Schaffner, Julie.

Development Economics: Theory, Empirical Research and Policy

Analysis. John Wiley and Sons. 2014.

Economic growth theory in

historical perspective
Chapter 4
10/2013

1
Introduction to growth theory
Growth theorists:
 Look at the world and make educated guesses about the features
of the world that are most important for understanding growth
 Build those features into models (using math or graphs) as the
assumptions that define the model
 Examine the logical implications of the assumptions for how and
why growth might vary over time within countries or differ across
countries
 Compare the patterns suggested by their models to empirical
patterns and seek to understand what they observe
 Repeat

Introduction to growth theory 2

Introduction to growth theory:
components
 Technological assumptions
o Options for turning inputs into outputs

 Behavioral assumptions
o How production factors are allocated across uses
o How the quantities of production factors evolve over time
o How technology evolves over time

 Exogenous parameters versus endogenous outcomes

Introduction to growth theory 3

The 1940s and 1950s
 Empirical observations
o Apparent success of rapid physical capital accumulation in the Soviet Union
o Long-term unemployment during Great Depression
o Large, poor agricultural sectors in developing countries

 Theoretical themes
o Capital fundamentalism
o Structuralism
o Dualism

The 1940s and 1950s 4

The Harrod-Domar growth model
 Technological assumptions
o We need u units of labor and v units of capital to produce one unit of output
• no human capital
• rigid technology
• constant returns to scale
• constant marginal product of capital

 Behavioral assumptions
o Investment in capital is financed by saving
o People save a fixed proportion s of income
o There is no technical change
o Labor grows at rate n
o We always have more labor than can be employed by the current quantity
of capital

The 1940s and 1950s 5

The Harrod-Domar growth model
• Technology and excess labor
• Same assumption re-expressed

• Saving behavior, investment,

capital growth

• Combining 2nd and 3rd above

• Re-expressing

The 1940s and 1950s 6

The Harrod-Domar growth model

Implications and interpretation:

 To increase growth, policymakers must increase s or reduce v

o And if v doesn’t change, then increasing s is the only way to speed growth
(capital fundamentalism)
 Rigidities mean that the government has a big direct role to play in
increasing s (structuralism)
 But this model also has some predictions that don’t accord well with
experience:
o knife-edge property
o steady growth in income is possible, but not steady growth in income per capita

The 1940s and 1950s 7

The Lewis model
 Technological assumptions
o Marginal product of labor virtually zero in traditional sector
o Standard production function in modern sector
• No human capital
• No technical change

 Behavioral assumptions
o Producers in traditional sector do not maximize profits and do not save
o Producers in modern sector maximize profits and save all of income
o Wages are stuck at some traditional level in traditional sector
o Workers migrate to the modern sector if paid the traditional wage plus a premium

The 1940s and 1950s 8

The Lewis model
VMPL in the Capitalist Sector

Capitalist
Income

Wage level

VMPL after reinvestment

Payments to Labor in of capitalist income
Capitalist Sector
Initial VMPL

L1 L2 Labor Employed in
Capitalist Sector (L)

The 1940s and 1950s 9

The Lewis model
Implications and interpretation:
 Growth is only possible through accumulation of capital in
modern sector.
 Producers in the modern sector see a wage that is “too high,”
so government should encourage them to expand
employment further.
o Since the modern sector produces substitutes for imports, encourage it
by protecting it from international competition.

 Agriculture plays a purely passive role in growth.

The 1940s and 1950s 10

The 1960s and 1970s
 Empirical observations
o Human capital and technical change are important (countering
capital fundamentalism)
o Even peasants respond to economic incentives (countering
structuralism)
o Growth through capital accumulation and technical change is
possible even in agriculture (countering some versions of dualism)

 Theoretical themes
o Technical change is critical for understanding growth
o Human capital is important, too
o Neoclassical decision-makers deliver ideal growth, so it might be
better to pull government out of the economy

The 1960s and 1970s 11

The simple neoclassical growth model
 Technological assumptions
o Neoclassical production functions Y=F(K,L) *

 Behavioral assumptions
o Investment in capital is financed by saving
o People save a fixed proportion s of income
o There is no technical change
o Labor grows at rate n
o All factors are always fully employed *

The 1940s and 1950s 12

Simple neoclassical growth model
Neoclassical technology

Re-express on per-worker
basis

Same as in Harrod-Domar

Fundamental equation used

to work out implications of
assumptions for growth

The 1960s and 1970s 13

Simple neoclassical growth model
k(n+d)
Output per Worker (k) k(n+d)
Output per Worker (k)
f(k)

s2f(k)

sf(k)

s1f(k)

ks Capital per Worker (k) Capital per Worker (k)

k1 k2

The 1960s and 1970s 14

Simple neoclassical growth model
Implications and interpretation:

 No growth in steady state equilibrium

 An increase in the saving rate increases growth in the short

run, and raises the long-run level of income per capita, but
does not lead to a sustained (for the long run) increase in the
growth rate.

 This model must be incomplete, because we do observe

sustained growth over long periods.

The 1960s and 1970s 15

Neoclassical growth with technical
change
Labor-using tech. change

Re-express on “per
effective worker basis”

Key equation

The 1960s and 1970s 16

Neoclassical growth with technical
change
Output per Effective Worker (y*)
k*(n+g+d)

sf(k*)

k*s Capital per Effective Worker (k*)

The 1960s and 1970s 17

Neoclassical growth with technical
change
 Steady state growth at the rate of technical change
o is constant but A is rising at the rate g.
o This implies K/L and Y/L are rising at the rate g, too.

 Neoclassical actors make good choices and deliver optimal

growth (assuming all markets, including financial markets,
work perfectly).

 After controlling for s and n+g+d (which determine steady

state k*), poorer countries should grow faster (i.e. we should
see conditional convergence), because where k* is lower, the
growth rate of k* is higher.

The 1960s and 1970s 18

Neoclassical growth with technical
change and human capital

 When we add human capital into the model, the story is much
the same, except that we must worry about the rate of
investment in human capital as well as the rate of investment
in physical capital.

The 1960s and 1970s 19

The 1980s to the present
 Empirical observations
o Rates of technical change and capital accumulation seem to vary across
countries.
o Some countries, and some groups within countries, appear stuck in
poverty.

 Theoretical themes
o Technical change must arise out of some people’s choices, and the most
compelling models that incorporate such choices suggest possible roles
for government intervention.
o Coordination/intervention might be necessary to help countries out of
macro poverty traps.
o Financial or other assistance might be necessary to help poor
populations out of micro poverty traps, and this might speed growth.

The 1980s to the present 20

Technical change as a by-product of
investment in human capital
 For example, let

 Holding A constant, economy is standard and works well.

 But when we account for effect of H on A:

o H has externalities, suggesting possible under-investment in the absence of government
intervention.
o Economies have increasing returns to scale, suggesting that poor countries might not
catch up.
Output per Effective Worker (y*)
k*(n+g+d) Output per Effective
k*(n+g+d)
Worker (y*)

sf(k*)
sf(k*)

k* s
Capital per Effective Worker (k*) k*s Capital per Effective
Worker (k*)
The 1980s to the present 21
Technical change as result of purposeful
activity: in technology leaders
 Technology is characterized by increasing returns to scale and producers
engage in imperfect competition.

 Monopoly profits serve as the reward for creating technologies.

◦ Though they may under-state the social benefit of invention, leading to under-investment in
the absence of intervention.

 Entrepreneurs keep devoting more labor to research and development

activities until the expected profits from technology creation fall to zero.

 The labor they devote to research and development has productivity (for
technology creation) that may
◦ fall with the quantity of labor devoted to research and development.
◦ rise with the initial level of A (if previous inventions improve research).
◦ fall with the initial level of A (if inventors run out of easy inventions).

The 1980s to the present 22

Technical change as result of purposeful
activity: in technology leaders

 Under these assumptions countries might grow at different

rates even in the long run, and

 Governments might have reason to intervene to improve

research and growth outcomes.
◦ Markets do many things well, but not everything.
◦ Governments might consider patent protection as a way to create the
profits that reward research and development.

The 1980s to the present 23

Technical change as result of purposeful
activity: in technology followers
 For technology followers, the challenge is to adopt advanced
technologies developed elsewhere.

 Technical change might be slowed by problems with

◦ human capital accumulation.
◦ intellectual property rights protection.
◦ trade barriers.

The 1980s to the present 24

Macro poverty traps

 Why do some countries seem stuck with low rates of

investment, innovation and growth?

 Is it possible that poverty is itself a barrier to growth?

The 1980s to the present 25

Macro poverty traps
Expected Return to Investment
Perceived by Any One Producer

Fraction of Other Producers

Who Invest

The 1980s to the present 26

Macro poverty traps
 Why might the investment returns anticipated by one investor
depend on how many other potential investors pursue
productive investment?
o One firm’s investment may increase the size of market for other firms, if
investment requires payment of a premium wage to draw workers out
of traditional sector.
o One entrepreneur’s decision to invest in production rather than
corruption may make corruption less attractive for others, by increasing
the probability of detection and decreasing cultural acceptability.

 What do we mean when we say that an economy in the bad

equilibrium is suffering from a coordination problem?

 What are the possible implications for policy?

The 1980s to the present 27
Micro poverty traps

 Why do some households seem to be stuck in deep poverty,

failing to grow their incomes, even while their richer
neighbors manage to save, invest and grow their incomes?

 Why might they fail to make high-return investments?

The 1980s to the present 28

Deeper determinants of growth

 If good policy is capable of encouraging investment and

productivity advance, what prevents some countries from
undertaking good policies?

 We’ll return to the political economy of growth and

development in Chapter 13.

The 1980s to the present 29