cross-posted from Voices on the Square, 21 October, 2013

It is something of a cottage industry at the moment to explain “why” gerrymandering is not responsible for the extremism of the Republican party. Take, for example, Nate Cohn at the New Republic, explaining

In fact, partisan gerrymandering usually reduces the number of extremely red districts. Why? Because the point of partisan gerrymandering isn’t to try and maximize the number of safe districts. The goal is to maximize the number of districts that are merely safe enough by packing as many of your opponents’ voters as you can into a small number of extremely partisan districts while safely distributing the rest throughout your own districts. In this way, gerrymandering may actually increase the number of moderate Republicans.

Of course, the argument is overblown. The primary electorate is, on balance, more extreme than the typical general election Republican voter, the check on that extremism dominating the Republican caucus elected to Congress is the ability of their opposition to take advantage of that extremism to win election against the partisan bias of the district. And in districts with an 8%+ partisan lean, it is it is difficul to recruit effective Democratic candidates. Without effective Democratic candidates, there is much less filtering of the extremism of the Republican primary base electorate.

As noted in The Gerrymander & Wave Election Floodplains, the gerrymander in Ohio created eight 8% Romney districts, so 2/3 of Ohio’s contribution to the House Republican Caucus comes from these safe districts.

However, more to the point, Nate Cohn is arguing over secondary impacts of the gerrymander, when the principle impact is given a nod and then forgotten:

Gerrymandering is undemocratic, and it did help consolidate the GOP’s House majority in 2012

We a political party in charge of the House with a majority of the Representatives depsite a minority of the vote due to the gerrymander. This is anti-democratic.

Imagining a world without heavy partisan gerrymanders

Now, I campaign for ending the gerrymander in Ohio, but between a lack of support from the Obama campaign and strong opposition by vested money interests, the issue that would have created a non-partisan redistricting system went down to heavy defeat in 2012.

And so we are left with the system that we have in place now. The state legislature district boundaries are drawn by a Districting Board consisting of the Governor, the State Secretary of State, the Auditor of the State, one Legislative Democrat and one Legislative Republican. This is a process that has been gerrymandered every ten years by both major parties since it was established in the 1960′s, and since the 1960′s, the Ohio State Assemble has only seen one change of control that was not the result of a change in partisan balance of the Districting Board. So in Ohio State Politics, once every ten years, three individuals elected on a statewide basis set down the Assembly majority for the coming ten years.

Then the (gerrymandered) state legislature sends a (gerrymandered) Congressional District map that has been passed through both houses, to be signed by the Governor. When, as in 2010, the two chambers are elected from gerrymandered districts from the previous decade, and the Governor is from the same party, the result can be a grotesque mockery of the democratic process, as in the current Ohio map above and to the right.

This essay is looking at opening up a second front in the fight against the gerrymander. The inspiration is the Australian bicameral (two chamber) system where at the Federal level and in most states, the lower house, the house of government, consists of single member seats elected by preference voting, and the upper house, the house of review, is elected on a proportional basis.

Having a bicameral legislature with both houses made up of representatives elected on a first-past-the-post basis, from districts heavily gerrymandered by whichever party held the balance of power tends to make the two houses largely redundant, except for the personalities of powerful politicians in either chamber protecting their individual prerogative. By contrast, electing the State Senate on a proportional basis provides a real check on the result of the State House of Representatives. And if that is regional proportional representation, that guarantees that every party contains a number of members who are “swing” members for a large enough partisan swing in the next State Senate election.

Having a Chamber of the Ohio Assembly that cannot be gerrymandered also means that one of the two chambers that must pass the Ohio Congressional District map is not itself a gerrymandered chamber.

So the plan here is to divide Ohio up into five regions: North & East Ohio, South & East Ohio, North & West Ohio, South & West Ohio, and Central Ohio. Thirty State Senators would be elected from the five regions in a party list election, in proportion to the vote for each party list. The final three State Senators would be elected statewide based on the spillover votes not required to elect the regional State Senators.

What Would These Regions Look Like

I don’t really know what the regions will look like. It is possible that 45% of the state population would end up in the North & East region. It is possible that 6% of the state population would end up in the South & East region. But I have set up a regional map that assumes that rural counties will tend to vote to be outside of the three regions including the three largest cities, and arrived at the map to the right.

The seeded counties were:

  • North & East: Cuyahoga County (pop 1.28m)
  • South & East: Washington, Athens, Meigs, Gallia and Lawrence Counties (pop 0.24m)
  • Central: Franklin Country (pop. 1.16m)
  • North & West: Williams, Fulton and Lucas Counties (pop 0.52m)
  • South & West: Hamilton County (pop 0.80)

Assuming that Canton, Akron & Youngstown joins Northeast Ohio and Dayton joins Southwest Ohio, the following population distribution would represent the rough balance of power between Northwest, Central and Southwest Ohio, with the strong disinclination of counties outside of the suburbs belts to be in a district with one of the Triple-C cities leading to a larger population in South & East and North & West Ohio.

I would propose that the three most populous counties are each seeded into a region, the counties bordering Michigan seeded into the North & West region, and a set of Ohio River counties seeded into the South & East region, and then residents of the remaining counties vote on their first and second preference for region. The regions would be allocated according to preference, limited by a requirement that they neighbor another county in the region. My notional map gives:

  • North & East Ohio, 34% of Ohio’s population, 10 State Senators
  • South & East Ohio, 11% of Ohio’s population, 3 State Senators
  • Central Ohio, 17% of Ohio’s population, 5 State Senators
  • North & West Ohio, 17% of Ohio’s population, 5 State Senators

South & West Ohio, 22% of Ohio’s population, 7 State Senators

State Senators are elected on a quota basis. For North & East Ohio, the quota is 1/11th of the electorate plus 1 vote, and every party receiving one or more full quotas elects one Senator from their party list, starting at the top and working down toward the bottom. If not all State Senate seats are filled, then the balance are filled by awarding one Senator to the parties with the largest partial quotas.

The final three seats are filled on a statewide basis. The votes from each region that were not used to fill a Senate Seat are pooled, and distributed with a quota of 1/4th of those votes plus 1. This ensures that if there is an “accidental gerrymander” in the district boundaries, the statewide vote is represented in the State Senate. This also provides an opportunity for a 3rd party that attracts sufficient support on a statewide basis to justify one or more seats in a 33 Seat Senate to gain democratic representation, even if they do not attract 1/11th to 1/4th support in any particular region.

Ballot Access for at least one candidate per region should be guaranteed to every party that gained at least 1% of the vote in the preceding Gubernatorial election. Ballot Access should also be guaranteed for parties with State Senate representation in that district, which could be guaranteed access to a party list equal to one plus twice the number of State Senate seats held.

Advocating For a Gerrymander Free State Senate

Obviously this is not a “quick fix” to the gerrymander, since gaining support for such a dramatic change to electoral institutions is likely a decade or longer process. That suggests that the first focus for advocacy would be at Universities, and in particular in 3rd parties on the left wing and right wing, where this reform would open up the hope of gaining representation in the State Senate independent of whichever two parties may be contesting to win the majority in the State House of Representatives.