Partisanship has a role in redistricting

I was an expert witness retained by plaintiffs, and I testified in the hearings before the special master on New Mexico redistricting after the 2010 census. I have been an expert witness in more than 50 voting rights and redistricting cases before the federal and state courts in the U.S. and Canada. I have drawn and evaluated districting plans for political parties, local units of government, civil rights organizations, the U.S. Department of Justice and the federal courts.

Because the new Redistricting Commission is prohibited from using partisan voting data in drawing the districting plans it will submit to the state Legislature, it is important for the public and our state representatives and senators to independently test proposed plans for partisan bias.

Plans drawn using only Traditional Districting Principles — TDP: one person, one vote; compactness; following local boundaries and major geographic features; community of interest, etc. — may have unintentional partisan bias. The way in which the commission prioritizes and balances the various TDPs will make partisan bias more or less likely.

Two assumptions may be involved in the provision prohibiting the Redistricting Commission from considering partisan voting data. Both assumptions, however, are wrong.

The first possible assumption is partisanship is not important in drawing districts. Partisan identification is not just an important determinant of how Americans vote, it is the most important determinant. Political scientists have known the importance of party ID and repeatedly confirmed the overwhelming and consistent strength of this relationship since the 1940s in election after election and for all offices from president on down the ballot.

In Congress and the state Legislature, party is clearly the most important determinant of how representatives behave. Whenever there is a controversial issue, the vote is usually more or less along party lines. This polarization has increased in recent years in Congress and in the state Legislature. Moreover, the Congress and the state Legislature are organized using partisanship with the leaders, chairs of committees and majorities on each committee going to the party with the most members.

It follows that the way in which a district plan translates votes into seats is the most important characteristic of any redistricting plan.

A second, more likely, assumption is that partisan fairness would automatically be obtained in a plan that follows TDPs. Commission districting plans may or may not provide a level playing field between the parties. In many states a plan drawn using only TDPs would favor the Republican Party candidates because Democratic voters are more concentrated in urban areas while Republican voters are more scattered. The possible bias in exclusively TDP plans in New Mexico are unknown.

Having a nonpartisan commission initially recommend plans to the state Legislature is an improvement over allowing partisan majorities in the state Legislature to draw plans using partisanship data to favor themselves. But partisanship data can also be used to draw districts that are fair — a level playing field between the parties.

More than a dozen statistical measures have been developed in the last decade for determining partisan bias. These statistical measures can easily be applied to proposed plans as they are developed and presented to the state Legislature by the commission. Information about the partisan bias or fairness of redistricting plans must be available to the public before they are adopted