We appreciated Pippa Norris’s generous review of our book Backsliding (“Voters Against Democracy,” May/June 2021). Unfortunately, however, it misses not only important aspects of our argument but also key dimensions of the very concept of backsliding.

Norris argues that we focus primarily on elites and thus miss larger changes in electorates and the failure of democratic institutions to deliver the goods. As she puts it, “Haggard and Kaufman treat demand-side factors, the forces that allow illiberal leaders to rise, as secondary. They assume a limited role for the public: voters provide a market for illiberal political appeals, sending illiberal leaders into office, but then are seen as passively accepting the consequences.” She even goes so far as to claim that we have a “‘great man’ theory of history.”

Nothing could be further from the truth. We do define backsliding as purposeful institutional change led by would-be autocrats, but precisely in contexts in which they have gained office in free and fair elections. The electorate is baked into the very concept of backsliding. As Norris notes, it is ultimately voters who acquiesce to majoritarian and even authoritarian appeals.

Moreover, the starting point in our causal story is polarization, and among publics as well as elites. Autocrats may stoke polarization, but the tinder has to be in place for the fire to start. It is precisely such polarization that contributes to the government dysfunction that Norris claims we ignore. Yet in our book, we made our storyline clear: polarization operates through stalemated institutions, growing public disaffection with democratic procedures, and voters’ support for autocrats who offer supposed solutions that cut through procedural niceties.

Norris also chastises us for our narrow and highly heterogeneous sample of 16 cases, but our selection was self-conscious. Backsliding is an incremental deterioration of democratic institutions and norms in countries that have attained a modicum of democracy to begin with. Not all forms of democratic regression fit this definition. What is striking is precisely that the diverse countries in our sample show common patterns: histories of polarization, a crucial role for legislative acquiescence—a key component of our argument that Norris ignores—and strategies of stealth and incrementalism. There is much more to be said on backsliding, including about its invidious foreign policy consequences. But Norris’s objections cover ground that is fully addressed in Backsliding.

Stephan Haggard

Lawrence and Sallye Krause Professor of Korea-Pacific Studies, University of  California, San Diego

Robert R. Kaufman

Distinguished Professor of Political Science, Rutgers University



As my review noted, Stephan Haggard and Robert Kaufman’s book makes an invaluable contribution to the debate over the deterioration of democratic institutions, norms, and practices. But as their letter suggests, Backsliding emphasizes top-down explanations focused on the role of autocrats (or would-be autocrats), who are seen to stoke polarization, and on the importance of “stalemated institutions.” They write, “We do define backsliding as purposeful institutional change led by would-be autocrats.” Right. In this view, the electorate is a facilitator of executive power but not a primary driver of it; voters “acquiesce,” but they are not the reason democracy is eroding. The public is treated as the Greek chorus in the wings—not the principal lead. This is a common view in the media.

That was the key point of my review. As the political scientist Ronald Inglehart and I demonstrate in our book Cultural Backlash, it is structural and cultural changes in mass society that are the primary long-term drivers of backsliding—resulting in consequences for institutions and opportunities for strategic elites.

There is clearly room for supply-side, demand-side, and institutional accounts in the literature on democratic regression. Analysts do not and should not have to artificially pick one or another in any comprehensive understanding. But there are profound differences in those accounts in theoretical emphasis and empirical evidence.