Anti-immigration reform protesters (Jose Luis Magana/Reuters) Anti-immigration reform protesters in Washington, D.C., in July. (Jose Luis Magana/Reuters)

Greg Sargent had a very nice look at the polling on immigration among Republicans today. Short version: comprehensive immigration reform needs to be sold, but most Republican voters are open to it.

It’s a useful post, but I’d be careful about what it’s telling us.

Remember, to the extent that members of the House are primarily concerned about their own reelections, the question isn’t so much whether a vote for immigration reform will cost them their seat, but whether it incrementally increases the chances of losing their seat. And the polling here is consistent that it presents at least some danger.

First of all, if the people who strongly dislike any path to citizenship are the ones who care the most about the issue, then they’re far more likely to vote on that issue. In a contested primary, it could be a big deal if 20 percent become single-issue voters opposing an incumbent over immigration, leaving the other 80  percent to vote on other issues; it’s an even bigger deal if intensity has turnout effects, and that 20 percent among Republicans becomes 30 percent or 40 percent of the primary electorate.

But an even bigger issue is whether immigration would be enough to get a primary challenger in the first place. No one loses renomination without opposition! So if immigration turns out to be the kind of issue that can push challengers to run, then it’s a big loser for Republican members of the House even if hardly anyone (else) in the district cares about it — because once a primary is contested, there’s always the chance of an upset loss.

All that said, what really matters in terms of results in the House isn’t so much whether a vote for comprehensive reform would yield primary challenges or even primary defeats. Anyone who studies House elections knows how rare those defeats are. No, what really matters is what Republican members think will happen. And, for that, the problem is that politicians are paranoid by nature about elections. They always overrate the effects of their actions — statements, votes, everything — on future electoral results. It’s actually a good thing that they do so; it’s part of what makes representation and, therefore, democracy work.

So what matters here is what potential threats that paranoia fixates on. Not what will really happen. Not whether they’ll be paranoid. But what they’ll be paranoid about.

And for that, the efforts of a generation of activists and organized groups to convince them that their greatest threat is conservative primary challengers turns out to be a really big deal. Most people now think that will be enough to sink immigration reform, even when the polling suggests the result should be otherwise, and even when the collective interest of the party might be to pass something that will make the issue go away.