I’ve had an idea for a couple of years for a project. The project would map the relationships between elected officials, their donors, and the issues that both care about.
So it would capture data on four distinct entities:
And then it would show the relationships between them.
Data capture might look like this:
Top issues (listed as pro-$SOMETHING, con-$SOMETHING)
Votes (counters for con-GUN, pro-LIFE, con-drilling, etc. based on voting record (see below)
Top donations (maps to Politicians)
Top Politicians that support them
Top Donors that support them
Name of legislation and goal, e.g., Markson-Billey(Con-GUN)
The Politicians that voted for (increments their counter for that)
The Politicians that voted against (increments their counter for that)
Connecting the dots
From here we ask questions:
Unsupervised Learning — Security, Tech, and AI in 10 minutes…
Get a weekly breakdown of what's happening in security and tech—and why it matters.
For a given Donor, what are their top Issues?
For each Donor, who are their top Politicians (that they give to)?
For each Politician, what are their top Donors?
For each Issue, what are the top Politicians and Donors that support it?
Do Politicians support the issues that they receive the most money on?
The next thing to track is voting, which is trickier. You basically have to make a new table called Votes, and then tag them as pro or con for a given issue. Then you note which side each Politician fell on.
Of course, to some degree this should be expected. But if the votes always follow the money you know you have an issue.
Ideally there would be a composite rating that could be built from a number of things, such as money taken in, percentage of similarity between donor goals and the politician’s goals, etc.
# Sample rating
Corruption Rating: 74% (Compromised)
There would be flaws in it, of course, and there would be infinite opportunity to improve it by removing subjectivity, but even a basic version could shine some much needed light on things.
Donors are sometimes people and sometimes entities, or lobbies. The Koch Brothers would be a donor, for example, as would the NAACP and the NRA.
The voting piece is the hardest, and requires the most subjectivity because you have to mark legislation binarily as pro or con a given issue, when arguments can be made for all sorts of interpretations.
I was re-inspired to start looking at this by a friend, who did something similar in Perl. God bless him.