Papers and other publications organized by research area.
Mexican Development and Regional Integration
Embedded Multinationals: When MNCs Choose Voice Over Exit (with Steven Samford). World Politics 79.1, 2027 (forthcoming). PDF · Multinational corporations (MNCs) operating in developing contexts have been broadly characterized as mobile investors that benefit from “passive” structural power; so why might they engage in resource-intensive efforts at everyday politicking? To address this mostly unnoticed phenomenon, we trace the extensive efforts of MNC subsidiaries in two starkly different industrial sectors in Mexico—electronics manufacturing and agribusiness—to participate in politics through domestic trade associations. We develop a causal model to explain why these and other multinational foreign firms would derive political power from investing in “voice” rather than threatening to “exit” from the economy as expected. Specifically, locational inflexibility of operation makes disinvestment unappealing, creating incentive to foster social embeddedness. Embeddedness, in turn, provides resources for MNCs to employ instrumental political power. The implications of this argument speak to a range of political economy scholarship by linking MNC embeddedness to regional economic integration and rethinking the role of MNCs in coalitions for economic upgrading in their host countries.
Embedded Capital: Multinational Firms, Business Associations, and the Politics of Upgrading in Mexico With Mariana Rangel Padilla and Steven Samford. · Under what conditions do multinational firms become embedded in the host-country institutions that mediate claim making and upgrading? Conventional accounts depict MNCs in middle-income countries as footloose actors whose leverage derives primarily from the threat of exit and as unwilling partners in local industrial coalitions. We argue instead that, in regionally integrated middle-income countries such as Mexico, multinational firms vary in their degree of embeddedness in host-country institutions and that this variation can be traced back to their degree of locational inflexibility. Sectors with greater economic embeddedness—in the form of sunk costs, geographic specificity, specific labor needs, and production-network dependence that make capital less mobile—have greater incentive to embed themselves in host-country institutions to shape their local environment rather than resorting to disinvestment. Empirically, the paper combines a new sector-level dataset with qualitative sector narratives on textiles, electronics, berries, steel, and auto parts. This paper represents the first step in a broader research program by theorizing economic embeddedness of MNCs and mapping its variation across Mexican manufacturing.
Beyond Expansion: Spatial Permanence and Turnover in Mexico’s Protected Berry Agriculture With Iskar Waluyo-Moreno and José Mauricio Galeana-Pizaña. · Agricultural booms are commonly assessed through changes in cultivated area, production, or exports, yet these indicators do not reveal whether newly established production systems persist spatially over time. We examined the expansion, turnover, and persistence of protected agriculture between 2016 and 2025 in nine major berry-producing municipalities in western Mexico. Using Sentinel-2 imagery, annual Random Forest land-cover classifications, trajectory analysis, and a discrete-time survival model, we tracked transitions among natural land cover, open-field agriculture, urban land, and protected agriculture. Across the two interperiod transitions, natural land cover declined by approximately 115.1 km², while protected agriculture recorded a net increase of 72.5 km². Direct conversion from natural land cover to protected agriculture accounted for only 5.5 km², whereas transitions from open-field agriculture accounted for 137.1 km², indicating that open-field cultivation frequently preceded the establishment of covered production. Gross gains nevertheless concealed substantial spatial turnover, particularly during the later transition, when expansion and exit from protected agriculture were nearly equivalent. The discrete-time model showed that the conditional exit hazard increased with distance from production nodes identified in 2016, after accounting for episode duration, temporal period, and patch-size class; the magnitude of this spatial relationship varied across periods. These findings indicate that the expansion of protected agriculture in western Mexico involved not only net growth but also recurrent entry, exit, and spatial redistribution. Agricultural-boom assessments should therefore consider where covered production persists, where it turns over, and how these dynamics relate to the spatial organization of established production areas.
The Moral Economy of the Entrepreneur: Land Reform and Rural Upgrading in Mexico · Land reforms that individualize property rights are often promoted as a means of encouraging investment and entrepreneurship. Yet critics argue that such reforms can undermine the collective institutions on which rural livelihoods depend. This paper revisits this debate through the case of Mexico’s PROCEDE program and the rapid expansion of export-oriented berry production. I argue that successful smallholder upgrading depends on the interaction between two dimensions of rural social organization: community dynamism and individual embeddedness within local social networks. Using an integrative multi-method design that combines administrative and spatial data, an original survey and survey experiment, and comparative case studies, I show that upgrading is most likely in ejidos (collective land-management associations) that embraced land reform and exhibited active institutional adaptation, and among producers who are deeply embedded in community institutions. The findings suggest that individualized property rights can facilitate upgrading, but only when accompanied by social institutions that enable coordination and risk-sharing.
Social Capital and Rural Upgrading: Cocoa in West Africa With Alexander Jelloian.
Community and Craft: Traditional Governance in the Mezcal Industry
Interest Organizations and Economic Policy
Work and Demand Making: Productionist and Consumptionist Politics in Latin America (with Ruth Berins Collier). Comparative Political Studies 55.10: 1631-62, 2022. PDF · Online Appendix · How does the world of work in Latin America affect the way workers act to defend their interests? To what extent have “productionist” demands—those concerning jobs, work conditions, and wages, which are highly salient across the region—been displaced by consumptionist or political demands? While the literature has distinguished formal and informal work broadly, we explore individual traits of work that crosscut the formal–informal distinction. Analyzing survey data from four Latin American capital cities, we find, not surprisingly, that both work-based atomization and insecurity depress demand making in the work arena. But these traits of work also affect demand making on the state, albeit in somewhat different ways. Insecurity is associated with a shift from productionist to consumptionist and political demands, while atomization is associated with a more generalized demobilization across issues. These findings have implications for the representation of worker interests in light of current labor-market restructuring and raise the question of whether labor can reclaim an important voice in that restructuring process.
Incentives for Organizational Participation: A Recruitment Experiment in Mexico (with Candelaria Garay and Mathias Poertner). Comparative Political Studies 54.1: 110-43, 2021. PDF · Online Appendix · While the presence of a strong civil society is recognized as desirable for democracies, an important question is what motivates citizens to join organizations. This article presents novel experimental evidence on the conditions under which citizens join interest organizations. We presented 1,400 citizens in two Mexican states with fliers promoting a new local interest organization. These fliers contained one of four randomly selected recruitment appeals. We find evidence that both brokerage of state patronage and demand making for local public goods are effective recruitment appeals. The effect for patronage brokerage is especially pronounced among respondents with prior organizational contact, supporting our hypothesis of a “particularistic socialization” effect wherein organizational experience is associated with greater response to selective material benefits. Our findings suggest that under some conditions, rather than generating norms of other-regarding behavior, interest organizations can reinforce members’ individualistic tendencies.
Organizational and Partisan Brokerage of Social Benefits: Social Policy Linkages in Mexico (with Candelaria Garay and Mathias Poertner). World Development 136, 2020. PDF · Online Appendix · The expansion of large-scale nondiscretionary social policy has been one of the most important policy innovations in Latin America in recent decades. While these benefits have reduced the political manipulation of low-income citizens, discretionary social programs—whose distribution follows opaque criteria and is often allocated according to political considerations—continue to exist. Employing an original survey in Mexico, we explore how citizens, both beneficiaries and nonbeneficiaries, experience and perceive access to discretionary social programs. While the literature on clientelism emphasizes the distribution of discretionary benefits by party agents in exchange for electoral support, recent studies have found that access can also operate through community associations or interest organizations. We conducted a list experiment to detect whether, in respondents’ experience or perception, social benefits are allocated on the basis of partisan campaign support or organizational participation. Our findings reveal that organizational brokerage is at least as important as the much-studied role of party-mediated clientelism for access to discretionary benefits.
Evading the Patronage Trap: Organizational Capacity and Demand Making in Mexico. Comparative Political Studies 52.13-14: 2097-134, 2019. PDF · Online AppendixHonorable Mention, 2019 Latin American Studies Assoc. Mexico Section Social Sciences Essay Award. · When do organizations broadly represent the interests of their economic sectors and when do they narrowly represent the interests of members? This article investigates how agricultural and small-business organizations in Mexico make demands for programmatic policies or patronage benefits. Contrary to explanations based on the class of members, I show that the source of organizational capacity shapes demand-making strategies. Organizations that generate selective benefits internally are able to engage in programmatic policies that shape sectoral competitiveness, whereas organizations that fail to solve membership challenges internally are vulnerable to the patronage trap, a self-reproducing cycle wherein they become specialized in demand making for discretionary private goods. I generate this argument through process tracing of two agricultural organizations in Mexico. Analysis of an original survey of economic interest organizations provides broader evidence that organizational capacity is a better predictor of policy demands than social class.
Interest Organizations and Distributive Politics: Small-Business Subsidies in Mexico, World Development 84: 97-117, 2016. PDFWinner, 2016 Latin American Studies Assoc. Mexico Section Social Sciences Essay Award. · Politicians often use distributive programs to build and maintain electoral linkages with interest organizations. This project analyzes the factors that lead Mexican state governments to allocate distributive benefits to small-business organizations rather than directly to entrepreneurs or to other groups. Theoretically, I address how electoral threats to the ruling party shape the incentives of politicians to deliver benefits to core and non-core interest organizations. Using an original dataset of decentralized small-business subsidies, I deploy a novel subnational design, comparing states governed by parties with different relationships to small-business organizations and facing different levels of electoral threat. I find that electoral competition shapes politicians’ incentives for incorporating organizations in distributive spending. Governors from the pro-small-business PAN increase allocations to these organizations when they are electorally strong, while governors from the centrist PRI allocate more to them when they face an electoral threat. Case studies provide additional evidence that ruling politicians allocate benefits to small-business organizations when linkages with these groups contribute to their most pressing electoral objectives.
Working Papers
Forging Common Fate: How Organizational Membership Shapes Political Participation With Mathias Poertner and Candelaria Garay. Under review. Honorable Mention, 2023 Kendra Koivu Paper Award, QMMR Section (APSA). · With the decline of labor unions, scholarly attention to mass-based organizations and their role in shaping political behavior has waned. Yet territorial, identity-based, and economic organizations have proliferated across the Global South. Despite their relevance, political scientists have largely neglected how these groups shape members’ political identities and participation. We develop a framework explaining how different types of organizational membership—particularly mandatory versus voluntary—generate feelings of “common fate” among members toward the organization, a key driver of political engagement. Drawing on original surveys and focus groups in Mexico, we show that mandatory membership fosters stronger organizational attachments and higher political participation. We identify two mechanisms: longer time horizons and encompassing membership scope. Our findings highlight the underappreciated importance of organizational design in cultivating collective identity and participation. We also introduce new measures of organizational attachment and call for renewed attention to mass-based organizations in comparative politics and political behavior.
Transparency and Accountability Institutions
Accountability in Time: Evolution and Expertise in Participatory Institutions (with Jésica Tapia, Daniel Berliner, Aaron Erlich, and Benjamin Bagozzi). World Politics 77.3: 468-514, 2025. PDFWinner, 2022 Latin American Political Institutions Section (LAPIS) Best Paper Award; Honorable Mention, 2026 Latin American Studies Assoc. Mexico Section Social Sciences Article Award. · How do participatory institutions change over time? Previous research has focused on exogenous changes, such as legal reform or leadership replacement. But institutions also evolve endogenously, through processes of behavioral and compositional change on the part of citizen claimants and government officials. These processes can gradually reshape institutions to become more responsive to either expert or nonexpert claimants. The authors refer to such processes as brokered and grassroots models of social accountability. In the context of Mexico’s access-to-information system, the authors employ new machine-learning-generated measures to analyze nearly two million information requests and responses filed between 2003 and 2019. They find evidence that shows claimants becoming more sophisticated over time, and officials becoming more responsive to these expert claimants—both findings consistent with a brokered accountability model. Quantitative and qualitative evidence reveals mechanisms of behavioral and compositional change by citizen claimants and government agents.
Measuring Human Rights Abuse from Access to Information Requests (with Sarah Ellington, Benjamin Bagozzi, Daniel Berliner, and Aaron Erlich). Journal of Conflict Resolution 66.2: 357-84, 2022. PDF · Existing measures of human rights abuses are often available only at the country-year level. Several more fine-grained measures exhibit spatiotemporal inaccuracies or reporting biases due to the primary sources upon which they rely. To address these challenges, and to increase the diversity of available human-rights measures more generally, this study provides the first quantitative effort to measure human-rights abuses from textual records of citizen-government interactions. Using a dataset encompassing over 1.5 million access-to-information requests made to the Mexican federal government from June 2003 onward, supervised classification is used to identify the subset of requests that pertain to human-rights abuses of various types. The results are validated against gold-standard requests pertaining to past human-rights abuses in Mexico and several accepted external measures of subnational and subannual human-rights abuses. We demonstrate that measurement from citizen-submitted request texts can provide measures of human-rights abuse that exhibit both high validity and notable spatiotemporal specificity relative to existing datasets and variables.
Media Attention and Bureaucratic Responsiveness (with Daniel Berliner, Benjamin Bagozzi, and Aaron Erlich). Journal of Public Administration Research and Theory 31.4: 687-703, 2021. PDF · How does media attention shape bureaucratic behavior? We answer this question using novel data from the Mexican federal government. We first develop a new indicator for periods of anomalously heightened media attention, based on 150,000 news articles pertaining to 22 Mexican government ministries and agencies, and qualitatively categorize their themes. We then evaluate government responsiveness using administrative data on roughly 500,000 requests for government information over a ten-year period, with their associated responses. A panel fixed-effects approach demonstrates effects of media attention on the volume of outgoing weekly responses, while a second approach finds effects on the queue of information requests already filed when anomalous media attention begins. Positive or neutral attention is associated with reduced responsiveness, while the effects of negative attention vary: attention to government failures leads to increased responsiveness, but attention to corruption leads to reduced responsiveness. These patterns are consistent with mechanisms of reputation management, disclosure threat, and workload burden, but inconsistent with mechanisms of credit claiming or blame avoidance.
The Political Logic of Government Disclosure: Evidence from Information Requests in Mexico (with Daniel Berliner, Benjamin Bagozzi, and Aaron Erlich). Journal of Politics 83.1: 229-45, 2021. PDF · Online Appendix · When citizens ask questions, how does their government answer? Requests for government information confront officials with incentives both for and against disclosure. We argue that officials seek to manage political risks in ways that favor requests from government-aligned regions. We study responsiveness in the context of Mexico’s access-to-information law, using publicly available data from several hundred thousand information requests filed with Mexican federal government agencies between 2003 and 2015. Our empirical strategy makes comparisons only among requests sent to similar agencies on similar topics at similar times, while accounting for the complexity, sophistication, and sensitivity of individual requests. We find that requests filed from locales with higher governing-party vote shares receive more favorable responses, across multiple indicators of the nature and timing of responses. Further, we find bias only for requests on publicly relevant topics, providing evidence in favor of a mechanism of mitigating political risks over one of rewarding supporters with greater access to benefits.
What Information Do Citizens Want? Evidence from One Million Information Requests in Mexico (with Daniel Berliner and Benjamin Bagozzi). World Development 109: 222-35, 2018. PDF · Online Appendix · While scholars have emphasized the importance of information for accountability, little research has addressed the demand for government information by real citizens. We study the totality of information requests filed with Mexican federal government agencies from 2003 to 2015—over one million requests in all. We use unsupervised methods to categorize requests, revealing diverse topics including the environment, security, budgets, government procurement, and government employees. While many topics have clear public accountability-seeking purposes, others focus on more private, micropolitical goals. Analysis over time and across states reveals links between information demand and issues of public interest such as environmental impacts and criminal violence. Our results demonstrate that, given functioning access-to-information institutions, citizens in a transitional democracy really do demand information relevant to public accountability.
Delivering Transparency: To What Degree does the Mexican Government Respond to Public Information Requests? (with Jonathan Fox and Libby Haight). Gestión y Política Pública 20.1: 3-61, 2011. [versión publicada en español] PDF · To what degree do transparency reforms actually make information public? This study documents patterns of agency responses to citizen requests for information. It draws on both official and independent assessments during the first years of implementation of Mexico’s 2003 open-government reform. The data show substantial progress overall, combined with significant variation across federal agencies and a growing trend toward official denials of the existence of requested information. The findings indicate that support for and resistance to open government are unevenly distributed across the public sector. Future analysis of varying patterns of agency compliance should address both agency-specific incentives and institutional cultures.
Working Papers
Covert Non-Responsiveness: Security and Citizen-State Interactions in Mexico With Daniel Berliner, Aaron Erlich, and James Korman. · Accurate performance reporting is essential for democratic accountability, yet officials often have incentives to appear responsive even when they are not. We examine a form of misleading reporting we call “covert non-responsiveness,” in which officials document fulfilling public needs while substantively failing to do so. We study this behavior using over 1.5 million information requests in Mexico and develop text-based measures to distinguish substantive responses from how they are officially recorded. We argue that covert non-responsiveness becomes more prevalent when officials face especially sensitive citizen claims, focusing on security-related requests as our primary test and examining security-oriented agencies and periods of heightened violence as extensions. We find that covert non-responsiveness is widespread: an estimated 26.8 percent of responses officially recorded as providing information are in fact wholly or partially nonresponsive. Most robustly, security-focused requests are more likely to receive covert denials claiming that the requested information does not exist, while evidence for broader agency- and context-level effects is more mixed. These patterns show how misleading reporting can conceal bureaucratic failure and undermine core institutions of democratic accountability.
Party Linkages and Clientelism
After Corporatism: Party Linkages with Popular Sector Organizations in Neoliberal Latin America. In Diana Kapiszewski, Steven Levitsky, and Deborah J. Yashar (eds.) The Inclusionary Turn in Latin American Democracies. Cambridge University Press, pp. 327-61, 2021. PDF · What role do popular-sector associations play in party building in contemporary Latin America? While corporatist ties to hierarchically organized labor and peasant associations have diminished, a bevy of “dissident” popular-sector organizations have emerged, such as neighborhood associations, Indigenous movements, and informal-sector associations. These organizations have played contrasting roles in party building for new left-wing parties such as Brazil’s PT, Bolivia’s MAS, Mexico’s PRD, Uruguay’s FA, and Venezuela’s AD, varying from durable and programmatic neo-corporatism to contingent patronage ties. Based on these cases, this chapter inductively identifies founding party traits that shape linkage type. Parties that established formal rules for incorporating organizational allies in party leadership and included a major segment of the labor movement in their founding coalition were most successful at institutionalizing spaces for allied organizations’ programmatic influence. This argument is tested through a subnational analysis of Mexican state governments under the PRD.
Gerrymandering Opposition: Minority-Concentrated Districts and Electoral Competition in Mexico (with Anne Meng). Studies in Comparative International Development 52.1: 64-86, 2017. PDF · Can institutions that are designed to improve minority representation also affect electoral competition? We address this question by examining how minority-concentrated districts—designed to empower Indigenous populations—affected minority participation and party competition in Mexico. Using an original dataset and a matching design that helps alleviate causal-inference problems inherent to observational studies, we find that these districts had no effect on minority participation but enhanced electoral competition. Field research reveals that minority-concentrated districts weakened one-party dominance by assembling minority voting blocs that were amenable to opposition-party appeals. More broadly, our results suggest that the mobilization of minority voting blocs can promote electoral competition in transitional democracies.
Beyond the Machine: Clientelist Brokers and Interest Organizations in Latin America (with Alisha Holland). Comparative Political Studies 48.9: 1186-223, 2015. PDFWinner, 2013 Latin American Political Institutions Section (LAPIS) Best Paper Award. · Organizational membership is one of the strongest, yet overlooked, predictors of vote buying across Latin America. We argue that this relationship is driven by the fact that politicians outsource some of their vote-buying efforts to interest associations. In contrast to the existing literature that focuses on party brokers, who are loyal to a single political machine, we introduce the concepts of organizational brokers, who represent interest associations and renegotiate ties to political parties between election cycles, and hybrid brokers, who split their loyalties between an interest association and a single political party. We illustrate the operation of these alternative broker types through case studies of street-vending organizations in an uninstitutionalized party system, Colombia, and peasant organizations in an institutionalized party system, Mexico. Attention to the role of brokers and the organizations they represent has implications for political accountability, collective action, and the persistence of clientelism.
Clientelism, Declared Support, and Mexico’s 2012 Campaign (with Simeon Nichter). In Jorge I. Domínguez, Kenneth F. Greene, Chappell H. Lawson, and Alejandro Moreno (eds.) Mexico’s Evolving Democracy: A Comparative Study of the 2012 Elections. Johns Hopkins University Press, pp. 200-26, 2015. PDF · Most studies of clientelism emphasize elite strategies. This chapter shifts attention to citizen behavior by examining “declared support”: publicly signaling support for a candidate by campaigning, attending rallies, displaying banners, or wearing party apparel. Drawing on the Mexico 2012 Panel Study and ethnographic fieldwork in Michoacán and Jalisco, we analyze when citizens declare support and how risk shapes their choices. Supporting a winning candidate may improve access to selective benefits and social services, while backing a losing candidate may bring penalties; remaining undeclared trades potential rewards for protection from unfavorable treatment. We develop and test hypotheses about individual and contextual predictors using panel survey responses and local-level aggregate data, triangulated with interviews and campaign observation. The analysis shows why citizens’ strategic calculations are essential to understanding clientelism.
Working Papers
Worker Inclusion without Union Empowerment: MORENA and Organized Labor in Mexico With David Foust. Revise and resubmit. · This paper examines labor inclusion under Mexico’s Movimiento de Regeneración Nacional (MORENA), the most successful new left-wing party in twenty-first-century Latin America. Governing as a populist party without organic ties to labor, MORENA delivered substantial gains to workers and adopted a landmark labor reform designed to strengthen union democracy. We depict this outcome as worker inclusion without union empowerment: workers are incorporated principally as beneficiaries and voters rather than through autonomous collective organizations. We characterize this pattern across three dimensions of union inclusion: (1) elite representation in party and public office, (2) electoral support among unionized workers, and (3) policy implementation. We attribute this outcome to MORENA’s organizational trajectory, persistent institutions of labor control, and external pressure surrounding the USMCA. Transformative left projects have historically relied on organic ties to popular-sector constituencies; MORENA’s weak organizational incorporation may limit the depth and durability of its promised Fourth Transformation of Mexican society.
Methodology
Multi-Label Prediction for Political Text-as-Data (with Aaron Erlich, Stefano Dantas, Benjamin Bagozzi, and Daniel Berliner). Political Analysis 30.4: 463-80, 2022. PDF · Online Appendix · Political scientists increasingly use supervised machine learning to code multiple relevant labels from a single set of texts. The current best practice of individually applying supervised machine learning to each label ignores information on interlabel associations and is likely to underperform as a result. We introduce multilabel prediction as a solution to this problem. After reviewing the multilabel prediction framework, we apply it to code multiple features of access-to-information requests made to the Mexican government and country-year human-rights reports. We find that multilabel prediction outperforms standard supervised-learning approaches, even in instances where the correlations among multiple labels are low.