While the low overall 2014 turnout of 36.4 percent has been widely reported as the lowest percent of voting eligible population since 1942 (Macdonald), the wrong culprit has been identified for the Democrats sub-par performance. That is, low turnout among key Democratic groups like the young, non-white, and non-male voting blocs has been mistakenly targeted with all the blame. Actually, these groups performed better than believed in a difficult year for Dems. Further, poor attitudes about Presidential performance and economic doldrums aren't highly correlated with poor Democratic results. Rather, 2014 can better be seen as a typical Mid-Term election. Democratic performance is startlingly similar to all previous in-party vote levels. This and other exit poll data (presented in a subsequent diary) bode well for future Democratic performance in mid-terms and Presidential elections.
A look at three voting measures of mid-terms shows a more positive outlook for Dems than is portrayed in early analyses. These measures include data for the last nine mid-terms, e.g., 1982 through 2014. The data include turnout (Macdonald Turnout 2014), Democratic versus Republican share of the vote (NYTimes) and exit poll data (Exit Polls: Surveying the American Electorate 1927-=2010, NYTimes). In all cases this analysis computes the mean values for mid-terms in the last 32 years and compares the actual 2014 results to this average to measure how different 2014 votes, voters and partisan splits were from the average, or, the presumed baseline.
Results are summarized in Figure 1.
First, the national turnout rate of 36.4 percent was 3.1 percent below the 39.5 percent average of the last nine mid-terms. In terms of total votes, 82.7 million votes were cast in 2014 out of a voting eligible population of 227 million. If instead 2014 had experienced average mid-term (MT) turnout, 89.7 million votes would have been cast, or, 7,000,000 more than the actual vote. Clues to who these folks were can be found in exit polls and voting results.
According to 2014 national exit polls, the party split for the 82.7 million votes was: Democrat 47%, Republican 53%. Converting this to total votes yields: Democrat 38.8 million, Republican 43.8 million, a 5 million vote victory for Republicans. This partisan split differed from the average of the nine previous mid-terms (MT). This historical average MT percentage vote for Democrats is 50.1 and for Republicans is 49.9. While the relative party performance differs from MT to MT, the overall votes cast show just how balanced the parties are. Since 1982, according to exit polls, Democrats have received 340.4 million votes in MTs and Republicans have received 341.7 million votes. Wow. You can’t get much closer than that!
By simple math (Figure 1 Column 3), applying the average partisan split (50.1D/49.9R) to the average expected turnout (89.7 million votes), we could have expected a slight Democratic victory: 44.9 million Dem votes/44.7 million Rep votes. Overall, the actual Democratic vote was 6.1 million below the average MT performance. The actual Republican vote was only 900,000 below the average MT performance, a much more typical result.
Why the Dems did so poorly can be traced to several factors. There is evidence to support the six-year-itch theory, or, the tendency of voters in the sixth year of a presidency to trend toward the out-of-office party. In this data there were four MTs that were sixth year elections: 1986, 1998, 2006, and 2014. The average out of power party vote for these four MTs was 52.5 percent, which was almost identical to the Republican vote share, 53 percent, in 2014. In fact, this theory can be extended to all MTs since 1982 because the party holding the Presidency has never received a majority of the Federal Congressional vote. The President’s party receives on average 46.7 percent of the MT vote, which was also almost identical to the Democrats 47 percent total in 2014.
There is no clear evidence to support the Presidential approval theory which states that a party’s performance in a MT tends to align with their President’s approval rating. In five of the nine MTs the party holding the presidency finished eights point or more below the Presidential approval rating. In 2014, the Dems did three points better than their President’s approval rating among 2014 MT voters.
There is evidence to contradict the ‘It’s the economy, stupid.’ theory which states that a party’s performance in a MT tends to align with the President’s party if the economy is doing well. In 2014 exit polls voters seemed to agree with objective measures (GDP, employment, and stock market—all rising) of the economy’s improvement. In 2014, 28 percent of voters said that their financial situation was getting better, which matched the MT average for this measure, 28.0. This number was a dramatic improvement over 2010 when only 15 percent of those polled believed the economy was getting better. Additionally, the percent of voters who believed that the economy was getting worse, 25 percent, substantially below the MT average of 41.6 percent. In fact, this percentage was the lowest recorded in any exit poll. Therefore, even though voters had resoundingly positive views about the economic status and prognosis, Democrats received little credit. Out of the 25 percent who believed the economy was improving, 42 percent voted for the Republican candidate. So, it wasn’t the economy.
In conclusion, the aggregate exit poll data provides evidence that the 2014 Republican victory was typical of both elections in the sixth year of a presidency and Mid-term elections in general. The belief that the 2014 results were an unusual indictment of Presidential performance or economic trends have little support in the actual exit poll data. 2014 was not a wave election. 2014 was an ebbing tide, e.g., cyclical, in decline and ready for its successor.