Can India industrialize?
The most important economic development question in the world.
And so my series of posts on developing-country industrialization comes to an end. There will be one more post linking to all the others (update: here it is), but the last country I’m going to look at is India. And I saved it for last because it’s the most important one.
Why is India the most important development story on Earth? Two reasons. First of all, sheer size; as of this year, India is the world’s most populous country, overtaking a now-shrinking China. And its population dwarfs that of every development success story I’ve written about in this series so far, combined.
But India isn’t just incomprehensibly enormous; it’s also still quite poor. Despite its remarkable 250% growth over the past three decades, it’s still poorer than almost any other country I’ve discussed so far.
Together, India’s massive size and low income levels mean that it has far more poor people than any other country on the planet. This is true if you look at extreme poverty (people living on less than $2 a day), and it’s even more true if you look at the number of people living on less than $3.65 a day:
Remember that the primary task of economic development is to bring people out of poverty. So if you could pick one country to industrialize in order to do the most good for humanity, it would be India. And the experience of China — the only country comparable to India in size — is very encouraging here. China’s industrialization over the four decades starting in 1979 brought about the largest and most dramatic poverty reduction in the history of the human race; if that feat could be repeated in India, it would be absolutely incredible. When Malaysia or the Dominican Republic reaches upper middle income status, it’s a good thing and a useful success story; if India did the same, it would shake the world.
India’s rapid growth in the 1990s, 2000s, and early 2010s was driven by a spate of policy changes — pro-business policies in the 1980s, followed by economic liberalization in 1991. But it’s widely believed that those policies have reached the limit of their ability to drive rapid growth on their own, and the economy saw a deceleration in the mid 2010s. Something more is needed in order to propel India onto a China-like trajectory. So let’s talk a bit about where India’s development currently stands, and then think about they might apply some of the lessons from other countries.
India lags in manufacturing but does OK on exports
The stereotype of India’s economy is that it relies on services more than on manufacturing. That’s actually pretty accurate; manufacturing is a smaller percent of the economy than Bangladesh or Vietnam, and the trend line is headed in the opposite direction:
But according to what I’ve been calling the “Chang-Studwell” theory of development — based on the book How Asia Works, any of Ha-Joon Chang’s books, and this IMF paper — what matters most is not manufacturing per se, but exports. And I don’t mean “net exports”, i.e. trade surpluses vs. trade deficits — I simply mean the amount of stuff a country sells overseas. The basic idea is that exporting forces companies to raise their productivity levels and learn foreign technologies (by hiring foreigners, by maintaining overseas offices, or just by stealing intellectual property). In this theory, manufacturing is important simply because manufactured goods are easy to export, and because manufacturing industries have opportunities for rapid productivity growth.
And India actually does a pretty good job exporting — about as good as China, and as good or better than China in the 1990s.
Looking at the composition of exports, meanwhile, allows us to get an idea of what a country specializes in. India’s exports of physical goods are pretty evenly divided between low-value manufactured goods on one hand, and raw materials and agricultural products on the other hand:
In fact, this doesn’t look too bad to me. Yes, there are some natural resources on here, and you never want to be stuck as a natural resource exporter if you can help it. But there’s a fair amount of manufactured stuff here as well. Yes, it’s mostly cheap stuff (though cars are pretty high-value, and there’s some electronics in there). But a country has to start somewhere.
What does worry me a bit more is that there’s not much labor-intensive manufacturing here — garments, toys, furniture, and electronics assembly. Making these sorts of products is usually where countries start on their road to industrialization — even Britain famously started with the garment industry when it first industrialized. Labor-intensive manufacturing is useful for generating employment, for moving poor people from farms to cities, and — at least, if you believe Ha-Joon Chang — for developing a widespread culture of manufacturing. And yet India doesn’t really specialize in this; its manufacturing exports are more likely to be capital-intensive things like chemicals. Contrast this with Bangladesh, which has focused single-mindedly on the garment industry. Bangladesh has moved people out of agriculture faster than India has, and has urbanized faster:
Labor-intensive manufacturing probably helped with that. So this is one area India could try to improve.
But anyway, India has one more huge category of exports that isn’t shown on the OEC chart — services exports. World bank numbers suggest that India’s exports of services are 60% as large as its exports of goods. That’s really big! The stereotype that India is unusually dependent on services for its development appears true, at least as far as exports are concerned.
The thing about services is, we don’t really know how well they contribute to development. Services have only really been exportable en masse for a short amount of time, thanks to the advent of the internet, so there isn’t a long record of countries that exported a bunch of services, as there is for manufactured goods. There are at least two main reasons to think services are less useful.
First, development runs on agglomeration effects — producers, suppliers, and customers all wanting to locate near each other — and services don’t require nearly the same supply chains that manufacturing does. This means we might expect to see services generate a smaller local multiplier effect (sometimes called an “external multiplier”), leading to less urbanization and product diversification.
Second, it’s generally considered harder to improve productivity in services than in manufacturing. This might not be true for all kinds of services — software engineering and various types of research, for example, count as a service export but are probably amenable to some of the same types of rapid technological and organizational improvements that manufacturing enjoys. But for service exports that are basically just local services beamed overseas by the magic of the internet — for example, India’s famous call centers — there probably isn’t as much room for rapid technological upgrading.
Anyway, this section was a bit of a long-winded, graph-heavy review of stuff that most Indian people have been hearing for many years — India needs to manufacture more, services aren’t a complete substitute, and yada yada. But I just wanted to point out that the situation isn’t nearly as stark as you might hear; manufactured exports are actually a decent percent of the total, and India is starting to have a presence in the industries like cars and electronics that tend to feature heavily in countries that boost themselves to upper-middle-income status. So there is a base to build on here.
Now that we’ve talked a bit about where India stands, let’s look at some of their recent efforts to improve the situation.
Big progress on infrastructure, less on education
At this point I should note that about a year ago, I interviewed Arvind Subramanian, a former Chief Economic Advisor to the Indian government, and co-author of some of the papers I cited above. Subramanian of course knows infinitely more about the Indian economy than I do. The interview is worth reading in full, but I’ll cite bits of it as we go along.
The most important thing we talked about was the policy efforts made by the Modi government since 2014. India needs to go beyond the pro-business and liberalizing reforms of the late 20th century, and since Modi and his political movement remain incredibly popular, they’re going to have to be the ones to do it.
One of the most important tasks for boosting Indian manufacturing is to improve infrastructure. To make physical goods, you must to be able to move parts and components to factories, finished products to markets, and workers to their jobs. India has lagged in this department for a long time, but as Subramanian noted in our interview, things have started to improve in the 21st century:
On infrastructure, since the late 1990s, things have really stepped up. It began with the government constructing rural roads and highways connecting the major cities. This was followed in the 2000s by a major private sector-led push across the board not just in roads but ports, airports, power and telecommunications. And the Modi government has continued this infrastructure thrust.
The combined length of roads in the country more than doubled between 2003 and 2019. The length of national highways has accelerated sharply since Modi came to power, going from 79,000 km in 2013 to 141,000 km in 2019. The volume of cargo handled at major Indian ports rose by almost 50% from 2011 through 2020. Electrical generation has been growing at a steady clip, and railways and villages have been electrified en masse over the past decade. Tap water access and access to toilets have increased substantially, thanks to a big push from Modi.
This is a huge change for India — not only for basic quality of life, but for the country’s productive potential. Access to reliable transportation, electricity, and water is one of the most crucial factors for manufacturing investment, and it’ll also help speed up India’s painfully slow rate of urbanization. And though India hasn’t yet built infrastructure on the same scale as China, its glaring deficits in that area look like they’re being rapidly remedied.
The other big resource manufacturers need is a populace with basic education, especially literacy and numeracy. We tend to think that low wages are the main thing that lures manufacturing investment, but cheap labor doesn’t matter if the workers can’t read or do basic math. And here India has traditionally lagged, with low high school enrollment rates and chronic widespread absenteeism by schoolteachers themselves (though some have challenged the absenteeism numbers). As a result, India’s years of schooling and literacy rates are considerably lower than peers like Vietnam, to say nothing of richer countries like Malaysia:
Again, note the impressive recent surges in Bangladesh.
In keeping with his general infrastructure push, Modi has put a lot of money into developing school facilities. And high school enrollment rates are slowly increasing. But it seems to me that a lot more could be done to boost education and literacy.
“Make in India” shouldn’t be “Make for India”
The Modi government has done more than just try to improve infrastructure and education; it has also engaged in industrial policy. The “Make in India” plan, launched all the way back in 2014, was a major effort to boost the manufacturing sector, basically involving a combination of deregulation, subsidies, and wooing of foreign investors. So far, however, the results have been underwhelming. As you can see in the chart in the last section, India’s manufacturing sector has actually shrunk as a percent of the economy over the last decade. Even India’s climb up the “ease of doing business” rankings appears to have been due mostly to a change in the way the rankings were calculated.
Why has “Make in India” failed? It’s possible that the Indian regulatory environment and business culture just weren’t suited to this kind of big push for manufacturing, and that it would have succeeded in China or Vietnam. But even China, for all its legendary manufacturing prowess, has failed to get much traction with its own “Made in China 2025” subsidies. So perhaps there’s a more systematic reason why Modi’s initiative didn’t work out.
Channeling Chang and Studwell, my guess is that one big mistake was to focus on making things for the domestic market. If you look at the list of projects associated with “Make in India”, most of them seem to be about paying both foreign and domestic companies to make things in India for the Indian market. Subramanian confirms that this was the focus:
But it is not obvious that the government's new strategy goes in the right direction. Its new industrial-cum-trade policy involves promoting domestic champions (mostly in non-tradables unlike the Korean chaebols which had to prove the test of efficiency by exporting), [and] offering production subsidies (mostly in capital and technology-intensive sectors rather than labor-intensive ones)…This is unlikely to be a recipe for a labor-intensive export boom.
In fact, this approach is a very well-known policy called “import substitution”, and it was quite popular in the 1960s in Africa and Latin America. Douglas Irwin has a good paper tracing the rise and fall of the import substitution idea, which by the end of the 20th century was widely considered a failure.
Why did import substitution fail? Chang and Studwell’s answer would probably be that making things for the domestic market doesn’t force companies to increase their productivity. It doesn’t help them discover their comparative advantage relative to foreign companies. It doesn’t push them to develop new products. It doesn’t give them much of an incentive, or even much of an opportunity, to absorb foreign technologies. The domestic market is safe, familiar, and uncompetitive, and it’s often possible to dominate it through political cronyism rather than through brutal technological competition.
In this sense, India’s giant domestic market is a long-term asset, but in some ways it’s also a short-term liability. If you were an electronics maker or a carmaker in Singapore or South Korea or Poland back when those countries were poor, you just didn’t have a huge safe domestic market to hide in. But if you’re an Indian manufacturer right now, even though India is still poor, the domestic market is still so big that there’s less incentive to take the icy plunge into the international waters. If you think this theory is right, then India will need to work especially hard to push its companies to make things for the world instead of just for India.
Fortunately, the Modi administration may have (belatedly) gotten this message. Last year he announced a new slogan: “Make for the world.” Presumably this means a shift from production incentives to export incentives. That would be a welcome shift, but it has to be done right. As Studwell writes, export subsidies can’t be permanent; not all companies have what it takes to succeed in international markets, and the ones who fail shouldn’t be kept alive as zombies. This is what Studwell calls “export discipline”.
SEZs, FDI, and India’s golden opportunity in electronics
In my last post on Poland and Malaysia, I talked about an idea that goes a little beyond Chang and Studwell’s recommendations: wooing foreign direct investment. The Modi administration is obviously all-in on the idea of FDI, having seen how effective it was in China. And India has already been trying to use one of the most successful tools that Poland, Malaysia, and China all used to lure FDI: special economic zones, or SEZs. These are areas with special infrastructure, lower regulation, tax incentives, and other incentives for foreign and domestic businesses that set up shop in the zone.
There are many advantages to SEZs. If your country is over-regulated (as India probably is), an SEZ gives the government an excuse to essentially deregulate within a small circumscribed area. If your country has patchy infrastructure (as India still does), at least you can have great infrastructure in and around an SEZ. SEZs also harness the economics of industrial clustering — they concentrate talent within a small area, and they locate suppliers near to the companies that will need their products. Finally, they serve as a method of advertising — famous SEZs are places that multinational companies know they can go set up shop.
In 2000, India created several “export processing zones”, which it upgraded to SEZs in 2005. These zones did have a positive economic effect; a 2018 paper by Hyun and Ravi found that India’s SEZs “boosted economic activity within areas several times the size of the zones”, and “drove a structural change in the local economy”. But overall, the consensus is that India’s SEZs have been a disappointment, especially compared with China’s blockbuster success. Overall, India has attracted some FDI, but not nearly as much relative to its economy as recent development stars like Malaysia, Poland, and Vietnam:
One complaint about the SEZs is that they encouraged investment in service exports. Sachin Menon of KPMG writes:
The SEZ scheme was introduced with the intention of boosting foreign direct investment in export-oriented manufacturing. However, the sector which actually gained the momentum in recent few years is [the] IT/ITEs sector….[T]he highest number of operational SEZs [relate] to IT…electronic hardware, [and] the semiconductor and telecom equipment sectors, comprising ~60.42 per cent of [the] SEZs in India. Thus, the initial focus to boost the manufacturing sector did not materialise, and a rather unprecedented jump was seen in service exports, IT software, electronic items, assembled parts such as printed circuit boards (PCBs), etc.
In fact, this sounds like a big win to me! Electronic hardware, semiconductors, and telecom equipment are exactly the manufacturing sectors India should be trying to promote!
Electronics are much lighter than, say, auto parts or steel or construction equipment, so they’re easier to ship around the world. India is located far from the key markets of Europe, the U.S. and developed East Asia, meaning that exporting heavy products might not be economical. But electronics pack a ton of economic value added into a very small, light object, and so they’re perfect for globally integrated supply chains. Just look at how Malaysia succeeded with electronics after famously failing in the auto industry.
Not only that, but electronics offer plenty of opportunity for technological upgrading and climbing up the value chain. The most high-tech, important, expensive manufacturing industry in the world is not autos, but semiconductors. Instead of regarding electronics as a lesser form of manufacturing, India should jump at the chance to get in on the ground floor.
And now is exactly the time to do it. The U.S. rivalry with China offers India a golden opportunity to capture some of the all-important electronics supply chain. Apple and its contract manufacturers, desperate to diversify out of China, are moving some iPhone and iPad production to India. Other electronics companies from the developed democracies are also eager to make the shift. The push for “friendshoring” is a golden opportunity for India, which is seen as politically much safer than China.
So while service exports are probably not the most useful for development, and it would be fine to restrict SEZ incentives so that they only apply to manufacturing, India’s SEZs are absolutely right to focus on electronics hardware.
But still, they could be doing a better job of it. One problem seems to be the small size of the Indian SEZs. China has only 20 such zones, and each one is a whole city or province; India has hundreds, and they’re tiny. Small size defeats the basic purpose of SEZs in a number of ways. If you want to make an end run around red tape and regulation, it’s much better to do this in a whole city than just in an office park. If you want to harness the benefits of clustering, you should make sure the cluster is big enough to accommodate a whole ecosystem of suppliers. And if you want the managers of multinational companies to know the names of your SEZs, there should be only a few names instead of a few hundred.
So expanding a few Indian SEZs to the size of whole cities would be a good idea.
This would also represent an opportunity to give those cities, or their states, more local control over the specific incentives in the SEZs. In our interview, Subramanian talked about how the idea of “competitive federalism” in India has fallen by the wayside in recent years, as power has been centralized in the hands of the national government. For many government functions, such as infrastructure and taxation, centralization is a good thing. But when it comes to industrial policy, like the incentives in SEZs, local governments can be better at experimenting with different ways of luring foreign investment. This strategy of localized industrial policy was highly effective in China in the 1990s and 2000s.
Rural areas: Time for another round of land reform?
So the most promising industrial policies for India would seem to be:
keep improving infrastructure, and make a big push to improve public education
“export discipline” through limited-time export incentives
encouraging FDI, especially in the electronics industry, through the use of improved SEZs, targeted deregulations, and taking advantage of “friendshoring”
These policies should help to grow labor-intensive manufacturing industries in India’s cities, which will pull rural “surplus labor” off the farms, increasing income in rural areas. But I wonder if India should also try to help its rural areas through another round of land reforms.
Export discipline is the most famous part of the development model that Joe Studwell advances in How Asia Works, but it’s not the entire model. The first step, Studwell argues, is land reform. Citing Taiwan, Japan, and South Korea as successful examples, Studwell argues that poor countries should buy farms from big landlords and give it to the tenant farmers who work on that land — a policy sometimes known as “land to the tiller”. India itself has a history of land reforms, some of them in this general vein. But the country had 144 million landless farmers in 2011, and that number is believed to be increasing, so there’s plenty of scope for another policy like this.
“Land to the tiller” policies are believed to have a number of beneficial effects. First, when small farmers own their own plots, they tend to farm them more intensively, which can raise agricultural output (this effect was famously observed by Indian economist and Nobel prize winner Amartya Sen). Second, giving small farmers land gives them a bit of property, which they can then sell off in order to start a small business or move to the city and become a factory worker. Also, in countries like India that have a lot of informal rural property rights, land reform can be a way to formalize landownership, which paves the way for greater rural entrepreneurialism later on.
But potentially the most important benefit of a “land to the tiller” policy is that it forces landlords to do something more entrepreneurial with their time and money. Landlords tend to be business-savvy, educated people whose talents could be better used starting export-oriented businesses than kicking back and collecting rent checks. After their landholding is liquidated, their best move is to move to the cities and start other kinds of businesses.
Anyway, this is a fairly radical idea, and would certainly take some major political will. But Studwell presents it as a key feature of what made the East Asian economies rich, so I thought it deserved a mention here.
Optimistic about India
All in all, I have to say I’m optimistic about India. Development seems to have a momentum that’s as much psychological as economic — once the people of a country know that they can achieve rapid growth, their hunger is whetted for more. The reforms of the 1980s and 1990s didn’t get India all the way to developed status, but they gave Indians a golden 25 years during which they started to realize just how great their country could become. And I think much of the rest of the world realized it as well.
So now is the time to keep pushing, with bold, persistent experimentation. There is no grand unified theory of economic development — every country’s situation and resources and challenges are unique, and though they can take useful examples and ideas from each other, in the end each one’s path to riches will look a bit different. What worked for India will not be quite the same as what worked for China, or what works for Vietnam or Bangladesh. But I believe that something will work. The world’s largest country will find its way out of the darkness of deprivation, and take its rightful place among the industrialized nations of the world.