The Problem With Comparative Advantage And Justin Yifu Lin's CAF Strategy of Economic Development

My friend recently posted on my Facebook an article which quoted Bill Gates saying that people don’t realise how many jobs are going to be automated in the near future. The implication being that my previous essay ‘Technology growth will lead to mass unemployment’ was correct. However, I’ve been spending a lot of time thinking about the idea of ‘what would convince me that I’m wrong?’ as a way to stay open-minded and avoiding being emotionally attached to my ideas, something that Michael Burry, of Michael Lewis’ ‘The Big Short’ impressed on me. Apparently he hated writing quarterly letters to investors for exactly this reason because in defending his ideas he would become attached to them and may not be able to leave them even if they were wrong. One unexpected fallout of this way of thinking is that an expert, someone you respect or even the person paying your cheques saying your wrong (or right!) does not qualify as a legitimate reason for changing your mind. Of course, if they present good arguments then that’s fine but just trusting in another persons intelligence or superior understanding is unfortunately not allowed. So although, I have to admit the absurdity of me critiquing an economist of Justin Yifu Lin’s reputation I nonetheless I have to persist until someone can convince me that my arguments are wrong. Having said that, to test my own understanding of his theory of Comparative Advantage Following (CAF) Economic Development the first part of this essay will simply be an attempt to relate an unadultered narrative of his theories and ideas. This may also serve as a way to get you up to speed on his ideas in case you are not familiar with them. The second part of the essay will be when I share some of my own ideas and critiques.


So here is the problem. You’re a poor, largely agricultural developing economy and you want to grow. Fast. What should you do? Well the obvious thing would be to look and see what your most successful neighbours have that you don’t. What you would find is they tend to have well-developed capital-intensive manufacturing industries. Therefore much of economic theory is built upon the idea that poor countries need to build up their industrial base. However, this won’t just happen on its own and so requires government policy to incentivise capital accumulation. Justin Yifu Lin argues that this can have the opposite of the desired effect crippling the poor country for decades just like we saw with China in the 1950s and 60s. In Yifu Lin’s view prematurely developing capital intensive manufacturing industries is like trying to walk before you can crawl. Instead countries should focus on industries in which they have a comparative advantage and can develop a surplus in. These industries he describes as viable because they can compete and survive without government support. Over time the developing country can invest its economic surplus in capital and gradually become more capital intensive. As the economy changes so will the industries in which the country has a comparative advantage in and thus you will have economic development smoothly transitioning to increasingly capital intensive industries. In ‘Part One – Justin Yifu Lin’s CAF Development Strategy’ I shall first outline more specifically the pitfalls of artificially developing a manufacturing base or what Justin Yifu Lin calls ‘Comparative Advantage Defying’ theory of economic development with specific reference to China’s economic story. Then, I will describe Yifu Lin’s ‘Comparative Advantage Following’ economic theory .


As previously described poor countries that want to quickly develop an industrial base need to intervene in the market because otherwise heavy industry capital investment will not naturally occur. This is for three reasons in particular.

  1. Capital intensive heavy industries require long construction periods and this in turns means large interest rate costs. To counteract this the government needs to artificially lower the interest rate.
  2. Key technology and capital needs to be imported but as we’re talking about poor countries with limited exports they do not have enough FX to buy the capital required. Therefore governments need to intervene in the exchange rates to make imports cheaper.
  3. Huge barriers to entry because of large initial capital outlays make investing in capital intensive heavy industry very expensive. For agragrian economies with limited surplus this is a particularly serious problem. Therefore to help incentivize capital accumulation by increasing expected profits need to grant monopoly status to firms as well as lower productive input costs namely lowering wages.

Such policies however lead to many unintended consequences which ultimately make the efforts counter-productive. Taking China as an example although there was remarkable capital accumulation: 24.2% of GDP in the 1st Five Year Plan and 30.8% in the 2nd economic growth in the period of 1952 to 1981 was a pitiful 0.5% a year at best. This is because although China had developed capital intensive manufacturing industries they were terribly inefficient and uncompetitive, only kept alive by government support. Support that was funded by the already poor Chinese farmers.

The reason why this happens gets a little messy but the key thing to keep in mind is that the government is trying to push the economy into industries its not naturally suited for, the market however pushes back which leads to lots of problems. Specifically

  1. The aforementioned strategy of lowering interest rates also lowers savings and as the level of savings help determine the investment level it results in a short-run shortage of capital.
  2. In manipulating the exchange rate although imports become cheaper exports simultaneously become more expensive which lowers foreign exchange reserves, which means less capital.
  3. For a country that is already poor paying for the huge capital costs becomes prohibitively expensive. Essentially you end up taxing poor farmers to subsidize inefficient manufacturing. In particular, the policy of lowering wages in turn reqruires lowering the prices of daily necessities which leads to shortages. In the China story the government therefore had to intervene in the agricultural products markets with disastrous effect. In fact the China’s effort to control the agricultural markets through the Peoples’ Commune suffered a 15% reduction in grain production both in 1959 and 1960 and the deaths of more than 30 million people.

As 1. changing the exchange and 2. the interest rates mean that not only are heavy industry capital costs cheaper but so are light industry and agricultural costs there is a danger that the already limited capital is diverted away from heavy industries to the less expensive light and agricultural industries. Therefore the government in pursuit of its policy of prioritizing heavy industries is forced to nationalize. But in doing so suffers from all the negative incentive effects of government control with the potential for corruption and poor management and ultimately inefficient uncompetitive manufacturing.


Instead of trying to artificially upgrade the industrial and technological structure Yifu Lin argues that countries should seek to upgrade the underlying endowment structure. With this successfully done the desried industrial and technological structure will naturally arise. As land and natural resources cannot be changed and labour growth differences between countries are minimal the key thing is to focus on capital acccumulation. This you will recognise is the same goal as the ‘CAD’ strategies. What Yifu Lin disagrees with though is the not the aim but the method. In particular he argues the best way to accumulate capital is to increase the economic surplus or profit that is made in each period as well as increase the percentage of that surplus that is invested in capital.

At the heart of ‘CAF’ strategies is by definition the concept of comparative advantage. Ever since its conception in 1817 by David Ricardo comparative advantage has been the bedrock of international trade. What comparative advantage says is that even if one country, let’s say America, is more efficient at producing goods than another country, let’s say Russia there can still be gains from trade. If you imagine a two good world of grain and ipods. In this world America produces both grain and ipods cheaper than Russia. However America is excellent at producing ipods but only very good at producing grain, in other words its relatively better at producing ipods. Russia in turn although it’s worse than America at producing everything it is good at producing grain but bad at producing ipods thus it is relatively better at producing grain. Therefore if America focuses all its effort on producing ipods and Russia all its effort on producing grain world output increases because each countries’ factors of production have been utilized doing what they are relatively best at doing.

A developing country following ‘CAF’ strategy essentially is picking its battles. Rather than trying to go up against the developed countries in industries they have huge advantages in, by following comparative advantage developing countries can pick industries in which they don’t have to compete against developed countries. CAF strategy in contrast involves competing against developed countries companies in industries they are more efficient in. Of course developing countries can try and support their companies through subsidies etc but an inefficient company backed by a poor country is still going to lose against a developed countries efficient and more technologically advanced companies. Even though for a developing country, focusing on its comparative advantage industries, the spoils of victory might not be as great companies competing in these industries will be what Yifu Lin calls ‘viable.’ Which he defines as companies that are normally managed with no government support that can achieve a normal profit in an open, free and competitive markets. Then over time countries will shift from the labour intensive industries to the more capital intensive ones as companies invest their profits competing in their comparative advantage industries into accumulating capital.


There I think three primary problems with comparative advantage and therefore Yifu Lin’s ‘CAF’ strategy.


The first problem I think is that although through ‘CAF’ strategies countries can avoid competing with more developed countries in industries they are inherently inefficient at unfortunately there are still lots of poor countries whose comparative advantage is in labour intensive industries left to compete with. Countries in an effort to support their domestic industries may employ protectionist measures or risk having them out-competed. Even if a country’s countries are able to survive the competition between so many companies operating in the same industry inevitably erodes the profit margins. This is crucial because it is exactly this economic surplus from which future capital accumulation is supposed to arise. This competition not only occurs between poor developing countries but also developed countries because as the cost of copying technology is lower than the cost of innovating countries – at least to some extent – can converge. This is usually viewed as a good thing but it has the unwanted by product of resulting in economies with similar economic structures and therefore increased competition.


Yifu Lin assumes that through focusing on the industries in which countries have a comparative advantage countries can gradually shift the isocost line and move away from labour intensive industries towards more capital intensive industries. This picture however is too simplistic because capital cannot be viewed as a continuous line. The reality is that accumulating capital in car manufacturing doesn’t mean that you can then smoothly jump to accumulating capital in ship manufacture. They don’t transfer that easily. This is also true of expertise and human capital. In reality each industry is relatively discrete. This means that countries can get trapped in a specific set of industries which they are viable at but cannot switch to different and perhaps more capital intensive industries without huge investment. Which of course leads us to exactly the same problem we were hoping ‘CAF’ would solve; that of a poor country trying to transition to a more capital intensive industrial structure but not having the capital to do it. Yifu Lin argues that because China enjoyed little economic growth in the 20 years until 1978 therefore China’s plan of transitioning to heavy industry was a failure, despite the incredible capital accumulation numbers. I would argue that perhaps China just showed an incredible resolve for a poor country to make itself poorer in the short-run so that in the long-run it can fundamentally shift its industrial structure and build the foundation for future runaway economic growth. For countries not willing to take this Faustian bargain the future may be bleak as the country gets trapped specializing in labour intensive industries and other countries become more and more efficient in the capital intensive industries.


This last critique is perhaps the most controversial. To take the previous example of Russia and the United States even though America enjoys an absolute advantage over Russia in the production of both ipods and grain America specializes in its comparative advantage ipods and Russia in its comparative advantage grain. At the heart of comparative advantage though is the assumption that America doesn’t produce grain because its limited factors of production would be better spent producing ipods. However if America had sufficient factors of production to produce enough ipods and grain to satisfy world demand then the United States with its absolute advantage in both industries would out-compete Russia and Russian workers and capital would be left unemployed. It sounds of course implausible because surely the Russian factors of production would be employed doing something else but as we have seen in the Great Depression it is possible for economies to last long periods without fully utilizing all its factors of production. In fact it could be argued that most of economic theory is about jump-starting economies out of low employment equilibriums into high employment equilibriums. I think this critique is particularly significant because as I argued in another essay titled ‘Technology growth leads to mass unemployment’ I think there are compelling arguments to believe that unemployment of factors of production and in particular labour could soon be a widespread problem. Of course, this is against our economic intuition because our economic history is one of our economies and labour forces shifting from labour intensive agriculture to labour intensive manufacturing to labour intensive services. Thus although much of our economic growth has been fuelled by the automatization of the efforts of labour there have always been new industries to soak up the unemployed workers. I think now for the first time in human history that is potentially not going to be the case. In particular this oncoming robotization points to a fundamental decoupling of labour and capital such that in many manufacturing processes the concept of a diminishing return to capital because of limited labour to work that capital will become meaningless.