New growth theory — also called endogenous growth theory — is the third major wave in the history of economic growth thinking. Where classical and neoclassical models treated technological progress as something that arrives from outside the economy, new growth theory insists that technology and knowledge are deliberately created within it, by firms and individuals responding to economic incentives. For the first time, growth theory could tell policymakers not just that technology matters, but what they could actually do about it.
The Arc of Growth Theory: From Classical to Neoclassical to New
Economic thought advances incrementally, and growth theory is no exception. Classical economists — Smith, Malthus, Ricardo, Mill — identified the central tension between population growth and technological progress but worked with essentially static models, unable to trace how an economy evolves over time. Neoclassical theory, above all the Solow model of 1956, introduced dynamics: capital accumulates through investment, labor grows with population, and the interaction of these forces with diminishing returns shapes the economy’s long-run trajectory.
The Solow model was a decisive step forward. But it left a critical question unanswered. Economic growth, in Solow’s framework, ultimately depends on technological progress — and yet the model says nothing about where technological progress comes from or how it could be encouraged. It appears as an external input, handed down from outside the system, and the economy simply grows at whatever rate it is given.
This was the gap that new growth theory set out to fill.
The Black Box at the Heart of the Solow Model
In neoclassical growth theory, technological progress is captured by what economists call total factor productivity — the portion of output growth that cannot be attributed to increases in capital or labor. It is, in essence, a measure of everything we cannot explain. Solow himself described it as a residual: what is left over after accounting for the inputs.
The trouble is that a residual is not an explanation. The Solow model represents technological progress as the variable “A” — a placeholder, a label on an empty box. It tells us that A grows at some exogenous rate, but it cannot tell us why, how, or what determines that rate. For a policymaker asking how to accelerate growth, the model offers no handle to pull.
This is why the Solow model is classified as an exogenous growth theory. Because the engine of long-run growth — technological progress — is determined outside the model, the model cannot generate predictions about how policy choices affect the pace of growth. It is a theory of the consequences of growth, not a theory of its causes.
Endogenous and Exogenous: A Distinction That Changes Everything
The vocabulary of endogenous and exogenous variables is central to understanding what new growth theory achieves. An endogenous variable is one determined inside a model — it responds to the choices and incentives described by the model itself. An exogenous variable is determined outside — it is taken as given, a fact about the environment that the model does not explain.
In the Solow model, technology is exogenous. Firms and households make decisions about saving, investment, and labor supply, but none of those decisions affect the rate at which technology improves. Technology just happens, at whatever pace it happens, independently of what anyone inside the economy does.
New growth theory makes technology endogenous. Firms invest in research and development because they expect it to generate profits. Individuals accumulate education and skills because they expect higher earnings. These decisions, driven by ordinary economic incentives, generate the knowledge and innovation that fuel growth. The rate of technological progress is not a gift from outside the system — it is an outcome produced inside it, shaped by policy, institutions, and the incentives facing economic actors.
Romer and Lucas: Two Routes to Endogenous Growth
The intellectual foundations of new growth theory were laid in the mid-1980s, primarily by Paul Romer and Robert Lucas, working from different angles but arriving at a shared conclusion: knowledge and human capital are the ultimate engines of sustained growth.
In January 1985, at a lecture for doctoral students at the University of Chicago, Robert Lucas laid out the ideas that would become endogenous growth theory — material he was preparing for the Marshall Lectures at Cambridge University that summer. It was a historic moment in the development of economic thought.
Romer published his landmark paper, “Increasing Returns and Long-run Growth,” in 1986. Lucas followed with “On the Mechanics of Economic Development” in 1988. Together, these works redirected growth economics and earned both economists lasting recognition — Romer received the Nobel Prize in Economics in 2018.
Romer: Knowledge Does Not Diminish
The Solow model’s central mechanism is diminishing returns to capital: each additional unit of capital added to a fixed supply of labor produces less output than the one before it. This is what drives the economy toward a steady state and causes growth to slow.
Romer’s insight was that knowledge behaves differently. Individual firms’ investment in research generates knowledge, and through knowledge spillovers, that knowledge diffuses to other firms and raises productivity across the economy. Unlike a piece of machinery, a new idea is non-rival: it can be used by many firms simultaneously without being depleted. And unlike a physical input, knowledge tends to build on itself — each discovery makes the next one more tractable.
Romer’s endogenous growth theory demonstrated that technological progress can generate increasing returns to scale, and that a competitive market equilibrium is possible even under these conditions. This made it possible to measure the effect of technological innovation on economic growth, and it gave much stronger theoretical support to the argument that government investment in R&D and intellectual property institutions such as patent systems are essential for sustainable growth.
The policy implications are direct. If knowledge generates increasing returns and positive spillovers, private markets will tend to underinvest in research — because individual firms capture only part of the social return on their R&D spending. This creates a justification for public investment in basic research, support for universities and research institutions, and patent systems that allow innovators to capture enough of the returns to make innovation worthwhile.
Lucas: Human Capital as the Engine of Growth
Where Romer focused on firms’ investment in knowledge through formal R&D, Lucas emphasized the accumulation of human capital — the skills, competencies, and productive knowledge embodied in people. Education, on-the-job training, learning by doing: all of these build human capital and, crucially, generate external effects that benefit others.
By being in proximity to many people, individuals can acquire high-quality information and accumulate new knowledge, skills, and capabilities at almost no cost — a dynamic that, in Lucas’s analysis, makes cities a primary source of national wealth. When skilled people cluster together, ideas flow more freely, collaboration becomes more productive, and the rate of learning accelerates for everyone.
Lucas established endogenous growth theory in an economy where human capital accumulation takes place as a social activity, and this theory opened the possibility of unlimited economic growth. Because human capital generates external effects that do not diminish as the stock grows, an economy that invests steadily in education and skills can sustain rising productivity and living standards indefinitely — without hitting the ceiling that the Solow model predicted.
The Core Message: Invest in Knowledge
Taken together, Romer and Lucas transformed the policy conversation around growth. The message of neoclassical theory was essentially passive: invest, save, and wait for the economy to reach its steady state. The message of new growth theory is active: the rate of long-run growth is something an economy can influence, through the choices it makes about education, research, and the institutions that govern innovation.
R&D spending matters directly: it creates new knowledge and ideas, and those ideas fuel productivity growth. But the ecosystem around R&D matters as well. The quality of education systems, the openness of scientific communities, the strength of intellectual property protection, the availability of skilled labor, the density of industry clusters — all of these shape how rapidly knowledge accumulates and diffuses through an economy.
Accumulated knowledge, in this view, is not just an input to production. It is the source of the inputs themselves. Each generation of ideas makes the next generation possible. Once the research agenda opened up by Romer and Lucas took hold, economists began examining technology itself — how it develops, how it spreads across countries, what determines the level of technology in a given economy, what the relationship between R&D and technological advance actually looks like. The field of growth economics became substantially richer.
The Limits of New Growth Theory
No economic theory provides a complete account of how real economies work, and new growth theory is not exempt from this rule.
The neoclassical prediction that poor countries should grow faster than rich ones — and eventually catch up — was not borne out by the data. But new growth theory produces the opposite prediction: countries that can invest heavily in R&D and human capital should grow fastest. Yet advanced economies, despite their enormous research capacity, are typically the ones experiencing the slowest growth. As economies mature and the stock of accumulated knowledge grows, growth rates tend to decline rather than accelerate — a pattern the theory struggles to explain cleanly.
There is also a structural tension at the theory’s core. For firms to invest in innovation, they need to expect profits from it — specifically, returns above the competitive norm. But the standard neoclassical framework assumes perfect competition, which drives profits to zero. New growth theory, in practice, requires imperfect competition: firms need market power, at least temporarily, to recover their R&D costs. This is why Romer’s models feature monopolistic competition, where innovators earn temporary quasi-rents before competitors catch up. The assumption is realistic, but it marks a significant departure from the neoclassical foundations that the rest of the framework inherits.
These tensions do not invalidate new growth theory. They define the frontier — the set of questions that remain open and that subsequent research continues to address.
What the Evolution of Growth Theory Tells Us
The progression from classical to neoclassical to new growth theory traces a single deepening question: what is technological progress, and where does it come from? Classical economists recognized that technology mattered. Neoclassical economists measured how much. New growth theorists asked how it is produced — and that question has reshaped both economics and economic policy.
The answers matter especially for sectors where knowledge and innovation are central to productivity. Agriculture is one of them. The dramatic gains in agricultural output per acre and per worker over the past century were not accidents: they resulted from investment in agricultural research, extension services, seed development, and the diffusion of new techniques across farms and regions. New growth theory provides the conceptual vocabulary for understanding why those investments were so consequential and why sustained commitment to agricultural R&D is essential for long-run food security.
The same logic applies to the AI transformation now reshaping economies across sectors. When a new AI capability is developed, it does not stay confined to the firm that built it — it spreads, inspires adaptations, and raises the productivity of everyone who learns to apply it. That is knowledge spillover in action. Whether in agriculture, manufacturing, or services, the economies that invest most seriously in the knowledge base of their industries are the ones best positioned to sustain growth over the long run.
References
- Paul Romer, “Increasing Returns and Long-run Growth,” Journal of Political Economy, 1986
- Paul Romer, “Endogenous Technological Change,” Journal of Political Economy, 1990
- Robert E. Lucas Jr., “On the Mechanics of Economic Development,” Journal of Monetary Economics, 1988
- KDI Economic Education & Information Center, “Solow: Building the Academic Foundation of Economic Growth” (eiec.kdi.re.kr)
- Korea Institute of Science and Technology, Technology Policy Research Institute (TePRI), “The Reality and Solutions for Appropriate Scale of Government R&D” (tepri.kist.re.kr)
- joohyeon.com, “Convergence Debate I — Romer and Lucas: Emphasizing Knowledge and Human Capital”
- Namu Wiki, “Endogenous Growth Theory”
- PenN Mike, “Implications of Paul Romer’s Endogenous Growth Theory” (pennmike.com)
For research-grounded analysis connecting economic growth theory to agriculture, food systems, and AI transformation in Korea and beyond, visit KAFI’s Economics & Growth Theory section.