[PPT] - Inequality and Development January 2011 () Inequality January PowerPoint Presentation

SLIDE 1

Inequality and Development

January 2011

() Inequality January 2011 1 / 23

SLIDE 2

SLIDE 3

SLIDE 4

SLIDE 5

SLIDE 6

Inequality Data

Historical data on income shares of top 20% population relative to bottome 40 % Early estimates of Gini coe¢cent by Jain (1975) “High quality” Gini coe¢cients over income and land distribution , ! …rst compiled, analyzed and discussed by Deininger and Squire (1998)

() Inequality January 2011 2 / 23

SLIDE 7

The Evidence — Inequality and Per Capita Income

Kuznets (1955) — inverted U–shaped relationship , ! con…rmed in cross–country data (Summers, Kravis and Heston, 1984) , ! true for several developed economies with long historical time series BUT little time–series evidence to support Kuznets hypothesis in LDCs. , ! Fields and Jakubsen (1994) Deininger and Squire (1998) Barro (2001): , ! Kuznets curve emerges as a clear empirical regularity in panel data (after controlling for other factors) , ! per capita income does not account for much of the variation in inequality across countries or over time.

() Inequality January 2011 3 / 23

SLIDE 8

0.2 0.3 0.4 0.5 0.6 0.7 5 6 7 8 9 10 11

Gini coefficient

log(GDP)

Scatter of Gini against log(GDP) Figure 4

SLIDE 9

Table 6 Determinants of Inequality Variable No fixed effects Fixed effects log(GDP) 0.407 (0.090) 0.407 (0.081) 0.437 (0.078) 0.415 (0.084) 0.132 (0.013) log(GDP) squared

0.0275

(0.0056)

0.0251

(0.0051)

0.0264

(0.0049)

0.0254

(0.0053)

0.0083

(0.0014) Dummy: net income or spending

0.0493

(0.0094)

0.0480

(0.0087)

0.0496

(0.0094)

0.0542

(0.0108) Dummy: individual vs. household data

0.0134

(0.0086)

0.0143

(0.0080)

0.0119

(0.0087)

0.0026

(0.0078) Primary schooling

0.0147

(0.0037)

0.0152

(0.0036)

0.0161

(0.0037)

0.0025

(0.0091) Secondary schooling

0.0108

(0.0070)

0.0061

(0.0070)

0.0109

(0.0070)

0.0173

(0.0099) Higher schooling

0.081

(0.034) 0.072 (0.032) 0.082 (0.034) 0.102 (0.030) Dummy: Africa

0.113

(0.015) 0.135 (0.016) 0.113 (0.015)

Dummy: Latin

Amer.

0.094

(0.012) 0.089 (0.012) 0.092 (0.012)

Rule-of-law index
0.040

(0.019)

Democracy index
0.003

(0.015)

Number of
bservations

49, 61 68, 76 40, 59 61, 70 40, 57 56, 67 35, 59 61, 70 36, 56 57, 59 R-squared 0.12, 0.15 0.18, 0.22 0.52, 0.59 0.67, 0.67 0.50, 0.58 0.78 0.72 0.56, 0.59 0.67, 0.67

SLIDE 10

0.3
0.2
0.1

0.0 0.1 0.2 5 6 7 8 9 10 11

Gini coefficient (unexplained part)

log(GDP)

0.3
0.2
0.1

0.0 0.1 0.2 5 6 7 8 9 10 11

Gini Coefficient versus log(GDP) Figure 5

SLIDE 11

Evidence — Inequality and Productivity Growth

Negative impact of inequality on subsequent growth in per capita income in long term Persson and Tabellini (1994) , ! historical data for 9 developed countries (from 1850 at 20 year intervals) , ! post-war cross-sectional observations for 56 countries (1960-1985) Alesina and Rodrik (1994) , ! negative impact of land inequality using 70 countries (1960–1985) Perotti (1996) — robustness to di¤erent measures of inequality and speci…cations Deininger and Squire (1998) con…rm using “high quality” data set

() Inequality January 2011 4 / 23

SLIDE 12

Table 3 Ž . Growth regression 1960–1992 with income and land inequality All countries Developing

a

countries Intercept 2.614 1.346 2.949 2.379 4.738 3.389 4.246 3.906 Ž . Ž . Ž . Ž . Ž . Ž . Ž . Ž . 2.94 1.40 4.12 2.39 4.47 2.17 2.93 1.51 Investment 0.132 0.122 0.134 0.123 0.107 0.115 0.130 0.148 Ž . Ž . Ž . Ž . Ž . Ž . Ž . Ž . 6.15 5.09 6.38 4.77 4.68 4.00 3.94 3.59 Initial GDP y0.302 y0.205 y0.288 y0.264 y0.308 y0.248 y0.301 y0.338 Ž . Ž . Ž . Ž . Ž . Ž . Ž . Ž . 3.70 2.23 4.39 3.49 4.50 3.06 1.39 1.54 Income Gini y0.047 y0.019 y0.025 y0.019 y0.018 y0.045 Ž . Ž . Ž . Ž . Ž . Ž . 2.80 0.95 1.34 0.86 0.60 1.27 Land Gini y0.034 y0.022 y0.037 y0.027 y0.039 y0.053 Ž . Ž . Ž . Ž . Ž . Ž . 4.07 1.95 3.85 2.09 2.43 2.10 Latin Dummy y0.530 y0.432 0.018 2.765 Ž . Ž . Ž . Ž . 0.85 0.87 0.03 1.83 Africa Dummy y0.214 y0.254 0.324 2.191 Ž . Ž . Ž . Ž . 0.32 0.46 0.46 1.52 Asia Dummy 1.320 0.668 0.798 1.882 Ž . Ž . Ž . Ž . 2.32 1.36 1.46 1.51 R2 adj 0.3781 0.468 0.549 0.564 0.550 0.547 0.576 0.585

No. Obs.

87 87 64 64 55 55 27 27

aOnly developing countries with a population of more than two million have been included.

Here and in all subsequent tables, figures in brackets denote t-values.

SLIDE 13

Positive relationship between inequality and growth in the short–term Forbes (2000) uses panel data for a cross–section of countries (5 year intervals) , ! positive relationship at 5 year intervals , ! no signi…cant relationship at 10 year intervals , ! negative relationship in cross–section Barro (2001) …nds weak overall impact of inequality on growth in panel data (10 year intervals) , ! negative relationship for poorer countries and a positive relationship for richer ones. , ! depends on inclusion of fertility rate — negative once this is dropped

() Inequality January 2011 5 / 23

SLIDE 14

Table 1 Panel Regressions for Growth Rate Independent variable Estimated coefficient in full sample Estimated coefficient in Gini sample log(per capita GDP) 0.124 (0.027) 0.103 (0.030) log(per capita GDP) squared

0.0095 (0.0018)
0.0082 (0.0019)
govt. consumption/GDP
0.149 (0.023)
0.153 (0.027)

rule-of-law index 0.0172 (0.0053) 0.0102 (0.0065) democracy index 0.054 (0.029) 0.043 (0.033) democracy index squared

0.048 (0.026)
0.038 (0.028)

inflation rate

0.037 (0.010)
0.014 (0.009)

years of schooling 0.0072 (0.0017) 0.0066 (0.0017) log(total fertility rate)

0.0251 (0.0047)
0.0306 (0.0054)

investment/GDP 0.059 (0.022) 0.062 (0.021) growth rate of terms of trade 0.165 (0.028) 0.124 (0.035) numbers of observations 79, 87, 84 39, 56, 51 R2 0.67, 0.48, 0.42 0.73, 0.62, 0.60

SLIDE 15

Table 2 Panel Regressions for Investment Ratio Independent variable Estimated Coefficient in full sample Estimated Coefficient in Gini sample log(per capita GDP) 0.188 (0.083) 0.121 (0.111) log(per capita GDP) squared

0.0110 (0.0053)
0.0077 (0.0070)
govt. consumption/GDP
0.271 (0.072)
0.353 (0.104)

rule-of-law index 0.064 (0.020) 0.070 (0.025) democracy index 0.072 (0.078) 0.047 (0.123) democracy index squared

0.086 (0.068)
0.057 (0.103)

inflation rate

0.058 (0.027)
0.022 (0.028)

years of schooling

0.0013 (0.0058)

0.0045 (0.0065) log(total fertility rate)

0.0531 (0.0140)
0.0592 (0.0187)

growth rate of terms of trade 0.052 (0.067) 0.129 (0.114) numbers of observations 79, 87, 85 39, 56, 51 R2 0.52, 0.60, 0.65 0.35, 0.64, 0.69 Notes: The dependent variable is the ratio of real investment (private plus public) to real

GDP. The measure is the average of the annual observations on the ratio for each of the

periods 1965-75, 1975-85, and 1985-95. See the notes to Table 1 for other information.

SLIDE 16

Table 4 Effects of Gini Coefficients on Growth Rates and Investment Ratios Gini Gini* log(GDP) Gini (low GDP) Gini (high GDP) Wald tests (p- values) Growth rate regressions 0.000 (0.018)

0.331

(0.141) 0.043 (0.018) 0.059

0.033

(0.021) 0.054 (0.025) 0.011, 0.003* fertility variable

mitted
0.037

(0.017)

0.367

(0.156) 0.043 (0.020) 0.012

0.036

(0.018)

0.036

(0.034) 0.085, 0.99* Investment ratio regressions 0.060 (0.070) 0.54 (0.47)

0.062

(0.062) 0.39 fertility variable

mitted
0.027

(0.066) 0.50 (0.48)

0.068

(0.062) 0.51

SLIDE 17

0.10
0.05

0.00 0.05 0.10 0.2 0.3 0.4 0.5 0.6 0.7

growth rate (unexplained part)

Gini coefficient

Growth Rate versus Gini Coefficient Figure 1

SLIDE 18

0.10
0.05

0.00 0.05 0.10 0.2 0.3 0.4 0.5 0.6 0.7

growth rate (unexplained part)

Gini coefficient Figure 2 Growth Rate versus Gini Coefficient (taking account of Gini-log[GDP] interaction)

SLIDE 19

“Income Distribution, Market Size and Industrialization”

Murphy, Shleifer and Vishny (1989)

“Demand side theory” of the interaction between inequality and development Formalizes idea of industrialization as a self–reinforcing “big push” in response to increased agricultural productivity or trade , ! initial rise in income allows …rms to overcome …xed costs ! pro…t multiplier e¤ect For this to occur, the distribution pro…ts must be su¢ciently equal , ! need large enough “middle class” so that market size is big enough.

() Inequality January 2011 6 / 23

SLIDE 20

Assumptions

“Potential” manufactured goods indexed by q 2 (0, ∞), and ordered by decreasing marginal utility. L households with varying income y that supply a unit of labour, and buy 1 or 0 units of each good Households have a maximum food requirement given by z , ! if y z, household just buys food , ! if y > z, spends z on food, then spends the rest on industrial goods , ! given price of industrial goods is p, household buys all goods up to ˆ q, where z + p ˆ q = y.

() Inequality January 2011 7 / 23

SLIDE 21

Agricultural sector

Consists of landowners and labourers. , ! income of a labourer is w(LA), where w 0(LA) < 0 , ! rent per hectare of a landowner is πA(LA), where π0

A(LA) > 0

() Inequality January 2011 8 / 23

SLIDE 22

Manufacturing sector

Traditional technology: , ! small scale, no …xed costs, labour productivity = 1/α < 1 , ! pro…ts: πT = p l α wl. , ! competition ) p = αw. Modern technology: , ! large scale, …xed labour costs C, labour productivity = 1 , ! monopolistic …rms set “limit price”: p = αw , ! pro…ts: π = pl wl wC = w[(α 1)l C] ) modern technology introduced only if demand is large enough: l C α 1.

() Inequality January 2011 9 / 23

SLIDE 23

Labour is mobile between sectors: w = w(LA) The distribution of shares among households is given by G(γ). , ! some households (the lower classes) own no shares, γ = 0, , ! the rest (the middle and upper classes) owns at least ¯ γ shares , ! total number of shareholders is N = L(1 G( ¯ γ)). If a household owns γ …rm shares, it also owns γ shares of land. , ! household incomes are given by y = w if γ = 0 w + γ(π + πA) if γ > ¯ γ

() Inequality January 2011 10 / 23

SLIDE 24

Equilibrium

Suppose that all sectors up to Q industrialize. , ! zero pro…ts in sector Q, and # of households that buy this good is N = C α 1. , ! the demand for the Qth good comes from the N richest households (the upper class), where N = L(1 G(γ)). , ! Q is determined by z + pQ = w + γ(π + πA) Q = w + γ(π + πA) z αw

() Inequality January 2011 11 / 23

SLIDE 25

The demand from these N richest housholds pays for the …xed costs in every sector q Q. , ! additional demand coming from the middle class ( ¯ γ γ < γ) translates into pure pro…ts For q > Q , demand does not cover the …xed costs ) traditional technologies. It can be shown that the pro…ts of modern …rms are given by π = α 1 α

[(π + πA)Γ (N N)(z w)] ,

where Γ = R γ

¯ γ γdG(γ)

() Inequality January 2011 12 / 23

SLIDE 26

SLIDE 27

Solving for π yields the MM-Curve: π = α1

α

[πA(LA)Γ (N N)(z w)] 1 Γ α1

α

.

Equilibrium in agriculture: w(LA)LA + πA(LA) = zN + w(LA)(L N) , ! this AA-Curve pins down LA and is not a¤ected by π

() Inequality January 2011 13 / 23

SLIDE 28

SLIDE 29

SLIDE 30

Implications

Agriculture as the leading sector: , ! increase in agricultural productivity ! increase πA for any given LA. , ! AA-Curve shifts left , ! MM-Curve shifts up — depends on the size of the multiplier e¤ect , ! “big push” to industrialization. Inequality and the Failure of Industrialization: , ! if fewer than N households own all claims to pro…ts and rents (i.e. the middle class is small), then demand cannot cover the …xed costs. ) all manufacturing production remains in the traditional sector. ) high inequality countries may fail to industrialization, even if there is a boom in agriculture (e.g. Argentina)

() Inequality January 2011 14 / 23

SLIDE 31

SLIDE 32

“Enterprise, Inequality and Economic Development”

Lloyd–Ellis and Bernhardt (2000)

“Supply-side theory” of interaction between wealth distribution, credit constraints and the distribution of talent Generates an equilibrium development process like that described by Lewis (1954) and Fei and Ranis (1966). Shows how income inequality endogenously traces out a Kuznets curve

() Inequality January 2011 15 / 23

SLIDE 33

Assumptions

Homogeneous preferences: u(Ct, Bt+1) = C α

t B1α t+1

Individuals inherit wealth (e.g. land) b ! endogenous distribution Gt(b) Individuals endowed "talent”: set–up cost x ! exogenous distribution H(x) Given (b, x), individuals must make occupational choice , ! entrepreneur , ! wage laborer , ! subsistence agriculture.

() Inequality January 2011 16 / 23

SLIDE 34

Workers receive wt, and invest their inheritance, earning rb. , ! subsisters receive γ. , ! urban cost of living: ν 0. ) reservation wage for workers: w = γ + ν Entrepreneurs who pay x produce according to f (kt, lt).

() Inequality January 2011 17 / 23

SLIDE 35

Borrowing limited by a moral hazard problem: , ! to borrow L entrepreneurs put up their inheritance b as collateral. , ! after production they can abscond, lose rb but gain rL. , ! apprehended with probability p, ! punishment imposes disutility d , ! borrowers would renege if βrb + pd < βrL ) maximum loan size = b + ∆, where ∆ = pd/βr. Net pro…ts earned by a type (b, x) entrepreneur: π(b, x, wt) = max

kt,lt f (kt, lt) wtlt r(kt + x)

s.t.

kt b + ∆ x.

() Inequality January 2011 18 / 23

SLIDE 36

Occupational Mapping , ! start–up cost of an agent with b who is just willing and able to become an entrepreneur: z(b, wt) = min[b + ∆, xm(b, wt)] where π(b, xm, wt) = wt , ! z(b, wt) is increasing and concave in b. , ! as wealth increases, number of entrepreneurs increases, but at a decreasing rate. , ! redistribution from rich to poor increases the supply of entrepreneurs.

() Inequality January 2011 19 / 23

SLIDE 37

Fig. 1. Occupational choice map.

SLIDE 38

Dynamics: Dual economy phase

At t = 1 agents with su¢ciently low x migrate, become entrepreneurs and employ additional agents , ! surplus labor remains in the rural sector and wage settles at w. , ! in generation t = 2 the distribution of inheritances grows in the …rst–order stochastic sense. , ! more agents become entrepreneurs and on a greater scale. , ! output rises and the distribution of income grows in the …rst–order stochastic sense. When wages are low, ine¢cient projects are worth undertaking if su¢cient capital can be employed , ! wealth (not e¢ciency) is the primary determinant of occupational choice , ! wealth inequality persists across generations

() Inequality January 2011 21 / 23

SLIDE 39

Investment

5 10 15 20 25 Generation 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 F r a c t i

n
f

O u t p u t Start-Up Capital Working Capital

Firm Size Distribution

5 10 15 20 25 Generation 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 C

n

s u m p t i

n

U n i t s Optimal Firm Size Average Firm Size Variance

Kuznets' Curve

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Per Capita Income 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 G i n i C

e

f f i c i e n t

Wages and Per Capita Income

5 10 15 20 25 Generation 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 C

n

s u m p t i

n

U n i t s Per Capita Income Wage

Occupations

5 10 15 20 25 Generation 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 F r a c t i

n
f

P

p

u l a t i

n

Farmers Entrepreneurs Wage Laborers

Income Shares

5 10 15 20 Generation 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 F r a c t i

n
f

V a l u e A d d e d Profits Wage Income

1.Dynamics of the Economy

SLIDE 40

Advanced economy phase

Competition drives the equilibrium wage above w, , ! pro…ts fall, and income and wealth inequality start to decline. , ! past wage increases e¤ectively transfer wealth from the rich to poor , ! supply of entrepreneurs and their demand for labor continue to rise. , ! wages are bid up further, output expands and the distributions of wealth and income grow in the second–order stochastic sense. As wages rise, less e¢cient agents prefer wage labouring and production becomes increasingly e¢cient , ! gradually, e¢ciency replaces wealth as the primary determinant of

ccupational choice

, ! wealth inequality becomes less persistent across generations

() Inequality January 2011 22 / 23

SLIDE 41