Nate Silver of FiveThirtyEight.Com: Public Option Likely Popular in BlueDog Districts

Yesterday, Arkansas' Mike Ross, an influential Blue Dog Democrat, stated that he was opposed to a public option in the Democrats' health care reform package. "An overwhelming number of you oppose a government-run health insurance option, and it is your feedback that has led me to oppose the public option as well," Ross asserted in a letter to his constituents.

Ross may well have gotten a significant number of letters and e-mails against the public option. He may have hosted a town hall forum before an audience who was skeptical of such a provision. But if Ross had actually polled his district, it's unlikely he would have found overwhelming opposition to the public option. Instead, he might even have found a that a plurality or majority of his constituents supported the public plan.

Ross's district itself, the Arkansas 4th, has not been polled publicly. But some others like it have been -- and have usually revealed decent levels of support for the public option. In particular, Research 2000, via Daily Kos, has now surveyed the public option in Kentucky, Montana, Nebraska, Nevada, and the Tennessee 5th Congressional District occupied by fellow Blue Dog Jim Cooper. Research 2000 also conducted a nationwide poll on the public option. These six polls reveal a strong relationship between support for the public option and support for Barack Obama last November:



This would not appear to be good news for the public option in Ross's district, which gave Barack Obama just 39 percent of its vote. However, there also appears to be a secondary relationship between support for the public option and the poverty rate. Kentucky and Nebraska, for instance, each gave Barack Obama 41 percent of their vote. But in Kentucky, the public option is supported (barely) at 46-45, whereas in Nebraska it's opposed 39-47. What's the difference? Kentucky is much poorer than Nebraska -- 17.0 percent of its residents are impoverished, versus 11.5 percent in the Cornhusker state. Likewise, Nevada gave Barack Obama 55 percent of its vote, whereas Cooper's TN-5 gave him 56. But in Nevada, the public option is supported 52-40, whereas in TN-5, the margin is much larger: 61-28 in favor. TN-5's poverty rate is about 50 percent higher than Nevada's.

While Arkansas-4 does not have a lot of Obama voters, it does have a lot of people in poverty: 20.5 percent of its population, which ranks it 50th out of the 435 Congressional Districts. It is basically like an exaggerated version of Kentucky where, according to the Research 2000 poll, 46 percent support the public option and 45 percent oppose it. That the public option is "overwhelmingly" unpopular in such a district is unlikely.

We can systematize these results by means of a regression analysis that accounts for the Obama vote share and the poverty level in each district. (Technically, we'll be using a logistic regession, treating each of the voters included in one of these surveys as a separate data point.) This analysis finds that support for the public option nationwide is about 55 percent, against 36 percent opposed, similar results to what I believe to be the most reliable polls on the subject.

What's more interesting, though, is where we project the public option in individual districts. We find that:

-- The public option is estimated to have plurality support in 291 of the 435 Congressional Districts nationwide, or almost exactly two-thirds.
-- The public option is estimated to have plurality support in 235 of 257 Democratic-held districts.
-- The public option is estimated to have plurality support in 34 of 52 Blue Dog - held districts, and has overall popularity of 51 percent in these districts versus 39 percent opposed.

Obviously, there is a margin of error inherent to this analysis when applied to any individual district. The polls that inform this analysis themselves have a margin of error, and there is an additional layer of error introduced by the statistical process that we apply to the data. But in Ross's district, for what it's worth, the projected numbers are 49 percent in favor of a public option and 41 percent opposed.

The districts represented in blue in the map below are those where we'd project the public option to have plurality support if a poll were conducted there; those in red are where we expect a plurality to oppose it:



The raw projections for all 435 districts can be found below. Blue Dog Democrats are indicated by a lower-case 'd' and a lighter shade of blue. In Congressman Sanford Bishop district, 75% favor a public option while 21% oppose. John Barrow district, 69% favor while 25% oppose. But in a surprise, Lynn Westmoreland district, 48% favor, while 42% oppose.