The "Party Convention Effect" on Presidential Election Politics

Yesterday, Republican National Committee announced Cleveland, Ohio as its choice for holding 2016 Republican National Convention. As Ohio holds the record of longer streak (Since 1960) of picking the right president than any other state in the Union, the political calculus behind this decision is clear: to sway the voters of this bellwether state by holding the most energizing event of the presidential election campaign in the state. Republican Party has the streak of losing the states they held national conventions in since 1992 when the incumbent Republican president won the convention state but lost the national election. Picking the 2016 runner-up, Dallas could have easily ended the 20-year losing streak as Texas has been solidly red but will not move the electoral vote equilibrium one way or the other.

The premise of RNC's rationale is that by holding the quadrennial event that is essentially the climax of presidential election cycle in the state, a center of organization, resources and media attention would be formed around the convention city and state to energize the party base in the state to turn out to vote.

Prior research studies on subject of party convention have agreed that the “fuzz” generated by party conventions would give convention holding parties “bumps” in national polling numbers in the immediate aftermath. However, how much of that enthusiasm is materialized into votes specifically in convention-holding states is still very much an unexplored territory.    

  

Both major parties have the tendency to hold their national conventions in swing states in recent presidential elections. Republicans held the convention in Saint Paul, Minnesota in 2008 and Tampa, Florida in 2012. Democrats held the 2008 convention in Denver, Colorado and 2012 convention in Charlotte, North Carolina. Republicans lose both of those states they held conventions in and Democrats won in 2008 but lost the convention state in 2012. To examine this “Convention Effect”, the ability of party conventions to sway the voters in convention–holding states that it, simply looking at whether the party actually win the state might not be as telling as looking at the percentage of votes that is actually statistically “swayed” toward a specific party.

To examine the extent to which party conventions “swayed” the votes or the “Convention Effect” so to speak, I have chosen the 3 most recent presidential elections, the 3 presidential elections in which their preceding ones did not have none-major-party candidates gathering more than 5% of the votes. The extent of the “sway” is calculated by comparing gap of the convention-holding states’ popular vote shares for the convention-holding with national vote share to this same gap in the previous presidential election. For example, Republicans held their 2012 convention in Tampa, Florida. In 2012 Presidential Election, their vote share in Florida was 49.13% and 47.20% in national popular vote generating a +1.93% “gap”. In 2008, when Florida was not a convention-holding state, GOP’s vote share in Florida was 48.10% in 45.65% nationally resulting in +2.45% “gap”. The “sway” of vote shares for Republicans for holding convention in Florida is therefore +2.45% -- +1.93% = -0.53%.  This means that, statistically, the effect of holding convention in Florida did not sway voters in favor of Republicans. In contrary, the Florida’s Republican “lean” dwindled by half a percentage point from 2008 to 2012.

Republican
Year	State	State %	National %	Gap	Previous   Gap*	Sway
2004	New York	40.08%	50.73%	-10.65%	-12.64%	+1.99 %
2008	Minnesota	43.82%	45.65%	-1.83%	-3.12%	+1.29 %
2012	Florida	49.13%	47.20%	+1.93%	+2.45%	-0.52 %
Average		44.34%	47.86%	-3.52%	-4.44%	+0.92 %

                       

Democratic
Year	State	State   %	National   %	Gap	Previous   Gap*	Sway
2004	Massachusetts	61.94%	48.27%	+13.67%	+11.42%	+2.25 %
2008	Colorado	53.66%	52.93%	+0.73%	-1.25%	+1.98 %
2012	North Carolina	48.35%	51.06%	-2.71%	-3.23%	+0.52 %
Average		54.65%	50.75%	+3.90%	+2.31%	+1.58 %

 * Previous Gap refers to convention state's vote share gap in the previous previous presidential election.

  

In 2008 and 2012, both parties' choice of convention state was more conservative than the previous one. In other words, Republicans tended toward picking more " friendly" while Democrats' convention states just got progressively less "friendly" to them. On average statistically, Republicans have "swayed" +1.92% of convention state vote share in their favor while Democrats have "swayed" +1.58%. Both parties have managed to push the convention states to their party's favor in all but one instance. That is Republicans' 2012 Tampa, Florida convention failing to push Florida to be more Republican relative to national vote.

The unique American system of Electoral College in presidential elections renders more importance to strategically allocate resources according to landscape of political geography. In a system that difference of 500 something votes in a specific state could alter the outcome of presidential election regardless of aggregate national votes, partisan strategists would definitely tell their candidate and their party to direct as many resources as possible to those critical states like Ohio. Of those resources at each party's discretion, the choice for location of national convention is probably the one single snap-decision that would have the most profound impact on the outcome. RNC's choice of Cleveland, Ohio to hold its party convention is one, major one step in the right direction toward winning. Since Republicans' have chosen Cleveland, it left Democrats less likely to pick Cleveland as well and more likely to land on New York, a state that is solidly blue but holds special significance to Hillary Clinton and women's rights movement should she be nominated. If Democrats choose New York, it is at the expense of not using this discretion to woo voters in states that could tip the election. Democrats has had very successful record of doing so in 2008 when Colorado was turned blue and in 2012, to a lesser extent as North Carolina was brought closer to national popular vote share, but not enough to keep it in the blue column.