I share Sam Roggeveen's enthusiasm for the iconoclastic approach of Joe Studwell's How Asia Works (his previous book on Asian Godfathers was a great read too). I also share Studwell's scepticism about the 'magic of the market', his views on the IMF, and his admiration for the achievements of the South Koreans.

But I'm unconvinced by Studwell's three-step development prescription, not because it is intrinsically wrong but because it is too hard to implement successfully.

The Koreans might have done so, but the strategy requires a level of sustained administrative competence, single-minded toughness and luck which are rare. Just as important, there are alternative development strategies, less demanding of skilled policy-making and administrative competence. The growth outcome won't match Korea's, but will be more feasible for countries like Indonesia (which Studwell sees as a development failure).

Let's go through the three elements of the Studwell strategy. The first stage requires land reform and a boost to agricultural productivity.

It's an old and sensible idea that agriculture has to provide the investable surplus which will propel the rest of the economy along the path of development. Fifty years ago, Clifford Geertz (Agricultural Involution) despaired about Indonesia's failure to follow the example of Japan, which shifted surplus agricultural labour into factory work to create a modern urban/manufacturing sector. This failure would lead the excess population to atrophy, farming progressively more Lilliputian plots.

But things turned out better. With the average size of farms on Java around half a hectare, the opportunity for land reform couldn't play the key role that Studwell advocates. But Soeharto, with his roots in agriculture, gave rice production high priority (extension services, high-yield seeds, fertilizer, pesticides and attractive terms-of-trade between agriculture and urban consumers via an active price stabilisation authority). Not very free-market, but big yield increases and self-sufficiency were speedily achieved.

What about a vigorous industry policy, the second Studwell requirement? Despite inheriting the usual disaster story of failed prestige projects from Sukarno, Soeharto was ready to have a go at 'picking winners'.

Cement, fertilizer, textiles, paper production, food processing and petroleum refining all fitted Indonesia's comparative advantage and made sense. Others were less defensible: Krakatau Steel, Tommy Soeharto's national car and Ibnu Sutowo's tankers. Habibie's IPTN aeroplane fits the Studwell strategy and might have succeeded if it hadn't been stopped by the Asian crisis: ex-aeronautical engineer Habibie was well-qualified to lead this project, plane construction is quite labour-intensive (all those rivets) and the Indonesian archipelago needs lots of them (one airline recently ordered several hundred in one hit).

Whether IPTN would have succeeded is not the issue here: the point is that Indonesia, for better or worse, did try the sort of hot-house industrialisation Studwell advocates, and the IMF wasn't able to stop this, at least until the 1997 crisis. Planning retained a central role, just as Studwell wants, and state-owned enterprises did the government's bidding. Where Indonesia had comparative advantage, this often worked out well, and where the industry didn't suit Indonesia's attributes, generally it was a failure.

Indonesia's development experience doesn't fit the Studwell formula. Java's rice production has done well without relying on his key element of land reform, and industry policy based on domestic entrepreneurship has been tried without much success.

Governments attempting to steer the process of development need effective administrative capacity; in a follow-up post, I'll expand on the idea that market failure is common enough, but so too is government failure.

Photo by Flickr user macloo.