The Perils of Seeking Perfect Information:
How many roadway signs is enough?
Is there a public participation process that can help find out?
24 Jul 2007 in Regulatory Economics, Regulatory Policy
A conventional example of market failure often used to justify regulation is the absence of perfect information. This is explicitly set forth in the Office of Management and Budget's guidance on the conduct of Regulatory Impact Analysis.
While it is true that the strictest form of the economic theory that is the foundation for benefit-cost depends includes this assumption, it is made for expository convenience rather than necessity. Information need not be perfect for economics to explain or predict behavior. Indeed, it works fine when knowledge is limited; the pursuit of knowledge is one of the instruments through which markets improve general welfare, and for some people gaining knowledge is the very activity that brings them satisfaction.
So how much knowledge is enough -- or, in the language of economics, "optimal"? Assuming that it's true that ignorance generally is bad and knowledge is good, is there an amount of knowledge that the government can provide to "solve" market failure, beyond which people start to get worse off? Today's A-hed in the Wall Street Journal gives an example showing that there is an optimum.
Mike Esterl reports (available temporarily without a subscription) that Germany has too many highway signs. Even taking into account the historic German passion for order, too much information can undermine the purpose for which information is provided -- in this case, traffic safety:
The sign should be taken down because there's no room to pass anyway, said the lawyer. It could use a cleaning, said a civil servant. "It's a nice day," said the policeman, keeping the peace as he moved the van a few yards down the road to inspect the next sign.
For nearly a decade, Germany's 15 million-member ADAC automobile association has been curb-crawling the nation's streets with municipal officials in an effort to persuade them to get rid of as much as half the country's estimated 20 million traffic signs. Many Germans believe the country's signage has become so dense that it's a safety hazard. A recent study concluded that the distracting signs keep drivers from watching the road.
Germans have coined a term for the phenomenon -- Schilderwald, or sign forest. But as the van-load of officials touring Troisdorf for surplus signs discovered on a recent morning, parting with them isn't proving easy.
"Germans like clear rules," said Joachim Adam, one of the three civil servants in the van.
There's a sign for toad crossings (Kröten-Wanderung!). There's one that tells drivers when they're on a "dirty road" (Verschmutzte Fahrbahn). Another warns them not to drive into lakes or rivers. There are 32 different signs regulating how to park at a curb. There's even a special parking sign for hikers where they leave their cars.
A recent ADAC-funded survey in nine European countries found that 45% of respondents thought their roads had too many signs, compared with 75% of German drivers.
Esterl quotes a spokesman for the American Automobile Association saying that this is not a problem in the US, but Andrew Delmege's dismissal of the issue does not appear to be founded on data. The phenomenon observed in Germany ("There are people sitting in government offices who simply like to put up signs") is surely not unique. The Washington Post's Eric M. Weiss says that after the $676 million Springfield, Virginia, "mixing bowl" project was completed there continue to be demands for more (and presumably better) signage. Early evidence is that the problem local drivers have with the new design is that they aren't paying attention -- a problem more signage cannot solve.)
Highway signage serves multiple purposes, such as improving safety, providing routing direction, and announce legal rules. Too many signs can clutter the visual field so that each message is diluted and drivers are more likely to miss (or misinterpret) the important ones. There is an optimal amount of highway information; it can be difficult to know how much is too much, though Esterl's story suggests a reasonable way to involve the public in reducing signage clutter. Signs proposed for removal are covered with a yellow bag; final decisions can be made based on the nature and scope of public reaction.
Public participation in regulatory decision-making seems to be wildly popular, even though the evidence for its effectiveness seems altogether lacking. There is an ad hoc committee of the National Research Council charged with examining this evidence in the case of environmental assessment and policymaking. One complication facing this committee is that many members are individuals whose prior work or research suggests they are committed to such processes irrespective of whether they improve environmental outcomes -- that is, public participation succeeds when the public has an opportunity to participate. This is a tautological statement, not a scientific inference.
The German highway sign example may be instructive as a successful public participation program. The small group that drove the highways included representatives of the dominant interest groups but not "the public." The group reached a consensus to remove 20% of the highway signs they inspected. By bagging the signs designated for removal, they also created a de facto appeal mechanism whereby the public (which hardly ever is genuinely represented in so-called public participation schemes) actually had an opportunity to object to proposed removal decisions. (Of course, the public had no way to appeal proposed non-removal decisions, and there is no evidence from Esterl's article that the public was asked to contribute suggestions to the hit list.)
Whether or not by design, the German public participation process on highway signage was biased in favor of a policy outcome of sign retention. The group consisted of one representative of the faction desiring to reduce signage, three civil servants representing the status quo, and a policeman to keep order. That they were able to agree on removing 20% of the signs they inspected despite this imbalance in representation strongly suggests that a balanced group would have recommended a much high removal rate. That does not make the German sign review program a failure; it means that its prospects for success were limited at the outset, but that it appears to have been quite successful given this design constraint.
