If you like wonky political analysis, Dan “Baseball Crank” McLaughlin has an excruciatingly detailed look at what the historical record portends for the 2016 political landscape.
McLaughlin’s conclusion based on evaluating every presidential election since the American Civil War is that a party that enjoys presidential incumbency for eight straight years almost always sees a drop-off in support in the following election. While this isn’t an earth-shattering result–almost all political analysts recognize that after two successful elections, it is very hard to win a third–McLaughlin shows that the mathematics behind this phenomenon are nearly unanimous. Even when parties do win third elections (FDR in 1940, and George H W Bush in 1988, for example), there is a pronounced drop-off in the amount of support that the incumbent party gets compared to the previous election’s results.
While I urge you to read the article, that’s not what this post is about. Instead, I saw something in one of McLaughlin’s charts that has caused me to question a bit of conventional wisdom that I have always before accepted.
Conventional wisdom holds that higher voter turnout favors Democrats over Republicans. So take a look at this chart from McLaughlin’s analysis. What this shows is the total presidential vote count for each party as well as the total number of voting eligible age non-voters from 1980 to 2012.Look at the blue line for Democrats. Aside from 2008 it is remarkably steady. So steady, in fact, that with an R-squared of .978, a Democratic candidate for president can expect to receive 3.87 million more votes than the Democratic nominee did four years before. Again, aside from 2008, in the last ten presidential elections, the Democratic nominee never over or underperformed this projection by more than 2.2 million votes. I’ve reproduced just the Republican and Democratic data below and added the linear trend that is fitted to 1980-2012 Democratic vote totals except for the outlier year of 2008.
In 2008 Barack Obama overperformed the trend line by an astounding 8.5 million votes as he turned out an anomalously large number of youth, minorities, and independents. Four years later, Obama overperformed the Democratic trend by a mere 1.1 million votes, a much more historically normal result.
Now look at the red line for the number of Republican votes. It is all over the place. In 1984 and 1988, when Democratic performance was almost exactly in line with the historical trend, Republicans outdistanced them by 17 million and 7 million votes. In 1992 and 1996, when Democrats only slightly underperformed their trend, Republicans fell woefully short by 6 million and 8 million votes.
Wondering how far back the trend line goes, I plotted the same data for all post-war presidential elections. Prior to 1980 there is more Democratic variability than in subsequent years, but the extreme linearity of the trend disappears at least before 1972. (If we consider, back to 1972, the 1976 election joins the 2008 election as an outlier. These two elections, Carter’s post-Watergate election and the first election with a black man on the ballot, might indicate that modern Democrats have very little independent or cross-over appeal, and that when it does occur, it happens only as a result of an uncommon circumstance.)
What’s interesting about the period between 1972 and 1980 when this pattern appears to have established itself is that this was the time when the Democratic Party completed its realignment. Previously it had been a coalition of geographies: white Southerners and urban Northern Catholics. Bouncing in and out of the Democratic coalition before then were Midwestern farmers, Northern Protestants, and blacks.
What has changed since is that the Democratic coalition is now almost entirely demographically based and includes nearly all black voters, a solid majority of Hispanics, and a steady percentage of women. Aside from these groups, there is very little bouncing in and out of the Democratic coalition since the 1970s. Again, with the exception of 2008 when Barack Obama greatly expanded his party’s vote before it collapsed back to its normal self four years later, the only growth Democrats have seen over the last four decades looks to be entirely dependent on the population growth of its constituencies.
Of course, many Democratic pundits and strategists would take comfort from such a conclusion. After all, at least one portion of the Democratic triad (Hispanics) is growing faster than the general rate of population growth. (McLaughlin addresses this point in another post here.)
On the other hand, the record of the last ten elections indicates that Republicans have far more upside potential, having bested the Democratic trend line by five million or more votes on four occasions. Democrats have never beaten that same trend by even half that amount except for once.
Of course, Republicans have also far underperformed the Democratic trendline. In 1992 and then again in 1996 voters who might have voted Republican either didn’t vote or voted for H. Ross Perot, and in 2012 a milquetoast Mitt Romney couldn’t even manage the Republican vote total amassed eight years before when the country’s population was nearly 20 million people fewer.
Sean Trende has made this point before:
” . . . census data and exit polls reveal that some 6 million white voters opted to sit out [the 2012] election. The data show these non-voters were not primarily Southerners or evangelicals, but were located in Northeast, Midwest and Southwest. Mainly, they fit the profile of “Reagan Democrats” or, more recently, a Ross Perot supporter. For these no-shows, Mitt Romney was not a natural fit.”
Less hitched to demography, Republicans are a coalition of ideas that in some years have more or less overlap and appeal than in other years and are more or less represented by a particular nominee.
In other words, the electorate includes really only two types of people: reliable Democrats who always vote no matter the nominee, and potential voters who will almost never vote Democrat but might vote Republican if their nominee can persuade them. Furthermore the number of the latter vastly exceeds the former. While in most years Democrats achieve a very predictable result based almost entirely on the mechanics of population growth, this raw political landscape allows Republicans the opportunity to enjoy earth-shattering landslides when it chooses well, or to suffer soul-crushing defeats when it does not.
And that brings us back to the question of turnout. The conventional wisdom is that because the Democrat’s voter base includes groups with historically lower turnout rates, they are the ones to benefit from higher turnout. However, the record of the last forty years places nearly the entire burden for victory on the ability of the GOP presidential nominee to excite that portion of the electorate that was never going to show up to vote for the Democrat anyway.