Michael Woldemariam’s generous review of my book raises a number of thought-provoking questions for further exploration.

Noting that I focus primarily on interpreting protests, Woldemariam asks what causes them. Causal inference is trickier in the study of social movements than in the study of, say, economic development, given pronounced practical and ethical barriers to experimentation. Nevertheless, there are several plausible reasons why third-wave African protests erupted precisely when they did. These warrant more attention in future research, possibly using the kinds of process-tracing methods that Woldemariam expertly demonstrates in his own book.

One catalyst of unrest could be the expiration of executive term limits. Most African countries underwent democratic transitions in the 1990s and forged new constitutions that often included two-term limits on the presidency. This meant that numerous African presidents were reaching their term limits around the time that the third-wave of protests broke out. By 2015, 27 African countries had reached a two-year limit a combined 36 times. Of those 36 cases, all but one (Tanzania in 1995) occurred between 2000 and 2015, and in 16 cases, the president ignored term limits or tried to amend the constitution (see Daniel N. Posner and Daniel J. Young, “Term Limits: Leadership, Political Competition and the Transfer of Power,” in Nic Cheeseman, ed., Institutions and Democracy in Africa, 2018). These breaches of democratic bargains seemed to be major triggers of activists’ decisions to launch protests. Examples that I highlight in the book include Niger in 2009 and Burkina Faso in 2015. Yet the relationship between term limits and protests is not one-to-one, as evidenced by the fact that protests unfolded even in countries not facing term limits.

Indeed, another protest trigger could be transnational contagion, as Woldemariam suggests. My book casts doubt on the interpretation of third-wave African protests as Arab Spring offshoots, but contagion might have occurred within sub-Saharan Africa. The heavy involvement of Y’en a Marre protesters from Senegal in social movements of other African countries is consistent with this hypothesis.

With respect to protest participation (as opposed to onset), I welcome Woldemariam’s advice to explore whether ethnicity mediates the relationship between economic grievances and protesting. By “bringing class back in” to the study of African politics, I do not mean to ignore ethnicity altogether. Adding interaction terms to regression models could be in order.

Woldemariam raises two final points, regarding protest leader effectiveness. I agree on the first one, that there are various ways to define movement success. Perhaps a successful movement is not one that lasts, but rather one that renders itself obsolete. I contend that success can mean both, if one considers long-term goals such as holding government accountable alongside short-term goals such as winning representation in government. In new projects, I am seeking to explain the determinants of multifaceted movement outcomes.

Lastly, Woldemariam asks why middle-class protest leaders would ever choose not to mobilize poorer compatriots. Returning to a thread of my initial argument, the answer has to do with “how middle class” the protest leaders are: Are they of the new African middle class that is autonomous from state power? If so, they will align more naturally with ordinary citizens than will actors from the older “state class.”