World Income Inequality: past, present and future'

1
Lecture 10

• World Income Inequality past, present and
  future.
• Read Outline to Chapter 11.

2
The Inequality of Nations

• The richest to poorest ( at 400 ) nation
  indicator tells you about the opportunities lost
  in poor nations
• It reflects an increasing gap between
  technological and institutional potential and the
  consequences of not being able to absorb advanced
  technology because of malfunctioning
  institutions, poor educational standards and bad
  government and poor advice from international
  aid agencies?
• Most of the Divergence Big Time is a post 1800
  phenomenon but you can trace the beginning back
  to c.1500.

3
(No Transcript)
4
Is our wealth based on exploitation of the poor?

• Not much, but
• Rich worlds agricultural protectionism lowers
  world market price for poor peasants world wide
• International division of labour has given poor
  countries exportable commodities with high price
  volatility
• Foreign investments in poor countries smaller
  than expected
• Major problem is unlimited supply of labour
  which pushes wages down to subsistence and
  translates technological progress into falling
  commodity prices and potentially falling terms of
  trade for the poor relative to the rich world,
  the W. A. Lewis hypothesis

5
The sources of inequality

• Inequality of personal income is linked to
• natural talent
• skill acquired by education and on the job
  training
• accumulated wealth
• inherited wealth
• market imperfections
• discrimination

6
Measuring Inequality

• Standard dispersion measures will do such as,
  standard deviation or coefficient of variation
• We will concentrate on the so-called
  Gini-coefficient
• The intuition behind G it measures the deviation
  of observed income distribution from the ideal
  state of perfect equality

7
The Gini explored
8
Interpretation of the numbers

• A Gini coefficient of 1 means absolute or perfect
  inequality, total income is held by the richest
  person.
• As the coefficient get smaller than 1 the economy
  gets less unequal. At zero we have perfect
  equality everybody earns the average income
• Ginis are not consistently used as fractions of 1
  but often as the fraction, say, 0.5 multiplied by
  100 50.
• Next table indicates considerable inequality in
  medieval European cities which reflects a
  sophisticated division of skills and labour as
  well as unequal distribution of wealth

9
(No Transcript)
10
The inequality possibility frontier

• Absolute inequality is not possible since there
  is a minimum subsistence income.
• Assume that income to be 400PPP (of 1990).
• If 99.9 per cent of the population earns that
  income and the rich 0.1 per cent extracts the
  rest we get an increasing inequality possibility
  frontier as average income rises.

11
A curve of maximum inequality
12
Implications

• Rich contemporary nations are far away from the
  maximum inequality even if the Ginis are similar
  to those of Roman and the Byzantine Empires.
• Pre-industrial and Early Modern (16th and 17th
  centuries) economies were more unequal that
  modern economies.

13
Inequality in the modern world
14
Why do trends differ?

• Un-weighted inequality increases each nation is
  represented by its average GDP per head and all
  nations have the same weight 1
• Population weighted inequality controls for the
  fact that nations differ in population size, the
  larger the nation is the larger weight its GDP
  per head gets

15
World income inequality paradox

• Population weighted inequality falls because
  many of the fast growing economies, such as China
  and India, have large populations
• But early modernizing economies have skill
  shortages increasing the so called Kuznets
  inequality
• As a consequence global inequality which
  controls for changes in each nations inequality
  has not fallen

16
Kuznets at work
17
Long term trends

• Next figure follows the long run trends in
  un-weighted (concept 1)and population weighted
  inequality
• Increase in pop. weighted inequality (concept 2)
  when a small number of economies join the growth
  club but decrease when a large number of
  economies join the growth club by 1950 and
  converge to income levels of the richest nations

18
(No Transcript)
19
China makes a difference!
20
Predicting the future from the past

• Next table estimates observed transition
  probabilities, for example the probability of
  getting from low initial income to high growth,
  from medium level initial income to high growth
  into high growth etc.
• Assuming that these probabilities are stable a
  Markow process predicts a stable long run outcome

21
Conclusion

• Inequality will fall but it will not disappear
• The very poor nations will fall in numbers and
  the very rich will increase
• But the estimates are based on historical
  transition probabilities and these probabilities
  will change
• A nation with a low transition probability to
  high growth might change institutioins and policy
  to improve chance

22
Back to the Future
23
The Lucas simulation

• Assume that all economies can sooner or later
  get into the growth club
• The number of economies joining the club is
  initially small, then increases until all
  nations are in the club
• The larger the income gap to the leading
  economies the faster will initial growth be for
  the newcomer
• In the long run all economies will grow at the
  same rate which is lower than the newcomer rate
• From these assumptions the evolution of average
  world growth rate and income dispersion can be
  predicted
• World average growth rates will reach a maximum
  and then fall in the future and so will (
  un-weighted) income inequality

24
More equality in the future!
25
Conclusion

• World inequality has probably peaked.
• When emerging economies expand their educational
  systems skill shortages will ease and domestic
  inequality will fall.
• Modern economies have Ginis similar to the Roman
  empire but lower than Earl modern Europe but they
  are far from their maximum inequality.

26
(No Transcript)
27
(No Transcript)