I blogged about this five years ago, but reader Dilan Esper suggested I repost it; so here it is, with a few small changes:
Many people are opposed to “party line” voting (by which I mean voting in partisan general elections based almost entirely on candidates’ party affiliation, rather than focusing more on each candidate’s own particular politics or character). Such party line voting, they argue, shows laziness, stereotyping, or lack of independence. I, on the other hand, think that in most situations voting for the party and not for the candidate is the most sensible approach. A few words on why.
1. Elections of legislators (federal or state). Which particular people are in the legislature definitely matters — but which party has a legislative majority matters far more. Legislative power is generally exercised by organized legislative party blocs, not by individual representatives who make up their own minds.
I generally think the country would be better off if the Republicans (for all their warts) are in control than if the Democrats are. So if I and those like me vote for a Democratic candidate over a Republican because we think that this particular Democrat is better (smarter, more honest, or even more in agreement with us on many issues, despite his party affiliation), and this candidate’s election ends up giving Democrats control of the relevant legislative chamber, then we’ve hurt the causes that we favor: By electing this candidate, whom we like, we’ve essentially elected a party that we dislike. And even if the candidate breaks with the party in some cases (which may be part of why we voted for him), in most situations — both when voting on legislation, and, as importantly, voting on whom to put on various legislative committees and the like — he’ll follow party discipline.
2. Elections of Presidents and governors. Presidents and governors ostensibly exercise executive power by themselves, so I may well imagine that an honest, smart Democrat may do a better job than a dishonest, dumb Republican.
But in reality, electing a President or governor also means electing his party, and not just him. First, he’ll probably select a cabinet that’s drawn from a wide chunk of his own party (since, among other things, he needs to maintain good relations with the party faithful). He may well appoint some judges that he might not much like, but that help cement relations with various wings of his party. And a Democratic President may let a Democratic Congress get its way on more issues (even ones on which he doesn’t fully agree with them), or may block the Republican Congress’s proposals (even ones which he doesn’t much disagree with), because that’s what his party base will want. (Naturally, all this applies equally to Republican Presidents.)
This suggests to me that one should basically ask “Which party do I want to see in power?,” and then vote for candidates of that party nearly all the time — because you are in effect electing a party, more than you are electing a person. There are, of course, some exceptions (“Intentionally voting for split government” is one that I’ve heard most talked about this year):
A. The truly awful candidates. Some candidates are so bad — either so personally dishonest or corrupt, or so far from your own views — that you might refuse to vote for them, even if you generally support their party. First, you might just feel that you can’t endorse this person. Second, even if your view is purely instrumental, you might think that it’s better for your party if this member of the party is defeated than if he wins. I’d have voted against David Duke in Louisiana for both of those reasons.
B. Trying to affect the position of your own preferred party. In some situations, you might feel more strongly about sending a message to the leadership of your preferred party than you do about winning this particular election. I take it that this is what some (though not all) Nader voters who otherwise preferred Democrats were thinking in 2000, as were some Libertarian voters who otherwise preferred Republicans. I think this is a dangerous game to play, but if that’s what you’re self-consciously trying to do, then I can see why you’d vote against your otherwise preferred party. Again, though, the focus should be on which party you want to elect, and not whether you think this particular candidate is a better person than the other one.
C. Nonpartisan elections and primaries. Naturally, none of this applies to nonpartisan elections or to primaries where you’re choosing from several candidates within the same party.
D. Intentionally voting for split government. You might conclude that it’s better for the country if control of the House, Senate, and the Presidency are split; if that’s so, then you might intentionally vote for one party in one race and another in another.
E. Mixed preferences for a major party and a third party. You might want to vote Republican if the election is close and your vote might matter to the outcome (see item I below on that), but Libertarian if the election isn’t close and you want to express support for Libertarians; likewise for Democrats and Greens and similar combinations. But this too is a reflection of the core strategic principle, which is usually “vote for the party, not for the candidate.”
F. Different preferences at different levels of government. My suggestion that you should choose your preferred party operates at each level of government; you might, for instance, conclude that you want the Republicans to run the country, but you want the Democrats to run the state government of your own state. That’s consistent with my general theory. (Who controls the states may affect who controls the federal government, and vice versa, but the effect is weak enough that supporting different parties at different levels may make sense.) I would just say that you should then vote for Republican candidates virtually all the time at the federal level, and for Democratic candidates virtually all the time at the state level.
G. The voter who is really in the middle. If you are really so much in the middle that you are indifferent whether the Democrats or the Republicans win, then of course it makes sense to vote for the particular candidate that you like most: The strategic party power considerations are by hypothesis irrelevant to you — you’d rather elect (say) a smart Republican over a dumb Democrat because you just don’t care whether the Republicans or the Democrats are in power. I think few people really feel that way, though; I think that even if you aren’t a consistent supporter of one party or the other, and even if you disagree with much that either party would do, you can probably identify what you think is the lesser of two evils (or the greater of two very slight goods) in any particular election.
H. The voter who is near the middle, and who really cares a lot about a particular candidate. This is a more common variant of G — say you mildly prefer the Democrats over the Republicans, but you think that this Republican candidate is much better than the Democrats, in a way that really matters: Either he has the brilliance or charisma to really add a lot to legislative deliberations, or he is running for an executive office, where one person’s character, intelligence, and policy preferences really can make a huge difference. This too may be reason enough for you to abandon party-line voting; but again, you’d have to really feel that this one candidate can do so much good that the expected value of this good really exceeds the expected harm of the wrong party (from your perspective, of course) getting power.
I. Expressive voting. Of course, there’s been something of a fiction behind this whole discussion of strategic voting — the fiction that your vote will make much of a difference. In fact, your one vote is highly unlikely to affect things (though in a few elections, including even Congressional ones, the result does turn on a handful of votes). We therefore vote largely because of how it makes us feel about ourselves (e.g., because we’ve been taught that to be a dutiful citizen, one should vote, or because we feel good about voting for a particular person we really like). So under that approach, if it makes you feel good to think that you vote based on each candidate’s individual merit, that’s what you should do, and never mind the practical consequences, since as a practical matter, your one vote is extaordinarily unlikely to actually have any practical consequences.
Still, I find it hard to feel good about casting a vote that, if duplicated by many others, would actually lead to a result that I dislike. That might ultimately be more an aesthetic rather than rational judgment here, but that’s what expressive voting necessarily involves. Thus, so long as I feel that I ought to vote, based on the fiction that my vote does have some practical effect, I prefer to go through with the fiction, and ask what practical effect that I’d like to have. And since my preferred practical effect is having the Republican Party be in power, I almost always vote party line.
Finally, I’m sure that there are other reasons to depart from the “vote the party, not the candidate” approach, though I suspect that most are variants on some of the ones I identify above. I’m not saying that party line voting is always the right approach (right in the sense of being most likely to lead to the results that you, as a voter, like). I just think there is enough reason for me to adopt it as a very strong presumption in my voting choices.