November 17, 2019 📊 Obama is probably right about Democrats’ leftward drift

According to several studies, drifting too far left (or right) hurts a candidate's chances of victory

Welcome! I’m G. Elliott Morris, data journalist for The Economist and blogger of polls, elections, and political science. Happy Sunday! This is my weekly email where I write about politics using data and share links to what I’ve been reading and writing. Thoughts? Drop me a line (or just respond to this email). Like what you’re reading? Tap the ❤️ below the title and share with your friends!


Dear Reader,

This week’s main read: Former president Barack Obama has officially waded into the 2020 Democratic primary, suggesting that a candidate that is too liberal might hurt the party’s chances in next year’s general election. I think the scholarly literature affirms Obama’s argument, and I review the evidence in this week’s newsletter.

Plus, I’m trying out a new format for sharing the things I read and enjoyed this week. It’s shorter and includes only the cream of the crop. It’s also a bit more free-wheeling than past editions of this email. Do let me know what you think!

Thanks all for reading my weekly email. Please consider sharing online and/or forwarding to a friend. The more readers, the merrier! If you’re shy, the best way you can support my newsletter is to press the heart button below the title (this makes it rank higher in Substack’s curation). If you’d like to read more of my blogging I publish subscriber-only content 1-3x a week on this platform. Click the button below to learn more!


Obama is probably right about Democrats’ liberal turn

Political science research tells us that voters punish candidates who are more ideologically extreme than others

The New York Times published former president Barack Obama’s comments to a room of wealthy donors on Friday:

“Former President Barack Obama offered an unusual warning to the Democratic primary field on Friday evening, cautioning the candidates not to move too far to the left in their policy proposals, even as he sought to reassure a party establishment worried about the electoral strength of their historically large primary field.”

The party is drifting too far to the left in their policy proposals, Obama argues, pushing them further from their maximally-electorally-competitive position. This line of thinking is not entirely new; in some ways, the entire 2020 Democratic primary has been structured around the question of whether the party’s leading politicians are becoming too liberal for their own good. But it is still significant to hear this coming from Obama, who, you know, actually has experience winning the presidency.

I do not want to get into the squabbling over what exactly constitutes “too far” to the left for the Democrats. I prefer my debates over policy to happen in other venues, both online and off.

Instead, I want to make the claim that there is some risk to nominating a candidate who is more liberal than the replacement-level candidate. Simply, I think the chance that Democrats win the 2020 general election decreases as their candidate moves more to the left, all else being equal (which, of course, it’s not, but bear with me). Let us now turn to the evidence that supports this claim.

Political science has long established a punishment for ideological extremity in American election. The paper that gets cited most often these days is a 2015 piece in the American Political Science Review from Stanford political science professor Andrew Hall. Hall analyses election results and candidate ideologies from 1980 to 2010 and concludes that extreme candidates motivate opposing partisans to turnout against them, a tax that moderates don’t face. He writes:

“… general-election voters punish the nomination of extreme candidates from contested primaries, on average. In fact, the average electoral penalty to nominating an extremist is so large that it causes an observable ideological shift in the district’s subsequent roll-call voting in the opposite direction, towards the opponent party’s ideology. Primary voters thus face a clear tradeoff, on average, between supporting a more “electable” candidate or voting based on ideology and risking the loss of the seat to the other party.

Hall supports his analysis with a number of graphics that I’d like to share with you all. Figures 2 and 3 here show that extreme candidates who win slim victories in their primaries perform substantially worse than moderate candidates, and that this effect might be larger in elections with no incumbents (but not in uncompetitive districts).

Hall estimates a roughly 20 percentage point decrease in the probability of general election victory for a typical ideologue. This effect grows with extremity.

This is not all too surprising. The traditional theory of American elections states that candidates closer to the “median voter” will do better in elections because they have broader appeal (a mythical theory when applied today, by my reckoning—but I will save this superfluous and technical rant for another time). Occupying fringe positions across a spectrum of policies is likely to turn some people off of your campaign, especially relative to voters who aren’t so extreme.

However, one weakness of Hall’s work is that it ends in 2010. Still, we do have evidence that Democratic candidates for the US House suffered for their ideological extremity in 2018.

According to Emory political science professor Alan Abramowitz, Democrats who endorsed Medicare for All in the most recent midterm elections fared worse than those that didn’t. This relationship holds up to a variety of controls, including for partisanship. According to Abramowitz’s work, “support for Medicare for All cost Democratic candidates in competitive districts almost five points of vote margin — a substantial effect in a close election”.

All this analysis is not just limited to House elections. We do indeed have some evidence that Democratic voters are penalizing ideologues in the 2020 primary today. According to polling from the New York Times and Siena College, moderate Democrats beat more liberal ones across several general election battleground states:

Hall has expanded upon his original findings in a 2018 paper with Daniel Thompson, also at Stanford. In it, the two authors re-frame the moderates-vs-ideologues debate as a question of a candidate turning out their base, supposedly by taking ideologically extreme positions, or courting moderate voters (sound familiar?). They combine a study of survey data on voter turnout with data on candidate ideology and find that that extremist candidates who win by small margins under-perform their moderate counterparts particularly because they actually decrease their party’s turnout in a general election, “skewing the electorate towards their opponent’s party”.

This actually suggests to me that the entire media conversation about “electable” moderates vs ideological candidates is formulated entirely in the wrong fashion. While some analysts argue that a polarized political era means Democrats and Republicans can run more ideologically extreme candidates and not be penalized, and others argue that swing voters (and the courtship of said voters) are becoming more important to outcomes, the empirical evidence presented by Hall suggests a tangential theory entirely. Moderate candidates, this paper suggests, enjoy a turnout advantage over their supposedly base-riling ideological alternatives. 

Although the idea that polarization is driving campaigns to be more geared to turnout is correct, in this revised worldview, the consequences are opposite to what some Democratic analysts have posited so far. This doesn’t mean candidates should be more extreme in order to win, it means quite the opposite! 

Importantly, Hall suggests this is even more important when the extremity of a candidate is well-known. Of course, this suggests that the “ideology tax” likely increases for nationalized federal elections over localized US House and state legislative and executive elections.

…Elections such as, dare I suggest, the 2020 Democratic primary?

Given the evidence, it’s no wonder President Obama believes that a nominee will have to sacrifice some liberal positions in order to win enough votes to beat Trump in 2020. And anyway, didn’t the guy like do this twice already? Y’all should probably listen to him.

And here are some selected links to the work I read and wrote last week:

Posts for subscribers:

Links and Other Stuff

  • Nate Cohn had a good Upshot article to start the week in which he reviewed some of the findings from their polling that “might change the way you think about electability”.

    • One of those findings is that white working class voters have seemingly stayed entirely put in their political preferences since 2016:

      If I were a Democratic Party strategist I would be particularly worried about my party’s strength in Wisconsin right about now. Investing in places like Texas and Arizona is great if the electoral map is already tilting your way, but it looks like they’ll need resources in other states.

  • I also enjoyed this piece from FiveThirtyEight’s Perry Bacon Jr. on why Deval Patrick is jumping into the 2020 nomination contest.

What I'm Working On

I wrote this week about why poorly educated voters hold the keys to the White House:

I finished a long project on how musical preferences map onto Americans’ socio-political-cultural divides:

I also ran the numbers on what exactly has changed Donald Trump’s presidential approval ratings, advancing a theory that taking orthodox Republican positions (IE: unpopular ones) on policy has hurt him more than the more partisan squabbles or daily drip, drip, drip of political scandal. In the end, we think this tells us that impeachment likely won’t hurt him all that much so long as Republican elites stick by the president’s side.

Next week, I’m writing for you all on Uber and binge drinking and I’ll be continuing the slog of work that is the 2020 primary.

Until then,


Thanks for reading!

Thanks for reading. I’ll be back in your inbox next Sunday. In the meantime, follow me online or reach out via email. I’d love to hear from you!

If you want more content, I publish subscribers-only posts on Substack 1-3 times each week. Sign up today for $5/month (or $50/year) by clicking on the following button. Even if you don't want the extra posts, the funds go toward supporting the time spent writing this free, weekly letter. Your support makes this all possible!