Home > News > Larry Evans on the House Votes on Health Care
137 views 7 min 0 Comment

Larry Evans on the House Votes on Health Care

- November 10, 2009

At my request, Larry Evans sends the following, based on his analysis of the House vote on health care reform:

bq. Saturday night’s roll calls on the House health care bill were fascinating for Congress watchers. If you looked closely at the endgame as reported on C-SPAN, the number of Democratic yes and no votes steadily rose until about 10 minutes were remaining in the roll call, at which point the no side basically locked in at 36 and then 37 votes. Yes votes continued climbing until 6:51 was remaining, at which point Maxine Waters put her party over the top. Next, two of the three remaining Democratic nonvoters immediately voted no. And then the lone GOP aye was cast by Joseph Cao.

bq. Based on how the last-minute decisions broke, it looks like the Democratic leadership had some votes in their pockets, yes if needed, and then called in just enough of these chits to put them over the top. GOP whip Eric Cantor also kept Cao from casting his yes vote until the very last minute, thereby forcing Pelosi to get one more recalcitrant Democrat to cast an uncomfortable vote in favor of the package. Republicans are probably already thinking about the commercials they are going to run in that district next year.

bq. The Democratic vote is especially interesting, especially in light of new district-level data about health insurance coverage at home. As the Urban Institute reports in a fascinating and troubling new policy brief (October 2009, Kenney et. al.), based on recent Census figures there is enormous variation across congressional districts in the proportions of kids and non-elderly adults who have private health insurance, government-provided health insurance, and who lack any health insurance at all. In Texas’s 29th district (parts of Houston), for example, a full 43 percent of non-elderly residents lack health insurance according to the Census Bureau, while in Massachusetts’s 3rd district (central and southeastern portions of the state), the figure is just 3 percent.

bq. You might expect some relationship to exist between district insurance coverage and the final passage votes cast on Saturday night by Democratic members on health care, right?

bq. Think again. Simple models of Democratic votes including the percent uninsured at home, ideology, and 2008 district support for Obama produce strong relationships in the expected direction for the last two variables, but no statistically significant relationship for percent uninsured or any other measure of health insurance coverage in the district. As mentioned extensively in the media, the Democrats who voted against their party on final passage tended to be junior moderates from districts that went for McCain in 2008. And just about every Republican voted no, so there is not much relationship on that side of the aisle between district health insurance coverage and the final passage roll call.

bq. Of course, there may have been an indirect relationship between district insurance coverage and the final passage roll call, one primarily exerted through party. Perhaps poorly covered districts are more likely to elect Democrats in the first place, whereas districts with high proportions of covered residents are represented by Republicans. Since the final passage vote divided the parties, district insurance characteristics may have influenced the outcome via party of the member.

bq. That’s easy enough to check. I modeled the party of the incumbent House member as a function of the percent uninsured, percent with private insurance, and percent with public insurance. This produced some interesting results. Indeed, as you would expect, districts with high public insurance rates are especially likely to have Democratic House members, while districts with high private insurance rates are statistically associated with Republican representation.

bq. But here’s the kicker ­ As the proportion lacking any insurance rises, the likelihood of having a Republican member actually appears to go up, not down.

bq. Why is that? A look at the map and the location of the “high uninsured” districts is instructive. These districts disproportionately are in Texas, Florida, and elsewhere in the South; places that in recent decades have been sending Republicans to Congress. The districts with low rates of uninsured residents, in contrast, tend to be in the Northeast and the Midwest, and often elect Democrats. Relatively low levels of unionization in the South play a role here, because a lot of people owe their health insurance to collective bargaining. State programs that provide significant insurance coverage (e.g., Massachusetts) are also less expansive in Republican-leaning areas. And more generally, the uninsured may be less likely to participate in politics, with obvious consequences for the quality of representation that they receive.

bq. One thing is clear, however. The considerable momentum in the House for extending health insurance coverage to the uninsured primarily results from broad ideological support that is spread throughout the constituencies with Democratic representatives in the chamber, rather than the districts with high proportions of uninsured residents, or exactly the people who probably have the most to gain personally from the passage of “Obamacare.” Now, on to the Senate.

See also Brendan Nyhan for some visualizations of the vote. If anyone else has some relevant analysis, please send it to me and I’ll be happy to post it.