The general argument that “institutions matter” for growth, while correct, is incomplete. A more refined perspective, first articulated in the historical work of Alexander Gerschenkron (1962), suggests that the institutions most conducive to growth change as an economy develops. Arrangements that are highly effective for a technologically backward economy may become obstacles to growth as that economy approaches the frontier. This section formalizes this concept of “appropriate institutions” and explores its profound implications for why some countries succeed in catching up while others get stuck in the middle.

The Fundamental Trade-Off: Imitation vs. Innovation

Productivity growth stems from two distinct activities, each favored by a different set of institutional arrangements:

• Imitation: The process of adopting and implementing technologies developed elsewhere (at the world frontier). This is the primary engine of growth for countries far from the frontier.
• Frontier Innovation: The process of creating entirely new technologies and products. This becomes increasingly important as a country approaches the frontier and the potential for catch-up growth through imitation diminishes.

These activities correspond to two distinct institutional strategies: 1) An imitation-enhancing strategy, which favors long-term investment, managerial experience, and large incumbents, is optimal far from the frontier where adopting existing technologies yields the highest marginal return. 2) An innovation-enhancing strategy, which promotes competition, firm entry, and stringent talent selection, becomes essential closer to the frontier where growth depends on creating novel technologies.

The Growth-Maximizing Institutional Path

The model reveals that the optimal institutional strategy is state-contingent, hinging on a critical threshold of technological development. There exists a proximity to the frontier, denoted â, that marks an inflection point.

• For a country below this threshold (a_t-1 < â), the growth-maximizing strategy is to adopt imitation-enhancing institutions.
• For a country that develops past this threshold (a_t-1 > â), sustaining high growth requires a switch to innovation-enhancing institutions.

A country that successfully navigates this institutional transition will maintain a high growth rate and ultimately converge to the world technology frontier.

The Non-Convergence Trap

A critical risk arises when a country fails to make this institutional switch. A country may become stuck in a “sclerotic equilibrium,” maintaining its imitation-enhancing institutions for too long. These institutions, which were once the engine of a successful catch-up phase, become a drag on growth as the economy matures.

This institutional inertia leads to a non-convergence trap. As the country approaches the frontier, its growth rate slows dramatically because its institutions are ill-suited for the frontier innovation that is now required. The growth equation for the economy (describing proximity a_t as a function of last period’s proximity a_t-1) is represented by a concave curve that lies below the 45-degree line for high levels of proximity. This curve intersects the 45-degree line at a point a_Trap that is strictly less than 1 (full convergence). This intersection represents a stable equilibrium from which the economy cannot escape, thus trapping it below the frontier. The country is trapped, not because its institutions are inherently “bad,” but because they have become inappropriate for its new economic reality.

Empirical Grounding for the Theory

This theoretical framework is supported by compelling empirical evidence.

• Research by Acemoglu, Aghion, and Zilibotti (2006) finds that policies promoting openness and low barriers to entry (innovation-enhancing policies) are more strongly and positively correlated with growth in countries that are already closer to the U.S. productivity frontier.
• Evidence on education reveals a similar pattern. Investment in higher education (which supports frontier R&D) is more growth-enhancing for countries near the frontier, whereas investment in primary and secondary education (which builds the absorptive capacity for imitation) is more critical for countries further behind.

The most significant barrier to long-run convergence may not be a lack of good institutions, but rather the persistence of institutions that were once successful but have become inappropriate for a country’s new stage of development. This institutional inertia represents a formidable challenge on the path to prosperity.