Can single-member districts ever be truly fair?

red and blue smily and frownyRecently I’ve been challenged in a very forceful but thoughtful way by a WordPress user who uses the handle realrepresentation. This reader advocates Proportional Representation as the system the United States should use to select its delegates to the House of Representatives.

I lived in Germany for a couple of years and watched a variant of proportional representation in action there, and I have to say I was impressed with the way it worked. Not only are there multiple viable political parties involved, but there is a fluidity among the fortunes of the parties that does indeed seem to reflect the changing opinions of the electorate, and friends of mine who were able to vote in Germany seemed to think that their vote mattered and that voting was important.

“RealRep” also makes the case that, not only is proportional Representation far superior to the gerrymandered single-member district system we have today – with which I heartily agree! – but it is also far superior to any single-member districting scheme that can ever be devised including OPRA or other “hands-off” variants – with which I don’t agree for a variety of reasons. I think my two areas of disagreement with RealRep probably fall into two broad buckets: first regarding fairness (i.e. is one system inherently more or less “fair” than the other), and second regarding practicality (i.e. given the political situation today, is one inherently more likely to be implemented than the other.) Both are intricate and interesting topics in and of themselves and worthy of detailed analysis, but I’d like to dip my toe in the water today focused only on the topic of fairness.

RealRep uses my own new smiley-/frownie-face graphics to make his point.* To quote him/her directly:

“Note the three examples of districting in the hypothetical rectangular state at the very top of this page. The voting population is as follows: 60% (9/15) black voters and 40% (6/15) red voters. One example shows how the state can be gerrymandered to give all three elections to the black voters, and two examples show how the state can be gerrymandered to give the red voters one or two of the three seats.

“Shouldn’t 40% of the voters be represented in the government? Of course. Should they have the majority of the three seats? Of course not.

Fair Representation Voting would establish the entire state as one district with three seats, and the red voters with 40% of the vote would win one seat while the majority, the black voters, would win two seats REGARDLESS OF WHERE IN THE STATE THE PEOPLE CHOSE TO LIVE.

“The beauty of FRV is that election outcomes are determined by like-mindedness of political belief and are independent, or mostly independent, of geography with very few wasted votes (i.e. votes which don’t elect anyone).”

Fair points, to be sure. But I’d like to kick the tires on one point RealRep makes when he/she implies that, in the hypothetical example, 40% of the voters are without representation. That does happen in the current system, but I think that is an artifact of gerrymandering and the related fact that, largely due to gerrymandering, the US is stuck in a two-party rut.

It DOES matter whether you win or lose, and whether you win or lose depends on how you play the game.

Let’s look at the diagram RealRep is referring to and get some terms straight. The diagrams show how two outcomes (a pink or a gray, indicated by which outcome “wins” in one of three districts) relate to two types of voters (a black symbol who wants gray to win, and a red symbol who wants pink to win). Given that there are three Congressional districts, and nine voters prefer gray while six prefer pink, the outcome of this election could be (from left to right) an undeserved 3-0 landslide by gray, a reasonable 2-1 victory for gray, or an underserved 2-1 victory by the minority position held for pink. The faces smile or frown according to whether their point of view prevails in the district in which they live. The larger point is that, with the exact same voting result, you may be happy or sad and your position may win or lose based only on how the lines were drawn. It is, in fact, a thematic way to describe what Pennsylvania’s State Senator Daylin Leach more comprehensively described as what happens in Pennsylvania.

In this example blue faces agree with “I prefer Democrats to Republicans”, red faces disagree.

The obvious comparison in the US is to equate pink and gray with the Republicans and the Democrats, and indeed if you live in a gerrymandered district and your party doesn’t win, you probably do feel like your vote doesn’t count and you have no meaningful representation for your views. But that is largely due to the fact that your opinions carry no weight with another party member who is safely installed in your district. In fact, one can argue your opinions don’t matter that much even if you’re representative is in the same party you are in: he/she is focused much more on what the party bosses want than what you want. This reality is usually masked but occasionally abruptly intrudes into the open, as when prior to Bill Clinton’s impeachment national polls found the general electorate very strongly opposed to impeachment, but faced with this strong preference only five Republicans broke from their party on the article of impeachment that finally passed. (Only five Democrats broke ranks with their party, presumably due to the heat they felt in their districts, though in doing so they were breaking from the solid national consensus.)

sacked smiliesThe way we want it to work, I think, is for representatives to broadly reflect what the people in their district think over the variety of all the issues they have to vote on. So, for example, a representative should know and respond to how his/her constituents think about several issues, and either vote as they would vote, persuade them they are wrong, or just vote against them confident that on balance the voters approve the representative’s voting record.

Getting back to RealRep’s question: yes, today if the election comes down only to “Do you prefer a Republican or a Democrat” and you’re on the losing end, the 40% who vote in the minority will not be represented. (And I actually don’t think the 60% majority are going to be well represented either.) But if your representative fears losing because he will lose his constituents on too many issues, your 40% minority view on any single issue doesn’t mean you don’t have a voice in government, it means your representative, on that issue, is faithfully representing his district. And when it comes time to vote for a new representative, if you’re in the minority you lose on that one issue also, because on balance the voters in your district prefer someone else.

The real test is whether, across a variety of issues, the 435 members of the House would tend to vote approximately as the millions of voters in the United States would vote. I suspect both hands-off single member elections and proportional representation, if implemented properly, would achieve that result.There are obviously big differences between the two, but in my opinion broadly-defined “fairness” is not one of them. Though on the issue of fairness I suspect there is interesting thinking to be done around realrepresentative’s phrase “where in the state they choose to live”.

What do you think? And what do you say, realrepresentation?



*These diagrams, btw, come from wikipedia commons and were used in an article on gerrymandering.

This entry was posted in Boundary mathematics, minority rights, Pennsylvania, Politics of Redistricting, Proportional Representation and tagged , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *