Princeton Election Consortium

Innovations in democracy since 2004

Outcome: Biden 306 EV (D+1.2% from toss-up), Senate 50 D (D+1.0%)
Nov 3 polls: Biden 342 EV (D+5.3%), Senate 50-55 D (D+3.9%), House control D+4.6%
Moneyball states: President AZ NE-2 NV, Senate MT ME AK, Legislatures KS TX NC

How a lazy media narrative has missed the boat on the NJ redistricting story

December 15th, 2018, 9:37am by Sam Wang

Much news coverage of the New Jersey redistricting amendment has characterized it as a Democratic power grab. But this is a lazy dependence on the national narrative of partisan warfare. Here in the Garden State, this is an intra-party squabble. Republicans are bystanders – and minorities are missing a chance at getting much-needed protection from the Supreme Court.

As fun as it is for the press to claim “both sides do it,” that’s actually not true in this case. In fact, there is so much confusion that proponents and opponents of the legislation are both citing Princeton Gerrymandering Project analysis – though in some cases, somewhat selectively.

There’s a wrong idea going around. The legislation *doesn’t* cement a Democratic advantage. Both parties can still gain an advantage, by persuading the independent commissioner to accept their plan.

This legislation is mostly not about partisanship. Instead, it is a power grab by the Democratic General Assembly – to take power away from Governor Murphy,  who is also a Democrat. The grab consists of changing the commission away from being appointed by the state parties and governor. Instead, commissioners will be appointed by legislative leaders.

The commission is currently half Democrats, half Republicans, plus an added independent tiebreaker appointed by the state Supreme Court. That’s the good part! The amendment doesn’t change that fact. But it also doesn’t do anything to mandate that competing plans from the two parties be closer together. It’s a lost opportunity.

Much of what has appeared in the news media about the mathematical formulas is wrong. Here’s a recent example by Mark Joseph Stern of Slate. As much as I generally like Stern’s work, there are some factual issues with this particular piece. He’s not alone.

Here is a quick FAQ on what the legislation does and doesn’t do.

Does this legislation lock in a Democratic majority? No, not at all. Under the formula that half of districts be more Democratic than average, and half more Republican than average, a guaranteed majority for Democrats is only true if their local candidates match their statewide candidates, and get more than 50% of the statewide vote. Which, um, is a popular majority. This is not exactly a problem. It’s a good thing!

Under this plan, can Democrats really create a plan that wins 70% of seats with 57% of the vote? Sort of. Mr. Stern and others have quoted the Princeton Gerrymandering Project as saying this. That’s true – but we also find that Republicans can do something quite similar. Amusingly, this latter fact was mentioned (unattributed to us) in a Thursday hearing in Trenton by New Jersey Democrats, who argued this was a good point about the bill. In other words, the bill doesn’t hand Democrats an advantage. It just doesn’t restrain bad acts at all by either side. That fact that we are giving ammunition to both sides illustrates the confusion surrounding the legislation.

By the way, the 70% figure is not as scary as it sounds (nerdy footnote 1).

The law can bite Democrats back, hard. Under the same rules, Republicans can craft a plan that gives them the legislature with a popular minority. We think this can happen with as little as a five-point swing against Democrats, which could easily happen in 2021 if Democrats win the presidency in 2020 (incumbents generally suffer losses the next year). That’s a loophole that Team Blue probably didn’t intend to leave.

Does the law suppress competition? No, this is totally wrong, on two counts. The wrong criticism is that the legislation “only” mandates that one-fourth of districts be near the average, which is how the legislation defines competitiveness. It lets the rest be more Democratic or Republican-leaning. But in a natural distribution of districts, that’s about what one would expect (nerdy footnote 2). So in a sense, the legislation is par for the course.

But there’s something sillier. Because the state average tilts Democratic, these pseudo-competitive districts are safe-ish for Democrats. Making more districts like that is exactly how Democrats would draw a gerrymander!

Is the bill better than no bill at all? No. Aside from the black eye that New Jersey Democrats are currently getting, it misses an opportunity to fix real problems. Remaining protections for minorities under the Voting Rights Act are one Supreme Court ruling away from disappearing. NJ legislators could write those protections into the state constitution…but they didn’t.

Another fix to the legislation would be to add language requiring that a statewide plan not favor either party unduly. That would be a bit vague, but at least it would give the tiebreaking commissioner an anchoring point.

But to tell the truth, New Jersey’s existing mechanism for drawing the lines is imperfect…but it’s not the worst out there, not by a long shot. So a relatively good solution is to simply leave the law alone!

Whatever one thinks of the press coverage, it is undeniably true that Senator Sweeney is in a bad political situation. We had a previous suggestion on how to repair the legislation, but at this point fiddling with the math might be politically untenable.

However, he can salvage it pretty easily, and still get what he wants. He could strip out the math and replace it with a guarantee of voting rights for minorities, echoing the Voting Rights Act. That still allows him to stick it to Murphy, and silence many critics.

Nerdy footnote 1: It is commonplace to complain that percentage X of the vote leads to much more than percentage X of seats. This complaint is mostly lame. Winner-take-all elections lead to that naturally. For example, in 2012 Barack Obama won 52% of the two-party vote – and 62% of the electoral votes. It’s a nonlinear relationship.

In the example cited by the media, 70% of New Jersey legislative seats is more than they would “deserve.” But under neutral principles, 57% of the vote might typically get them only about 60-65% of seats. 70% is too many, but it is not exactly a runaway gerrymander, like one sees in North Carolina.

Nerdy footnote 2: The legislation calls for at least one-quarter of districts to be within five points of the state average. But in most states, the standard deviation on partisanship is about 15 points. So the legislation mandates how many districts should fall within one-third of a standard deviation. The answer is 26%, almost exactly one-quarter. In other words, the legislation mandates typical behavior of a normal process. By itself, this is not a partisan move!

However, having lots more such districts might be a problem, as we have documented before.

Tags: Redistricting

3 Comments so far ↓

Leave a Comment