The map whisperer: A conversation with Cory McCartan, winner of the 2021 Iowa redistricting challenge

Images courtesy of Cory McCartan

Every 10 years, the political commentator Dave Wasserman holds an “Iowa redistricting challenge.” The goal of the challenge is to use new census data to draw a congressional map of Iowa distributing the population as equally as possible into districts.

This year, the challenge was won by Cory McCartan, a Grinnell College graduate currently pursuing his Ph.D. in Statistics from Harvard University. Within two hours, McCartan had submitted a map in which the difference between the largest and the smallest congressional district was only five people.

Little Village talked with McCartan about that map, Johnson County’s claims of being undercounted in the 2020 Census and what we might expect Iowa’s new congressional maps to look like.

Congrats on winning the challenge! In layman’s terms, how did you do it?

I’m in grad school studying redistricting, so I’ve been working on computer algorithms that can generate many, many redistricting plans at a time. The way these redistricting programs work is by drawing a district one at a time, trying to split off a district from a blank map, and it keeps doing that until it comes up with a district that comes close to your population target. Then once it’s done that, it tries to find another district from the remainder of the map, and keeps going that way until it has a whole map.

In the case of Iowa, you do congressional redistricting by counties. Ninety-nine counties doesn’t seem like that many, but there are trillions and trillions of ways you can put 99 counties together into a plan. I didn’t try all of them, that would have taken too long, but I tried enough to zero in on this particular map.

The reason that redistricting is in the news is that the Census Bureau released their latest population counts. Some places here in Iowa, like Johnson County, are claiming that they were undercounted in the census. Do you think the census numbers are accurate, considering the pandemic and the other difficulties the U.S. Census Bureau faced?

We’re not going to get a sense of how well they did their job until sometime next year, when the Census does a quality check survey that determines what they think the undercount was. But the characteristics that have traditionally led to undercounts are places with more diversity, places with college students – so it’s not out of the question that Johnson County was a perfect storm of these factors. It was very hard for the census to track down someone’s actual location during the pandemic.

When you look across the country, in areas where you might expect to see undercounts – cities with a lot of immigrants, for instance – we didn’t see a big drop. So there’s some reason to think the census count was OK, but we won’t know more until next year.

Is there any redress for places that think they were undercounted?

There might be lawsuits on the census’s new privacy protections, but that’s not going to substantially change the count of people there. All the federal formulas on funding, et cetera — those are locked into last week’s numbers.

Back to redistricting. What do you think about the rise of tools like Dave’s Redistricting App that allow members of the public to draw maps?

That’s definitely new. So far, the increase of computing power has mostly been on the map-drawers’ side. In 2010, there were a lot of computers being used, but it was by people wanting to gerrymander. Now, the public side has caught up a bit, and people can draw their own maps and ask, “Why didn’t you do it this way?” There are also sites that evaluate maps, to say this map is packing and cracking specific populations.

You’re even seeing states use public map-drawing software to solicit input in the drawing process. In Colorado, there’s a website where you can draw your community; in this area, we all have the same job, or this is an immigrant community or linguistic group, and it’s really important you keep this together.

There was a ski resort area, for instance, where people said it’s important that this area has the same representative in Congress, so their interests are advocated. The commission took all that information, so that the districts preserve these communities and not split them up. That’s something technology and data have enabled, and that’s a positive development for building good districts.

You post a lot of your maps on Twitter. Are you of the opinion that the more we put out there and discuss these maps in public, the better?

If a redistricting official wants to create a gerrymander, they are able to do that. If you have enough computing power and the tools, and you get someone with experience in doing this, they’re going to be able to draw a pretty potent gerrymander. It’s an asymmetry of attention and salience and resources. If we aren’t making a big deal about this, you’re just going to have people with a vested interest and the resources to further that interest drawing the maps.

So being able to quickly say, “Here are 10,000 other ways you could have done this, and they’re all more fair than the one you did,” and being able to do that quickly after they put out proposed maps, starts to shift the momentum back to better, fairer districting. In the past, that would have played out inside a statehouse or in a courtroom alone.

Do you plan on doing that kind of analysis of the officially proposed Iowa maps?

As states around the country release proposed and final maps, I and another Ph.D. student here plan to quickly analyze them, by comparing them with other examples of what might have been. If the proposal is right in the middle of the other examples, that’s good, that’s fair – if it’s not, we can point out the ways that these maps are unfair.

In Iowa, our redistricting laws narrow the range of possible partisan outcomes for congressional maps. What do you think are the most Republican or most Democratic maps that you think we could see here?

If you know a little about Iowa political geography, you know there are four heavily Democratic vote centers: Polk, Story, Johnson and Linn counties. You can either keep those together or separate them, and that’s the primary determinant of the partisan outcomes of that map.

The partisan shifts we’ve seen in the state obviously don’t bode well for Democrats. Traditionally, there was a lot more competition in rural areas, particularly in the east. That hasn’t completely cratered but it’s a much more challenging landscape for Democrats. So winning congressional districts comes down to racking up those vote margins in the cities.

It’s possible to draw a map that is 2-2 with margins that in a normal year are comfortable. But that also means that, maybe in 2022 there’s a swing against Democrats, and those two could be wiped out. So maybe even better for Democrats would be a 3-1, where that one seat is a little safer, and protected against midterm swings.

On the flip side, Republicans could also take that one Democratic seat and make it a lot more competitive. You would have two safe Republican seats, or even three safe Republican seats, and one more competitive seat. It’s possible Republicans could try for four safe seats, but they really can’t be too safe just because of the vote center in Des Moines.

I don’t want to place bets, but it seems you could have anything from 4-0 to 2-2, but more importantly, how safe are those boundary seats going to be? Are Republicans going to get greedy and try to take all four and risk a 2018-sized backlash? Or are they just going to lock things in as they are at 3-1 and preserve their majority for the rest of the decade?

Obviously no one saw 2016 coming back in 2011, so no one knows what future alignments are going to hold. That’s the challenge for redistricting: not just knowing what’s going to work for the next midterm, but down the road.

Evan Burger is a longtime political organizer who lives in Slater, Iowa. He was the Iowa caucus director for Bernie Sanders’ 2020 campaign, and is now a partner at the progressive consulting firm Hegemony Strategies.

Thoughts? Tips? A cute picture of a dog? Share them with LV »


World of Bikes presents:


Come talk with our experts about ways to stay on the bike this winter! Find out more at World of Bikes.

Get Started

The Future is Unwritten

You look to Little Village for today’s stories. Your sustaining support will help us write tomorrow’s.


$10/mo or $120/year
The cost of doing this work really adds up! Your contribution at this level will cover telephone and internet expenses for one month at the LV editorial offices.


$20/mo or $240/year
$240 is enough to cover one month’s costs for sending out our weekly entertainment newsletter, The Weekender. Make a contribution at this level to put a little more oomph on your support and your weekend.


$30/mo or $360/year
(AUTO-RENEW) connects eastern Iowa culture with the world. Your contribution at this level will cover the site’s hosting costs for three months. A bold move for our boldest supporters!

All monthly and annual contributors receive:

  • Recognition on our Supporters page (aliases welcome)
  • Exclusive early access when we release new half-price gift cards
  • Access to a secret Facebook group where you can connect with other supporters and discuss the latest news and upcoming events (and maybe swap pet pics?) with the LV staff
  • Invitations to periodic publisher chats (held virtually for now) to meet with Matt and give him a piece of your mind, ask your burning questions and hear more about the future plans for Little Village, Bread & Butter Magazine, Witching Hour Festival and our other endeavors.