2016: The Turning Point

Followup to "Why Is Everything Liberal?"

My article “Why is Everything Liberal?” has gotten a great deal of attention. See in particular thoughtful commentary from Bryan Caplan and Robby Soave at Reason.

This post is a followup, with two main goals. First, I’ve discovered additional evidence that liberals care more about politics, which I will just add on to what already was an extremely strong case.

Second, some people criticized the piece for not addressing what has changed recently. I think I’ve found the answer to that too, which is that the mobilization gap increased precipitously in 2016. It is at that time that we see Democrats overtake Republicans in fundraising, liberals overtake conservatives in signing petitions, and the left’s already sizable lead in protesting become much larger. While it seems that liberals have always cared more about politics if we are looking at the tail end of the distribution–i.e., those who become activists, journalists, or academics–it is only in 2016 that we see more noticeable and significant gaps open up in the next level down in the pyramid.

Since 2016, liberals have achieved true mass mobilization in a way conservatives never have in the modern era.

Additional Evidence for the Theory

Charles Fain Lehman (follow here) recommended I look at the ANES for more evidence for my argument that liberals care more about politics. The following chart is from 2016, and shows whether respondents had joined a protest or signed a petition in the last year (data in this post drawn from here for 2020, and here for all other years).

In 2016, fewer than 1% of conservatives had been to a protest in the last year, compared to 15% of extreme liberals, 10% of regular liberals, and 5% of the slightly liberal. Even moderates, at 2.4%, protested more than conservatives. Remember, this was before the Women’s March and the peak of BLM! The estimates for protest size used in the original post were pretty crude, but it’s nice to see self-reported data match what we see in the real world. Petitions tell the same story, but the differences are not as extreme: 61% of very liberal individuals had signed one in the last year, compared to just 26% of the very conservative.

I wrote the following in the last post:

People who engage in protesting care more about politics than people who donate money, and people who donate money care more than people who simply vote. Imagine a pyramid with voters at the bottom and full-time activists on top, and as you move up the pyramid it gets much narrower and more left-wing. Multiple strands of evidence indicate this would basically be an accurate representation of society.

This is actually not exactly correct, as it seems that more Americans protest than donate money, at least on the left, so donating money may be higher up on the pyramid for them. Regardless, as will be seen below, data on protesting, petitions, and donations are consistent with liberals becoming much more mobilized across a wide variety of fronts in 2016.

What Changed?

Few have taken issue with my conclusion that liberals care more about politics, but some critiqued my piece for not looking at what has changed over time. This is fair enough. As I wrote, my original essay had started as an investigation into Woke Capital, which has become a lot more extreme in the last several years. Other institutions besides corporations have also gone left, and this includes, well, everything: the dictionary, knitting clubs, board games, and literally almost anything you would think to call an institution. So “Why are corporations woke?” is still not an interesting question, but “What has changed in the last decade?” is.

My theory that institutions are captured by whatever segments of the population are most mobilized would predict that the mobilization gap between liberals and conservatives has increased over the last several years.

Luckily, the ANES has data on protests or signing petitions from four different election cycles. Unfortunately, the data are not always in a consistent format. For example, in 2012 they asked respondents whether they had been to a protest in the last 4 years, instead of the last 1 year as in 2016. Nonetheless, we can learn something about how the protest gap has changed over time.

Liberals already tended to protest more in the years leading up to 2012. But conservatives used to at least hold their own. This matches what we know from the real world, as this was the height of the Tea Party. Glenn Beck’s largely forgotten “Restoring Honor Rally” in summer 2010, for example, drew a lot of people, though nobody really knows how many. Wikipedia says “a scientific estimate placed the crowd size around 87,000, while media reports varied wildly from tens of thousands to 500,000.” This was also the time of Occupy Wall Street, so liberals weren’t exactly sitting on their hands, but conservatives at least made a showing. By 2016, conservative protesting had collapsed to practically nothing, while liberal protesting stayed at similar levels or, more likely, increased (hard to know for sure because of the time frame of the 2012 question being different).

Actually, the chart on protests above underestimates how much the mobilization gap has expanded, because it doesn’t take into account the fact that more people considered themselves liberal in 2016 compared to 2012.

The chart below shows weighted numbers, that is, how many people of each ideology had been to a protest in the last 1 year (2016), or the last 4 years (2012) in each survey, instead of the percentage of people within each ideological group who had protested. Unlike the data in the chart above, this figure takes into account how many people are identifying as conservative or liberal at the point of each survey. Note in the figure below that 2016 and 2012 not only cover different time periods, but also have different sample sizes, so one should not use the chart to draw conclusions about how mobilization has changed for each side across time. Nonetheless, we can compare liberals to conservatives within each survey.

What does this show? Here are the estimated ratios for total number of protesting liberals to each protesting conservative across the entire population for 2004, 2012, 2016, and 2020.

Liberal: Conservative Ratio

Protesting 2004 (last 1 year): 2.6

Protesting 2012 (last 4 years): 1.5

Protesting 2016 (last 1 year): 9.7

Protesting 2020 (last 1 year): 3.6

2004 was the height of the debate over the Iraq War. Liberals were more mobilized at that point, but by 2012 conservatives had largely closed the gap to a ratio of “only” 1.5:1. I feel confident in saying that pre-2016, the year 2004 should’ve seen one of the largest gaps of the last several decades given the debate over Iraq, so we can take that 2.6:1 ratio as the height of difference in willingness to protest throughout the oughts, and maybe the 1990s and 1980s too (Frustratingly, there is no 2008 question about joining a protest in the last year, or 4 years, only a question about whether the respondent has ever joined a protest, which isn’t very useful). By 2016, liberals opened up an advantage that was larger than ever.

The 2020 survey was in the midst of the pandemic, and liberals were more likely to engage in social distancing. Thus, the 3.6:1 ratio for that year should not be taken as conclusive proof that the mobilization gap has closed to any significant degree since 2016. (For ANES nerds, I used the post-election measure of ideology for every year but 2020, which only has a pre-election measure of ideology. This doesn’t appear to matter much in most places, except that if we use the pre-election question on ideology for protesting in 2016, the liberal:conservative ratio for that year is 4.2. So few conservatives have been protesting since 2016 that small changes in survey response can make a large difference for the liberal:conservative ratio. The important point to take away is that a relatively small protest gap in the years leading up to 2012 has transformed into a much larger disparity since 2016.)

We can also look at the “petition gap” through the same lens. Again, the data are not perfectly consistent, but we can calculate ratios at different points in time. As with protests, 2012 asks about signing a petition in the last 4 years, while 2016 and 2020 ask about the last 1 year. Annoyingly, 2012 has two questions about petitions, one on paper petitions and one on online petitions, while 2016 and 2020 ask one question about either kind of petition. Still, all the numbers are there, so I can calculate a paper petition ratio (2012), an online petition ratio (2012), and overall petition ratios (2016, 2020). They are as follows.

Liberal: Conservative Ratio

Paper Petition (2012, last 4 years): 0.9

Online Petition (2012, last 4 years): 0.9

All Petitions (2016, last 1 year): 1.6

All Petitions (2020, last 1 year): 1.7

In 2012, liberals were more likely to sign petitions than conservatives, but the gap was pretty small and there were many more conservatives in the country, which meant the right actually had more total people signing petitions. By 2016, more Americans than before were calling themselves liberals, and liberals were more mobilized, giving the left a substantial advantage.

Another thing we can do to see how relative mobilization has changed over time is to look at campaign donations. In the previous essay, I went all the way back to 2012, and showed that for every recent presidential election cycle Democrats brought in more money. I didn’t go back to 2008, as I was sure Obama outraised McCain, and I was of course right.

However, if you expand the analysis to midterm elections and all federal candidates, we see the Democrat advantage does not open up until 2016. Here are numbers I’ve gathered from Open Secrets for every election from 1990, as far back as data go.

In the following chart, I put together data on total funds raised for candidates for House, Senate, and President from 1990 to 2020.

Let me make a technical point here. In the last post, I took a shortcut in the analysis by conflating “more donors” and “raised more money” as if they were the same thing (feel free to skip this paragraph if you don’t care about campaign finance law and trust me on the point that they usually are). They tend to be, because individuals can only donate a certain amount to a candidate under federal law, with the cap right now being $2,900 (you can donate unlimited amounts to Super PACs that are officially unaffiliated with the campaign). This means that somebody can’t donate a billion dollars to a candidate and single-handedly tilt the numbers. Therefore, the side that gets more money usually has more individual donors. So in the discussion about 2016, I hadn’t actually confirmed who had more donors, only pointing out Hillary had raised more money. Turns out, she did in fact also have more individual donors, with Trump only winning the category of “small donors,” which means those giving less than $200. In other words, Trump donors were really cheap, consistent with the idea conservatives care less. Nonetheless, whether you want to take total money or number of individual donors as your measure of cardinal utility in 2016, the story is the same (see donor demographics for Trump and Clinton). So when the graph above shows one side raising more money in a particular cycle, it probably means they had more donors, though it is not a mathematical necessity that this must be true in every case, and it would be too complicated and mostly pointless anyway to check.

Age and Education Polarization Don’t Explain Woke Institutions

In response to my piece, Ezra Klein argued that liberal domination of institutions was better explained by age and education polarization than liberals caring more. This is an argument I’ve seen him make elsewhere before (see also this and this from Josh Barro on Woke Capital).

Romney won college educated whites by somewhere between around 5% and 15%, while according to CNN’s 2020 exit polls, Biden won the same demographic by 12%. CNN actually has Trump barely winning college educated whites in 2016 (48%-45%). Education polarization is real, and the fact that college educated whites vote something like 15-30% more Democrat than they did in 2012 should be having some effects on board rooms and the larger mobilization gap. Yet educational differences do not seem nearly massive enough to explain the total liberal domination of institutions, as Republicans hold their own well enough with degree holders.

As far as the age gap, it can cut both ways. When I was growing up in the 1990s, the stereotype was that retirees had a lot of time on their hands and were therefore politically powerful, while young people were largely indifferent. Old people certainly have more money, and so you’d expect age polarization to actually give Republicans an advantage in donations. Yet since 2016 the trend has been the opposite. As parties have polarized more by age, Democrats have started winning the competition over fundraising. Maybe young people are inherently more likely to protest, but wouldn’t you expect old people to be just as capable of signing petitions? Thus, I’m pretty confident that age and education gaps are less important than the simple fact that liberals care more about politics.

People are often surprised how well Republicans have maintained their vote share among high earners in the Trump era. Here are the results of one exit poll from 2016 and 2020.

According to the graphic, in 2016 Trump practically tied Clinton among those making more than $100K a year, and won this group by a substantial margin 4 years later. There’s therefore little evidence that the mobilization gap or woke institutions can be explained by socioeconomic differences. Republicans are not too poor to donate as much money as liberals, and signing a petition seems a pretty low cost way to become involved. Rather, liberals simply care more, and the gap between the parties in caring appears to have gotten larger over time.

2016 is When Everything Changed

The left has always had an advantage in committed activists. Yet, no matter whether you look at donations, protests, or signing petitions, the mobilization gap increased in 2016. Liberals had always protested more, but in 2016 the ratio was absolutely massive, being around 3.7x larger than it was around the time of the invasion of Iraq. This was before an upsurge of liberal protest activity that has included BLM, March for Our Lives, and most importantly, the Women’s March. Finally, the parties raised about an equal amount of money from 1990 until 2016, when Democrats took a lead that has now lasted three straight election cycles (2008 was an exception to the rule of parity in the pre-2016 era, when Democrats ran a fresh faced Barack Obama against John McCain, who seemed good at exciting Republican elites and MSNBC pro-war centrist types but not actual voters).

So what about “Woke Capital”? In many ways, business was the last domino to fall. Yes, liberals have always had more noisy activists, and corporations tended to bow to them on some issues when they got really agitated, like MLK day. But big business is more directly answerable to a wider swath of the population than are schools or non-profits, and so held out the longest. Coca-Cola and Walmart care more about what the median citizen thinks than does Harvard, The New York Times, or the ACLU. Yet after 2016, when the mobilization gap exploded, almost nothing in society could remain neutral, and pressure has come from both within and outside corporations for them to take a stand on almost all hot button issues.

Why was 2016 the year everything changed? Take a wild guess.

Up to this point, I have carefully gone out of my way to avoid mentioning a certain Orange Man, letting the numbers speak for themselves. But the conclusion is unavoidable.

It would be wrong to say that Trump hasn’t mobilized a lot of people on the right too. Republicans raised 1.76x more money in 2020 than in 2012. Over the same time period, however, Democrats increased their fundraising by 2.1x. Both times he ran, Trump got more total votes than Mitt Romney. In 2016, that support was well distributed enough for him to barely win the presidency, when Hillary Clinton actually got fewer votes than Obama did in 2012. (I don’t speak of the Trump era in the past tense, because we shouldn’t assume that it’s over. As of this writing, Trump is leading the betting markets to be the Republican nominee for 2024, and I think as a former president with a substantial hold on the party’s base he is massively underpriced at 18%. But that’s an essay for another day.)

Just as the previous post raised further questions, this one does too. The most interesting thing to me is not simply protests and donations, but why one side has for over half a century now drawn more idealistic people who want to dedicate their lives to changing the world. The journalist-academic-activist complex is ultimately where power lies, and it has grown much stronger in the last 5 years because it has started to engage many more people at the intermediate level in the mobilization pyramid, among those who give money, sign petitions, and go to protests, and who find themselves between true elites on top and the mass of the largely indifferent voting public at the bottom.

If the rise of Trumpism explains the last five years, why did the left begin with such a strong built-in advantage? I hope to explore this question soon.

The irony of the Republican turn towards Trumpism is that it was supposed to make conservatism stronger by broadening its appeal among a wider swath of the population. That it has, but it may also be responsible for a backlash that has overwhelmingly shifted the culture to the left (I say the “Republican turn towards Trumpism” instead of “turn towards populism” because I doubt there has been a move towards anything that can be called populism on the right that is distinguishable from support for Trump as an individual).

Moreover, right-wing protest culture has collapsed since the time of the Tea Party. It’s hard to know for sure, but other forms of conservative activism may have fallen off too. So even the degree to which Trump has actually mobilized the right must come with a caveat: he has turned out more Republican voters and gotten more people to donate small amounts of money, but few seem to want to make more substantial sacrifices, even compared to 2012.

Overall, the Trump era has provided mixed electoral results for Republicans. They won unified control of government in 2016, lost the House but kept the Senate in 2018, and came extremely close to winning again in 2020. Yet it has been an awful 4 years for conservatives who care about controlling institutions, or at least keeping them neutral, although even here it hasn’t been a complete loss. After all, the Trump era has given conservatives a comfortable majority on the Supreme Court, probably the most important single institution of all.

Federal court appointments last until death, while the widening of the mobilization gap is relatively new. Best case scenario for Republicans is that Amy Coney Barrett and Brett Kavanaugh live for a very long time, while the Trump era ends up being an anomaly in mobilizing the left to an unusual degree, with things going back to something resembling the pre-2016 historical norm. Worst case scenario is that things continue as they have for the last 4 years, with anti-Trump hysteria combining with the Great Awokening having created a class permanently mobilized for confronting racism and other evils, plus Republicans not even getting the mobilization on their own side that Trump gave them. A generation shaped by the experience of Trump and a party currently led by such uninspiring figures as Kevin McCarthy and Liz Cheney may end up giving conservatives the worst of all worlds.

[Note: This is the second in a series of articles on woke institutions. See the previous article “Why is Everything Liberal?” and the followup “Woke Institutions is Just Civil Rights Law.”]