TOO MUCH INFORMATION
Understanding What You Don’t Want to Know
By Cass Sunstein
The announced thesis of Cass Sunstein’s “Too Much Information” is simple: People should be given information only when it would significantly improve their lives. Arguing that the welfare of citizens should be the yardstick for deciding whether or not to disclose, Sunstein sets out to re-examine everything from mandatory food labels to rules about declaring the origins of “conflict minerals” to the government’s broad authority to ask people for their data (a burden he calls “sludge”).
The book represents a change of heart. Sunstein administered the Federal Office of Information and Regulatory Affairs for most of President Obama’s first term. Then he was on the side of full disclosure, including forcing restaurants and movie theaters to let customers know exactly how many calories were in their food. That achievement prompted an email from a friend: “CASS RUINED POPCORN.” “Too Much Information” is an attempt to take that offhand remark seriously; if your measure is whether information helps or hurts people’s lives, much current disclosure, including mandatory calorie counts in movie theaters, starts to look more complicated.
Sunstein launches his argument using data from a small survey he conducted. Among his respondents, preferences for more information varied widely: Knowing the date of their death appealed to a little over a quarter of those surveyed, knowing whether their partners were cheating appealed to over half, knowing if there is life on other planets appealed to nearly three-quarters. There was also great variation in reported willingness to pay for that information, with median bids ranging from $1 for credit card late-fee disclosure to $200 to know if heaven exists.
Willingness to pay is a key metric, allowing the balancing of otherwise hard-to-compare priorities. Learning that your favorite food is bad for you will upset you, but might prolong your life; present mood versus future utility has to be considered somehow. Unfortunately, Sunstein’s own data on willingness to pay is more evocative than convincing. This is a bit stats-geeky, but the median answers versus average answers from his survey vary wildly, a variation that does not show up in nationally representative data later in the book. The lack of consensus among his respondents makes Sunstein’s survey watery concrete for the foundation he is trying to pour.
This is a distraction but not a fatal one; the book actually delivers something stranger and more interesting than the announced thesis: a tour of human biases that end up creating “behavioral market failures.”
These are things like our failure to notice missing information: A movie released with no reviews is probably terrible. Or how about blithely accepting advice from salespeople without considering that they will benefit from our spending more? The “telltale heart effect” leads businesses to clean up their various acts, even if customers have not changed their behavior. (Some restaurant hygiene labels seem to work this way.) The “halo effect” means that seeing kale salad on the menu can make people feel better about the chili cheese dog they actually order. And so on. The range and direction of these effects makes it impossible to conclude that behavioral market failures point in one particular direction.
While Sunstein’s analysis of consumer psychology is illuminating, his examples are sometimes oversimplified. He rightly mocks the F.D.A.’s food pyramids and plates, without noting that their uninformative design — lentils and bacon are both “proteins” — is a result of agricultural lobbying. (Regulatory capture, the risk that agencies may feel more beholden to the industries they regulate than they do to the public interest, is never mentioned, curious for a book by a regulator.) Similarly, he speculates about the effects of disclosing the dangers of smoking to addicts, without noting that recent victories over smoking came from revealing the effects on bystanders, not smokers. And there is no mention of organizations like ProPublica or The Markup (or The Times, for that matter) that rely on public data to find hidden stories.
To want Sunstein to account for regulatory capture, secondary harms and data journalism, though, is to want more of what he is already offering. Though it presents itself as a new solution to a host of current problems, “Too Much Information” ends up presenting a host of new problems to one current solution: transparency. Among government reformers and progressive regulators (like Sunstein himself, a decade ago), increasing access to information has been regarded as an obvious goal since Watergate. The book doesn’t replace that generational certainty with a new one, but it does make it impossible to continue regarding information disclosure as an uncomplicated good.