A major debate in the literature on the political economy of development centers on the relationship between regime type and economic development. This debate has been heavily influenced by the East Asian development model, where authoritarianism has often gone hand in hand with high growth rates. In South Korea, Taiwan, Singapore, Malaysia, and Indonesia, development has been propelled by authoritarian or semidemocratic regimes. One key element of this argument is that the repression of labor under these authoritarian regimes has been especially helpful in states' pursuit of high growth rates because it has ensured political stability and checked societal demands.

Mobilizing Restraint wades into this debate by challenging the view that authoritarianism and the repression of labor benefit economic development. In this book, Emmanuel Teitelbaum asks how states in the developing world manage the tension between economic growth and the demands of labor. Through a comparative study of South Asian states, he shows that some states have managed industrial conflict more effectively than others, and have thereby ensured more sustainable economic development.

What explains this variation is regime type, specifically the depth of democracy. Those states that have more competitive democracies and provide freedom of association and collective bargaining rights are better able to manage industrial conflict and, therefore, have better economic outcomes. The overarching point is that democracies ensure that workers are represented and that representation works to mitigate violent conflict.

The argument hinges especially on democratic competition, and it is therefore crucial to understand the underlying logic. Democratic competition forces parties to take into consideration the interests of the working class if they want to win elections. To do this, parties court labor unions and workers by providing them material benefits and political representation. The close ties that develop between parties and unions provide parties with the means to channel the behavior of unions, in particular by moderating, or “restraining,” their mobilizational urge. Hence, the thought-provoking title of the book. Crucially, parties are able to restrain unions and workers because they are encompassing organizations—an idea borrowed from Mancur Olson. With broader developmental concerns, they thus internalize the externalities associated with aggressive behavior. Furthermore, parties provide union leaders with inducements and channels for political advancement that steer leaders in the direction of party pragmatism.

The take-home message of Mobilizing Restraint is that democracy is better than authoritarianism for managing industrial conflict and, therefore, better for economic development. Since democracies can restrain aggressive and violent behavior, they provide a more conducive economic climate for productivity, industrial output, and foreign investment. On the other hand, repression of labor under authoritarian regimes leads to its radicalization, thereby undermining economic development.

Teitelbaum's book is a worthy defense of democracy in developing countries and of the importance of giving labor its due. The central problem in this study, however, is that democracy is lauded much more than the evidence merits. First, the stipulated mechanism for mitigating industrial conflict in a democracy, namely, competitiveness, does not hold theoretically and empirically. Teitelbaum's assumption about what democracy does—forcing parties to court labor—is a theoretical position that works only if one presupposes that certain types of institutions are already present. In the absence of particular types of unions and parties with specific organizational capacities and ideological proclivities, democracy can have no positive effect on industrial conflict. My second point follows from the first: What really matters in explaining the outcomes of industrial conflict is the character and origin of institutions. Third, Teitelbaum's broadest claim—indeed, the apparent motivation of the study—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.

The critical problem with the claim that democracies are effective at managing industrial conflict and, therefore, at fostering a better climate for economic development is the assumption that competition compels parties to be concerned with the interests of the working class. As Teitelbaum puts it: “Democracy requires parties to mobilize workers as a base of support” (p. 11; emphasis added). This statement assumes two things: first, that in their search for votes, parties will necessarily seek to court labor; and second, that there actually is a constituency worth courting that identifies itself as a working class. These assumptions may perhaps be empirically correct in the cases examined by the author. But the sweeping way in which this argument is made—and it bears emphasizing that the broad goal of this study is ultimately theoretical—is deeply problematic. If democracy really has such a powerful effect on party behavior, should we not expect all types of parties to court labor? Surely, this is not the case.

In the developing democracies that I study, it is by no means clear that parties seek to court interest groups in order to win elections. In Southeast Asia, at least three countries today are procedurally democratic: Thailand, the Philippines, and Indonesia. But in none of them do parties seek the votes of labor qua labor in order to win elections, and in none is labor considered a major social force. Indeed, the problem with these democracies, particularly Thailand and the Philippines, is that parties are primarily clientelistic organizations that show minimal interest in the struggles of the working class. In neither party system—where elections are extremely competitive—have parties ever imagined that the route to an electoral victory passed through labor. Furthermore, labor has historically not organized or articulated itself in a manner that could elicit the interest of parties.

The general claim that democracy “requires” parties to cater to labor hinges on the assumption that a particular institutional landscape is present: that parties have some programmatic base, that strong unions exist to represent labor, and that labor exists as a sociological group that can make demands on competing parties. This may be true in parts of Latin America and South Asia, but it is certainly not true in much of Southeast Asia, Africa, or the Middle East. Strangely enough, it is the European historical experience that appears to fit this general assumption better than the developing world.

This takes us to the second issue: the role of institutions. Institutions function in the argument as the intermediate variable through which democracies can restrain labor. The comparative framework is based on contrasting states where the types of unions and party–union linkages covary with outcomes of industrial conflict. In the states where the outcome has been minimal labor violence—Kerala and West Bengal—unions were linked to major parties. In the state with an intermediate outcome—Maharashtra—“narrow interest unions” (NIUs), unions that were not affiliated with major parties and therefore had narrower, parochial interests, were more dominant. These NIUs engaged in acts of violence that hurt development. Finally, in the negative case—Sri Lanka—NIUs gained prominence once state repression had deligitimized the unions linked to major parties. Thus, the covariation between a particular type of union and party–union linkage and the outcome of industrial conflict suggests that institutions are central to the analysis.

Given this variation in the institutional characteristics of these states, it would appear imperative to answer questions about institutional structure and origin. Why have Kerala and West Bengal been characterized by particular types of unions and party–union linkages? Why have Maharashtra and Sri Lanka historically had more narrow interest unions, as opposed to unions tied to major parties? What is the origin of variation in types of unions? Could the presence of programmatic, institutionalized, and left-leaning parties help build particular types of unions and then structure a tight relationship between parties and unions? The choice of Kerala and West Bengal as positive cases in this study reinforces the view that it is historical-institutional patterns—and not democracy—that drive the argument, since these are the only two states in India where communist-Marxist parties have been dominant.

The problem, then, with the author's study is that institutions, especially parties, are left undertheorized. Teitelbaum assumes that institutions in the developing world are of one ilk: “[M]ajor political parties are highly encompassing organizations in low- and middle-income countries” (p. 67). But the developing world is too complex for such statements. Simply looking at a few countries in Southeast Asia undermines such sweeping generalizations. If Teitelbaum had not made broad assumptions that are unsupported by empirics—that democracies force parties to court labor and that parties are encompassing organizations—we would have ended up scrutinizing the institutional variation among the South Asian states much more closely. This would have meant looking deeply at these states' institutional origins, at the way parties and unions were shaped, and at the way they interacted with each other historically. Ultimately, we would have had a richer understanding of party–labor ties across South Asia.

The final point I wish to make addresses the broad contrast advanced in this book: that democracies are inherently better than authoritarian regimes in managing labor relations and, therefore, in facilitating economic development. There are two problems with this claim. First, the negative case employed to illustrate the developmental and normative failures of authoritarianism—and therefore of the East Asian benign dictatorship model—is Sri Lanka in the 1980s and 1990s. But Sri Lanka was never an authoritarian regime in any way comparable to South Korea or Taiwan. Indeed, Sri Lanka is described in various ways that complicate the argument: “fragile democracy,” “brutally repressive authoritarian regime,” “labor-repressive,” and a “democratically mobilized setting.” If Sri Lanka is in fact not an authoritarian regime but a fragile democracy, then it is analytically problematic to use this case as a means for rebutting the East Asian model.

Beyond the evidence presented in the book, we should also be careful about unqualified generalizations regarding how authoritarian regimes deal with labor. Authoritarianism is, after all, a residual category. In Mexico—for a long time a single-party authoritarian regime—the Partido Revolucionario Institutional (PRI) was a solid ally of labor. Only as the country liberalized economically and politically did the PRI begin to shed its prior support for labor. Singapore repressed labor in the mid-1960s and then co-opted unions, but contrary to Teitelbaum's theory, labor did not become more radicalized. In fact, the high growth rates that Singapore achieved under single-party rule improved the conditions of the working class, thereby reducing the impetus for radicalization. These examples of Mexico and Singapore reinforce the importance of giving institutional structures and origins greater analytical weight than regime type.

The wider debate that Mobilizing Restraint engages in—how regime type relates to economic development—is an important one. The study is understandably motivated by unease with claims that authoritarian repression of labor is beneficial for development. But in the effort to debunk theories that purportedly support authoritarianism, it is also important not to laud democracy beyond what it deserves. Democracies vary immensely in their relations with labor, just as do authoritarian regimes. Crucially, the historical-institutional landscape may help explain the predicament of labor and the poor better than does regime type.