Via Noah Pinion, an article on Indonesia:

I’ve been writing a series of posts about economic development. The last 20 years have seen a marked acceleration in the rate at which poorer countries — not just China, but many countries — are catching up to richer ones. Among the success stories I’ve profiled so far are Bangladesh, the Dominican Republic, and Poland. But today I want to talk about the country that might impress me more than any of the others: Indonesia.

Indonesia’s growth has been solid, but not amazing. Compared to its Southeast Asian neighbors, it’s pretty middle-of-the pack in terms of both the growth rate — an increase of 160% since 1990 — and in terms of per capita GDP. At around $11,000 a person, Indonesia is solidly in the middle-income category. (Note: There was a typo here, now fixed!)

So why do I find Indonesia to be such an impressive growth story? Well, several reasons. But first of all, the deck was really stacked against Indonesia in a number of ways.

Difficult geography, difficult history

First of all, Indonesia is geographically huge and fragmented. Its land area is only about a fifth that of the U.S., but it stretches across a vast area.

This inherently makes it hard for people to move around the country. It also creates differences in the demand for public goods, which tends to paralyze governments.

An even bigger factor affecting the latter is the size and diversity of Indonesia’s population. With 274 million people, Indonesia is the fourth largest country on Earth (the U.S. is third at 330 million). And this enormous population is made up of an absolutely dizzying array of ethnic groups — 1340, by the official count.

A large population has its advantages and disadvantages (see here for a great list, with research links). But a large diverse population, in a post-colonial state, tends to be a recipe for dysfunction. As Alesina et al. (2006) showed, European empires often created “artificial” states by lumping a bunch of ethnic groups together within fairly arbitrary borders, and this practice tended to lead to a bunch of ethnic conflict that made it hard for these countries to grow (at least, for a while). Indonesia somewhat fits this bill — much of it used to be part of various Javanese empires (Java being the most populous island in the archipelago), but the modern state is a creation of Dutch colonialists.

As one might expect from this history, the country has often been riven by ethnic conflict, and there’s still some of this today. The most severe right now is the Papua Conflict, a low-level civil war between the government and a separatist movement.

Indonesia’s post-colonial history is violent and turbulent. In the 1960s the army, with support from the U.S. CIA, committed a mass slaughter of half a million suspected communists. In 1998 there was a huge series of riots against ethnic Chinese people, which ended up toppling the country’s dictator at the time.

Given its natural challenges and this bloody and chaotic history, it’s all the more remarkable that Indonesia is now one of the freest and most democratic in its region, with a higher Freedom House ranking than Malaysia, the Philippines, or Singapore. Since the fall of the dictator Suharto in 1998, the country has had four peaceful transitions of power (though one of these was an impeachment of a president who tried to dissolve the parliament). Minority rights have improved over time. Though the country is about 87% Muslim, and did have a bit of Islamist terrorism for a while, this has mostly faded in recent years, and the country is generally looked upon as a model for peaceful Islam.

Increasing stability and freedom is always impressive, but at a time when global authoritarianism and intolerance are gaining strength, it’s all the more so. In a world growing steadily darker, Indonesia has bucked the trend. And there’s some evidence that Indonesia’s move toward democracy has been good for its economic growth.

But what’s also impressive about Indonesia is that it has maintained long-term economic growth while switching its basic development model — and now may be about to switch models a second time.

Industrialization, deindustrialization, reindustrialization

The first thing I do when trying to figure out a country’s development model is to go to the Atlas of Economic Complexity and look at its exports — basically, a measure of what the country specializes in. Roughly speaking, there are two types of developing countries — natural resource exporters and industrial exporters. The former typically have all kinds of problems — political dysfunction, overvalued currency, vulnerability to resource price swings, and so on. This is known as the Resource Curse. The industrial exporters generally start out poorer, since they don’t have much to sell. But as long as they maintain the right policies they tend to experience steadier growth; eventually, they usually end up richer than the resource exporters.

When we look at Indonesia, it basically looks like a natural resource exporter:

The top exports involve a lot of chopping down trees and digging up rocks.

But what’s especially interesting about Indonesia is that it wasn’t always this way. Go back to 2003, and while fossil fuels still take the top spot, electronics and clothing are numbers 2 and 3. Logging and mining industries, the biggest in 2020, were way down the list in 2003.

Thus, over the past two decades, Indonesia seems to have undergone a process of deindustrialization. And when we look at manufacturing as a percentage of overall GDP, this is confirmed:

What happened here? Why did Indonesia go in for manufacturing and then turn back? Here is a good overview of the relevant history by Aswicahyono et al., and here is another by Puspitawati.

Industrialization started in the 1960s. Suharto, a military dictator, came to power after the upheavals and bloodshed of 1965-66. In the late 60s, like many other authoritarian modernizers in Asia, he attempted to establish political legitimacy by promoting rapid economic growth. Encouraged by a group of Berkeley-trained economists, Suharto invested in infrastructure and education, stabilized the macroeconomy, and promoted free trade and entrepreneurship.

At first Suharto’s policies mainly benefitted the agricultural sector, but by the late 70s, Indonesian manufacturing began to ramp up. Though Indonesia did maintain a significant role for state-owned enterprises — especially in natural resources and simple primary industries — foreign investment in manufacturing became significant. (Note that this is a challenge to the theories of Ha-Joon Chang, who sees manufacturing FDI as a liability.) By the 1980s, Indonesia was becoming a powerhouse in exactly the kind of labor-intensive light manufacturing industries that countries are supposed to do early on in their arc of industrialization — clothes, fabrics, electronics assembly, and so on. Remember America’s collective freakout in the early 1990s over Nike’s “sweatshop” factories? That was Indonesia!

Then everything came crashing to a halt in the Asian Financial Crisis of 1997. Indonesia was hit hard, and the resulting turmoil ended up throwing the dictator out of power. Indonesia’s democracy had begun, but its industrialization stalled. In particular, Aswicahyono et al. note that before the crisis, it was common for small manufacturing companies in Indonesia to become big companies, but this scaling-up process stopped after 1997-8:

Thus the crisis and its immediate aftermath appear to have marked a turning point in this process of firm mobility. Until the crisis, smaller firms continued to display the dynamism evident in the pre-crisis period. However, after the crisis, the pace of graduation slowed, and the small firm share in both series declined.

The authors blame financial constraints, which is plausible given how hard the crisis hit the country’s banking system. But starting in the early 2000s, there was another factor, which Aswicahyono et al. mention only briefly — the rise of China. If you were a multinational company who wanted to make shoes cheaply in 1991, you went to Indonesia; in 2003, you went to China. FDI into Indonesia recovered and even accelerated in the mid and late 2000s, but much was now flowing into resource-extractive industries. A third factor might have been what economists call Dutch Disease — rising commodity prices in the 2000s made it more lucrative for Indonesian entrepreneurs to sell commodities instead of manufactured products.

In other words, Indonesia went from making shoes and computers to chopping down trees and digging up rocks. Aswicahyono et al. sum up the post-crisis, post-democratization situation:

Agriculture returned to positive growth, reflecting Indonesia’s diverse natural resource advantages in the context of high commodity prices…Manufacturing is unique in that it is the only major sector to experience more than a halving in both its output growth and its output elasticity.

But now here’s a surprise. Despite this steady deindustrialization after the turn of the century, Indonesia kept growing at about the same rate as before! Here’s a long-term GDP series from the Angus Maddison database:

The Asian Financial Crisis is easy to see on this graph, but can you spot the collapse in manufacturing as a share of GDP and the shift back to natural resource industries? I sure can’t. Indonesia switched its whole development model while barely missing a beat. And it did this even though global commodity prices experienced wild ups and downs over that period.

This kind of smooth, even growth is not typical for a commodity exporter. Economic models would struggle to explain how huge shocks to the terms of trade would leave GDP effectively untouched. How can we explain it?

Well, one potential answer is urbanization. While Indonesians’ movement into cities did slow a little bit after the turn of the century, it kept going at a decent rate. Another possible answer is regional agglomeration — essentially every country in the region has experienced smooth growth over the last two decades, despite big differences in industries and development models:

Indonesia’s return to growth in the 2000s and 2010s might thus be a story of neoliberal success. Trade liberalization — which continued after Suharto’s fall, with prodding from the IMF — meant that Indonesia could always find something to sell to the countries around it, even if its comparative advantage shifted from manufacturing to logging and mining.

But in the long term, it’s still no good to have an economy based mainly on logging and mining, for several reasons. First, no country has ever reached the ranks of the developed nations that way, except for a few small petrostates. Second, eventually you start running out of easily accessible trees and rocks. And third, chopping down trees and digging up rocks is very hard on the environment, which tends to stoke popular anger.

So Indonesia needs to shift back toward manufacturing, and resume the industrialization that paused after 1997. And incredibly, it does seem to be moving in this direction! Manufacturing’s decline as a share of GDP has been arrested. You can see a slight uptick in 2020 on the graph above, and preliminary statistics show that the increase accelerated in 2021, with investment rising in the sector. The pandemic has disrupted everything, of course, so it’ll be a while before the smoke clears. But before Covid, there were indications that Indonesia was benefitting from the U.S.-China trade war, as producers looked to diversify outside China. And it’s possible that “friend-shoring” (also called “ally-shoring”, though Indonesia is not a formal U.S. ally) will motivate additional investment shifts from China to Indonesia. Beyond those political pressures, it’s also the case that Chinese manufacturing wages have risen tremendously, to the point where China is much less competitive in labor-intensive industries than in the 2000s.

Indonesia might thus see a return to the manufacturing heyday of the 1980s and early 1990s, if it plays its cards right. This lengthy 2019 report by the Asian Development Bank and Indonesia’s National Development Planning Agency has a long list of policy suggestions to revive the manufacturing sector. Many of the recommendations involve helping smaller manufacturers scale up like they used to, and reintegrating Indonesia into global manufacturing value chains.

Given Indonesia’s remarkable history, I am optimistic that this program can succeed. Indonesia has already partially industrialized, democratized, and then maintained steady growth despite substantial deindustrialization. And the overall growth of Southeast Asia, especially with the increasing exodus from China, has created very favorable conditions for manufacturing in the region. If any country is versatile and flexible enough to come back from deindustrialization, it’s Indonesia.