How China Escaped the Poverty Trap

Yuen Yuen Ang

• Publisher: Cornell University Press
• Year: 2016
• Print ISBN: 9781501764561
• eBook ISBN: 9781501706400
• Topic: History
• Subtopic: Development Economics
• Index: History

WINNER OF THE 2017 PETER KATZENSTEIN BOOK PRIZE
"BEST OF BOOKS IN 2017" BY FOREIGN AFFAIRS
WINNER OF THE 2018 VIVIAN ZELIZER PRIZE BEST BOOK AWARD IN ECONOMIC SOCIOLOGY

"How China Escaped the Poverty Trap truly offers game-changing ideas for the analysis and implementation of socio-economic development and should have a major impact across many social sciences."
? Zelizer Best Book in Economic Sociology Prize Committee

Acclaimed as "game changing" and "field shifting," How China Escaped the Poverty Trap advances a new paradigm in the political economy of development and sheds new light on China's rise.

How can poor and weak societies escape poverty traps? Political economists have traditionally offered three answers: "stimulate growth first," "build good institutions first," or "some fortunate nations inherited good institutions that led to growth."

Yuen Yuen Ang rejects all three schools of thought and their underlying assumptions: linear causation, a mechanistic worldview, and historical determinism. Instead, she launches a new paradigm grounded in complex adaptive systems, which embraces the reality of interdependence and humanity's capacity to innovate.

Combining this original lens with more than 400 interviews with Chinese bureaucrats and entrepreneurs, Ang systematically reenacts the complex process that turned China from a communist backwater into a global juggernaut in just 35 years. Contrary to popular misconceptions, she shows that what drove China's great transformation was not centralized authoritarian control, but "directed improvisation"—top-down directions from Beijing paired with bottom-up improvisation among local officials.

Her analysis reveals two broad lessons on development. First, transformative change requires an adaptive governing system that empowers ground-level actors to create new solutions for evolving problems. Second, the first step out of the poverty trap is to "use what you have"—harnessing existing resources to kick-start new markets, even if that means defying first-world norms.

Bold and meticulously researched, How China Escaped the Poverty Trap opens up a whole new avenue of thinking for scholars, practitioners, and anyone seeking to build adaptive systems.

Part 1

FRAMEWORK AND BUILDING BLOCKS

1

MAPPING COEVOLUTION

We do not need to be complex, however, just for the sake of being complex, but we do need to get over our simplicity hang-ups. Obviously, our theories will always be simpler than the worlds we study, or we are trying to reproduce these worlds rather than a theory of these worlds.

—Elinor Ostrom and Xavier Basurto, “Crafting Analytical Tools to Study Institutional Change”

My interviews with local bureaucrats in China deliberately included a question that has long been debated in academic circles: In your locality, do you think it was effective governance that led to growth or growth that enabled the government to improve? The bureaucrats were consistently astonished by the naiveté of this question. To them, the answer is obvious: causation runs both ways. One regular cadre in a city-level agency, who had no scholarly training whatsoever, gave a memorably insightful reply:1

The economy and the bureaucracy interact and change together. If the economy is poor, then, inevitably, it will be difficult to improve governance. . . . In reality, we do what we must and then adjust as we go along. . . . There must be a process. It’s impossible for a government to reform overnight.

More bluntly, another bureaucrat griped that the question posed was flawed. In his words, “To say that we should grow the economy and then improve the business environment or vice versa are both misguided. Obviously, we must pursue both at the same time, like the pursuit of material and spiritual development.”2

These replies suggest that development as a coevolutionary process is a plain reality to many practitioners and probably lay observers too. It is no wonder that they dismissed the question as academic. I concurred heartily with their practical insights and then pressed further: “Could you tell me, from the beginning, how did this process of interaction unfold? How did growth affect the bureaucracy, and in turn, how did bureaucratic reforms affect the economy?” This time, even seasoned bureaucrats were stumped. “It’s too complicated!” was the usual, exasperated response. Instinctively, the practitioners knew that the process of development is interactive and coevolutionary, but laying out and making sense of this entire process step-by-step is anything but easy.

The purpose of this chapter is twofold. First, I preview the steps I will take throughout this book to map coevolutionary processes of development. In order to chronicle exactly how states and markets coevolve, we must first spell out the steps of analyses. Second, I highlight new and important insights we can learn from a coevolutionary perspective that are obscured in standard linear and path-dependent accounts. At the outset, let me underscore three key insights:

1. Conventionally good institutions like formal property rights and professional bureaucracies are institutions that preserve markets after markets have already been built. Building markets, however, calls for drastically different institutions.

2. Market-building institutions typically look “wrong,” that is, inconsistent with best practices, seemingly backward, and prone to corruption. Their developmental potential or role is therefore easily missed or even dismissed.

3. Institutions, once established, do not always self-reinforce. If initial institutions succeed to spur markets, increased wealth changes preferences and resources, which in turn motivate further institutional adaptation.

Before we survey the coevolution of states and markets across thirty years of reform and in thick contextual details, it helps to preview this complex process in a small and stylized setting. Following the style of analytic narratives,3 I will apply my empirical approach to map the steps of a particular case of coevolution—the mutual emergence of industrial markets and professional bureaucracies—in a modeled local polity that I nickname Glorious County (in salutation to Deng’s famous maxim “to get rich is glorious!”). Given the deliberately narrow scope of my analysis in this chapter, I call this a mini–analytic narrative. From this exercise, we’ll extract general patterns of state-and-economic coevolution that will play out on a much larger scale in the remainder of the book.

Four Steps of Mapping Coevolution

The words “evolve” and “coevolve” are frequently invoked in social science analy-ses.4 Coevolutionary narratives, however, are more than just statements that things evolved together or that incremental change occurred. These are intuitive statements that any lay observer can make, as my interviews with the local bureaucrats suggest. Coevolutionary narratives with teeth strive to reveal mutual causal influences between two (or more) domains of interest: for example, how a particular bureaucratic adaptation affects the economy, how subsequent economic changes feed back to the bureaucracy, so on and on, in a zigzag causal chain.5 In particular, a useful coevolutionary account should pinpoint why particular institutions or strategies were selected at given junctures and whether or not these selections continued to fit the environment at later periods.

For the benefit of a general readership, I keep my methodological discussion in this chapter brief. For an elaboration, appendix A details my procedures of data collection, highlighting the design of fieldwork and interview questions to document changes across multiple institutional domains and over time.

The logic of coevolutionary analysis is straightforward: array the traits of selected domains over multiple time periods and then examine the influence of change in one domain on change in the other domain at each juncture. More concretely, I follow four basic steps:

1. Identify two (or more) domains of significance.

2. Identify significant time periods of analysis.

3. Identify dominant traits of each domain in the significant periods.

4. Identify the mechanisms of mutual influence at significant junctures.

While the logic of analysis is straightforward, the hard work lies in implementing these steps and collecting data to map coevolutionary paths. For a start, measuring qualitative changes across institutional domains, across regions, and over time is a formidable task in any national context, not to mention in developing and authoritarian countries, where there are few ready-made and precise datasets to download. In my study of China, nearly all the data had to be collected from scratch through extensive fieldwork and interviews, which generated more than a thousand pages of transcripts in Chinese (for more details, see appendix B). Using this cache of historical-qualitative data, I propose causal links on the grounds of plausibility. Precise testing of each of these numerous links is beyond the scope of this book, but the paths I sketch here lay the necessary groundwork for future ambitious efforts at collecting fine-grained panel datasets that may allow us to test more precisely each proposed link in each causal chain.

Any historical study in search of long-term interactive patterns will run into challenges of data collection. Such challenges are magnified when attempting to trace multistep and multidirectional—rather than linear—paths of change. Nevertheless, these obstacles should not stop us from reaching for a dynamic theory of development. My efforts at systematically mapping the coevolution of states and markets using the qualitative data I collected, I hope, suffices for readers as an essential and worthwhile step toward revisiting the fundamental question of how development actually happens.

The Norm: Good Bureaucracies and State-Led Growth

Studying the coevolution of states and markets is an exceedingly complex task. Both states and markets are large cluster domains, each containing multiple attributes. To preview a coevolutionary approach to the study of development, it makes sense to zoom in on one strand of each cluster. On the state cluster, I prioritize the bureaucracy. More specifically, I will focus on the emergence of professional bureaucracies organized along Weberian precepts or popularly regarded as “good” bureaucracies. Then, with respect to the economy, I will focus on the industrial sector, setting aside other higher-order spheres such as financial markets and corporate governance.

Especially in the study of late-developing economies, there is good reason to focus on good bureaucracies and industrial growth. This is because in poor and rural economies, the conversion of agriculture to industry is a (if not the) principal source of growth. And among late developers, interventionist state policies are crucial for accelerating the process of industrial catch-up, as persuasively argued by the developmental state school.6 But in order for state intervention to successfully foster markets rather than breed corruption, proponents of developmental states stress that the right kind of bureaucracy must first be established.7

So what is the right bureaucracy for state-led development? The answers can be traced a century back to Weber’s monumental essays, in which he spells out the characteristics of legal-rational, professional bureaucracies that depart fundamentally from—and more importantly are superior to—bureaucracies that prevailed in premodern times.

Among the list of attributes Weber identifies, two stand out.8 First, professional bureaucracies perform specialized functions through technically qualified personnel. While nonspecialization was the norm in premodern and patrimonial settings,9 modern corporations and governments all embrace functional specialization as a norm. Thus, Herbert Simon states as a matter of fact, “The administrative organization is characterized by specialization—particular tasks are delegated to particular parts of the organization.”10 Indeed, in public administrations across countries—including in China—every office is formally assigned a specialized function with a delineated scope of expertise and responsibilities (for example, the Commerce Bureau regulates commercial affairs, the Education Bureau administers education services, the Environment Protection Office enforces environmental regulations, and so forth). That specialization is superior to nonspecialization seems uncontroversial. Offices should be more efficient if they specialize in well-defined tasks than if they juggle multiple, overlapping responsibilities. Weber’s observations are aligned with Adam Smith’s emphasis on efficiency gains from the “division of labor” in capitalist economies.11

Second, inseparable from specialized functions is impersonality. Impersonality refers to the conduct of duties and exchanges with parties with whom we are not personally acquainted or related. When we say “keep it professional” in modern parlance, we mean such norms as not recruiting family members into workplaces and not discussing private affairs while at work.12 For public agents, impersonality is associated with integrity. Pu...

Table of contents

• List of Figures and Tables
• Preface
• Acknowledgments
• Introduction: How Did Development Actually Happen?
• Part 1 Framework and Building Blocks
• Part 2 Direction
• Part 3 Improvisation
• Conclusion: How Development Actually Happened beyond China
• Appendix A: Steps for Mapping Coevolution
• Appendix B: Interviews
• Notes
• References
• Index