Redistricting

"Redistricting" establishes how your vote will be counted to elect Representatives to the US Congress. Already, political professionals are making plans to respond to the 2010 census to shape districts to their liking. You need to know about this to advocate a fair plan for your own state.

"As a mapmaker, I can have more of an impact on an election than a campaign, than a candidate." says Republican consultant David Winston, who drew House seats for the GOP after the 1990 census. "When I, as a mapmaker, have more of an impact on an election than the voters, the system is out of whack." quoted in Slate

(Following the 2010 census) the population of California, the most populous state in the union and larger than all but 34 nations, will grow nearly 8 percent from 2000 to 2010 - but California will lose a seat in the House.

Redistricting is perhaps the most complicated and mysterious political process because 1) it may differ from state to state due to law and which party controls what arms of government, 2) it often involves horse-trading out of the public eye, and 3) it is habitually been inadequately covered by the press because of the previous two reasons.

Political pundits and commentators dismiss 'process issues' by claiming they are of no interest to Americans. They are wrong. ... Today, in Washington, process is the name of the game, and those who do not understand this fact are operating in ignorance. Political observers who do not make an effort to understand process matters will remain uninformed. Scholars & Rogues, December 2008.

I have developed an application to help you see understand and compare districting plans - current and alternatives. Using election statistics and census data, you can answer these questions: What does it mean for you? Would election results change? What are the characteristics of the current group of people your vote is pooled with - and how would that change? The application currently works for Massachusetts - showing the current districts and an alternate plan with more compact districts. Links to download the application and see a screen-cast demonstration are at the bottom of this web-page. In the future, I will be enhancing this to add more states and present better alternative plans.

Historically, district plans are developed by each state. There is little guidance from the US Constitution. In the past, this has led to some abuses, and spawned the term "Gerrymandering", when some interest group, often the party in power, attempts to arrange districts to influence the outcome in their favor. The US Supreme Court, reluctant to interfere in the affairs of another branch of the Government, has made some decisions that over-turn district plans. One guiding principle is "one voter - one vote". Within any one state, anyone's vote should have an equal affect on any election. As a result, districts try to have equal populations.

However, there still is no agreed-upon, objective way to divide states or evaluate the results. There have been several attempts to develop computer systems to "automatically" create the districts. Some people have advocated compact districts with minimum perimeters. George Clark has applied this to California. Some people have advocated districts where the average distance to the center of each district is a minimum. Brian Olson has used this approach to develop plans for all the states. These alternative plans usually ignore city boundaries, and just group contiguous clusters of census tracts together. When alternative plans are presented, it is hard to see what difference they would make to any individual. Would any elections be changed? For any specific city, what would the new district look like, and how would the population differ from the current district?

I am developing algorithms that will divide states into compact districts. I don't think cities should be split - voters in the same city should have the same Representative in the US Congress. There is an understanding that districts should include voters with shared political interests. So, I would like districts to be as equal as possible, be as compact as possible, and have the division be proposed without political intervention. More importantly, I want to make the underlying information available and readily understandable to citizens. They can look at the current and alternative plans and decide what is best for them - and choose to influence their elected representatives if necessary.

This application portrays the current districts for Massachusetts, as well as one alternate plan - currently only generated manually. You can compare the two plans, comparing district populations and one measure of compactness. You can also see summary census data for the entire state, each district and any city.

Specifically - the comparison panel in the center - larger image shown below - graphically portrays the population in each of the current and alternative districts. It also shows the average distance in each district between each city center and the center of the district, weighted by the population of each city. I believe the voters of a District should share a common political interest, and that may manifest itself in relative proximity. By contrast, a District with a bizarre and stretched-out shape, like the current Massachusetts District 4, while having a good population count, may not help citizens represent their interests in the Congress. For the populations, the data is summarized as range and deviation. Based on its total population of 6,349,097 in the 2000 census, Massachusetts is "apportioned" 10 seats in the US House of Representatives. Hence, a "perfect district" would have 10% of that, or 634,910. Range is the difference between the highest and lowest district populations, expressed as a percent of the "perfect district". Similarly, deviation is the average difference between each district's population and the perfect size (ignoring sign) again expressed as a percent of the perfect district. I found this article from the Minnesota Senate legal department a useful guide in developing these measures.

There are a lot of ways to interact with the application to gain more information. I prepared a brief screen-cast showing you the details, and here is a link to see it. Application screencast. The screen-cast will run in your browser, takes about 6 minutes. When you're done, you can use the back-button to return here.

The application will run on your desktop. Once downloaded, you can run it at any time - you don't need to be connected to the Internet. If you run it again in the future and you are on the Net, the application will check to see if there is a later version available, and download that, so you stay up-to-date.

I will be enhancing the application in the future - using algorithms to generate better redistricting plans, and expanding to handle other states.

What you don't know about Redistricting can hurt you.