Projects‎ > ‎

Buying elections: a causal analysis of money in politics

  • Team Sequoia
  • Location: Stanford
  • People: Joshua Loftus, Weijie Su, Xiaoying Tian, Qingyuan Zhao
  • Emails: joAftBCius, suABweiCjie444, catABherinetCxy1991, qinAgyBCzhao respectively, remove ABCs and append gmail
  • Category: Insight
Description: It is easy to see a strong correlation between spending and winning elections, but everybody knows correlation is not causation. Some candidates are just more popular and their popularity could explain both their financial advantage and the election outcome. The field of observational studies attempts to answer questions about causality using only observational data. Our team will apply simple techniques from this field to answer the question: Can I buy an election?

After the latest election there were a lot of news stories focusing on specific wealthy donors who spent a lot of money but their candidates lost anyway: http://openchannel.nbcnews.com/_news/2012/11/08/15001252-money-cant-buy-happiness-or-an-election

But these are individual stories, not statistical inferences. We will try to reach a general, statistically valid answer.

We are starting with electoral data sets from Sunlight, but may need to add others. We are using R, the open source statistical programming language.


ċ
House_of_Representatives_-_2012_Competitive_Election_Dataset.csv
(14k)
joft aloft,
Feb 3, 2013, 11:58 AM
ċ
code.R
(7k)
joft aloft,
Feb 3, 2013, 11:58 AM
ċ
final2010.csv
(5k)
joft aloft,
Feb 3, 2013, 11:59 AM
Ċ
joft aloft,
Feb 3, 2013, 11:57 AM
ċ
sequoia.Rmd
(3k)
joft aloft,
Feb 3, 2013, 11:53 AM
ċ
sequoia.html
(29k)
joft aloft,
Feb 3, 2013, 11:52 AM
Comments