What explains differences in the margin of victory in Florida legislative races? Data is collected for all 160 legislative races from 2016 (120 house contests and 40 senate elections) and two sets of analyses are conducted: all races and just competitive races. When looking at all races, five variables are statistically significant in the multivariate model. District party registration difference has a positive effect on margin of victory and the greatest relative impact. Races involving two major party candidates have a much smaller margin of victory, a contest with a major party versus a minor party (or No Party Affiliation candidate) has a somewhat smaller margin of victory, and a race involving only a write-in candidate as an opponent has a wider margin of victory. House races had somewhat smaller margins of victory compared to senate races. When examining just competitive races (contests that had at least two candidate names listed for an office) only two variables were statistically significant in the multivariate model. District party registration difference had a positive relationship with margin of victory and races involving a minor party candidate as the main challenger had higher margins than contests between two major party candidates.

I examine the relationship between the electoral security of congressmen, measured as vote margins in the previous election, and the support Members of Congress offer to their party. I develop a theory that predicts safe members will be more willing to support than vulnerable members and leaders demand more loyalty from safe members than vulnerable. This arrangement is rational and beneficial for leaders and both types of members. Using an OLS regression, I find basic support for my theory.

I examine the relationship between the electoral security of congressmen, measured as vote margins in the previous election, and the support Members of Congress offer to their party. I develop a theory that predicts safe members will be more willing to support than vulnerable members and leaders demand more loyalty from safe members than vulnerable. This arrangement is rational and beneficial for leaders and both types of members. Using an OLS regression, I find basic support for my theory.