Thursday, February 25, 2010

Wow, its been a long time with no posts

It's been months and much has happened. I will give a quite update on a few things and then hopefully post something more detailed later.

Life moves on here in Uganda. Things at my office are tough because the organization still doesn't have funds so the project is on hold at the moment. However, my bosses in the states have gotten me involved in a few other related projects that are, I think, pretty exciting. One is testing how the scorecard impacts voters' opinions of their MPs (members of parliament). We have data from a survey experiment that we are using to analyze this relationship. In some (randomly selected) surveys, people were asked their opinion of their MP and then shown the scorecard for their MP and then asked again, while others were asked without seeing the scorecard. We want to see if the scorecard (given performance of your MP: high, low, or moderate) changes one's opinion of the MP in question.

The second project is trying to identify policy positions of each of Uganda's MPs. There is literature out there that says African politicians don't have policy positions (patronage and clientelism play a larger role than actual policy). We are trying to do this as objectively as possible. Therefore, we are using what each MP says in Parliamentary sessions to form their policy positions. We have a program that counts the words used by each MP and then we will compare this to "reference texts" in order to see how that MP's language compares to the benchmarks. Maybe and example will help: if we are trying to find how pro government of pro oppposition an MP is, we will use the government and opposition leaders contributions in parliament as the reference texts and compare the word frequencies to see where on the spectrum that MP sits. This is pretty cool because it take the human bias out of it without being meaningless; usually things like this are done by people evaluating speaches etc but if you are evaluating a speach and know which party the person belongs to, then you are already biased to put them more on that side of the scale. This program simply treats the text as data. One other scale we are going to try to do, which i think is really interesting given current Uganda politics is pro or anti homosexuality: there is a current bill up for debate called the Anti-Homosexuality Bill of 2009. And it is a disaster and I feel is a violation of Homosexuals' human rights (for more reading on that see my friend Avner's blog:; I would rather you read his blog than for me to write a post about it, I would just repeat him anway). It would be pretty exciting if we can develop these and other policy position scores and would give great incite into Ugandan politics!