Development economics has a long history of grand prescriptions. Structural adjustment, the Washington Consensus, good governance reforms, human capital investment—each era produced its own checklist of what poor countries supposedly needed to do. The trouble is that countries following similar advice ended up with wildly different outcomes.

This puzzle motivated Ricardo Hausmann, Dani Rodrik, and Andrés Velasco to propose something different in the mid-2000s. Rather than assuming every economy suffers from the same maladies, they suggested we should diagnose each patient individually. What constraint is actually holding this specific country back, right now?

Growth diagnostics is not a theory of development. It is a decision framework—a way of prioritizing reforms when governments cannot do everything at once. Its appeal lies in acknowledging what practitioners have always known: political capital is finite, administrative capacity is limited, and the sequence of reforms often matters more than the reforms themselves.

The core insight is deceptively simple. In any economy, many things are wrong. Infrastructure is inadequate, education is uneven, courts are slow, credit is scarce, and corruption is present. But not all of these problems constrain growth equally at any given moment. Some are binding constraints; others are merely suboptimal conditions the economy has learned to work around.

The framework proposes a decision tree. Is growth low because returns to investment are low, or because the cost of financing investment is high? If returns are the issue, is it because of low social returns (poor infrastructure, weak human capital) or low private appropriability (high taxes, weak property rights, macroeconomic instability)? Each branch narrows the diagnosis further.

The trick is looking for telltale signs. If finance is the binding constraint, we should observe high real interest rates, credit-worthy borrowers being turned away, and firms self-financing at unusual rates. If poor infrastructure binds, we should see high logistics costs eating into margins and firms clustering near ports or highways. The absence of these symptoms suggests the constraint lies elsewhere.

This approach flips conventional thinking. Instead of asking what should a developing country have, it asks what is currently preventing this economy from growing faster. The two questions sound similar but generate very different reform agendas.

Takeaway

A problem is not the same as a priority. The binding constraint is the one whose removal would produce the largest gain in growth—everything else can wait its turn.

El Salvador in the 2000s provided one of the first practical applications. On paper, the country had done many things right: stable macroeconomy, dollarized currency, open trade regime, reasonable institutions by regional standards. Yet growth remained sluggish. Standard prescriptions had little left to offer.

The diagnostic pointed to something less obvious: low returns to investment driven by weak self-discovery. Salvadoran entrepreneurs had few models of what tradeable goods the country could produce profitably. The constraint was not capital or education but the informational externality of pioneering new export activities. Reform priorities shifted toward industrial policy and export promotion rather than more macro stabilization.

Brazil, by contrast, showed the opposite pattern. High interest rates, deep financial repression, and firms starved of credit despite strong investment opportunities. Here the diagnosis pointed toward financial constraints and fiscal dominance rather than deficits in human capital or infrastructure, which were also imperfect but not binding.

The comparison matters. Two middle-income Latin American countries, similar in many respects, required nearly opposite reform priorities. A generic checklist would have prescribed the same medicine to both and helped neither.

Takeaway

Comparative diagnosis reveals that the same symptom—slow growth—can have opposite causes in different economies. The right reform is context-specific by design.

The framework's greatest strength is its humility about universal prescriptions. It forces analysts to look at actual prices, quantities, and firm behavior rather than benchmarking against idealized institutional templates. It also provides a language for negotiating reform priorities with governments that have limited bandwidth.

But growth diagnostics has real limitations. The decision tree assumes constraints are separable and hierarchical, when in reality they often interact. Weak courts, thin credit markets, and unclear property rights may reinforce each other in ways that make identifying the binding constraint artificial. Removing one without the others can produce little change.

There is also a political economy blind spot. The framework asks what economic constraint binds, not what political coalition would tolerate its removal. A binding constraint that cannot be politically loosened is a curiosity, not a policy guide. Some critics argue the approach smuggles in technocratic assumptions about reform feasibility.

Finally, diagnostics can be misapplied when practitioners work backward from a preferred reform to a symptom that justifies it. Any framework that requires judgment is vulnerable to motivated reasoning. Used carefully, it disciplines thinking; used carelessly, it dresses up prior beliefs in analytical language.

Takeaway

A framework is only as good as the honesty of the person applying it. Diagnostics discipline analysis but cannot substitute for political judgment about what is actually reformable.

Growth diagnostics did not solve development. What it did was shift the conversation from universal prescriptions to context-specific analysis. That shift, modest as it sounds, has changed how many practitioners approach their work.

The framework works best when treated as a discipline rather than a doctrine. It forces you to look at evidence before recommending reform, and to justify why one constraint matters more than others. Those habits are valuable whether or not you accept every branch of the decision tree.

Perhaps the deepest lesson is that development strategy is fundamentally a question of prioritization under uncertainty. Diagnostics offers no guarantees—only a more honest way to think about where to start.