Thad Cochran's unexpected victory on Tuesday night was unexpected, in part, because very few polls showed him beating Chris McDaniel. (It was also unexpected because there's very little precedent for incumbents who finish second in a primary in a runoff state to surge and win the runoff.) However, rewind the tape back to Eric Cantor's unexpected loss a few weeks ago, where something very similar happened. In Mississippi as in Virginia, it was an open primary, and the pollsters weren't thinking outside the box. However, the victorious campaigns did think outside the box; they targeted voters outside the universe of likely voters in a Republican primary.
In Cochran's case, that meant targeting Democratic voters who, while they might not want to vote for Cochran in a general election, might still easily prefer Cochran, an old-school appropriator, to McDaniel, a shrink-the-government-at-all-costs type with alleged neo-Confederate ties. Misssissippi is one of the most racially polarized states in terms of voting patterns (in 2008, 98 percent of blacks voted for Barack Obama and 88 percent of whites voted for John McCain), so that meant explicitly trying to make inroads among African-American voters.
There were anecdotal reports before the election, of a Cochran-affiliated PAC hiring a Democratic operative to do GOTV in the African-American community, and the McDaniel campaign clearly saw the threat too, with their last minute push to put poll intimidators watchers in place. It seemed like a Hail Mary at the time ... but looking at Tuesday night's election results, it very clearly worked.
As you can see in the maps above, there's a clear relationship between the counties where Cochran greatly increased his vote between the two elections, and the counties with the largest black populations. The map on the left is based on Cochran's vote gain, not in terms of raw numbers (which often weren't very large; it was often a gain of just a few hundred votes, but when Cochran got only a few hundred votes in those counties in the first round, in some cases it was a gain of double or more), but in terms of percentage. The map on the right is based on the percentage of black residents in each county. (If you'd prefer to see the data in scatterplot form, several other observers have done that, though the maps make it look more striking.)
There's more over the fold ...
If we're talking about gains of only a few hundred votes per county, though, why does that matter in the big picture, you might ask? For starters, in a game of inches like this (McDaniel beat Cochran by 1,314 votes in the primary, and Cochran beat McDaniel by 6,373 votes in the runoff), every vote counts.
In addition, you can see the effect just by zooming in one county: Hinds County, where Jackson is located. This is the state's most populous county, and with most of the Jackson area's white residents having decamped to exurban Rankin and Madison Counties in the last few decades, it's now 69 percent black and also the state's largest pool of black votes. Cochran got 10,928 votes here in the primary, but that shot up to 17,927 votes in the runoff. McDaniel also increased in Hinds, but only from 5,621 to 6,962. In other words, Cochran gained 6,999 votes, while McDaniel gained 1,341, a net of 5,658 in Hinds County alone. (Cochran improved his performance from 66 percent here in the primary, already one of his strongest counties, to 72 percent.)
Statewide, Cochran gained 36,979 votes from the primary to the runoff, while McDaniel gained 29,220 new voters. That's a Cochran net of 7,759 statewide. As you can see, the Cochran gain in Hinds County is almost as large as his gain statewide! In other words, much of the foundation for Cochran's win is found in this one, mostly-black county.
Now take a look at Jones County, which is McDaniel's home county and his stronghold. McDaniel got an almost-unreal 85 percent here in the primary, with 11,025 votes to Cochran's 1,816. But when it came to the runoff, McDaniel gained only 1,615 votes to go to 12,640, while Cochran gained 948 votes, going up to 2,764. Had McDaniel been able to build on his previous performance in Jones the way that Cochran did in Hinds, McDaniel would have won ... but instead, he was already almost maxed out on his friends and neighbors. (McDaniel in fact lost ground, dropping down to "only" 82 percent of the vote in Jones in the runoff.)
County |
Black % |
Cochran
primary
vote |
Cochran
runoff
vote |
Cochran
gain |
Cochran
gain % |
McD
primary
vote |
McD
runoff
vote |
McD
gain |
McDaniel
gain % |
Hinds |
68.8 |
10,928 |
17,927 |
6,999 |
64.0 |
5,621 |
6,962 |
1,341 |
23.9 |
Jones |
28.2 |
1,816 |
2,764 |
948 |
52.2 |
11,025 |
12,640 |
1,615 |
14.6 |
Total |
36.9 |
153,654 |
190,633 |
36,979 |
24.1 |
155,040 |
184,260 |
29,220 |
18.8 |
In fact, of the 10 counties that had the biggest percentage gain for Thad Cochran from the primary to the runoff, eight of them are also among the 10 counties with the highest percentage of African-American residents (and the other two are both at least 69 percent black). The single biggest gainer, Jefferson County, is also Mississippi's blackest county. Again, with only about 8,000 residents, Jefferson County by itself isn't a difference-maker. But when all the counties in the Delta make a similar move, it adds up quickly.
Interestingly, though, the number 11 county in terms of Cochran gain was Jones County, McDaniel's home turf! (It's the one county in the southeast that stands out in bright red on the left map.) Jones has a low African-American percentage (28 percent) by Mississippi standards; perhaps McDaniel's few local detractors were motivated to get out and vote against him after having shrugged off the primary.
County |
Black % |
Cochran
primary
vote |
Cochran
runoff
vote |
Cochran
gain |
Cochran
gain % |
McD
primary
vote |
McD
runoff
vote |
McD
gain |
McDaniel
gain % |
Jefferson |
85.4 |
121 |
321 |
200 |
165.3 |
95 |
111 |
16 |
16.8 |
Humphrey |
74.1 |
391 |
811 |
420 |
107.4 |
181 |
214 |
33 |
18.2 |
Claiborne |
84.0 |
161 |
295 |
134 |
83.2 |
112 |
112 |
0 |
0 |
Holmes |
83.0 |
365 |
688 |
303 |
83.0 |
257 |
296 |
39 |
15.2 |
Sharkey |
70.7 |
320 |
557 |
237 |
74.1 |
86 |
115 |
29 |
33.7 |
Coahoma |
75.3 |
611 |
1,050 |
439 |
71.8 |
298 |
345 |
47 |
15.8 |
Hinds |
68.8 |
10,928 |
17,927 |
6,999 |
64.0 |
5,621 |
6,962 |
1,341 |
23.9 |
Washington |
71.1 |
1,455 |
2,345 |
890 |
61.2 |
606 |
844 |
238 |
39.3 |
Tunica |
73.2 |
235 |
363 |
128 |
54.5 |
86 |
121 |
35 |
40.7 |
Sunflower |
72.5 |
771 |
1,186 |
415 |
53.8 |
358 |
428 |
70 |
19.6 |
The 10 counties where Cochran had the lowest gain—in fact, in six of them, Cochran actually got fewer votes in the runoff than he did in the primary—by contrast are mostly among the state's whitest counties. It's not as sharp a relationship; only three of the bottom 10 counties are also among the 10 whitest counties, but all but one are in the bottom quarter of white counties.
County |
Black % |
Cochran
primary
vote |
Cochran
runoff
vote |
Cochran
gain |
Cochran
gain % |
McD
primary
vote |
McD
runoff
vote |
McD
gain |
McDaniel
gain % |
George |
8.1 |
1,233 |
967 |
- 266 |
- 21.6 |
2,220 |
2,410 |
190 |
8.6 |
Hancock |
7.0 |
3,631 |
3,100 |
- 531 |
- 14.6 |
3,221 |
3,301 |
80 |
2.5 |
Greene |
25.9 |
418 |
389 |
- 29 |
- 6.9 |
983 |
1,113 |
130 |
13.2 |
Perry |
20.0 |
697 |
667 |
- 30 |
- 4.3 |
1,188 |
1,385 |
197 |
16.6 |
Stone |
19.1 |
1,259 |
1,251 |
- 8 |
- 0.6 |
1,412 |
1,509 |
97 |
6.9 |
Pearl River |
12.3 |
2,140 |
2,136 |
- 4 |
- 0.2 |
4,618 |
5,758 |
1,140 |
24.7 |
Clarke |
34.4 |
926 |
936 |
10 |
1.1 |
1,245 |
1,518 |
273 |
21.9 |
Webster |
19.9 |
815 |
867 |
52 |
6.4 |
758 |
919 |
161 |
21.2 |
Choctaw |
30.1 |
574 |
613 |
39 |
6.8 |
620 |
601 |
- 19 |
- 3.1 |
Harrison |
21.9 |
13,113 |
14,203 |
1,090 |
8.3 |
9,480 |
10,355 |
875 |
9.2 |
If you look at the map of which counties have the lowest percentage of African-Americans, you'll notice that they tend to fall in two corners of the state, in the southeast corner (the core of MS-04) and the northeast corner (the core of MS-01). However, the counties in the southeast corner are the ones where Cochran didn't gain (like George, Greene, Hancock, and Pearl River). The counties in the northeast corner (like Tishomingo—the state's whitest county, at 3 percent black—Itawamba, and Alcorn) don't follow the runoff's pattern, though. In the MS-01 counties, as you can see from the left map, Cochran gained votes at a 20 to 25 percent rate, pretty typical of the state as a whole. This divergence among counties at the white end of the slope shows why there isn't a terribly strong correlation between Cochran gain and black percentage (0.19), even though it shows up quite well on the map.
In other words, Cochran has some regional appeal in the northeast that he doesn't in the southeast. It's possible that the northeast, which is considered the "Appalachian" part of the state, has a different attitude toward public works than the southeast, maybe thanks to the influence of the Tennessee Valley Authority; however, the southeast certainly sees its share of federal dollars as well, with a lot of Navy shipbuilding activity there. Alternatively, it may be that McDaniel had some unique appeal in the southeast; he's from Jones County, which is in MS-04, though it's further north and not in the stubby little panhandle area, where the counties where Cochran gained the least are found.
Now that Cochran no doubt knows that he eked out one last term thanks to his African-American constituents, does he start becoming more attentive to their needs (if not out of gratitude, then at least out of an enhanced awareness of their existence)? Here's to hoping that he does, although he might feel that his recompense is to keep on whittling down the government at a much slower pace than the Chris McDaniels of the world would have liked to.