6.
THEORIES OF
ECONOMIC GROWTH:
SOLOW
Karthikeya Naraparaju
Overview
Solow Growth Model
Steady State
Parameters affecting Steady State
Level and Growth Effects
Technological Progress
Convergence
Solow (1956) Growth Model
Solow’s model relies on the possible endogeneity of another
parameter in Harrod-Domar model: the capital-output ratio, θ.
Solow’s twist on the Harrod-Domar story is based on the law
of diminishing returns to individual factors of production.
Capital and labour work together to produce output.
If there is plenty of labour relative to capital, a little bit of
capital will go a long way.
If there is a shortage of labour, capital-intensive methods are
used at the margin: capital-output ratio rises.
Solow Growth Model – The Solow Equations
K (t+1) = (1-δ) K(t) + sY(t)..............(3.8)
Dividing through population, we have,
(1+n) k(t+1) = (1-δ)k(t) + sy(t).......(3.9)
Where ‘k’s and ‘y’ denote per-capita magnitudes.
Intuition: RHS has two parts: per-capita capital (net of
depreciation) and current per-capita savings.
Adding these two should give us the per-capita capital stock in
k(t+1).
But population is also growing (at rate n), thus this exerts a
downward drag on per-capita capital stocks.
Solow Growth Model
Larger the rate of growth of population, the lower is per-
capita capital stock in the next period.
We now relate the per capita output at each date to the per
capita capital stock, using the production function.
With constant returns to scale, we may use the production
function to relate per-capita output to per-capita input.
Moreover, as we know, production function also exhibits
diminishing marginal returns to each input.
In our case, it is diminishing marginal returns to per-capita
capital.
Solow Growth Model – Production Function
Evolution of Capital Stock
(1+n) k(t+1) = (1-δ)k(t) + sy(t).......(3.9)
The Steady State
If the initial stock of per-capita capital is “low”, the output-
capital ratio is very high and so the per capita capital stock
can expand rapidly.
The growth of per-capita capital slows down over time and it
settles down to k*.
Why is this slowing down happening?
Diminishing marginal returns to per-capita capital.
For each subsequent increase in k, the increment in y is lower,
which in turn implies that the growth of k in the next period is
lower (from eq. 3.9).
The Steady State
Growth loses momentum if capital is growing too fast relative
to labour, as is happening to left of k*.
The growth of capital is then brought in line with the growth of
labour.
Thus long-run capital-labour ratio is constant – k*.
The Steady State
Similarly, if a country is starting with a ‘high’ initial capital
stock, the output-capital ratio is low.
So the rate of expansion of capital is low, relative to the rate
of growth of the population.
This implies that the per-capita stock of capital falls.
The per-capita stock of capital continues to fall until it reaches
k*, where the rate of growth of capital is equal to the rate of
growth of population.
The Steady State
If the per-capita capital stock settles down to some “steady-
state”, so must per-capita income!
Thus, in this version of Solow model, there is no long-run growth
of per capita output, and total output grows exactly at the
rate of growth of the population.
The savings rate has no long-run effect on the rate of growth.
Sharp contrast with Harrod-Domar model.
Solow v/s Harrod-Domar
This discrepancy is coming because of diminishing returns to
capital, thus creating endogenous changes in capital-output
ratio.
Smaller is the diminishing returns, closer is the curve to a
straight line, larger is k*.
The different predictions of these models are driven by
different assumptions about technology.
Parameters and Steady State
The rate of savings does not affect the long-run growth rate of
per-capita income (which is zero).
But savings rate affects the long-run level of income.
Similarly the rate of depreciation of capital (δ) and the growth
rate of population (n) will have an effect on the steady-state
level of per capita output.
At steady-state, k*/y* = s/(n+ δ).
Parameters and Steady State
An increase in s, will increase k*/y*.
This means lower output-capital ratio which can happen only at
a higher level of k*.
On the other hand, a higher depreciation (δ) or population
growth rate (n) implies a lower steady-state k*.
This can be seen graphically but also through reasoning.
Level and Growth Effects
A growth effect is an effect that changes the rate of growth of
a variable.
A level effect, leaves the growth rate unchanged while shifting
the entire path, up or down.
Level and Growth Effects: Population
The parameter of population growth (n) has an interesting
double effect.
An increase in ‘n’ lowers the steady-state per-capita income,
i.e. It has a level effect.
But at steady-state the rate of total income should equal
population growth rate.
Which means as population growth increases, the total income
should also grow at faster rates in the steady-state!
Population is both an input as well as a consumer of final
goods.
Level effects of Savings Rate
Savings rate on the other hand, has only level effects.
It does not have any growth effects on the total income in the
long-run.
Higher savings rate pushes the economy to a higher trajectory
in the short and the medium run but ultimately, in the long-run,
per-capita income settles down to a steady-state level.
Savings only has a level effect in Solow’s model, unlike the
Harrod-Domar model!
Solow Model – so far
Solow model has a strong prediction:
Regardless of the initial per-capita capital stock, two countries
with similar savings rate, depreciation rates, and population
growth rates, will converge to similar standards of living “in the
long-run”!
How seriously should we take the Solow model?
The real world is actually different: with growth and also
different standards of living across countries.
But that is no reason to discard a model!
Models are only pointers to important aspects of reality.
Solow Model With Technical Progress
We can think of economic growth as having broadly two
sources:
technical progress – better and more advanced methods of
production
Through build-up of plant, machinery, etc.
The Solow Model claims that without the first, the second
component alone cannot generate growth.
With technology improvements, we can think of our production
function to keep on moving upwards in each period.
We will then have some growth to be sustained in the steady-
state.
Solow Model With Technical Progress
We can think of technical progress as a way of improving the
efficiency, or economic productivity of labour.
We can make a distinction between working population P(t)
and the amount of labour in ‘efficiency units’ L(t) used in
production.
L(t) = E(t) P(t).............(3.11).......Effective Population
E(t): efficiency or productivity of an individual at time t.
With this we can amend the capital accumulation equation.
Solow Model With Technical Progress
Solow Model With Technical Progress
Capita per efficiency unit of labor produces output per efficiency unit
of labour.
Just as in the original model, if there too much of Capita per
efficiency unit of labor, then we have a shortage of (effective)
labour and the output-capital ratio falls.
That is there are diminishing returns to efficiency units of labour.
If the capital per effective labour rises, it means physical capital is
growing faster than the rate of population growth and technical
progress combined.
Solow Model With Technical Progress
The analysis runs exactly parallel to the earlier case.
However, the interpretation of the steady-state level of capital
per effective labour changes.
Even though the capital per effective labour converges to a
stationary steady state, the amount of capital per member of
the working population continues to increase!
At what rate will this increase?
Thus there is long-run growth in the model!
Convergence - Unconditional
The strongest prediction of the Solow model is called
unconditional convergence.
Suppose countries, in the long-run, have same rates of technical
progress, savings rate, population rates and depreciation
rates,
then in all countries, capital per efficiency labour will converge
to a common value, k^*.
History in terms of countries’ initial conditions does not matter.
Convergence - Unconditional
Convergence – Unconditional (Baumol
1986)
Convergence – Unconditional (De Long
1988)
