The Washington PostDemocracy Dies in Darkness

Republicans are holding their 2016 convention in Cleveland. Which may mean they will do worse in Ohio.

So, it's settled. The 2016 Republican presidential convention will be held in the great state of Ohio. (Perhaps because the Dallas people were a little too sassy.) Does this mean that the Republicans will win Ohio? Well, if a century of history has taught us anything, it's that the party will likely do worse in the state than it did in 2012.

We'd previously looked at how the three most recent convention locations affected the vote for each party's presidential candidate in the host cities. Between the election preceding the convention and the convention year, the Democrats saw increases in support in their last three convention cities. Results for the Republicans were … mixed.

That is interesting. But we were more curious about the long-term trend and, more importantly, what happened in the electoral-vote-holding states. So we reviewed election results for every convention since 1912 and compared the state's vote percentage for the host party to:

  1. The national percentage for the party.
  2. How that difference compared to the difference in the previous election.

Here are the convention locations, for reference.

Year Democrat city Republican city
2012 Charlotte, NC Tampa, FL
2008 Denver, CO St. Paul, MN
2004 Boston, MA New York, NY
2000 Los Angeles, CA Philadelphia, PA
1996 Chicago, IL San Diego, CA
1992 New York, NY Houston, TX
1988 Atlanta, GA New Orleans, LA
1984 San Francisco, CA Dallas, TX
1980 New York, NY Detroit, MI
1976 New York, NY Kansas City, MO
1972 Miami, FL Miami, FL
1968 Chicago, IL Miami, FL
1964 Atlantic City, NJ San Francisco, CA
1960 Los Angeles, CA Chicago, IL
1956 Chicago, IL San Francisco, CA
1952 Chicago, IL Chicago, IL
1948 Philadelphia, PA Philadelphia, PA
1944 Chicago, IL Chicago, IL
1940 Chicago, IL Philadelphia, PA
1936 Philadelphia, PA Cleveland, OH
1932 Chicago, IL Chicago, IL
1928 Houston, TX Kansas City, MO
1924 New York, NY Cleveland, OH
1920 San Francisco, CA Chicago, IL
1916 St. Louis, MO Chicago, IL
1912 Baltimore, MD Chicago, IL

And now, in graph form, how the party did versus the national percentage in each of those years.

When compared to how their party's candidate did nationally, Republicans actually did substantially better in convention states than the Democrats. Sixteen times, they outperformed the national percentage, versus the Democrats' 15. But more significantly, their candidates did 2.64 percentage points better in convention states, while Democrats only did 0.37 percentage points better.

There's a reason for that. Here's the count for each of the 26 elections we looked at on three key metrics.

Comparison Democrats Republicans
Convention state beats national vote percent 15 (57.7 percent) 16 (61.5 percent)
State beats national percentage in previous election 13 (50 percent) 19 (73.1 percent)
Convention state beats previous election 15 (57.7 percent) 12 (46.2 percent)

That second row is significant. In 19 of the 26 elections, the Republicans already were doing better in the states that hosted their conventions. In other words, they picked hosting locations where they did well.

When you compare national vs. host state in the election year with national vs. host state in the previous election, removing (to some extent) that bias, the picture changes.

On average, Democrats see a 0.23 percentage point jump in states that host their conventions over the previous election. Republicans see a 1.41 percentage point drop. You'll notice some outliers in the graph above -- the result of picking very friendly or hostile states, or skewing from hometown bias. If you take out the outliers (which we set at anything over 20 percentage points off the norm), Dems still did better: 0.58 percentage points to the Republicans' -0.54.

The point is this: Elections are not won or lost with convention picks. But if they were, the Republicans have a lousy track record on where they host their conventions.