On March 2014, Mexico adopted laws outlining the use of direct democracy. Direct democracy promised to give citizens greater control of government by reducing the gap between citizens and government. However, the first four referendums were not proposed by citizens but one each by the four major parties. Using process tracing, I examine the congressional debate of direct democracy and demonstrate that its design and rules leave little room for citizens to use the process but allow large organizations, like parties, ample room to use it. The lesson is that direct democracy should not be studied independently from the political parties that adopt it. En marzo de 2014, México aprobó leyes que delineaban el uso de la democracia directa. Ésta prometía otorgar a los ciudadanos un mayor control sobre el gobierno reduciendo la distancia entre ciudadanos y gobernantes. Sin embargo, los cuatro primeros referendos no fueron propuestos por ciudadanos, sino por cada uno de los cuatro partidos principales. Mediante el seguimiento de procesos, analizo el debate del Congreso en torno a la democracia directa y demuestro que el diseño y las reglas dejan poco espacio para que los ciudadanos hagan uso del procedimiento, y sin embargo, permiten que organizaciones como los partidos tengan amplio margen para utilizarlo. La lección es que la democracia directa no debe estudiarse independientemente de los partidos políticos que la adoptan.

Direct democracy (DD)—including initiatives and referendums—is increasingly used to decide policy around the world. Based on a wide‐ranging review of literature published over the past 25 years, this article offers theoretical and methodological recommendations to advance the study of DD. First, more focus should be placed on the stages of DD and how they affect one another. Richer theories can only flow once the origins, rules, and context of DD are considered—including the endogeneity of DD. Second, a stronger analysis of how DD is justified by democratic theory is needed; I demonstrate that some mechanisms of DD are clearly supported by democratic theory while others lack a coherent foundation. Finally, historical (e.g., ancient Athens) and contemporary DD are often used interchangeably. An overview of past institutions reveals that contemporary DD is not a return to something that once was. Each lesson provides fruitful avenues for future research. Related Articles Adams, Brian E. 2012. “Citizens, Interest Groups, and Local Ballot Initiatives.” Politics & Policy 40 (1): 43‐68. https://doi.org/10.1111/j.1747-1346.2011.00336.x Stephan, Mark. 2004. “Citizens as Representatives: Bridging the Democratic Theory Divides.” Politics & Policy 32 (1): 118‐135. https://doi.org/10.1111/j.1747-1346.2004.tb00178.x Waheduzzaman, Wahed, and Sharif As‐Saber. 2015. “Politics and Policy in Achieving Participatory Governance in a Developing Country Context.” Politics & Policy 43 (4): 474‐501. https://doi.org/10.1111/polp.12121 Related Media The Centre for Research on Direct Democracy (C2D). N.d. C2D is a database with information on direct democracy institutions around the globe. http://c2d.ch LATINNO is the first comprehensive and systematic source of data on new institutions for citizen participation evolving in Latin America. https://www.latinno.net/en/project-information/ The Initiative and Referendum Institute (IRI). N.d. IRI at the University of Southern California tracks all direct democracy mechanisms in the United States. http://www.iandrinstitute.org