If and when Republicans gain a modest number of House and Senate seats in November 2014, the Beltway set may be tempted to interpret the results as a sixth-year itch rebuke of President Obama on everything from IRS/DOJ/Benghazi to rocky implementation of the Affordable Care Act. To do so would be to ignore simple demographic math: likely GOP success in 2014 has much more to do with a shifting electorate than shifting opinions.
In 2012 - according to exit polls - House Democrats won 60 percent of voters ages 18-29, 51 percent of voters ages 30-44, 47 percent of voters ages 45-64, and 44 percent of voters ages 65 and older. That combination was enough to win them a 49 percent to 48 percent plurality of all votes cast for House (even if Republicans still kept a comfortable majority thanks to Democrats' inefficient distribution on the map and redistricting).
But had Democrats won the same levels of support among each age group in 2010, Republicans would still have won a clear plurality of all votes cast that year. How? Voters under the age of 30 were 19 percent of all voters in 2012, but just 12 percent of all voters in 2010. Likewise, voters 65 and up were 17 percent of all voters in 2012, but 21 percent of all voters in 2012. Herein lies the biggest danger for Democratic candidates in 2014.
Midterm elections have always drawn older voters, and usually drawn white voters, to the polls in disproportionate numbers. Older voters are less transient, have grown deeper roots in their local communities, and pay much more attention to non-presidential elections than their younger counterparts. In the 1980s, that didn't hold partisan consequences. Today, that amounts to a built-in midterm turnout advantage for Republicans.
A smart question might then be, what about 1998 and 2006, two midterm elections in which Democrats did quite well? The answer: everything is relative. The "impeachment" midterm of 1998 wasn't necessarily a banner Democratic year; it just wasn't the bad Democratic year history would have predicted a president's sixth year would be. It's easy to forget Republicans kept the House in 1998, and by a bigger margin than they did in 2000.
In 2006, Democrats' 30 seat gain in the House was monstrous compared to the tiny swings of the past five election cycles before it, but it didn't produce a historically large Democratic majority. In 2006, Democrats won Independents 57 percent to 39 percent to win a 233 to 202 seat majority. In 2008, Democrats won Independents by a much smaller 51 percent to 43 percent, but a surge in youth and minority turnout helped produce a massive 257 to 188 seat majority.
Republicans' built-in midterm turnout advantage really began to emerge in the early part of the last decade but has ballooned in the Obama era. That's because partisan voting patterns are more polarized by age and race than they ever have been, and Obama's coalition is more highly dependent on young and non-white voters than any presidential coalition before it. The four charts below explain how this has fostered "boom and bust" outcomes every two years.
Age Breakdown of the House Electorate, 1992-2012
Democratic Share of Two-Party House Vote by Age, 1992-2012
|Year||18-29||65 and Older||Age Gap|
As recently as 2000, House Democrats did just as well with voters over the age of 65 as they did among voters between the ages of 18 to 29. But beginning in 2002, Democrats started performing much better with the youngest voters than the oldest voters, and in both 2010 and 2012, House Democrats performed a whopping 16 points better with 18-29 year olds than voters over 65. This gap spells big 2014 trouble for Democrats running in marginal states and districts.
Racial Breakdown of the House Electorate, 1992-2012
Much as the American electorate has grown older, it has also rapidly grown more diverse and less white. In 1992, white voters comprised 87 percent of the electorate; today they comprise 72 percent. However, only in 2002 did there begin to appear a marked difference in the share of non-white voters between midterm and presidential elections. In both 2006 and 2010, the midterm electorate was three points whiter than the immediately preceding presidential electorate.
Democratic Share of Two-Party House Vote by Race, 1992-2012
Between 1992 and 2012, House Democrats have always done between 30 and 39 percent better with non-white voters than white voters. However, whereas a widening generation gap between the parties has deemed existing generational turnout patterns more dire for Democrats in 2014, the story of race is a mirror image: a widening racial turnout gap between midterms and presidential years has made an existing partisan racial gulf more dire for Democrats in 2014.
Is there any hope that Democrats can find the silver bullet to turn around their midterm turnout woes with minority and younger voters? In truth, the forces that drive more older voters to the polls in midterm years are beyond the control of either party's smartest micro-targeting squad. About the best case scenario Democrats can hope for is to try to make the drop-off among youth and minority voters just a little less dramatic.
There are some efforts underway: as Bloomberg's Joshua Green wrote just today in BusinessWeek, the whiz-bang Obama Big Data geek squad that surgically targeted casual voters to generate a five million vote reelection margin has set up shop as Civis Analytics, catering to both corporate clients and 2014 Democratic candidates. The DCCC has expanded its targeting department, and has reached out to Civis and others for help in turning out "casuals."
But so far, there doesn't exactly appear to be a Democratic "Manhattan Project" geared towards tackling this problem, in part because not all Democrats seem to be focused on how dire it is. As Cook Political Report National Editor Amy Walter wrote two weeks ago, "Organizing for Action, the Obama campaign operation-turned-grassroots organization, is designed to promote the Obama Administration's agenda, not to turn out Democratic voters in 2014."
Are there House Democrats who could actually perform better in 2014 than they did in 2012? Yes, a few. An important distinction between midterm and presidential electorates is that midterm electorates tend to know their incumbents more personally. This is good news for personally popular Democratic Reps. John Barrow (GA-12), Mike McIntyre (NC-07), and Jim Matheson (UT-04), who nearly lost amid 2012 presidential turnout in heavily GOP districts.
But there are also places where Democrats are likely to face especially severe headwinds in 2014. These include highly transient districts where Hispanic and other minority turnout gyrates wildly between presidential and off-years. In fast-growing CA-36, for example, Democrat Raul Ruiz lost to then-GOP Rep. Mary Bono Mack by 16 points in the initial June 2012 race, but won by six points when Hispanic turnout exploded in November.
For most Democratic House candidates, a good rule of thumb might be to subtract two to three points from the 2012 Democratic percentage in the district to come up with a reasonable approximation of a "starting point" for a 2014 race. That means most Democrats probably need to perform about five percent better among Independent voters in 2014 just to stay afloat at 2012 levels, and would need an even higher share of Independents to pick up seats. Wow.
Cook Political Report Web Editor Loren Fulton contributed to this report.
May 9, 2013If Democrats Can't Beat Sanford, Who Can They Beat?
April 25, 2013Sanford Sideshow Aside, Three Questions About 2014
April 15, 2013The Republican Advantage