> ## Content Index
> Fetch the complete content index at: https://www.mastersinpublicaffairs.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Polarization, by Nolan McCarty
- URL: https://www.mastersinpublicaffairs.com/episodes/polarization-by-nolan-mccarty/
- Published: 2026-10-14T13:26:17.000Z
- Updated: 2026-10-14T13:26:17.000Z
- Description: Polarization changes who we engage, how we engage, and how that engagement is perceived.
- Author: Joseph Lavoie
- Tags: Polarization, Nolan McCarty, 2019, #episode, #ai-4

Nolan McCarty is a political scientist at Princeton who has spent more than 20 years measuring how far apart the two American parties have moved. This book is him taking the common explanations for polarization one at a time and checking each against decades of data. You are probably familiar with many of these explanations: Polarization is caused by redrawn electoral maps, party primaries, cable news, the internet, corporate money, and the voters themselves. It turns out, most of these are wrong.

The finding underpinning the whole book is that American polarization started at the top. Elected officials, activists, and donors moved apart in the mid-1970s; voters didn't change until the 1990s, and when they did, most didn't become more extreme. They sorted themselves into whichever party already agreed with them. In fact, the share of Americans giving extreme answers on the issues hasn't moved in 40 years.

So what gives? That’s what we cover in this episode.

## What We Cover in the Episode

- My own story from the pandemic: someone close to me who held my views on COVID until the day places of worship stayed closed while bars reopened, and then changed sides almost overnight. It kick-started a growing obsession I have with this topic.
- Two votes 45 years apart. Medicare passed in 1965 with a majority of House Republicans and 14 of 32 Republican senators voting yes. Obamacare passed in 2010 with no Republican votes at all.
- Polarization, partisanship, and sorting: three different things that produce the same picture in a poll, and why we shouldn’t conflate them.
- The immigration example. The gap between Democrats and Republicans on whether immigrants strengthen the country went from 2 points in 1994 to 42 points in 2017, while the whole country moved in the same direction, from 30% to 65% agreeing. Nobody's views had to move for that gap to open.
- The partisan overlap: in the 1960s and 70s, some of the most liberal members of Congress were Republicans and some of the most conservative were Democrats. Today every Republican sits to the right of every Democrat.
- Why the same district elects a liberal or a conservative depending on who wins, and why that means redrawing the maps can't fix it.
- The four incentives that pull a legislature apart: individual donors now supply 80% of a candidate's money, up from 20%; moderates have stopped running; an opposition that can win the next election has no reason to cooperate; and parties too weak to protect members from a primary challenge.
- Four mental models: before you believe a gap, check the middle; one district, two very different representatives; weak parties, strong partisanship; and the system decides what polarization does to you.
- Three misunderstandings: "it's worse than ever and I can feel it," "the algorithms did it," and "the voters did it." On the algorithms, what the browsing data from 2004 to 2016 shows, and why the Americans who hardened fastest were the ones least likely to be online.
- The Canadian numbers, from a July 2025 survey of 2,250 Canadians by Digital Public Square and Abacus Data, and where I think polarization does its damage in a parliamentary system.
- The mastery lesson: on polarization, gather the data to understand the environment rather than to make a case, and the hopeful reading of what that data shows.

## Bonus: ideas that didn't make the episode

**Uncooperative federalism**

McCarty's chapter on what gridlock does to government ends in three places: the president acting alone, the courts, and the states. For example, Congress never changed the federal ban on marijuana; the states legalized it anyway and the Justice Department decided not to prosecute. On the environment, tobacco, and immigration, states sued the federal government and shaped national policy by winning. Two legal scholars, Jessica Bulman-Pozen and Heather Gerken, call this uncooperative federalism. For a practitioner it's a map of where the work moves when the legislature stalls: to the people who run departments and to the lawyers.

**Local news makes legislators less partisan**

James Snyder and David Strömberg compared members of Congress whose districts line up with a local television market, and so get covered, against members whose districts don't. The covered members do more constituency work, bring more money home, and vote in a less partisan way. McCarty reads the decline of local news as a likely contributor to polarization. Of course, we have seen this decline in Canada and to a lesser extent the UK, which has me wondering if this is a contributing factor in these two countries as well.

**Inequality, and the restraint McCarty shows about it**

The strongest pattern in his data is that polarization in Congress and income inequality have risen and fallen together since the introduction of the income tax. He writes that the relationship might run the other way, with polarization causing inequality by stopping government from acting, or that both might be caused by something else, and that nobody has been able to tell. 

**Ranked ballots and the Australian Senate**

McCarty reviews the reforms people hope will reduce polarization: proportional representation, ranked ballots, multi-member districts. His verdict on ranked ballots is that they make campaigns less negative and voters happier, and not much else; Australia's lower house uses them and is still a two-party system. Australia's Senate, which uses multi-member districts, had seven small parties holding 20 of 76 seats after 2016\. The broadening of representation comes from the multi-member districts, not the ranking. His closing advice to reformers: sell these changes on their own merits, because promising they'll reduce polarization won't hold up against the evidence.

##