Harrod and Domar left growth balanced on a knife edge. Robert Solow and Trevor Swan changed one assumption, letting capital and labour substitute, and the knife edge turned into a stable resting point. This lesson builds the model from its assumptions to its one equation and its one picture, which is probably the most-used diagram in growth theory.
The basic Solow model is a closed economy producing one good. It has five ingredients:
Compare this with Harrod–Domar. There the capital–output ratio was fixed. Here it adjusts: if capital is plentiful relative to workers, its marginal product falls, and the economy uses it differently. That one change removes the knife edge.
Capital grows by what is invested and shrinks by what wears out:
Now convert to per worker, . Taking the growth rate of k, it is the growth rate of K minus the growth rate of L: . Substitute and use Y/K = y/k:
Read it as a budget. The first term is how much capital is added per worker. The second, break-even investment, is how much must be added just to keep k constant: δk to replace what wore out and nk to spread the capital over the new workers. If saving exceeds break-even investment, k rises (capital deepening). If it falls short, k falls.
Draw the two terms as functions of k. The saving curve s·f(k) bends over, because of diminishing returns. The break-even line (n + δ)k is straight. They cross once, at the steady state k*. To the left of it saving is higher, so k rises; to the right the line is higher, so k falls. The lower chart follows k through time from the starting point you choose. Move the sliders and watch where the curves cross.
Capital per worker is below the steady state, so saving exceeds the investment needed to keep up and k rises.
The steady state is the value k* at which k̇ = 0:
In the steady state k is constant, and so is y = f(k). But the workforce grows at n, so the totals must grow too:
| Variable | Growth rate in the steady state | Why |
|---|---|---|
| Capital per worker, k | 0 | The definition of the steady state |
| Output per worker, y | 0 | y = f(k) and k is constant |
| Consumption per worker, c | 0 | c = (1 − s)y |
| Total capital, K | n | K = kL, with k constant |
| Total output, Y | n | Y = yL, with y constant |
Three exam results follow. Output grows at the population growth rate: IIT JAM 2021 asks what happens to aggregate output growth when population growth rises to a new value: it settles at the new rate (and GATE 2026 gives the same answer, n). The saving rate has no effect on the long-run growth rate, only on the level (UGC NET 2025: a higher saving rate raises growth in the short run, during the transition). And growth per worker is zero without technical progress (JNU 2018: faster labour-force growth has no effect on the long-run growth rate of per-capita income, because that rate is zero whatever n is).
Why does k converge to k* from any starting point? Look at the growth rate of capital per worker:
Capital per worker grows when saving per unit of capital, , exceeds n + δ. With diminishing returns, f(k)/k falls as k rises: capital is spread over the same workers and each unit yields less. So when k is small, saving per unit of capital is high and k grows quickly; when k is large it is low and k falls. At k* the two sides are equal. A growth rate that falls as the variable rises is a force that always pulls it back to a steady state.
In Harrod–Domar a departure from the warranted path fed on itself. Here it dies away. That is what “stable” means, and it is why Solow’s model displaced the knife edge. JNU 2019 poses the contrast as an assertion and reason: Solow is a major improvement because it builds a long-run growth model without fixed proportions. And JNU 2016: capital accumulation drives long-term growth in Harrod–Domar but not in Solow.
Diminishing returns are what make the Solow economy settle. Remove them, so that capital never loses its productivity, and the economy never settles: that is the AK model of Lesson 10.
Questions on the steady state, its stability, and what grows at what rate, from CUET-PG, Hyderabad, JNU, IIT JAM and GATE. Questions on saving, population, transitions and technology are in Lesson 6.
Definitions, one-step arithmetic and which-model questions. Most CUET and Hyderabad questions sit here.
Growth accounting, steady-state algebra, statement questions and matching, up to JNU, IIT JAM and UGC NET level.
This topic’s hardest questions, from the DSE entrance, are worked through in Lesson 14: The steady state and what moves it.
These are questions from the TestStop bank. A free account answers 10 a day, each with its worked solution; All-Access opens every set with no daily limit.
The Solow–Swan model is the standard neoclassical growth model. Output is produced from capital and labour with constant returns to scale and diminishing returns to each input. A constant fraction s of output is saved and invested, capital depreciates at rate δ, and the labour force grows at rate n. The economy converges to a steady state in which capital per worker is constant.
In per-worker terms, k̇ = s f(k) − (n + δ) k, where k is capital per worker, f(k) is output per worker, s is the saving rate, n the growth rate of the labour force and δ the depreciation rate. The first term is saving (investment) per worker; the second is the investment needed to replace depreciated capital and to equip new workers with the same capital per worker.
Set k̇ = 0, so that saving per worker equals break-even investment: s f(k*) = (n + δ) k*. For a Cobb–Douglas function y = k^α this gives k* = (s/(n + δ))^(1/(1 − α)). Output per worker in the steady state is y* = (k*)^α and consumption per worker is (1 − s) y*.
Capital per worker and output per worker are constant, so growth per worker is zero. Total capital and total output grow at the rate of labour-force growth, n. A change in the saving rate changes the level of output per worker but not its long-run growth rate, which is why the model needs technical progress to explain sustained growth per worker.
Yes. Because of diminishing returns, saving per worker rises more slowly than the break-even line. Below k* saving exceeds break-even investment and capital per worker rises; above k* it falls. So the economy converges to k* from any positive starting point. This is the main difference from the Harrod–Domar knife edge.
Harrod–Domar has fixed proportions of capital and labour and growth equal to s/v, which makes growth depend permanently on saving and makes the warranted path unstable. Solow lets capital and labour substitute with diminishing returns to capital, so the economy settles in a stable steady state where a higher saving rate raises the level of income but not the long-run growth rate.