Protest in America

Kieran has a great discussion about the significance of current anti-war activity in the U.S. I couldn’t pass up some comment on it, because the data he discusses is the data I have been collecting and laboriously cleaning for three+ years.

Kieran’s overall point is that, in the period for which we have clean protest data, protest events with many tens of thousands of participants are rather rare, making contemporary protests that draw huge numbers of people a significant phenomenon—even given, of course, that the protest is wrapped up in a set of conditions much broader and more complicated than the already-difficult task of counting participants.

The comments to Kieran’s post raise many of these issues. For instance, are the mass marches of today even comparable to those of the 60s, given arguments about the “banalization” of protest and the institutionalization of professional movement organizations? Kieran rightly points out the tendency to dismiss contemporary protest as either an aping of the 60s or an embarrassing abandonment of decorum—as if protest were neither informed by the drama of past events nor subject to the kinds of tactical innovation (and yes, escalation) that characterized protest in the 50s and 60s. Plenty of research and theory points to the transition of movement activity to “paper memberships” in the 1970s and beyond; that is, some scholars argue that the institutionalization of the social movement “sector” means fewer people actively engage in protest, chosing instead to write letters or donate money, or, more frequently, let SMOs (social movement organizations) do it for them through lobbying and other tactics. But it’s clear that dismissing contemporary protest based on the fact that one part of the movement society is based in organizations that don’t assemble mass protests, just doesn’t make sense. Comparison of political participation data shows that in 1976, 3% of national survey respondents reported taking place in some active protest activity; in 1990, it was 5.7%. The moderation of certain protest activity via SMOs doesn’t appear to short-circuit participation, although raw effectiveness of those activities is certainly a good question.

Social movement outcomes are very difficult to track. How do we definitively connect collective action to some desired outcome? Good data and techniques such as event history analysis may help answer some of these questions (Sarah Soule, my advisor in the realm of movements, has done a good bit of work with colleagues on this very subject). Timothy Burke comments at Kieran’s blog that “size doesn’t matter” because contemporary protesters are out of touch with the fact that protest has a different relationship with policy and the electoral process than it did in the 60s. I’ll grant that the relationship has changed, and that some activists may not have caught on to that fact completely, but too often comments about small protests are used to denigrate the anti-war movement. This is an almost daily occurrence at InstaPundit and the coterie of mock-the-protesters blogs, where the assembly of a couple dozen protesters is an opportunity to point out that anti-war sentiment is a minority opinion held by motley, dirty, socialist foreigner-lovers. On the face of it, that just isn’t true; when you compare the numbers between now and the 1960s, it’s dramatically untrue. This is significant popular mobilization, regardless of its effectiveness at actually influencing policy.

The context of mobilization has undoubtedly changed since the 60s, as well—a point raised by one of the commenters to Kieran’s post. A good deal of social movements research has explored the ways in which civil rights-era protests were affected by the development of tactics (both of protest and of suppression) and influenced by the ability of organizers to communicate with one another, mobilize participants, and activate protests in additional nearby locations. New communications technology are already changing the way at least some of this mobilization works (Howard Rheingold’s Smart Mobs has an interesting account of this, and plenty of contemporary sociological research has taken up the issue, as well).

A couple of comments on the event data itself:

  • The estimation of protest size was most often done with the aid of clues in the text. Accounts frequently reported “a handful,” “dozens,” “hundreds,” or “between 4,000 and 7,000” participants, for example. Only rarely were there absolutely no clues to the number of participants.
  • As Kieran points out, there’s a ton of data in there. We’ll ultimately be able to control for group composition, geographical location, etc.
  • The data captures events that are less “visible” than street protests (e.g., boycotts and letter-writing campaigns) and those that are much more confrontational (e.g., attacks, take-overs, etc). Again, lots of data means lots of possibilities for the kinds of complicated analyses we want to do. We’re social scientists, after all.

Much later update: In Fall 2004 I used some of this protest data to write more about protest and protesters.