Thursday, January 31, 2013

Taking A Gerrygander

I don't need to tell you about gerrymandering. At least I shouldn't need to. It's a basic reality of electoral politics, American politics in particular, a tradition literally as old as the republic. Ever since the days of namesake Elbridge Gerry- who participated in the Constitutional Convention, signed the Declaration of Independence, and was part of the very first US Congress- as long as there has been a need for regionally-based representation in government, there has been someone looking to make those regions as favorable to themselves as possible. And so every time districts are drawn, there is inevitably someone- probably a bunch of someones- who will create a district in a shape previously unseen outside of H.P. Lovecraft stories. Even before Gerry, Patrick Henry- he of "Give me liberty, or give me death"- asked the Virginia legislature in 1788 to give him redistricting and give political enemy James Madison headaches by drawing him into the same district as James Monroe in the election for the first Congress. (It didn't work. Madison won.)

With the advent of computers, necessary to ensure each district has precisely the same amount of people if at all possible, gerrymandering can be taken further than ever before, with voter data and algorithms being able to maximize advantages beyond what was previously possible. In 2003, Jeffrey Toobin told the story of Frank Mascara, a Democrat elected to Congress in 1994 as a member of the House from the Pittsburgh suburbs. After the 2000 Census, Republicans had control of the Pennsylvania state legislature, and therefore the redistricting process. They drew the districts such that Mascara's house was in one district and his driveway in another. The district with his house was now in the district of House mainstay- and fellow Democrat- John Murtha. Toobin referred to Mascara as having been "kidnapped". Murtha defeated Mascara in the 2002 elections, and Mascara's political career was over.

The 2012 elections being the first since the 2010 census was finalized, gerrymandering has come into the conversation once again, most notably in the recent Republican chatter about converting certain states from winner-take-all in the Electoral College to a circumstance where they award electoral votes by Congressional district. This was considered for Virginia, Michigan, Wisconsin, Ohio, Florida and Pennsylvania- all considered swing states in the 2012 election, and all of which went to Barack Obama. News of the plan quickly spread, and so did the outrage over a plan to essentially gerrymander the Electoral College. Opponents were quick to note how, had the plan been carried out for 2012, Mitt Romney would have won the Electoral College despite substantially losing the popular vote. The plan quickly fell apart as Republicans backed away one by one.

Mostly lost in the discussion is the fact that the states themselves have been pre-gerrymandered the entire time. Nobody set out to do it that way, but what is the Electoral College but a mechanism to turn the states into 51 districts, each with a different weighted importance at election time? Beyond that, Congressional districts can't spill over state lines, nor can state legislative maps, and Senators and governors never see their constituent borders change. The Senatorial circumstance matches that seen in Alabama from 1901 until 1972, stemming from a Jim Crow mentality under which, given that changes in population and movement would have given blacks more power under any new map, the white-dominated legislature simply refused to redistrict at all. 1972 marks the year that the U.S. District Court finally forced them to do so.

There is a movement to remove the districts on the Presidential level, the National Popular Vote Interstate Compact, in which states- which are allowed to award their electoral votes in any manner they see fit- enter into an agreement with each other that they will award their electoral votes to the candidate that wins the national popular vote. The compact states, though, that they won't do that until enough states have agreed to total 270 electoral votes- enough to override whatever the stragglers are doing. As of now, nine states carrying 132 electoral votes are on board: Maryland, New Jersey, Illinois, Hawaii, Washington, Massachusetts, the District of Columbia, Vermont, and California. (You'll note it's all safely Democratic states so far. The Republican base is more rural, and so the party takes advantage of rural areas having disproportionate power in the Electoral College. It's not in their best interest to join in.)

In Toobin's article, he interviewed Nathaniel Persily, then an assistant professor at the University of Pennsylvania.

“There used to be a theory that gerrymandering was self-regulating,” Persily explained. “The idea was that the more greedy you are in maximizing the number of districts your party can control, the more likely it is that a small shift of votes will lead you to lose a lot of districts. But it’s not self-regulating anymore. The software is too good, and the partisanship is too strong.”

I want to play with this quote a bit. Persily's already laid out the idea: once your party has control, the goal becomes to get your party as many seats as you can. Usually, this is going to happen when you have the majority anyway. But you don't have ALL the votes. The other party has voters out there too, and therein lies the challenge. If you had all the voters, there wouldn't be much point in gerrymandering, now, would there? The goal behind gerrymandering, as you know, is to spread your opponent across as many districts as possible while giving them as few seats as possible. The two main ways to do that are packing, in which you cede one district in order to take as many voters as possible out of the other districts, and cracking, in which you make as many districts as possible in which the opponent has just barely not enough voters to win, in an effort to deny them any seats at all. Packing creates one 90-10 win for the opposition and landslide losses everywhere else, cracking creates a lot of 55-45 wins for your side that, if done right, stand little chance of turning into 51-49 losses. The risk is that you go for too many seats, spread your supporters too thin, and that a shift in the electorate creates a gigantic swing in seats. And that risk, as much as Persily wishes to minimize it, does still exist. Software is only as good as the people who use it.

Which brings us to the alternative of a majority maximizing their majority: a majority being made out of a minority. We can go back to Gore/Bush or Tilden/Hayes, but we don't need to go back in time. It's happening right now. In 2012, the Republicans- funny how they keep benefiting, isn't it?- took a total of 58,283,036 votes in all elections nationwide to the House of Representatives, against the Democrats' 59,645,387 votes. That means the Republicans got 49.42% of the vote nationwide. However, the districts were arranged such that they won 234 seats to the Democrats' 201- 53.79% of the seats.

Now, Persily noted that things were too partisan and the software too good to have to worry about that number dropping much. And given the number of articles about how the Republicans are going to be in control of the House for the next decade, some would agree with him. But hang on a minute. Didn't we just get done- aren't we still- talking about how the Republican support structure is steadily slipping further and further as more and more of their supporters become alienated, and as Republican-friendly demographics are inexorably shrinking while Democratic-friendly demographics are inexorably growing? What happened to that?

Sure. 49% of the votes can become 50+1% of the seats. Heck, 49% of the votes has turned into 54% of the seats. But gerrymandering can't save you forever. Alabama's Jim Crow legislators thought it could, but things work differently from their day. You have to redraw the lines to ensure equally-sized districts, and if your minority shrinks enough, there comes a point where no amount of redrawing will give you the gavel. 49% can be enough; 46% is surely enough as well given the circumstances in the House. But how far down can you dip? Will 45% do it? 44%? 43%? How about 39%? 29%? 19%? 9%? Where's the line?

Few are interested in the answer, instead focusing their energies on various proposals to redraw borders in such a way that gerrymandering is reduced or impossible. But given that the party in power is rarely if ever in a mood to limit their own gains, the party out of power- the party much more likely to want a fairer system, so their losses might be minimized- is not in a position to do anything about it, and areas with split control are usually most interested in saving the jobs of all involved- jobs which would surely be imperiled under a new system- it seems less practical to study the science of fair redistricting than it seems to study the science of its limits.

And unless the Republicans find a way to grow their support, they may find out those limits soon enough.

No comments: