Sabato's Crystal Ball

Demographics Overtakes Economy as Prime Presidential Election Indicator

Geoffrey Skelley, Political Analyst, U.Va. Center for Politics April 25th, 2013

One of the questions we asked prior to the 2012 election was whether or not state-level unemployment figures would matter much on Election Day. As it turned out, the answer was “probably not much.”

Throughout the 2012 election cycle, the Bureau of Labor Statistics’ monthly jobs report became a highly anticipated event. Politicians, journalists and election handicappers (the Crystal Ball included) waited eagerly every first Friday of the month, anticipating what the national unemployment number would say about President Obama’s reelection chances. Then, toward the end of each month, the BLS released the state-by-state numbers, sparking further debate about the relative importance of state numbers, particularly in the swing states.

In hindsight, the relentless focus on jobs numbers, at least in the states, was overwrought. Political science research indicates that while the national economic situation influences presidential races, the state-by-state rates do not appear to affect the outcome all that much. That the states swung relatively uniformly in their election results from 2008 to 2012, regardless of the changes in employment numbers, gives even more credence to the argument that individual state-level employment data are relatively unimportant.

Just look at the case of Nevada. As Chart 1 shows, Friday’s state jobless report finds Nevada has the highest unemployment rate in the nation. From the September 2012 data, one can see that the Silver State also held this dubious honor when the last pre-election state-level report was released in October 2012. Yet despite Nevada’s poor economic condition, President Obama won its six electoral votes by a little less than seven percentage points. If they voted today, Nevadans would likely choose the president once again. Something else seems to matter more than unemployment right now: demographics.

Chart 1: State-by-state unemployment rates, September 2012 and March 2013

Source: Bureau of Labor Statistics

As in many other states, Nevada would have supported Mitt Romney if only white voters had had their way last November. In the Silver State, whites supported Romney 57%-41% over Obama — but whites only made up 64% of Nevada’s electorate. Conversely, nearly three-fourths of the 36% of the Nevada electorate that wasn’t white supported Obama, assuring the incumbent of victory there. As we discussed last week, only two swing states’ white voters supported Obama over Romney: Iowa and New Hampshire. In all the other key battleground states that he won, minority voters made the difference for Obama, overcoming margins as large as 61%-37% for Romney among Florida’s white voters.

As long as the major parties are worlds apart in terms of the diversity of their respective bases, the “demographics as destiny” storyline will prove to be substantially true. While the economy will remain a vital factor in presidential elections, with the national unemployment rate playing a major role in driving public opinion, the demographics of individual states may have as much, or greater, impact on the accumulation of Electoral College votes.