Table 1
The effect of international remittances on poverty
| Model | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) |
|---|---|---|---|---|---|---|---|---|---|
| OLS | OLS | OLS | FE | FE | FE | IV + FE | IV + FE | IV + FE | |
| DV | P0−1.9 | P1−1.9 | P1−3.2 | P0−1.9 | P1−1.9 | P1−3.2 | P0−1.9 | P1−1.9 | P1−3.2 |
| R | −0.950*** (0.052) | −0.362*** (0.020) | −0.673*** (0.039) | −0.575*** (0.156) | −0.139*** (0.037) | −0.430*** (0.118) | −1.012*** (0.196) | −0.475*** (0.100) | −0.665*** (0.144) |
| log_IC | −19.677*** (0.773) | −7.485*** (0.406) | −14.974*** (0.507) | −12.789*** (1.334) | −3.558*** (0.559) | −10.919*** (0.950) | −11.245*** (1.578) | −2.372** (0.896) | −10.090*** (1.031) |
| GN | 0.297*** (0.049) | 0.177*** (0.033) | 0.213*** (0.030) | 0.503*** (0.084) | 0.355*** (0.056) | 0.407*** (0.067) | 0.497*** (0.111) | 0.350*** (0.080) | 0.404*** (0.084) |
| INF | 0.009 (0.027) | 0.029** (0.012) | 0.009 (0.023) | 0.012 (0.026) | −0.004 (0.015) | 0.008 (0.022) | 0.052 (0.032) | 0.026 (0.019) | 0.029 (0.024) |
| Observations | 345 | 345 | 345 | 345 | 345 | 345 | 336 | 336 | 336 |
| Adjusted R2 | 0.814 | 0.729 | 0.835 | 0.535 | 0.400 | 0.609 | 0.362 | 0.056 | 0.490 |
| Hausman (γ) | 6.43 (0.000) *** | 15.32 (0.000) *** | 6.70 (0.000) *** | ||||||
| F-test of excluded instruments | 30.850 (0.000) *** | 30.850 (0.000) *** | 30.850 (0.000) *** | ||||||
| Kleibergen–Paap LM Test | 11.820 (0.003) *** | 11.820 (0.003) *** | 11.820 (0.003) *** | ||||||
| Hansen J test | 2.176 (0.140) | 0.075 (0.785) | 1.625 (0.202) | ||||||
| Endogeneity test | 4.347 (0.037)** | 9.311 (0.002) *** | 2.779 (0.096) * |
Notes: Driscoll–Kraay SEs are in parentheses under coefficient estimates. p-values are in square brackets. Per capita income (IC) is transformed into a logarithm because it is measured by units of money (USD). Hausman shows robust Hausman test results. Kleibergen–Paap LM statistics show robust underidentification test results. Hansen J statistics show robust overidentification test results. Robust endogeneity test results are shown by chi2-statistics. Some observations in Models 7–9 are dropped because singleton groups were detected by Stata. P0−1.9, poverty headcount ratio at USD 1.90 per day; P1−1.9, poverty gap at USD 1.90 per day; P1−3.2, poverty gap ratio at USD 3.20 per day.
DV, dependent variable; FE, fixed effect; IV, instrumental variable; LM, lagrange multiplier; OLS, ordinary least squares; SE, standard error.
Table 2
The effect of international remittances on poverty by controlling for political variable for Models 7–9 in Table 1
| Model | (7) | (8) | (9) |
|---|---|---|---|
| IV + FE | IV + FE | IV + FE | |
| DV | P0−1.9 | P1−1.9 | P1−3.2 |
| R | −1.005*** (0.206) | −0.471*** (0.104) | −0.661*** (0.153) |
| log_IC | −10.864*** (1.543) | −2.244** (0.908) | −9.844*** (1.012) |
| GN | 0.499*** (0.116) | 0.350*** (0.082) | 0.405*** (0.087) |
| INF | 0.040 (0.033) | 0.022 (0.020) | 0.022 (0.025) |
| POL | −1.489** (0.589) | −0.490 (0.351) | −0.954** (0.359) |
| Observations | 333 | 333 | 333 |
| Adjusted R2 | 0.377 | 0.064 | 0.500 |
| F-test of excluded instruments | 29.79 (0.000) *** | 29.79 (0.000) *** | 29.79 (0.000) *** |
| Kleibergen–Paap LM test | 11.330 (0.004) *** | 11.330 (0.004) *** | 11.330 (0.004) *** |
| Hansen J test | 0.757 (0.384) | 0.004 (0.949) | 0.689 (0.406) |
| Endogeneity test | 5.735 (0.017) ** | 9.351 (0.002) *** | 3.361 (0.067) * |
Notes: Driscoll–Kraay SEs are in parentheses under coefficient estimates. p-values are in square brackets. Per capita income (IC) is transformed into a logarithm because it is measured in units of money (USD). Kleibergen–Paap LM statistics show robust underidentification test results. Hansen J statistics show robust overidentification test results. Robust endogeneity test results are shown by chi2-statistics. Some observations are dropped because singleton groups are detected by Stata and lacked data for the political variable. Table only reports FE-IV estimation results. P0−1.9, poverty headcount ratio at USD 1.90 per day; P1−1.9, poverty gap at USD 1.90 per day; P1−3.2, poverty gap ratio at USD 3.20 per day; R, total remittance c as a share of GDP (current USD).
DV, dependent variable; FE, fixed effect; GDP, gross domestic product; GN, Gini ratio; INF, inflation; IV, instrumental variable; LM, lagrange multiplier; POL, political stability; SE, standard error.
Table 3
Results from one-step SGMM estimation
| Models | (1) | (2) | (3) | (4) | (5) | (6) |
|---|---|---|---|---|---|---|
| One-step SGMM | One-step SGMM | One-step SGMM | One-step SGMM | One-step SGMM | One-step SGMM | |
| DV | P0−1.9 | P1−1.9 | P1−3.2 | P1−1.9 | P1−1.9 | P1−3.2 |
| L1.P0−1.9 | 0.699*** (0.161) | 0.697*** (0.171) | ||||
| L1.P1−1.9 | 0.567*** (0.100) | 0.589*** (0.112) | ||||
| L1.P1−3.2 | 0.707*** (0.150) | 0.707*** (0.164) | ||||
| R | −0.374* (0.191) | −0.064 (0.066) | −0.308* (0.150) | −0.386** (0.180) | −0.080 (0.057) | −0.324** (0.147) |
| L1.R | −0.123 (0.129) | −0.015 (0.041) | −0.064 (0.099) | −0.124 (0.130) | −0.018 (0.043) | −0.062 (0.102) |
| log_IC | −4.068* (2.025) | −1.122** (0.537) | −4.028* (2.084) | −4.196** (1.772) | −1.245** (0.551) | −4.250** (1.972) |
| GN | 0.049 (0.051) | 0.046** (0.022) | 0.027 (0.043) | 0.048 (0.057) | 0.042** (0.020) | 0.026 (0.048) |
| INF | 0.014 (0.023) | 0.005 (0.005) | 0.004 (0.024) | 0.015 (0.025) | 0.007 (0.006) | 0.005 (0.026) |
| POL | −0.076 (0.606) | −0.046 (0.179) | 0.015 (0.596) | |||
| AR (2) | 0.61 (0.545) | 0.60 (0.546) | 0.27 (0.789) | 0.64 (0.523) | 0.60 (0.546) | 0.31 (0.757) |
| Hansen J test | 4.91 (0.672) | 4.09 (0.770) | 5.02 (0.658) | 3.15 (0.871) | 2.32 (0.940) | 3.99 (0.781) |
| Difference-in-Hansen test | 1.36 (0.851) | 4.27 (0.370) | 1.79 (0.775) | 2.99 (0.560) | 4.98 (0.290) | 4.28 (0.370) |
| Instruments | 27 | 27 | 27 | 28 | 28 | 28 |
| Observations | 209 | 209 | 209 | 209 | 209 | 209 |
Notes: Robust SEs are in parentheses. p-values are in square brackets. Per capita income (IC) is transformed into a logarithm because it is measured in units of money (USD). Hansen J statistics show the validity test of external instruments. Difference-in-Hansen statistics show the validity test of GMM-type instruments. Some observations are dropped because singleton groups are detected and lacked data for the political variable.
P0−1.9, poverty headcount ratio at USD 1.90 per day; P1−1.9, poverty gap at USD 1.90 per day; P1−3.2, poverty gap ratio at USD 3.20 per day; R, total remittance c as a share of GDP (current USD).
AR, autoregressive; DV, dependent variable; GDP, gross domestic product; GN, Gini ratio; INF, inflation; POL, political stability; SE, standard errors; SGMM, system-generalized method of moments; L1, first lag.
Table A1
List of countries
| Countries | Years | Countries | Years | Countries | Years |
|---|---|---|---|---|---|
| Albania | 2008, 2012 | Gambia | 2010, 2015 | Nepal | 2003, 2010 |
| Armenia | 2002–2016 | Georgia | 2003–2016 | Nicaragua | 2005, 2009, 2014 |
| Azerbaijan | 2005 | Ghana | 2012, 2016 | Niger | 2005, 2007, 2011, 2014 |
| Bangladesh | 2005, 2010, 2016 | Guatemala | 2006, 2014 | Nigeria | 2003, 2009 |
| Belarus | 2002–2016 | Honduras | 2010–2016 | Pakistan | 2004–2005, 2007, 2010–2011, 2013, 2015 |
| Benin | 2003, 2011, 2015 | India | 2004, 2009, 2011 | Peru | 2008–2015 |
| Bhutan | 2012 | Indonesia | 2002–2014 | Philippines | 2003, 2006, 2009, 2012 |
| Bolivia | 2002, 2004–2009, 2011 2016 | Iran | 2005, 2006, 2009, 2013–2016 | Russian Federation | 2010–2015 |
| Bosnia and Herzegovina | 2007, 2011 | Iraq | 2012 | Rwanda | 2005, 2010, 2013, 2016 |
| Botswana | 2009, 2015 | Jamaica | 2002, 2004 | Senegal | 2005, 2011 |
| Brazil | 2012–2016 | Jordan | 2002, 2006, 2008, 2010 | Serbia | 2012–2015 |
| Bulgaria | 2006–2014 | Kazakhstan | 2009–2016 | Sri Lanka | 2002, 2006, 2009, 2012, 2016 |
| Burkina Faso | 2003, 2009, 2014 | Kenya | 2005, 2015 | St. Lucia | 2016 |
| Burundi | 2013 | Kyrgyz Re-public | 2002–2016 | Tajikistan | 2009, 2015 |
| Congo, Dem. Rep | 2012 | Malawi | 2004, 2010, 2016 | Tanzania | 2007, 2011 |
| Costa Rica | 2002–2016 | Malaysia | 2004, 2007–2008, 2011, 2013, 2015 | Togo | 2006, 2011 |
| Cote d’Ivoire | 2002, 2008, 2015 | Maldives | 2009 | Tunisia | 2005, 2010, 2015 |
| Dominican Republic | 2002–2016 | Mali | 2006 | Turkey | 2005–2016 |
| Ecuador | 2008–2014 | Mexico | 2004–2005, 2006, 2008, 2010, 2012, 2014, 2016 | Uganda | 2002, 2005, 2009, 2012, 2016 |
| El Salvador | 2002–2016 | Moldova | 2002–2015 | Ukraine | 2002–2008, 2011–2015 |
| Ethiopia | 2015 | Morocco | 2006, 2013 | Zambia | 2006, 2010, 2015 |
| Gabon | 2005 | Namibia | 2009, 2015 |
Table A2
Descriptions of the variables
| Variables | Definitions | Source | Expected Sign | Mean | Std. Dev. |
|---|---|---|---|---|---|
| Dependent variables | |||||
| P0−1.9 | The percentage of the population that is living under 1.90 USD per day (constant 2011 USD adjusted for PPP) | WDI | 11.80 | 17.45 | |
| P1−1.9 | The percentage of mean shortfall in income from 1.90 USD per day (constant 2011 USD adjusted for PPP) | WDI | 4.06 | 7.18 | |
| P1−3.2 | The percentage of mean shortfall in income from 3.20 USD per day (constant 2011 USD adjusted for PPP) | WDI | 10.15 | 13.01 | |
| Explanatory variables | |||||
| R | Total remittance consisting of personal transfers and compensation of employees as a share of GDP (current USD) | IMF | Negative (−) | 7.22 | 7.97 |
| Control variables | |||||
| IC | Income (GDP) per capita (constant 2011 USD adjusted for PPP) | WDI | Negative (−) | 9,029.94 | 6,055.61 |
| GN | Gini Index varies between 0 (equality) and 1 (inequality) | WDI | Positive (+) | 39.32 | 8.27 |
| INF | Percentage change of consumer price index in annual average | WDI | Positive (+)/Negative (−) | 7.23 | 7.27 |
| Instrumental variables | |||||
| TOP | Total import and export as a share of GDP | WDI | Positive (+) | 77.79 | 31.39 |
| EX | LCU per USD | WDI | Positive (+) | 896.54 | 3,430.19 |
[i] Notes: Std. Dev, Standard deviation. Number of observations is 345 for all variables.
[ii] GDP, gross domestic product; IMF, International Monetary Fund; LCU, local currency unit; WDI, world development indicators.
Table A3
First-stage regression results for Models 7–9
| Models | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| DV | R | |||
| Excluded instruments | ||||
| TOP | 0.102*** (0.018) | 0.107*** (0.018) | 0.107*** (0.018) | 0.100*** (0.016) |
| log_EX | 0.126 (0.483) | 0.245 (0.497) | 0.226 (0.470) | 0.893 (0.214) |
| Included exogenous variables | ||||
| log_IC | 4.182** (1.985) | 4.021** (1.979) | 4.173** (1.752) | |
| GN | −0.021 (0.049) | −0.022 (0.049) | ||
| INF | 0.025 (0.027) | |||
| Observations | 336 | 336 | 336 | 336 |
Notes: Driscoll–Kraay SE are in parentheses. p-values are in square brackets. Per capita income (IC) and exchange rate (EX) transformed into logarithm because they are measured in units of money (USD) and local currency unit/USD, respectively). Some observations are dropped because singleton groups were detected.