Redistricting RI: Congressional Maps

Today’s post is the first in a 3-part series on the legislative redistricting process in Rhode Island, starting with the Congressional maps. Compared to larger states, redistricting on the Congressional level in Rhode Island is fairly low-stakes, but because of past abuses by entrenched interests in both parties, I believe that the redistricting process is worthy of attention at every level. Also, Rhode Island was notably close to losing its second Congressional seat, so this exercise is one we should have fun with while we still can!

To start, I would encourage everyone to read this explainer from Sam Howard about the various redistricting proposals that were introduced during last year’s legislative session. Obviously, the only proposal with any teeth is the one adopted by the General Assembly leadership that is now playing out. But Howard makes a valuable point that I wish to emphasize here, which is that redistricting plans prioritize different aims. These guiding priorities can produce different maps, sometimes drastically different, even in a small state like Rhode Island.

My game plan for this series is to produce maps (using Dave’s Redistricting App) for the Congressional, State Senate, and State House districts that each roughly comport with the redistricting plans introduced last year, each of which had a slightly different set of priorities. Of course, all of the plans make at least some mention of the need to follow federal law and court precedents on redistricting. Again, I would encourage reading the post from Howard linked above, but my somewhat-oversimplified rundown of the various plans is as follows:

• Ruggerio/Shekarchi Plan (Adopted): Prioritizes compact, contiguous districts that reflect communities of interest, very broadly defined.
• Cortvriend Plan: Prioritizes fair representation of minority communities, and more narrowly defines communities of interest as “including, but not limited to, racial, ethnic, economic, social, cultural, geographic, or historic identities”
• Euer Plan: Prioritizes partisan fairness, contiguity, and geographic integrity.

So, with these three plans, we have something like three different sets of guiding priorities, which we can translate into different maps at every level.

To start, for anyone unfamiliar, the current RI Congressional map looks like this:

The notable stats (partisanship scores are from the presidential elections, not the Congressional elections themselves) are below. It is also important to note that, based on the 2020 Census numbers, D1 has 5,617 more in population than D2.

• Ruggerio/Shekarchi plan

Because this plan is the one that will produce the final maps, the map below is a de-facto prediction of what I think the districts will look like.

The conventional wisdom is that there is a need to shore up the 2nd district to make it safer for Rep. Jim Langevin and any future Democrats. As you can see from above, Trump was within single digits of Clinton in D2 in 2016, before Biden bounced back. Even still, Biden was still notably far behind Barack Obama’s 22 point victory in D2 in 2012, so a long-term rightward trend is still visible here. Overall, the Cook Political Report puts D2’s Partisan Voter Index at D+6, meaning it is 6 points more Democratic than the nation on average. In a state with just 2 seats that Biden won by 20, Dems will want a safer seat here.

The proposed map moves the rural, solidly Republican towns of Burrillville and Glocester into D1, and puts more of Providence into D2. The results of this are easy to predict — a more Democratic D2 and a less Democratic but still completely safe D1. Minority representation becomes slightly worse in D1, but representation is North of 25 percent in each district. I also reconfigure the Providence split to make it smoother and more compact. I personally believe that splitting Providence is bad for representative democracy in this state, but it is one easy way to ensure control for Democrats, and it is not a stretch to predict that this is what the Dems will do. If you are going to do a split, I think the West/South vs. North/East split in this map, with Smith Street as the dividing line, is about as good as it gets.

Cortvriend Plan

With minority representation as a guiding priority, my aim was to create a district (D1) that maximizes the opportunity for representation, without excessive community splits. Of course, this is impossible to do while still splitting Providence, so it moves totally to D1 here. The split occurs instead in Cranston, which I don’t love but feel better about due to the demographic and geographic differences between Eastern and Western Cranston. The heavily -white Bristol and Newport counties move to D2, as does the town of North Smithfield. The county shifts are significant changes, but in a geographically odd state like Rhode Island, I am choosing to count contiguity by bridge as a viable option. I kept Smithfield in D1 in the interest of representation for the 1,700 Asian residents living there.

Doing this, you can get D1 to 43.7% minority, and with the exception of falling short on Asian representation, Hispanic and Black residents are over-represented in D1 compared to their share of the overall RI population. Interestingly, the partisanship of D1 changes none from 2016 to 2020, even though Biden won the state more handily, perhaps due to rightward swings among Hispanic voters in 2020. It is still solidly safe D regardless, while D2 becomes more competitive but still solid D in a good Dem year.

Euer Plan

A plan that prioritizes partisan fairness can be approached in different ways, and I will choose to do it 2 ways here and a different way when it comes to the General Assembly maps.

The Dem-friendly way to interpret this would be 2 districts that each match the partisan lean of the state overall. The Republican friendly interpretation is that 1 of the 2 seats should be made as competitive as possible. I present both here:

For the top district, I used DRA’s 2016–2020 composite results, which compiles all statewide elections from those years to give a broader measure of partisan lean, to put Rhode Island’s overall lean at D+23. I then set out to make 2 districts that are roughly D+23 using this measure. There are likely many other ways to accomplish this, but all require some of the weird moves that I made above, such as splitting apart Pawtucket and East Providence from the rest of Providence County, and putting a split in Coventry. Even with losing Pawtucket, D1 has higher minority representation in this configuration than in the current map. Republicans are pretty well shut out from both districts in this map.

The second map is as close to a “reasonable” Republican gerrymander that you could get, and it actually requires no split communities (Providence and South counties are split). Again, without excessive splits and/or sacrificing compactness, D2 here is probably the only district you could draw that went for Donald Trump in 2016. Interestingly, the two maps here have the exact same minority representation despite very different geographies.

I hope the way I have laid out these plans reinforces a general point about the complicated nature of the redistricting process. Although there was never any doubt about which redistricting plan would be chosen, this exercise is worthwhile to show that the priorities we choose to guide this process have real consequences for how Rhode Islanders are represented at the state and federal levels. My next post will cover the State Senate maps, again proposing 3 different maps guided by different priorities.