As far as I can remember, the issue of whether the Democrats in Florida should adopt issues on the right or left of the political spectrum has been prominent. I have even held viewpoints advocating both a move to the right, and more recently a move to the left (which I somewhat take back). But what if issues do not determine vote choice? If we look at the Columbia and Michigan models of voting behavior, they advocate that issues really do not matter, and that other factors really determine vote choice. In the Michigan model, partisanship is adopted from a voter’s parents. However, in the ever-changing state of Florida, the Michigan model might hold true. Even the Columbia model, which looks at socioeconomic factors, might not fully explain Florida voters. Purely, Florida is one of the trickiest places to run an election.
One of the authors of the Michigan model, Donald E. Stokes, came up with another approach to voting behavior in 1963, and created the valence model of voting. Valence voting doesn’t really look at left-right issues (though it can), but instead how voters evaluate each of the parties competing and make their vote choice based on those evaluation. Non-political factors can play a role in valence voting. For example, if we look at the 1984 and 2004 elections, the idea of patriotism played a role in the election of Ronald Reagan and George W. Bush.
Stokes did not test his theory in his 1963 article, titled Spatial Models of Party Competition, but a few decades later, Harold Clarke et al. (2004) did test this valence model in their book Political Choice in Britain. The authors found that valence voting did have an impact when it came to vote choice in Britain, and that voters who might not have agreed politically with a candidate, such as Margaret Thatcher, went ahead and voted for her anyway because they saw her are more competent than her opponents. This could also possibly explain the concept of Reagan Democrats and Republicans that voted for Bill Clinton in 1992.
Of course, this model relies on a number of things, one of those being party leader. In the case of Florida, we will treat the gubernatorial candidate as the party leader (since it does not apply in the traditional sense as it does in Britain or Canada). When a leader is seen as being competent, then voters in the valence model are more likely to favor that candidate. But, again, voting is an act of choosing one party over another. In this case, the opposition party has to be seen as less competent.
With that, let us bring in the 2014 governor’s race in Florida. Rick Scott was not the most popular of governors. However, was Charlie Crist seen as being less competent? After leaving the 2010 Republican primary for Senate, then running as an independent in the general election, only to then switch to the Democratic Party to run for governor in 2014, Crist’s credibility might have been damaged, and voters were less likely to vote for Crist. Also, while left-right issues might not really matter in this model, Crist’s deviation from his viewpoint on issues while a Republican might have led to additional questions regarding credibility. Because of these issues, Crist might not have been the best candidate for the Democratic Party, as voters see him as being less competent, even compared to Rick Scott.
Using the valence model, we can create some hypotheses which would argue that the competency issue might have played a role in this race (though testing requiring survey research would be needed). And if we look at the previous statewide races, we might observe the same thing. Democratic candidates for statewide office have usually been weak, which makes the Republican candidates look more competent, purely on the basis that voters are unaware of the Democratic candidates.
But what about the elections for President?
As has been mentioned elsewhere, Bill Clinton won the state in 1996, and Barack Obama won Florida twice. However, the difference is that these candidates are running against national Republican candidates and parties and not local candidates. In 1996, voters were happy with the direction of the country under Bill Clinton, which made him look like a more competent choice than Bob Dole. As for Barack Obama, McCain’s mistakes, such as picking Sarah Palin, might have led to voters seeing Obama as a more competent choice. As far as Mitt Romney in 2012, the former Massachusetts governor always had an issue with credibility regarding his views on the issues. In 2004, voters saw George W. Bush as more competent than John Kerry, which was aided by the Swiftboat Veterans for Truth commercials.
Looking at the valence model, where we remove left-right issue positioning, we see why Democrats continue to perform poorly on the state level, yet are able to succeed on the national level in Florida. Florida Democrats have not been able to recruit the type of candidates that are viewed as more credible than their Republican opponents. However, this is not a problem for the national Democrats. Therefore, whether the Democrats move to the right or to the left of the spatial model of voting behavior really will not matter…at least not in this model.
Next week, I will discuss how issues might have an impact on vote choice, and how Florida Democrats continue to have a problem. I will continue to argue that it does not matter whether the Democrats move further right or left, but how people view the Democrats and Republicans on certain issues, and further explore the differences between the state and national Democrats.