Supplementary Appendix to the paper Income and by Markus Brueckner and Hannes Schwandt November 2013 downloadable from: https://sites.google.com/site/markusbrucknerresearch/research-papers
Table of Contents Table Page I List of Countries in Sample......1 II III Excluding Outliers...,.2 Using PWT 8.0 and WDI (2013) PPP GDP Data...3 IV Using 5-Year GDP Growth......4 V Rich vs. Poor Countries...5 VI Excluding Large Oil Importing Countries...6 VII Using Initial Net Exports Shares of Oil in GDP...7 VIII Oil Net-Export GDP Shares, Demographic Structures, and Income Inequality...8 XI Effects of the Oil Price Instrument on Female Labor Participation and Immigration...9 X Controlling for Female Labor Force Participation and Immigration...10 XI Pre- vs. Post-Fertility Transition Countries...11
Appendix Table 1. List of Countries in the Sample Afghanistan Croatia Ireland Norway Albania Cuba Israel Oman Algeria Cyprus Italy Pakistan Angola Czech Republic Jamaica Panama Argentina Denmark Japan Papua New Guinea Armenia Djibouti Jordan Paraguay Australia Dominican Republic Kazakhstan Peru Austria Ecuador Kenya Philippines Azerbaijan Egypt Kiribati Poland Bahrain El Salvador Korea, Republic of Portugal Bangladesh Equatorial Guinea Kuwait Qatar Barbados Estonia Laos Romania Belarus Finland Madagascar South Africa Belize France Malawi Spain Benin Gabon Malaysia Tajikistan Bolivia Gambia, The Mali Tanzania Bosnia and Herzegovina Georgia Malta Thailand Brazil Ghana Mauritania Togo Bulgaria Greece Mauritius Trinidad &Tobago Burkina Faso Guatemala Mexico Tunisia Burundi Guinea Mongolia Turkey Cambodia Guinea-Bissau Morocco Turkmenistan Cameroon Guyana Mozambique Uganda Canada Haiti Nepal Ukraine Central African Republic Honduras Netherlands United Arab Emirates Chad Hungary New Zealand United Kingdom Chile Iceland Nicaragua United States China India Niger Uruguay Colombia Indonesia Nigeria Uzbekistan Congo, Dem. Rep. Iran Russia Venezuela Congo, Republic of Iraq Rwanda Vietnam Costa Rica Latvia Samoa Yemen Cote d`ivoire Lebanon Senegal Zambia Ethiopia Libya Sierra Leone Zimbabwe Fiji Lithuania Slovenia 1
Appendix Table 2: Effects of Income Growth on (Excluding Outliers) (5) (6) 2SLS 2SLS 2SLS LS LS LS Excluding Top and Bottom 1st Percentiles of: Pop Growth GDP Growth Oil Shock Pop Growth GDP Growth Oil Shock 0.05** 0.06** (0.03) 0.06*** -0.00-0.00 Kleibergen Paap F-stat 17.27 27.66 30.16... Yes Yes Yes Yes -0.00 Observations 4337 4337 4337 4337 4337 4337 Note: The method of estimation in columns (1)-(3) is population-weighted two-stage least squares; columns (4)-(6) population-weighted least squares. The instrumental variable in columns (1)-(3) is the change in the international oil price between year t and t-10 multiplied with countries' average GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. Columns (1) and (4) exclude observations in the top and bottom 1st percentile of population growth. Columns (2) and (5) exclude observations in the top and bottom 1st percentile of GDP per capita growth. Columns (3) and (6) exclude observations in the top and bottom 1st percentile of the oil price shock. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 2
Appendix Table 3: Effects of Income Growth on (Robustness to Using PWT 8.0 and WDI (2013) PPP GDP Data) Panel A: PWT 8.0 PPP GDP Data 2SLS 2SLS 2SLS 2SLS Excluding Top and Bottom 1st Percentiles of: None GDP Growth Oil Shock Pop Growth 0.040*** (0.009) 0.098*** (0.037) 0.027*** (0.008) 0.028*** (0.008) Kleibergen Paap F-stat 45.56 14.97 41.82 44.35 Observations 4096 4014 4014 4014 Panel B: WDI 2013 PPP GDP Data 2SLS 2SLS 2SLS 2SLS Excluding Top and Bottom 1st Percentiles of: None GDP Growth Oil Shock Pop Growth 0.190** (0.081) 0.196** (0.083) 0.102* (0.057) 0.097* (0.058) Kleibergen Paap F-stat 13.36 13.07 11.97 11.96 Observations 2118 2076 2076 2076 Note: The method of estimation is population-weighted two-stage least squares. The instrumental variable is the change in the international oil price between year t and t-10 multiplied with countries' average GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. Column (2) excludes observations in the top and bottom 1st percentile of population growth. Column (3) excludes observations in the top and bottom 1st percentile of GDP per capita growth. Column (4) excludes observations in the top and bottom 1st percentile of the oil price shock. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 3
Appendix Table 4: Effects of Income Growth on (Robustness 5-Year ) [5-year Average] (5) (6) (7) (8) 2SLS 2SLS 2SLS 2SLS LS LS LS LS None 0.03*** (0.01) Pop Growth 0.02* (0.01) Excluding Top and Bottom 1st Percentiles of: GDP Growth 0.04** Oil Shock None Pop Growth 0.04* -0.01-0.01 GDP Growth -0.01 Kleibergen Paap F-stat 43.82 42.78 32.24 12.90.... Oil Shock Yes Yes Yes Yes Yes Yes Yes Yes -0.01 Observations 5119 5009 5009 5009 5119 5009 5009 5009 Note: The method of estimation in columns (1)-(4) is population-weighted two-stage least squares; columns (5)-(8) population-weighted least squares. The instrumental variable in columns (1)-(4) is the change in the international oil price between year t and t-5 multiplied with countries' average GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. Columns (2) and (6) exclude observations in the top and bottom 1st percentile of population growth. Columns (3) and (7) exclude observations in the top and bottom 1st percentile of GDP per capita growth. Columns (4) and (8) exclude observations in the top and bottom 1st percentile of the oil price shock. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 4
Oil Price Shock Appendix Table 5: Effects of Income Growth on (Rich vs. Poor Countries) 2SLS 2SLS 2SLS 2SLS Richest 50th Percentile of Countries 0.109*** (0.027) 0.533*** (0.141) Poorest 50th Percentile of Countries 0.029* (0.018) Excluding Richest 10th Percentile of Countries 0.044** (0.019) First Stage for 0.363*** (0.08) 0.366*** (0.089) Excluding Poorest 10th Percentile of Countries 0.056*** (0.021) 0.360*** (0.084) Observations 1964 2464 4047 3896 Countries 70 70 126 126 Note: The method of estimation is population-weighted two-stage least squares. The instrumental variable is the change in the international oil price between year t and t-10 multiplied with countries' average GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 5
Appendix Table 6: Effects of Income Growth on (Excluding Large Oil Importing Countries) [5-year Average] Kleibergen Paap F-stat 2SLS LS 2SLS LS 0.049** (0.022) 0.029*** (0.008) 0.035*** (0.011) 0.019*** (0.006) 12.17. 32.55. Observations 4124 4124 4475 4475 Note: The method of estimation in columns (1) and (3) is population-weighted two-stage least squares; columns (2) and (4) population-weighted least squares. The instrumental variable in column (1) is the change in the international oil price between year t and t-10 multiplied with countries' average GDP share of net oil exports; column (3) the change in the international oil price between year t and t-5 multiplied with countries' average GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. The excluded countries are China, France, Japan, Netherlands, South Korea, United States and the United Kingdom. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 6
Appendix Table 7: Effects of Income Growth on (Robustness to Using 1970 Net Export Shares and Restricting the Sample to the Post-1970 Period) [5-year Average] Kleibergen Paap F-stat 2SLS 2SLS 2SLS 2SLS All Countries 0.081*** (0.019) Excluding Large Oil Importers 0.080*** (0.022) All Countries 0.035*** (0.013) Excluding Large Oil Importers 0.039** (0.016) 18.87 16.47 16.94 14.96 Observations 3959 3639 4039 3719 Note: The method of estimation is population-weighted two-stage least squares. In columns (1) and (2) the instrumental variable is the change in the international oil price between year t and t-10 multiplied with countries' 1970 GDP share of net oil exports. In columns (3) and (4) the instrumental variable is the change in the international oil price between year t and t-5 multiplied with countries' 1970 GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. The excluded countries in columns (2) and (4) are China, France, Japan, Netherlands, South Korea, United States and the United Kingdom. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 7
Appendix Table 8. Oil Net-Export GDP Shares, Demographic Structures, and Income Inequality Panel A: Bivariate Correlations Share of Population Aged 0-14 Share of Population Aged 15-64 Gini Income Coefficient Oil Net-Export GDP Shares 0.076 0.004-0.057 Panel B: Descriptive Statistics Mean Median Sdv Min Max Share of Population Aged 0-14 0.357 0.394 0.101 0.130 0.500 Oil Exporting Countries Only 0.363 0.387 0.090 0.138 0.492 Oil Importing Countries Only 0.355 0.397 0.105 0.130 0.500 Share of Population Aged 15-64 0.583 0.567 0.066 0.470 0.822 Oil Exporting Countries Only 0.584 0.573 0.065 0.474 0.822 Oil Importing Countries Only 0.582 0.565 0.067 0.470 0.746 Gini Income Coefficient 0.438 0.438 0.092 0.210 0.674 Oil Exporting Countries Only 0.434 0.436 0.094 0.233 0.628 Oil Importing Countries Only 0.440 0.438 0.102 0.210 0.674 8
Appendix Table 9: Effects of Oil Price Instrument on Female Labor Force Participation, Employment, and Immigration Oil Price Instrument Female Labor Force Participation Rate Ratio Female-to-Male Labor Force Participation Ratio Female-to-Male Employment Immigration LS LS LS LS -1.06 (2.02) -0.018 (0.026) -0.029 (0.026) Observations 2138 2138 2138 932 2.25 (1.87) Note: The method of estimation is population-weighted least squares. The dependent variable in column (1) is the female labor force participation rate; column (2) the ratio of female-to-male labor force participation; column (3) the ratio of female-to-male employment; column (4) immigration. Huber robust standard errors (shown in parentheses) are clustered at the country level. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 9
Appendix Table 10: Effects of Income Growth on (Robustness to Controlling for Female Labor Force Participation and Immigration) (1) (2) (3) 2SLS 2SLS 2SLS Panel A: Controlling for Female LFP Immigration Female LFP and Immigration Female Labor Force Participation 0.048** (0.024) 0.032** (0.013) 0.048** (0.023) Immigration 0.139* (0.081) 0.042** (0.019) 0.004 0.337* (0.189) Kleibergen Paap F-stat 39.04 12.91 37.03 Time FE Yes Yes Yes Country FE Yes Yes Yes Observations 2137 922 372 Panel B: Without Panel A Controls But Same Sample 0.047** (0.024) 0.057*** (0.019) 0.040** (0.020) Kleibergen Paap F-stat 39.28 14.03 34.68 Time FE Yes Yes Yes Country FE Yes Yes Yes Observations 2137 922 372 Note: The method of estimation is population-weighted two-stage least squares. The instrumental variable is the change in the international oil price between year t and t-10 multiplied with countries' average GDP share of net oil exports. The dependent variable is total population growth. Huber robust standard errors (shown in parentheses) are clustered at the country level. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 10
Oil Price Instrument Appendix Table 11: Effects of Income Growth on Fertility (Pre- vs. Post-Fertility Transition Countries) ΔFertility Rate 2SLS 2SLS LS LS Pre-Fertility Transition Countries 1.02** (0.46) 0.41*** (0.10) Post-Fertility Transition Countries 1.22* (0.70) Pre-Fertility Transition Countries 1.38** (0.64) First Stage for 0.73** (0.30) Post-Fertility Transition Countries 0.43*** (0.14).. Observations 2741 1709 2741 1709 Countries 80 60 80 60 Note: The method of estimation in columns (1) and (2) is population-weighted two-stage least squares; columns (3) and (4) least squares. Columns (1) and (3) report estimates for pre-fertility transition countries. Columns (2) and (4) report estimates for post-fertility transition countries. Following Cervellati and Sunde (2011) we call a country pre-transitional if in 1960 the country's life expectancy is equal or below 50 years or the average crude birth rate is equal or above 30/1000; a country is classified as post-transitional if in 1960 the life expectancy is above 50 years and the average crude birth rate is below 30/1000. The instrumental variable in columns (1) and (2) is the change in the international oil price between year t and t-10 multiplied with countries' average GDP share of net oil exports. The dependent variable is the change in the fertility rate. Huber robust standard errors (shown in parentheses) are clustered at the country level. *Significantly different from zero at 10 significance, ** 5 percent significance, *** 1 percent significance. 11