Is China a free market success?

I don’t know why I get so annoyed by these endless debates over the “lessons” of the “Chinese miracle,” but I do.  Here is Ezra Klein (and Matt Yglesias):

“Something that emerges quite quickly and a bit unexpectedly from being taken around on an economics-focused tour of China is that the Chinese economic miracle is really a great deal less of a ‘free market’ miracle than the conventional understanding in the United States would suggest,” writes Matthew Yglesias. That’s definitely correct, though it’s worth wondering why anyone thinks of China as a free-market success story.

It would really help if people stopped calling one of the poorest economies in East Asia a “miracle.”  The problem seems to be that people insist on thinking in binary terms; free market or communist, successful or unsuccessful.  Here’s what really going on.

1.  In 1979 China had an incredibly inefficient communist system and was poorer than India, poorer than sub-Saharan Africa.  It was an almost completely statist economy with incomes below 10% of US levels.

2.  China has moved to a mixed economy.  These reforms might allow China to eventually reach something like 60% of America’s per capita GDP (which isn’t very impressive.)  During the transition from Maoism to this mixed economy, China can expect to grow really fast.   There is no miracle here; all the other capitalist East Asian economies also grew fast during earlier decades.  Chinese incomes will plateau well below US levels without further market reforms.  If they take further market reforms (and it seems almost certain they will) then they may plateau at 80% of US incomes, like some other developed economies.  If they go toward a low tax, relatively free market model (such as Hong Kong), they might even surpass the US some day.

3.  East Asian countries with Confucian cultures that have much fewer SOEs are much richer than China, just as you’d expect.

4.  East Asian countries with more SOEs than China (North Korea and Vietnam) are poorer, just as you’d expect.

5.  Regions of China dominated by SOEs (i.e. the Northeast) have grown slower, just as you’d expect.

6.  Regions of China dominated by private business, rather than SOEs (Zhejiang) have grown fastest, just as you’d expect.

What’s so hard to understand?  SOEs are usually inefficient.

Of course there are many other complications:

1.  The inefficiency of SOEs is not due to government ownership, but rather government subsidies and barriers to entry.  Without those factors, SOEs can be highly efficient (see Singapore.)

2.  There may be areas (such as transport) where the government should be involved.  Ezra Klein mentions how good the Chinese government is at building infrastructure, and I agree.  But that has nothing to do with whether the Chinese system of heavy state involvement in manufacturing and services is wise.  It isn’t.

3.  In a related post, Tyler Cowen makes a very good point:

It’s also possible that the successes of state ownership “decay” with time, as was arguably the case with the French model before the privatizations and has been the case with NASA in the United States.

I strongly agree.  This is one factor that hurt the Soviet economy over time.

So to summarize, to the extent that China is a free market, it is an economic success, and to the extent it is statist, it is mostly a failure (excluding some sectors like transport.)  But the question “Is the Chinese miracle due to a free market economy?” is nonsensical.  It isn’t a miracle at all; it is a country rapidly transitioning from being extremely poor to having a so-so economy.  That is all.

I want to personally apologize to Ezra Klein for the exasperated tone of this post.  His post is no worse than 1000 other similar posts; I’m not even sure he disagrees with me.  Indeed the conventional view in the US is wrong, just as he says.  It’s just that I just get annoyed seeing the debate constantly framed this way.

BTW, Tyler also links to this Yglesias post:

By the lights of the American conventional wisdom, this whole situation seems mostly like a warning sign””beware! you’re violating the terms of The Washington Consensus!””but nobody can doubt that they’ve had a great run for the past 20 years. What’s more, though the prevailing policy consensus in the United States would lead you to believe that a country with France’s policies would necessarily be a basketcase, France is itself a very successful and prosperous nation and society. A crucial difference, however, is that the French economy takes place against the backdrop of the kind of social welfare provision that you see in all different kinds of European countries, something that’s overwhelmingly lacking in China.

There is some truth to this, but one shouldn’t push this comparison too far.  The French state does seem to handle its duties better than almost any other bureaucracy.  But let’s not forget that France also has a very strong private sector.  China would have to do a massive amount of privatization to approach French levels of state ownership in manufacturing and services.  France has private banks that lend to private firms (something pretty rare in China.)  And the fact that even Mitterrand privatized French SOEs in the 1980s suggests that the French recognized that the old model was slipping.  At some point the Chinese will have built up their infrastructure, and investment will slow sharply.  Consumer spending will rise.  At that point they’ll be much better off with a free market model, as compared to an SOE-dominated model. 

One other point.  Yasheng Huang (who knows much more about this than I do) argues that the infrastructure projects that impress Western businessmen, and even Western liberals, are very anti-egalitarian.  He argues that more free market reforms are needed to improve life for the rural population, whereas instead the rural people are being exploited to build fancy 325km trains, maglevs, and ultra-stylish airports.  Something to think about.  Anyone who reads Huang will come away with a much different impression of China from what’s presented in the news media.



32 Responses to “Is China a free market success?”

  1. Gravatar of Alex Alex
    1. June 2010 at 13:19

    “East Asian countries with Confucian cultures that have much fewer SOEs are much richer than China, just as you’d expect.”

    Did you read the paper I just ran 4 million regressions by Xala-i-Martin?

  2. Gravatar of Alex J. Alex J.
    1. June 2010 at 13:51

    I’m going to guess that you know enough about China to know you know very little. That makes you annoyed with people who act like they know a lot when you can see they know less than you.

  3. Gravatar of david david
    1. June 2010 at 14:41


    Oh my god, I can’t believe that’s a paper called that. That is awesome.

  4. Gravatar of Mike Sandifer Mike Sandifer
    1. June 2010 at 15:25


    To quote you: “1. The inefficiency of SOEs is not due to government ownership, but rather government subsidies and barriers to entry. Without those factors, SOEs can be highly efficient (see Singapore.)”

    You are far from being a conservative in the American sense. Maybe it’s a measure of how far right the country’s moved since the early ’80s. Obviously, many conservatives today would call you a socialist, but then nuance doesn’t seem to matter anymore, if it ever did.

  5. Gravatar of Paul Zrimsek Paul Zrimsek
    1. June 2010 at 18:52

    You’ve been around long enough to know how this game is played! China is a centrally directed economy, and will continue to be until it suffers an economic collapse– for which its free-market policies will be to blame.

  6. Gravatar of Matthew Yglesias » Brief Reply on the Chinese Model Matthew Yglesias » Brief Reply on the Chinese Model
    1. June 2010 at 20:01

    […] offer a brief reply to some lengthy posts by Tyler Cowen and Scott Sumner I don’t intend anything I’ve said about China to be read as saying “China has a […]

  7. Gravatar of Sean Sean
    1. June 2010 at 21:09

    Matt’s reply: “I’m simply saying that I think most American discussions of China overrate the extent to which the Chinese economy has been liberalized.”

    He has no clue how much it has liberalized. I’ve only been in the country a decade, but talking with friends about how they would go get their weekly vouchers as teenagers gives me a very slight idea. Even in the time I’ve been here the government interference in most day-to-day business has significantly declined. The influence of the party representative is much less now.

    Matt sounds like one of the idiots who goes to North Korea, assuming that it is one big prison. After seeing people walking around freely in 1950s-level architecture they come to the conclusion that “most American discussions of North Korea overrate the extent to which North Koreans are held hostage.”

  8. Gravatar of johnleemk johnleemk
    1. June 2010 at 21:13


    Scott has avowed before that he supports the redistribution of wealth in principle. There is no doubt in my mind that to many on the right and perhaps even the left (at least in America), Scott is a socialist.

  9. Gravatar of Sean Sean
    1. June 2010 at 21:17

    Actually, I’m being a bit unfair. I think he does have a better-then-average grasp on the Chinese economy. But what truly irks me is the assumption of what people think of the country. Just because he made an erroneous assumption, he shouldn’t attribute that misunderstanding to “most American discussions.”

  10. Gravatar of scott sumner scott sumner
    2. June 2010 at 04:21

    Alex, No, but maybe the title turned me off.

    Alex J. I suppose that’s human nature.

    Mike, Yes, but I should add that I don’t recommend government ownership. But if you are going to do it, the Singapore way is better than any other. I consider myself a right wing liberal–someone with liberal values who generally thinks the free market is the best way to achieve those values.

    Paul, That’s a very astute observation. There is also this: The left says the Dickensian working conditions and pollution are the free market China, the shiny new infrastructure is the statist part. (Actually, living standards are highest in the more market-oriented parts, but you’ll never hear that.)

    Sean, That’s a good point. China has liberalized just as much as people think, where Matt is right is that they haven’t become as liberal as many people think. That’s because they started with a system that was almost unimaginably statist (and totalitarian) by our standards. First time visitors don’t see the changes. BTW, many of Matt’s posts on China are very good.

    johnleemk, I support almost anything in principle, as long as it makes society happier. Even communism. As a matter of fact, however, I think the happiest societies have Danish values, Swiss democracy, and Singapore-style economic policies (low taxes, forced saving, free trade, etc.) I think communism is about the worst system.

  11. Gravatar of Paul Zrimsek Paul Zrimsek
    2. June 2010 at 05:07

    My observation was really just a special case of the Iron Law of Mixed Economies: It is impossible to regulate a mixed economy so stringently that any economic crisis it undergoes will not be blamed, by nearly everyone, on “the free market” instead of the mixed economy.

  12. Gravatar of david david
    2. June 2010 at 07:24

    Since we’re now discussing developing nations rather than nations-that-have-gotten-there like Singapore, it’s worth noting that Singapore’s own SOEs were very efficient even when they were explicitly nationalized and government-managed through the 70s and 80s. The quasi-privatization process did not significantly alter the economic landscape.

  13. Gravatar of johnleemk johnleemk
    2. June 2010 at 07:37


    Although I’m from the region, I’m not terribly familiar with the workings of Singaporean SOEs in the 70s and 80s. However, my impression is that even then, the executives of these SOEs were held accountable by the government to a much larger degree than their counterparts were in other countries.

    My pet theory is that because Singapore is so blatantly undemocratic (in the sense that its leaders are more than willing to ride roughshod over public sentiment), they are able to make a large number of decisions that leaders in more democratic societies would find difficult to make.

    The flip side, of course, is that there’s no guarantee that the state will continue to make the right decisions, and there are obviously many cases where they have erred. But in general, there are few sacred cows in Singapore relative to other countries, which does make it easier to shift gears if something is going wrong.

  14. Gravatar of Ed Dolan Ed Dolan
    2. June 2010 at 10:58

    Scott writes . . .”These reforms might allow China to eventually reach something like 60% of America’s per capita GDP . . . Chinese incomes will plateau well below US levels without further market reforms. ”

    I think this is an important point. It reminds me of taking my first course on the Soviet Economy in my grad student days way back in the 1960s. The textbook we used featured a graph showing US GDP higher than Soviet GDP but rising more slowly, and projecting a crossover sometime in the 1980s. It didn’t happen because there were no reforms and the Soviet economy “plateaued” (we usually stay “stagnated”) just as Scott expects the Chinese economy to do. By the 1990s, Gorbachev realized what was going on, but his version of reform–perestroika–completely misunderstood what a market economy was all about, so it ended in catastrophic failure rather than renewed growth.

  15. Gravatar of Alex Alex
    2. June 2010 at 12:07

    It´s a fun paper to read and INHO a very honest paper too. Xala-i-Martin basically gets something like 60 variables that are commonly used in the Barro style regressions that try to find the relationship between growth rates and X factors. Then he starts running all the possible regressions he can with combinations of those variables. He finds that the only variable that always shows as having a positive effect on growth is the share of people of confucius religion. It is also a great paper to cite when you have to write a referee report on a paper that presents yet another Barro style regression. If you don´t like the paper you can just say it´s been done before by Xala-i-Martin.


  16. Gravatar of Tom Waye Tom Waye
    2. June 2010 at 14:41

    Yasheng Huang’s work on China certainly is very interesting and at the same time fairly readable to a lazy undergrad like me. I have to disagree with Professor Sumner’s assertion that China will almost certainly embark on further reforms; as Huang writes, government policy much more strongly favoured free enterprize in the 80s, often explicitly over SOEs (as such incomes grew a lot faster during that decade). Many reforms were reserved in the early 90s as the Chinese-state made a massive power-grab. Naturally income growth was slower in the 90s than the 80s. My point here is that if the government was willing to reverse reforms once to tighten the reigns it might well be willing to do so again rather than continue liberalization.

  17. Gravatar of Joe Joe
    2. June 2010 at 18:04

    Great post. I was waiting for your response to this.

    I follow that with a total non-sequitur…

    Will you at any point be responding to the recent debate concerning the oil disaster and whether regulation or 100% liability torts are better?

    Krugman, had a post arguing for regulation, with good commentary by readers,

    Jeffrey Miron, had one arguing for forcing companies to post “insurance” bonds,

    Most libertarians are for 100% liability tort law, liberals for regulation.

    It would be great to hear your views.



  18. Gravatar of david david
    3. June 2010 at 00:39

    Johnleemk, yes, that is also my theory. Operating at high efficiency or privatization both reduce rent captured by given interest groups, so countries that can privatize easily probably also could operate SOEs at high efficiency (if nothing else, by merely dangling the threat of privatization).

    China today is likely more subject to interest-group-driven pressure than Singapore in 1980, which makes both privatization and efficient SOE operation politically untenable.

  19. Gravatar of Mattias Mattias
    3. June 2010 at 03:32


    Where do you think China is today then?

    Is it close to the level where it will stagnate if it doesn’t reform or can it grow fast for another decade before it reaches that critical level?

  20. Gravatar of scott sumner scott sumner
    3. June 2010 at 05:09

    Paul, I completely agree.

    David, Yes, and am I correct that even when government-owned they were forced to operate in a more business-like way by not having a lot of protection from competition and subsidies?

    johnleemk, I agree, and would add that the elections there are a sort of safety valve. I understand that Singapore is not democratic in the Western sense. But voters do have an opportunity to voice dissatisfaction. The key test will be how the government acts if that dissatisfaction ever reaches levels that suggest they’d rather have another government. My understanding is that thus far the voters have been reasonably satisfied with the government (at least overall.)

    Ed Dolan, Yes, the Samuelson text predicted that the Soviets would surpass us in GDP/person by 1990.

    Alex, That’s very funny. I guess I would like it after all. The Cunfucian thing reminds me of a happinessness regression I did with 32 developed countries. I put in a dummy for the 5 Asian countries, and the coefficient was negative 40. Happiness was measured on a arbitrary scale of 0 to 100 where zero was the least happy and 100 the most happy. So an Asian country with the economic characteristics of a country at the 70th percentile, would have an actual happiness level of 30.

    Tom Waye. My impression is that after the post-Tiananmen reversals, reforms started up again in the 2000s. There have certainly been lots of privatizations, and I think changes were made around the time of joining the WTO that could be viewed as reform. So I am a bit more optimistic, although I concede we don’t know for sure. The cities seem vastly more market-oriented to the human eye. When I visited Beijing in the early 1990s we were herded into the drab government-owned “Friendship store.” Now Beijing is full of huge, glitzy shopping malls with Western and Chinese stores. Urban housing privatization is far advanced from the early 1990s.

    Joe, I commented earlier somewhere, but only in the comment section. I had linked to Tabarrok’s post criticizing Krugman’s claim that there was a $75 million liability cap. Hasn’t BP stock lost $10s of billion in value? That suggests that Tabarrok was right. It’s not clear that there is a market failure here. It seems to me there is a failure of creative thinking. I like Robin Hanson’s idea of nuking the well. It’s generally not a good idea to draw up regulations based on an extreme event. (Note the Finreg fiasco.)

    David#2, I agree. But I would add that China’s huge size offers one advantage–it makes for a more competitive market. State-owned auto firms must compete for customers with lots of other state-owned and private firms. (There are some internal trade barriers, but China is still pretty competitive.

    Mattias, My sense is that it can grow fast w/o further reforms, although the pace will gradually slow. I mentioned the 60% figure, but note that China is currently only about 15% of the US GDP/person, even on a PPP basis. That’s why I don’t consider myself a “China booster” when I say rapid growth is likely to continue. The real China boosters are people like Fogel–who said China would double Western European living standards later this century.

    BTW, I think that China is continuing to reform its economy, although I’d like to see much faster reforms.

  21. Gravatar of Karthik Karthik
    3. June 2010 at 09:13

    Unless there’s elitist intentions, it would help if acronyms are expanded the first time they are used. What’s SOE?

  22. Gravatar of david david
    3. June 2010 at 12:47

    David, Yes, and am I correct that even when government-owned they were forced to operate in a more business-like way by not having a lot of protection from competition and subsidies?

    Not quite. Buses and a variety of other domestic-facing industries did not allow competition for a period; licenses were required (and not granted). Wages were also set by the government via the National Wages Council, and prices also set by fiat via another since-dissolved statutory board whose name escapes me at the moment. But generally the motive here was not leftist capture but a suspicion of both employees (often Chinese-speaking and leftist) and employers (often Chinese-speaking and collusive in the view of the English-speaking government – which had nearly been toppled from power by the aforementioned Chinese-speaking community twice by 1963). So the government seized the reins and prevented the creation of new competitors.

    In the mid 70s the government had relatively little need to seek democratic approval – less than post-Goh Singapore today – and was focused on efficiently serving any foreign investment that might be attracted. So the government had an motive in maintaining efficiency rather than serving an interest group it was quite intent on destroying.

    By the time privatization restarted in the late 80s, the managerial corps of most industries had been replaced with English-speaking Western-educated people, and the institutional power of the Chinese-speaking social elite had been sharply curtailed (just in time for Lee to change his mind and declare that the campaign to westernize Singapore had gone too far. Asian values! Asian values! Change the road names back!).

  23. Gravatar of david david
    3. June 2010 at 14:06

    And I should add that the notion of Singapore’s government having to adjust legislative rules for competition, etc. in order to give SOEs advantages or create competition would be quite strange, given Singapore’s tiny size – as I said in a comment to another of your posts, the people leading the businesses were often also the government.

    Essentially, imagine if the US were small enough for the Secretary of Transportation to be CEO of every bus, train, and taxi service by himself. And let’s suppose that he has enough pull within the entire Justice Department to make your life horrible in so many little ways, which may not be explicitly cruel but not really enjoyable either. Plus, he also has pull within every single union and labor group, and any dissent there is also subject to Justice Department attention.

    Would he really need Congress to declare United States Bus Service a protected monopoly for it to deter competition, if he wanted it to be a monopoly? And if he wanted to operate USBS efficiently – presuming that the equilibrium number of such businesses really happens to be one in Tiny United States – would Congress need to change anything for it to happen?

  24. Gravatar of ssumner ssumner
    4. June 2010 at 05:51

    Karthik, Sorry, I meant State Owned Enterprises.

    David, Thanks for the info. I guess Singapore moved toward the market just as most other countries in the 1980s.

    On your second point, I think you misunderstood my argument. I certainly consider all those petty roadblocks you mention (which are abuses of government power) to be barriers to entry. That is the issue that is being debated. I had read that Singapore is far less corrupt than other Asian countries. Now that may be wrong. But if it is right then it implies there are far fewer of those abuses in Singapore than elsewhere. You see it score very high on things like “ease of doing business.”

    By the way, Singapore is about the size of Denmark, and I consider Denmark to be a real country. Indeed I think 5 million is close to the optimal size for a country.

  25. Gravatar of david david
    4. June 2010 at 10:43

    There are potential barriers to entry; the point is just that the Singapore government chooses not to use them to grant minor interest groups rent. It does wave its powers around if it feels its own overarching economy-related political goals are being threatened, like the SIA strike incident.

    Most countries have many influential interest groups; Singapore has exactly one, and it is the government. It doesn’t want or need to act under the table to extract rent; for this reason Singapore is transparent. Investors in Singapore know what they’re getting right off the bat, and Singapore extracts relatively little tax to boot.

    This isn’t some banana republic where the people in government try to grab whatever they can before they get run out of power. Singapore doesn’t pretend it isn’t authoritarian. Let’s say you’re the government of Singapore, and you want to capture more rent. Why would you bother with demanding bribes and hidden fees, or protecting some business that you own on the side? Just announce in parliament “okay, we’re gonna raise the tax rate, and the amount we pay our MPs and civil servants”. Sometimes the most efficient way to extract rent isn’t via corruption. And sometimes a state-linked enterprise really has the state as an investor envisioning an opportunity and expecting dividends rather than a means of being beholden to some labor group.

    The point here is that Singapore did invoke barriers and abuses – albeit legitimate abuses, because all it was doing was accusing local people of being communists, eh? – when it had other domestic potential power competitors to deal with, then ‘privatized’ once it had removed them from the political sphere. But the powers still exist and are exercised.

    Are Singapore’s state-linked businesses efficient? Yes, they are. Is this because of the wonderful untrammeled hand of competition? No, not really. They were efficient before they faced competition. The state just had and has an explicit interest in their efficiency, and the state continues to poke at the system to ensure it remains so, up to and including (still) suppressing collective bargaining outside permitted channels or breaking up and merging businesses to see what might work or designing whole new investment opportunities from scratch based on what it predicts will work.

  26. Gravatar of scott sumner scott sumner
    5. June 2010 at 09:41

    David, I don’t know enough to argue with you on this–you may be right. I found your comment about them being a monopoly rent seeker kind of interesting. I think someone once did an article talking about how if you monopolized corruption you’d get a much more efficient outcome. Does that sound right?

  27. Gravatar of david david
    5. June 2010 at 14:12

    Yeah, I’m not the source of the idea – Mancur Olson came up with the model based on a stationary bandit versus a roving bandit; a stationary bandit with a longer time-discounting period faces a greater incentive to reinvest toward economic growth because he expects to be able to demand a share of the pie in the future. This was originally intended to explain moves toward settlement and civilization, but it extends neatly towards explaining how dictators behave.

    Perhaps not coincidentally, the highest paid political executive in the world is Singapore’s PM. The second is Hong Kong’s Chief Executive. Singapore has the previously described strategy of carefully destroying competing interest groups, thus ensuring that the government remains in control. Hong Kong just takes the direct approach of putting industry representatives into legislature (only half of Hong Kong’s LegCo is geographically elected; the other half are elected by business groups). Either method creates a group of cross-industry business shareholders and managers with legitimized control and a stake in wider economic growth instead of individual business rent-seeking.

    Both Singapore and Hong Kong also pursue policies of paying civil servants rates roughly based on what private-sector managers of organizations of similar size are paid.

    All this does fit the monopolized-corruption story neatly. Albeit that it isn’t corruption if it’s transparent, I guess.

  28. Gravatar of ssumner ssumner
    6. June 2010 at 12:05

    David, Thanks, that is quite interesting information.

  29. Gravatar of cuibono cuibono
    6. June 2010 at 15:05

    Singapore, like China, gets discussed a lot by Western intellectuals who selectively see what they want and use it as a poster boy for their desired narrative (free market success stories, ‘liberal autocracies’ etc), while the real story (as David’s explanations demonstrate) is usually a lot more complex and nuanced. The Singapore regime, being mostly educated in Oxbridge and Ivy schools, is also excellent at telling Western audiences what they want to hear. China’s regime is getting a lot better at this too.

  30. Gravatar of cuibono cuibono
    6. June 2010 at 15:15

    To get a less biased view of Singapore, I would suggest having a conversation with a thoughtful, well-informed Singapore national who is i) not in the pay of the regime (such as a “government scholar”) and ii) does not belong to the ruling class (or at least is at the periphery). Unfortunately, for various reasons this group is almost a null set, but a few exist. The same principle should apply to learning about China.

  31. Gravatar of scott sumner scott sumner
    7. June 2010 at 08:29

    cuibono, I have had many conversations with Chinese people, but find that they often don’t have the best perspective on the economic picture. In the 1990s I was much more optimistic about Chinese growth than they were, and I was correct in retrospect.

    In another post I also criticized simplistic views of countries. China is neither a free market paradise, nor an example of successful central planning.

    I like certain things about Singapore, but not others.

  32. Gravatar of – Development of China's Economics – Free Essays on Economics - Development of China's Economics - Free Essays on Economics
    13. February 2022 at 01:14

    […] 2010. The money illusion. [Online] Available at: [Accessed […]

Leave a Reply