Book Review: How China Escaped the Poverty Trap, by Yuen Yuen Ang

Update: this review is now available in Chinese

Following on from yesterday’s book review on an account of Bangladesh’s success, here’s a great book on another developmental superstar – China.

The macro-story on China is well known, but always bears repetition. Emerging from the carnage of the Mao era, China in 1980 had a GDP of $193 per capita, lower than Bangladesh, Chad or Malawi. It’s now the world’s second largest economy, with a thirty fold increase in GDP per capita, based on a textbook-defying combination of one party Communist state and capitalism – in the words of one tongue in cheek official ‘no capitalist state can match our devotion to the capitalist sector.’

Success on this scale inevitably finds many intellectual fathers claiming paternity – China is variously portrayed as a victory for a strong state; free markets; experimentation and for central planning. How China Escaped the Poverty Trap blows the conventional explanations away, drilling down into what actually happened, reconstructing the history of different cities and provinces through years of diligent research.

This book is a triumph, opening a window onto the political economy of China’s astonishing rise that takes as its starting point systems and complexity. Its lessons apply far beyond China’s borders. The author, Yuen Yuen Ang (originally from Singapore) is Associate Professor of Political Science at the University of Michigan.

Ang starts with a classic developmental chicken and egg problem – which comes first, good institutions or economic prosperity? Different camps within academia and the aid business urge developing countries either to ‘first, get the institutions right’ or ‘first, get growth going’ – and then the rest will follow.

Using China as an elephant-sized case study, Ang takes a systems sledgehammer to this kind of linear thinking, and argues that development is a ‘coevolutionary process’. Institutions and markets interact with and change each other in context-specific ways that change over time. The institutions that help to achieve take off are not the same as the ones that preserve and consolidate markets later on.

Perhaps her most explosive finding is that for countries just at the start of their development trajectory, so-called ‘weak’ institutions are often better than ‘strong’ ones. The weak/strong description is imposed by experts from already developed countries who conclude that their institutions are obviously the ‘strongest’, since their countries are the richest. Echoes of Ha-Joon Chang’s work on the history of trade policy here (and he provides a fulsome plug for the dust jacket).

It’s impossible to do justice to the book in a brief review, but here’s a taster:

Firstly, there are three distinct patterns to China’s take off: the changes made by the state are broad, bold and uneven. Broad in that the state makes changes across the whole economy: she has little time for Dani Rodrik’s attempts to identify specific bottlenecks and tackle them one at a time – that is thinking about the economy as a complicated, rather than a complex, system. Bold in that when the state changes direction, it does so big time, and the whole country, or at least its 50 million civil servants, jump and do whatever it takes to achieve the new goals. Uneven because the changes play out differently according to time and place within China, and the leadership is happy with that.

Secondly, to explore China through the lens of complexity, she unpacks three signature processes of co-evolution, summarized in the graphic:

  • Variation: how does the system throw up alternatives and options to deal with particular problems?
  • Selection: how does it select among the variants to form new combinations?
  • Niche Creation: how does it craft new, distinct and valuable roles within the system?

China’s answer to the variation question is a fascinating one. The phrase ‘one party state’ conjures up an impression of a regime intent on command and control. An autocracy like China has plenty of red and black lines (don’ts and dos), of course, but what Ang reveals is a third tier – deliberate ‘directed improvisation’- a ‘paradoxical mixture of top-down direction and bottom-up improvisation’. The state sets broad, and often very vague, parameters and then it is up to officials to improvise within them, often coming up with solutions and innovations that astonish the big cheeses in Beijing. I love this phrase, and it could easily become as prevalent and useful as Peter Evans’ ‘embedded autonomy’ to describe the technocracy of developmental states.

She paints a picture of millions of study groups of civil servants poring over the latest Delphic utterances from the Central Committee (e.g. ‘socialism with Chinese characteristics’), trying to work out what they can get away with, then entrepreneurially beavering away in responses. She has a partial explanation of why the vague rhetoric of leaders has such a galvanizing effect on officials – she doesn’t touch on the stick – the potentially lethal consequences of getting things wrong/overstepping the new line.

But she graphically demonstrates the carrot – officials on misery wages (the state did not have the capacity to pay properly) were allowed/encouraged to take a cut of any economic activity they could generate. Corruption in the form of ‘profit-sharing’ by officials was an integral part of the model (Ang even worries that the current crackdown could hamper further progress). More broadly the state operates a decentralized ‘franchising mode’, with local government at various levels keeping much of the proceeds of growth.

She shows that China did not follow the East Asian tiger route of highly technocratic ‘developmental states’ ‘picking winners’, not least because of a chronic shortage of technocrats – under Mao, anyone showing signs of technical expertise was branded a capitalist running dog (a term we really need to start using again). Instead primitive accumulation – the first steps on the road to take off – built on what China had in abundance: social networks. Every department and senior official was required to hustle for investment from friends, relatives and contacts, and incentivised with juicy bonuses.

That initial open season led to a flood of chaotic, unplanned primitive capitalism and plenty of graft. Over time both


the chaos and the graft changed: Once economic activity was under way, the state started to regulate and shape, becoming more interventionist about encouraging things like complementarity of industries and specialist clusters as the economy progressed; corruption moved from petty (apparently involving lots of free meals for officials) to the big money variety, hogged by senior figures.

The detail is often fascinating – the history of a single city in Fujian; the way she gets hold of and compares performance criteria for township leaders in Shanghai in 1989 v 2009, showing how an initial overwhelming emphasis on economic growth has transformed into a laundry list of political, social, environmental as well as economic targets that no human can possibly meet. Is the model starting to creak?

And then in her final chapter, she leaves China and briefly applies her co-evolutionary analysis to trade in late Medieval Europe, taxation in 19th C United States and – wait for it – the rise of Nollywood as Africa’s global cinema hub. Which brilliantly shows that aspects of her analysis of China can apply more broadly – clear links to the Doing Development Differently movement, which she name checks (DDD guru Lant Pritchett gave her a glowing review).

I’m running out of space. I give up. Just read the book. (or if you like podcasts you can listen to the author being interviewed by Alice Evans)

And what does all this mean for development organizations? She won a Gates Foundation-sponsored ‘New Horizons’ prize for her essay on the future of aid, so I will just have to read it and get back to you.

Tomorrow, the fun bit – comparing the two books and their authors

Subscribe to our Newsletter

You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, please see our .

We use MailChimp as our marketing platform. By subscribing, you acknowledge that your information will be transferred to MailChimp for processing. Learn more about MailChimp's privacy practices here.


18 Responses to “Book Review: How China Escaped the Poverty Trap, by Yuen Yuen Ang”
  1. Alice

    Yuen Yuen’s book is absolutely fantastic, for all the reasons you mention: (1) it explains radical poverty reduction; (2) through rigorous research, e.g. the subnational comparative element; (3) highlights a process that could be supported more broadly, in other emerging economies; and (4) is beautifully written, it’s an absolute page-turner.

    But I have misgivings about some interpretations and reviews of the book. Most commentators have highlighted improvisation and adaptation, and used this book to support the anti-logframe movement. While improvisation and adaptation were clearly central in China, we will not accelerate poverty reduction more broadly merely by financing more DDD/ PDIA style reforms. Instead we need to critically reflect on Development Studies, and why we never saw this. Why were we so blinded to China’s success?

    I think we need to take a step back and recognise the damaging influence of Global North group think, our normative assumptions, and neocolonial insistence on micro-management. These biases had blinkered us to China’s success, its mechanisms, and comparable processes in other places. By mandating specific reforms and processes; curbing organisational autonomy; measuring countries by our standards and telling them of their awfulness; vilifying ‘corrupt networks’, we may have constrained directed improvisation elsewhere. That’s the big problem we need to grapple with.

    For me, the real story of this book isn’t China’s success, but our failure to see it, understand it, and support similar processes more broadly.

    Thank heavens for the brilliant Yuen Yuen Ang for helping us see a bit more clearly.

    • Duncan Green

      Good points Alice, but don’t see an either/or choice between decolonising development studies and taking systems thinking more seriously (what you call the ‘anti-logframe movement’)

      • Alice

        Absolutely, they’re complementary!

        In order to support politically-smart, locally-led improvisation (engaging with complexity, via PDIA/ DDD/ and all other relevant acronyms), we need to decolonise development: broaden our communities of knowledge, question our assumptions, be prepared to do things differently.

        Decolonisation is a precondition.

        • Alice

          Another point, which I added to your review of Naomi Hossain’s book, is:

          How can international development community support pro-poor directed improvisation?

          [The CCP promoted a particular kind of improvisation to achieve its end. That created an ‘authorising environment for experimentation’ – to borrow a term from Matt Andrews, Lant Pritchtett and Michael Woolcock’s book. How do you support that kind of pro-poor authorising environment more broadly?]

          If used carefully, aid is certainly one mechanism.

          But there are other options in the toolkit.

    • Alice, thanks a lot, I agree with your broader points. But I am very hesitant to label China as a “success.” That’s because once labeled as a “success,” any flaw or failure will give skeptics (who don’t read and don’t want to listen) reasons to say, “But look, there’s corruption! Smog! Growth is slowing down!”, and thereby discredit the entire experience.

      This was exactly what happened to the 1990s literature on East Asian tigers (developmental states). They were praised as successes before problems came, but once the Asian financial crisis hit, the entire model was booed down the stage.

      We don’t label or think about the United States as a “success,” do we? We embrace all the good and bad that comes with capitalism and democracy. The US is allowed to exist as it is without being stuffed into black-and-white boxes: success or failure. Hence, even after the 2008 financial crisis, started in the US and that nearly wiped out the global economy, this massive disaster still did not negate market capitalism. People were more realistic about the shortcomings of the system. That’s all.

  2. The review by Pritchett that Duncan’s mentions looks at Yuen Yuen Ang’s ideas through the PDIA lenses:

    The review includes links to the slides of her presentation on Complexity & Development 2.0 ( ) and to a video of her presentation of the book at the World Bank (,+2016./1_70wkqd8l/29528271 ).
    All very interesting!!

    Clearly, Ang’s notion of “directed improvisation” can be seen as an effort to induce the emergence and amplification of Positive Deviance, but at an enormous, chinese scale.

  3. Many thanks to Duncan for a fabulous review, not only for your positive words, which I very much appreciate, but for capturing the essence of a chunky book so aptly and so brilliantly.

    Effective reviews are like mirrors, reflecting what’s most important, so I learned a lot from your review. I was struck that China is probably the single best illustration of “systems thinking” and “complexity economics,” not in the fuzzy sense but with plenty of concrete, creative mechanisms. This is paradoxical—but then, isn’t our world today full of paradoxes?—that the textbook case of adaptive development in the face of complexity is an authoritarian, communist party-state.

  4. Thank you Alice for your kind words!

    I am optimistically, maybe naively, hopeful that change will come along the lines you describe: “broaden our communities of knowledge, question our assumptions, be prepared to do things differently.” I’m thinking Lant and Michael’s point about “when the solution [good governance] is the problem” (written 13 years ago), PDIA, Doing World Development Report 2017, Duncan on systems thinking…

    The tide has just begun to shift.

    BUT the speed of change will depend on what we say, debate, and do in years to come. It’s not automatic. Right now, I would say there’s still a lot of skepticism and resistance. This happens whenever an original system (assumptions, ideas, and institutions) is being challenged. We also need to lay a rigorous foundation for a new way of thinking, so that creation follows destruction (hence “creative destruction”), and I am hopeful that’s what academics like me are supposed to do.

  5. Thanks for reviewing what I agree is a groundbreaking (and hugely readable!) book. Yuen Yuen Ang’s analysis, and China’s success in lifting 800 million people out of poverty – highlights the need for thinking, policy and practice to move beyond the “Good Governance” mantra, and adds much to the emerging outlines of an agenda that focuses on local innovation, learning and adaptation. A recap of the event held at the OpenGov Hub in Washington last week, with Yuen Yuen Ang, Edouard Al-Dadah from the World Bank’s World Development Report 2017 team, and Shanthi Kalathil from the National Endowment for Democracy, including links to the video and my notes from the book.

    Anyway, two questions for you Duncan.

    First, were you not itching to hear more about the role of citizen action in China’s development, or of how accountability plays out without democracy, in Lily Tsai’s terms?

    Second, partly stimulated by Winnie Byanyima’s powerful contribution at the Open Government Partnership’s meeting in New York earlier this week: how did you feel about the apparent marginalization, or at least postponing, of the argument for democracy and rights that seems to be implied by Yuen’s analysis of China’s escape from the poverty trap?

    Video of Winnie’s remarks at 53:40 here – with collection of pieces put together by the Open Government Partnership on trust, here

    Finally, this piece from AidData colleagues, in the Washington Post, directed at USAID, is awesome, including on the importance of assessing the functional effectiveness rather than the form of institutions.

    • First of all, Alan, thank you for having me at GI. It was both intellectually enriching + FUN, which is the best experience any speaker could ask for.

      Your point about “postponing democracy and rights” raises a very important debate. I wouldn’t, however, think of the issue as “postponement.” That would take us back to the linear thinking of “growth first” or “democracy first.”

      Instead, I would suggest thinking along these lines.

      [1] Carve out pockets of openness, innovation, and freedoms within an existing authoritarian government. The CCP is authoritarian, and yet there is bottom-up initiative, deliberation & creativity within the bureaucracy.

      [2] “Civil society” may not only mean citizens—it can encompass the civil service (in China, this is an organization of 50 million people, size of South Korea’s population!).

      [3] It’s possible to expand rights and constrain power without formal democratization. Many people are puzzled why China hasn’t democratized despite rapid growth? This is a false puzzle, because political changes have in fact occurred within Chinese society. To give a simple example, in the past, police officers could easily harass and extort citizens. Well, try that in Shanghai today, and see what urban and educated Shanghainese will do. Their rights may not be complete, but citizens do exercise them.

      That’s not to say that we shouldn’t aim for democracy. Rather, I’m suggesting that steps 1 and 2 pave the foundation for a rooted democracy. We have seen what happened when dictatorships are abruptly uprooted and replaced with the formalities of democracy in a big bang.

      • Alice

        On civil society, I’m looking forward to reading Diana Fu’s forthcoming book: “Mobilising without Masses”.

        I think you (Alan and Yuen Yuen) both make important points about people’s expectations: how do urban and educated (as contrast with precarious migrant, or rural) Chinese expect of government? Do they perceive it as tolerant, responsive and capable of providing various public goods – e.g. education, health, employment rights, property rights? Those norm perceptions will shape how they interact with bureaucrats, police and other government officials. For

          • There are probably 3 libraries worth of literature on state-society relations in China! Dickson’s book is a highly readable and balanced, using surveys of public opinion, urban and rural. He finds, “while the Chinese people may prefer change, they prefer that it occurs within the existing political framework.”

            One point is worth highlighting. There are different types of rural residents and farmers. They’re not a homogeneous group. Some are bullied and acquiescent, some are strategic and skilled protestors (e.g., they know how to use central slogans to combat local officials and get their way), and a small number are known as “nails” (recalcitrant, even extorting).

          • Alice

            Yuen Yuen [replying to your comment below],

            I didn’t convey myself clearly, I meant that perhaps the subnational differences documented in your book (i.e. the lagged inland process of experimentation and belated improvements in governance and economic development) were influenced by a wider set of factors than the ones you mention (i.e. economic opportunities, supply of inward investment).

            Maybe subnational processes of governance and economic development were also influenced by ‘ideas’ & ‘norms’: e.g. local expectations of government; how rural campaigners, hardened collectivists, upwardly mobile farmers, successful business people etc anticipated the government to react to various demands; their perceptions of state tolerance, capability etc for various issues [echoing Duncan’s nod to norms in the trilogy blog post]

    • Duncan Green

      thanks Alan, great additional links – it’s these accumulating conversations that make blogging useful! On the citizen action point, I was just happy to read something original, based on deep and prolonged scrutiny of the actual system. But yes, it would have made a good additional lens – the book really only deals with the state machinery (and is quite open about that). Yuen Yuen’s next topic?

      • Thank you all for this wonderfully stimulating conversation.

        Yes, state first, and next book is on corruption, in which citizens may either participate and/or resist.

        Also, the next stage of China’s adaptive power will need to come more from society than state, as I write in the conclusion. The first 35 years had to be state dominant, as China started out as a communist-authoritarian system; hence, bureaucrats were the main actors of adaptation, innovation, and change. The next lap will be a different story… to play out in decades to come!

  6. A bit late to this conversation, but enjoyed Duncan’s book review and the discussion here which it created. The link to the WB presentation is really good as well – excellent presentation.

    Duncan’s insights and Yuen Yuen’s analysis have prompted me to share something that I hope is of interest and resonates with the theme of the book – but from a different angle. In May this year two colleagues and I reviewed a DFID funded education project that we helped implement in four very poor counties in Gansu, western China, between 2000 and 2006. So, an unusual opportunity to look back ten years after completion and reflect on what “sustainability” might actually look like (something we talk a lot about in development) and to see what kind of changes had taken root.

    Much of what we found in terms of what was sustained and what was not – and why – echoes the points Yuen Yuen is making about the room given to experiment while working within a (loosely) defined policy umbrella (Chinese has an oft quoted phrase : Shang mian you zhengce, Xia mian you dui ce – up there you have your policy, down here we have our way of implementing it). I particularly like the term “directed improvisation” which I think captures this well and which for us too was really the modus operandi of this project.

    Although we were looking at education development rather than markets – the role of weak institutions and how they become stronger is a common theme. How that strengthening is influenced by good leadership and by organisational culture is key.

    Of course, in all of this there is a view of China as a development outlier whose success is too specific to hold lessons for others. Like Yuen Yuen I disagree with that and think there is lots to learn as long as we separate out the generic lessons from the ones with only “Chinese characteristics”.

    If anyone is interested, our review from this trip (called “The ripple effect” ; is a reflection piece rather than research and is told from a practioners angle.

    We also made some short films about the visit and some of the children who received financial support to go to or complete schooling. We’ve followed them over a 13-year period (a bit like Yuen Yuen’s time slice methodology) so can show some interesting perspectives both on how education broke the intergenerational cycle of poverty – but, also how it takes more than education to escape the poverty trap.

    The films are :
    Escaping Poverty (,
    Educating a Generation (
    and Gansu Revisited (

    Thanks for this book review, the book and the insights. Looking forward to Yuen Yuen’s new book on corruption.