Discussions about whether government should interfere with the workings of an otherwise free market tend to take on a dogmatic flavor; frequently, assertion meets counter-assertion with nary an empirical fact in sight. In part, this is because such discussions are forward-looking, and the future is by definition data-free at any present moment. To make matters worse, historical experience is often of limited usefulness in macroeconomic matters, either because there are simply too many economic variables to account for (and no way to control for irrelevant factors), or because some party to the discussion claims that “this time is different.”
Nonetheless, there are times when the balance of empirical evidence is so overwhelming that some conclusions are inescapable. For example, anyone who claims that government can fix the prices of goods and services without courting calamity is rightly dismissed as ill-informed (unless he’s talking about the price of money, in which case he’s appointed to the Fed).
As a further contribution to the list of Topics on Which We Have Very One-Sided Evidence, I offer this link to an interesting study by Clifford Winston called “Government Failure versus Market Failure.” Winston, who was interviewed a while back on Russ Roberts’s EconTalk podcast, surveyed empirical literature from the last forty years or so and found a surprising degree of consensus that the social cost of failed government policies at the federal level exceeds the social cost of the market failures the government interventions were intended to address. This tended to be so either because the original case for government intervention was flawed (i.e., there was no market failure after all), or because there was a market failure but government addressed it in a way that was clearly inferior to some other solution. (For those who think they hear an axe grinding somewhere, I should perhaps say that Winston’s study was sponsored by the AEI-Brookings Joint Center for Regulatory Studies, and Brookings at least is not normally considered a hotbed of laissez-faire liberalism.)
Winston’s study is too voluminous and detailed to summarize very well, particularly because he tries admirably to distinguish the kinds of government interventions that never work (farm price supports, for example) from those that seem to be justified by a cost-benefit analysis. But here are some results I found particularly interesting:
- “Antitrust policies toward monopolization, mergers, and collusion have done little to raise consumer welfare, while economic regulation of agricultural products and international trade has produced large deadweight losses in the process of transferring resources from consumers to producers.” The only major antitrust case Winston ranks as a “possible exception” to his generally negative opinion of antitrust effectiveness is the 1984 breakup of AT&T.
- “Generally, [farm] subsidies mainly go to Big Agribusiness corporations and the richest farmers. . . . [I]n 2003 the top 6 percent of recipients collected 55 percent of all subsidies. Although subsidies used to be inversely related to farm prices (that is, they would increase when farm prices fell), that relationship has not been true in recent years. For example, according to the U.S. Department of Agriculture, farm earnings in 2004 reached a record $74 billion, while direct government payments were as high as they were in years when farmers earned much less.” Moreover, while farm subsidies were originally justified in significant part by the fact that farmers had lower and less stable incomes than non-farmers, that began to change in the 1960s, and for the last thirty years farmers have been, on average, wealthier than non-farmers. Researchers in 1989 and 1992 estimated the welfare loss from these policies at between $3 billion and $12.4 billion per year. In 2003, President Bush signed a ten-year, $190 billion extension of the program.
- Winston likewise found “a large volume of empirical evidence . . . that trade protection has mainly generated gains to established U.S. industries that fall far short of the losses to consumers.” No surprise there, really.
- FTC advertising regulation appears to be largely unnecessary because consumers for the most part don’t fall for the advertising in the first place. One pair of researchers found that in every decade from the 1930s to the 1980s, “70 percent of consumers thought that advertising was often untruthful and sought to persuade people to buy things they did not want. [Editorial comment: Only 70%? I hope the other 30% never operate heavy machinery.] . . . The authors argued that the stability of consumers’ beliefs about advertising through time — especially during the 1970s when advertising regulation moved from extreme laxity to unprecedented force and the 19080s when regulation receded — was inconsistent with the view that advertising regulation increased the credibility of advertising.”
- “Until the mid-1980s, manufacturers were prohibited by law from promoting the health content of their food products through advertising. When the prohibition was lifted, the consumption of fiber cereals increased and the consumption of fat and saturated fat decreased.”
- “With the single exception of BART in the San Francisco Bay area, every U.S. transit system actually reduced social welfare.” (I’d love to know what’s different about BART, but Winston doesn’t say.)
Winston also includes some great examples of the “unintended consequences” problem. For example, he notes that researchers have found that Corporate Average Fuel Efficiency or “CAFE” standards for cars “compromised safety by distorting the mix of large and small vehicles.” In other words,
“CAFE caused consumers to shift to light trucks (vans, minivans, and SUVs) instead of small cars and estimated that 50 percent of the increase in the share of light trucks since the 1970s could be attributed to CAFE, thereby offsetting 75 percent of the vehicle weight that would have been lost otherwise. The shift to light trucks increased fuel consumption, and although light trucks increase their occupants’ safety, they increase injury severity from a collision to occupants of smaller vehicles and to pedestrians, bicyclists, and motorcyclists.”
So what works? Government programs that use market-like incentives to address unwanted externalities, like regulations addressing air pollution and aircraft noise. In these situations, “[g]overnment can increase efficiency by using pricing or quantity policies to make consumers and firms account for the social costs (or enable them to accrue the social benefits) of their actions.” There is a trap here, to be sure: Sometimes regulatory programs are sold in market-like lingo but they do not actually impose the discipline of the market — school voucher programs that do not actually allow bad schools to close would be an example. Still, it would be a great thing if those who peddle top-down, centralized solutions to complex social problems had to at least pretend to understand the incentives affecting individual behavior at the micro level.
Amidst all these particulars, at least one general conclusion seems warranted: “[T]he existence of government failure suggests the absence of an incentive to reconcile an intervention’s costs and benefits to policymakers with its social costs and benefits. In contrast, it appears that in at least some instances market participants have greater incentives to correct market failures than the government has to correct these failures.” But of course reasonable minds already knew that.