To many people, “development economics” is synonymous with the randomized controlled trial: randomly assigning some group of people to an intervention, like cash transfers or malaria protection nets, in order to tell if it really makes the recipients better off. RCTs have revolutionized development economics and international aid policy. The RCT pioneers Esther Duflo, Abhijit Banerjee, and Michael Kremer won the Nobel Prize in Economics in 2019, and the chief economist of USAID, Dean Karlan, is a prominent RCT advocate. RCTs are influential because they are a powerful tool to cut through uncertainty and preconceptions. For example, international development practitioners long believed that providing textbooks in underfunded schools would help students learn more: but an RCT in Kenya showed that most students did not benefit from getting textbooks, primarily because they were already learning very little in school. RCTs can help us identify what works. Yet nearly everyone who thinks about development has, at some point, had a crisis of faith.

Skeptics argue that RCTs distort the focus of policymakers toward technocratic questions that lend themselves to studies of limited scope, and away from the hard problems that are ultimately more important. After all, aren’t RCT-driven solutions just Band-Aids over the fundamental reality that poor countries are poor? Shouldn’t we spend our time and money focusing on how to bring economic growth and prosperity to countries, rather than tinkering with partial solutions for a tiny number of people?

This argument is appealing, but it faces difficulties when the rubber meets the road. What would it look like to adopt a “growth strategy” for doing good? One policy commonly proposed by international agencies like the World Bank or the International Monetary Fund is trade integration between developing countries and global markets. Trade almost certainly increases growth, so reducing trade barriers could be a huge win. Thus, a candidate growth-first intervention might be funding a think tank that advocates for reducing trade barriers in a large, relatively protectionist country like Nigeria.

But are we certain that a large-scale increase in trade integration is the right goal? It does likely increase growth, but it may have increased rural poverty in India, caused lasting unemployment in Brazil, and led to political polarization in the U.S. These costs eat away at the benefits. Even if trade integration is the right goal, how would we pick which sectors to focus on? And how likely would our think tank be to succeed in its advocacy for lower tariffs when other political interest groups benefit from high tariffs? As the costs and uncertainties stack up, our growth-first policy starts to look less appealing. The intervention could still be the best way to promote economic growth, but you can start to understand why this approach isn’t so appealing to many people.

The modern, RCT-driven style in development economics arose in part as a response to challenges like these. Macroeconomic policy is hard. Every country has different needs and constraints — and the poorest countries in the world are usually the ones that struggle the most to design and implement good policy. 

But the right response to this difficulty is not to abandon growth as a goal and go back to narrow household interventions. Instead, we should search for a “missing middle.” Are there interventions that are likely to impact growth, but which have a much stronger evidence base, and are narrow enough in scope to be easily implementable? There are different ways to conceptualize what this middle ground looks like, but I’ve chosen to focus on one approach that is naturally relevant to international aid spending: interventions for firms. 

The founding question of development economics was “How do we make poor countries rich?” It was a powerful motivation, but it was too abstract to generate credible prescriptions. Instead, it led to a lot of dubious cross-country correlations presented as causal evidence, and broad statements about the causes of growth (“institutions,” “ideas,” “human capital”) that might be correct, but didn’t translate into sharp policy recommendations.

But you can make a remarkable amount of progress with a simple observation: A country’s economic growth can be understood as the aggregate of the growth of its individual firms. ¹ Specifically, a country’s growth rate is mechanically determined by three factors:

1. Growth in the number of firms

2. The average growth rate of each firm

3. The market share of high-growth firms relative to low-growth firms

Accelerating any of these three factors will increase a country’s growth rate. I can’t emphasize enough how important this reframing is. Firms aren’t the only part of the development process. Bad institutions, geographic disadvantages, low agricultural productivity, colonial legacies — these are all factors that can keep countries poor. But it is almost a mathematical truism that if a country’s firms are growing on average, then that country’s economy will grow. 

And while “How do we make countries grow?” produces platitudes, “How do we make firms grow?” has real answers. Economists can study the impact of economic policies on companies in the same country without having to make messy cross-country comparisons, or use traditional RCTs to study which policies help firms grow. As a result, development economists have written hundreds of papers about firm growth, and what policies could accelerate it. This flavor of development research marries the credibility of RCTs with a focus on bigger questions of growth. Below, I’ve summarized five key takeaways from this research.

Above is the figure that radicalized me on firm growth as a key lens through which to view development.

In the U.S., companies either grow or they die. If a company survives for 10 years, its employment has on average tripled from when it started. Of course, most don’t survive 10 years: The growing firms are gaining employees from all the firms that exited the market, which is why exit is as important to this figure as growth. In contrast, firms in India are almost completely stagnant. The average Indian firm that survives to year 20 has only added 20% more employees compared to when it started. The life cycle of firms in Mexico is somewhere in between, but closer to India. Subsequent research has found similar patterns in Colombia and Ghana. Firms grow very slowly in developing countries, even when the country itself is growing. 

This stagnation also explains why firms in developing countries are overwhelmingly small. If you started with a collection of small firms and had the same “grow or die” pressures present in the U.S., you would quickly end up with a smaller number of large firms — which is exactly what we see in rich countries.

Instead, most progress in developing countries seems to take the form of new firms entering the market, and staying small. Can that be a viable strategy for genuine growth? Probably not. Many modern goods can’t be produced without scale. (Imagine a five-person steel mill!) When firms stay small, countries are stuck with low-value manufacturing, stunting their economic output. 

97% of firms in India, 96% of firms in Indonesia, and 91% of firms in Mexico have fewer than 10 employees. Of these, most are just a single owner-operator, or perhaps a household enterprise. 55% of employment in developing countries is self-employment, rising to a staggering 77% in sub-Saharan Africa. These individuals operate firms, producing goods or providing services. Indeed they operate most firms. If we want to enable firms to grow, how should we think about these self-employed people?

One possibility is that self-employed people are “micro-entrepreneurs.” They would like to grow their enterprises, but don’t have the resources to do so. This is the premise of microfinance, the most popular development intervention of the 2000s. Microfinance is the practice of giving households small loans that they can use to set up or grow a home business. If self-employed people are really microentrepreneurs, then the key to firm growth is giving them more access to capital. However, many different lines of evidence suggest that this view of self-employed people is inaccurate, and that it is more accurate to think of them as workers looking for wage employment than as entrepreneurs. 

In developing countries, self-employed people transition to wage employment at similar rates as unemployed people — and earn similar wages when they do. This isn’t what we would expect to see if self-employed people intended to grow their businesses as “microentrepreneurs.” In that case, they would be reluctant to quit their enterprise and take a wage job. This is exactly the behavior we see in rich countries, where self-employed people transition to wage employment much less frequently than unemployed people, and do it for higher wages. 

Microfinance studies also reveal that microloans have very little average impact on household or business outcomes. Most businesses run by an individual or a household are just not designed to scale. All of these facts point to self-employed people behaving more like unemployed workers than like entrepreneurs — which is to say, looking for jobs rather than aiming to create them.

If self-employed people act more like unemployed workers than business owners, that implies that we should not design policy to bolster the growth of microenterprises. These microenterprises are desperate measures in the absence of wage employment, and will melt away if and when formal-sector firm growth creates more jobs. Formal-sector firm growth is key to making developing countries grow.

Another urgent implication of this fact is that there is an unemployment crisis in developing countries that isn’t captured by official statistics. The typical approach to measuring unemployment is to ask whether people want to work but are unable to find any opportunities to do so — including self-employment. By this measure, the official unemployment rate in developing countries is 5% and 6% — around the same as in developed countries. However, if self-employed people are unemployed “in disguise,” this number could be much higher. One study estimated that at least 24% of self-employment during India’s agricultural lean season occurs solely because workers cannot find jobs. If we (loosely!) extrapolate this to a sub-Saharan African country with 77% of workers being self-employed, then the true unemployment rate jumps from 6% to 25%! Even if only 50% of workers are self-employed, then the true unemployment rate is still 18%. That level of unemployment is a catastrophic failure, and a crisis that cuts against both poverty alleviation for individuals and aggregate growth. 

Firms can only sell to the people they can reach. In developing countries, mobility is very low, even in urban areas. This means that people are more likely to buy from stores near them, regardless of quality. When gas prices increased in Mexico City last year, retail shops proliferated, suggesting that the new shops arose only because high gas prices led people to travel less. Moreover, developing countries have high internal trade costs (according to data from Ethiopia and Nigeria), both because of poor infrastructure and because market power in the transportation sector drives up freight rates (as we see in Colombia). Thus, companies face barriers to selling to large markets. If they are not in big cities, it’s expensive to transport their goods there, and even if they are nearby, consumers will prefer closer firms.

It can also be hard for firms in developing countries to let potential customers know they exist. If consumers are only aware of the firms closest to them, then they will never switch to purchasing from more productive competitors. We can see this dynamic at play in a natural experiment in the Indian state of Kerala. Before the spread of mobile phones, fishermen primarily bought boats from the boatbuilder in their village, regardless of quality. But after mobile phones became common, fishermen were able to exchange information about the price of fish in different villages. As a result, they traveled more to other villages to sell their catch, and encountered more boatbuilders. In six years the number of boatbuilders fell by 60%. Employment and sales were reallocated to the survivors. Thus, information spread led the most productive firms to expand.

This second barrier to firms selling across markets — the lack of information — could be addressed by interventions that spread price information to buyers and sellers. This is what happened after the introduction of cell phones in Kerala. Another study found that building an electronic platform for agricultural trade in Uganda increased trade between villages, caused prices to converge between them, and increased farm revenues.

The direct effect of these two frictions is that firms sell less, so they grow less. But their indirect consequences are possibly even more important: without prospects to sell to a large market, firms don’t have any incentive to invest in productivity improvements. In one RCT, Ugandan maize farmers only invested in increasing the quality of their harvest after they were given the opportunity to sell to markets that paid a quality premium. Without market access, investing in productivity won’t actually make companies more profitable, so they don’t do it. And because these frictions create little monopolists in their corner of the market, they have no outside pressure to increase productivity either. There are ways to help firms adopt better technologies and practices, but those are useless if firms don’t have an incentive to learn. This makes fragmented markets possibly the biggest barrier to firm growth.

Physical market size has a clear solution: build better transportation infrastructure. When India built the Golden Quadrilateral, a national highway network connecting major economic, industrial, and agricultural centers across the country, it increased competition and boosted the market share of productive firms. And more competition within the transportation sector further brings down transportation costs, making it easier for the best firms to “export” to other areas of a country. E-commerce logistics chains can have a similar effect: E-commerce expansion in China increased the exporting of goods from cities to the countryside.

Firms in developing countries use less advanced technology than firms in rich countries in the same industry. Some of this gap is explained by differences in firm size — small firms will naturally not benefit as much from high-fixed-cost technologies — but even large firms in developing countries use frontier technology at a much lower rate.

One simple way to address this gap is simply to import advanced technology from rich countries. This can work: When Chinese automakers enter joint ventures with Western automakers, the quality of their cars improves. But for this to work, the firms need to get knowledge along with machines. A different study found that when Chinese manufacturing clusters received Soviet steelmaking machines, they became more productive, but this effect died out within 20 years — after the machines wore out. In contrast, when they received both machines and training from Soviet engineers, their productivity growth was higher for decades more. Successful technology transfer requires knowledge transfer. This is a key lesson for governments trying to facilitate technology transfer: They have to ensure that foreign firms train domestic firms, not just give them technology.

This also means that it’s important for developing countries to invest in skilled workers. If a country has a severe shortage of good engineers, then it may not be possible for firms to learn how to use advanced machines. Leapfrogging — where firms go from primitive technology straight to advanced technology —doesn't seem to work. When firms upgrade their technology, they tend to move incrementally up in complexity rather than leaping to the frontier, suggesting that there are difficulties in absorbing technology that is too advanced. 

Firms in developing countries also tend to have weak management practices. In an RCT of Indian textile firms, researchers found that management training from a global consultancy reduced wastage of materials, increased labor productivity, and improved the quality of outputs — ultimately increasing profits by more than the consulting fees. This is surprising, given that most of the firms in the experiment had been in business for over 20 years and the management practices they learned were no secret. Yet they were unaware of many of the new practices they were trained on, and they were skeptical of the practices they did know about. So a key barrier to improving firm productivity in developing countries is helping firms learn about better practices.

Nonetheless, it seems to be true that firms in developing countries are not well-managed, and this management gap matters. Just like frontier technology, there is a lot of promise in importing management practices as an intervention. The RCT in India provides the strongest evidence for this possibility, and a clear model of what to do: Subsidize or otherwise encourage firms to take up management consulting. There’s also good evidence that these effects can persist over the long run. Italian firms that participated in an eight-week management training program in the U.S. became much more productive than firms that were initially enrolled in the program, but couldn’t attend because of a budget cut. This effect persisted for decades. So it seems quite likely that training firms in developing countries on management practices is effective, though exactly how to best operationalize that idea — and most importantly, scale it — is still an open question.

One insight from the East Asian growth miracles — countries like Japan, South Korea, Taiwan, and Singapore — is that interacting with the global market is a catalyst for productivity growth. Trading internationally helps domestic companies learn from global firms with frontier technologies and practices.

When Argentina and Brazil liberalized bilateral trade, firms in Argentina that gained more export opportunities increased their spending on technology, suggesting that access to foreign markets incentivized firms to upgrade. And when rug manufacturers in Egypt were randomly given access to export opportunities, they became more profitable — not by producing any faster or cheaper than the control firms, but by making rugs of higher quality. When the researchers asked them to make rugs of a standard quality, they did so faster than control firms — showing that they could have chosen to produce normal rugs faster, but preferred to translate their productivity gains into higher quality rather than lower cost.

This distinction matters: Quality upgrading is the holy grail of firm productivity growth. Why? Because in most manufacturing sectors, there’s a hard limit on how much you can reduce costs, but there’s no limit on how high-quality you can make a product. Apple didn’t become a trillion-dollar company by making phones and computers cheaper than every other company. Having a higher-quality product makes firms harder to replace. Many companies can make great phones, but only one can make the iPhone. 

In contrast, when a country’s firms specialize in low-cost manufacturing, there’s always a risk that they will be displaced by competitors in another country with low labor costs. China began as a low-cost manufacturing destination. “Made in China” used to be a pejorative; while that connotation hasn’t fully disappeared, it’s improved dramatically as China has invested in moving up the value chain and producing higher-quality goods. Quality upgrading is everything! So the fact that exporting increases firms’ quality is very important, and suggests that creating export opportunities is one of the most effective ways to help firms grow.

“Making businesses big” is not a sexy intervention. Its beneficiaries are hard to plaster on donor brochures. But they are no less real. When a textile manufacturer in Mexico can export to the U.S., it hires more people to help meet that demand. When more and more people can get these stable jobs in the formal sector, rather than having to scrape by with self-employment, they have a path to genuine prosperity for themselves and their children. Even helping one firm grow can provide this path to a hundred people — far more leverage than we can get by targeting individual households. 

This idea is not revolutionary. But seriously reckoning with it requires a fundamental shift in how we view international development. We cannot endlessly triage temporary solutions for a few people, while countries stumble on the path to prosperity. But neither can we treat growth as a black box, feeding in “growth advocacy” and hoping that it spits out a higher growth rate. The RCT revolution in international development created a culture of empiricism about which policies work. We need to apply that rigor to more ambitious questions. One way to do that is to understand that economic growth is intrinsically linked to the growth of firms. That allows us to draw on years of careful research about firms in developing countries to design interventions that can help firms grow — and with them, everyone.