My friend Micah Sifry has a must-read article in The Nation entitled “How the Sanders Campaign Is Reinventing the Use of Tech in Politics.” He interviews two key staffers, Zack Exley and Becky Bond, who reveal a lot about the way their campaign has engaged its supporters.

I’ve written before about an oscillation: campaigns go back and forth between using technology to empower volunteers and accumulating Big Data to make their centralized outreach efforts more precise. Bond is explicit about which direction the Sanders campaign wants to move:

We’re shifting the focus away from a small number of sophisticated data and technologists engaged in a kind of Election Day arbitrage that ekes out incremental advantages by using micro-targeting algorithms to identify and turn out voters based on a model. Instead, we’re putting hundreds of thousands of volunteers to work, and in some states have literally called every single voter who will pick up the phone to identify everyone who supports Bernie or is undecided. Then we have other volunteers persuade the undecideds and turn out those who indicated support.

The article repeatedly describes the campaign as a “movement.” For instance, Exley says, “When Claire and I first arrived at the campaign, we knew that a movement was already way out ahead of the campaign. We believed it was our job to set up structures and tools to … help grow the movement.” A campaign fueled by volunteer hours and small donations that encourages its activists to recruit and lead certainly has a movement “feel.” But what would qualify the Sanders campaign as an actual movement–or as part of one?

Some would say that it’s already a movement because it has engaged a lot of fired-up people in unpaid political activity. Exley describes “a massive volunteer organization that’s making more than 1 million calls every day right now, knocking on countless doors and doing so much more.” Those accomplishments are typical of big, grassroots-based campaigns–not only partisan electoral campaigns but also bursts of grassroots energy in civil society. According to the late Charles Tilly and his colleagues, such campaigns are components or activities of social movements. But one campaign–even a large one–does not itself constitute a social movement.

Others would say that Sanders is part of a movement because he belongs to a loose, evolving, open network of academics, cultural figures, union leaders, organizers, and a few politicians that originated in the New Left and that supports democratic socialism in the United States. Not only is that network called a movement, but it is sometimes called The Movement–as in, “I grew up in the movement, you know,” or “I got to know Bernie through the movement back in the ’70s.”

I personally do not identify with this network, in part because I haven’t done anything worthy of admission to it and in part because my actual political beliefs are too eclectic. (I am not sure you can love Hayek and be in The Movement.) But I’ve known and admired bono fide participants all my adult life. The questions are … Does this strain of political thought and activism really qualify as a movement, even as it has spanned multiple decades? Has it shown enough signs of motion to be a movement? And how much of a movement activist is Senator Sanders? My sense is that he has been a solo voice on important issues, but not much of a movement-builder. He is not known for training organizers or leading organizations. As a voice and a vote in the Senate, he may be an asset to the movement–much like a noted author or musician who supports the cause–but I’m not sure he’s a movement person.

His campaign could nevertheless be an important episode in a movement that spans a longer time horizon and that has many more leaders than Bernie Sanders. It’s too early to say whether that’s the case, because everything depends on what happens after the 2016 election.

Another question is what movement his campaign is part of, if it turns out to be part of a movement at all. Sanders’ own roots are in ’60s-style US-based democratic socialism (see the Port Huron Statement), but other currents are feeding his campaign. Bond says, “First of all, I want to take this opportunity to say that the movement to defend black lives is fundamentally changing the terrain of social-change organizing. After recognizing that, yes, the young people and working-class folks, many of whom are from communities of color, who are leading the movement behind Bernie Sanders as volunteers on the ground are changing American politics.” That comment sounds somewhat aspirational to me–Sanders would be closer to the nomination if he had engaged the Black Lives Matter movement more effectively. But a large coalition could still form after his campaign concludes. Influenced in part by Tilly, I’d look for these features as evidence that a movement is afoot:

• A set of campaigns–such as the Sanders presidential run and the civil disobedience in cities like Ferguson–that gain rather than lose momentum over a span of years and that look increasingly interrelated.
• A characteristic repertoire of political acts, which might encompass everything from viral “memes” on social media to people shutting down highways.
• A diverse, not completely consonant, yet overlapping and interacting set of prominent leaders, some involved in politics and some outside of it.
• Cultural manifestations, such as very popular music in support of the cause.
• A set of increasingly specific demands that begin to be implemented by major institutions.

See also: questions for the social movement post Ferguson.

Political scientists and data-crunchers were almost unanimous in their authoritative predictions all summer and fall that Donald Trump was going nowhere. (Daniel Drezner has a nice summary.) Meanwhile, several political theorists and political philosophers were alarmed by Trump from the start (e.g, Jason Stanley, and others whom I follow on social media). It seems they were right, so score a point for political theory. But this case actually reveals interesting strengths and weaknesses of two ways of thinking about politics.

Empirical social science is based on data, which is by definition from the past (although sometimes including the very recent past, like this morning’s polls). To the extent that it is predictive, it derives patterns or trends from what has already happened. That is a very broad definition that can encompass research at any geographical or historical scale. It can describe research that is meant to be descriptive or predictive of the existing regime and power structure or research that looks for openings for radical change. So it’s unfair to stereotype political science. However, there is a dominant style of research on American politics that has the following features:

• It focuses on the US case, presumably because empirical generalizations are difficult across national lines. Trump strikes me as highly similar to current European right-wing leaders. Mainstream political science could explore this resemblance, but it would be very hard to incorporate data from Europe into a predictive model of US elections.
• It restricts itself to recent political history, because models of election outcomes that include data from distant times are irrelevant. But, as often acknowledged, a study of presidential elections since 1960 or since 1972 is based on a problematically small number of cases. Just as elections have changed fundamentally at turning points in American history, so they can change again.
• It discounts the significance of rhetoric and narrative, because empirical studies of the impact of discourse usually find small results. For instance, one can usually predict the results of a presidential election based on economic conditions a few months earlier. Likewise, the presidential bully pulpit is found almost never to affect public opinion. Such research suggests that rhetoric and ideological positioning are unimportant. Yet a broader look at the differences among regimes (and among eras in our own history) make ideas and ideologues look central again.
• It discounts the impact of Margaret Mead’s “small groups of thoughtful and committed citizens,” because empirical research typically finds larger effects from demographic changes, market conditions, and other impersonal forces. Yet Nate Silver calculates that Trump won just 2.0% of the eligible adult population in Iowa, 9.7% in New Hampshire, 6.5% in South Carolina, and 1.8% in Nevada. That’s why he’s winning the nomination. Silver adds, “A few passionate supporters can go a LONG way.”
• It takes the basic structure of the regime as a given. We have, for instance, a two-party system with privately funded elections and a certain ideological spectrum. But obviously, the regime could–and probably should–change.

Although you can study the current regime empirically with a critical intent, I think focusing tightly on the way things actually are creates a bias in favor of the status quo. It makes the discipline conservative. Theodore Lowi concludes his great book The End of Liberalism (1969, revised in 1979) by saying:

Realistic political science is a rationalization of the present. The political scientist is not necessarily a defender of the status quo, but the result is too often the same, because those who are trying to describe reality tend to reaffirm it. Focus on the group, for example, is a commitment to one of the more rigidified aspects of the social process. Stress upon the incremental is apologetic as well. The separation of facts from values is apologetic.

There is no denying that modern pluralistic political science brought science to politics. And that is a good thing. But it did not have to come at the cost of making political science an apologetic discipline. But that is exactly what happened. … In embracing facts alone about the process, modern political science embraced the ever-present. In so doing, political science took rigor over relevance.

Now, to be clear, political scientists are not apologists for the 2016 election, which most would depict as a nightmare. But, Lowi would argue, they were apologists for the fundamentally unstable and indefensible system that produced it.

Compared to empirical political scientists, theorists have been more attuned to the possibility of disruptive events, because:

• They are interested in the regime, not just concrete behavior and events. They recognize contradictions within the regime that may presage radical change.
• They have other regimes in mind–from ancient Greece to fascist Italy and beyond.
• They are highly attuned to ideas and ideology, and therefore quick to see that Trump might have an unprecedented popular appeal.
• They mostly don’t much like the status quo. Instead of being apologists for it, they are quick to expect and even celebrate its demise.

These predilections can mislead, though. It is very important to take into consideration the findings of empirical social science. Otherwise, what you want (or fear) can too strongly color your interpretation of events.

Indeed, I can imagine that the 2016 election will vindicate mainstream models of American politics. It seems highly likely that Hillary Clinton will beat Donald Trump, albeit with limited effects on down-ballot races because the country is so polarized along partisan lines. Clinton holds similar policy views to almost all the Democrats in Congress, so the election may reinforce their central place in US politics. Leftish critiques of the Sanders variety will then struggle for attention and traction. However, once the first Hillary Clinton administration nears its end, Democrats will have held the White House for 12 straight years, and voter fatigue may set in, perhaps compounded by a recession. Republicans will realize they can win with a more mainstream, Romney-type candidate. They will nominate such a person, and the parties will rotate as usual, restoring the system that we know.

That remains a plausible scenario. But so does a political realignment, or a constitutional crisis, or a meltdown. It’s to political theory that you must turn to assess not only the possibility of such events but their desirability.

In his most recent book, Public-Spirited Citizenship: Leadership and Good Government in the United States, Ralph Ketcham tells how leading American political scientists of the early 1900s decried education that took the form of “sermonizing and patriotic expostulation” (p. 105). The only alternative they recognized was a rigorous, detached, disenchanted study of politics as it was. In keeping with that goal, they advocated specialization and expertise. Political science meant training for professors and technocrats in basically the current system. Ketcham argues for a broad liberal education that is “profound,” “integrated,” and “radical.” But positivist social scientists tend to gravitate to education as specialized empirical training for the status quo. If you hope to navigate a time such as ours, you need do data and empirical models. But you also need a bit of profundity and radicalism.

From CIRCLE’s release this morning: “An estimated 1.8 million young people participated in Super Tuesday’s primaries and caucuses, almost a million youth in the Democratic contests and around 900,000 in the Republican contests. With a number of strong showings across many states, young people continued this year’s trend of high participation that rivals the numbers from 2008, when youth turnout in some cases tripled that of previous years. Young Republican participation, especially, has increased compared to 2008, sometimes by dramatic amounts. And in both parties young people are still not rallying around the frontrunners.”

CIRCLE also has a nifty new interactive tool that allows you to compare recent presidential campaigns’ youth support. One takeaway for me: Sanders is mobilizing almost as many young voters as Obama did in ’08. (Sanders’ percentage is larger, but the actual number is a bit smaller.) The young Obama voters in ’08 were on a bus that drove all the way to the White House. The Sanders voters will not have such a smooth ride. What difference will that make to their development as citizens and activists?

The current homepage of NPR news is a feature article by Asma Khalid about CIRCLE’s Youth Electoral Significance Index (YESI). My CIRCLE colleagues have identified the top 10 states and 10 congressional districts where the youth vote will matter most in Nov. 2016.

Politicians, campaigns, educators, and civic leaders should reach out to young people everywhere. But we recognize that political actors with limited resources will want to invest where they can have the biggest impact, and reporters may want to cover the youth vote where it counts for the most in electoral terms. Hence the YESI.

Young Americans are not very loyal to parties. Many young people hold political beliefs that may make them almost guaranteed to vote for one party rather than the other–true “swing” voters are very rare–but they don’t identify with parties as organizations or devote their energy to parties as opposed to candidates and causes. I think that is partly why young people have so far been happy to vote for Bernie Sanders over Hillary Clinton by 4:1 margins. Remember that the primary election process was set up to allow party members to choose their party’s nominee, but young people who vote in the Democratic primaries don’t blink an eye to support a candidate who has chosen not to be a Democrat during his political career. I am not saying they should behave differently; I just think it’s interesting.

I have often heard cultural/generational explanations of this trend. Supposedly, Millennials are less favorable to organizations of all kinds. They prefer and expect looser and less hierarchical networks. There may be some truth to that, but I would suggest a different hypothesis. Young people are less loyal to parties than their predecessors were because parties don’t do anything any more.

Parties used to have functions, such as recruiting volunteers, paying workers, and organizing events (not to mention controlling patronage). Parties are now labels for clusters of entrepreneurial candidates and interest groups. The change occurred because the campaign finance reforms of the early 1970s defunded the parties, and then the deregulation of the 2000s allowed vast amounts of money to flow to entities other than parties. The Koch Brothers’ political network, for instance, employs 3.5 times as many people as the Republican National Committee does.

If parties do nothing for or with young people, it is easy to explain why youth don’t care about parties.

The General Social Survey asks about partisan ID at least every other year. The proportion of younger people who are Independents has grown, but most political scientists argue that that trend is misleading since the number of undecided or swing voters has actually shrunk. More to the point are questions that the GSS has asked only twice, about membership and active participation in parties. We know that parties didn’t do much to engage youth in 2004, because Dan Shea surveyed local party leaders that year, and “Only a handful of [county] party chairs mentioned what we might call significant activities, programs that require a significant amount of time or resources.” The parties were already hollow compared to decades earlier. He also asked an open-ended question: “Are there demographic groups of voters that are currently important to the long term success of your local party?” Just eight percent named young voters.

Nevertheless, the GSS indicates that the proportion of youth who actively participated in parties was 3.6 times higher in 2004 than it was in 2014. The hollowing-out continues.

My colleagues at CIRCLE are producing a stream of detailed and almost instantaneous analysis of the caucuses and primaries. Keep checking the CIRCLE homepage for the latest.

Here I use CIRCLE’s evidence to illustrate how Sen. Sanders’ dominated the youth vote in the Iowa caucuses. Consider the Democratic and Republican caucuses as one event: the voter first chooses which party to caucus with, and then selects a candidate. By that reasoning, about 50,000 young Iowans (ages 17-29) caucused, and about 58% of them chose the Democratic side. Sanders drew 84% of the Democratic youth, while the Republican youth split their support. As a result, Sanders drew about 49% of all the young caucus-goers put together. Cruz came in second with about 11% of all the youth, followed very closely by Rubio, then Clinton, and then Trump.

Sanders got about eight times as many votes as his main opponent on the Democratic side, and about eight times as many as Trump, with whom he is sometimes paired as a supposed enemy of “the establishment.”

That raises such questions as: Can Sen. Sanders do better among older people in other states? Can he perform as well among youth in states where young Democratic voters are far more diverse than they are in Iowa? Can Sec. Clinton narrow the generation gap, and can she get out the youth vote if she wins the nomination? (She only drew about 4,000-5,000 young Iowans on Monday and came in fourth in that age bracket, which ought to ring some alarms.) Finally, where will young Republicans land as their field narrows?

Below is CIRCLE’s press release from this morning. Additional data can be found on the website.

Young Democrats Propel Sanders to Virtual Tie in Iowa; Record-breaking Participation Among Young Republicans, who Choose Cruz, Rubio Over Trump

Medford/Somerville, MA – Youth turnout in last night’s Iowa caucuses is estimated to be 11 percent, according to youth vote experts from the Center for Information and Research on Civic Learning & Engagement (CIRCLE) – the preeminent, non-partisan research center on youth engagement at Tufts University’s Jonathan M. Tisch College of Citizenship and Public Service.

Highlights of the youth vote in Iowa include:

• An estimated 11.2% of eligible Iowan youth aged 17 to 29 participated in last night’s Republican and Democratic caucuses.
• On the Democratic side, the youth choice was decisive. Of the estimated 31,000 young people who participated in the Democratic caucus, 84% supported Senator Bernie Sanders, contributing to a virtual tie between Secretary Hillary Clinton and Senator Sanders.
• Young Republicans selected Senator Ted Cruz as their top candidate (with 26%), closely followed by Senator Marco Rubio (23%). Youth support for Donald Trump, which came in at approximately 20%, trailed the support he received among older Republicans.
• Since 1996, youth turnout in Iowa has exceeded 4% only twice: in 2008 (14%) and yesterday (11.2%).
• A record-breaking 22,000 young people voted in the Republican caucus.
• About 31,000 young people participated in the Democratic caucus, the second highest level since 1996 (behind 2008).

“Last night’s Iowa caucuses demonstrated the potential power of young people to shape elections,” said Kei Kawashima-Ginsberg, Director of CIRCLE. “In the Democratic caucus, young voters helped to propel Senator Sanders to a virtual tie, and Republican youth broke their own record of caucus participation. One message is clear: when candidates and campaigns ask young people to participate and inspire them to get involved, they respond.”

For CIRCLE’s full Iowa caucus analysis, please see here. Throughout this election season, CIRCLE’s 2016 Election Center will offer new data products and analyses – such as a preview of youth participation in the NH primary – providing a comprehensive picture of the youth vote, both nationally and in targeted states and congressional districts across the country. You can view trend data on youth turnout via CIRCLE’s interactive maps.

Washington Monthly has a cover story by Phil Kiesling arguing that voting by mail would raise turnout substantially and also produce a more representative electorate, especially in primaries, thus reducing partisan polarization. Kiesling makes many good points, and his argument is definitely worth reading. Among other things, he is right to be frustrated by the research on voting-by-mail, although I wouldn’t characterize the problem as he does–as “the tyranny of dated, superficial, and/or irrelevant, academic research.” In fact, my colleagues and I are responsible for some of that research, and there are reasons that we analyze the data as we do. It is genuinely hard to measure the effects of a policy that has been adopted in only one or a few states. But that methodological challenge can limit our imaginations as reformers. Even though I don’t think he characterizes the research fairly, Kiesling helpfully challenges our imaginations.

Before getting into the methodological weeds, let me explain why I am not yet convinced that vote-by-mail would increase turnout. Below is Oregon’s turnout trend compared to the national trend. I show the presidential years and off-years as separate lines, because otherwise you get a confusing zigzag pattern. Oregon abolished polling places in 2000 and has run its elections entirely by mail since then. Turnout in the 2004 presidential election was strong, but not so great in 2008 and 2012. Turnout in the off-year elections since 2000 has been lower than it was in the 1980s and 1990s. I don’t see evidence of positive impact here.

Kiesling calls it a “problematic assumption” to measure turnout as I do above: as the percentage of all eligible citizens who voted. (My data come from Michael McDonald.) Kiesling would rather define turnout as the percentage of registered citizens who vote. But if votes-per-registered rises, and yet votes-per-eligible-citizens does not, then there is only one mathematical possibility: the percentage of the population that registered must have fallen. That’s not a good thing. So the only way to assess the net impact of a reform on our democracy is by measuring turnout as votes per eligible population. We always do this, not in a “selective manner … to overlook (or even disparage)” vote-by-mail.

My graph simply allows you to see whether turnout has risen or fallen in Oregon before and after the vote-by-mail reform. That is not really a satisfactory method because many other factors are obviously at work. Nor can we conduct randomized experiments with new voting laws. Thus the way almost all academics study state policy reforms is by building large datasets that include statistics from all 50 states over many years. Then we try to isolate the apparent impact of a given change, controlling for everything else that we have measured, including the baseline turnout rates for states prior to the reform.* That method has yielded a rough consensus in favor of Election Day Registration–mentioned by Kiesling and exemplified in our own research.

This method cannot, however, assess the impact of a reform that has only been implemented in one or a handful of states. The math just doesn’t allow estimates of impact when there are very few cases in the category of interest. Thus, frustratingly, we omit Oregon’s vote-by-mail system from most of our models. Kiesling is right to complain about that gap, but I know of no satisfactory solution. One potential option would be to lump universal vote-by-mail together with other reforms, to increase the number of cases, but that would hide any special advantages of universal vote-by-mail.

Kiesling says that the research “lazily tries to extrapolate absentee ballot-based data, inappropriately uses VEP-based yardsticks, and/or focuses on presidential contests only—ignoring midterm and primary elections, where the approach is best suited to show dramatic results.” Those are not fair complaints about our models.

I’d still vote for and recommend universal vote-by-mail, as an experiment. The graph shown above makes me a little pessimistic about its impact (particularly in midterm elections). And I do lament the loss of a secret ballot when elections move to mail. The secret ballot was a hard-won reform designed to prevent voters from being coerced by employers and relatives or from selling their votes. (If you promise to pay me for voting for A, but my ballot is secret, I can take your money and vote for B, which ruins the market. In an election-by-mail, however, your boss, your spouse, or a bribe-giver can check your ballot.) Kiesling says the Oregon Secretary of State is aware of “just a dozen documented cases,” and I guess we could live with that–although each case of coercion is a human rights violation.

More broadly, I resonate with Kiesling’s frustration. If a state does something bold and different, the standard methods of social science really struggle to assess its impact. When a bunch of states implement the same rather modest change, we can measure impact, but often it’s disappointing. For instance, we have found that the variation in states’ education laws does not seem to matter for students’ media literacy or electoral participation. I don’t conclude that reforming education is a waste of time, but rather that the prevailing reform strategies are too weak. A few states have done more interesting things with civic education policy, but we can’t assess their impact using multivariate models.

As citizens, we must keep our imaginations vivid and hopeful. Statistics are always about the past, and the future can be different. I welcome the Washington Monthly’s cover article as a spur to creativity. A 21st-century election system should look very different from the clunky mechanisms we have in place today. Universal vote-by-mail could be the answer. I just don’t read the existing research evidence as Kiesling does.

—

*Kiesling writes: “But point out that in 2014, the nation’s three UVBM [universal vote-by-mail] states’ voter turnout—based on the identical VEP denominator—was 16 percent higher than non-UVBM states, and the response is often something along the lines of, ‘Yes, but that’s irrelevant because Oregon (and presumably now Washington and Colorado) have always been high-turnout states.’ As if Minnesota, Maine, and Wisconsin are any different?” This is overlooking the power of a multivariate model, which takes into account the historical turnout rate as well as many other variables. We favor Same Day Registration not because turnout is simply higher in states that have it, but because it is higher controlling for other factors.