Abstract: Robert Lucas's vision of global convergence emphasized the diffusion of modern knowledge and industrial capability from frontier economies to latecomers. Once modern growth had emerged somewhere,poorer countries could grow rapidly by adopting technologies,organizational forms,and production methods already developed at the frontier. Under this benchmark,world income differences would eventually narrow as more countries joined the process of modern economic growth. This paper argues that the Lucasian mechanism remains fundamentally correct but incomplete because it treats the international diffusion environment as exogenous. The central distinction introduced here is between income convergence and power convergence. Standard growth theory focuses primarily on relative per capita income. Yet international order responds not only to living standards,but also to aggregate economic scale. For a populous follower,aggregate power can converge long before income converges. If a follower has four times the population of the frontier country,aggregate parity emerges when its per capita income is only one quarter of the frontier level. Such an economy may still view itself as developing in terms of domestic productivity and living standards,while simultaneously becoming systemically consequential from the perspective of the frontier country. This arithmetic changes the political economy of diffusion. When a large follower is still relatively small in aggregate terms,openness is mutually beneficial:it expands markets,deepens specialization,strengthens supply chains,and reinforces a frontier-centered international order. However,as the follower's aggregate scale rises,the same openness increasingly generates strategic costs for the frontier country. Trade,foreign investment,scientific collaboration,technology transfer,financial integration,and production networks may accelerate the rise of a potential systemic rival. The paper therefore endogenizes the openness of the international diffusion environment. Rather than treating diffusion as a fixed background condition,the framework models openness as a strategic choice made by a forward-looking frontier country. As the follower approaches aggregate-power parity,the frontier country rationally reduces openness,especially in strategically sensitive sectors. The result is what the paper calls endogenous diffusion congestion:the very success of catch-up weakens the openness of the international order that made catch-up possible. The mechanism produces several broader implications. First,the congestion generated by a large follower spills over to other developing economies. Smaller followers may not themselves pose strategic threats,but they depend on the same frontier-centered channels of diffusion. As openness contracts in response to the rise of a large follower,these countries inherit a more restricted global learning environment. Second,the rise of a large follower may simultaneously create alternative channels of diffusion. A large follower that has recently industrialized can become an “intermediate frontier”, supplying development-relevant technologies,infrastructure,manufacturing systems,and organizational capabilities to poorer economies. The future of global convergence may therefore depend not only on whether the old frontier-centered order remains open,but also on whether multiple diffusion systems can coexist and remain interoperable. Third,the paper argues that the world economy is likely moving away from frictionless globalization toward a world of selective openness. Globalization does not necessarily disappear; rather,openness becomes differentiated across sectors,technologies,and political relationships. Some channels remain open,while others become increasingly restricted,politicized,or fragmented. The framework contributes to three strands of literatures simultaneously. It extends the growth and convergence literature by making the diffusion parameter endogenous to international order. It contributes to international political economy by formalizing how aggregate-power convergence alters the incentives of leading states to sustain openness. It also contributes to the fragmentation literature by explaining why restrictions emerge specifically when large followers approach systemic significance. The paper's broader implication is that the central challenge of the twenty-first century is no longer simply how to achieve convergence,but how to govern convergence once it becomes systemically consequential. The future of global development will depend not only on domestic absorptive capacity,but also on whether the international system can preserve broad channels of knowledge diffusion under conditions of growing aggregate-power competition.