Lots of people have shared a lot of analyses on Tuesday’s elections, much of it on a national scale. They try to make guesses about what the elections say about Obama, or the Republican agenda, on certain subgroups of voters, and so on. I don’t have any great insights into any of these things. I am, however, interested in the flip of the Minnesota House to the Republicans. And while I can’t offer a lot of insight into the why, I do have numbers that appear to tell a tale of turnout. If the question is whether the Republicans won their races on Tuesday due to a higher Republican turnout, a great switch of voters from the Democratic column to the Republican column, or Democratic voters not showing up, it appears that the latter issue may be the important one.
I’ve collected data from the 2010 elections and the previous midterm elections in 2006, as the 2006 election is most directly comparable to this year (the high-turnout presidential election makes things tough to analyze). This results in the following, data-filled table:
| 2006 Totals | 2010 Totals | |||||||||||
| District | GOP | DFL | Total | GOP | DFL | Total | GOP Raw Increase | GOP % Increase | DFL Raw Increase | DFL Decrease | Turnout Change | DFL Decrease>GOP Increase? |
| 1A | 6495 | 7632 | 14127 | 8119 | 5707 | 13826 | 1624 | 25.00% | -1925 | -25.22% | -2.13% | Yes |
| 1B | 6139 | 8191 | 14330 | 6528 | 6397 | 12925 | 389 | 6.34% | -1794 | -21.90% | -9.80% | Yes |
| 2B | 7859 | 9228 | 17087 | 8448 | 7668 | 16116 | 589 | 7.49% | -1560 | -16.91% | -5.68% | Yes |
| 3B | 0 | 13282 | 13282 | 8519 | 8107 | 16626 | 8519 | N/A | -5175 | -38.96% | 25.18% | N/A |
| 8B | 7499 | 8230 | 15729 | 8673 | 6786 | 15459 | 1174 | 15.66% | -1444 | -17.55% | -1.72% | Yes |
| 12B | 8033 | 8615 | 16648 | 8940 | 6721 | 15661 | 907 | 11.29% | -1894 | -21.98% | -5.93% | Yes |
| 13B | 7163 | 8524 | 15687 | 7998 | 7197 | 15195 | 835 | 11.66% | -1327 | -15.57% | -3.14% | Yes |
| 16A | 8752 | 8293 | 17045 | 9319 | 7468 | 16787 | 567 | 6.48% | -825 | -9.95% | -1.51% | Yes |
| 25B | 9233 | 9293 | 18526 | 8898 | 8867 | 17765 | -335 | -3.63% | -426 | -4.58% | -4.11% | Yes |
| 27A | 8332 | 8617 | 16949 | 7509 | 7451 | 14960 | -823 | -9.88% | -1166 | -13.53% | -11.74% | Yes |
| 30B | 8928 | 9634 | 18562 | 9711 | 8823 | 18534 | 783 | 8.77% | -811 | -8.42% | -0.15% | Yes |
| 37B | 14143 | 0 | 14143 | 10910 | 7844 | 18754 | -3233 | -22.86% | 7844 | N/A | 32.60% | N/A |
| 38A | 7544 | 7601 | 15145 | 7606 | 6829 | 14435 | 62 | 0.82% | -772 | -10.16% | -4.69% | Yes |
| 38B | 8282 | 8119 | 16401 | 8323 | 7680 | 16003 | 41 | 0.50% | -439 | -5.41% | -2.43% | Yes |
| 40A | 6548 | 7457 | 14005 | 6675 | 6203 | 12878 | 127 | 1.94% | -1254 | -16.82% | -8.05% | Yes |
| 41B | 9340 | 8147 | 17487 | 8750 | 7887 | 16637 | -590 | -6.32% | -260 | -3.19% | -4.86% | No |
| 42A | 8131 | 9963 | 18094 | 8689 | 8582 | 17271 | 558 | 6.86% | -1381 | -13.86% | -4.55% | Yes |
| 49B | 7919 | 6637 | 14556 | 7171 | 6760 | 13931 | -748 | -9.45% | 123 | 1.85% | -4.29% | No |
| 53A | 9141 | 9192 | 18333 | 10130 | 7851 | 17981 | 989 | 10.82% | -1341 | -14.59% | -1.92% | Yes |
| 56A | 8970 | 9214 | 18184 | 9166 | 8259 | 17425 | 196 | 2.19% | -955 | -10.36% | -4.17% | Yes |
| 56B | 9591 | 10039 | 19630 | 10778 | 9494 | 20272 | 1187 | 12.38% | -545 | -5.43% | 3.27% | No |
| Averages | 8002 | 8377 | 16379 | 8612 | 7551 | 16164 | 396 | 3.80% | -1052 | -13.63% | -4.08% | |
There’s a lot of data here, so here’s the rundown: You have 2006 totals, then 2010 totals, then a comparison of the raw vote changes, percentage change, turnout change, and whether the DFL vote drop is greater than the GOP increase. This data is for the house seats where the DFL incumbent lost this year.
One thing that makes the data directly comparable is the fact that many (but not all) of the DFL incumbents that lost were first elected in 2006, so for most of the districts, the DFL candidate that ran in 2006 also ran in 2010, making direct comparisons possible. In addition, of the few races in 2006 in this table that the DFL candidate did not win, such as in 16A, the DFL candidate ran again in 2008 and won. There are a few anomalies in this table, such as the two races that had no opposition in 2006 (3B and 37B), but by and large, this table is a pretty good comparison of how DFL candidates performed in 2006 as opposed to 2010.
Several things jump out. First, turnout was lower than in 2006. Second, the Republican candidate gained an average of slightly less than 400 votes from 2006 to 2010. But most importantly, the DFL candidate lost an average of over 1,000 votes (I removed the no-candidate races from these calculations). This DFL dropoff is especially evident in races in 38A, 38B, 40A, and 56A: in those races, the Republican candidate barely improved from 2006, but the DFL candidate lost substantial support. In fact, in the majority of races, the DFL dropoff was greater than the increase in GOP support.
What does this mean? Mainly, that the overall move from Democratic to Republican support by voters can’t entirely be explained by an outright switch in most of these races. Some voters undoubtedly did change their support, but not enough to explain why the Democratic incumbent lost. Instead, it seems clear that the enthusiasm gap was certainly evident in many close races this year: Democrats did not turn out like they have before to vote for their candidates. Had the dropoff not occurred, then many of the incumbents that lost Tuesday would have won.
Why did this dropoff occur? That’s the question. The economy probably has something to do with it. In any case, though, people running for office in two years will be looking to answer this question, and prevent a repeat.