Maybe we're not all that polarized

Partisanship is a very muddled concept these days. That’s the only conclusion you could draw from a comparison of recently released data on partisanship in the 50 states. Two respected national polling organizations, Gallup and Annenberg, have graciously shared the partisanship data gleaned from polls they conducted in each state throughout 2004. For Gallup, this was the fruit of 37,000 interviews. Annenberg’s data are based on an astounding 67,777 interviews.

Partisanship is a very muddled concept these days. That’s the only conclusion you could draw from a comparison of recently released data on partisanship in the 50 states.

Two respected national polling organizations, Gallup and Annenberg, have graciously shared the partisanship data gleaned from polls they conducted in each state throughout 2004. For Gallup, this was the fruit of 37,000 interviews. Annenberg’s data are based on an astounding 67,777 interviews.

Partisanship — or party identification, as it is sometimes labeled — is not the same as party registration. Party registration is an official status in 28 states whereby voters declare a permanent party affiliation when registering, but that’s not what we’re considering here. Partisanship is more of an attitude. A common question designed to ferret out partisanship asks, “Do you think of yourself as being a Republican, Democrat, independent or something else?”

Supposedly today we have a highly partisan and polarized electorate that is easily broken down into red and blue states based on partisan leanings. This hardening of leanings was seemingly reflected in polls conducted throughout 2004 showing historically low percentages of undecided voters in the presidential contest. Voters had chosen sides, either Republican or Democrat, and that was all there was to it.

Based on that, I would have expected that the Gallup and Annenberg data would be roughly comparable. If a state is mostly Republican in one poll’s view, I would expect it to be mostly Republican in the other one, too. If partisanship is as polarizing and divisive as pundits contend, then I would expect it to harden attitudes into a firm foundation. But that’s not the case.

I should point out some dissimilarity in ways the two organizations collected and released the data. First, Gallup asks voters who might at first declare independence to lean toward a party. That reduces Gallup’s percentage of independents to single digits in all but one state, Michigan. Annenberg is happy to leave independents in the middle box, so its results typically show much higher percentages. Annenberg also differs in conservatively omitting eight small states from the report on account of the few interviews there. To make apples-to-apples comparisons, I concentrated on rank-order differences in the two studies.

The results show similarities, but dissimilarities are more notable. Take Nebraska for example. Both organizations have Nebraska in their top 10 states for GOP partisanship. Gallup ranks Nebraska the seventh most Republican state, while Annenberg awards the Cornhuskers a national championship in Republicanism, giving it the No. 1 ranking that has eluded the state’s legendary football team lately. But despite the ranking difference, the two organizations are nearly identical on the actual numbers. Gallup says that Republicans have a 13-point advantage over Democrats; Annenberg sees a 17-point margin. Close enough, for sure.

Utah is a more curious case. Both polls rank Utah high in Republican partisanship — No. 1 in Gallup and No. 2 in Annenberg’s rankings — but they are very different on the magnitude of the GOP advantage. Gallup shows Republicans with a 36-percentage-point margin over the Democrats, while Annenberg’s margin is just 25. That 11-point spread is enough to matter.

There is even less agreement between the reports regarding the most Democratic and independent states. Massachusetts, the state Annenberg ranks No. 1 in independents, doesn’t even make Gallup’s top 10. But it is Gallup’s most Democratic state. While Annenberg sees 47 percent of Massachusetts voters being independents, Gallup says that only 6 percent are. Gallup’s movement of leaners to a partisan column partially explains this difference, but it cannot wholly reconcile the vast gulf separating these two reports.

Comparing only those 42 states covered by both reports, Gallup and Annenberg agree only 40 percent of the time on the top 10 Republican, Democrat and independent states. That tells me that partisanship is much less a stable force in American politics than is often supposed.

Hill is director of Hill Research Consultants, a Texas-based firm that has polled for GOP candidates and causes since 1988.