Book Review — How Asia Works (Part 1)

Rice growing in Chihshang, eastern Taiwan (Financial Times)

How Asia Works by Joe Studwell is an economic history of several successful Asian countries and attempts to explain how their economies grew so quickly, focused mainly on China, Taiwan, Japan, and Korea. The book is organized by policy area rather than country, with three main sections focused on land reform, industrial policy, and finance.

One organizing principle of the book is nuanced skepticism of neoliberalism and the Washington Consensus. The first two sections defend coercive redistribution of land and (well-designed) tariffs, which are both policies that are at odds with a pro-free market stance. Nonetheless, even as a generally pro-free market person, I found this book pretty persuasive. Studwell is generally fair and empirically-minded rather than polemical. One meta-lesson from the book is that policy paradigms (such as neoliberalism) shouldn’t be totalizing. For example, even if you think market liberalization is generally good, there are probably specific instances where the best policy contradicts or at least is in tension with your general policy worldview.

Land Reform

Land reform refers to redistributing farmland from landlords to their tenants, with the aim of increasing the number of farmers who own their own farms. Studwell identifies several Asian countries (China, Taiwan, Japan, Korea) who did aggressive land reform, and credits these reforms with large increases in agricultural productivity and economic growth.

The story for how land reform improves productivity is that farmers who own their land have a greater incentive to make investments and use techniques that improve crop yields. Studwell argued that landlords generally tried to make money by extracting rent (in both senses) from farmers and charging them high interest rates rather than improving yields.

Studwell notes that small-scale, garden-style farming is the most productive on a per-land basis because there are many labor-intensive techniques that farmers can use to squeeze more yields out of a given amount of land. Since they are labor-intensive, they do not make sense for developed or even middle-income countries, but these techniques are a good way to improve agricultural yields in poor countries where wages are still low and the majority of the population works in agriculture. Studwell claims that land reform encourages or enables farmers to use these better techniques, thus improving growth.

China, Taiwan, Japan, Korea all implemented land reform after the end of WWII. In China, land reform was tied to communist ideology — the communists led by Mao Zedong were aligned with the common farmer over landlords, and seized the land (often violently) to redistribute among the farmers. Unfortunately, in the 1960s the communists then undermined land reform by replacing small farms with large communal farms, which removed the incentive for farmers to do garden-style farming. Agricultural yields then plummeted and millions of people died from the resulting famines. Meanwhile, in Japan, Taiwan, and Korea, land reform was an attempt to head off communism by getting farmers invested in the existing system and thus less tempted to join a potential revolution. Pressure from the US was instrumental: After WWII, the US and other allies played a heavy role in establishing the new Japanese government, and land reform was one policy reform pushed by the US due to the US’s desire to prevent communism from spreading in Asia.

Overall, this period was associated with large increases in crop yields of around 50% in the land reform countries. This led to a large increase in GDP and human well-being and set the stage for further economic gains from manufacturing and industrialization. Studwell also discusses land reform attempts in the Philippines and Indonesia, which were much less successful at redistributing land than in China, Taiwan, Japan, or Korea. Crop yields failed to increase in these countries by nearly as much, providing something of a natural experiment that suggests that land reform did cause the yield increases observed in the countries that implemented it.

Why does land reform boost productivity?

Studwell’s case for land reform partially depends on misaligned incentives in tenant farming. Landlords don’t just take a large share of the pie, they shrink the pie compared to if their tenants owned the land themselves by failing to improve crop yields. (Obviously, there’s also an equity case for land reform independent of crop productivity). This seemed a little bit counterintuitive to me at first — why can’t landlords increase the crop yields of their land in a way that leads to higher profits for them?

Unfortunately, Studwell doesn’t go into too much detail about why land reform works, which I think is the biggest weakness of this section. His explanations are pretty brief — for example, in the introduction to this part of the book, he says the following:

Landlords could make the investments to increase yields, but they make money more easily by exacting the highest possible rents and by usury, which adds to their land holdings when debts cannot be paid and they take over plots that have been pledged as colatteral. A situation arises where ‘the market’ fails to maximise output.

This is about as detailed as this part of the book gets as far as analyzing incentives. He spends most of the time talking about the specific case studies, discussing the political history leading to land reform and the results of land reform. But given that this is a book about policy, I wish he spent more time convincing me that the agricultural productivity gains can actually be attributed to land reform and that land reform is uniquely key to achieving those gains.

I think this is a result of the nature of tenancy arrangements and liquidity constraints. Landlords don’t capture all of the gains of their tenants’ productivity, since their money comes from the rent they charge, not directly from the tenants’ crop yields. I think that large productivity increases from land reform are due to tenant farmers’ liquidity constraints — they have narrow margins that they can use to invest in their farms. Landless farmers in poor countries are barely above subsistence level at best. If their rent is equal to 20% of their crop yields, they have so little money left that they cannot afford to make useful investments. Maybe if rent were cut to 10% of crop yields, that would give farmers the liquidity to make those investments and improve their crop yields. But Studwell claims that land reform only increases yields by 50% overall, so the landlord loses 25% of their rental income if they do this. (I think in reality landlords usually charged a flat rent, not a percentage of the tenant’s income. But you can think of landlords charging a flat amount that happens to be equal to 20% or 10% of crop yields).

If liquidity is the limiting factor to improving yields, then land reform isn’t about land, it’s just about redistribution in general and you could achieve the same goal by giving farmers cash. So it seems plausible that the benefits of land reform are not about land ownership per se but just a general case of economic inequality preventing the poor from making fruitful investments in themselves and their businesses (which you can see by, for example, the high returns on investment from unconditional cash transfer programs like GiveDirectly). In particular, this has implications for land reform implementation. Some land reform programs included bans or restrictions on the newly-landowning farmers selling their land and becoming tenants again. These provisions probably were motivated by something like preventing former landowners from coercively or predatorily reclaiming their land, but absent those considerations, it’s not clear that giving farmers the fair market value of the land in cash is such a bad idea.

Additionally, I’d be interested in knowing if there are alternative tenancy agreements that can solve or mitigate this incentive problem, and if they don’t exist, why those alternatives don’t work. For example, why can’t landlords run their farms as a business and employ farmers instead of charging farmers rent? That’s how farms (and most businesses) in developed economies work, as far as I know. The vast majority of people are not self-employed business owners, but instead get paid by the owners of capital to use that capital to produce things. And the capital owners generally seem well-incentivized make investments to improve their capital and increase their employees’ productivity. If such an arrangement is feasible, the landlord makes more money as crop yields increase. The land owners are incentivized to maximize yields because farmers earn wages rather than growing their own crops and paying rent, so the landlord captures a greater share of marginal productivity gains from investment. This is obviously worse than land reform from an equality standpoint (and accordingly I don’t actually endorse this as an alternative to land reform even if it worked), but the gains to productivity perhaps could still happen, the country’s GDP would still increase, and food prices would still fall, allowing a greater portion of the population to urbanize and continue the cycle of economic growth.

There must be a good reason this didn’t happen — maybe because the management overhead of making sure the farmer employees are working productively despite not getting to keep the crops they grow would be too expensive. In fact, this seems somewhat similar to the (very unsuccessful) collective farms in Maoist China, except with the existence of a profit motive for the person who owns the collective. I expect this would outperform the Chinese collectives, but maybe the labor-intensive “gardening” approach to agriculture just isn’t suitable for an employee-employer relationship.