Saturday, October 24, 2015

How China got rich

I do not understand this Tyler Cowen post about Chinese catch-up growth:
It seems obvious to many people that Chinese growth is Solow-like catch-up growth, as the country was applying already-introduced technologies to its development. 
But how many other economies have grown at about ten percent for so long?  Was there not a secret ingredient added to the mix? 
Increasing returns to scale?...Something about Communist Party governance which enabled the corruption to be channeled productively into building more infrastructure rather than holding up progress?  Tiger Mom parenting combined with a relatively meritocratic exam system?... 
More radically, is there some “natural,” culture-neutral rate at which innovations trickle down from the world leaders to the poorer countries?  The diversity of growth rates would seem to indicate not.  Is each country then not innovating — with varying degrees of success — by building its culture-specific net for catching and transmitting global innovations throughout the nation?
I don't know about Tiger Mom parenting, effectively channeled corruption, or "culture-specific nets", whatever those are. I do know, however, that Solow catch-up growth is not about TFP increase or the diffusion of innovation!

There are basically two kinds of catch-up growth. The first is where you cheaply copy innovations from countries at the technological frontier. That's what Tyler is talking about. But the second, Solow catch-up growth, is far simpler - it's just about capital accumulation.

Countries don't start out with capital - they have to build it. So they save some of their production each year and use it to build productive capital (or trade it to other countries for capital). They do this until they have so much capital that they hit the point of diminishing returns, and depreciation starts to get so costly that it doesn't make sense to build any more capital (per worker).

Really simple. Save money, build capital. And that's exactly what China has been doing. Here's a graph from a Federal Reserve Board discussion paper from 2013:

A very smooth exponential increase in the capital-to-labor ratio.

Now of course, in the Solow model, that red line should flatten out and turn into an S-curve eventually, as the K/L ratio reaches the steady state. That day of flattening will be delayed by productivity improvements - i.e., by the other kind of catch-up growth (plus any original innovation China does along the way).

But a lot of Chinese catch-up growth - about two-thirds, according to a standard decomposition - looks just like the classic save-money-and-build-capital thing. China has many more roads, many more railroads, many more cars, many more trucks, many more buildings, many more machine tools, and many more computers per person than it did in 1980 or 2000.

We don't need Tiger moms, culture-specific nets, or increasing returns to scale to explain that.

In fact, though China's growth looks more productivity-driven than some other Asian economies, It is probably less productivity-driven than Japan's catch-up was. If you want to look for Asian countries with cultural secret sauces, I'd start with Japan instead of China. Remember, when we talk about China, we're talking about a country whose per capita GDP is still only about a quarter of the countries at the frontier. And it's not yet clear how rich China will get before its growth slows dramatically.

The question of how technology diffusion happens is, of course, one of the great unsolved and under-studied mysteries of economics.

Update: This paper seems relevant.


  1. And the, diminishing, benefits of putting resources to work more effectively and increasing inputs into the system (as Krugman pointed out in a slightly different context:

  2. But how many other economies have grown at about ten percent for so long? Was there not a secret ingredient added to the mix?

    China was coming off a very low base where the economy had effectively been artificially suppressed by really really bad policy. Stop doing really stupid stuff. Invite in some ethnic Chinese businesses from Taiwan and Hong Kong to transfer semi-modern techniques. Piggy back on Chinese speaking banking in Taiwan and Hong Kong. That seems to be the secret sauce.

    The Chinese progress in the forty years from 1975 to 2015 is probably not as remarkable as the Japanese progress in the forty years from 1865 to 1905.

    1. Anonymous10:35 PM

      While transforming Japan from a feudal farmer society to an industrial one was impressive, the actual rates of growth of the Japanese economy in the Meiji period were lower than in the post war period.

  3. Noah, here we have an answer for you and Tyler
    for more detail can also look to academic version

    1. Thanks!! Do you have a link to the academic version??

    2. Anonymous10:56 PM

      It's the note in the essay, NBER WP21397, from July, (I'm not Unknown, from above, fwiw.)

  4. I'm not sure where to start on this. You say about two-thirds of China's growth is standard catch-up -- that's close to what I suggest in my post! (I say we might think of two to three percentage points as something else.) Somehow you're pretending I got this wrong. On the rest, savings doesn't guarantee capital value. It is well known for instance that China's exports are especially technologically sophisticated for the country's per capita income (see Rodrik). How did that happen? We are back to considering IRS, education, family, and other factors, exactly as I suggest. Plenty of countries accumulate capital but don't move so well to the frontier as China did, that still requires explanation.
    And it is a perfectly valid and indeed standard point to worry whether capital accumulation vs. embodied innovation is always so well-defined a distinction, as I do at the end of my post. Simply repeating the distinction doesn't clear the matter up.
    I don't doubt that the Japanese case is worth considering also.
    Overall you're reacting too quickly and not thinking through enough what I actually wrote.

    1. You say about two-thirds of China's growth is standard catch-up -- that's close to what I suggest in my post! (I say we might think of two to three percentage points as something else.)

      I'm not trying to be adversarial or pull a "gotcha" here, just pointing out that good ol' saving and investment were a huge part of the story.

      You said that two or three percentage points might come from "innovation and increasing returns to scale in some manner". But that's very different from absorption of foreign tech. You seem to suggest that much of the rest is absorption of foreign tech, which would mean that a large fraction of Chinese growth - much greater than 1/3 - was due to productivity increases instead of simply to capital accumulation.

      That was my reading of it.

      It is well known for instance that China's exports are especially technologically sophisticated for the country's per capita income (see Rodrik). How did that happen?

      Yeah, that's the Solow residual. If we look at growth accounting, we see that increases in the residual are not always a big part of the catch-up stories of places like Singapore, for instance. In China it appears to have been a bigger part, but not as big as in places like Japan. That was my point.

      And it is a perfectly valid and indeed standard point to worry whether capital accumulation vs. embodied innovation is always so well-defined a distinction, as I do at the end of my post.

      Agree that it's not a well-understood distinction. Couldn't find it in your post.

      Overall you're reacting too quickly and not thinking through enough what I actually wrote.

      Perhaps. As I saw it, you were musing about cultural determinants of technology diffusion and innovation, and I was trying to point out that simple capital accumulation was being given short shrift, since I saw no reference at all to it in the post...

    2. Anonymous8:53 AM

      "...It is well known for instance that China's exports are especially technologically sophisticated for the country's per capita income (see Rodrik). How did that happen?..."

      [I just cannot figure what Cowen is questioning. Is it that there is too little production of technology goods under US patents being produced in China by Foxconn or that there is too much or just that Foxconn underpays Chinese workers? Then one could ask the same question about GE as Foxconn but Foxconn produces exclusively "technologically sophisticated" goods and I am not sure if GE washers and dryers qualify. Multinational firms were exporting capital to China in their offshoring initiatives. So, domestic capital accumulation was only part of the capital formation for offshoring of MNCs production. ]

    3. Anonymous7:37 AM

      Ignorant question, but I'm confused over the words 'money,' 'savings' and 'capital.' I don't know the real savings rate, in terms of cash money, in China. I do know that they invested heavily in infrastructure but they don't book said expenditure as 'deficit' spending. From that point of view, capital can be accumulated without an accumulation of 'savings.' Anyway, I'm confused because I don't believe 'savings' of cash money has much to do with the accumulation of 'capital.' Maybe this is a distinction without a difference, but not from my perspective.

  5. China's exports are especially technologically sophisticated for the country's per capita income

    Having Taiwanese manufacturers come to China to employ Chinese peasants to assemble telephones made to a foreign design, using chips designed and made outside China and running foreign software, does not make China a high tech exporter.

    The Chinese are exporting hi tech routers but they are using designs stolen from Nortel.

  6. Check out St. Louis Fed Reserve Yi Wen's largely Solow take on China's growth: The Making of an Economic Superpower - Unlocking China's Secret of Rapid Industrialization.

  7. Anonymous8:09 PM

    China's growth not so special :

    China has hit their debt limit like Japan did in 1990 , at a lower level of development , so in time I think they'll be perceived as decidedly unspecial.

    The trick to "special" catch-up development is to get as much done as possible before the infestation of financial parasites overwhelms you. China got a nice finance boost which sped things up over the last couple decades , but now those debilitating parasites are embedded.

    Sad , really.

    1. Anonymous8:12 PM

      Forgot to add this :

  8. There's not one type of growth or two - there are lots. "Capital" is an easy word to say, hard to define, even harder to measure. The Mao period laid a lot of foundations (internal peace, a reasonable degree of governance, mass health and education, lots of basic infrastructure). The current regime were smart enough to realise that this plus geopolitical and world economic changes gave them an opening they could exploit. If, for instance, China had remained in semi-quarantine (as the Soviet Union was for most of its life), then the policy options would have been very different.

  9. Anonymous10:33 PM

    Cowen gets his racial ideas from the same spot that his frequent commentators do. The idea that races and culture are a deterministic model of success is just part of conservative thought. Of course the great irony is that a 1920s or 50s would have written about how China's economic retardation was caused by its Confucian culture that is too conservative.

  10. This seems mostly a misunderstanding of a somewhat confusing Tyler post. But without disputing your points re 1/4 of the frontier or Japan, there's obviously something to the view that 1980 China had a store of submerged cultural economic proficiencies. Likewise there's a reason why Chinese people are on average wealthier than others in SE Asia. To start with, on average they work harder and save more.

  11. Anonymous8:07 AM

    I don't know about mainstream economists but historians and development economists have long asked why have East Asian economies moved to the frontier before African and South American ones did. Cultural factors are likely to play some role. But so are other things: eg the effects of European colonialisation. (I stress European; Japanese colonisation of East Asia was very different in nature to that of the Europeans.) There is also agreement that important things happened in both China and Japan when they were closed economies, and they way they transited into the global one which are important in understanding their successful industrialisation.

    This is an area better understood by historians and area specialists - you really need good linguistic and other skills. Look at this and consult them before you think about models or neo-classical theory - that is the least important thing to get what happened.

  12. Did China decide that if China builds it, they will come; or did they come, so China decided to built it. I'm not an economist (I received an undergraduate degree in economics in 1973, when the teaching of developmental economics was devoted mostly to theory (comparative advantage, etc.), and those interested in the subject were economists like JKG), so I don't default to models for an explanation of all economic phenomenon, including the extraordinary story of China. An historian might suggest a connection to the cultural revolution: if not for the CR, China may not have changed course so radically and quickly. Or does my allusion (if they built it) an important factor? My comment at MR was that it was a perfect storm: China's capital investment coincided with western firm's (i.e., U.S.) decision to betray U.S. labor and shift production overseas and the anti-tax movement in the U.S. which allowed U.S. firms to avoid U.S. tax with schemes with little or no substance. Anyway, I'm a regular reader of MR and as everyone else who is a regular reader knows the blog posts have an ideological slant. That's not a criticism but an observation; indeed, an ideological slant can be right.

  13. Also I don"t understand the distinction you're trying to make between a behind country accumulating capital and importing technology. The two are always one and the same and I don't think there was any intent by Solow to distinguish.

    1. It's the difference between knowing better road construction techniques and actually building a bunch of roads using those techniques.

    2. Which is exactly zero practical difference, no?

      Putting aside China, which was obviously FDI-led catch-up, is there any practical situation where a low-capital country could build capital without importing technology? I really don't think "Solow growth" makes any practical sense if it excludes importing technology.

  14. The usual way an undeveloped nation exploits the gradient between its internal conditions and the world market is by selling its labor and resources, but the profits from these are usually spent on symbolic goods. The money goes for palaces, decorative armies, art collections, and exotic financial instruments. It might show up in the national accounts as capital accumulation, and it will produce income for the elite who commerce in it, but it will not increase the productive capacity of the nation.

    Historians often note that Rome built an empire, but China built a nation. China's leadership, for all its autocracy and corruption, is nationalistic. That means that the government can plow money into infrastructure and business leaders are forced to plow money into capital plant. A businessman who starts impinging on government prerogatives will find himself slapped down, broke, in exile or worse.

    In the west, we are no longer nationalists, so we have an ideology of minimal taxation and government spending that cripples our economic development. We have a wealthy elite that trades symbolic goods which have been subject to an ongoing inflation. Look at the prices of market center apartments, fine art paintings or corporate ownership. (Twenty years ago I put my money where my mouth is on this and I've profited handsomely. I do see this inflation slowing, but not ending any time soon.)

    There is nothing racial about China's economic success. The sheer size of China and the large gradient meant that it could maintain a high rate of growth for decades. The gradient has been decreasing, so one would expect the rate of growth to decrease without some form of internally driven stimulus. China does not have an overseas empire or large chunks of land to divide, traditional stimulus techniques. However, it does print its own money. It could take the obvious next step to maintain its growth levels, but that has serious political risks.

  15. I think capital accumulation via financial repression is underrated as a development tool. Forcing up savings and investment. Obviously if the capital is just completely wasted it's a disaster, but even the USSR managed to find things to invest in until the 1970s. Singapore, with some slightly nepotistic government corporations driving investment, has reach a per capita GDP of $75k and is still going.

  16. Anonymous8:50 PM

    I think this depends upon what you call capital. Is financial capital the only important capital?

    Clearly foreign capitalists provide a lot more an financial capital. They provide access to markets, worker knowledge and business knowledge. None of these are easily shared.

    It isn't just financial capital that builds foreign economies. In fact, financial capital is useless without the other connections to markets, both global and domestic.

    Capital is not just about money. The main deficiency in most emerging economies is the proper ratio of intelligent capitalists vs. financial capital.

    Our current model assumes that our money is what powers foreign economies. Not true. The main power is in our knowledge and our market power to other inputs.

  17. Anonymous9:08 PM

    In other words, it isn't just about saving money. It is about developing internal sustainability given the need for external inputs provided by trade.

    Domestic capital markets cannot be sustainably created without a gift of knowledge regarding current technology and external market requirements.

    China has almost all US IP in possession. We should have given most of it to them without requiring hacking.

    Most emerging economies can't compete with the US because of IP. China did it because they stole it. Give it to them.

    But in addition, the US has enormous buying power in global markets that many emerging economies don't have. Oil for one. Food is another.

    If we allow companies to profit from foreign economies we should also provide a transfer of other types of capital that allows them to develop on their own to fit in with our global economy. They shouldn't have to depend upon US companies.

    This goes against everything taught in business school, nevertheless it is the right way to do global business. It is the only ethical way to do global business given our clear advantage in technology and business knowledge.

  18. Anonymous9:53 PM

    So a practical policy would be something like this:

    If your company utilizes labor in another country, you must share your IP with that country.

    Obviously there's a lot of room for military IP held in the US if US labor is used to produce it.

    Somebody give me a reason this is not a good policy?

  19. "Increasing returns to scale?...Something about Communist Party governance which enabled the corruption to be channeled productively into building more infrastructure rather than holding up progress? Tiger Mom parenting combined with a relatively meritocratic exam system?... " So how would this applied to countries like Thailand, Vietnam, Indonesia the Filipinos and Malaysia. Malaysia in particular. Very different than China in terms of culture, ethnicity, governance, scale, natural resources, . Yet it has transformed itself into a modern society very quickly. Was the pathway to grow for these countries similar to that of China?

  20. Oops, forgot the notify me.

  21. In most growing, developing economies, far and away the strongest driver of growth is the urbanization rate. China has been able to drive a faster urbanization rate than any other country in the past, largely due to a command economy structure I/T capabilities to rapidly organize work opportunities for rural immigrants. China is approaching the point where further increases in urbanization will be harder, so of course they are slowing down.

    1. I have looked at a chart on the level of urbanization for a number of Asian countries for the year 2014. Philippines 49%; Thailand 49%; Indonesia 53%; China 54%; Malaysia 74%. Malaysia stand out as the most urban society. My question is to what extent has China's urban growth rate be faster that these other countries.

    2. This comment has been removed by the author.

    3. This comment has been removed by the author.

  22. Anonymous11:20 PM

    You can save and invest and get zero returns. You can also get tremendous returns. It is not just about savings, but about how those savings are deployed.