How Science Can Make Voting Fair

Published Mar 5, 2018

Voting Rights expert Dr. Michael Latner uses science to level the playing field for voters.

In this episode
  • Colleen asks what gerrymandering is and if it’s even legal
  • Mike explains how the science of elections deals with representation and inequalities
  • Mike shares the ways electoral reform is difficult, why it’s “like working on your car while you’re driving down the road”
  • Colleen learns why there are multiple voting systems used across the US instead of just one, universal system
Related content
Full transcript

Welcome to the got science podcast. I’m your host Colleen MacDonald. Do you ever wonder if your vote counts? With so much political maneuvering like redrawing district lines for political gain there has to be some way science can help us out. Well that’s what we’re talking about today on the got science podcast.

As we approach the midterm elections a few states have been making news for rulings on how they draw maps. Specifically how voters in this state are grouped and counted by congressional district. For example North Carolina was forced to redraw its congressional districts last year. And earlier this month the United States Supreme Court ruled that the Supreme Court of Pennsylvania had the right to demand that lawmakers in the state redraw their own congressional maps. In fact they’re working on new maps as I speak. Hopefully when they’re done they’ll have solved the problem that brought them to the Supreme Court in the first place and landed North Carolina in the news, the problem of gerrymandering.

Fun fact about gerrymandering: it was named after Elbridge Gerry, the ninth governor of my home state of Massachusetts. While he was in office, a particular district in the commonwealth was redrawn under new laws. Because it had been so oddly reshaped it ended up looking like a salamander on the map. Elbridge Gerry had approved the laws allowing the new map, so people stated calling the district, and the practice of redrawing electoral maps gerrymandering. Why did the “G” changed from Gerry to Jerry? Who knows. What we can answer is why gerrymandering is so bad for our Democracy. Joining the podcast today is my colleague, Michael Latner, our Kendall voting rights fellow here at the Union of Concerned Scientists. Mike is on sabbatical from California polytechnic state university where he’s an associate professor of political science and public policy. He joined me to talk about redistricting, our friend the gerrymander, different systems of voting, and most importantly for our podcast, how science can inform truly fair and proportional representation.

Colleen: Mike, thanks for joining us on the "Got Science?" podcast.

Mike: Thank you for having me.

Colleen: I want to start right off with gerrymandering. Tell us what it is and why it’s legal.

Mike: Sure. Well, so some gerrymandering right now is legal, some is not. So the idea of gerrymandering can mean a lot of things. And in its essence, it means drawing electoral district boundaries in such a way so that one group of voters is advantaged over another group of voters. And so, there are a number of ways that you can gerrymander or advantage some voters over others. You can gerrymander in such a way so that rural voters are over-represented, and urban voters are under-represented. You can gerrymander in such a way that certain racial groups have an advantage, and certain racial groups have a disadvantage.

Colleen: So, it sounds like it should be illegal, but it's not. What's the up-side?

Mike: Right, so in the case of racial gerrymandering, right? So typically historically, gerrymandering was one of the methods in the south that the caste system was maintained, where African-American voters were disenfranchised. The value of their votes were diluted relative to white voters, in addition to a lot of other electoral mechanisms as well.

But one of the ways that this was primarily done is what's known as mal-apportionment, right? And mal-apportionment occurs when you have very un-equal populations in electoral districts. And so what would typically happen in the south was they would draw electoral districts around a black community, they'd put all the African-American voters or African-American residents in that large, over-populated district. And then, say, you have five other districts in a six-district city, those rural, under-populated districts would all elect white representatives, even if the composition of the town was 50-50, right? So, that was made illegal after the Voting Rights Act of 1965. There were a number of cases that have come to be known as the re-apportionment and re-districting revolution, which basically outlawed racial gerrymandering of any kind.

In addition to outlawing mal-apportionment, the method of cracking up or packing African-American populations in such a way that their vote was diluted at the state level, that was also outlawed. And so we now have a one-person-one-vote rule, which means you have to have equal populations in districts or roughly equal populations depending on what level of government you're talking about, which is fairly easy to measure. That's why we use the census, and that's one of the primary uses of the census is to apportion populations to districts.

In addition to that, vote dilution, is illegal under the 14th Amendment. So it's a violation of political equality because you're not treating each voter equally. And in the case of racial gerrymandering, we science has developed statistical tests that can test the magnitude of equality or inequality, the amount of bias in racial voting to the extent that different racial groups vote as a block and to the extent that other groups tend to vote with or tend not to vote with those specific racial groups. And those are racial minorities and linguistic minorities that are protected under the Voting Rights Act.

Colleen: So who gets to draw these lines and when?

Mike: So, that's one of the big questions, and one of the major factors that my research has uncovered in terms of where do we find biased re-districting and gerrymandering? And the historic method of drawing districts has been left to state legislators, right? So as part of our federalist constitutional structure, as you may recall, the framers of the constitution couldn't even agree on what counted as a human being, much less a registered voter. And so the responsibility of administering elections was largely left to the states. And that means the state legislators are responsible for drawing electoral districts, both congressional districts, and their own districts. Basically, every 10 years has been the traditional deadline for re-drawing districts, and that's now mandated by the constitution and by the equal population requirements.

So we have a census, we get a count of how many people there are and where they are, and then district lines are re-drawn every 10 years to account for population shifts. Now, in some cases, states may take the action of re-drawing districts in the middle of the decade, right? So there's no requirement that states can only re-district every 10 years or they have to re-district every 10 years. I mean, now, they basically have to re-district every 10 years because of the population requirements. But one of the lawsuits in Texas that has to do with a racial and partisan gerrymanders in the 2000s, Texas, when the party control switched hands in the state legislature, the new Republican majority then decided that it was a good time to redistrict the state and re-draw congressional districts and other districts.

And you might recall from that period that the Democratic members of the state legislature literally left town. They left the state in order to prevent a quorum from being called so that they could test the map. So there's a lot of electoral drama.

Colleen: I mean, the reason that we do this is because populations shift. You might say, "Well, why don't you just leave it the way it is if it was good 10 years ago?"

Mike: Sure. Well, and that is indeed what lead to some of the grossest forms of discrimination in the south and in other states. Even outside the south, is there were some states that didn't change their districts for 40 or 50 years. And so you had massive inequalities, in particular, urban populations being under-counted, right? Because if you...One way to think about mal-apportionment is to think about the place where it's actually mandated by the constitution. And that's in the US Senate, right?

So in the US Senate, states are actually electoral districts and every state gets two senators. So the state of Wyoming which has fewer people that live there than live in Orange County, California. California has 40 million voters, but they both get two senators, right? So, California...If you vote in California, your voting power is diluted to about one 70th of the strength of someone who votes in Wyoming, who still gets their two senators.

And the apportionment process also determines how many congressional seats each state gets. Because since 1914, the size of the house of representatives has been capped at 435, and so that means as populations shift around, we have a minimum, right, the Wyoming rule, every state gets at least one representative. But then the larger states, Texas, California, right, they may lose or gain seats depending on how population shift.

Colleen: So it seems to me that there has to be a strong role for science here to help us figure out the best way to do this. And tell me a little bit about the research that you've been doing and the role that data plays in this.

Mike: Sure. So the role of social science has been central, frankly, to most of the judicial decisions that have been made enforcing the Voting Rights Act of 1965. So, all the racial gerrymandering and re-districting laws suits relied heavily on the use of expert testimony, of social scientists to actually demonstrate the harm that was done, and to demonstrate not just whether an in-equality exists or not, but to actually measure the extent of inequality, which allows judges to then answer the question, "How much is too much?" And so in the case of partisan gerrymandering, where we have political parties, a political party in control of a state legislature...and this is, frankly, where we find the worst gerrymandering, right? So when one party controls the entire re-districting process...and what that means, typically, is unified government. So you've got the governor, the state legislature all under the control of one party. So there's no veto point or there's no way that the minority can stop the majority from implementing whatever plan they want.

And those are the conditions under which we tend to find, surprisingly enough, the worst gerrymandering, right? And so we've got state legislators that are maximizing the partisan advantage that they have, largely in my view, and there's some controversy over this, but again, this is where social science comes in. My view is that there was a specific Supreme Court decision, the Vieth versus Jubelirer decision which basically in a 4/4 plurality ruling with Justice Kennedy siding with the Conservatives argued that there was no justiciable standard, right? So there was no standard that could be used to say how much is too much in the case of partisan gerrymandering.

And after that ruling, once the 2010 census was completed, and the 2012 districts were drawn, there was no reason of first-day legislators to look over their shoulder and be afraid of judicial reprimand. And so, what we saw in our research was that not only did we see states that changed party hands engage in extreme partisan gerrymandering, but even states where party control didn't change.

That is you had a republican majority in 2002 and you had a republican majority in 2012 or a democratic majority. We saw the partisanship and the level of gerrymandering become worse, even in those states.

So they really did, and we have pretty clear evidence that shows that they maximized their partisan advantage. And we do that through a statistical technique, right? We use what is known as the symmetry test, which is essentially a measure of how efficiently each party's votes are being allocated in terms of how many seats they get.

And so it looks for an inequality in that inefficiency. So we get disproportionate outcomes. And one of the reasons why we get disproportionate outcomes in any election is because we use single-member districts. And so it's not unusual in my home state of California, for example, the Democratic Party is the biggest party? They're the majority party and they tend to win about 70% of the seats with about 60% of the votes? So there's a clear dis-proportionality between the share of votes and the share of seats that they get. What the symmetry essentiality test is it says, "Well, what happens? Let's simulate an election where the Republicans get the same vote share as the Democrats did in this past election. Would they, too, receive that same level of bonus? Or if they didn't, then that suggests and it shows that the lines are drawn in such a way that both parties are not being treated equally."

And there are a number of techniques that have been developed over the last few decades that allow us to use this symmetry standard as a constitutional standard. That is, we can use this test to measure the degree to which one voter's vote is diluted relative to a voter who's voting for a different party.

Colleen: That raises two questions for me. One is why not take the parties out of it completely? And my second question is, why...I mean, it's different for every state, why shouldn't it just be the same methodology for all states?

Mike: Sure. So, on the one hand, you can't remove politics from what is inherently a political process, right? So, one of the, frankly, I would say it's a feature, not a bug of our electoral systems in the US, most of them, at least, is that there's dis-proportionality that I'm talking about, that's actually a feature of the system,?

So, that sort of dis-proportionality occurs because we elect members from single-member districts rather than multi-member districts, and using more proportional formulas which try to minimize the difference between vote share and seat share. And so because we use that type of electoral system which we just inherited from the Brits, that creates very strong incentive to manipulate the lines? If it didn't matter, if the stakes weren't so high, parties wouldn't be engaging in this sort of behavior. But because of this institutional structure, both parties have very strong incentive. And so to some extent, you can't take the politics out. What you can do, though, is you can remove the sort of naked partisanship that we're seeing when legislators are essentially allowed to choose their own voters?

So this condition where you've got one party in charge of the process...well, of course, they're going to draw it to maximize their advantage. And so what we have found is that in states that have adopted re-districting commissions, particularly re-districting commissions that either have a veto over a legislative plan or in states like California, you've got this very deliberative sort of a process where you've got citizen input, you've got people suggesting maps, and then a bi-partisan and...some would say non-partisan, but I'm not sure, I believe in non-partisanship. A bi-partisan commission then chooses a map. And so we have taken control over the map-making process out of the hands of the legislature. And that turned out to be a pretty good idea. Many of the states that were using re-districting commissions had some of the lowest levels of bias out of all the 2012 plans.


Colleen: So if you could wipe the slate clean and re-factor our whole system to make it better, to make it more fair, what would you do?

Mike: If I was king?

Colleen: Yes.

Mike: Well, an easy answer would be to simply move to proportional representation, right? I mean, this is what most of the worlds' democracies do, they have some sort of proportional system. But, of course, nothing is ever that easy, and as I tell my students, doing electoral reform is sort of like working on your car while you're driving down the road. That is, caution should be warranted, and so while I think that it's certainly possible, and we are seeing many reforms. So Maine, for example, is the first state to adopt ranked-choice voting, which is a type of more proportional electoral formula that still keeps single-member districts. But it does away with primaries which tends to produce more partisan candidates. It's less expensive, it's easier to administer in terms of not having as many elections, and it relies on people's rank choice of ordering candidates in the order that they would like to see them represented. And so it is more proportional in that sense.

There're a number of cities throughout the United States that have adopted ranked-choice voting. So you're seeing some of these reforms happen. One of the problems with having a sort of full proportional system, something like the Netherlands or Israel has, where the entire legislature is a single electoral district, and basically, whatever percent of the vote that a party gets, that's the percentage of seats that it gets. That might not work so well with a super-majoritarian governing structure like we have. That is, our system is really not designed for coherent, parliamentary-style parties. And indeed, one of the reasons that we're seeing so much dysfunction in Washington now is because we have these very disciplined, coherent parties that our system is not designed to operate with, right?

Our system is super-majoritarian in the sense that we have a broad separation of powers, we defuse power between state and federal governments. And within the federal government, we've got two chambers and a separately elected president. We've got judicial reviews, so we've got a judicial veto, so there are lots of places to stop legislation in our system. It is designed to work with loose coalitions and creating a proportional system where no party has a majority. So typically, in a PR-style system, you have a government that is built through a coalition of parties, right? So you'll have one party that gets most of the votes, 30%, 40% and they will build a government, right?

In a system that has a lot of veto points like ours, that could be rather disastrous, right, because you could have a lot of parties that are holding out because, basically, it would very much bias the status quos, right? So small parties that would rather see the status quo over-changed could basically hold our system hostage because it's very super-majoritarian. So it's a complicated answer...

Colleen: I can't imagine anybody holding our system hostage. That would never happen.

Mike: Right, we've never seen that.

Colleen: I wanna just unpack that a little bit. So, I go into the voting booth and let's say we had it for the, you know, for the presidential election.

Colleen: I'm going off on a tangent here why isn't there one voting system that is used everywhere?

Mike: So this is largely a function of where the authority rests to administer elections and that's in the states. And so, we don't have a single electoral system, we've got 50 electoral systems that operate under their own rules, and have their own bureaucrats that are responsible. I mean, I would go even further to suggest that a lot of the reason has to do, not just with state officials, but actually with local officials because it's the local county administrators and we have a very de-centralized electoral system in most states. And so the people making decisions over voting machines and over types of ballots aren't even necessarily the state officials.

Colleen: This is a hugely complicated problem to solve. Are there things happening that you feel positive about that we’re moving in the right direction despite the current trump administration?

Mike: Absolutely. I'm...Well, in addition to being an optimist by heart, I'm consciously optimistic about a number of decisions that are being made. So you've got these several cases having to do with partisan gerrymandering, which in the case of North Carolina and Pennsylvania you've already got decisions that have been made and standards that have been set and are now developing. The Supreme Court's going to come out with a decision on a Republican gerrymandering in Wisconsin and a Democratic gerrymandering in Maryland. And that's going to set the standard. One way or another, there's going to be a lot of activity after that decision is made.

Once again, we see science playing a really important role in terms of informing citizens and administrators about what actually works. And so one of the things that we see are that automatic voter registration is a way of increasing participation by having them the option to opt out rather than opt into registration. The more permissive and more open you make the system, the higher turn-out you tend to get. And we've already seen a few studies that have shown that automatic voter registration works. In addition to that, it also increases security because once you know who all your eligible voters are, you can manage that database much better than you can a database where you're trying to decide who's eligible and who's not. And so, the social science and the election science, frankly, is making a big difference in pushing reforms that we know work.

Colleen: Great. Well, Mike, it's been great talking to you today. Thanks for coming up here to Cambridge to hang out with us.

Mike: Well, thank you for your time.


Editing and music: Brian Middleton
Research and writing: Pamela Worth
Executive producer: Rich Hayes
Host: Colleen MacDonald

Related resources