Annexes—The Solow Model and Long-Run Growth

the two founding articles of the model, published one year apart, which durably defined the program of growth macroeconomics.

Solow (1956)

Robert M. Solow, "A Contribution to the Theory of Economic Growth" (Quarterly Journal of Economics, vol. 70, no. 1, February 1956, pp. 65-94). Theoretical article that sets out the neoclassical model as a response to the Harrod-Domar model (unstable, with a single possible capital-output ratio). Solow introduces the aggregate production function with factor substitutability (Cobb-Douglas), demonstrates the existence and stability of an equilibrium path, and establishes that capital accumulation alone leads to a steady state due to diminishing returns. This is the result that justifies the crucial role of technological progress in sustaining long-run growth.

Solow (1957)

Robert M. Solow, "Technical Change and the Aggregate Production Function" (Review of Economics and Statistics, vol. 39, no. 3, August 1957, pp. 312-320). Empirical article that introduces the growth-accounting decomposition method. Solow applies his procedure to U.S. data 1909-1949 and finds that total factor productivity (Solow residual) explains about 87% of growth in output per hour worked, with capital accumulation accounting for only about 13%. This empirical result lastingly reshaped the profession and oriented growth macroeconomics toward the study of technological progress as the central variable.

Nobel

Solow receives the 1987 Nobel Prize "for his contributions to the theory of economic growth".

six empirical regularities formulated by Nicholas Kaldor in "Capital Accumulation and Economic Growth" (in F.A. Lutz and D.C. Hague (eds.), The Theory of Capital, Macmillan, 1961, pp. 177-222), which the Solow model precisely aims to reproduce.

Six stylized facts

(1) per capita output grows at a roughly constant rate without secular decline ; (2) per capita capital grows over time (capital deepening) ; (3) the capital/output ratio (K/Y) is roughly constant in the long run ; (4) the real rate of return on capital is roughly constant ; (5) the shares of capital and labor in national income are roughly constant ; (6) growth rates differ substantially across countries.

Importance of the framework

these facts combine decade-long stability (facts 1-5) and cross-national dispersion (fact 6) that any growth theory must explain simultaneously. Solow shows his model naturally reproduces facts 1, 3, 4 and 5 in the steady state with exogenous technological progress, and that fact 6 results from differences in structural parameters (s, n, g, A₀) — which motivates the conditional rather than absolute convergence documented by Barro and Sala-i-Martin (see vocabulary entry "Convergence speed").

Contemporary critique

Charles Jones and Paul Romer ("The New Kaldor Facts", American Economic Journal: Macroeconomics, vol. 2, no. 1, January 2010, pp. 224-245) update and extend the list in the era of globalization, notably adding the widening international gaps in idea/output ratios and the role of human capital.

extension of the Solow model that endogenizes the savings rate via intertemporal optimization of a representative household.

Founding sources

Frank Ramsey, "A Mathematical Theory of Saving" (Economic Journal, vol. 38, no. 152, December 1928, pp. 543-559) ; David Cass, "Optimum Growth in an Aggregative Model of Capital Accumulation" (Review of Economic Studies, vol. 32, no. 3, July 1965, pp. 233-240) ; Tjalling Koopmans, "On the Concept of Optimal Economic Growth" (in The Econometric Approach to Development Planning, North-Holland, 1965, pp. 225-300).

Difference with Solow

in Solow the savings rate s is exogenous (arbitrarily set, often calibrated to 20-30%). In Ramsey-Cass-Koopmans, a representative household maximizes ∫₀^∞ u(c(t))·e^(−ρt) dt subject to the capital accumulation constraint, where ρ is the rate of time preference. The savings rate then becomes a function of deep parameters (preferences, technology).

Keynes-Ramsey rule

the dynamic optimality condition reads (ċ/c) = (1/θ) × (MPK − δ − ρ), where θ is the coefficient of relative risk aversion (the inverse of the intertemporal elasticity of substitution). Consumption grows as long as the net marginal product of capital exceeds the time preference rate.

Consequences

the steady state differs from Solow's and coincides with a modified golden rule (MPK = ρ + δ + θg instead of MPK = δ + n + g). The equilibrium path is Pareto-optimal by construction. The model has become the backbone of modern macroeconomics (DSGE), of consumption theory, and of optimal economic policy.

overlapping-generations model proposed by Peter Diamond in "National Debt in a Neoclassical Growth Model" (American Economic Review, vol. 55, no. 5, December 1965, pp. 1126-1150), as an alternative to the Ramsey-Cass-Koopmans framework.

Central difference

instead of an infinite-horizon representative household, the economy contains a succession of finite-lived generations (typically 2 periods, working age and retirement), each generation coexisting with the previous one.

Theoretical consequences

the competitive equilibrium is no longer necessarily Pareto-optimal. There can be dynamic over-accumulation ("dynamically inefficient over-accumulation") where capital per capita exceeds the golden-rule level. Above all, Ricardian equivalence fails. Public debt has real effects because it redistributes across generations that do not internalize each other via altruism.

Major applications

analysis of public debt and pension financing (Auerbach and Kotlikoff, "Dynamic Fiscal Policy", Cambridge University Press, 1987), rational asset bubbles, intergenerational transfers. The OLG framework is today the standard for studying any question with an intergenerational dimension (population aging, public pensions, long-run sovereign debt, climate change). Diamond receives the 2010 Nobel (with Mortensen and Pissarides) for his work on market frictions, but the OLG model constitutes his central contribution to growth macroeconomics.

simplest endogenous-growth model, formulated by Sergio Rebelo in "Long-Run Policy Analysis and Long-Run Growth" (Journal of Political Economy, vol. 99, no. 3, June 1991, pp. 500-521).

Functional form

the aggregate production function becomes linear in broad capital, Y = A·K, where K includes physical, human, and knowledge capital considered as a single accumulable factor.

Conceptual breakthrough

the diminishing returns to capital that constrained long-run growth in Solow disappear. The marginal product of capital remains constant instead of tending toward zero, so accumulation can sustain permanent growth without recourse to exogenous technological progress.

Equilibrium growth rate

g = s·A − δ, where s is the savings rate, A the productivity of broad capital and δ the depreciation. The savings rate and economic-policy choices (capital taxation, public investment, education) directly affect the long-run growth rate, unlike Solow where they only affect the level.

Position in the debate

first modern formalization of endogenous growth, just before the more sophisticated models of Paul Romer ("Endogenous Technological Change", Journal of Political Economy, 1990) and Aghion-Howitt ("A Model of Growth through Creative Destruction", Econometrica, 1992). The AK model remains pedagogically central for showing how broadening the concept of capital (notably to human capital) can justify constant returns at the aggregate level, where diminishing returns are valid only for physical capital alone.

endogenous-growth model based on creative destruction, proposed by Philippe Aghion and Peter Howitt in "A Model of Growth Through Creative Destruction" (Econometrica, vol. 60, no. 2, March 1992, pp. 323-351).

Central mechanism

growth results from a succession of innovations that improve product or process quality. Each innovation grants its inventor a temporary monopoly rent, but this rent is progressively destroyed by the next innovation that renders the previous technology obsolete. This is the direct formalization of Joseph Schumpeter's "creative destruction" (Capitalism, Socialism and Democracy, 1942).

Quality ladders

production uses a continuum of intermediate goods on "quality ladders". Each jump on the ladder multiplies productivity by a fixed factor γ > 1. Innovation arrivals follow a Poisson process whose intensity depends on R&D invested.

Equilibrium growth rate

g = λ·ln(γ), where λ is the arrival rate of innovations per unit of time (γ being the productivity multiplier of each innovation). The rate depends on R&D incentives, hence on public policy (patents, R&D taxation, education).

Link to the cycles explanation

the article provides the modern microfoundation for the creative-destruction concept discussed in the C10 explanation (business cycles and crises) in its vocabulary entry "Creative destruction (Schumpeter)" and in its "Austrian theory" annex. Aghion and Howitt distinguish accumulation-driven growth (Solow, AK) from innovation-driven growth (Schumpeterian), with different policy implications.

Position in the literature

together with Romer (1990, product varieties), Aghion-Howitt forms the second major branch of 1990s "New Growth Theory". Aghion and Howitt later published a reference synthesis ("Endogenous Growth Theory", MIT Press, 1998).

founding article of "development accounting", published by Robert Hall and Charles Jones as "Why Do Some Countries Produce So Much More Output per Worker than Others?" (Quarterly Journal of Economics, vol. 114, no. 1, February 1999, pp. 83-116).

Central question

adapted from Solow's growth accounting (1957) but applied cross-sectionally at a point in time rather than temporally. Why do rich countries produce 30 to 40 times more per worker than poor countries? Method — log-linear decomposition of output per worker into three components, capital intensity (K/Y), human capital (measured by years of education), and productivity (TFP).

Central empirical finding

productivity (TFP) explains most of the cross-country variation, not factor accumulation. Rich countries are not rich primarily because they have more capital or more education, but because they use these factors more efficiently.

Institutional interpretation

Hall and Jones attribute TFP differences to "social infrastructure", i.e. the set of institutions and policies that determine the economic environment (security of property rights, trade openness, political stability, governance).

Link with Solow

the article is the cross-sectional application of the Solow residual and confirms at the international scale what Solow had found temporally for the United States. Productivity explains most of growth. It orients development macroeconomics toward the study of deep determinants of TFP (institutions, geography, culture).

major article by Daron Acemoglu, Simon Johnson and James Robinson, "The Colonial Origins of Comparative Development — An Empirical Investigation" (American Economic Review, vol. 91, no. 5, December 2001, pp. 1369-1401), which empirically establishes the causal role of institutions in long-run growth.

Central question

why is conditional convergence conditional ? AJR answer — because the structural parameters themselves (s, n, g, quality of human capital) are determined by institutions, and institutions inherited from history vary massively across countries.

Causal identification

major methodological challenge, institutions and growth being mutually endogenous (rich countries develop better institutions, and vice versa).

AJR propose an innovative empirical instrument

the mortality rate of European settlers in the 18th-19th centuries in the colonies, measured from British military archives.

Argument

in high-mortality regions (sub-Saharan Africa, tropical America), Europeans could not settle en masse and set up "extractive institutions" designed to exploit resources without building a rule-of-law state. In low-mortality regions (North America, Australia, New Zealand), they reproduced European "inclusive institutions" (contract security, property rights, checks and balances).

Empirical finding

institutions installed between 1500 and 1900 persist to this day and account for a substantial share of cross-country differences in income per capita. According to their instrumental-variable estimates, institutions have a large effect on income per capita: stronger protection of property rights accounts for a substantial share of cross-country differences in wealth.

Link with Solow

AJR provide an explanation for the TFP differences documented by Hall and Jones (1999) by tracing them back to their deep cause, political-economic institutions.

Extensions

Daron Acemoglu and James Robinson, "Why Nations Fail" (Crown Business, 2012), general-audience synthesis ; 2024 Nobel Prize to Acemoglu, Johnson and Robinson "for studies of how institutions are formed and affect prosperity".

major theoretical controversy over the very nature of aggregate capital, pitting "Cambridge UK" (Joan Robinson, Piero Sraffa, Pierangelo Garegnani, Luigi Pasinetti at the University of Cambridge) against "Cambridge MA" (Paul Samuelson, Robert Solow at MIT, in Cambridge Massachusetts).

Central question

can aggregate capital K be measured independently of the interest rate ? Neoclassical growth theory (Solow) assumes a production function Y = F(K, L) where K is a scalar quantity. But to add heterogeneous goods (machines, buildings, computers) into a single quantity, one must value them at a price, which is itself a function of the interest rate — a logical circularity pointed out by Joan Robinson as early as "The Production Function and the Theory of Capital" (Review of Economic Studies, 1953-1954).

Founding articles

Piero Sraffa, "Production of Commodities by Means of Commodities" (Cambridge University Press, 1960), reconstructs value theory without resort to aggregate utility ; the "Symposium on Paradoxes in Capital Theory" (Quarterly Journal of Economics, November 1966), with contributions by Samuelson, Levhari, Pasinetti, Morishima and others, marks the culmination of the controversy.

Reswitching

phenomenon where a given production technique can be optimal at a low interest rate, become suboptimal at an intermediate rate, then optimal again at a high rate. This contradicts the monotonic decreasing relation between capital intensity and interest rate that neoclassical theory presupposes to justify its aggregate production function.

Samuelson's partial concession

in the 1966 symposium, Paul Samuelson acknowledged the logical validity of reswitching and wrote that the neoclassical "parable" of homogeneous capital is not generally valid.

Contemporary status

the controversy was never formally resolved. Mainstream macroeconomics continues to use aggregate production functions of Cobb-Douglas type as operational tools, on the assumption that the issues raised by Cambridge UK are quantitatively minor in practice. Heterodox economists (post-Keynesians, Sraffians, neo-Ricardians) maintain that this controversy remains a fundamental unresolved critique of the research program inherited from Solow.

Pedagogical importance

for a hard-level explanation, it is essential to mention that the K aggregate concept used throughout the Solow model has been subject to a serious theoretical challenge whose implications remain debated.