In a recent New York Times op ed, former FBI director James Comey urges that "Every American should be speaking about our nation's values. Every American should be voting those values, which are far more important than even the most passionate policy differences." I agree with much of what Comey says in this piece. He's right that values are an important component of making good voting decisions, and I think he's also right to argue that many of the Trump Administration's policies are a menace to fundamental American values. Some of Comey's opponents on the political right also believe that values are the key to good voting decisions. Consider, for instance, the social conservative "Values Voter Summit."

Many take this kind of reasoning a step further and suggest that good values are really all you need to be a responsible voter. If so, perhaps I and other critics are wrong to worry about the problem of widespread voter ignorance. Even if voters don't know much about government and public policy, that does not matter much, so long as they have good values. They can then vote for candidates who espouse the same laudable principles and leave the policy decisions to them. Political philosopher Thomas Christiano even argues for a system under which values and policy decisions are separated, and voters only decide the former.

While values do matter, they are not a substitute for voter knowledge. In some respects, their importance actually implies the need for more knowledge, not less. I addressed these issues in some detail in a post I wrote last year, which I think remains relevant today:

[M]eans and ends in government are often closely entangled. Many issues – health care, environmental policy, welfare policy, education, and others – involve complicated combinations of facts and values. It isn't easy to see how the two can be disaggregated.....[T]he the size and complexity of government makes things more difficult. Among other things, the policy choices at stake involve tradeoffs between many different aims – such as those between increasing economic growth and raising environmental standards, for example.

The above points also apply to oft-made claims that voters can use "values" as an effective information shortcut.... Instead of comparing opposing policy platforms, the voters can just support whichever party or candidate seems to care more about their values, and leave the details of policy to the experts. The fly in the ointment here is that many of the issues at stake in most elections involve competing views about how we can best achieve widely shared values. Both Democrats and Republicans claim that they will grow the economy, create better opportunities for the poor and middle class, improve the quality of health care, protect us against terrorism, and so on. On these and many other issues, the big question facing the electorate is whether the parties' policies are likely to actually achieve these goals, and at what cost. Evaluating that requires factual knowledge, not just values.

A few controversial political issues, most notably abortion, may primarily come down to differences over ultimate values. But many others involve disagreements over facts and policies.

Knowledge is also essential for those issues where values really are at the heart of our disagreement. An electorate that knows little about factual matters will often do a poor job of evaluating values, too:

[I]f voters do a poor job of evaluating means, it is not clear that they are likely to do a much better job of evaluating ends. For the same entirely understandable reasons that few voters spend much time thinking about how public policy works, most also do little or no rigorous thinking about values. For example, how many people think seriously about the relative merits of utilitarian consequentialism as opposed to rights-based theories of morality? For most people, ultimate values are almost like religion: they accept those prevalent in their family or community without giving the matter much systematic thought – in part because there is little incentive to do so.

In my view, the values of the average voter in most Western democracies are not so bad. If the voters had a better understanding of how to achieve them, the quality of public policy would be vastly better. But many political theorists would argue otherwise. If you're a hard-core utilitarian like Peter Singer, a Rawlsian egalitarian, or a Catholic natural law theorist, your view of the average American voter's values is likely to be far more negative than mine. And even I think much of the electorate has significant moral blind spots, such as the assumption that it is morally permissible to make immigration policy with little or no consideration of the rights of potential immigrants themselves.

Some reject criticism of the voters' values by arguing that we can't really know which values are best. Who can say what's right? We have no reason to believe that one person's values are any better than another's. So why not let each citizen's preferences weigh equally, as "one person, one vote" democracy is supposed to do? This sort of relativist argument can be broadened into a critique of concern about voter ignorance generally. Who's to say what kind of knowledge matters and what kind doesn't?

The problem with this kind of argument is that its attempt to rescue democracy from its critics ultimately destroys any basis for preferring democracy in the first place. If we have no idea which values are best, how can we know that the values promoted by democracy are any better than those advanced by dictatorships or oligarchies? The true moral relativist has no reason to believe that the governments of the US or Canada are preferable to those of North Korea or Saudi Arabia......

The "one person, one vote" principle underlying democracy is not a morally neutral standpoint. It is based on certain assumptions about who should be allowed to participate in government and why – assumptions that can lead to multiple different interpretations of the idea.

If we think we know enough about values to conclude that democracy is superior to to dictatorship, then we should also be able to use the same principles to evaluate how good the voters' values are, and whether it might be desirable to limit their power over at least some issues. You can't logically be a moral absolutist about the comparison between democracy and dictatorship, and a relativist about competing values within a democracy. If, for example, Western democracies are superior to North Korea because they provide greater freedom and happiness to their people, then those same values can be used to evaluate the performance of both voters and other participants within the democratic political process.

Ultimately, there are good reasons to question voters' judgments about both factual issues and values. That doesn't mean we should do away with democracy entirely. Dictatorship is likely to be much worse on both fronts. But it does indicate we should be open to proposals for restructuring democracy to reduce the harm caused by public ignorance. In some cases, that may mean making fewer decisions at the ballot box, and more in other settings where people have better incentives to become informed.

I would add that, at this point, I probably take at least a modestly more negative view of the values of the average voter than I did when I wrote the above-quoted post last year. There are, I now believe, some issues on which we badly need a significant improvement in values, perhaps greater than I was previously inclined to think.

In another part of my 2017 post, I offered some additional criticisms of Christiano's proposal. See also this post on the same subject by Jason Brennan.