Growth Theory

Growth theory is the branch of macroeconomics that asks why some economies produce vastly more per person than others, and why output per person rises over time.

7 frameworks, from 1776 to today · Macroeconomics · Economics

still a working view ended · each leads to its entry below

Overview

Growth theory is the branch of macroeconomics that asks why some economies produce vastly more per person than others, and why output per person rises over time. Its central object is aggregate output—the total goods and services an economy makes—and its central puzzle is that small differences in growth rates, compounded over decades, produce enormous differences in living standards1. An economy growing at 2 percent a year doubles its output in about 35 years; one growing at 7 percent doubles in a decade. Growth theory tries to explain what determines that rate: saving and investment, technology, population, institutions, and the way these interact. It is the part of economics most directly concerned with the long run, and its models are the ones governments use when they argue about how to raise a country's growth rate.

Classical Political Economy: Growth as Accumulation and Its Limits

The first systematic growth theory came from the classical political economists of the late eighteenth and early nineteenth centuries. Adam Smith's Wealth of Nations (1776) argued that growth came from the division of labour and capital accumulation, and that it was self-limiting: as capital accumulated, the return on it would fall. David Ricardo, in On the Principles of Political Economy and Taxation (1817), formalized a sharper version. Growth required investment, investment required profit, and profit was squeezed by rising rents as population pressed on scarce land. Ricardo's model predicted that an expanding economy would eventually reach a stationary state—a plateau where accumulation stopped because profits had been eaten by landlords.

The mechanism mattered because it made a political claim: the interests of landowners and of growth were opposed. Thomas Malthus added the population side, arguing that population would rise to consume any improvement in wages. These classical models were the first to treat growth as a system with feedback—accumulation raising output, output raising population, population lowering wages and profits—and the first to ask whether growth could continue indefinitely. Their answer, broadly, was no.

Harrod–Domar: Saving, Capital, and the Knife-Edge

The classical tradition faded, and growth theory was rebuilt in the mid-twentieth century around a simple accounting idea. Roy Harrod (1939) and Evsey Domar (1946), working separately, produced what became the Harrod–Domar model2. Its core is an identity: the growth rate of output equals the saving rate divided by the capital–output ratio—how much capital it takes to produce a unit of output. If a country saves 20 percent of income and needs 4 units of capital per unit of output, it grows at 5 percent.

The model's difficulty was stability2. Harrod and Domar assumed production used capital and labour in fixed proportions, with little substitution between them. In that world, the economy had a "warranted" growth rate at which saving and investment were consistent, but any deviation pushed it further away—a knife-edge. If growth ran slightly above the warranted rate, capital became scarce and the gap widened; slightly below, capital became glutted and the economy slid into stagnation. The model implied that capitalist growth was inherently fragile, a conclusion that fit the interwar experience of depression and that Keynesian economists found congenial. But it could not explain the sustained growth of the postwar decades, and its fixed proportions made it hard to reconcile with the fact that economies do substitute labour for machines.

Solow–Swan: Substitution, Diminishing Returns, and the Residual

Robert Solow (1956) and Trevor Swan (1956) independently fixed the knife-edge by allowing capital and labour to be substituted for one another. In the Solow–Swan model, output is produced by capital and labour with diminishing returns: each extra unit of capital adds less output than the one before1. This single change made the model stable. An economy with little capital has a high return on it, so it saves and invests and grows fast; as capital accumulates, returns fall, investment slows, and the economy converges to a steady state where capital per worker is constant and growth per worker stops.

The model's striking implication was conditional convergence: poor countries should grow faster than rich ones, catching up, provided they have similar saving rates and technology1. It also implied that in the long run, growth of output per person could not come from capital accumulation at all—only from technological progress, which the model treated as exogenous, arriving from outside1. When Solow tried to account for actual US growth, he found that most of it was not explained by capital or labour at all. The unexplained remainder, the Solow residual (later called total factor productivity), became the measure of technology's contribution. Solow's 1957 paper attributed roughly seven-eighths of US growth per hour to technical change, a result that reframed the field: growth was mostly about knowledge, not machines.

Optimal Growth: Ramsey, Cass, and Koopmans

Parallel to the descriptive Solow model, a normative tradition asked what a society should save. Frank Ramsey (1928) had posed the problem of choosing a consumption path that maximized welfare over time, but his formulation was difficult to apply. In the mid-1960s, David Cass (1965) and Tjalling Koopmans (1965) independently solved a version using a discounted welfare criterion—future consumption counts less than present consumption—and a utility function that values each additional unit of consumption less as consumption rises. The resulting Ramsey–Cass–Koopmans model describes an economy that saves and consumes optimally, with a steady state determined by the discount rate and the productivity of capital.

This framework became the workhorse of dynamic macroeconomics. Its apparatus—maximizing a discounted sum of utility subject to a capital accumulation constraint—was later borrowed by real business cycle theory and by modern heterogeneous-agent models. It also made precise the trade-off at the heart of growth policy: consuming more now means investing less and consuming less later, and the "right" balance depends on how much a society discounts the future.

Endogenous Growth: Bringing Technology Inside the Model

The Solow model left technology unexplained, and by the 1980s economists found this unsatisfying1. If growth was mostly technical change, a theory that assumed technical change was the weakest possible foundation. Paul Romer's 1986 paper and Robert Lucas's 1988 paper launched the endogenous growth literature, which tried to explain the growth rate from within the model.

Romer's key move was to treat knowledge as a special kind of capital. Unlike machines, ideas can be used by many firms at once without being used up—they are non-rival. A firm that discovers a better design can license it to others, and the cost of the discovery does not rise proportionally with the number of users. This generates increasing returns to scale: a larger economy can support more research and grow faster. Romer's 1990 paper, with a explicit research sector, made the growth rate depend on the number of researchers and on the strength of intellectual property protection. Lucas emphasized human capital—education and skills—as a source of growth that could spill over between workers.

The policy implication was large: growth could be raised by subsidizing research and education, because the private return to discovering ideas is lower than the social return. The models also predicted scale effects—bigger countries or bigger research sectors should grow faster—which sat awkwardly with the data, since growth rates in the twentieth century did not obviously rise with population. Later work, notably by Charles Jones, refined the models to remove the scale effect by assuming ideas get harder to find as they are discovered.

Schumpeterian Growth: Creative Destruction Formalized

A distinct endogenous tradition drew on Joseph Schumpeter, who had argued in Capitalism, Socialism and Democracy (1942) that capitalism advanced through creative destruction: new products and processes continually displaced old ones, and the economy's dynamism came from that churn rather than from smooth accumulation. Philippe Aghion and Peter Howitt (1992) formalized this as a model of quality-ladder growth, in which each innovation raises the quality of a product and temporarily gives the innovator market power, until the next innovation destroys it.

The Schumpeterian framework made a different prediction from Romer's: growth depends not just on how much research is done but on the turnover of incumbent firms. It also generated a subtle result about competition—some competition encourages innovation by raising the reward to winning, but too much destroys the rents that fund research. This "inverted-U" relationship between competition and innovation became a testable hypothesis and a bridge between growth theory and industrial organization.

Institutions and the Empirical Turn

From the 1990s, the question shifted from the mechanics of growth to its causes across countries. The Solow model's convergence prediction was tested and found wanting in its simple form: poor countries were not, on average, catching up1. The field turned to cross-country growth regressions, and to the quality of institutions—property rights, contract enforcement, constraints on government—as the deep determinant. Daron Acemoglu, Simon Johnson, and James Robinson (2001) used the mortality rates of European settlers in former colonies as an instrument for institutional quality, arguing that where settlers faced high disease mortality, colonial powers set up extractive institutions that persisted and retarded growth1. The result was influential and contested, and it pushed growth theory toward history and political economy.

The empirical turn also brought new data and methods: the Penn World Table assembled comparable national accounts across countries; the growth accounting tradition decomposed growth into capital, labour, and productivity; and later work used firm-level data and natural experiments. The field increasingly asked not just what the model says but which model the world resembles.

The Present: Convergence, Divergence, and the Frontier

Today growth theory is a set of overlapping frameworks rather than a single model. The Solow–Swan model remains the baseline for thinking about capital accumulation and convergence, and the Ramsey–Cass–Koopmans apparatus is the standard dynamic framework34. Endogenous growth models in the Romer and Schumpeterian traditions explain why technology and institutions matter and why policy can affect the growth rate. Empirical work on institutions, human capital, and firm dynamics tests and refines these models.

The open questions are large. Why do so many poor countries fail to converge, and why do some grow rapidly for a while and then stall? How much of the growth slowdown in rich countries since the 1970s reflects slower idea production, and how much reflects measurement? Does automation and artificial intelligence change the returns to capital and labour in ways the models do not capture? And the oldest question—whether growth can continue indefinitely given resource limits—remains unresolved, now framed in terms of climate and energy rather than land and rents. Growth theory has never settled these questions; it has given economists the tools to argue about them precisely.

References

  1. Economic growth (Long-term growth, Solow–Swan model, Endogenous growth theory, Political institutions), Wikipedia
  2. Harrod–Domar model, Wikipedia
  3. Solow–Swan model, Wikipedia
  4. Ramsey–Cass–Koopmans model, Wikipedia

The frameworks of Growth Theory

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1776-1870

    Classical economics

    What limits the growth of wealth?

    Fly to its tower in Nousfarer

  2. 1939-1960

    Harrod–Domar Model

    How fast does saving make output grow?

    Fly to its tower in Nousfarer

  3. 1956-Present

    Solow–Swan model

    What determines output per worker over time?

    Fly to its tower in Nousfarer

  4. 1965-Present
  5. 1986-Present

    Endogenous Growth Theory

    What makes technology grow from inside?

    About Endogenous Growth Theory
    Why do some economies enjoy rising living standards for decades, while others seem to stall after an initial burst of growth? By the 1980s, the dominant Solow-Swan model had a frustrating answer: long-run growth depends on technological progress that the model itself treats as a mysterious external force—the famous 'Solow residual.' Endogenous growth theory emerged to bring that engine inside the model. It asks: what if the rate of technological progress is itself the outcome of deliberate decisions by firms and individuals? The theory's central move is to treat knowledge, innovation, and human capital as produced inputs that do not suffer from the diminishing returns that limit physical capital accumulation. This shift opened a new way to think about why policy—subsidies for education, patent protection, R&D tax credits—can permanently raise a country's growth rate, not just its level of output.
    Core curriculum. A first course in endogenous growth theory typically builds from a simple workhorse model to richer accounts of how innovation happens. The sequence runs as follows:
    1. AK model – The simplest endogenous growth model, where aggregate output is proportional to the capital stock (Y=AKY = AK). Because capital includes knowledge, it does not face diminishing returns, so the economy can grow indefinitely without an exogenous技术进步.
    2. Constant returns to capital – The key aggregate property that distinguishes endogenous from neoclassical models: at the economy-wide level, the marginal product of capital does not fall to zero, allowing sustained growth.
    3. Human capital – Education, training, and health improve workers' productivity. Models treat human capital as a produced factor that can be accumulated, and its accumulation does not run into diminishing returns at the aggregate level.
    4. Learning by doing – Workers and firms become more productive through experience. This is a form of knowledge creation that happens as a by-product of production itself.
    5. Knowledge spillovers – When one firm invests in new knowledge, other firms benefit without paying for it. These positive externalities are a central mechanism for generating aggregate constant returns to capital.
    6. Externalities in growth – Because knowledge spills over, the social return to investment in knowledge exceeds the private return. This creates a rationale for government intervention.
    7. Endogenous technological change – The rate of innovation is itself determined by economic incentives, not by an exogenous time trend. Firms decide how much to invest in R&D based on expected profits.
    8. Innovation incentives – The profit motive drives firms to create new products or processes. Monopoly power (temporary) is the reward that makes innovation worthwhile.
    9. Monopolistic competition – Many endogenous growth models depart from perfect competition. Firms have some market power over their unique products, which allows them to earn the profits that fund R&D.
    10. R&D-based growth – A class of models in which firms deliberately invest in research and development to create new technology. The growth rate depends on the share of resources devoted to R&D.
    11. Product variety expansion – One way innovation occurs: firms create new types of intermediate or final goods. Growth comes from an ever-widening set of products.
    12. Quality ladders – Another innovation channel: firms improve the quality of existing products. Growth comes from a sequence of quality improvements that make older products obsolete (creative destruction).
    13. Scale effects – An early prediction of many endogenous growth models: larger economies (in terms of population or R&D workers) should grow faster because they have more people to generate ideas. This prediction proved controversial and led to later refinements.
    14. Patent protection – A policy tool that strengthens innovation incentives by granting temporary monopoly rights. The optimal patent length balances the reward for innovation against the static loss from monopoly pricing.
    15. Education subsidies – A policy lever that encourages human capital accumulation, raising the long-run growth rate by making it cheaper for individuals to invest in skills.
    Intellectual history. Endogenous growth theory crystallized in the mid-1980s as a direct response to the limitations of the Solow-Swan neoclassical model. In the Solow-Swan framework, long-run growth is driven by exogenous technological progress—a force that the model takes as given. By the early 1980s, growth economists had grown uneasy with this 'black box' treatment of technology. Paul Romer's 1986 paper 'Increasing Returns and Long-Run Growth' is widely regarded as the founding text. Romer showed that if knowledge is a non-rival good (one person's use does not prevent another's) and if firms can capture only part of the benefits of their research, then the economy can exhibit increasing returns to scale at the aggregate level, generating sustained growth. Robert Lucas's 1988 paper 'On the Mechanics of Economic Development' provided a parallel foundation by focusing on human capital accumulation and its spillover effects. Lucas showed that if workers invest in skills and those skills raise the productivity of everyone around them, the economy can grow without bound. These two papers launched the two main strands of the endogenous growth literature: the R&D-based approach (Romer) and the human-capital approach (Lucas).
    A second wave of models in the early 1990s, often labeled Schumpeterian growth theory, refined the innovation mechanism. Philippe Aghion and Peter Howitt's 1992 paper 'A Model of Growth Through Creative Destruction' introduced quality ladders, where each new innovation makes the previous one obsolete. This framework explicitly modeled the role of monopoly power and the possibility that too much or too little competition could affect growth. Gene Grossman and Elhanan Helpman's 1991 book Innovation and Growth in the Global Economy synthesized product-variety and quality-ladder approaches and extended them to open economies. These later developments are sometimes classified as a separate framework (Schumpeterian growth theory) but are also taught as advanced topics within endogenous growth theory.
    The theory's relation to its predecessors is one of supplementation rather than clean replacement. The Solow-Swan model remains the standard starting point for teaching growth accounting and the role of capital accumulation. Endogenous growth theory added the 'how' and 'why' of technological change. It coexists with optimal growth theory (the Ramsey-Cass-Koopmans framework), which provides the intertemporal optimization tools that many endogenous growth models use. The Cambridge growth theory, with its focus on income distribution and effective demand, addresses a different set of questions and has not been absorbed into the endogenous growth tradition. Unified growth theory, which emerged around 2000, attempts to explain the transition from Malthusian stagnation to modern growth by incorporating endogenous fertility and human capital—a project that builds directly on the insights of Lucas and Romer.
    Key insights. First, endogenous growth theory shifted the policy conversation: if growth is endogenous, then subsidies for education, R&D tax credits, and patent protection are not just efficiency measures but potential drivers of the long-run growth rate itself. Second, the theory provided a framework for understanding why convergence (poor countries catching up to rich ones) is not automatic—if knowledge spillovers are local rather than global, or if human capital is scarce, differences in growth rates can persist indefinitely. Third, by placing innovation at the center, the theory connected growth economics to industrial organization and the economics of knowledge, creating a rich interdisciplinary dialogue.
    Common pitfalls. A frequent misunderstanding is that endogenous growth theory 'proves' that government intervention always raises growth. In fact, the models highlight trade-offs: patent protection encourages innovation but creates monopoly distortions; education subsidies must be financed by distortionary taxes. The theory provides a framework for analyzing these trade-offs, not a simple policy prescription. Another pitfall is conflating the AK model with the whole theory. The AK model is a pedagogical starting point, but most of the theory's insights come from models with explicit microfoundations for innovation, human capital, and market structure.
    The 13 concepts its tower teaches
    • AK model
    • Constant returns to capital
    • Endogenous technological change
    • Externalities in growth
    • Human capital
    • Innovation incentives
    • Knowledge spillovers
    • Monopolistic competition
    • Non-rival knowledge
    • Product variety expansion
    • Quality ladders
    • R&D-based growth
    • Scale effects

    Fly to its tower in Nousfarer

  6. 1990-Present
  7. 1992-Present

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Classical economicsHarrod–Domar ModelSolow–Swan modelRamsey–Cass–Koopmans modelEndogenous Growth TheoryInstitutional and Cross-Country Growt…Schumpeterian Growth Theory

Beyond the field

Where Growth Theory's ideas came from in other fields, and what they led to there: the strongest 12 of 19 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Growth Theory is a planet in Nousfarer

Its city has a tower for each of these 7 frameworks. Land, walk in, and climb one concept at a time.

Land on Growth Theory

Related fields

Fields its ideas are linked to

More in Macroeconomics

See where Growth Theory sits on the Map of Human Knowledge