Solow Growth Model Explained

Table of Contents

    The Solow growth model explains why investment can help an economy catch up quickly, yet eventually deliver slower growth. More machinery raises output per worker, but each additional unit of capital contributes less than the previous one. Sustained growth in output per worker requires continuing technological progress in the standard model.

    The distinction matters when judging whether rapid growth reflects rebuilding, a higher saving rate, or lasting productivity improvement. This article develops the model’s diagram and calculations; the broader guide to economic growth determinants covers the wider roles of education, institutions, trade, and policy.

    What Is the Solow Growth Model?

    The Solow growth model describes how saving, capital accumulation, labor growth, and technology shape an economy’s productive capacity. Robert Solow’s foundational paper appeared in 1956.[1] Its central question is how much growth can come from equipping workers with more capital, and what happens as that process matures.

    Key Concept

    With technology fixed, an economy approaches a steady state where investment just maintains capital per worker. Saving more raises the eventual level of output per worker, but the extra growth during the adjustment eventually fades.

    Capital means productive assets such as equipment, buildings, and tools. Investment means creating or adding to those assets. Buying an existing financial security does not, by itself, add a new machine to the model’s capital stock.

    Start with total output Y, capital K, and workers L. Output can be understood as a simplified measure of real gross domestic product. Dividing by workers gives capital per worker, k = K/L, and output per worker, y = Y/L.

    Production with Technology Held Fixed
    Y = F(K, L)   and   y = f(k)
    Output per worker depends on the capital available to each worker.

    The standard production function has diminishing returns to capital: adding capital raises output, but by progressively smaller amounts when labor and technology stay fixed. This is consistent with constant returns to scale: increasing capital and labor together in the same proportion increases output proportionately.[2]

    Watch: Solow Growth Model Explained: Why Japan Grew 3x Faster Than America

    The Solow Growth Model Equation

    Investment adds capital, while depreciation removes it. When the workforce grows, investment must also equip the additional workers to prevent capital per worker from falling.

    Change in Capital per Worker
    dk/dt = s × f(k) − (δ + n) × k
    Capital accumulation per worker equals saving-investment minus the investment required to maintain capital per worker.
    • s is the fraction of output saved and invested.
    • f(k) is output per worker, so s × f(k) is investment per worker.
    • δ is the depreciation rate; δk replaces capital that wears out or becomes obsolete.
    • n is the growth rate of workers; nk equips additional workers at the existing capital-per-worker level.
    • dk/dt is the rate at which capital per worker changes.

    This is a continuous-time equation, with technology fixed. The basic closed-economy setup assumes a constant saving fraction and that saving finances investment.[2] The loanable funds market provides a related explanation of saving and investment financing.

    Call (δ + n)k break-even investment. Depreciation and population-related dilution are different: depreciation reduces the capital stock, while a growing workforce spreads available capital across more workers. Here, population and labor are assumed to grow together.

    How to Read the Solow Diagram

    The horizontal axis shows capital per worker. The vertical axis shows investment per worker per unit of time. Saving-investment rises with capital, but its curve flattens because production has diminishing returns. Break-even investment is a straight line because its required amount is proportional to capital.

    Investment and break-even investment determine steady-state capital A hypothetical economy saves 20 percent of output, with output per worker equal to the square root of capital per worker. The investment curve is 0.20 times the square root of k. The break-even line is 0.05 times k. They intersect at capital per worker of 16 and investment of 0.80. At capital of 4, investment is 0.40 and break-even investment is 0.20. Capital rises below 16 and falls above 16. Investment per worker (units per year) sf(k): investment (δ + n)k: break-even 0 0.80 2.00 Steady state +0.20 change in k 0 4 k* = 16 40 k rises k falls Capital per worker, k
    Hypothetical Solow diagram: investment equals break-even investment at k = 16. The calculations below use the same curves and units. On small screens, swipe horizontally to view the full diagram.

    The intersection identifies the positive steady state:[2]

    Steady-State Condition
    s × f(k*) = (δ + n) × k*
    The star identifies the capital-per-worker level that remains constant when the model’s parameters stay fixed.
    • Below k*: investment exceeds break-even investment, so capital per worker increases.
    • At k*: the two amounts match, so capital per worker is constant.
    • Above k*: investment is insufficient to maintain that capital intensity, so capital per worker decreases.

    With fixed technology, output per worker also becomes constant. Total capital and total output can still grow as the workforce expands. A steady state therefore does not necessarily mean that total GDP stops growing.[1]

    Solow Model Example: Finding the Steady State

    A Hypothetical Economy

    Suppose output per worker is y = k0.5, the square root of capital per worker. The economy saves 20% of output, depreciation is 4% per year, and the workforce grows at 1% per year. Technology is fixed. Capital and output use normalized units, not actual country data. Flow amounts are continuous-time rates with time measured in years.

    Break-even investment is (0.04 + 0.01) × k = 0.05k.

    At the positive steady state, 0.20 × (k*)0.5 = 0.05 × k*. Dividing by the positive square root gives (k*)0.5 = 0.20 / 0.05 = 4. Therefore, k* = 42 = 16 and y* = 160.5 = 4.

    Check the result: investment is 0.20 × 4 = 0.80. Depreciation is 0.04 × 16 = 0.64, and equipping additional workers requires 0.01 × 16 = 0.16. Thus, 0.80 − 0.64 − 0.16 = 0 change in capital per worker.

    Capital adjustment at different starting levels; hypothetical continuous-time flows
    Capital per worker k Output per worker Investment Break-even investment Net change rate dk/dt
    4 2 0.40 0.20 +0.20
    9 3 0.60 0.45 +0.15
    16 4 0.80 0.80 0
    25 5 1.00 1.25 −0.25

    Each row follows the same calculation. At k = 4, investment is 0.20 × 2 = 0.40, break-even investment is 0.05 × 4 = 0.20, and the difference is +0.20. At k = 25, the difference is 0.20 × 5 − 0.05 × 25 = 1.00 − 1.25 = −0.25.

    The diagram’s gap measures an amount of capital per worker per unit of time. To obtain the instantaneous percentage growth rate of capital per worker, divide by k. At k = 4, this is 0.20 / 4 = 5%. At k = 9, it is 0.15 / 9 ≈ 1.67%. These are capital growth rates; output growth depends on the production function.

    What Happens When the Saving Rate Rises?

    Suppose the saving rate permanently increases from 20% to 30%, with everything else unchanged. At every positive capital level, more output now becomes investment. The investment curve moves upward, while the break-even line stays fixed.

    Higher saving raises steady-state capital per worker The solid investment curve uses a 20 percent saving rate and intersects the unchanged break-even line at capital per worker of 16. The dashed curve uses a 30 percent saving rate and intersects at capital per worker of 36. The economy accumulates capital between the old and new steady states. Investment per worker (units per year) Saving 20% Saving 30% Break-even 0 0.80 1.80 Old steady state New steady state 0 16 36 40 Capital per worker, k
    A higher saving rate moves the positive steady state from k = 16 to k = 36, without changing depreciation or workforce growth. On small screens, swipe horizontally to view the full diagram.

    At the old capital level of 16, investment becomes 0.30 × 4 = 1.20, while break-even investment remains 0.80. Capital per worker therefore starts increasing again at a rate of 1.20 − 0.80 = 0.40 normalized units per year.

    The new steady state solves 0.30 × (k*)0.5 = 0.05 × k*. The square root of capital is 0.30 / 0.05 = 6, giving k* = 36 and y* = 6. Investment and break-even investment both equal 0.30 × 6 = 0.05 × 36 = 1.80.

    A Higher Level, Followed by Slower Growth

    Output per worker rises from 4 to 6, a (6 / 4 − 1) × 100 = 50% level increase. During the transition, it grows. Once the new steady state is reached, its growth rate returns to zero because technology is still fixed.

    Saving also has an opportunity cost. Before capital or output can respond, consumption falls from (1 − 0.20) × 4 = 3.20 to (1 − 0.30) × 4 = 2.80. At the new steady state it reaches 0.70 × 6 = 4.20. That eventual consumption increase is specific to this example; higher saving does not always maximize steady-state consumption.[3]

    Conditional Convergence and Japan’s Catch-Up

    With the same saving behavior, technology, depreciation, and workforce growth, an economy starting with less capital per worker grows proportionately faster toward the same positive steady state. Diminishing returns make the opportunity for further capital deepening smaller as it approaches that destination.

    This is conditional convergence. Economies can have different steady states. A low-income economy already close to its own low steady state need not grow quickly, while changes in its saving or productivity conditions can change its destination.[3]

    Historical Illustration: Japan and the United States

    Angus Maddison’s 2001 OECD publication estimates that over 1950–1973, real GDP per person grew at an average compound rate of 8.05% per year in Japan, compared with 2.45% in the United States. The ratio is 8.05 / 2.45 ≈ 3.29.[4]

    The comparison illustrates rapid catch-up, but does not isolate its causes. Reconstruction, investment, technology adoption, and other developments can contribute together. Japan’s subsequent growth experience also involved changes in productivity and labor-force growth; it cannot be explained fully by reaching a single fixed Solow steady state.[7]

    These are explicitly dated real GDP comparisons from a historical estimate vintage. They are not nominal exchange-rate gains. Also, GDP per person and output per worker are different measures: translating between them requires accounting for workers as a share of the population.

    Capital Deepening vs. Technological Progress

    Capital deepening means more capital per worker. Technological progress improves what a given combination of inputs can produce. A single improvement raises productive capacity; continuing improvements can sustain growth in output per worker in the standard Solow model.

    Capital Deepening

    • Gives workers more productive equipment.
    • Faces diminishing returns with other conditions fixed.
    • A higher saving rate raises the eventual income level.
    • Its extra growth fades as the economy approaches its steady state.

    Technological Progress

    • Raises the productive capability of available inputs.
    • Can support continuing growth when improvement persists.
    • Allows capital per worker to keep increasing along a balanced growth path.
    • Enters the baseline Solow model as an assumed external process.

    Why the Definition of Steady State Changes

    To represent continuing progress, let A measure labor efficiency and let g be its growth rate. Production becomes Y = F(K, A × L). The product A × L is called effective labor.[5]

    Use a separate capital measure, keff = K / (A × L), for capital per effective worker. Its equation is:

    Capital per Effective Worker
    dkeff/dt = s × f(keff) − (δ + n + g) × keff
    Investment must keep pace with depreciation, more workers, and rising labor efficiency to maintain capital per effective worker.

    On the resulting balanced growth path, capital and output per effective worker are constant. Capital and output per ordinary worker grow at g, while total capital and output grow at n + g.[5] For hypothetical continuous-time rates of g = 2% and n = 1%, this means 2% growth per worker and 1% + 2% = 3% total growth.

    The extra g term belongs to the effective-worker equation. Adding it to the earlier diagram while continuing to define k as K/L would mix two different measurements.

    What Solow’s U.S. Evidence Measured

    In his 1957 study of U.S. private nonfarm output per labor hour over 1909–1949, Solow attributed 87.5% of the increase to technical change and 12.5% to increased capital use.[6] This was a growth-accounting decomposition. His technical-change category was broad, and the estimated productivity component can reflect unmeasured influences and measurement error. It is not a universal percentage of GDP growth caused by inventions.

    How to Use the Solow Model in Analysis

    1. Choose the measure. Distinguish total output, output per person, output per worker, and output per effective worker.
    2. State the assumptions. Identify saving, depreciation, workforce growth, and the assumed technology path.
    3. Compare investment with maintenance needs. With technology fixed, compare sf(k) with (δ + n)k to determine the direction of capital adjustment.
    4. Separate transition from sustained growth. Ask whether the forecast assumes temporary capital deepening or continuing productivity improvement.

    When forming capital market expectations, this framework helps make growth assumptions explicit. The model does not, by itself, turn an economic growth forecast into a security-return forecast.

    Common Mistakes in the Solow Growth Model

    • Treating a higher income level as permanently faster growth. Saving can lift output during a transition without changing its eventual growth rate.
    • Calling all break-even investment depreciation. Replacing worn-out capital and equipping additional workers are separate requirements.
    • Reading the diagram gap as a percentage. The gap is dk/dt. Divide by k for the capital growth rate.
    • Assuming every poor economy must catch up. Convergence depends on the relevant steady state and underlying conditions.
    • Assuming capital becomes useless at the steady state. Additional capital still raises production; the saving flow simply matches the investment needed to maintain capital per worker.
    • Mixing ordinary and effective workers. The technology extension changes the denominator from L to A × L.

    Limitations of the Solow Growth Model

    The Source of Technology Growth Remains Unexplained

    Technology is exogenous in the baseline model: its path is supplied as an assumption. The model identifies the importance of continuing progress without explaining how research, incentives, or institutions generate that progress.[5]

    The basic model also assumes a fixed saving fraction and a simplified aggregate production process. Separate human-capital accumulation, institutional differences, and other influences require additional structure.[3] Its conclusions depend on the assumptions used; they are not universal laws covering every possible growth model.

    Solow studies long-run productive capacity. Short-run unemployment and demand fluctuations require other tools, such as the aggregate demand and aggregate supply model. Average output also does not reveal how income is distributed or establish that a particular saving policy improves everyone’s welfare.

    Frequently Asked Questions


    With technology fixed, it is the capital-per-worker level where investment exactly covers depreciation and the needs of additional workers. Capital and output per worker stay constant, although total output can grow with the workforce.


    Higher saving initially makes investment exceed break-even investment. Capital and output per worker rise, but diminishing returns reduce the additional output generated by further capital. With technology fixed, the economy eventually reaches a higher steady-state output level with zero per-worker growth.


    No. Catch-up is conditional. Countries with comparable underlying parameters converge toward a comparable steady state, but different saving, workforce-growth, and productivity conditions can imply different destinations. Low income alone does not establish a large growth opportunity.


    When population and workers grow together, more investment is needed to equip the additional workers. Holding saving, technology, and depreciation fixed, a higher n raises the break-even line and lowers the positive steady-state capital and output per worker. This is a model comparison, not a complete judgment about demographic policy.


    Yes. With continuing labor-augmenting technological progress, the steady state is defined per effective worker. Those quantities remain constant while output per ordinary worker grows at the technology rate g. Total output grows at n + g along the balanced growth path.


    No. Higher saving reduces current consumption before output responds. It raises eventual output, but sufficiently high saving can also reduce steady-state consumption. In the fixed-technology model, the Golden Rule identifies the saving rate that maximizes steady-state consumption, rather than the highest possible saving rate.[3]

    Sources

    Sources consulted October 20, 2026. Model examples and diagrams use hypothetical parameters. Historical estimates retain their stated periods and source vintage.

    1. Robert M. Solow, A Contribution to the Theory of Economic Growth, February 1956, especially Sections II–III, pp. 66–70. University-hosted scan. The original paper defines output net of depreciation; this article’s explicit depreciation equation follows the modern treatment in Source 2.
    2. Daron Acemoglu, MIT, Solow Growth Model: Theory and Data, October–December 2025, especially assumptions, equations (25)–(26), and the Steady State With Population Growth diagram, PDF pp. 42–49.
    3. George-Marios Angeletos, MIT, The Solow Growth Model (and a look ahead), especially Sections 2.1.7, 2.4–2.5, and 2.6.1 on steady states, conditional convergence, and consumption.
    4. Angus Maddison, OECD, The World Economy: A Millennial Perspective, June 12, 2001, Table B–22, p. 265: annual average compound growth of real GDP per capita over 1950–1973.
    5. Christian Wolf, MIT, The Solow Model with Exogenous Perpetual Growth and Growth Accounting, Fall 2025, especially PDF pp. 18–25 on effective labor, balanced growth, and exogenous technology.
    6. Robert M. Solow, Technical Change and the Aggregate Production Function, August 1957, pp. 312, 314, 319, and 320: definition, measurement scope, error discussion, and summary findings. University-hosted scan.
    7. Masaaki Shirakawa, Bank of Japan, Japan’s Economy and Innovation, May 31, 2010, Section II.2, The Rise and Fall in the Growth Rate.
    Disclaimer

    This article is for educational purposes and does not provide investment or economic policy advice. Hypothetical model parameters are not country estimates or forecasts. Historical growth estimates depend on their source, period, and measurement method.