Voting Behavior in the 2008 Election

An instructional resources project sponsored by the APSA, ICPSR, and SETUPS.

Party identification as a possible confounding variable

There is ample evidence that issue positions vary according to party identification, and we might argue that those who consider themselves to be Republicans would generally be more favorable to Social Security privatization, perhaps because they have heard Republican leaders endorse such a proposal, or perhaps because they have a general ideological disposition toward reducing the economic role of government. Republicans, of course, would be very likely to vote for McCain, as we have already seen that the relationship between party identification and voting is very strong. For exactly the same reasons, we might expect that those who consider themselves to be Democrats would generally be opposed to Social Security privatization and also would be likely to vote for Obama. Thus, we must consider the possibility that attitudes on Social Security privatization had no direct effect on the vote in 2008 and that the relationship between these two variable is due entirely to their common connection to party identification.

In this case, we are assuming that voters did not form their party identification on the basis of their attitudes on Social Security privatization. If attitudes on this issue influence or affect party identification, then party identification would not be a confounding variable in this relationship. Instead, it would be an intervening variable, a concept discussed in the next set of exercises. The assumption that attitudes on Social Security privatization have little or no effect on party identification seems reasonable on the basis of what we know about party identification. It is a more basic and stable attitude. While it can be affected by issues, it usually takes highly salient and fundamental issues to change party identification, and Social Security reform did not fall into that category. The more reasonable assumption in this case is that Republicans, who generally have more ideologically conservative dispositions, would be led to favor Social Security privatization on the basis of their established partisan and ideological attitudes, just as Democrats would be led to oppose it.