Most discourse on the issues of democracy is plagued by this phenomenon (from Przeworski 1999), although usually asserted in the form "Democracy *requires* X" or "This threat to X is a threat to democracy". Political scientists are as guilty of this (maybe more?) as everyone else

It looks like Mamdani will win. One underrated but obvious reason why RCV doesn't matter much: it usually produces the same winner as plurality vote. Given that Mamdani is leading in the first round, he very likely would have won in a normal plurality election too.

RT @MargRev: The Effects of Ranked Choice Voting on Substantive Representation

RT @qjps_editors: The new issue of the Q is now out. We have great papers from @David_J_Hilden, @MichaelRKistner; Karekurve-Ramachandra; @J….

Check out the paper here: Here is an ungated version:

It's also possible that we would need to see bigger electoral changes (multi-member elections, parliamentary system) to get big policy change. That's possible, but we'll have to wait and see.

My broader takeaway from this project is that electoral rules aren't the reasons why we have the policies we have. The reason we don't have socialism or any other ideological program is due to what voters and elites want. Those things can change—but RCV won't change them!.

So while RCV might change the outcome of an election here and there (maybe it will push Mamdani over the finish line!), it's not likely to lead to systematic effects on the ideology of the people we elect or the policy outcomes we get.

But maybe RCV improves representation? That is, maybe cities wanting more liberal policies move left and those wanting more conservative policies move right? I don't find any evidence that RCV improves policy relative to public opinion either.

The ideological effect of RCV is smaller than the shifts that >99% of non-RCV cities naturally saw from 2012 to 2020. *At most*, the effect size of RCV is still below the average shift in non-RCV cities b/w 2012 and 2020. These are very small effects!.

I look at whether a) city councils change ideologically and b) fiscal policy changes after a city adopts RCV. I don't find evidence that either claim is true.

New York City is using RCV tomorrow to choose its nominees. Over the past decade, reformers have hoped that RCV would improve representation. Perhaps it helps someone like Mamdani win?. My new article in @qjps_editors (linked @ end) finds that RCV doesn't improve representation.

RT @joshmccrain: Was on a panel here a few years ago w/ a bunch of RCV advocates, and it was a huge shock to everyone when I said a) empiri….

RT @leedrutman: Yet another null finding for RCV's effect. Time to accept that RCV is too marginal to make any meaningful difference on pol….

What can we conclude from this? Voters' policy attitudes are not just a function of their partisanship; they develop from previously-held principles. Additionally, voters can use their values to figure out where they stand on new or hard-to-understand policies.

Finally, I use cases of new/evolving policy domains: welfare reform in the 1990s and transgender policies in the 2010s. I find that values shaped policy attitudes; for example, traditionalism in 2016 predicts anti-trans attitudes in 2020 more than 2016 partisanship or ideology

When we look over time, we see that people's policy stances 2 or 4 years later move to be in line with their previously held values

I do find that there is a high degree of omitted variable bias: using individual-level fixed effects reduces the effect size of values on issue stances, but they remain significant. They're also larger than the effect sizes of partisanship & ideology

We also want to compare these effects to other drivers of attitudes. I compare values' effects to those of partisanship and left-right ideology.

Past cross-sectional work finds that political values such as traditionalism & egalitarianism correlate with policy attitudes. But it's possible that these findings stem from omitted variable bias. I use panel surveys to test the relationship over time.

A short thread on my article which just came out in @BJPolS. I use panel studies to find that values drive the development of policy stances, and these effects are at least as large if not larger than the effects of voters' partisanship or ideology.