Human Capital and Growth

1
Michaelmas Term 2010
Part IIB. Paper 2
Economic Growth
Lecture 4 Human Capital Models of Economic Growth
Dr Tiago Cavalcanti
2
Readings and Refs
Texts Jones chs. 3 8 Articles () Mankiw,
Romer D. Weil (1992) A contribution to the
empirics of economic growth Quarterly Journal of
Economics, 107, 407-437. Lucas (1988) On the
mechanics of economic development Journal of
Monetary Economics, 22, 3-42. () Lucas (1990)
Why doesnt capital flow from rich to poor
countries American Economic Review, 80 (2),
92-97.
3
Checking Predictions 3 Regressions
MRW ran regressions of the form

• first regression tests for absolute convergence.
  Too demanding! It requires that income gap
  between two countries decline, irrespective of
  what type of policies, institutions, and
  investment behavior these two countries have.
• second regression tests for conditional
  convergence
• get estimate of speed of convergence??

4
Testing for conditional convergence incorporate
terms to take account of steady state
differences directly.
BX include a whole range of possible
determinants, including the original investment
rates and population growth but also variables
such as educational attainment levels, life
expectancy, public spending on education, and on
consumption, black-market for.ex. premium,
political instability etc.
? estimates of convergence rate ? remain around
2 - so they conclude that strong and robust
evidence for conditional convergence at rate
around 2 percent ? Convergence rate of 2 only
half of value predicted by neo-classical theory
5
Prediction 4 Interest rates, capital flows

• Neo-classical model Cobb-Douglas technology

• Stylised facts ? 1/3, and yUS 10 ?
  yIndia

• If technology transferable, capital investment
  about 100 times more productive in India. Should
  see LARGE capital flows.
• Equality achieved only if labour 10 times more
  effective in US

6
Neo-Classical Growth Model Reality Check

• Neo-classical growth model works qualitatively,
  but quantitative predictions do not fit
• estimated capital share too high (near 0.6)
• estimated the rate of convergence near 2 per
  annum, about half the rate predicted by the model
• observed interest rates differentials and
  international capital flows much lower than model
  predictions

7

• Capital appears far more effective/important in
  explaining per capita GDP than predicted by
  Neo-Classical theory

• Two primary approaches to account for this
• Include another factor Human Capital into
  Neo-Classical production function
• Incorporate spill-over effects from capital
  investment
• Investment in new capital has not only directly
  increases output, but also introduces positive
  externalities for related industries ? Endogenous
  Growth Models

8
Human Capital

• Observe quality of labour varies and labour
  productivity increases with education training
• However, education and training use resources
  that could have been allocated to alternative
  uses -- such as consumption or investment in
  physical capital
• decision to increase stock of education/training
  within labour force conceptually similar to
  investment in stock of physical capital (Becker
  Human Capital)

9
Augmenting Neo-classical model with Human Capital
1. Mankiw, Romer and Weil (MRW 1992) include
human capital as input in Cobb-Douglas technology
2. Lucas (1988) use investment in human capital
to increase effective labour input
ut proportion of labour allocated to
education/training
10
MRW Approach

• Two dynamic equations

• Steady state solution

11
MRW Approach
12
Unique Steady-State
13
Some Implications

• Higher saving rate in physical capital not only
  increases k, but also human capital, h.
• Same applies for higher investment in human
  capital.
• The relative contribution of physical and human
  capital in output depends on the share of each
  factor.

14

• substitution of steady state solutions in
  production function

So

• For country i, assuming that ln Ai C ?i ,
  gives regression

15
MRW run an OLS regression on this equation, using
the same 98 non-oil exporting countries, and
obtain coefficient estimates (std errors in
parentheses)
Giving estimates for ? 0.31 and ? 0.28
MRW conclude augmented Solow model, with ? 1/3
and ? 1/3 consistent with their findings
16
(No Transcript)
17
MRW augmented Solow model

• start with dynamic equations

• assume physical and human capital earn same
  interest rate r

18
Speed of Convergence

• non-linear fundamental dynamic equation

19
Speed of Convergence

• speed of convergence

20
Speed of Convergence

• Calibrate ? 1/3, ? 1/3 and (n g ? )
  0.06, have ? 0.02
• Augmented Solow model implies realistic speed of
  convergence

21
Interest rates (revisited)
The augmented model gives
Using stylised facts that ? ? 1/3, and yUS
10 yIndia have
Differences in human capital likely to be greater
than differences is technology - so predicted
interest rate differential smaller - but still
uncomfortably large.
22
Review of MRW specification
- steady state similar to original Solow
model - increases in investment rates of either
human or physical capital increase steady state
level of y, but not growth rate - Long run growth
rate of y determined by exogenous growth rate of
technical progress g - Empirically, generates
realistic values for capital share ? and the rate
of convergence ? - still predicts high interest
rate differentials between rich and poor countries
23
Some Concluding Remarks

• Solow model Simple and tractable framework,
  which allows us to discuss capital accumulation
  and the implications of technological progress.
• It shows us that if there is no technological
  progress there will be no sustained growth even
  with human capital.
• Generate per capita output growth, but only
  exogenously technological progress is a black
  box.
• Capital accumulation determined by the saving
  rate, the depreciation rate and the rate of
  population growth. All are exogenous.
• Need to dig deeper and understand what lies in
  these black boxes.

24
Next lecture

• Technology and Endogenous Growth
• Growth Accounting
• Endogenous Growth
• Learning by Doing and AK model