Saturday, February 16, 2008

Memory as art

I saw something recently that reminded me of an earlier post (Your Brain Is Not a Closet). I attended a gallery opening at BUIA in Chelsea. Artist Eve Tremblay’s show was called “Becoming Fahrenheit 451”. Ms. Tremlay actually recited Ray Bradbury’s Fahrenheit 451 from memory (ok, with some starts and stops, but still, what an amazing accomplishment). Like the characters in the novel who memorize books, Ms. Twombley reminded us of how powerful memory is if we develop it.

Wednesday, January 9, 2008

Does greater spending produce more votes?

After the first caucus, in Iowa, comments were made to the effect of: greater spending does not equate to greater political popularity. At the time, Huckabee's popularity and low spending seemed to show that this was true.

And shouldn't it be a positive message greater spending did not equate to greater political popularity? If the claim is true, then it would be difficult to, in effect, buy votes.

So I took a look at the numbers and ran a regression for spending before the Iowa vote against the percentage of votes received. I found an R-square of less than 1%. In other words, in Iowa spending had almost nothing to do with percentage of votes gained.

I repeated the exercise for the New Hampshire caucus. Here results were wildly different and a regression run on spending up to that day showed an R-square of over 63%, a relatively strong relationship. Had things changed?

I know there are many problems with this study. First, I didn't have perfect spending data. My data for the Iowa caucus was actually the dollar amount of what was spent on TV ads in the state, while my data for New Hampshire was for total spending up to that time. I'm sure better data is out there, I just haven't found it yet.

My Iowa data set was also only for the top 5 candidates while I had data for the top 10 in the New Hampshire study. (All data are from NY Times: http://politics.nytimes.com/election-guide/2008/finances/index.html)

However I found that if you take Huckabee out of the equation in Iowa, the R-square goes up to over 27%. I would expect that early on, fringe candidates are able to gain more using less spending, while over time the larger spenders rise to the top.

We'll see what happens. I'll post periodic updates. And if anyone knows a better data source please let me know.