To what extent does Mobilizing Restraint provide insights that are useful to the study of developing countries beyond South Asia? This question deserves serious attention, but in raising doubts about whether my study has broader applicability, Erik Kuhonta does not engage key arguments in the book as well as much of the empirical evidence that I present to support them.

Mobilizing Restraint takes as its starting point the scholarly consensus that labor repression is an ineffective tool for promoting economic development (pp. 6–7). Previous studies have demonstrated that labor repression was not essential to economic development in East Asia and elsewhere, and that where wage repression was practiced (as in Singapore) the strategy backfired by producing labor market distortions that temporarily hampered growth (e.g., Banerji and Ghanem Reference Banerji and Ghanem1995; Fields Reference Fields1994). In a number of Asian countries, repression also generated higher levels of industrial protest and a more unstable industrial relations environment over the long term (e.g., Freeman Reference Freeman1993).

Although my study provides further evidence of the incompetence of authoritarian regimes when it comes to dealing with industrial conflict, its central question is more specific than whether democracies or authoritarian regimes are better at managing labor unrest. The book asks: What tools are available to developing democracies in managing industrial conflict that authoritarian regimes do not have, and under what circumstances do these tools become effective? Because democratic governments rely on popular support to remain in power, they cannot as easily repress organized labor and must develop more sophisticated methods for managing industrial unrest. I argue that democratic politics gives rise to two unique tools for managing industrial conflict: synergistic state–labor ties and industrial relations institutions such as collective bargaining, labor courts, and labor tribunals (Kuhonta notes the first tool but not the second). Further, I show how democratic competition and the respect for labor rights that can only be found in a democracy have been essential to the functioning and the effective use of these tools in postreform India. As India's democracy became more competitive, major political parties faced increased pressure from their constituents to attract private sector investment in industry. Consequently, major parties drew on ties with unions and formal labor relations institutions to restrain worker protest.

Kuhonta sees little merit in my study's empirical findings. He states that “Teitelbaum's broadest claim … that democracy is better than authoritarianism at managing industrial conflict is untenable. It is untenable given the evidence provided in the book, as well as cursory comparative references across the developing world.” However, he does not engage or refute any of the analysis presented in the book. Let us look briefly at where I find support for my arguments.

First, I find substantial evidence that political competition and synergistic ties improve industrial relations outcomes in an analysis of interview and survey data collected during two years of fieldwork in South Asia. For my field research, I selected four regional cases based on the political structure of the union movement in each case: Kerala, Maharashtra, Sri Lanka, and West Bengal. The qualitative interview data collected in these regions show how the incentives and behavior of major party unions differ from those of politically independent unions (Chapter 3). This anecdotal evidence is further confirmed in a time-series cross-sectional analysis of industrial protest in 15 major Indian states. Here, I find a negative relationship between political competition and industrial protest. A different analysis of industrial violence among 148 Indian firms finds a negative relationship between union–party ties and various types of industrial conflict, including industrial violence (Chapters 4 and 6).

Second, I find substantial evidence that expanding the authority of industrial relations institutions promotes better industrial relations outcomes. For example, an industry-level analysis of industrial disputes legislation and economic performance (34,000 observations) presented in Chapter 5 shows how industrial relations institutions in India promote better economic performance, whereas repressive legislation that limits the ability of unions to go on strike results in worse economic outcomes over time.

On the basis of these four chapters of qualitative and quantitative analysis, I disagree with Kuhonta that there is insufficient evidence to demonstrate the relationship between democracy and industrial relations outcomes in South Asia.

Does the study have broader applicability? While my book examines only South Asia, there are reasons to believe that its findings may be relevant for other regions. As Kuhonta notes, we see similar political dynamics governing industrial relations in regions that have had democracy for a long time, as in northern Europe, and there are some parallels (and some differences) in the Latin American corporatist experience. Further, the relationship that I find in India between democracy and industrial relations outcomes appears to be present throughout the developing world as a whole. In a recent study, Graeme Robertson and I show how democracies do a better job of managing industrial conflict in the context of economic globalization than do nondemocracies (Robertson and Teitelbaum Reference Robertson and Teitelbaum2011). The results are not driven by any one region in the developing world; we find that they are robust to the inclusion of regional dummies, the dropping of each major region from the analysis, and an analysis limited to only Asian countries.

Kuhonta argues that only particular types of parties (e.g., left programmatic parties) will court the votes of unions, and that these parties do not exist in most developing countries. This is an important point, and one I address in my book. I note that in South Asia, parties across the ideological spectrum have forged ties with unions. In other words, it is not simply left parties in Kerala and West Bengal that forge state–labor ties. We do see a greater preponderance of state–labor ties in Kerala and West Bengal, but such ties have been forged in Maharashtra and Sri Lanka as well. And, as I show, the results of my analysis are not dependent on the presence of left political parties or the political dynamics prevalent in the two states with left parties. The political ideology of an affiliated party is never a significant predictor of union protest behavior (p. 109). The centrist Indian National Congress and the right-wing nationalist Bharatiya Janata Party, for example, are just as likely to restrain worker protest as are left parties. The political determinants of strikes are also significant when I drop Kerala and West Bengal from the analysis and when I conduct the analysis for Maharashtra alone (pp. 110–11). The results are also robust to an analysis of union–party ties and strike protest in Sri Lanka (Chapter 6).

It is important to note, however, that my theory does not predict that the political dynamics seen in South Asia should be found in all countries at all times. Kuhonta says that the relationship between democracy and better industrial relations outcomes does not appear to exist for industrial relations dynamics in Africa, in the Middle East, and in some countries in Southeast Asia. One reason might be that these countries lack competitive democracies that respect labor rights, industrial economies, or both. But going forward, in countries that experience a deepening of democracy and an expansion of the industrial sector, my theory would predict an increased emphasis on institutional and political solutions to industrial unrest and less reliance on repression, and a subsequent improvement in the industrial relations environment. It will therefore be fascinating to compare the experiences of South Asian countries with those in other developing regions in the coming decades.