Do Economists Wear Seatbelts?

Stubborn beliefs in the economist's research market

Posted Mar 02, 2009

I once shared a cab to LaGuardia with a freshly minted Economics Ph.D. On the scent of fame, he identified himself as part of a new breed of "rock-and-roll economist," prepared to draw an audience with findings that jar and awe.

In the art of rhetoric, there are two ways to capture an audience: Tell people their worst fears are overblown, or tell them that a burden can be guiltlessly shed. Don't worry. Be happy. This cynical pose has led distinguished scholars in other fields, like Yale psychologist Paul Bloom, to decry the morality of the modern economist: "The problem is not that economists are unreasonable people, it's that they're evil people... They work in a different moral universe."

The economist's typical response strikes many as obtuse, and robotically affect-free. It is not the economist's job to seek to improve human welfare, or even seek the truth. After all, research is a market, no more and no less. Whether the incentive is fame or money, the market will sort out its quality. That's the official view, anyway. In practice, it's pretty hard to miss the ideology. But if you want to pretend that the economics research market filters out ideology, you shouldn't be surprised when an economist from the Ivy League tells us that U.S. citizens suffer "climate change hysteria", worrying pointlessly about global warming:

"[F]or the average resident of developed, industrialized countries, a warmer climate will bring net health benefits rather than any significant health costs."

Great. Sit back with a mint julep or buy that new tennis racquet (you'll have no need for skis). Global warming is coming to the U.S.! But there's more:

"In the U.S., a warmer climate will likely not only bring health benefits, but also quite sizeable recreational benefits. Early studies of the impact of climate warming in the 2.5 degree centigrade range focused on skiing and unsurprisingly found that a warmer climate would mean a potentially large decrease in ski days and a correspondingly large welfare loss. But skiing is of relative economic insignificance compared to summertime recreational activities such as boating, camping, fishing, golfing, hunting and wildlife viewing, with only $2.5 bill spend annually on skiing compared to $76 billion on the summertime activities. With either a modest 2.5 centigrade increase, or even larger 5 degree centigrade increase in temperature, recent economic work estimates very large net recreational benefits from global warming in the U.S., with net benefits perhaps reaching over $25 billion under the 5-degree increase scenario ." [footnotes omitted]

You will look in vain for cues of satire in this passage. This is a serious contention, complete with the blind trust that a destabilized climate will be gracious enough to allow fiscal year projections for recreational suppliers and real estate developers. The article talks as though national borders magically block global warming, or at least that the average industrialized nation can safely ignore the impact of global warming on less developed nations ("with moderate climate change predominantly benefitting, rather than harming, the U.S."). As it is, this is "A Modest Proposal" without the irony. Or, it could be a pretty effective Carnival Cruise ad.

Could modern economics also show I've been foolish for spending money on people worse off? Malthus, of course, famously argued that it was inefficient to give Irish workers better-than-subsistence wages - they would just squander the increase on the expense of more children - offsetting their gains with greater consumption. So you see, this coerces employers into subsidizing private choices. The coercion is even worse if it is governmentally mandated by a minimum wage.

Malthus's is an argument from "offsetting behavior". It assumes that, by some mysterious alchemy, there is a "natural amount" of risk that we will seek and absorb no matter what the governments does. It is an empirical speculation about human behavior, and there is only occasional evidence for it in psychology. It is, however, the canon in contemporary economics.

The sacred text in this canon is the classic 1975 study by economist Sam Peltzman that reported that automobile safety laws - like seatbelt laws - make people feel safer. Feeling safer, people believe they can "afford" to drive more recklessly (the so-called "Peltzman effect"). But by 2001, the celebrated Peltzman effect had been unmasked as not quite an urban legend, but surely a grossly overstated instance of offsetting behavior. The first hit came from Steven Levitt and Jack Porter, who showed that whatever small increase in recklessness seatbelts introduced was swamped by their power to restrain drivers and occupants. Wearing a seatbelt reduces the likelihood of death by 60% and saved as many as 15,000 lives in 1997 alone. The second blow came from Alma Cohen and Liran Einav in 2003, who found that seat-belted drivers are not more reckless. And these analyses are still unrebutted. Mandatory seatbelt laws reduce driver and occupant deaths, with no increases in pedestrian deaths. But like the Darwin awards, the Peltzman effect is too deliciously ironic or tantalizing to keep to yourself in casual conversation. (For more on the resilience of belief in the Peltzman effect specifically, and in offsetting behavior more generally, see The Empathy Gap.)

Did the research market in economics respond to these novel findings? No. Instead, economists largely ignored these findings, while continuing to ceremonially cite the Peltzman effect. The finding that seat belts do not produce compensating risky behavior might cause the finding to be buried, and professional ties severed. This would seem to be a big crack in its façade of objectivity. In the 2007 edition of his widely used textbook, Greg Mankiw, Chair of the Council of Economic Advisors under George W. Bush, fails to report the findings that disconfirm the canon, "[T]he net result is little change in the number of driver deaths and an increase in the number of pedestrian deaths." (2007, p.8)

Mankiw describes the "Peltzman effect" as evidence for the "general principle that people respond to incentives". But he is too modest. Of course we respond to incentives; that's why a sale at Walmart can produce a stampede. The general assumption of offsetting behavior ventures far more - that government regulation is futile.

But in an efficient research market, the existence of offsetting behavior should be open to empirical test. And so far, the evidence is uneven. We can't assume that behavior will always fully erase the effects of a subsidy. To the extent that it is assumed, it is a beard for conservative economic policy that the economists' "research market" has failed to efficiently sort out. The sheer persistence in the academy of the Peltzman effect myth, for years beyond its disconfirmation, could only show that ideology is alive and well in economics.

Any theory that treats global warming as a growth opportunity, and government social programs as futile on the grounds of offsetting behavior is, shall we say, in the grips of an ideology. Either an unacknowledged value system is driving the economists' judgments of acceptable risk - in which case their inquiry is not value-neutral, or a false assumption of offsetting behavior has been insulated from refutation, in which case the research market is not free. There is nothing sinister going on here; these are all honest oversights of economists. But many economists do not choose the simplistic and mechanical psychology that lends comfort to these conclusions. In fact, Daniel Kahneman, Richard Layard, and Robert Frank champion a psychologically informed economics, one in which people routinely honor sunk costs, take losses to gain unproductive vengeance, and make inefficient decisions to vent moral outrage. Of course, economics is not the only science driven by ideology. But is it the only one that, on the whole, disavows with verve the relevance of value judgments.