly thin.—

Your problem would be with Vernon Smith here: he’s the one who designed and ran it. I remember reading about these before, and they did their best to work against the possibility that the subjects felt “watched” and “judged” by the experimenters (in addition to demonstrating that people gave more when they were clearly being watched, and less with the precautions in place).

—They may think the money, already allocated to the experiment, will simply be wasted if it stays with the experimenters. In other words, regardless of whether it came from taxpayers or not, it is not likely to go back to the taxpayers.—

As is pointed out, even if this is the case, it still doesn’t explain why people would pick this particular allocation of the money over any other.

People seem to be acting as if it were a chance to actually create value that they feel like they shouldn’t pass up: i.e. that its somehow more important, or does more good, to give up more of their own money in the case where the experimenter offers to double their cash. But, knowing nothing about who’s getting the cash, why would they be more inclined to give even if they thought the experimenter would otherwise pocket the money?

—Anyway, more information is needed about the actual experimental set up.—

Indeed, but then, this thread isn’t entirely premised on any particular reading of the experiment or even its validity, it’s just inspired by some of the potential implications. Vernon Smith seems to think that it is fairly conclusive: Landsburg disagrees with his interpretation, without quibbling over whether or not it’s conclusive.

11

Probably because the very offer to triple the amount sends a strong (if unconscious) signal to the subject that A probably needs the money.

I agree with hawthorne that the way this offer is made is important. In any case it would be natural for the subject to make certain assumptions about A, absent any evidence. Thrust into this role of giving money directly to another person, the subject would likely relate this to other similar situations s/he has encountered in the past: the homeless people he’s seen begging for money, the friend who lost a job and needed a quick loan. In other words the subject is likely to typecast A as being needy even without information about A due to the nature of the experiment. Then add the experimenter with his offer to triple the amount given, and I see potentially strong social pressure to act. Especially since the choice appears to be a simple dichotomy: give to A/don’t give to A. To me it seems equivalent to a leading question on a survey. (This is my impression, I will try to hunt down more info on the actual setup) From a social psych perspective I see several potential threats to validity.

To me the validity is critical if we’re to consider the implications. Absent more info on the setup, I’m not confident the experiment measures what Landburg (or Smith) thinks is being measured.

12

This experiment seems to demonstrate what you might call the particularisation of human empathy. The subjects do know something about the “other person” in the experiment - they know that they are a single individual and people are always more empathetic towards a single individual than a vast amorphous mass of people (in this case, probably “the University” or “the research team” - I doubt that most of the subjects would be looking further than that towards “the taxpayer”). The fact that the subject is told nothing about the other person just makes them a stronger target for empathy, because the subject can project any personality they like onto this stranger, and will probably think of them as a person “just like me”

Furthermore, the subject has another reason to feel empathy with the Random Cash Recipient - they are both subjects in this experiment. So, they are probably similar in circumstances. Aren’t most experimental subjects students? (unless the experimenter makes a big effort otherwise) A student being a subject in this experiment would have a very good reason to think that the RCR was a) similar to them and b) more in need of cash than the researchers.

I’d be really interested to know:

  • if there’s a difference in the outcome of the experiment depending on whether the subjects are students, or psychology researchers, or professionals, or unemployed.

  • were the subjects ever asked to describe what sort of person they thought would be getting the money? I suspect if they were, the descriptions would look pretty much like the subjects themselves

  • what was Smith actually trying to measure in his initial experiments? Landsburg doesn’t really go into this, either in the OPs link or in his earlier Reason article (which gives a few more details, albeit in a highly melodramatic style).
    I don’t know that the experiment shows anything much about redistribution as such. It does show that we aren’t very good at dealing with judgements involving large groups of people - we’re wired for one-on-one relationships. But then, I think your daily newspaper can provide you with any number of examples of the same thing. Think of the old saying “one death is a tragedy, a thousand deaths is a statistic” Or think of a situation where a new road or a train line which may benefit thousands of people runs up against the problem of one person living in the way who doesn’t want to give up their home. People’s sympathy will naturally lie with the home owner - the single individual.

13

Apos: please clarify this for me. You say this, but neither you nor the Author mentions what Vernon’s experiment DOES puport to show–if anything. WHAT WAS THE ORIGINAL POINT OF VERNON’S EXPERIMENT?

14

I think hawthorne already discussed it, toad.

—Probably because the very offer to triple the amount sends a strong (if unconscious) signal to the subject that A probably needs the money.—

I’m not sure if I buy that, and if I did, I think it would make things even MORE unsettling. The more you are told you can take makes you think that the anonymous recieving third party deserves more?

The question is: why is A more deserving than anyone else who might end up with the money? Why give it all to A instead of spreading it around to more people: the equally anonymous C,D, and E?