For quite some time now the ‘new institutionalism’ has dominated the study of Latin American government bureaucracies. It has taught us much, although mostly about their pathologies – such as obstacles to getting incentives right – with the occasional nod to ‘islands of efficiency’. Because the quest for the Holy Grail of Weberian legal-rational bureaucracy seems so daunting, it is easy to conclude that the Latin American public sector is, with few exceptions, woefully, if not pitifully, dysfunctional. These perceptions become amplified when examining smaller agencies charged with oversight of issue areas relatively low on the list of national priorities, such as the environment. The steady drumbeat of deficiencies has left me, and others I am sure, wondering: Well then, how do these institutions work? For, perhaps in more instances than we imagine, and assuming the bureau's function is to address a public issue, the business of the state somehow gets done even in the absence of Weberian bureaucracy. How does that happen?

Greening Brazil, a breakthrough book, makes an outstanding contribution to this puzzle. It demonstrates how small agencies in low salience issue areas confronting powerful detractors survive, expand and make a difference. Kathryn Hochstetler and Margaret Keck persuasively argue that extensive interpersonal and professional networks carefully cultivated by key leaders, along with their finely honed discernment over which battles to fight and how to fight them, are the key explanatory factors.

This analytic insight courses through the book which, more faithfully characterised, focuses on the role played by activists in state and society in the shaping of Brazilian environmental politics. It is the story of how networks of environmental activists, beginning from practically nothing, became involved with the state, and turned a fledgling, weak agency into a player in Brazilian environmental policymaking. It is also the story of the transformation of Brazilian environmental activism and its relationship to the state over time.

Without neglecting the importance of transnational networks in Brazilian environmental politics, Hochstetler and Keck forcefully argue that domestic politics played the more central role. The key networks involved national, state and local government institutions and their interplay with activists, interest groups and nongovernmental organisations in civil society. These linkages between activists in government bureaucracy and in civil society, along with transnational networks, provided critical resources at pivotal moments of challenge and opportunity. In the process, Brazilian environmental politics experienced a transformation. In Amazonia its focus shifted from nature conservation during the military dictatorship to a distinctive socio-environmental approach – which addresses both environmental destruction and social injustice – during the transition to democracy and beyond. In urban areas the discussion shifted from pollution control to sustainable cities.

Another refreshing innovation permitted these insights. Hochstetler and Keck drew on a burgeoning literature in contentious politics that examines the role of networks, political opportunities, and transnational connections in episodic, discontinuous social movement trajectories. This approach takes into account both internal movement development and a movement's relationship to the state. Thus, the contentious politics approach addresses head on the question of what happens to social movements when their causes become institutionalised politics. Hochstetler and Keck take that one step further. How do state institutions born of social movements capitalise on that relationship for their functioning?

To be sure, cultivating dense networks of professional and personal relationships across many levels of government, civil society and transnational organisations is time consuming and requires a unique set of personal skills to manage effectively. Therefore this arrangement probably is less ‘efficient’ than one that built on universal rules and procedures with carefully defined boundaries of authority and power. Fortunately, Hochstetler and Keck are not constrained by a priori notions of efficiency. They are interested in what is. As a result, at least in this one instance, researchers have illuminated the actual practices by which an underprivileged government agency in Latin America survives and, within its limitations, strives to effect some positive change. We have learned something both theoretically informed and real about the functioning of Latin American institutions and the role of activism in that process. That is quite an achievement.

Greening Brazil also opens the door to exciting new directions in research. Do comparable networks generally emerge in the transition from ‘new social movement’ to more institutionalised policymaking in Latin America? Do they perform similar functions? The book also invites more systematic research on the role of personal and professional networks in Latin American bureaucracies at all levels of government and their possibly synergistic relationship to civil society. Is there overlap with the concept of ‘embedded state autonomy’?

Hochstetler and Keck have performed yet another great service with their choice of methodology. They not only demonstrate (again) that qualitative research produces useful work, but that insightful results also emanate from approaches that do not ‘reflect a classic research design with a well-formulated initial hypothesis, investigated in the field, and analysed and reported promptly [but that it can also come from the synthesis] of a whole series of research efforts … over a number of years’ (p. ix). Indeed, their approach is probably as close as political scientists can get to a political anthropology of the relationship between movements and the state – with participant observation and all.

This is compelling stuff, and graduate students should read it. Although most of us may counsel them not to follow this method, it is useful for them to reflect on personal experience as a source of insight into problems in Latin American studies. Moreover, the book is a vivid example of how to advance knowledge, informed by theory, on the real workings of Latin American institutions beyond deductive analyses of pathologies in institutional design followed by prescriptions on how to fix them. It also encourages us to place the accomplishments of social movements in perspective rather than dismiss them (or become disillusioned) because they seemingly lack the capability to transform the society they contest.