Chapter Four

Theories of Growth
4.1. The Solow-Swan growth model

By: Haile Girma (Assistant Professor)

Department of Economics
Salale University

1
 1. Introduction
Questions
1. Why are some countries rich and others poor?
2. Why do some countries grow faster than others?
3. How can economies exhibit sustained economic growth?
 The growth models(e.g. Solow growth model) predicts that
countries with:-
– high investment (savings) rates,
– low population growth rates and
– low rates of capital depreciation are likely to have more
capital and output per worker.

2
 In addition, countries that have:
– a high level of technology (A) or
– a high growth rate of technology (g) are also likely to
have more capital and output per worker in the
steady state.

3
 Stylized Growth Facts

 Before Industrial Revolution (1750) standards of living was very

similar across countries.

 Large differences in per-capita incomes across countries

appeared after IR.

 Positive correlation between per capita output and per capita

investment.

 Negative correlation between per capita output and population

growth.
 Large variations in growth rates
 Growth miracles, e.g. Japan, NICs
 Growth disasters, e.g. Sub-Saharan Countries, Argentina
 Over the modern era, cross-country income differences has
widened – enormous differences in human welfare across
different part of the world.
 Rich countries: Convergence
 Poorest countries: No Convergence

5
 Solow-Swan growth model
 Solow-Swan model is named after Robert Solow and Trevor
Swan, or simply the Solow model.
The Solow model focuses on four variables:
• Output (Y)
• Capital (K)
• Labor (L)
• Knowledge/Effectiveness of labor (A)

6
 The production Function takes the following form
Yt = F (Kt, AtLt)…………………………………(1.1)
 Where t denotes time, which enters indirectly into the
production function through K, L and A

Some properties of the production function:

1. Output changes over time only if input changes over time
2. Technological progress only if the amount of knowledge
increases.

7
 Assumptions of production function

1. Homogeneous of degree one (CRS).

• CRS: F(cK, cAL) = cF (K,AL) ……………………..(1.2)
• Implicitly assumes that other production factors (e.g. land) is
relatively unimportant
• With CRS, an intensive form of the production function is
easily specified.
• Set
…………………………..(1.3)

8
 Hence (1.3) can be rewritten in intensive form production function
as :
………(1.4)
Where

2. Essentiality condition (Both factors are necessary)

i.e., F(0, AL) = F(K, 0) = 0, for any K, AL.
3. Positive and diminishing marginal products.
 Intensive production satisfy,

9
4. Inada conditions (named after a Japanse economist Ken-Ichi
Inada (1963)):
and

→ MPK becomes large when capital stock sufficiently small and

vice-versa.
 Generally, the intensive form production function is
assumed to satisfy the following conditions:

10
 Examples of Production Function: graphically and mathematically.

Fig. 1.1. Production Function

11
Example: Cobb-Douglas technology
……………..(1.5)

………..(1.6)
……………(1.7)
(First Order Condition)
It is straightforward to check that this expression approaches
infinity as k approaches zero, and that it approaches zero as k
approaches infinity

< 0 (Second Order Condition)

It is negative for all k values

12
The Evolution of the Inputs into Production
The remaining assumptions of the model concern how the stock of
labor, knowledge, and capital change over time.
The model is set in continuous time (i.e., the variables are defined
every point in time)
Labor and knowledge (technology) is assumed to grow at a
constant rate over time:
…………..(1.8)
……………(1.9)
where n and g are exogenously given constant growth rates
and a dot over a variable denotes a derivative w.r.t time (that is,
(t) is shorthand for .

13
 The growth rate of a variable refers to its proportional rate of
change.
 That is, the growth rate of X refers to the quantity .
 Thus equation (1.8) implies that the growth rate of L is constant
and equal to n, and (1.9) implies that A’s growth rate is constant
and equal to g.

14
 In this model it is also assumed output is used to either for
consumption or investment .
 The saving rate (s) is assumed to be constant and exogenously
given.
 For simplicity one can assume that one unit of investment is
equal to 1 unit of new capital.
 Existing capital depreciates at a rate δ.
 These assumptions imply that the capital stock grows
according to.
………………………..1.10
 Although no restrictions are placed their sum is assumed to be
positive (i.e., n+g+δ)
 This completes the description of the model.

15
The Solow model is grossly simplified in a number of ways.
•There is only one good
• No government
• Fluctuation in employment are ignored
• Production is described as aggregate production function with just
three inputs
• The rate of saving, depreciation, population growth and
technological progress are constant.
• The model omits many obvious features of the real world which
are important for growth

16
 The Dynamics of the Model
Here we want to determine the behavior of the economy we have
just described in above.
The evolution of two of the three inputs into production, labor
and knowledge, is exogenous.
Thus, to characterize the behavior of the economy, we must
analyze the behavior of the third input, Capital.
a. The Dynamics of Capital
Because the economy may grow over time, it is much easier to
focus on the capital stock per unit of effective labor , k, than on
the unadjusted capital stock K.
Since i.e. a function of K, L and A, which are all functions of t,
the chain rule applies and we can find the intensive form of the
capital growth equation form.
17
………(1.11)
Where is simply k ,
 Substituting these facts into (1.11) using (1.10) equation yields.
…………..(1.12)

18
And using the fact that
……….1.13

actual investment - Break even investment

19
.

20
 Equation (1.13) is the key equation in the Solow model.
 It states that the rate of change of the capital stock per unit of
effective labor is the difference between:
• sf (k) ( actual investment per unit of effective labor) and
• (n+g+δ)k (breakeven investment) or the amount of investment
that must be done just to keep k at its existing level.
 There are two reasons that some investment is needed to prevent
k from falling.
First, existing capital is depreciating at (δk) this must be replaced to
keep the capital stock from falling.
Second, the quantity of effective labor is growing at n + g, thus, in
addition to replacing depreciation, capital stock must grow at
rate n + g to hold k constant.
Note: The fact that the growth rate of the quantity of effective labor,
AL, equals n + g is an instance of the fact that the growth rate of
the product of two variables equals the sum of their growth 21
• When sf(k(t) > (n+g+δ)k, k is rising.
• When sf(k(t)< (n+g+δ)k , k is falling.
• When sf(k(t)= (n+g+δ)k, k is Constant.
 In other words,
• If k begins at some level other than k*, it will move toward k*
• For k below k*, saving > the amount of investment needed to
keep k constant, so k rises.
• For k above k*, saving < the amount of investment needed to
keep k constant, so k falls.
 From Figure 1.2
• Since f (0) = 0, actual investment and break-even investment are
equal at k = 0.
• The Inada conditions implies
o As k approaches zero, f '(k) becomes large- the slope of the
actual investment line is greater than the slope of the break-even
investment line. 22
o As k becomes large, f '(k) falls toward zero -the slope of the actual
investment line is less than slope of the break-even investment
line.
• At k∗ value of actual investment and break-even investment are
equal- equal slopes.
• Finally, since f ''(k) < 0 , the two lines intersect only once for k > 0.
• From (Fig. 1.3),
• when k < k* , actual investment exceeds break-even investment as
a result is positive- k is rising
• When k > k*, the actual investment is less than break-even
investment and therefore is negative- k is falling
• Finally for k =k*, is zero- k is constant and is called the steady
state value of k.
• Thus, regardless of where k starts, it converges to k∗ and
remains there.
23
.

Figure 1.3: Steady State

24
 b. the Balanced Growth Path
• How output, capital and consumption growing in this economy
when k = k*(i.e. at steady state level).
• L and A are growing at exogenously given rates - n and g,
respectively.
• The capital stock K = ALk, and since k is constant at k*, the
aggregate capital stock of the economy is growing at the rate
(n+g).
• Since both capital and effective labor are growing at the same
rate (n+g), based on assumption CRS aggregate output is also
growing at the rate (n + g).
• Since consumption is (1 − s)Y , where s is constant, consumption
also grows at the same rate as output.
• Finally, capital per worker, K/L, and output per worker, Y/L, are
growing at rate g.
25
• Thus, the Solow model implies that regardless of its starting
point, the economy converges to a balanced growth path, a
situation where each variable grows at a constant rate.
• On the balanced growth path, the growth rate of output per
worker is determined solely by the rate of technological
progress.

26
 Comparative Dynamics: Impact of a Change in
Saving Rate

 The parameter of the Solow model that policy is most likely to

affect is the saving rate.
 What is the effect of (unanticipated) change in the saving rate s?
 Consider Solow economy that is on a balanced growth path, and
suppose that there is a permanent increase in s.
 The increase in s shifts the actual investment curve upwards,
thereby resulting in an increase in k*.

27
 a. The Impact on Output

28
 Initially, when s increases and the curve shifts up, at the initial
steady-state value of k, the actual investment exceeds break-
even investment.
 Thus, is positive which result in rise of k, which continues till
it reaches the new steady-state value of k*.
 This is depicted in the figure 1.4
 , we concluded grew at rate g when k = k*.
 However, when k is increasing, as the economy moves from
one steady-state to another, Y/L grows at a rate higher than g
(because the source of growth is not only g], but also growth
rate of k (which was constant when k = k* hence y* = f(k*)
before s increased to s*).
 Once k reaches its new steady-state value, growth rate of Y/L
falls back to g.

29
 Thus a permanent increase in s produces temporary
increase in the growth rate of output per worker, k rises
for some time but eventually it reaches a level at which
additional savings are devoted to maintaining the higher level
of k.

30
.

31
 By assumption, s jumps up at time and remains constant
thereafter.
 Since the jump in s causes actual investment to exceed break-
even investment by positive amount, jumps from zero to a
positive amount.
 k rises gradually from the old value of k∗ to the new value, and
falls gradually back to zero.
 Y/L equals Af (k).
 When k is constant, Y/L grows at rate g,
 When k is increasing, the growth rate of Y/L is greater than g
because both A and K are increasing.
 When k reaches the new value of k∗, however, again only the
growth of A contributes to the growth of Y/L, and so the growth
rate of Y/L returns to g.

32
 Thus, a permanent increase in the saving rate produces a
temporary increase in the growth rate of output per worker
because k is rising for a time, but eventually it increases to the
point where the additional saving is devoted entirely to
maintaining the higher level of k.
 The growth rate of output per worker, which is initially g, jumps
upward at and then gradually returns to its initial level.
 Thus, output per worker begins to rise above the path it was on
and gradually settles into a higher path parallel to the first.

33
 In sum, a change in the saving rate has a level effect but not a
growth effect: it changes the economy’s balanced growth
path, and thus the level of output per worker at any point in
time, but it does not affect the growth rate of output per
worker on the balanced growth path.
 Indeed, in the Solow model only changes in the rate of
technological progress have growth effects; all other
changes have only level effects.

34
 b. The Impact on consumption
 Since consumption per unit of effective labor c = (1 − s)f(k), an increase
in s at the initial steady-state level of k results in an initial decrease in c
and then as k rises to its new level c also rises.
 Whether or not c exceeds its original level can be seen by writing down
the expression for consumption per unit of effective labor.
 Steady-state consumption is given by:
…................... (1.14)
 is determined by we can write
………(1.15)
 An increase in s raises .
 It is known that ∂k∗/∂s is positive.
 Thus, c will rise in response to an increase in s if , is greater than (n + g +
δ).

35
 Intuitively when k rises investment must increase by (n + g + δ)
times k in order to sustain the new level of k.
 If f'k is less than (n + g +δ ), then the additional output from a
higher k is not enough to support the higher level of k.
 As a result c must decline in the long run to maintain the stock
of capital.
 On the other hand, if f'(k) exceeds (n+g+δ) there is more than
enough output to support the higher level of k, and therefore c
increases in the long run.
 However, if the steady-state value of k to start with is changed
due to s in such a way that f'k = (n + g + δ) then a marginal
change in s does not change c in the long run, and consumption
is at its maximum possible level among balanced growth
paths.

36
 This value of k∗ is known as the golden-rule level of the capital stock.
and the associate saving “saving gold”.
 At the golden rule level of capital, consumption is at its maximum level.
 Since s is exogenous in the Solow model, there is no guarantee that k
will be at its golden rule level.
This cases are depicted in the figures below.
 f’ (k∗) can be either smaller or larger than n + g + δ.
 This is shown in Figure 1.6.
 The figure shows not only (n+g+δ)k and sf(k), but also f (k).
 Since consumption on the balanced growth path equals output less
breakeven investment, c* is the distance between f(k) and (n+g+δ)k at
k = k*.
 The figure shows the determinants of c* for three different values of s
(and hence three different values of k*).

37
 In the top panel, s is high, and so k* is high and f '(k*) is less
than n + g + δ.
 As a result, an increase in the saving rate lowers consumption
even when the economy has reached its new balanced growth
path.
 In the middle panel, s is low, k* is low, f '(k*) is greater than n +
g + δ, and an increase in s raises consumption in the long run.
 Finally, in the bottom panel, s is at the level that causes f’(k*)
to just equal n + g+δ—that is, the f'(k) and (n + g+δ)k lines are
parallel at k =k*.
 In this case, a marginal change in s has no effect on
consumption in the long run, and consumption is at its
maximum possible level among balanced growth paths.

38
 This value of k* is known as the golden-rule level of the capital
stock.
 Since s is exogenous in the Solow model, there is no guarantee that
k will be at its golden rule level.

39
.

Figure 1.6. Output, investment and consumption on

balanced growth path
40
 The Central Questions of Solow model growth
theory
• Why are some economies much richer than others?
• Are income levels converging across nations?
 The Solow model identifies two potential source to why output
per worker varies across countries and over time:
1. Differences in capital per worker (K / L)
2. Differences in the effectiveness of labor (A)
 Due to convergence of k to k* changes in the effectiveness of
labor is the only factor that lead to permanent changes in
the growth rate.

41
 Empirical Application

a. Growth Accounting
 In the Solow model, long run growth of output per worker
depends only on technological progress.
 But short run growth can result from either technological
progress or accumulation.
 Thus the model implies that determining the sources of short run
growth is an empirical issue.
 Thus the growth accounting relates to an empirical extension that
allows to distinguish between different sources of growth.

42
 Growth accounting is pioneered by Abramovitz(1956) and
Solow(1957).
 To see how it works, consider the production function given as
• . This implies
……………..(1.16)

……………..(1.17)
43
Note that
is the elasticity of output with respect to labor at time t, is the
elasticity of output with respect to capital at time t and .
 Subtracting from both sides and using the fact that αL(t)+ αK(t)
= 1, gives an expression for the growth rate of output per worker:

…….…..(1.18)

44
Note that:
• The growth rate of Y,K and L are straight forward to measure.
• If capital earns its marginal product, can be measured using
data on the share of income that goes to capital

 R(t) then can be measured as the residual in equation (1.18)

above.
 Thus equation (1.18) provides a way of decomposing the
growth of output per worker into the contribution of growth
of capital per worker and a remaining term, the Solow
Residual

45
 The Solow residual–some times interpreted as a measure of the
contribution of technological progress (TFP).
 Growth Accounting has been applied to many issues:
 Yong(1995)-used detailed growth accounting to argue that the
higher growth in the newly industrialized countries of East Asian
than the rest of the world is mainly due to rising investment,
increasing labor force participation, and improving labor
quality(in terms of education), and not to rapid technological
progress other forces affecting the Solow residual.

46
 Table 1.1: A Growth Accounting Exercise for Ethiopia in the Last
Decade (2000-2010) based on the result of econometric estimation
reported in Alemayehu and Befekadu (2005) and Alemayehu et al
(2008).
Year GDP Contribution Contribution Contribution of factor
growth of labor of capital productivity(TFP)

Alemayehu Geda (2008;2013)

47
 b. Convergence
 Are poor countries growing faster than the rich ones, i.e. is there
convergence?
• Are income levels converging across nations?
 Solow model suggests three reasons why countries are expected
to converge:
 First, the Solow model predicts that countries converge to their
balanced growth paths.
 Thus to the extent that differences in output per worker arise
from countries being at different points relative to their balanced
growth paths, one would expect poor countries to catch up to
rich ones.

48
Second, the Solow model implies that the rate of return on capital
is lower in countries with more capital per worker.
 Thus there are incentives for capital to flow from rich to poor
countries; this will also tend to cause convergence.
Third, if there are lags in the diffusion of knowledge, income
differences can arise because some countries are not yet employing
the best available technologies.
 These differences might tend to shrink as poorer countries gain
access to state-of-the-art methods.

49
 Conclusion

 Only differences in the productivity of labor can account for

vast differences in wealth across space and time
 The “technology factor” is, however, exogenous to the Solow
model – the model makes no prediction of what this factor
really is, how it behaves or how it grows

50
 Assessing the Solow model
Big successes
A simple, transparent neoclassical model with clear implications.
Has been critical to empirical work.
Use of Solow residual to estimate relative impact of technology
and capital accumulation on income growth.
Raises convergence issue and allows for clearer discussion.
Failures:
Convergence picture more complex than implied by the model.
Model is only able to explain non-convergence, and vast
differences in income across countries by technology.
But technology moves easily across borders.

51
 Technology is completely exogenous, yet it is the key
determinant of growth.
 Technology is really just a residual.
 All we’ve really found is that, under neoclassical assumptions,
accumulation of physical capital cannot explain long-run growth.
 It could be that human capital accumulation, learning by doing,
public investment in infrastructure, etc. is driving growth.

52
.

w M o d e l
d o f S o l o
En r y m u c h
n k y o u v e
Tha t t e n t i o n !
o r y o u r A
F

53