Fact-checked by Grok 2 weeks ago

Endogenous growth theory

Endogenous growth theory constitutes a body of macroeconomic models positing that long-run economic growth emerges from internal economic processes, notably investments in human capital accumulation and innovation-driven technological progress, rather than exogenous factors like population growth or unexplained technical change.^[1]^[2] Pioneered in the late 1980s by economists Robert Lucas and Paul Romer, the framework endogenizes the growth rate by modeling knowledge creation as a deliberate outcome of private and public decisions, such as research expenditures and education, which generate non-rivalrous ideas that spill over to enhance productivity economy-wide.^[3]^[4] Central to these models is the relaxation of neoclassical assumptions of constant or diminishing returns to reproducible factors, often via linear production functions like Y = AK where output grows proportionally with accumulable knowledge-augmented capital (A), permitting perpetual growth without convergence to a steady state dictated by exogenous parameters.^[5] This shift underscores the potential for government policies—targeting R&D subsidies, intellectual property rights, and schooling—to influence steady-state growth rates, contrasting with Solow-style models where such interventions affect only transitional dynamics.^[5]^[6] While theoretically appealing for rationalizing cross-country growth divergences and the observed role of ideas in historical prosperity, empirical validation remains contested, with evidence supporting human capital's growth effects but limited confirmation of strong scale dependencies or innovation externalities predicted by early formulations.^[6]^[7]

Historical Development

Origins in Response to Neoclassical Limitations

The neoclassical growth model, as formalized by Robert Solow in 1956 and Trevor Swan in 1956, posits that long-run per capita income growth is determined solely by exogenous technological progress, with capital accumulation subject to diminishing marginal returns leading economies toward a steady-state equilibrium where growth ceases absent external technological inputs.^[8] This framework implied conditional convergence, whereby poorer economies grow faster than richer ones until they reach similar steady-state income levels, conditional on similar savings rates and population growth.^[9] However, empirical observations from the postwar period, including persistent income divergences among OECD countries and limited evidence of rapid convergence in developing nations, challenged these predictions, as documented in studies showing low or negative correlations between initial income levels and subsequent growth rates.^[8] A key limitation was the "black box" treatment of technological change as an unexplained residual, often labeled the Solow residual, which accounted for the bulk of observed growth but lacked microfoundations linking it to deliberate economic agents or incentives.^[9] Neoclassical models thus failed to explain why technological progress itself might accelerate or vary systematically across economies, rendering policy interventions—such as investments in education or R&D—ineffective for sustaining long-run growth rates in theoretical terms.^[8] This exogenous assumption also clashed with evidence of increasing returns to scale in knowledge production, as highlighted in early critiques noting that ideas, unlike physical capital, exhibit partial non-rivalry and potential for spillovers.^[10] Endogenous growth theory emerged in the mid-1980s as a direct response, seeking to internalize technological progress within the model by modeling it as the outcome of purposeful investments in human capital, research, and innovation, thereby allowing for sustained growth without relying on external shocks.^[8] Pioneering works, such as Paul Romer's 1986 paper on increasing returns and long-run growth, argued that scale effects from knowledge accumulation could generate constant or increasing returns to reproducible factors, overturning the neoclassical steady-state constraint.^[11] Similarly, Robert Lucas's 1988 contribution emphasized human capital externalities as an endogenous driver, addressing the convergence puzzle by permitting persistent growth differentials tied to internal accumulation processes rather than mere catch-up dynamics.^[9] These developments reflected a broader empirical push from the convergence controversy, where data on cross-country growth variances underscored the inadequacy of exogenous explanations.^[8]

Key Contributors and Milestones

The development of endogenous growth theory in the late 1980s marked a shift from exogenous technological progress in neoclassical models to internal economic mechanisms capable of sustaining long-run growth without diminishing returns. This framework addressed limitations in Solow-Swan models by endogenizing factors like knowledge creation and human capital investment, allowing growth rates to depend on policy and incentives rather than external shocks.^[10] Robert E. Lucas Jr. provided a foundational impetus with his 1988 paper "On the Mechanics of Economic Development," which emphasized human capital externalities and learning-by-doing as drivers of persistent cross-country growth differences, challenging convergence predictions of neoclassical theory.^[12] Lucas's model integrated optimizing agents with endogenous skill accumulation, showing how average human capital influences marginal productivity, thus permitting scale effects and non-convergent growth paths.^[13] Paul Romer built on these ideas, formalizing endogenous technological change. In his 1986 paper "Increasing Returns and Long-Run Growth," Romer demonstrated that production processes exhibiting increasing returns to scale—particularly through non-rivalrous knowledge inputs—could generate sustained per capita growth without relying on exogenous factors.^[14] His seminal 1990 contribution, "Endogenous Technological Change," introduced a microfounded model where research and development (R&D) by profit-maximizing firms expands the variety of intermediate goods under monopolistic competition, linking innovation incentives to growth rates and policy variables like subsidies or intellectual property protection.^[15] Romer's work earned him the 2018 Nobel Prize in Economic Sciences for integrating technological innovation into long-run macroeconomic analysis.^[16] Subsequent milestones included Sergio Rebelo's 1991 AK model, which simplified endogenous growth via linear production functions yielding constant returns at the aggregate level, reinforcing the theory's tractability for policy analysis.^[17] In the 1990s, Philippe Aghion and Peter Howitt developed Schumpeterian variants incorporating creative destruction, where new innovations displace old ones, further endogenizing growth through quality-improving R&D races. These advancements collectively established endogenous growth as a paradigm emphasizing internal accumulation over exogenous shocks.^[18]

Core Concepts

Internal Factors Driving Sustained Growth

Endogenous growth theory posits that long-term economic expansion arises from internal mechanisms, including deliberate investments in human capital and research and development (R&D), which generate productivity gains through non-diminishing returns to knowledge accumulation.^[19] Unlike neoclassical models where capital accumulation faces diminishing marginal returns leading to steady-state convergence, endogenous approaches emphasize that internal factors like innovation create constant or increasing returns, enabling perpetual growth rates determined by policy and economic choices rather than exogenous shocks.^[20] These factors operate via externalities, such as knowledge spillovers, where the social returns to private investments exceed private returns, justifying sustained growth without relying on population expansion or unexplained technological progress.^[21] A primary internal driver is human capital accumulation, as formalized by Robert Lucas in 1988, where individuals allocate time between production and learning, with the latter enhancing productivity through positive externalities from the economy's average human capital stock.^[22] In Lucas's model, output per worker grows endogenously at a rate proportional to the fraction of time devoted to education, implying that policies boosting human capital formation—such as subsidies for schooling—can raise steady-state growth without diminishing returns, as human capital augments both current output and future learning capacity.^[4] Empirical calibration of this mechanism, using data from developing economies, shows that variations in human capital investment explain persistent cross-country growth differences, with higher education shares correlating to growth rates exceeding 2% annually in high-investment regimes.^[12] Innovation through R&D represents another core internal factor, particularly in Paul Romer's 1990 framework, where profit-maximizing firms invest in designing new intermediate goods varieties, drawing on a stock of non-rivalrous ideas that expand with cumulative research effort.^[19] This process yields sustained growth because each new idea partially spills over to all producers, increasing aggregate productivity without rivalry costs, such that the growth rate equals the marginal product of research labor times the research intensity, often modeled as g = \delta L_A / L, where L_A is researchers and \delta captures idea productivity.^[23] For instance, calibrations to U.S. data from 1950–1985 indicate that R&D intensity around 2.5% of GDP sustains growth rates near 2%, with scale effects amplifying returns in larger economies due to greater research labor pools.^[24] These internal factors interact to underpin sustained growth: human capital fosters better R&D outcomes, while innovations enhance human capital returns, creating feedback loops resilient to shocks. Policy implications include incentives for R&D monopolies to internalize spillovers via patents, as evidenced by post-1980 U.S. patent reforms correlating with accelerated total factor productivity growth from 1.1% to 1.5% per year.^[21] However, challenges arise from potential underinvestment due to externalities, with estimates suggesting social returns to R&D at 50–100% versus private returns of 20–30%, underscoring the need for targeted subsidies to maximize internal growth drivers.^[20]

Knowledge, Human Capital, and Innovation as Endogenous Elements

In endogenous growth theory, human capital—encompassing skills, education, and knowledge acquired by individuals—serves as a core endogenous driver of long-term economic expansion, as individuals can invest time and resources to augment it, generating positive externalities that prevent diminishing returns to scale. Robert Lucas's 1988 model posits that agents divide their labor between goods production and human capital accumulation, where the latter enhances overall productivity through spillovers, such that aggregate output grows perpetually if the fraction of time devoted to learning exceeds a threshold determined by production parameters.^[4] This framework implies that policies boosting education or training can sustain higher growth rates, contrasting with exogenous models where such investments merely shift levels without altering steady-state growth. Empirical cross-country regressions, such as those linking secondary schooling attainment to per capita GDP growth from 1960–1985, support a positive coefficient on human capital stock, though causality remains debated due to reverse causation from growth to education.^[7] Knowledge accumulation further endogenizes growth by treating ideas as non-rivalrous goods that, once produced, can be reused indefinitely across firms without depletion, fostering increasing returns at the economy-wide level despite private diminishing returns. Paul Romer's 1990 formulation emphasizes knowledge spillovers from research activities, where the stock of existing ideas lowers the cost of new discoveries, enabling sustained technological progress without relying on exogenous shocks.^[19] In this view, the partial excludability of knowledge—via patents or secrecy—creates incentives for private investment, but social returns exceed private ones due to uncompensated diffusion, justifying public subsidies for R&D to internalize these externalities. Data from U.S. patent applications, which rose from about 50,000 in 1980 to over 100,000 annually by 2000, correlate with productivity accelerations in knowledge-intensive sectors, providing suggestive evidence for spillover-driven growth.^[25] Innovation, modeled as the deliberate creation of new technologies or product varieties through R&D, integrates human capital and knowledge to propel endogenous expansion, often under monopolistic competition where firms profit from temporary exclusivity while expanding the production possibilities frontier. Romer's expanding variety approach (1990) demonstrates how R&D investment, proportional to the existing knowledge base, yields exponential growth in intermediate goods, with the growth rate given by \gamma = \frac{\lambda \bar{A} L_A}{\bar{n}}, where \lambda >1 captures research productivity amplified by spillovers, L_A is labor in research, and \bar{n} the number of varieties.^[19] Variants incorporating Schumpeterian creative destruction link innovation to quality improvements that displace obsolete technologies, with human capital enhancing the selection of high-impact ideas. Panel data analyses across OECD countries from 1970–2010 find that R&D intensity (as a share of GDP, averaging 2–3%) positively predicts total factor productivity growth, particularly when interacted with tertiary education levels, though scale effects—where larger economies innovate more per capita—face criticism for overstating small-country growth potential.^[26] These mechanisms collectively imply policy levers like intellectual property enforcement or immigration of skilled workers can elevate steady-state growth, grounded in causal channels from internal accumulation rather than serendipitous external advances.^[27]

Formal Models

The AK Model and Constant Returns

The AK model constitutes one of the simplest endogenous growth frameworks, featuring a production function Y = AK, where output Y derives linearly from a constant productivity parameter A and broadly defined capital stock K, which may incorporate physical capital, human capital, or knowledge as accumulable factors.^[28] This linear specification arises as a limiting case of the Cobb-Douglas production function Y = A K^{\alpha} L^{1-\alpha} when the capital elasticity \alpha = 1, yielding constant returns to scale solely to the reproducible factor K, with labor L absent or normalized.^[29] Consequently, the marginal product of capital remains constant at A, avoiding the diminishing returns that constrain long-run growth in exogenous models like Solow's.^[30] Under standard assumptions of competitive markets and rational agents, households maximize intertemporal utility \int_{0}^{\infty} e^{-\rho t} u(c_t) dt, where consumption c_t follows from savings decisions, while capital evolves via \dot{K} = s Y - \delta K, with savings rate s, depreciation \delta, and no population growth for simplicity.^[31] In balanced growth equilibrium, per capita output and capital grow at rate g = sA - \delta - \rho, endogenously determined by the savings rate and productivity rather than an exogenous technological progress parameter.^[28] This setup implies that policies elevating s or A—such as tax reductions on capital income—permanently raise the growth rate, contrasting with neoclassical predictions of level effects only.^[32] The constant returns to accumulable factors underpin sustained growth without scale exhaustion, as replication of K proportionally expands output indefinitely, fostering divergence across economies with differing savings propensities.^[33] Rebelo (1991) formalized this in analyzing long-run policy impacts, demonstrating how fiscal distortions, like capital taxes at rate \tau, reduce effective A to A(1-\tau), thereby lowering g proportionally.^[28] Empirical interpretations often broaden K to include human capital accumulation, aligning the model with observations of persistent growth differentials, though critics note its knife-edge reliance on exact linearity, where slight sublinearity (\alpha < 1) restores diminishing returns and convergence.^[30]^[34]

Romer's Expanding Variety Model

Paul Romer's expanding variety model, introduced in his 1990 paper "Endogenous Technological Change," explains sustained economic growth through the endogenous invention of new varieties of intermediate inputs, driven by profit-motivated research and development (R&D).^[19] The model features three sectors: a competitive final goods sector that integrates labor and a continuum of specialized intermediate goods; a monopolistically competitive intermediate goods sector where each variety is produced by a single firm under patent protection; and an R&D sector where agents create new designs using labor or human capital and the existing stock of knowledge.^[19]^[35] This structure endogenizes technological progress by linking the growth of knowledge (measured by the number of varieties, A) to economic incentives, contrasting with exogenous models where innovation occurs outside agent optimization.^[19] In the final goods sector, output Y is produced competitively using non-specialized labor L_Y and quantities x_i of each intermediate variety i from 0 to A_t, following the production function Y_t = L^1-α_Y ∫^{_A^t}₀ x_it^α di, where 0 < α < 1 reflects diminishing marginal returns to each input but increasing productivity from greater variety.^[35]^[36] Intermediate goods are produced using capital (or forgone output), with each monopolist facing constant marginal cost normalized to the rental rate r, leading to pricing p_i = r / α and a markup of 1/α over marginal cost due to market power.^[19]^[35] In symmetric equilibrium, all x_i equal some \bar{x}, yielding effective Y_t ≈ L^1-α_Y A^α_t *\bar{x}^α, where variety expansion A_t augments productivity without diminishing returns at the aggregate level.^[36] The R&D sector generates new varieties at rate ḣA_t = δ L_A A_t, where L_A is labor allocated to research and δ > 0 is a productivity parameter; the linearity in the existing knowledge stock A_t captures "standing on the shoulders of giants," enabling accelerating invention potential.^[19]^[35] Key assumptions include the non-rivalry of ideas—blueprints can be used costlessly by all without depletion—and partial rivalry in their physical production, which sustains monopoly rents to incentivize R&D.^[19]^[35] Free entry into design creation equates the expected value of a new invention (discounted monopoly profits) to R&D costs (wages w), ensuring zero net profits in equilibrium.^[19] Along the balanced growth path, per capita output grows at constant rate g ≈ ḣA*_t/A_t = δ (H_A/total human capital or fraction of labor in R&D), endogenously determined by resource allocation between production and research, influenced by parameters like the discount rate ρ and elasticity of intertemporal substitution.^[19]^[35] A prominent implication is the scale effect: larger population or human capital stock L or H raises g^* proportionally, as it expands the market size for intermediates (boosting invention value) and R&D labor supply, interpreting the model as describing world-level growth dynamics.^[35]^[36] The equilibrium features underinvestment in R&D due to positive externalities from knowledge spillovers and monopoly distortions, suggesting potential welfare gains from subsidies.^[19]

Human Capital and Schumpeterian Variants

In human capital-based endogenous growth models, sustained economic expansion arises from the endogenous accumulation of knowledge and skills embodied in workers, rather than exogenous technological progress. Robert Lucas's 1988 model posits that individuals divide their time between producing goods and investing in human capital formation, where human capital h enters the aggregate production function multiplicatively as Y_t = K_t^\alpha (h_t u_t L)^{1-\alpha}, with u_t denoting the fraction of time devoted to production and L the labor force.^[12] This structure generates constant returns to reproducible factors—physical capital K and human capital-augmented effective labor—avoiding the diminishing returns that constrain neoclassical models and permitting a balanced growth path where output, capital, and human capital per worker expand indefinitely at a constant rate determined by the time allocated to learning.^[22] Human capital accumulation occurs via an externality: each agent's learning benefits the aggregate stock, as productivity depends on average rather than individual h, leading to suboptimal private investment relative to social optimum unless externalities are internalized.^[4] Extensions of the Lucas framework, such as the Uzawa-Lucas variant, emphasize intertemporal trade-offs in human capital investment, where forgone consumption funds education that boosts future productivity, with growth rates rising in the elasticity of output to human capital and falling in population growth or depreciation rates.^[2] These models predict that policies enhancing education returns—such as subsidies for schooling—can permanently elevate growth by shifting time allocation toward accumulation, though empirical calibration reveals sensitivity to parameter assumptions like the intertemporal elasticity of substitution.^[7] Unlike AK models relying on linear production, human capital variants incorporate micro-foundations of individual optimization, highlighting how initial human capital endowments amplify long-run divergences across economies due to compounding effects.^[13] Schumpeterian variants integrate Joseph Schumpeter's concept of creative destruction, where growth stems from sequential innovations that obsolete prior technologies, driven by profit-seeking R&D rather than mere accumulation. In the seminal Aghion-Howitt model of 1992, a research sector generates vertical quality improvements in intermediate goods, each innovation raising economy-wide productivity by a fixed factor \lambda > 1 while displacing the incumbent monopolist's market, yielding a growth rate g = \mu \log \lambda where \mu is the Poisson arrival rate of innovations proportional to research labor.^[37] This framework features a "business-stealing" effect, tempering R&D incentives as innovators capture rents at incumbents' expense, balanced against a "preemption" motive where early innovation blocks rivals; equilibrium innovation thus depends on market size, with larger economies fostering faster growth via scale effects.^[38] Subsequent Schumpeterian developments refine these dynamics, incorporating firm entry, expansion, and exit: successful innovators expand operations post-innovation, while displaced firms shrink or vanish, generating resource reallocation akin to observed business cycles in growth data.^[39] Unlike human capital models' focus on broad skill enhancement, Schumpeterian approaches stress asymmetric outcomes—growth concentrates in frontier innovators, with laggards facing obsolescence—explaining persistent firm-level heterogeneity and the role of competition policy in curbing monopolistic barriers to entry without stifling R&D.^[40] Empirical implementations, such as calibrations to U.S. patent data, validate the mechanism's ability to match observed growth volatility and cross-country patterns, though they underscore challenges in disentangling creative destruction from expansionary variety models.^[41]

Comparison to Exogenous Growth Theory

Fundamental Theoretical Divergences

The primary divergence between endogenous and exogenous growth theories lies in the treatment of technological progress. Exogenous growth models, such as the Solow-Swan framework developed in 1956, posit that long-run economic growth stems from an externally driven rate of technological advancement that is not explained or derived within the model itself.^[42] In contrast, endogenous growth theory internalizes technological change as a byproduct of deliberate investments in research and development, human capital accumulation, and innovation processes modeled explicitly, thereby making sustained growth rates dependent on these endogenous factors rather than unexplained residuals.^[43] This shift addresses the Solow model's limitation in accounting for why technological progress occurs, attributing it instead to economic incentives and spillovers from knowledge creation.^[44] A second key theoretical split concerns returns to scale and the accumulation process. Neoclassical exogenous models assume diminishing marginal returns to reproducible factors like physical capital, leading to a steady-state equilibrium where per capita output growth converges to zero absent population growth or exogenous technical progress.^[33] Endogenous models, exemplified by the AK framework, introduce constant or increasing returns to broad capital aggregates—including human and knowledge capital—eliminating convergence to stagnation and permitting perpetual growth through ongoing accumulation without reliance on external shocks.^[15] For instance, in the AK model, output is linear in the capital stock (Y = AK), where A captures productivity from non-rivalrous ideas, contrasting sharply with the Cobb-Douglas form Y = AK^αL^{1-α} (α < 1) of exogenous variants that enforces diminishing returns.^[42] These foundational differences yield divergent implications for policy efficacy and economic convergence. In exogenous theory, variations in savings rates or factor accumulation influence transitional growth paths and steady-state levels but leave the long-run growth rate invariant, rendering fiscal interventions impotent for permanent acceleration.^[43] Endogenous theory, however, implies that policies enhancing R&D subsidies, education spending, or innovation incentives can raise the steady-state growth rate itself by amplifying knowledge spillovers and scale economies.^[15] Consequently, exogenous models predict conditional convergence among economies with similar parameters due to catch-up dynamics, while endogenous frameworks allow for persistent cross-country growth disparities or even divergence if initial conditions or policy choices sustain differential innovation rates.^[33] Empirical scrutiny of these predictions remains contested, with endogenous proponents arguing that observed non-convergence in global incomes challenges the universality of neoclassical assumptions.^[44]

Divergent Predictions on Convergence and Scale Effects

In neoclassical exogenous growth models, such as the Solow-Swan framework, diminishing returns to reproducible factors like capital predict conditional convergence: economies starting with lower capital-labor ratios experience faster per capita growth rates, conditional on identical savings rates, population growth, and exogenous technological progress, eventually converging toward a common steady-state path where long-run growth equals the exogenous rate of technological advancement plus population growth.^[9] This arises because capital accumulation faces diminishing productivity, pulling poorer economies toward the steady state while richer ones grow more slowly.^[20] Endogenous growth models, by contrast, often predict the absence of such convergence or outright divergence due to mechanisms like constant or increasing returns driven by non-rivalrous knowledge accumulation. In the AK model, for instance, production takes the linear form Y = [AK](/page/AK), yielding a per capita growth rate of sA - n - \delta (where s is the savings rate, n population growth, and \delta depreciation), which depends on economy-specific parameters like s or A (reflecting human capital or innovation efficiency); differing values thus sustain permanent gaps in growth rates, with no inherent catch-up mechanism.^[20] Similarly, models emphasizing R&D-driven innovation, such as those with knowledge spillovers, imply that initial advantages in human capital or institutional quality compound over time, fostering "club convergence" among similar economies but divergence between groups, as superior innovators pull ahead indefinitely.^[45] These predictions align with observed persistent income disparities, challenging the neoclassical assumption of uniform technological diffusion.^[46] Regarding scale effects, exogenous models treat long-run per capita growth as independent of population size or economic scale, attributing it solely to exogenous factors. Early endogenous models, however, incorporate strong scale effects: in Romer's 1990 expanding-variety framework, the growth rate of knowledge or intermediate goods varieties is proportional to the aggregate labor force devoted to R&D, implying that larger populations generate more innovations per capita, elevating steady-state growth rates with scale—doubling population size would roughly double the innovation rate and thus per capita growth.^[47] This stems from the partial non-rivalry of ideas, where fixed R&D costs are spread over larger markets, but it contrasts sharply with exogenous neutrality to scale and has prompted later theoretical refinements to mitigate empirically inconsistent implications.^[48]

Empirical Assessment

Evidence Supporting Endogenous Mechanisms

Empirical studies have identified positive associations between research and development (R&D) expenditures and long-term economic growth rates across countries and regions. For instance, an analysis of U.S. states from 1963 to 2007 found that a 1% increase in R&D stock leads to output growth with returns ranging from 83% to 213% to state GDP, with 77% of these benefits spilling over to other states, supporting the notion of knowledge spillovers central to endogenous models.^[49] Similarly, panel data from OECD countries indicate that public and private R&D investments enhance labor-augmenting technical change, confirming key assumptions of endogenous growth frameworks where innovation drives sustained productivity gains.^[50] Cross-country regressions further bolster the role of human capital accumulation as an endogenous driver of growth. Research utilizing time-series data has validated extensions of models like Lucas (1988), showing that increases in average years of schooling correlate with higher per capita income growth rates, independent of physical capital accumulation, consistent with externalities from skilled labor fostering technological progress.^[7] In developing economies, empirical validations highlight how human capital investments yield persistent growth effects, as evidenced by studies linking educational attainment to total factor productivity improvements over decades.^[51] Innovation metrics, such as patent filings, provide additional support for endogenous mechanisms through scale and variety effects. Econometric analyses reveal that R&D-intensive sectors exhibit higher growth rates, with innovation outputs explaining variations in GDP per capita across nations; for example, a positive and statistically significant relationship between R&D intensity and growth has been documented in post-2000 datasets, attributing up to 0.5-1% annual growth differentials to endogenous knowledge creation.^[52] These findings, drawn from firm-level and aggregate data, underscore how internal investments in ideas and human capital generate non-rivalrous benefits that evade diminishing returns predicted by exogenous models.^[53]

Empirical Challenges and Failures to Validate

Empirical tests of endogenous growth models have revealed inconsistencies with observed data, particularly in the predicted scale effects of innovation-driven growth. Standard R&D-based frameworks, such as Romer's expanding variety model, imply strong scale effects where growth rates rise proportionally with the population or research labor force size, as more researchers generate more ideas. However, regressions using post-1960 data from OECD and other economies estimate the long-run elasticity of total factor productivity growth to the research labor force at approximately 0.2 to 0.6, substantially below the predicted value of 1.^[6] This discrepancy indicates that idea production does not exhibit the constant returns assumed, undermining the core mechanism of sustained endogenous growth without external drivers.^[54] The AK model, a foundational endogenous framework assuming constant returns to accumulable factors like broad capital, fares poorly against time-series evidence on investment and growth dynamics. It predicts that long-run growth rates should vary directly with savings or investment rates, yet U.S. and cross-country data from 1870 to 1990 show investment shares fluctuating or trending (e.g., rising from 15% to 20% of GDP in the U.S. post-1950) without corresponding permanent shifts in per capita growth rates, which remain stable around 2%.^[6] Such patterns align better with diminishing returns in neoclassical models, where transient booms from higher investment fade, rather than the perpetual acceleration posited by AK specifications.^[55] Efforts to validate human capital or variety-expanding models empirically often falter due to reliance on unobservable parameters, such as the degree of knowledge spillovers or the elasticity of substitution between goods, rendering tests non-falsifiable. For instance, cross-country growth regressions incorporating R&D spending or education levels yield mixed results, with coefficients on endogenous variables frequently insignificant or unstable across specifications, failing to displace exogenous technical progress as the primary growth residual.^[56] Moreover, while endogenous theory accommodates growth divergences, robust evidence of conditional convergence—poorer economies catching up to richer ones when controlling for policies and institutions, at rates of 2-3% per year in postwar samples—supports augmented Solow models over unbounded endogenous accumulation.^[57] These shortcomings have prompted modifications like semi-endogenous growth, where long-run rates depend on population growth rather than levels, effectively conceding original predictions' empirical invalidity.^[47]

Policy Implications and Debates

Investments in Education and R&D

In endogenous growth models emphasizing human capital, such as Robert Lucas's 1988 framework, investments in education augment the stock of skilled labor, which generates constant or increasing returns to scale due to knowledge spillovers that enhance aggregate productivity without diminishing marginal productivity.^[12]^[58] These externalities imply that private agents underinvest in schooling, as individuals appropriate only a fraction of the social benefits from improved human capital, which propagates through imitation, collaboration, and technological adaptation across sectors.^[58] Policy responses typically advocate subsidies for education—such as vouchers, tax credits, or public funding—to internalize these spillovers and elevate equilibrium human capital levels, thereby sustaining higher long-run growth rates. For instance, calibrations in Lucas-style models suggest that doubling education investment could raise steady-state growth by 0.5-1% annually, depending on externality parameters estimated from cross-country data.^[59] Parallel arguments apply to research and development (R&D) in models like Paul Romer's 1990 expanding-variety framework, where private R&D generates non-rivalrous ideas that expand the range of intermediate inputs, fostering perpetual growth under monopolistic competition.^[26] However, knowledge as a public good creates free-rider incentives, leading to suboptimal innovation without intervention; governments thus promote R&D through direct subsidies, grants, or patent protections that capture partial rents from discoveries.^[26] Empirical estimates indicate that a 10% increase in R&D expenditure as a share of GDP boosts total factor productivity growth by approximately 0.1-0.2% over five years in developed economies, with stronger effects in sectors like information technology.^[26]^[60] Cross-country regressions support these mechanisms, revealing that countries with higher education attainment—measured by average years of schooling—and R&D intensity experience faster GDP per capita growth; for example, panel data from 1960-2010 show a coefficient of 0.3-0.5 on human capital stocks for growth rates, controlling for physical capital and initial income.^[59]^[61] Yet, debates arise over causality and magnitude: while vector autoregression analyses confirm positive shocks from education spending to growth in high-income nations, evidence from low-income settings often detects level effects (permanent income shifts) rather than rate effects, as predicted by semi-endogenous variants critiquing pure endogenous assumptions.^[62]^[62] Moreover, subsidizing education in stagnant populations may divert resources from quality-improving R&D, potentially contracting variety-driven innovation and long-term growth.^[63] Policy implementation faces trade-offs, including fiscal distortions from funding sources like progressive taxes, which can crowd out private savings, and selection inefficiencies if subsidies favor quantity over skill-targeted programs.^[64] Optimal subsidy rates, derived from Ramsey-style rules in these models, typically range from 20-50% of marginal costs for both education and R&D, calibrated to externality elasticities around 0.5-1.0, though real-world applications like U.S. R&D tax credits yield elasticities of 0.1-0.3 for private spending increases.^[65] Critics, drawing on microdata, argue that public investments often underperform private ones due to bureaucratic allocation failures, with returns diminishing beyond secondary education in developing contexts.^[62]

Critiques of Government Intervention

Critics of endogenous growth theory's policy prescriptions contend that government interventions, such as subsidies for research and development (R&D) to internalize knowledge spillovers, are undermined by fundamental information problems. Public officials lack the dispersed, tacit knowledge held by private entrepreneurs and markets, rendering centralized efforts to direct innovation inefficient and prone to misallocation.^[66] This aligns with Hayek's emphasis on the limits of planners' foresight, as markets aggregate signals like prices to guide resource use toward viable projects, whereas governments often oppose successful innovations, such as Japan's Ministry of International Trade and Industry initially discouraging Sony's transistor radio development.^[66] Empirical assessments of R&D subsidies reveal limited effectiveness, with low additionality indicating that public funds frequently displace or merely supplement private initiatives that would occur anyway. Mansfield's 1984 analysis of 41 U.S. energy R&D projects found that firms would have self-financed 80% without subsidies, while each dollar of federal funding induced only 12 cents in additional private R&D, with subsidized projects yielding half the productivity of privately funded ones.^[67] Norwegian case studies similarly showed 78% of subsidized projects proceeding independently despite grants covering up to 65% of costs.^[67] Some evidence points to outright crowding out, where subsidies reduce private R&D expenditures, as documented in studies by Lichtenberg (1984) and others.^[67] Incentive misalignments compound these issues, as government decision-makers face no personal downside for failures, unlike private investors who bear direct losses. The 2011 bankruptcy of Solyndra, after receiving a $500 million U.S. Department of Energy loan guarantee, exemplifies how such dynamics lead to backing unviable technologies.^[66] Political capture further distorts allocation, channeling resources to influential firms rather than meritorious innovations, as seen in Tunisia where regime-connected enterprises captured disproportionate profits while harming broader competitiveness.^[68] Overall, these government failures often exceed the market imperfections endogenous models seek to address, yielding net inefficiencies in growth-promoting efforts.^[68]

Criticisms and Limitations

Theoretical Assumptions and Internal Inconsistencies

Endogenous growth models, such as those developed by Paul Romer in 1990, assume that ideas and knowledge are nonrivalrous goods, meaning their use by one agent does not preclude use by others, while being partially excludable through mechanisms like patents. This nonrivalry underpins increasing returns to scale in aggregate production, as accumulated knowledge enhances productivity economy-wide without proportional input increases.^[15] Models further posit constant or increasing returns to broad capital, including human capital and R&D inputs, rejecting the neoclassical assumption of diminishing marginal returns to accumulation alone.^[11] Knowledge spillovers are treated as key externalities, where private investments in innovation yield public benefits, sustaining long-run growth endogenously. A core internal tension arises from the scale effects inherent in first-generation models like Romer's, where the per capita growth rate is proportional to the size of the research labor force or population, implying larger economies inherently grow faster due to linear idea production from scaled inputs. This conflicts with the partial excludability assumption, as nonrival ideas should not scale linearly with population without invoking implausible constant returns to research effort across indefinite sizes, potentially leading to explosive or unbounded growth paths.^[47]^[69] Endogenizing the growth rate also introduces multiple equilibria and indeterminacy, as small differences in initial conditions or self-fulfilling expectations can trap economies in low-growth states despite identical fundamentals, complicating unique steady-state predictions and rendering transitional dynamics analytically intractable. These models rely on extreme assumptions about frictionless spillovers and monopolistic competition in innovation markets, yet fail to reconcile micro-level profit maximization—where firms underinvest due to externalities—with macro-level sustained expansion, often requiring ad hoc policy interventions to achieve uniqueness.^[11] In AK variants, the linear production function Y = AK assumes away diminishing returns axiomatically, begging the question of why knowledge accumulation evades the replication constraints observed in physical capital.

Empirical and Predictive Shortcomings

Empirical examinations of endogenous growth theory have revealed significant challenges in validating its mechanisms through data. A primary shortcoming lies in the rejection of scale effects central to early models like Romer's (1990) R&D-based framework, which predict that larger populations or higher population growth rates should elevate per capita output growth via expanded idea production. However, time-series analyses of post-World War II data from OECD economies contradict this: Jones (1995) regressed per capita GDP growth rates on population growth and other variables for France, West Germany, Japan, the United Kingdom, and the United States over periods spanning 1950–1990, yielding coefficients on population growth that are statistically insignificant (often near zero) or negative, with no evidence of proportionality as theorized.^[70] ^[71] The AK subclass of endogenous models, positing constant returns to accumulable factors for perpetual growth without exogenous technical progress, fares poorly against production function estimates. Empirical capital elasticities derived from cross-country and time-series data typically fall between 0.4 and 0.6, well below the unity required to sustain growth absent diminishing returns, as shown in studies aggregating firm-level and national accounts data.^[72] R&D spillovers, a cornerstone for endogenous innovation-driven growth, lack robust economy-wide confirmation. While micro-level firm studies detect positive returns to R&D (e.g., elasticities around 0.15 for productivity), aggregate evidence is tenuous; Griliches (1988) highlighted difficulties in isolating externalities beyond private appropriation, and rapid industrialization in East Asian economies like South Korea (average annual GDP growth of 7.8% from 1965–1970) and Taiwan relied predominantly on imported technology and imitation rather than domestic R&D stocks, which remained negligible until the mid-1980s.^[72] Predictive power is undermined by the theory's expectation of permanent growth accelerations from policy-induced rises in investment or R&D shares, yet data indicate transient effects. For instance, cross-country regressions linking savings rates to long-run growth yield implausibly high implied elasticities or fail under robustness checks (e.g., outliers like Botswana skew equipment investment-growth correlations), and the post-1973 productivity slowdown in OECD nations persisted despite stable R&D-to-GDP ratios around 2–2.5%, pointing to unmodeled factors like organizational inefficiencies rather than endogenous drivers.^[72] Endogenous models' divergence predictions—persistent income gaps due to internal knowledge accumulation—clash with evidence of conditional convergence, where poorer economies among OECD peers close gaps toward richer ones at rates of 2–3% annually when conditioning on factors like initial income and investment, as in Barro (1991) analyses, suggesting catch-up dynamics overlooked in pure endogenous frameworks.^[72]

Extensions and Recent Developments

Integrations with Empirical Microdata

Empirical microdata, such as firm-level records on innovation and individual-level data on education and productivity, have enabled researchers to test the microfoundations of endogenous growth theory, including knowledge spillovers and human capital accumulation. These datasets allow for causal identification of mechanisms like R&D-driven technological progress and externalities, which aggregate models often assume but cannot directly observe. For instance, historical U.S. patent and census microdata from 1836–2004 reveal that states with higher innovation rates, measured by patents per capita, exhibited 26% greater per capita income by 2000 compared to low-innovation states, supporting localized spillovers from inventive activity.^[53] Firm-level studies further integrate microdata to quantify innovation responses to endogenous factors. Analysis of World Bank surveys from over 3,900 Chinese manufacturing firms in 2002–2003 shows that a 1% increase in skilled human capital (e.g., highly educated workers) raises patent applications by up to 27% in metropolitan areas, using negative binomial regressions to address count-data skewness. General manager education and tenure also enhance firm-level innovation, with postgraduate-educated managers linked to 113% higher patents in smaller cities, affirming human capital's role in endogenous technological change akin to Romer (1990).^[73]^[15] However, microdata evidence on human capital externalities—central to models like Lucas (1988)—reveals limitations. Instrumental variable estimates from U.S. Census microdata (1960–1980), leveraging compulsory schooling laws, yield external returns to education of 1–2%, statistically insignificant and far below levels needed to sustain perpetual growth in endogenous models. Confidence intervals rule out externalities exceeding 5–6%, suggesting that while private returns to human capital are robust, aggregate spillovers may not fully validate the theory's growth predictions without additional mechanisms.^[74] Schumpeterian extensions, informed by firm dynamics microdata, connect creative destruction to endogenous growth. Firm-level evidence indicates that innovation at the micro level, such as patenting and R&D intensity, drives aggregate technological frontiers, with declining research productivity (e.g., halving every 12 years) prompting refinements like semi-endogenous variants. These integrations highlight microdata's value in refining theory but underscore empirical tensions, as micro-level mechanisms do not always scale linearly to macro outcomes.^[15]^[53]

Applications to Digital and Knowledge Economies

Endogenous growth theory applies to digital and knowledge economies by positing that sustained expansion arises from internal processes of knowledge creation and diffusion, where innovations in information technology and human capital exhibit non-diminishing returns due to the non-rivalrous nature of ideas.^[15] In these sectors, technological progress is modeled as endogenous, driven by deliberate investments in research and development (R&D) rather than exogenous shocks, enabling perpetual productivity gains as firms accumulate patents, algorithms, and data assets that scale across users without proportional cost increases.^[75] For instance, Romer's 1990 framework highlights how knowledge spillovers in idea production—evident in software ecosystems and open-source collaborations—generate increasing returns, contrasting with neoclassical models' convergence predictions.^[15] Empirical evidence supports these mechanisms in digital contexts, where information and communication technology (ICT) infrastructure amplifies endogenous growth by facilitating rapid knowledge dissemination and human capital enhancement. A 2023 study on Poland's economy found that ICT adoption, as an internal growth factor, significantly influences long-term trajectories through channels like innovation networks, with econometric models confirming positive coefficients for ICT intensity on GDP per capita growth rates from 2000–2020.^[76] Similarly, panel data from Chinese provinces (2012–2018) indicate that digital economy indices—encompassing broadband penetration and e-commerce—boost high-quality development by 0.238 units per one-unit index increase, primarily via endogenous R&D and technological catch-up, with robustness checks using instrumental variables affirming causality.^[77] These findings align with firm-level observations in knowledge-intensive industries, where knowledge transfers, rather than capital deepening, account for productivity surges, as seen in East Asian tech hubs during the 1990s.^[72] Extensions of endogenous models to digital platforms incorporate data as a cumulative input, where user interactions and machine learning generate self-reinforcing innovation loops, akin to human capital accumulation but with network externalities.^[78] In knowledge economies, this manifests in sectors like biotechnology and AI, where collective invention—publicly shared algorithms improving via iterative feedback—drives growth rates exceeding traditional manufacturing; a 2022 theoretical extension posits AI as a non-rival input akin to Romer's ideas, though empirical validation remains preliminary due to data scarcity in pre-2010 periods.^[79] Overall, these applications underscore policy relevance for subsidizing digital R&D to internalize spillovers, though challenges persist in measuring intangible outputs accurately.^[80]

References

[1]
Endogenous Growth Model - an overview | ScienceDirect Topics
An endogenous growth model is defined as a theory that emphasizes the role of internal factors, such as human capital investment, innovation, ...
[2]
Endogenous Growth Theory - Defined, Example, Limits
The endogenous growth theory is the concept that economic growth is due to factors that are internal to the economy and not because of external ones.
[3]
[PDF] Introduction to Endogenous Growth Models David Romer's R&D ...
Paul Romer's 1986 model and Robert Lucas's (1988) human capital model. • These models get around the diminishing marginal returns to “capital” assumption by.
[4]
[PDF] The Lucas Growth Model - JHU Economics
Lucas (1988) presents a growth model in which output is generated via a production function of the form. Y = AKα(`hL)1-α.
[5]
Endogenous Innovation in the Theory of Growth
8 In some other examples of endogenous growth theory, such as Romer (1990), Grossman and. Helpman (1991a, ch. 3), and Young (1993), innovation serves to ...
[6]
Endogenous Growth Theory: Intellectual Appeal and Empirical ...
This paper examines whether the recent theoretical insights stemming from endogenous growth theory have provided a better guide to explaining actual growth ...
[7]
[PDF] A Time-Series Test of the Endogenous Growth Model with Human ...
From the viewpoint of endogenous growth theory, these findings provides us with an explanation for observed divergence between rich countries and poor countries ...
[8]
The Origins of Endogenous Growth - American Economic Association
This paper describes two strands of work that converged under the heading of 'endogenous growth.' One strand, which is primarily empirical, asks whether there ...
[9]
[PDF] Neoclassical vs. Endogenous Growth Analysis: An Overview
The resurgence of growth theory that took place in the 1980s, and in- volved the development of endogenous growth models, arose in response to a perception ...
[10]
[PDF] The Origins of Endogenous Growth - depfe
Both Robert Lucas (1988) and I. (Romer, 1986) cited the failure of cross-country convergence to motivate models of growth that drop the two central assumptions ...<|separator|>
[11]
(PDF) Endogenous Growth Theory: A Critical Assessment
Aug 7, 2025 · 1Solow (1991) marks the origins of endogenous growth theory with Romer's doctoral thesis of 1983,. emerging as Romer (1986), ...
[12]
On the mechanics of economic development - ScienceDirect.com
This paper considers the prospects for constructing a neoclassical theory of growth and international trade that is consistent with some of the main features ...
[13]
[PDF] ON THE MECHANICS OF ECONOMIC DEVELOPMENT* Robert E ...
This paper considers the prospects for constructing a neoclassical theory of growth and interna- tional trade that is consistent with some of the main ...
[14]
The Origins of Endogenous Growth - jstor
Both my model (1986) and Robert. Lucas's model (1988) included fact 4 without taking the final step and including step 5. In both of these models, the ...
[15]
[PDF] Paul Romer: Ideas, Nonrivalry, and Endogenous Growth
The substantive contribution of that paper was to build a model in which the long-run growth rate was determined endogenously, and to highlight that, because of ...
[16]
Nobel laureate Paul Romer's contribution to endogenous growth ...
Oct 24, 2018 · A major contribution of Romer's 1990 paper was to seamlessly embed the ideas of increasing returns and imperfect competition3 into a micro- ...
[17]
The Origins of Endogenous Growth - IDEAS/RePEc
Paul M. Romer, 1994. "The Origins of Endogenous Growth," Journal of Economic Perspectives, American Economic Association, vol. 8(1), pages 3 ...<|separator|>
[18]
[PDF] Historical Antecedents of Contemporary Endogenous Growth Theory
In summary, Adam Smith's main contributions to early thought include making explicit the positive effects of the division of labor for productivity and ...Missing: milestones | Show results with:milestones
[19]
[PDF] Endogenous Technological Change Paul M. Romer The Journal of ...
Jan 9, 2008 · For ex- ample, in an extension of this model (Romer 1990), I show how an increase in L could reduce research effort and the rate of growth.Missing: key | Show results with:key
[20]
[PDF] 5 THEORIES OF ENDOGENOUS GROWTH - Reed College
The neoclassical growth theory that we studied in Coursebook Chapters 3 and 4 largely evolved in the 1950s. There was considerable filling-in of details in ...
[21]
[PDF] Economic growth, technological change, and climate change
Oct 8, 2018 · In his focus on the fundamental endogeneity of technological change, Romer has em- phasized how the economy can expand the boundaries – and ...Missing: internal | Show results with:internal
[22]
[PDF] The Mechanics of Economic Development
This paper considers the prospects for constructing a neoclassical theory of growth and interna- tional trade that is consistent with some of the main features ...
[23]
[PDF] Endogenous Technological Change: The Romer Model - Karl Whelan
The Solow model identified technological progress or improvements in total factor productivity. (TFP) as the key determinant of growth in the long run, but did ...
[24]
Endogenous Technological Change | Journal of Political Economy
Growth in this model is driven by technological change that arises from intentional investment decisions made by profit-maximizing agents.
[25]
[PDF] Paul M. Romer - National Bureau of Economic Research
It defines an equilibrium that allows for both monopolistic competition and external effects arising from knowledge spillovers. Section 4 offers a brief ...
[26]
[PDF] R&D, Innovation, and Economic Growth: An Empirical Analysis
They find that innovation is positively related to human capital in the R&D sectors and national knowledge stock.2 They also show that there is a significant ...
[27]
Human capital, innovation, and growth - Wiley Online Library
May 16, 2022 · This paper explores the interaction between human capital and innovation in the process of economic growth. Using a model of endogenous ...
[28]
[PDF] Sergio Rebelo Working Paper No. 3325
This paper studies a class of growth models in which cross—country differences in economic policy can generate this type of heterogeneity. In growth experiences ...Missing: 1991 | Show results with:1991
[29]
[PDF] The AK Growth Model
The AK model, where Y=AK, is derived by setting α=1 in the Solow model. It has a shortcoming that capital is the only factor of production.
[30]
[PDF] Endogenous Growth: AK Model - Lutz Hendricks
Oct 14, 2024 · The best way of thinking about AK models: ▶ a reduced form for a model with multiple factors ▶ there may be transition dynamics, but it does ...
[31]
[PDF] The Rebelo AK Growth Model - JHU Economics
The Rebelo AK Growth Model. Rebelo (1991) examines a model in which a social planner maximizes the discounted sum of utility in an economy with an AK ...
[32]
Long-Run Policy Analysis and Long-Run Growth Sergio Rebelo - jstor
This model is used to study the effects of taxation and the influence of the rate of savings on the rate of economic growth. Section III expands this model to ...
[33]
[PDF] Lecture 8, Neoclassical Endogenous Growth - MIT Economics
Nov 22, 2022 · Two-sector endogenous growth models behave very similarly to the baseline AK model, but avoid this.
[34]
[PDF] The underestimated virtues of the two-sector AK model - EconStor
We show that the two-sector version of the AK model proposed by Rebelo. (1991) can be read as an endogenous growth extension of Greenwood, Hercowitz and Krusell ...
[35]
None
### Summary of Expanding Variety Models (Romer's Model)
[36]
[PDF] Romer's Model of Expanding Varieties
In contrast to Solow model, Romer's model is better interpreted as a model of the growth of the World economy. ! Or alternatively stated: how the world ...
[37]
A Model of Growth Through Creative Destruction - jstor
A model of endogenous growth is developed in which vertical innovations, generated by a competitive research sector, constitute the underlying source of ...
[38]
[PDF] NBER WORKING PAPER SERIES A MODEL OF GROWTH ...
INTRODLJCION. This paper presents a model of endogenous stochastic growth, based on Schuxnpecers ... Also as we showed in Aghion and Howitt (1988), the inclusion ...
[39]
[PDF] A Model of Growth Through Creative Destruction - Harvard DASH
Aghion, Philippe, and Peter Howitt. 1992. “A Model of Growth Through Creative. Destruction." Econometrica 60, no. 2: 323-351. Published ...
[40]
[PDF] What Do We Learn From Schumpeterian Growth Theory?
These models shed light on several aspects of the growth process which could not be properly addressed by alternative theories.
[41]
https://cepr.org/voxeu/columns/sustained-growth-through-creative-destruction-nobel-laureates-philippe-aghion-and
[42]
[PDF] Exogenous and Endogenous Growth Models: a Critical Review
The empirical evidence has shown that higher real interest rates are posi- tively associated with growth (Balassa 1989; Gelb 1989), while financial repres-.
[43]
(PDF) Exogenous and Endogenous Growth Models: a Critical Review
Proponents of exogenous growth models argue that technological progress is the key determinant of long-run economic growth as well as international productivity ...
[44]
[PDF] The Growth of Nations - Harvard University
Hence, this work goes by the name endogenous growth theory. The Basic Model. The idea behind endogenous growth theory can be seen most easily by considering ...
[45]
Slow Convergence? The New Endogenous Growth Theory ... - jstor
We argue that endogenous growth theory has important regional implications, but also major limitations when applied to a regional context. Key words: endogenous ...
[46]
[PDF] On Convergence in Endogenous Growth Models
In endogenous growth models, the rate of convergence to a balanced path depends on technological parameters, not preferences, and increases with human capital ...
[47]
[PDF] Growth: With or Without Scale Effects? - Stanford University
That is, the scale effect shows up in the level of per capita income instead of its growth rate.
[48]
A note on endogenous growth and scale effects - ScienceDirect
The first generation of endogenous growth models (Romer (1990), Grossman and Helpman (1991), Aghion and Howitt (1992)) exhibited scale effects in innovation: as ...
[49]
[PDF] The Impact of Research and Development on Economic Growth and ...
R&D has a large long-run effect on output and TFP, with returns to state GDP of 83% to 213%, and 77% of returns accrue to other states.
[50]
Evidence on an endogenous growth model with public R&D
May 20, 2024 · This paper confirms important assumptions and results of the model for OECD countries. Labour-augmenting technical change is enhanced through private and ...
[51]
[PDF] « Human Capital and Economic Growth »
In short, in contrast with Lucas' assumption, endogenous growth would seem to be based more on the existence of externalities resulting from human capital ...
[52]
The Impact of R&D Investment on Economic Growth - MDPI
Empirical research shows a positive and statistically significant relationship between R&D intensity and GDP growth. For instance, N. Bloom et al. (2013) ...<|separator|>
[53]
None
### Summary of Microdata Integration with Endogenous Growth Theory
[54]
Growth without Scale Effects | Journal of Political Economy
This paper modifies models of endogenous innovation to allow for the possibility that a rise in the profitability of innovative activity could lead to an ...
[55]
Time series tests of endogenous growth models - ResearchGate
Aug 9, 2025 · Firstly, Jones (1995) criticizes the one-sector AK model, since it fails to reconcile the empirical facts of trending real investment rates and ...
[56]
Endogenous growth models: a critique of empirical testability
The role of human capital in economic development: Evidence from aggregate cross-country and regional US data. C.V. Starr Working Paper No. 92-46 (1992).
[57]
Evaluating growth theories and their empirical support
Aug 15, 2006 · According to Pack (1994), the empirical literature on endogenous growth models may not find evidence in favor of such a class of models because ...
[58]
[PDF] NBER WORKING PAPER SERIffi HUMAN CAPITAL AND GROWTH
that the "human capital revolution" in development has been a disappointment, and that growth accounting measures of the effects of education do not help us.
[59]
[PDF] Education and economic growth - ERIC
In endogenous growth theory, human capital is an essential input into a research sector which generates new ideas and technologies (Romer, 1990; Aghion and ...
[60]
How do R&D and remittances affect economic growth? Evidence ...
Apr 22, 2024 · The study reveals a novel mechanism of R&D expenditures, remittances, and economic growth in MICs, shaping their mutual influence on this economic landscape.
[61]
[PDF] Measuring Human Capital Effects on Economic Growth in ...
The effect of education on human capital is the reason endogenous growth models (e.g. Lucas, 1988; Romer, 1990) have been used extensively to study the linkages ...
[62]
[PDF] Ideas and Education: Level or Growth Effects?
These critiques of endogenous growth theory seem to imply that policies aimed at increasing investment in education and/or research will not be successful in ...
[63]
[PDF] The Effects of Education Subsidy in an R&D Based Economy
Jun 15, 2024 · Education subsidies may hinder long-term growth if population growth is constant or decreasing, as they expand variety innovation but impede ...<|separator|>
[64]
[PDF] Coupling Education and Innovation Policies for Economic Growth
Sep 19, 2020 · This paper analyzes the various layers that connect R&D subsidies and higher education policy to productivity growth. We put the development of ...
[65]
Endogenous growth and welfare effects of education subsidies and ...
We study the welfare effects of education subsidies and pensions in an OLG model with physical and human capital. ... Endogenous growth, human capital and ...
[66]
Why Industrial Policy Fails - Hoover Institution
Oct 26, 2023 · A basic understanding of the economics of information and of incentives shows why industrial policy fails. Consider information. One of ...<|control11|><|separator|>
[67]
[PDF] How Effective are Government R&D Subsidies: The Empirical ...
This paper aggregates the evidence on direct. sUbsidies, research institutes, and tax credits. Although the empirical results are patchy some clear indications.
[68]
Three reasons why industrial policy fails - Brookings Institution
i) Existing distortions. The analytical case for industrial policies is based on the idea that there is a market failure that is preventing industrialization ...
[69]
[PDF] The Romer Model - Karl Whelan
The Solow model identified technological progress or improvements in total factor productivity (TFP) as the key determinant of growth in the long run,.
[70]
[PDF] Time Series Tests of Endogenous Growth Models - Charles I. Jones
Jun 6, 2005 · That paper shows that eliminating scale effects in a straightforward way also eliminates the hallmark of the endogenous growth literature: in ...
[71]
Time Series Tests of Endogenous Growth Models - jstor
This paper argues that the prediction of long-run growth effects provides a simple, intuitive test of endogenous growth models in a time series context. The ...
[72]
[PDF] Endogenous Growth Theory: Intellectual Appeal and Empirical ...
Most empirical research generated by endogenous growth theory has tested earlier growth models, rather than testing endogenous theory itself. Moreover, most of ...
[73]
[PDF] Firm-level Human Capital and Innovation: Evidence from China
While at the macro level human capital has been recognized as an important determinant of innovation in endogenous growth theory (e.g., Romer, 1990), there.
[74]
[PDF] How Large Are Human-Capital Externalities? Evidence from ...
Similarly, many recent endogenous growth models emphasize the link between human capital and growth. For example, in Lucas's. (1988) model, worker ...
[75]
Understanding Endogenous Growth Theory: Key Concepts and ...
Critics argue that endogenous growth theory relies on assumptions that are difficult to empirically validate. Paul Romer's work on endogenous growth was ...
[76]
Endogenous growth and the influence of... | Open Research Europe
Empirical evidence shows that FDI can foster economic growth by assisting economies by transferring capital, management expertise, technology, and innovation ( ...
[77]
Study on the effect of digital economy on high-quality economic ...
Sep 21, 2021 · According to the endogenous growth theory, endogenous R&D and innovation are the core elements that drive economic growth and technological ...
[78]
[PDF] A Model of the Data Economy - NYU Arts & Science
Nov 26, 2021 · Keywords: Data, growth, digital economy, data barter. ... From here on, we return to the quality equation (2), withouth endogenous growth.
[79]
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4327622
[80]
Full article: Influence of the digital economy on economic growth
The endogenous growth model emphasizes how innovation and technological advancements greatly impact growth, in industries within the digital economy realm like ...<|separator|>