This is very important, yet I have not seen anyone comment on the implications of an overrepresentation of Democrats. I have seen some people dismiss the finding that moral issues were the most important issue for a plurality of voters (22%), of whom 80% voted for Bush, simply because if the exit polls were wrong about Kerry winning, they cannot be trusted at all. However, there is no reason to assume people lied in these polls as there is no motive in doing so. Therefore let us postulate two assumptions. First, Democrats were overrepresented in the exit polls. Second, people were honest in their answers. I believe both assumptions are reasonable and supported by the available evidence. What does this mean?
It means that American voters, in terms of the number of actual voters, actually care more about moral issues and other conservative issues than the exit polls indicate. In other words, those that dismiss these findings because the exit polls did not accurately predict the winner, are making a huge mistake. Instead of throwing away the data, they should realize that the data accurately captures those polled. The exit polls did a great job capturing the differences between Bush and Kerry voters; its only flaw was in polling more Democrats than Republicans. Consumer behaviorists know attitudes change slowly, if at all, and any politician who ignores the only poll of actual voters is making a big mistake.
One caveat to this discussion, I have not seen the margin of error on the exit polls. An alternative explanation is that since this was a close election, the polls' results were within the margin of error. Given some of the exit polls I saw mentioned on election night (such as Bush losing Pennsylvania by 20 percentage points), I find this unlikely. However, if the polls results did fall within the margin of error, this would simply mean the sample actually represented the voting population and thus we should believe it. No matter which explanation is correct, the exit polls have provided very valuable and accurate data about the differences between Republican and Democratic voters.
Dean,
I agree. First of all, people drawing obviously false conclusions from data generally bothers me, even outside of any political context.
Politically, as a socially conservative independent, I have really mixed feelings. On one hand, so long as the Looney Left controls the Democratic Party, it is probably best if they continue to lose power through grossly misunderstanding the US people. On the other hand, I think it best for the country to have two strong parties who use different perspectives in a way to best serve their constituents.
Right now I expect the Democrats to continue to lose power for a few more elections. Then my crystal ball gets really hazy. The Dems will then either change and start regaining power or go the way of the Whigs. If the latter, I suspect the Republicans will eventually split into two parties (Libertarians and Social Conservatives by whatever names they pick – one of them will obviously keep the Republican mantle).
You know, if I were a highly loyal Republican I would have a hard time not taking glee in these exit polling discussions. Because anyone who has much understanding of statistical analysis knows that all the bloviating about how these "moral values" people were gay-haters and rabid fundamentalists and such only hurts Democrats. Not only is it insulting to many voters who might vote Democratic but it's also a good way for Democrats to scapegoat their losses rather than figure out ways to persuade more voters to support them.
Some Democrats obviously get it but the hysterics are sure drowning them out right nwo.