The Paradox of Punishing for a Democratic Future

By Rachel López and Geoff Dancy

Recently, the U.S. Supreme Court grappled with an issue of first impression in this country, but one familiar to other jurisdictions around the world—that is, whether a former head of state should be immune from prosecution for his criminal acts while in office. Those who argue in favor of criminal accountability, at home and abroad, often trumpet the democratic benefits of punishing state officials. Such justifications have also been consecrated in law, finding their way into judicial decisions as a basis for overturning amnesty laws. But is there any evidence to support these claims? Can a government really punish its way to a more democratic future?

Using empirical evidence from other countries that have prosecuted state officials for their crimes over the last three decades, this study sheds light on the possible effects of these prosecutions on democratic institutions and behaviors. First, it examines an in-depth case study of Guatemala, a country where this issue recently came to the fore, to develop a set of hypotheses about the democratic effects of punishing state officials. To determine whether the lessons gleaned from Guatemala are more generalizable, it tests these hypotheses using the most extensive global data set of prosecutions of government officials in domestic courts, which specifically focuses on human rights prosecutions.

Interestingly, the findings reveal a paradox. While criminal prosecutions of state officials for human rights violations are associated with some pro-democracy outcomes, like increased civil society activism and pro-democratic mobilization, they are also associated with greater political polarization and anti-system backlash. By contrast, they appear to have little effect on democratic institutions. Considering this data, a central takeaway is that the democratic effect of prosecuting political leaders tends to rest with the people. Whether punishing them helps to ensure a more democratic future depends more on how the populace responds—negatively or positively—than on the limited institutional effects resulting from punishment.

2025, 68p.