Inequality and Development January 2011 () Inequality January 2011 1 / 23
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
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
Figure 4 Scatter of Gini against log(GDP) 0.7 0.6 Gini coefficient 0.5 0.4 0.3 0.2 5 6 7 8 9 10 11 log(GDP)
Table 6 Determinants of Inequality Variable No fixed effects Fixed effects 0.407 0.407 0.437 0.415 0.132 log(GDP) (0.090) (0.081) (0.078) (0.084) (0.013) log(GDP) squared -0.0275 -0.0251 -0.0264 -0.0254 -0.0083 (0.0056) (0.0051) (0.0049) (0.0053) (0.0014) Dummy: net -- -0.0493 -0.0480 -0.0496 -0.0542 (0.0094) (0.0087) (0.0094) (0.0108) income or spending Dummy: -- -0.0134 -0.0143 -0.0119 -0.0026 individual vs. (0.0086) (0.0080) (0.0087) (0.0078) household data -- -0.0147 -0.0152 -0.0161 -0.0025 Primary schooling (0.0037) (0.0036) (0.0037) (0.0091) Secondary -- -0.0108 -0.0061 -0.0109 -0.0173 (0.0070) (0.0070) (0.0070) (0.0099) schooling -- 0.081 0.072 0.082 0.102 Higher schooling (0.034) (0.032) (0.034) (0.030) Dummy: Africa -- 0.113 0.135 0.113 -- (0.015) (0.016) (0.015) -- 0.094 0.089 0.092 -- Dummy: Latin Amer. (0.012) (0.012) (0.012) Rule-of-law index -- -- -0.040 -- -- (0.019) -- -- -- -0.003 -- Democracy index (0.015) Number of 49, 61 40, 59 40, 57 35, 59 36, 56 observations 68, 76 61, 70 56, 67 61, 70 57, 59 0.12, 0.15 0.52, 0.59 0.50, 0.58 0.56, 0.59 -- R-squared 0.18, 0.22 0.67, 0.67 0.78 0.72 0.67, 0.67
Figure 5 Gini Coefficient versus log(GDP) 0.2 0.2 Gini coefficient (unexplained part) 0.1 0.1 0.0 0.0 -0.1 -0.1 -0.2 -0.2 -0.3 -0.3 5 5 6 6 7 7 8 8 9 9 10 10 11 11 log(GDP)
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
Table 3 Growth regression 1960–1992 with income and land inequality Ž . All countries Developing a countries Intercept 2.614 1.346 2.949 2.379 4.738 3.389 4.246 3.906 2.94 1.40 4.12 2.39 4.47 2.17 2.93 1.51 Ž . Ž . Ž . Ž . Ž . Ž . Ž . Ž . Investment 0.132 0.122 0.134 0.123 0.107 0.115 0.130 0.148 6.15 5.09 6.38 4.77 4.68 4.00 3.94 3.59 Ž . Ž . Ž . Ž . Ž . Ž . Ž . Ž . Initial GDP y 0.302 y 0.205 y 0.288 y 0.264 y 0.308 y 0.248 y 0.301 y 0.338 3.70 2.23 4.39 3.49 4.50 3.06 1.39 1.54 Ž . Ž . Ž . Ž . Ž . Ž . Ž . Ž . Income Gini y 0.047 y 0.019 y 0.025 y 0.019 y 0.018 y 0.045 2.80 0.95 1.34 0.86 0.60 1.27 Ž . Ž . Ž . Ž . Ž . Ž . Land Gini y 0.034 y 0.022 y 0.037 y 0.027 y 0.039 y 0.053 4.07 1.95 3.85 2.09 2.43 2.10 Ž . Ž . Ž . Ž . Ž . Ž . Latin Dummy y 0.530 y 0.432 0.018 2.765 0.85 0.87 0.03 1.83 Ž . Ž . Ž . Ž . Africa Dummy y 0.214 y 0.254 0.324 2.191 0.32 0.46 0.46 1.52 Ž . Ž . Ž . Ž . Asia Dummy 1.320 0.668 0.798 1.882 2.32 1.36 1.46 1.51 Ž . Ž . Ž . Ž . 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 a Only 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.
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
Table 1 Panel Regressions for Growth Rate Independent variable Estimated coefficient Estimated coefficient in full sample 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 R 2 0.67, 0.48, 0.42 0.73, 0.62, 0.60
Table 2 Panel Regressions for Investment Ratio Independent variable Estimated Coefficient Estimated Coefficient in full sample 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 R 2 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.
Table 4 Effects of Gini Coefficients on Growth Rates and Investment Ratios Gini Gini* Gini Gini Wald log(GDP) (low (high tests (p- GDP) GDP) values) Growth rate regressions 0.000 (0.018) -0.331 0.043 0.059 (0.141) (0.018) -0.033 0.054 0.011, (0.021) (0.025) 0.003* fertility variable omitted -0.037 (0.017) -0.367 0.043 0.012 (0.156) (0.020) -0.036 -0.036 0.085, (0.018) (0.034) 0.99* Investment ratio regressions 0.060 (0.070) 0.54 -0.062 0.39 (0.47) (0.062) fertility variable omitted -0.027 (0.066) 0.50 -0.068 0.51 (0.48) (0.062)
Figure 1 Growth Rate versus Gini Coefficient 0.10 growth rate (unexplained part) 0.05 0.00 -0.05 -0.10 0.2 0.3 0.4 0.5 0.6 0.7 Gini coefficient
Figure 2 Growth Rate versus Gini Coefficient (taking account of Gini-log[GDP] interaction) 0.10 growth rate (unexplained part) 0.05 0.00 -0.05 -0.10 0.2 0.3 0.4 0.5 0.6 0.7 Gini coefficient
“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
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
Agricultural sector Consists of landowners and labourers. ! income of a labourer is w ( L A ) , where w 0 ( L A ) < 0 , ! rent per hectare of a landowner is π A ( L A ) , where π 0 , A ( L A ) > 0 () Inequality January 2011 8 / 23
Recommend
More recommend
