Transcript Emerging Income Inequality and Widening Economic Divide

Emerging Income Inequality and
Widening Economic Divide: the Case of
Sri Lanka
Upali Vidanapathirana
IDEAs
Conference on Policy Perspectives on Growth, Economic
Structures and Poverty Reduction:7-9, June, 2007
Tsinghua University, Beijing, China
The puzzle
Indicators
Country
category
Low middle
income
countries
Sri
Lanka
(2004)
$ 826-3255
$1010
Life
expectancy 58.1
(female)
69.8
74
Infant Mortality
79.1
24
13
Under 5 mortality
122.0
29.3
15
Adult Literacy
71.5
93.1
92
GNP per capita
/ Low
or income
countries
Les than $ 825
Introduction



Liberalization orthodoxy claims that
‘openness’ produces faster GDP growth
rate on a sustainable basis
Such growth engenders income equality
Sri Lanka’s liberalization experience of
about 30 years is adequate to test this
claim using context specific empirical
data
Theoretical base of inequality and ‘openness’



Inequality in ‘open
economies is explained in
terms of archetypical
Kuznet’ hypothesis where
initial spurt in GDP
growth increases
inequality (Kuznet, 1955)
Openness accelerates
growth rate but it is
distribution neutral
(Kraay, 2000)
HOSS theorem claims
that ‘openness’ increase
scope for jobs and
increases real wage rates
of unskilled workers to
cause equality.



Kuznet’s claims are
not archetypical;
inequality is harmful
to growth (Rodrik,
1996; Thorebecke,
2002)
Openness is not
distribution neutral
(Wood, 1999;
Anderson, 2005)
HOSS fails in many
countries to give jobs
to unskilled labour
(Carter and Barham,
1996; Lipton, 1985 nd
2007)
Proximate determinants of inequality
1
2
3
4
Economic growth is the most potent determinant of income distribution; its
distributive implication depends on whether the growth was pro-poor or pro-rich.
Growth performance of priority sectors become the second most potent
determinant. It tells us where the growth occur and whether it excludes some
sectors and communities. In the case of Sri Lanka the rural sector + estates
house about 77 percent of the population.
Employment generation is the third variable. If free trade serves unskilled
labour better, naturally that should be pro-poor.
The tenor of public policy is another crucial determinant; Fiscal contraction, for
instance, aggravate conditions of inequality. This includes pruning financial flows
to social welfare, physical and institutional infrastructure, and human
development.
5
5
6
Arising from 4 above, the status of infrastructure including roads, markets,
storage, irrigation canals, extension services etc undermines the distribution of
economic opportunities.
Education is considered an equalizer of income distribution; This depends on
considerations of access and equity. So does health!
Why reforms in Sri Lanka in 1977?

Reforms in SL was conveniently referred
to as ‘crisis driven’ although in practice it
was more of ‘crisis creating’ (The reasons for
reforms in Sri Lanka in 1977 much before its SA
counterparts include ostensibly to address problems of
low GDP growth, unemployment, poverty and
inequality and also to correct problems arising from
worsening economic fundamentals).


But in reality it was a political project of a
rightist government!
Has reforms in Sri Lanka result in crisis?
Many see a link between reforms and crisis in Sri
Lanka(Lakshman, 1996; GOSL, 1990; Jayasuriya,
2004).
Objectives of the paper



To evaluate the Sri Lankan case of economic
reforms (1977-2006) to ascertain whether the
reforms have produced desirable outcomes in
terms of equity and inclusiveness,
What specific factors have engendered and/ or
hindered such outcomes,
What are the lessons of experience Sri Lanka
provide to other developing countries
Methodological issues




The post-liberalization era is divided into two
phases i.e., the ‘first wave-1978-1988) and the
second wave (1989-2006);
The liberalization experience is compared with
the pre-reform era (1970-1977) to contrast the
development trajectories
Focuses on both income inequality and the
broader space of economic divide and to
identify some of the proximate determinants
Data sources include CFS (CBSL) and HIES
(DCS) series; here, one faces a huge challenge
as data are not consistent for a variety of
reasons. The paper uses other eclectic sources
of data too .
Income distribution by deciles
(1973-2003/04)




Decile distribution data are rather congested; Yet, they
show that income shares of the first two quintiles
declined relentlessly since 1978/79.
Conversely the top most deciles gained persistently.
The income share of the 1st decile fell from 2.79 in
1973 to 1.86 in 2004. The dividends of reforms have
by passed the poor.
The share of the 10th decile swelled from 28.03 in 1973
to 36.45 in 2004. The rich amassed the benefits of
reforms
Income inequality trends (income
data)
Quintiles
1973
1978/79 1981/82 1986/87 1996/97 2003/04
Bottom
7.2
5.7
5.7
5.0
5.7
5.1
2nd
12.1
10.3
9.5
9.1
10.0
9.1
3rd
16.2
14.3
13.3
13.5
14.1
13.4
4th
21.6
19.8
19.5
20.1
20.8
20.5
Top
42.9
49.9
52.0
52.3
49.4
52.0
Quintile distribution (1973-2004)


The gains to the top 40 percent have increased from 64
percent in 1973 to 72 percent in 2004.
However, what is more important perhaps is the
changes in the income share of the top 1 percent which
represent the ultra rich (For which we do not have
data).


Using the Mishra’s classification income share of 19.3
for the first 40 % in 1973 was permissible but this
condition deteriorated sharply during the post reform
period.
Figure 1- Lorenz Curves for years 1973, 1981 and 1996.
Source: Central Bank of Sri Lanka
Lorenz Curves for 1973-2004
100
90
Cumulative distribution of income
80
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
Cum ulative distribution of population
Line of perfect equality
Lorenz curve 1973
Lorenz curve 1981
Lorenz curve 1996
100
How Gini ratio changed over the period
1963-2004
Inequality Ratios based on CFS data 1963-2004
0.6
0.5
0.4
0.3
0.2
0.1
0
1963
1973
GINI coefficient for
1978/79
a)
Spending units
1981/82
b)
1986/87
Income receivers
1996/97
c)
2003/04
Decile distribution*
Other indices of income inequality
Description
1973
1978/79
1981/82 1986/87
1996/97
2003/4
0.35
0.41
0.43
0.50
0.45
0.52
0.46
0.52
0.43
0.48
0.46
0.50
(Q1+Q2)/Q5
Q5/Q1
0.45
5.9
0.32
8.7
0.29
9.1
0.27
10.4
0.32
8.6
0.27
10.1
Theil Index
0.28
0.35
0.39
0.39
0.33
0.38
Gini ratios
Spending units
Income receivers
Quintile Ratios
Discussion




Indicators are unanimous that income inequality has
worsen.
Inequality level was lowest in 1973; worst in 1986/87
(towards the end of the first wave). Second wave
marks signs of leveling off but this trend was reversed
again in 2004.
The figures quoted are comparatively worse and here
most of the SA countries perform much better. For
instance, Gini ratio is around 0.3-0.4 for all the SA.
The quintile ratio of 10 plus (meaning, the share of the
top income decile is about 10 times the bottom income
decile) compares with about 6 in India and Pakistan
Consumption data



Consumption data are generally supposed
to be better than the income data.
However, Sri Lanka’s consumption data
series is not as consistent as the income
data series of the DCS.
DCS data series became comparable
since 1990/91
Consumption Data
Quintiles
1980/81
1990/91
2002
Bottom
2nd
3rd
4th
Top
8.9
12.9
16.5
21.3
40.4
8.8
12.9
16.5
21.5
40.2
7.7
11.5
15.4
21.5
43.7
Cumulative % of Income for 1980 and 2002
Consumption data-Changing pattern of Gini ratios 1980-2002
100
90
80
70
60
50
40
30
20
10
0
1
2
3
4
5
Cumulative % households
Line of equality
1980 line
2002 line
6
Discussion of Income/expenditure
distribution trends

In the mid 1970s, poor gained both
absolute and relative terms (Fields, 1980).
These changes were directly policy driven.
The changes had far reaching
consequences (Land Reforms, 1972; Nationalization of
Plantation Companies, 1975; Ceiling on Housing property, 1973;
Compulsory savings for rich, 1994 and so on).

1977 reforms reversed this process (Tax
concessions, amnesties, stipulations to increase land rent for share
tenancy, privatization programme, removal of the universal food
subsidy, reduction of social welfare expenses, removal of fertilizer
subsidy, closure of PMB, closure of seed farms, elimination of state
monopolies, reduction of expenditure on irrigation and road
infrastructure and so on).
Widening the divide



Specific situations of reforms affecting
nutritional levels of adults in the first wave and
children under 5 in the second wave are
identified.
‘Drop-outs in education’ and the gaps in
educational quality by regions form another
divide.
Regional, sectoral and gender biases in terms
of income, health, educational and employment
opportunities cover another major problem
area.
Filters-GDP Growth



Growth rates falling and becoming
volatile
In booms the ‘top’ deciles gain but in
busts the ‘bottom’ loses
disproportionately .
Was the growth ‘pro-poor’? Who gain
and where?
Table 1.1-Economic growth as a filter
1970-1978
1978-1988
1989-2004
19702004
3.07
4.98
4.83
4.48
1.41
0.35 (1973)
2.95
0.46 (1987)
1.48
0.46(2004)
1.93
0.35-0.46
-22
31
0
31
Indicator/year
Average growth rate
Average rate of growth
of agricultural sector
Gini ratio
Reduction of income
inequality (%)[1]
Source: Computed using data from the annual report of the Central Bank
Negative figures indicate reduction of poverty (
Growth trends compared
Priority sectors and growth



In terms of distribution of income rural +
estate sectors matter a lot; 77 % of the
population and more than 90 % of the poor
dwell here
Agricultural income has fallen; minimum
procurement prices have fallen; credit to the
sector has fallen; irrigation investments have
fallen
Industrialization has centralized to the Western
province (which is the top gainer of reform
dividends); so does the booming services
sector
Table 1.2- Priority sectors (industrialization strategy)
1970-77
Indicators / Period
Annual growth rate of
Industrial GDP (deflated
using GDP deflator).
6
Annual growth rate of exports 14.4
(in US $)
3.06
Annual growth rate of GDP
1978-88 19892004
8.8
6.1
12.4
16.7
4.98
4.45
Spatial distribution of industrial units
Are food grains unproductive?
Falling real prices of food grains
Filters-Fiscal compression



Investments on physical and
institutional infrastructural have
fallen
Soft targets like education and
health sectors suffered the most
Subsidies including ‘food’ , fertilizer,
credit, are immediate casualties
% ag exp.
Ag exp.
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
Rs. millions (1996 price) / % ag
expenditure / Total Exp.
Falling public expenditure on
agriculture
Public expenditure in agriculture
60
50
40
30
20
10
0
Reemergence of Malaria in the second
wave
Impact of Financial reforms
Employment filter





‘Jobless growth’ – falling employment elasticity
of growth leaving many ‘un’- or ‘under’
employed.
Changes in the structure of employment
favouring unorganized/ informal/ casual
sectors.
Privatization leading to ‘lay-offs’ and
‘casualization’ of work
Wage levels falling in real terms
Conditions of work deteriorating
Employment as a Filter
Table 1.3 –Data on the elasticity of employment
Total
Employment
% Change
of GDP
% Change of
employment
1971 229
3649
4.09
1.08
1979 321
4647
40.1
27.34
Year
GDP (1996
factor cost
price)
Employment
elasticity of
agricultural
GDP*
Employment
elasticity of GDP*
1.77(0.22)
0.68 (0.08)
0.09(0.008)
1987 462
5271
43.92
13.43
0.30(0.03)
-0.16 (-0.01)
2003 929
6947
101
31.79
0.31 (0.02)
1.4-Percentage employed by status of Employment
1973 1978/79 1981/82
1986/97
1996/97
2003/04
60.9
-1.4
30.9
6.7
29.7
28.6
1.6
30.0
10.2
22.8
35.3
1.3
30.0
10.6
20.9
35.8
1.7
32.9
8.7
Category of
employment
Regular
Casual*
Employers
Self employed
Unpaid family workers
36.5
25.6
1.5
23.0
13.5
30.4
36.2
2.2
22.8
8.5
1.5-Dichotomies in weekly earnings by
organized and the unorganized sectors
Earnings for a Males in the
week (Rs)*
organized
private sector
Below 300
9.3 (9.3)
3001-600
30.9 (40.2)
601-1000
26.4 (66.4)
1001-2000
23.1 (89.7)
2001-3000
5.5 (95.2)
Over 3001
4.8
(100)
Females in the
organized
private sector
14.2 (14.2)
47.1 (61.3)
28.6 (89.9)
8.6 (98.5)
0.8 (99.3)
0.7
(100)
Males in the
unorganized
private sector
22.3 (22.3)
35.2 (57.5)
27.4 (84.9)
13.3 (98.2)
1.1 (99.3)
0.7
(100)
Females in the
unorganized
private sector
51.8 (51.8)
33.5 (85.3)
11.9 (97.2)
2.6 (99.8)
0.1 (99.9)
0.1 (100)
Inflation filter



Exchange rate-inflation link.
Inflation is specially bad for workers in the
agricultural and other unorganized sectors
(whose wage rates are not indexed).
It is also bad for producers whose bargaining
power is limited; it lowers internal terms of
trade for farmers when input costs rise at a
faster rate than output prices.
Exchange rate-inflation nexus
60
40
-40
-60
-80
-100
Infaltion
Bud defici
Ex rate %
Ex debt %
2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
-20
1973
0
1971
Percentage Change
20
Real wages falling for plantation
workers in the second wave
Overall welfare



Sri Lanka’s ‘initial welfare gains’ are lost;
she has failed to ‘increase the lead’ or
even to ‘maintain the lead’.
Other conditions of socio-political gains
also are fast disappearing
Suicides in the agricultural belts portray
this calamity.
Suicides in Sri Lanka
The lessons

Findings show that there is a marked divergence between the
theory and practice of liberalization orthodoxy.

In the case of Sri Lanka the dividends of reforms were ‘short-lived’
but the social costs were ‘deep’, ‘pervasive’, ‘long-drawn’.

‘Income inequality’ has undoubtedly increased; there is ample
evidence to show that the other forms of divides are widening.

These trends are directly linked to the reforms and various ‘filters’
of inequality also are identified.

Context specific data shows that reforms thus far have failed to
produce the promised results.

The dangers are not over. Pressure for ‘market driven’ reforms in
the ‘land, water, infrastructure, health and education markets’ are
building; they will remove the remaining safeguards and with that
‘inclusiveness’ will disappear from the skies of Sri Lanka forever.