Systems and politics

Electoral district design can affect politicians’ incentive to compromise. This is a systems problem, of course, and can be understand and fixed entirely without reference to political views.


  • Assume that people of similar views tend to live near each other.
  • Then if you can control the boundaries of political districts, you can create groups that are more likely to vote as a consistent block.
  • Now give the people whose jobs depend on those district boundaries the power to set them.
  • Finally, use a winner-take-all primary system in which only one candidate from each party is on the general election ballot.

Over time, this system should produce districts that look like Texas’ 29th district below, or California’s 11th district before the Citizens’ Commission re-drew it. There’s no obvious reason why these districts should look as they do … unless you wanted to ensure an outcome in which the general election vote in any district never gets close enough to be of any doubt.



Biasing of voter blocks is consciously done, yet results in an outcome that only serves the elected officials themselves and not the population taken as a whole.

It’s not that hard to design a better system. There are a lot of better systems, as it turns out. Either through giving to power to some other, less-biased group, or by using a simple, easily understood and by its nature unbiased algorithm, they all involve taking away from the regulated the right to regulate themselves – a pretty simple concept.

Here’s an example of a simple fix: The shortest splitline algorithm – in brief, dividing the state in half by population repeatedly until you have enough districts for that state’s population – would work fine. The result would look like this for California:

Screen Shot 2013-10-16 at 2.02.24 PM

That would leave you with districts based totally on geography and population. It would take politics entirely out of the decision-making process.

The main objection to an algorithm-based approach is that lines don’t follow normal terrain or cultural boundaries. But of course, that’s the point: you don’t want districts to group entirely like-minded individuals together. All that’s necessary is that the voter know which district they’re in and the candidates know who their constituents are. Definitely not rocket science.

Another approach, such as California actually enacted, is to create a non-partisan citizen’s commission. Here’s California’s 11th after the commission. It took a lot more work than an algorithm would have, but the result is definitely an improvement over the prior district map. It’s quite a bit more regular to say the least.


Here’s the point and why it matters in business: All of this was known decades ago. Everyone knew this was happening and could have predicted ending up where we ended up this month. No technology was ever required to fix this problem, and none is required now. All that is needed is for the decision makers (in this case, citizens; in business, management) to recognize that they’ve put in place a system – again, not a system of technology, but one of approach – that is designed to produce a poor outcome, and replace it with a simpler one that is not.