Receive our newsletter - data-led analysis, original reporting and insights
Government / Elections

How winner-take-all elections keep urban voters at a disadvantage

It's not just the US: left-wing parties suffer in many countries because their voters are heavily concentrated in urban areas.

The anti-democratic institutions of the US political system have been thrown into harsh relief over the past five years. 

President Donald Trump won without the popular vote. The Republican Party’s grip on the US Senate is enabled by states like Wyoming and Mississippi, which each get the same number of senators as California, a state with a population almost as large as the 25 smallest states combined. As a result of all this, hard-line ideological conservative judges now have an overwhelming majority on the Supreme Court. 

a pickup truck with a Trump sign and protesters in front of the Pennsylvania capitol building
Conservative protesters in front of the Pennsylvania capitol building in Harrisburg, where Republicans have controlled the legislature for much of the post-war period despite routinely losing statewide elections. (Photo by Mark Makela/Getty Images)

But the US political system is stacked against the Democratic Party, as it is currently constituted, because it is also stacked against city residents. The bulk of Democratic voters are heavily concentrated in dense urban centres, while districts for the House of Representatives, and most state legislatures, often divide urban populations and dilute them with rural or exurban interests. Even without gerrymandering, it is very hard to draw legislative maps that allow competitive elections. 

Unlike the US’s other political dysfunctions, this one is shared by most of the other democracies that descend from the British Empire. Geographically based, winner-take-all districts are why the British Labour Party has been out of power for the great majority of the UK’s post-war history and why Canada’s left-of-centre parties have often been relegated to the sidelines despite winning majorities of the vote. 

Stanford University professor Jonathan Rodden wrote Why Cities Lose in 2019 to explain why winner-take-all districts disadvantage left-wing parties, which have been anchored in urban areas since the 19th century. In the run-up to the 2020 US elections, City Monitor spoke with him about what this inequality means for US urban transportation policy, whether the Republican alienation of suburban voters will be a lasting phenomenon and New Zealand’s embrace of a fairer legislative electoral system. 

Over the course of the 2020 election, we’ve heard a lot about anti-democratic biases baked into US democracy – the Electoral College, the Supreme Court, the Senate. But your research shows there’s yet another layer to that. Why are legislative elections in the US – and the UK, Canada and Australia – biased against left-wing parties?

My interest is in countries that use a single-member-district, winner-take-all system. This includes Britain and several of its former colonies, including the US. The problem for parties of the left is that they’ve become urban parties since the Industrial Revolution, and they’ve only become increasingly so since then. This is especially true in the US, where the correlation between population density and Democratic voting is extremely high. 

Democrats are highly concentrated in urban districts, and they end up winning there by large supermajorities, in some cases 80% of the vote. On the other side, in the most rural districts, Republicans tend to win by smaller margins. What that means is that for Republicans, it’s possible to win a legislative majority in some cases without winning 50% of the votes.

You argue that, contrary to popular belief, this is not solely a gerrymandering issue. You point to the Pennsylvania State Senate, which in the last century has been under Democratic control only in the late 1930s, the 1970s and in 1993 – despite the Democrats’ winning a majority of the vote over half the time. And the Democrats were actually drawing the legislative maps a lot of that time!

We can look at lots of cases where gerrymandering is an important part of the story. But one of the things we can do when we look at computer simulations of districts is draw lots and lots of potential districts for a state like Pennsylvania. We can see that even if we leave partisanship out of the equation and we just draw compact, contiguous districts, and we draw thousands of them, we can see that the compact geography of Democrats and the diffuse spread of Republicans is such that the Democrats will fall short in the transformation of votes to seats. 

Another useful way to understand this is to go back over time, where we can look at a state like Pennsylvania in periods when the Democrats had the upper hand in the redistricting commission, and we see that this bias was relatively similar. And it’s been around ever since World War Two. It’s even harder to stick with that story when you see something similar in Canada, Australia and the UK, where districts are drawn by commission.

This bias has been apparent for a while, and many wealthy democracies tried to fix their systems to account for it. How did continental European political countries try to make their systems fairer to get away from this anti-urban bias?

Continental Europe had a similar system to the UK in the 1800s and early 1900s. But in a brief period around World War One, most of Europe transitioned from a single-member, winner-take-all system to proportional representation. This involves drawing larger districts and having multiple members in the seats, and those seats would be allocated proportionally to the parties according to their vote share. In European democracies, a party with 32% of the vote will have around 32% of the seats. 

At the end of your book, you highlight all the research showing that countries with proportional representation have stronger environmental regulation, healthier labour unions and more robust social safety nets. You note that Republicans in the US routinely vote against national transportation projects like high-speed rail or funding for urban mass transit in a way conservative parties in continental Europe rarely do. 

The whole nature of debate in the US has been structured by this urban-rural divide in the political process. We’re accustomed now to the idea that the party of the right would be opposed to any kind of investment in mass transit. What I think is so interesting about Europe is that this is not at all the case; this is not the way political battles are fought in Europe. There are parties of the centre right that win in many urban neighbourhoods throughout the continent. It would be bad for their constituents to push against mass transit, so there’s more of a cross-party coalition in favour of investment in urban infrastructure in these proportional democracies.

What can be done to make this element of the US political system fairer? We are hearing ideas about getting rid of the Electoral College, reforming the Supreme Court, adding states to balance the Senate. What about this issue?

In the hierarchy of problems, this one has received a lot less attention. There’s a lot of people who believe this to be primarily a problem of gerrymandering. I think they will continue to work on redistricting reform. But the idea of a larger-scale reform like proportional representation isn’t one that lots of people, at least within the Democratic Party, are willing to consider. 

When a party comes to power on a big wave of support – even a party that has been very critical of the existing system – once they get into power using the existing rules, they become more supportive of those rules. This is something that we have recent experience with in Canada. A big part of the Liberal Party’s campaign in the last election involved electoral reform. Then, once elected, they put together a royal commission, thought about some proposals and came to the conclusion that the existing system actually works fine. So I’m not optimistic that it will be in the interest of anyone in the Democratic Party to push for large-scale electoral reform. 

But it is worth noting that some of this problem is becoming less noticeable in recent elections. A lot of this gets papered over when you have a big blue wave. As the Republican Party has lost support in many suburbs, the entire problem we’ve been talking about may be diminishing. In that case, Democrats might come to the conclusion it’s not a problem at all anymore. Over time, they might be correct in that assessment. As the political geography changes – and it’s really changing quite dramatically – the problem may be going away. Or it might just be going into hibernation for a little while and then it shows up again a few years later. 

The BBC's exit poll results from the UK's 2019 general election are projected on a building
The BBC’s exit poll results from the UK’s 2019 general election are projected on a building in London. (Photo by Tolga Akmen/AFP via Getty Images)

That’s the real question for observers of legislative elections over the next couple of cycles. Is what we’re seeing in many suburban areas a short-term phenomenon? Or is it a longer-term realignment? Consider the experience of the Labour Party in the UK, where for a long time they had a very pronounced problem of overconcentration in its core urban support districts. It was very difficult for the Labour Party to win in the suburbs. But then, under Tony Blair, they were able to do that, and everyone proclaimed Labour’s geography problem to be a thing of the past. But it really only lasted a couple of election cycles, and then things went right back to where they were. Labour became a party that once again had trouble winning outside of its core urban bastions. 

You could imagine that after the Trump era is over, the Democrats’ geography problem could come back. That’s the big unanswered question.

Is the huge wave of suburban support for Democrats a question of anti-Trump sentiment or of Republican Party policy?

A lot will depend on what the Democrats’ platform becomes over the next decade. Because the Democrats have a perpetual battle on their hands – like labour parties in Britain or Australia – between their core, urban left-wing districts and the crucial suburban districts that they need to win in order to form legislative majorities.

It can be hard, once you form a majority, to tell your core supporters to always compromise. But if the party platform is perceived as moving too far towards the urban core, then they could easily be back in the position where they find it difficult to win crucial suburban districts.

But the dynamics of this presidential election are almost perfectly suited to cut against that dynamic. On the Republican side, the entire strategy is that Trump is trying to get out even more of his base than last time. He’s trying to rev them up even more, potentially alienating suburban moderates – while on the left, the Democratic primary ended with the most moderate candidate winning.

What’s especially destructive for the Republicans about this rural-based mobilisation strategy is that giving up that geographic advantage in the suburbs is like killing the goose that laid the golden egg. That could be a very costly change for the party. Is this just a short-term reaction to an unpopular candidate or will a generation of voters now view the Republican platform through the lens of the Trump era? That’s a harder question to answer. But the Republicans recovered from Nixon relatively quickly, so it may not be the case that this is something that lasts forever. 

You also write about how this anti-urban political dynamic is changing through new housing and population trends. States that are very sprawling, like Arizona, aren’t like Pennsylvania. In Phoenix, you say there isn’t a partisan identity that correlates with density in the way that there is in the Philadelphia area. Then, in wealthier, knowledge-economy cities, you have people being pushed out by high housing costs, Nimby regulations that make it hard to build new homes. But that’s having the effect of pushing Democratic voters to previously Republican locales. How much does that relate to recent Democratic success in the suburbs?

Suburbanisation is still an ongoing process that’s been happening for decades. In a lot of cities, it is a process that started with whites and now, in an era where housing is less segregated, now involves African Americans and other minorities. There is a suburbanisation of minorities going on even in some of the old industrial cities. This means that the very Democratic core, some parts of that are moving to the suburbs, and these areas are more heterogeneous. 

The other process is that people are moving away from high-priced, knowledge-economy cities, where they’ve been priced out, and moving to other places with good labour market opportunities, where the housing is cheaper. Many of these places are in the Sun Belt: Austin, Houston, Texas, Phoenix, Orlando. Many of those newer cities don’t have that same concentrated urban core, like the old industrial cities, and they tend to be much more of a chequerboard of Democrats and Republicans all spread out throughout the metro area. That problem of concentration of Democrats in the urban core is less pronounced in some of those places. 

The thing that gets all the attention is the Electoral College. Is Texas going to go blue? Is Georgia? What about Arizona? But to me, the more interesting and subtle question is, are Democrats becoming better distributed in states where they might actually win more House and state legislative seats? There is talk in Texas about the possibility of Democrats’ actually gaining control of the state legislature. That’s very much an issue of political geography.

In terms of even bigger reforms, you wrote a case study of the one English-speaking democracy that switched from this conservative-favouring, winner-take-all district system to proportional representation. How has that affected who wins elections in New Zealand, and how has it shaped policymaking?

Throughout the post-war period, the parties of the left in New Zealand, and the Labour Party in particular, suffered in the transformation of votes to seats. They even had two consecutive elections where they won the popular vote but had fewer seats. It’s like having two US presidential elections in a row where one party wins the popular vote but loses the Electoral College. In New Zealand, it really led to a strong movement in favour of reform in the 1990s.

New Zealand Prime Minister Jacinda Ardern at a podium
New Zealand Prime Minister Jacinda Ardern’s Labour Party had a blowout victory in the October 2020 elections. (Photo by Hannah Peters/Getty Images)

Before reform, the parties of the left were in government much less frequently than parties of the right even though they did pretty well in terms of votes. But after reform, it’s been pretty evenly split. With the Labour Party winning a big majority in the government in October’s election, it will then be the case that the left has been more powerful than the right in the period after the reform. But I think the best way to see it is really that it’s competitive. 

We can expect to go back and forth between the two major parties, usually with some kind of coalition partner. This is rare that the Labour Party was able to win an outright majority, because usually when you switch to proportional representation, one of the reasons the mainstream party of the left doesn’t like it is because it makes more room for Greens and other smaller parties they have to form a coalition with. But this was a very successful election for the Labour Party, and they are able to form a single-party government.

Jake Blumgart is a staff writer for City Monitor.