Conflict of interest? No, can’t be.

I recommend reading, Does Government Dependence Influence Voting Behavior? by David Waciski.

I’m always skeptical of data and statistics, even those that confirm biases. But, I found this interesting.

First, Waciski starts with a great quote from Alexis de Tocqueville:

The American Republic will endure until the day Congress discovers that it can bribe the public with the public’s money.

Later, he writes why looking at state level voting data to determine if voting is influenced by government aid can be misleading. Since there are lots of reasons why people vote, looking at too big of groups of people to answer the question, assumes that the noise caused by all the other reasons can be ignored. It can’t.

My favorite example of this aggregation phenomenon comes from academic research. Statistics always shows that individual-level data (like parent status or number of books in home) explains variations in individual student academic performance better than aggregated data like schools or school districts. No duh. Even in ‘good schools’, there are A students and F students, so just about any data that exists at an individual level will naturally have more individual level variations than any data that exists at a group level.

To picture this, imagine art with Legos. The smaller the Legos, the better the artist is at being able to reproduce the recognizable variations in their subject.

Finally, Waciski concludes from looking at precinct level data, instead of state level (smaller Legos), on voting and government aid:

On average, every one percent increase in the number of households receiving federal assistance resulted in a two percent increase in the vote received by President Obama.

I think “…resulted in…” is a bit strong here. He has found a relation between two sets of data, but that isn’t enough to establish cause. Though, it is enough to say that cause is a possible explanation for the relation.

And, I think it’s a probable explanation, because it is easy to identify the mechanism that might convey the cause: 

Unsurprisingly to those who understand human nature, the recipients of federal aid have a strong propensity to support the politicians who provide that aid.

In just about any other walk of life, most people aren’t surprised when other people act in their own self-interest. Yet, somehow, it couldn’t possibly be when it comes to voting?