Growth Diagnostics Explained: Dani Rodrik’s Method for Designing Development Policy

I've been studying development economics for years, and if you've been following along, you've probably read my posts on structural transformation, corruption, or institutions. Each of these streams of thought is focusing on a certain ‘part of the development problem’ and claim that fixing it is the path to success.

The structuralists point to South Korea's industrial policy. The institutionalists show how property rights transformed Botswana. The governance experts highlight Singapore's anti-corruption quest. After four years of reading their arguments I can tell that their arguments work in the cases they are focusing on. Often even cross country analyses suggests that the problem they are focusing on is the key to unlocking economic growth. Nevertheless, if they're all right, why does their evidence from country A, so often fail when transplanted to country B?

I just finished reading Dani Rodrik's 2005 paper on "Growth Strategies," and as often with his papers I was left feeling like I understand the world a little better now.  Rodrik doesn't side with one or another development strategy. Instead, he makes a nuanced argument about what to do with all of them. He argues that successful reforms are built around economic principles tailored to local capabilities, constraints, and opportunities, and the hard part is figuring out which constraint is the binding one in a specific time and place.

Meet Rodrik’s Martian

Rodrik opens with a thought experiment. Imagine an intelligent alien examines which countries grew rapidly over the past decades, then tries to match these outcomes to whether they followed Washington Consensus reforms (privatization, liberalization, deregulation, fiscal discipline, trade openness).

Our Martian is likely to be utterly confused.

The high performers like China, Vietnam, South Korea, Taiwan systematically violated many elements of the standard recipe. China maintained state ownership while growing at 10% for three decades. South Korea and Taiwan promoted exports through subsidies and import protection, not trade liberalization. On the other hand many Latin American countries in the 1990s implemented the Washington Consensus faithfully yet grew disappointingly slowly. If our intelligent martian can't match policies to outcomes, something about our economic story doesn’t add up. 

Four Inconvenient Truths

Rodrik looks at growth episodes across the developing world and finds four patterns that should make us rethink our approach.

First, growth accelerations require surprisingly little institutional change. Rodrik documents 83 cases of sustained growth accelerations since the 1950s. Most happened with minimal institutional reform. China in 1978 had terrible property rights, massive state ownership, and pervasive corruption, and yet grew exponentially. India's acceleration started in the 1980s, a full decade before the famous 1991 reforms. The lesson Rodrik shows is that the institutional threshold for igniting growth is lower than we might think.

Second, successful strategies mix orthodoxy with creative heterodoxy. Every successful developing country adhered to basic economic principles like property rights of some form, market-oriented incentives, sound money, and fiscal sustainability. But they implemented these through wildly different institutional arrangements. China created market incentives through a “two-track system” that kept central planning while allowing markets to emerge alongside it. Mauritius created an Export Processing Zone with near-zero tariffs for exporters while maintaining a heavily protected domestic market. What unites these cases isn't a standard economic policy structure but rather its creative packaging of economic principles around local constraints.

Third, institutional innovations don't travel well. Taiwan used tax incentives to promote industrialization. South Korea used directed credit. Both worked in their contexts, but when other countries copied them the results were heterogenous. The USSR tried China's two-track system and it failed. Dozens of countries created Export Processing Zones modeled on Mauritius and very few succeeded. Rodrik notes that successful reforms aren't about ‘copy-pasting’ policies but instead they're about matching economic principles to local knowledge and political constraints.

Fourth, sustaining growth is harder than starting it. To support his theory empirically, Rodrik examines 83 growth accelerations and shows that most of them didn't last. The initial reforms that spark growth aren't sufficient to sustain it. Many accelerations collapsed when countries faced external shocks because of weak institutions of conflict management. For example, he compares Indonesia and South Korea during the 1997-98 crisis to showcase that even though they were going through the same shock and had different institutional capacities, both experienced radically different outcomes.

So, where do these findings leave the economist dealing with development problems in the 21st century?

The Growth Diagnostics Approach

I can already hear the development consultants asking: ‘Fine, but if no universal recipe exists, what prescriptions are we supposed to put in our reports?’ Rodrik's answer is start thinking less like pharmacists and more like doctors.

Rodrik finds the answer in “growth diagnostics” which is a systematic framework for identifying the most binding constraint to economic activity. The core insight of this theory is that not all constraints to one’s growth are not made equal and resources for reform are limited. Focus your limited reform capacity on what actually matters.

Growth diagnostics can take many shapes and forms, but to simplify it lets take an example where we can ask: ‘why is private investment and entrepreneurship low?’ Then systematically narrow our questions down: is it low returns or high cost of finance? If low returns, is it infrastructure and human capital, or is it that investors can't capture what they produce? If the latter, is that because of government failures like corruption, or market failures like coordination problems?

How do you know which constraint binds? Where a constraint truly binds, its shadow price will be high. This means you should see high returns to relaxing it, and quick results when you do. Agents also work around binding constraints in costly ways meaning in an environment with bad property rights people tend to invest in cattle instead of machinery. Behaviors like these are diagnostic clues.

Diagnostics in Practise

This paper has made me take everything I have learned about economic policy and understand these frameworks are empty if not combined with a good understanding of what exactly is likely to work in a specific context. Not because institutions don't matter, they do. Not because corruption isn't harmful, it is. But because importance and urgency are different things.

Property rights matter everywhere, but they're not the binding constraint everywhere. Education matters everywhere, but it's not always the most urgent priority. Corruption is often part of the issue, but not always an addressable constraint. Rodrik as well as other scholars in the growth diagnostics circles acknowledge that development policy is as much of a science analysis under limited information as it is an art of prioritization under constraint.

This is why structural economists, institutionalists, and governance scholars can each be correct in their preferred cases, but attempting to turn their insights into a universal ‘manual for growth’ is likely bound to fail by design. The bottom line is, South Korea's industrial policy worked because it relaxed binding constraints around coordination and finance. Trying the same thing where governance or macro stability is the binding constraint would likely not work.

If you find this perspective on development compelling, I recommend exploring The Atlas of Economic Complexity by Ricardo Hausmann and César A. Hidalgo, an interactive resource that allows you to view countries through this analytical lens.

Bibliography

Rodrik, D. (2005). Growth Strategies. Chapter 14, in Aghion, P., and Durlauf, S. (editors), Handbook of Economic Growth, pp. 967-1014. North Holland.