Election analyses make leaps of logic

By now, you've probably read through myriad election analyses that mostly attempt to guess the motivation of voters without every really asking them.

Ah, the life of a pundit.

Some, of course, have numbers associated with them and start out as solid pieces of research but then often venture into guessing games.

The analysis, for example, that Obama won 70 percent of the Latino vote and that helped push him over the top is pretty spot on. The numbers can be shown to reveal that. But why Latinos voted for Obama may be a little less certain.

Some of it makes sense. Romney had to start so far right in the primaries on immigration that he couldn't recover to a moderate position even after that. Several national columnists have made that case. At the same time Obama offered the reduction in immigration enforcement plan a few months before the election that also may have looked good to Latino voters.

The biggest stretches in election analysis usually come from winners of elections who claim a "mandate" from the voters are based on just about anything they've ever said in their political careers.

These winners assume a variety of crazy things that would drive your average empiricist over the edge of their regression equation.

Just two years ago, Minnesota Republicans claimed all kinds of mandates from the voters -- one being their win was a mandate to not raise taxes. So how can one explain that they didn't raise taxes and they still lost. The voters must have pulled back their no-tax mandate.

In reality, voters cast their ballot for any number of wide ranging reasons, some, maybe many, may have very little to do with any policy. Maybe they just don't like how one party talked, or that one politician seemed too arrogant.

Democrats who defeated Republicans would do well not to draw any similar mandates -- for example, that voters cast their ballots because they wanted taxes raised on the wealthy.

Voting motivation can be much more complicated or much more simple than political leaders may think.

The most accurate conclusion on cause and effect in voting: You will never be able to prove one.