Readers of this blog know of my obsession with the topic of character-based voting. Suppose that we accept some workable distinction between matters of character and matters of policy with respect to politicians and political candidates, each a potential consideration for or against their continued stay in office or their candidacy.* What role should judgments of character play? Is it ever justifiable to vote for or against a candidate (or support or remove a sitting candidate) on grounds of character abstracted from considerations of policy? Clearest version of the question: can a person’s moral character ever be bad enough to disqualify him or her from office independently of anything we know about their views on policy, or even in defiance of the knowledge that they have the “right” views on policy?
A long-standing consensus in political science (and to some degree political philosophy) has asserted that “policy>character.”** The quasi-mathematical formulation seems to imply that while considerations of policy and character both have some weight, those of policy always exceed those of character in political deliberation. I can’t say that I’ve done an exhaustive study of the matter, but I’ve done my share of reading in both political science and political philosophy. I so far have not seen a credible argument for this claim. Political scientists seem to assume it a priori, and leave matters there. Political philosophers argue against character-based voting by over-simplifying the issues, and arguing against strawman versions of character-based voting. If anyone can find or make a sound, cogent argument for “policy>character,” I’d be much obliged.
It’s worth noting that almost no one in the world of ordinary work hires by adopting any equivalent of a “output>character” formulation. Matters of character (however ill-conceived the moral framework, however epistemically impoverished the information used) almost always play a significant, sometimes the determinative role. Hiring a person is different from–a little more complex than–buying a robot. No simple arithmetic formula is going to capture the normative considerations involved, unless the person doing the hiring is himself a robot, and/or is programmed by someone whose mental processes operate in robotic fashion. It’s not clear why political elections should be so different in this respect from ordinary hires.
The case of Mark Samsel strikes me as a QED for character-based voting. Here’s the specifically criminal part of the allegation against him:
WELLSVILLE — Kansas Rep. Mark Samsel was arrested on a charge of misdemeanor battery in connection with an incident that reportedly took place Wednesday, April 28, while he was working as a substitute teacher for Wellsville USD 289.
Samsel, 36, of Wellsville allegedly was involved in an incident with a student that school administration reported to authorities. An investigation was conducted by the Wellsville Police Department and the Franklin County Sheriff’s Office, which led to Samsel’s arrest on a charge of misdemeanor battery, according to a news release from the Franklin County Sheriff’s Office.
Samsel was booked into the Franklin County Adult Detention Center at 3:42 p.m. Thursday, April 29, and was later released on bond.
There’s video of the relevant incident here. More here.
I grant that some of the commentary on Samsel’s alleged behavior is as stupid as the behavior itself. My aim here is not to discuss the Samsel case per se (I regard him as innocent until conclusive proof of guilt emerges), but to use it as a realistic thought-experiment. Imagine that a hypothetical Samsel ends up being guilty of all the accusations and criminal charges made against him. Now imagine that this hypothetical Samsel is up for re-election, running against candidate Ramsel.
Five scenarios:
- Samsel runs against policy-equivalent candidate, Ramsel-1. Voters vote for Ramsel-1, using considerations of character as a tie-breaker.
- Samsel runs against a candidate, Ramsel-2, who has little or no track record on policy. Samsel’s own track record is mediocre or unremarkable, neither particularly bad nor particularly good. Further, there is no indication that Ramsel-2’s policy views are themselves particularly bad (or particularly good). What’s clear is that Ramsel-2’s character is superior to Samsel’s. Voters vote for Ramsel-2, using considerations of character as the relevant fact that resolves the epistemic indeterminacy they face.***
- Samsel runs against a candidate, Ramsel-3, whose views on policy are slightly worse than Samsel’s have so far been. But voters take Samsel’s behavior to be a proxy for future untrustworthiness and instability that calls Samsel’s past successes into question. Ramsel-3’s views are not so much worse than the prospect of Samsel’s untrustworthiness and instability to put Ramsel-3 out of the running. On the contrary, the reverse is true: Samsel’s untrustworthiness and instability is so egregious that it disqualifies him as a viable candidate, even against a policy sub-optimal candidate.
- Samsel runs against Ramsel-3, except that voters’ reason for voting for Ramsel-3 is a matter of moral desert. Ramsel-3 may be policy sub-optimal relative to Samsel (at least so far), but Samsel deserves punishment of some sort (or deserves the imposition of reparations that make it impossible for him simultaneously to pay the reparations and reap the rewards of office), whereas Ramsel-3 by hypothesis does not. So voters decide to punish Samsel by voting against him, treating Ramsel-3’s expected policy sub-optimality partly as a sunk cost, and partly as a problem to be remedied (on the supposition that Ramsel-3’s policy views are not so fixed that they can’t be influenced in a more positive direction).
- We can imagine a scenario that’s a complex hybrid of (3) and (4) above.
Defenders of a categorical “policy>character” view are committed to thinking that the voters in the preceding scenarios are doing something irrational. What is it supposed to be? Without an answer to that question, the view suffers from a major lacuna that calls into question the reasonability of the motivation for it. Suffice it to say that I haven’t seen an answer. Always looking, though.
*I’m not suggesting we should. I’m just supposing for present purposes that we do.
**The consensus is slowly unraveling, at least in the sub-specialty of electoral studies. As well it should. But that’s a topic for another post.
***Scenarios (1) and (2) are slight variations on one another. Both involve tie-breaking, but of subtly different kinds. In (1), voters have the relevant information of policy equivalence between Samsel and Ramsel. In (2), they lack it. Hence in (1), there is policy equivalence; in (2), there might be. In (1), character breaks a known tie. In (2), character serves as an insurance policy in the context of a gamble on policy.
Policy of Truth 
It seems wrong — obviously so once we think about it and as your cases show — that considerations of policy (even when the value/disvalue is small) always outweigh considerations of character (even when the value/disvalue is large) in political deliberation, voting, etc. I wonder, though, if there is not a more plausible and subtle general policy>character type of thesis? After all, one of the main points, if not the main point, of government is to enact (good, necessary) policy; and it is not one of the main points of government to have elected (or other) officials or officers that have excellent character, don’t have awful character, etc. Maybe these types of considerations point toward the right way of giving policy a certain pride of place — though I have not thought about these matters enough to have a specific formulation or suggestion here. But, clearly, even if some kind of pride-of-place policy-over-character thesis is true, clearly the application or conclusion should not be that considerations of character are always unimportant, never make the difference or never outweigh policy considerations.
LikeLiked by 1 person
Yes and no. It’s complicated.
Yes, at some level, political offices exist primarily to promote the best policies, not to reward the best people. In that sense, the policies a candidate is expected to promote should be more important to a voter than the character she’s thought to have.
But one problem here is epistemic. Often, we have better evidence of a person’s presently-existing character than we do of the policies we expect them to promote. So it can often make sense to make a decision on evidence X that has better epistemic credentials than evidence Y, even if Y is more directly relevant to the decision than X. It can make sense to go with what you actually know instead of what you half-expect, even if the half-expectation is more directly relevant to the outcome you seek.
Second, background assumptions about character mediate our expectations of what a candidate is going to do about policy. If I promise to promote policy P in office (where P is the best policy), it can only be rational to vote for me if you expect me to live up to my promise–a matter of trustworthiness or fidelity, that is, of character. If we’re talking about voting for political candidates, after all, we’re not talking directly about policy, as though voting for a candidate just was voting on a policy. We’re talking about treating the candidate’s intentions as a proxy for an expected policy. When I vote for Smith, I am voting on the supposition that by voting for Smith, who says she will promote P, I am in that (very) indirect sense voting for P. But if Smith is full of shit, all bets are off.
Third, even if Smith is fully sincere, she needs the skill to translate her intentions into successful policy. In a democracy, we’re not voting for autocrats who, once elected, simply push some button or move some lever, and thereby enact policy. Policy involves a long and complicated process. The process requires certain traits of character, and rules out others. So a good candidate has to be politically effective, which is a matter of both political and moral character (instrumental effectiveness regulated by justice). Take two candidates. Smith has the best policy views, Jones is policy sub-optimal. But Smith is such an asshole that you can’t expect her to work with anyone effectively to translate those views into actual policy. Whereas you can expect that with Jones. It’s rational to vote for Jones, then.
Fourth, candidates have worked-out views only on a small set of policy issues. But one can’t assume that those issues are the only ones they will confront while in office. Few expected (a year and a half ago) to have to confront COVID, for instance. So epistemic-moral character matters with respect to policy indeterminacy or open-endedness: does the candidate have the epistemic capacity to deal with conceptual novelty? Is she flexible enough, or smart enough, or saavy enough (etc.) to handle the unexpected, or simply those issues that haven’t been central to her own interests, but are still central to political life? Does she know how to make effective use of the epistemic division of labor? Is she open to new ideas, responsive to feedback, etc.? All matters of character.
So despite the commonsense claim you make, I’m still inclined to balk a bit at a quick “yes.” Discussion of this issue in the literature is remarkably unsophisticated, over-simplified, and frankly primitive. It takes a third-person perspective on voting that abstracts far too much from the real-world considerations that ought to go through the head, from a first-person perspective, of an actual voter voting for a candidate, i.e., person, not voting on policy in a referendum. The social science I’ve seen is hugely over-simplified, and focused on operationalizable variables, not on the full context behind them.
Yes, at some level, policy is fundamental. But the fact remains that voting is a mechanism for enacting policies that causally speaking go through candidates. So one is voting for policy via a candidate, not directly on policy. In that case, character becomes hard to wish away even as a primary deliberative consideration. Facile formulations like “policy>character” simply abstract away from real world complexity in a clueless way. Yes, some highly qualified version of that claim is likely to be true. But the Devil is in the details, and I haven’t seen anyone come up with them. That hasn’t stopped both philosophers and political scientists from making bold pronouncements about the normative fundamentality of policy over character when it comes to what the voter does on the ballot.
LikeLiked by 1 person
One last complication I’ll mention in passing: when we vote for legislators, we’re voting for candidates who are going to pass laws that lead to policy. Legislation is not by itself policy. So the causality involved is yet more attenuated than is usually thought. One needs to know whether the candidate has the sense to grasp how legislation relates to policy. Part of answering this question is asking whether the candidate has common sense, a robust sense of reality. If you deal with a brilliant policy wonk, but find that despite the brilliance of her ideas, she has no clue how anything actually works in the real world, that’s a good reason for not voting for her. Cluelessness is a disqualifying trait; clued-in-ness is a qualifying trait. That’s true for many reasons, but the worst conceivable candidate is the one who thinks, full time, like a lawyer, fixated on legalistic minutiae without bothering to grasp how legislation will be translated into de facto policy.
The social science on this issue does not even approach a rigorous, determinate answer to the relevant question here–no matter how much bluffing the defenders of quantitative social science would like to do to make the reverse pretense. The question is: when people judge someone lacking in common sense, does this judgment track an actual defect in candidates up for election? Can plain, politically ignorant people discern when someone is just a lawyerly policy wonk, full of brilliant ideas, but too clueless to understand how any of those ideas will actually play out in the real world?
I can’t claim to have done an exhaustive reading of the whole literature on electoral studies, but I’ve read enough to know that political scientists haven’t answered this question, and can’t. So when a Jason Brennan blusters his way into print by claiming that social science has discredited the idea of character-based voting, take that claim with a gigantic grain of salt. The citations in The Ethics of Voting on this subject don’t even begin to check out, and there’s nothing in his writing that addresses it, but to quote The Offspring, yes, in his own mind, he’s the deepest trip.
All I can say is: when they get that social science together? Give it to me, baby.
LikeLiked by 1 person
I agree with all of that. It is largely for these sorts of reasons that I am unwilling to go out on a limb with some kind of suitably nuanced policy>character principle. One thing I would add is that, often — and I suspect rationally — people vote on the basis of values, not policy (perhaps not knowing much about the specifics of policy). So, often, the route to policy is through values and that is often pretty orthogonal to general character. If one of your most important political values is the right of women to have non-late-term abortions, you might rationally vote for someone who is lacking (at least somewhat) in many of the virtues you outline above — if you have good evidence that they will stand up for this value in the political process (even if they will do so irrationally, beyond what you think is appropriate). It is complicated, but I agree that good policy is usually quite a bit “downstream” from advocacy and voting (and to some extent even the back-and-forth of normal politics, which is often as much a battle over values and mutual respect, or lack thereof, than a matter of policy debate).
It strikes me that Brennan’s stronger pro-expertise points concern things like actual policy not being economically (or otherwise logistically) illiterate. This speaks most directly to such expertise entering at some point in the process, before we get bad policy. But that is a pretty weak thesis and pretty far removed from things like the appropriate reasons for voting for this as against that candidate (and, again, perhaps pretty far from appropriate or effective everyday, retail politics).
LikeLiked by 1 person
Well, I agree with that.
I think the inference to draw is that justified voting is multiply realizable. In other words, there’s a multiplicity of very different sorts of considerations at work when you vote for a political candidate, ranging from issues of character to where the candidate stands on your pet issue or issues, and despite their apparent non-relation to one another, they can all be justifiable. You vote for Smith because she’s trustworthy; I vote for her because of her counterfactually rock stable position on abortion. Nothing wrong with either thing. And neither approach to voting makes the sorts of epistemic demands that Brennanite epistocrats seem to demand of voters. There is such a thing as wrongful voting, but Brennan’s standards of correct voting are far too demanding, and motivated less by a justifiable account of voting than by indignation at the economic illiteracy of voters.
I don’t disagree with Brennan’s pro-expertise point on economics. Too many voters are economically ignorant, and vote on their ignorance–which is certainly a problem. But he and Caplan have taken this legitimate insight and put weight on it that it can’t possibly bear. This one legitimate point is not going to get you to epistocracy, unless you confuse the issues so thoroughly that we lose the distinction between democracy and epistocracy. As I’ve said before (I think), I basically agree with Piero Moraro’s and Robert Talisse’s criticisms of Brennan’s views on epistocracy, and regard those criticisms as pretty much conclusive.
I would dispute the idea that the average political philosopher anywhere has a workable sense of the logistics of anything beyond academia. Microeconomics is one thing, but microeconomics is not logistics. A person who lacks first-personal knowledge of the logistics of X is unlikely to know much of value about the logistics of X, and is not going to learn it because he knows the microeconomics of X.
Time for lunch. I’m an expert on those logistics.
LikeLiked by 1 person