What do the immigration debate and the inequality debate have in common? Both seem to turn on zero-sum versus positive-sum framing of the issues.
Much opposition to immigration derives from the belief – correct or not – that immigrants create a burden on the welfare state. Since the dollars spent on welfare (including public health, education, and so forth) can’t simultaneously be spent on other projects (or left in the taxpayers’ pockets), the zero-sum framing is justified: the immigrants’ gain is our loss. But even if the immigrants-on-welfare argument is put to rest, either by laws restricting immigrants’ reliance on welfare or evidence demonstrating the problem is small, some people will still oppose immigration because of the perception that immigrants take American jobs. Here, the zero-sum framing is mistaken, relying on the bogus belief that the number of jobs is fixed, meaning the immigrants' job gain is our job loss. But given that belief, opposition to immigration follows naturally enough.
The debate over inequality follows a similar pattern. Why should we be concerned at all about inequality? The presumption is that gains to the rich necessarily imply losses to the poor – another example of zero-sum framing. The presumption is wrong, because it fails to consider the possibility that everyone might gain in absolute terms. It’s true, for instance, that the share of income received by every income quintile except the top one has decreased over time. But it’s also true that the average real income of every quintile has increased at the very same time. It’s just not true that the rich have gotten richer and the poor have gotten poorer; instead, the rich have gotten a lot richer and the poor have gotten a little richer. It’s a positive-sum game. Once this fact is understood, some opposition to inequality should dissolve. But the debate isn’t over, because the opponents of inequality fall back on a different argument: that the relative success of the rich makes the rich feel smug and the poor feel envious. Even if the absolute material well-being of the poor has improved, their relative well-being has declined. By means of this argument, the inequality debate is framed once again as a zero-sum game: if we care intensely about relative status, gains to the rich are ipso facto losses to the poor.
It should go without saying that I’m simplifying greatly. Probably the most important factor I’ve glossed over is that positive-sum is not synonymous with beneficial to everyone involved. Immigration may (I believe does) create gains that exceed the losses to all parties, making it a positive-sum activity; and in the long run, I think almost everyone shares in the gains (through, among other things, lower-priced goods and services). But almost doesn’t mean universal, and some people assuredly lose. They would have reason to oppose immigration even without a zero-sum perspective. Similar, the kind of economic system that generates high inequality along with widespread growth in real incomes doesn’t help absolutely everyone. There are certainly many short-run losers and some long-run losers as well. They, too, might have reason to oppose the system even without thinking in zero-sum terms. Still, I think it’s interesting how far the zero-versus-positive-sum distinction goes in explaining policy debates.
Saturday, September 09, 2006
Subscribe to:
Post Comments (Atom)
7 comments:
But immigration and inequality are highly disanalogous if we ask "what's the alternative?" Anti-immigration folks generally want to stop (or drastically reduce) immigration, which would shut down at least a large portion of the positive sum game. Anti-inequality folks generally think that it is possible to continue to grow at a similar (perhaps slightly slower) rate but to have a substantially larger portion of the growth go to benefit the poor (and the middle class), and for a substantially lower portion of the growth to go to the richest few percent. They want to keep playing a positive sum game, but to rejigger the rules a little bit so that the gains are more evenly distributed.
Blar -- Your disanalogy relies on framing immigration as a "big change versus no change" proposition, while framing inequality as a "small changes at the margin" proposition. But in fact, in both areas there is a wide spectrum of possibilities. Someone could support a mild reduction in immigration, for instance, rather than stopping or drastically reducing it.
In any case, I think your argument about inequality bolsters my point. You describe a sophisticated anti-inequality person who recognizes the positive-sum aspect of the game, but who nevertheless focuses on the zero-sum aspect of the game -- how we can change the rules to improve the position of some (the poor) while diminishing the position of others (the rich). And he's even willing to reduce the positive-sum character of the game (by accepting a slightly lower rate of growth) in order to alter the distribution of gains.
This doesn't prove the anti-inequality person is wrong. But it certainly supports my claim that zero-sum versus positive-sum framing explains a lot of the debate, does it not?
But the person who wants a mild reduction in immigration is still opting out of the positive sum game a little bit, rather than wanting to influence how the benefits of the game get distributed. My point is that, unlike these anti-immigration folks, anti-inequality people often are not ignoring the positive sum game.
To a large extent, anti-inequality sentiment is based on concerns that are unrelated to zero-sum or positive-sum thinking. Much of society (especially the poor) haven't been benefiting very much from our growth (or "haven't been benefiting as much as they could have" or "haven't been benefiting as much as similar people in other countries"), which is a problem. Realistically, dealing with this problem probably means reducing the enormous growth rate for the richest sliver of society, and maybe even slightly reducing the overall growth rate, rather than somehow producing a bunch of additional growth.
One could even make the argument that the anti-inequality people are focusing on a positive sum game while anti-anti-inequality folks are stuck with zero-sum thinking. You're thinking of reducing inequality as a zero-sum redistribution, while they're recognizing that it increases utility (because of marginal diminution; though obviously there are winners and losers).
blair: "Much of society (especially the poor) haven't been benefiting very much from our growth"
I hear this all the time, but I've seen no evidence to support this assertion. How do we determine that the poor from 1996 have not benefitted from the gains of the past decade? Wouldn't we need to follow the economic status of the same individuals across time?
How do we know that even those now in the bottom decile haven't gained significantly? Every low-skilled Latin American immigrant - legal or otherwise - realized a huge wage increase from the U.S. economy the moment he started working here.
I don't know, Blair. I'm reluctant to accept your assertion without seeing some real data on mobility.
But the person who wants a mild reduction in immigration is still opting out of the positive sum game a little bit, rather than wanting to influence how the benefits of the game get distributed.
They are trying to influence how the benefits of the game are distributed. They want to make sure native-born Americans get a larger share of the wages, jobs, etc. By doing so, they also reduce the total amounts of such benefits available.
To a large extent, anti-inequality sentiment is based on concerns that are unrelated to zero-sum or positive-sum thinking. Much of society (especially the poor) haven't been benefiting very much from our growth (or "haven't been benefiting as much as they could have" or "haven't been benefiting as much as similar people in other countries"), which is a problem.
But that is zero-sum thinking -- or, to be more precise, constant-sum thinking. (Zero-sum is a special case of constant-sum; any constant-sum game can be normalized to an equivalent zero-sum game.) The anti-inequality folks want to redistribute gains from one group to another; one group's increased share of the gains is offset by the other group's decreased share. Even if they are aware of the positive-sum aspect of the game -- the aspect that allowed these gains to exist at all -- they are focused on the constant-sum aspect of the game when they demand redistribution.
One could even make the argument that the anti-inequality people are focusing on a positive sum game while anti-anti-inequality folks are stuck with zero-sum thinking. You're thinking of reducing inequality as a zero-sum redistribution, while they're recognizing that it increases utility (because of marginal diminution; though obviously there are winners and losers).
Okay, now you're going to force me to get technical. In game theory, if a game is to be defined as zero-sum or positive-sum, the payoffs are generally stated in terms of money, resources, or other measurable external quantities -- not in terms of utility. For instance, chess is often given as the quintessential example of a zero-sum game, because for one player to win, the other must lose. But if we put payoffs in utility terms and assume utility is interpersonally comparable, then it's entirely possible that player A might get much greater enjoyment from winning than would player B. So we'd have to say chess is a positive-sum game! But if we did that, then zero-sum would become essentially a useless concept; there would be nothing we could definitively say is zero-sum. So I think the traditional approach is more helpful, because it gets at the notion of a situation in which one person's gain is necessarily another person's loss in objective terms.
Robert Frank shows many situations where inequality does turn into a zero sum game. His paper on positional externalities is particularly good (both the one in the Zeckhauser volume and a later one on the web called "Are Positional Externalities Different").
If people are bidding for a fixed resource, those at the bottom are hurt when the buying power of those at the top increases more rapidly than their own. I know this may seem like a relatively rare situation. But just consider the ways in which SUV's reallocate road safety, or the recent furor over Golden's book on college admissions, and you'll see the point.
Or just imagine a city where one bedroom apartments are all bought up by very wealthy people and converted to mansions.
PS; for an elaboration on the SUV point, see
http://madisonian.net/archives/2006/09/15/suvs-as-a-technology-of-regressive-redistribution-of-safety/
Frank -- I see Robert Frank's work as another example of the process described in the original post: finding a way to re-frame the debate as zero-sum by focusing on relative rather than absolute well-being. The only difference is that Frank's model relies on rivalry between the more-rich and less-rich, rather than between the rich and poor.
For responses to Frank, see Klein and Kashdan's article in the most recent Econ Journal Watch. For a shorter take-down, see Will Wilkinson's on the subject.
Post a Comment