Disobedient markets : street vendors, enforcement, and state intervention in collective action

Under what conditions do informal workers organize? Contrary to conventional wisdom, informal workers organize in nearly every major city on every continent and officials often encourage them to do so. I demonstrate that under certain conditions, governments offer private benefits to informal workers who organize self-regulating associations, which solves the workers' collective action problem. This leads to another puzzle: why do governments pay people to organize, especially people who routinely violate the law? I argue that where the state cannot stop violators, it may prefer to pay violators to organize a self-regulating group rather than enforce the law itself. The state can then bargain over legalization, regulation, and enforcement with a representative group.

The project challenges assumptions about collective action in marginalized communities and offers a new theory of collective action where informal workers interact strategically with the state. I argue that where enforcement is costly, states may take an active role in encouraging potential violators to organize and, once organized, regulate themselves. Recent work on the politics of enforcement demonstrates that governments reap political benefits by not enforcing laws against poor citizens in informal work and housing. However, foregoing enforcement can create additional political, public health, and material costs. I extend this work by using enforcement costs to explain why informal workers organize. I suggest that where governments successfully encourage potential violators to organize, governments keep the political benefits of forgoing enforcement while civil society organizations assume partial responsibility for enforcement.

My fieldwork included 14 months in Bolivia and Brazil working as a street vendor, gathering 92 interviews, and administering two surveys. Chapter 1 of the project develops the puzzle---why do informal workers organize, given barriers to collective action---in the context of current research. Chapter 2 presents the theory and then formalizes it in a game theoretic model of collective action. Chapter 3 justifies the model's assumptions and demonstrates how its dynamics work in an ethnography of street vendor organizations and their interactions with the city government in La Paz, Bolivia. Chapter 4 illustrates how the model explains variation across people and places by comparing street vendors in La Paz to their counterparts in the neighboring city of El Alto. I then compare the highly organized street vendors in La Paz to the sparsely organized vendors in São Paulo, Brazil. Chapter 5 tests the theory on out-of-sample data: 26,304 self-employed respondents in 17 countries from the Latin American Public Opinion Project. I analyze the data with logistic regressions and then move to a nonparametric machine learning framework to address concerns about identifying assumptions. I find similar patterns in different types of data across different analytic frameworks, lending support to the theory.