The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Recent events such as the unexpected rise of Donald Trump and the growth of partisan hatred have led many people to start taking the problem of political ignorance and bias more seriously than before. Two important new books offer insightful diagnoses and potential solutions for these dangers: Tom Nichols' The Death of Expertise: The Campaign Against Established Knowledge and Why it Matters, and Cass Sunstein's #Republic: Divided Democracy in the Age of Social Media.
Both works make important points and offer valuable insights, particularly when it comes to the role of the internet and social media in our political environment. Each is essential reading for anyone interested in this pressing subject. But both also have some drawbacks. In particular, both overlook the extent to which the ultimate cause of the problems they focus is the political process itself, and the terrible incentives it creates. Ignorance and bias may be less of a menace if we made fewer important decisions at the ballot box.
I. Nichols' Death of Expertise.
Nichols' book addresses an interesting paradox: education levels are higher than ever before, and—thanks to the internet and other technologies—expert insights on almost any topic are easier and cheaper to find than at any time in history. Yet public hostility to expertise is extraordinarily high, and political ignorance is widespread. It played a major role in the 2016 election, among other races. Nichols recognizes that skepticism of experts and exaltation of the supposed wisdom of the common man is hardly new. Nineteenth century populists decried elitist experts in many of the same ways as those of today. Likewise, today's political ignorance is actually quite similar to that which survey researchers have found over many decades.
Still, Nichols contends that the present situation is different because "the issue is not indifference to established knowledge; it's the emergence of positive hostility to established knowledge." His book is an excellent discussion of the ways in which modern culture and technology exacerbate this problem. For example, while the internet makes expert knowledge and accurate information more accessible than ever before, it also makes it easy to spread conspiracy theories, "fake news," and a variety of crude deceptions and misrepresentations. More subtly, the ability to "google" almost any topic often gives people a sense of greater knowledge and understanding than they actually possess, and a greater willingness to ignore genuine expertise. In a sense, everyone can now feel like an expert, even if their supposed knowledge is shallow and superficial. The sheer volume of information on the internet also makes things difficult. As Nichols puts it, "the countless dumpsters of nonsense parked on the internet" make it hard to separate useful information from dreck. The less knowledge you have to begin with, the more difficult it is to separate the wheat from the chaff.
These problems, Nichols contends, are exacerbated by a variety of cognitive errors, including confirmation bias (the tendency to overvalue information that confirms one's preexisting views, while ignoring that which cuts against it), herd behavior, and biases caused by identity politics. In an interesting chapter on higher education, he contends that rising college attendance, which beneficial in some ways, has also exacerbated the problem by treating students as "consumers" entitled to deference to their opinions, even on issues where they lack insight and knowledge. Nichols also indicts the campus culture of political correctness and "safe spaces," which he—like many other critics—rightly regards as an obstacle to true learning and expertise.
The problems Nichols identifies are not limited to any one side of the political spectrum. Both right and left ignore or deny expert knowledge that cuts against their preexisting views: the right, for example, on global warming (which experts overwhelmingly believe is a genuine problem) and the left on GMO foods (which a scientific consensus regards as no more dangerous than "organic" foods).
Nichols does not claim that experts are infallible or immune from criticism. He devotes an entire chapter to documenting their flaws. He does, however, powerfully argue that, within their fields, experts are far more likely to be correct than laypeople. Ignoring expertise or assuming that everyone's opinion on complicated issues is equally valid is a recipe for catastrophic mistakes.
Nichols makes many good points. But he oversimplifies by implicitly assuming that the problem of ignorance and contempt for expertise is largely uniform across issues. In reality, most of the really egregious examples he cites relate to political issues, not decisions made in the private sector and civil society. Many of the same voters who are ignorant of basic facts about government policy and angrily deny that expertise on such questions matters, also routinely go to doctors to cure their illnesses, hire mechanics to fix their cars, and follow expert advice on many other aspects of their lives. The same people who fall prey to political hucksters and demagogues who skillfully use the internet to their advantage, have little difficulty figuring out that you get cheaper and better products at, say, Amazon than on some shady fly-by-night site.
People do sometimes fall prey to ignorance and quackery in their private decision-making. Nichols notes the example of the growing anti-vaccination movement, popularized by celebrities with little or no medical expertise. He also discusses opposition to GMO foods in the face of scientific evidence to the contrary.
These are genuine problems. But the vast majority of parents still defer to medical expertise on vaccination for their children, and most Americans continue to happily consume GMO foods. In fairness, vaccination is a rare case where ignorance or irrationality by even a small minority can have a major negative effect on society as a whole. But that does not change the fact that most parents make relatively good decisions about it. By contrast, there are many examples of political misinformation spreading to the majority of the population. For example, majorities believe the bogus claim (popularized by Donald Trump, among others) that immigration increases crime, and President Obama's deception to the effect that, under Obamacare, "if you like your health care plan, you can keep it."
The college campus excesses Nichols rightly decries are another illustration of the same pattern. PC abuses are far more common in humanities and social science fields closely related to politics than in the hard sciences and other subjects more distant from it.
This divergence between political and private sector behavior is no accident. When we act as voters, we have little incentive to either acquire information or evaluate it in an unbiased way. Because the chance that your vote will affect the outcome of an election is infinitesimally small, most voters are "rationally ignorant": they devote little or no effort to acquiring information. They also generally make little effort to curb their biases or sift expert opinion from quackery and claptrap. Instead of acting as truth-seekers, they often behave as "political fans" cheering on their preferred candidate or party. By contrast, most people devote greater time and effort to deciding what car or smartphone to buy than to deciding who to support in a presidential election—or any election. They take the smartphone decision more seriously because it is far more likely to make a real difference.
Nichols briefly notes that political bias exacerbates some of the dangerous tendencies he highlights. But he largely fails to consider the key differences between political information and other types.
This problem also reduces the utility of his recommendations for reform. Most of these fall into the category of moral exhortation directed at voters, experts, and the media. For example, Nichols urges the public to "be humbler, be ecumenical, be less cynical, and be a lot more discriminating" in assessing expert statements and what they see in the news media.
This is excellent advice! Voters should indeed be more realistic about their own limitations, make more effort to consider opposing views, and be more discriminating about which information sources to trust. I have made similar suggestions myself. Voters can also make good use of expert knowledge without blindly deferring to it.
Nichols' recommendations to the experts are also well-taken: among other things, they should be more careful about the limits of their knowledge and more should be done to hold them accountable for their errors.
Unfortunately, in a world of rational ignorance, it is unlikely that many voters will act on such recommendations. Carefully assessing experts' claims and considering opposing views requires time, effort, and self-control that most people are unwilling to devote to the task. The potential collective benefit of a smarter, more discriminating electorate is enormous. But there is very little incentive for individual voters to improve their performance.
And as long as most voters are rationally ignorant and prone to "political fan" behavior, many real or imagined experts and pundits will have strong incentives to increase their ratings and income by pandering to popular prejudices rather than trying to correct them. Changing this unfortunate state of affairs is likely to require institutional reform, not just moral suasion.
II. Sunstein's #Republic.
Cass Sunstein is one of the nation's leading legal scholars, and also a major contributor to the literature on behavior economics. In #Republic, he zeroes in one important aspect of the broader problem that is also the focus of Nichols' book: the role of online media—especially social media—and their impact on political discourse. Sunstein recognizes that websites such as Twitter and Facebook have made it easier for people all over the world to make connections with each other, and have also increased access to information. But he argues they have major negative effects, as well.
Social media makes possible an impressive personalization of information streams, a kind of "daily me," as Sunstein puts it. That increases the likelihood that we will get information that genuinely interests us. But it also makes it more likely that we will only get information sources that reinforce our preexisting views. Liberal Democrats' Facebook and Twitter feeds are overwhelmingly left-wing, and conservative Republicans' the opposite. The degree of personalization that exists now may well be increased in the future, as technology advances. That might make people even more biased in their evaluation of political issues and even more close-minded than many are now.
As Sunstein explains in this book and in previous work, ideologically homogeneous groups tend to create feedback loops that make them become even more biased and even more extreme than they started out. This dynamic operates online as much or perhaps even more as in groups that meet in "real" space.
The problem is exacerbated by growing partisan and ideological polarization, which makes people even more unwilling to consider opposing views than previously—a pernicious pattern Sunstein himself warns about in the book and in previous work. In extreme cases, this phenomenon even leads to violence and terrorism. As Sunstein describes, self-reinforcing radicalization through online feedback loops is a recruiting technique effectively used by ISIS and other terrorist groups.
Although Sunstein is associated with the political left, and a former high-ranking Obama administration official, he—like Nichols—recognizes that the dangerous developments he identifies are not limited to any one side of the political spectrum. Both left and right-wing partisans demonstrate intolerance and close-mindedness online, and both tend to focus on information that confirms their preexisting views, while discounting anything that cuts the other way. Both sides are also susceptible to a variety of other political pathologies that the flaws of social media contribute to. The book is an excellent overview of these problems.
To a greater extent than Nichols, Sunstein recognizes that the problems he identifies probably require institutional solutions, not just moral suasion directed at individuals. In some of his other writings, Sunstein advocates delegating extensive authority to experts, so as to offset the negative effects of public ignorance. In this one, Sunstein's proposed remedies are much more modest.
For the most part, he advocates only very limited government intervention, and instead promotes voluntary reforms by architects of websites and social media. For example, he urges them to create mechanisms by which users can more readily access opposing views, such as a "button" that enables users to access a conservative article on the same issue as the liberal one they are reading (or vice versa). Political websites could carry links to similar sites with different orientations, and so on. He does, however, advocate subsidies for more balanced media, and possibly also requiring broadcasters (though not websites) to carry at least some more balanced programming about public affairs.
These proposals share some of the same weaknesses as Nichols' moral exhortations: they are unlikely to have much impact on rationally ignorant voters and biased "political fans." The former do not wan to pay more than minimal attention to political information, and more opportunities to see opposing views are unlikely to change their minds. The latter are likely to continue to prefer to reinforce their preexisting views rather than seek out new ones. Indeed, there are already a good many blogs, websites, and even broadcasters (like C-SPAN) that often link to opposing views, carry a wide range of positions, and so on. They have made only a modest dent in the dynamics Sunstein fears. Such efforts are better than allowing total online polarization to reign. But they seem unlikely to have more than a modest impact.
Throughout #Republic, Sunstein argues that we should seek ways to get people to act like "citizens" rather than mere "consumers." The latter merely want to satisfy their existing preferences, including the preference for information that conforms to their preexisting viewpoints. By contrast, citizens care about the public interest, and are willing to reexamine their preferences, and give opposing views more of a hearing.
There is some real merit to this perspective. But Sunstein overlooks another important way in which consumers differ from citizens. As noted above, consumers make decisions they know are likely to make a difference. That gives them stronger incentives to seek out information and consider it at least somewhat objectively. By contrast, when we are in "citizen" mode, we may indeed care about the public interest. But we also know that our choices are unlikely to make much difference, and therefore make very little effort to either assimilate information or ride herd on our prejudices. That is why studies indicate that even otherwise knowledgeable and sophisticated people tend to be far more biased in the way they process political information than in their approach to similar data about private sector decisions. We are usually smarter and less biased when we "vote with our feet" than when do so at the ballot box. Websites and other information sources catering to foot voters may also be more balanced and otherwise more helpful than those that cater to the political fan communities produced by ballot box politics.
Perhaps, therefore, the antidote to the perils so effectively outlined by Nichols and Sunstein is not less consumerism, but more. We should work to ensure that more of our decisions are like the ones we make about smartphones and cars, and fewer have the perverse structure of incentives that exists at the ballot box. In some cases, that may mean taking some issues off the political agenda and leaving them to the private sector. In others, the better path might be decentralizing more power to the local and state level, so that people will have more opportunities to "vote with their feet" between jurisdictions. Consumers are likely to act more responsibly if we give them more opportunity to make choices that actually matter.
Like private sector consumption decisions, our choices about where to live in a federal system are generally based on better information and reasoning than ballot box decisions. Decentralizing power could also help mitigate the partisan bias and hatred that Sunstein rightly identifies as a serious menace to political discourse.
No one solution is likely to fully overcome the dangers described in these two excellent books. The right strategy, if there is one, may well involve a variety of different measures. But those who worry about the perils of political ignorance and the sad state of political discourse would do well to consider the possibility that we should give consumer choice more of a chance.