First things first: Call me easily amused, but I found it funny that Mississippi has both a Lincoln County and a Jefferson Davis County (the two opposing commanders-in-chief of the Union and Confederacy during the Civil War).
Second thing: Have I gotten myself a track record for being just slightly wrong now?
- In Iowa's June 3 primary, I predicted that state senator Joni Ernst would win the necessary 35% to avoid a convention nomination, but wouldn't win a majority. She did end up avoiding the convention--by clobbering her nearest competitors by 38 points.
- In Mississippi's June 3 primary, I said that neither state senator Chris McDaniel nor incumbent Sen. Thad Cochran would receive a majority; Sen. Cochran would lead, but the election would be forced to a runoff. I was right about the runoff but wrong about the standings (although McDaniel's lead was very, very narrow).
Now, most recently, I predicted that McDaniel would prevail in the runoff election held Tuesday. This, as you may recall, was mainly because 1) McDaniel had come in first in the actual primary and 2) most polling showed McDaniel up, and the few polls that had Sen. Cochran up gave him only narrow leads well within the margin of error. So you can imagine my surprise when Sen. Cochran defied the odds to be renominated for a seventh term in the Senate.
How were the polls this wrong? The key might be buried in the text of some survey releases. Democratic pollster Chism Strategies, which conducted four polls between the June 3 primary and the June 24 runoff, includes the following caveat in their releases: "The survey was restricted to voters from the GOP primary on June 3rd. It does not include potential voters who did not participate in that election." However, turnout increased by over 60,000 votes between June 3 and June 24--meaning that the Chism polls almost certainly did not represent those new voters in their samples. Similarly, in a survey conducted by The Polling Company, less than 1% of respondents were black. In Mississippi, as across the country, African-Americans vote overwhelmingly Democratic--Mississippi's large black population mean that of the 11 states of the former Confederacy, Mississippians gave the fourth-largest share of the statewide vote to President Obama in 2012--so their absence from a GOP primary poll is not too egregious.
Still, though, Mississippi holds open primaries--meaning that whether you're a Republican, Democrat, Alaskan Independent, whatever, you're allowed to vote in the GOP primary runoff so long as you didn't vote in the Democratic primary on June 3. And those voters are apparently exactly the ones Sen. Cochran reached out to in the three weeks between the primary election and the runoff--specifically, black voters. While Republicans may have supported McDaniel over Sen. Cochran, Republicans were not the only ones voting in the Republican primary.
How can we be sure that it was black voters who turned the tables in favor of Sen. Cochran? The answer is that we can't, but we can come pretty close. Here's a by-county map of the results from the June 3 primary (red is for Sen. Cochran, green for McDaniel):
Those familiar with Mississippi's demographics will take note of the two candidates' strongest performances:
That doesn't tell us much, though, unless we somehow know that it's black Democrats who are giving Sen. Cochran the votes in the west on June 3 (which we don't). Skip ahead three weeks to the June 24 runoff and we find a similar map:
Pretty similar--Sen. Cochran prevails in the west, McDaniel in the southeast, but the important thing to note is the change in both candidates' performances by region. It's easier to see on this map, which shows the change in the margin of victory: red means the margin of victory widened in favor of Sen. Cochran; green means the margin widened in favor of McDaniel.
The deepest red indicates an increase of at least 600 votes in favor of Sen. Cochran; in populous Hinds County, which contains the capital and largest city, Jackson, Sen. Cochran picked up over 7,000 votes, increasing his margin over McDaniel in that county by 5,678 votes. At the same time, the deep green in the northwest, DeSoto County, indicates that McDaniel increased his margin over the senator by 3,783 votes there. To a lesser extent on the map, you can see that Sen. Cochran increased his margins in the majority-black western counties.
But perhaps the most convincing evidence that Sen. Cochran's efforts to turn out black voters in his favor succeeded is the following graph:
It demonstrates an exceptionally strong (R-squared = 0.8717) correlation between the number of black voters in a county and the number of votes Sen. Cochran picked up in that county in the runoff compared to the primary. We can't draw a causal inference for sure, but it does seem to suggest that black voters, dormant in the Democratic primary (those who voted in the Democratic primary would not have been allowed to vote in the Republican runoff), woke up on June 24 and gave their votes to Sen. Cochran.
What's next?
In the general election, where Sen. Cochran will face former Rep. Travis Childers (who won the Democratic primary handily), there's no reason to think black voters will stick with Sen. Cochran. In large part that uneasy alliance seems to have been more of a "better the evil you know" situation--sure, you don't like Cochran, but do you want the neo-Confederate? In November we should expect to see black voters return to voting Democratic. An establishment Republican being renominated by black Democrats was the story of the primary, but it shouldn't last.
Now, if Chris McDaniel follows the urgings of some of his Tea Party supporters and mounts a write-in campaign against both Sen. Cochran and Rep. Childers--that'd be a story.
Still, though, Mississippi holds open primaries--meaning that whether you're a Republican, Democrat, Alaskan Independent, whatever, you're allowed to vote in the GOP primary runoff so long as you didn't vote in the Democratic primary on June 3. And those voters are apparently exactly the ones Sen. Cochran reached out to in the three weeks between the primary election and the runoff--specifically, black voters. While Republicans may have supported McDaniel over Sen. Cochran, Republicans were not the only ones voting in the Republican primary.
How can we be sure that it was black voters who turned the tables in favor of Sen. Cochran? The answer is that we can't, but we can come pretty close. Here's a by-county map of the results from the June 3 primary (red is for Sen. Cochran, green for McDaniel):
 |
| 2014 Mississippi Senate GOP primary results by county and margin of victory. |
Those familiar with Mississippi's demographics will take note of the two candidates' strongest performances:
- McDaniel performed most strongly in the southeast, especially in Jones County, which McDaniel represents in the 42nd district in the Mississippi Senate.
- Sen. Cochran performed most strongly in the west, in the counties irrigated by the Mississippi River. These counties are largely black and correspondingly Democratic, and there's an amazing scientific explanation behind it, as explained by marine biologist Craig McClain. (It has to do with plankton back in the Cretaceous Era; give it a read.)
That doesn't tell us much, though, unless we somehow know that it's black Democrats who are giving Sen. Cochran the votes in the west on June 3 (which we don't). Skip ahead three weeks to the June 24 runoff and we find a similar map:
 |
| 2014 Mississippi Senate GOP primary runoff election results by county and margin of victory. |
Pretty similar--Sen. Cochran prevails in the west, McDaniel in the southeast, but the important thing to note is the change in both candidates' performances by region. It's easier to see on this map, which shows the change in the margin of victory: red means the margin of victory widened in favor of Sen. Cochran; green means the margin widened in favor of McDaniel.
 |
| Change in raw vote margin between 2014 Mississippi Senate GOP primary and runoff. |
The deepest red indicates an increase of at least 600 votes in favor of Sen. Cochran; in populous Hinds County, which contains the capital and largest city, Jackson, Sen. Cochran picked up over 7,000 votes, increasing his margin over McDaniel in that county by 5,678 votes. At the same time, the deep green in the northwest, DeSoto County, indicates that McDaniel increased his margin over the senator by 3,783 votes there. To a lesser extent on the map, you can see that Sen. Cochran increased his margins in the majority-black western counties.
But perhaps the most convincing evidence that Sen. Cochran's efforts to turn out black voters in his favor succeeded is the following graph:
 |
| Source: U.S. Bureau of the Census QuickFacts. |
What's next?
In the general election, where Sen. Cochran will face former Rep. Travis Childers (who won the Democratic primary handily), there's no reason to think black voters will stick with Sen. Cochran. In large part that uneasy alliance seems to have been more of a "better the evil you know" situation--sure, you don't like Cochran, but do you want the neo-Confederate? In November we should expect to see black voters return to voting Democratic. An establishment Republican being renominated by black Democrats was the story of the primary, but it shouldn't last.
Now, if Chris McDaniel follows the urgings of some of his Tea Party supporters and mounts a write-in campaign against both Sen. Cochran and Rep. Childers--that'd be a story.
No comments:
Post a Comment