Skip to main content
. Author manuscript; available in PMC: 2017 Feb 1.
Published in final edited form as: World Dev. 2016 Feb 1;78:541–553. doi: 10.1016/j.worlddev.2015.10.022

Table 3.

Fixed and Random Effects Panel Income Regressions: Dependent Variable is the Natural Logarithm of Real Household Income Per Person

Fixed
Effects		 Random
Effects
Age of HH Head 0.018
(0.014)		 0.011
(0.010)
Age of HH Head2 −0.023 *
(0.013)		 −0.011
(0.010)
Number of Dependents −0.166 ***
(0.021)		 −0.180 ***
(0.015)
Number of Working Age HH Members −0.031
(0.023)		 −0.019
(0.017)
Female-Headed HH (=1) 0.097
(0.088)		 0.060
(0.056)
Education of HH Head 0.006
(0.011)		 0.011
(0.007)
Avg. Education of HH Members 0.020
(0.014)		 0.050 ***
(0.010)
ln(Land Area) 0.179 ***
(0.039)		 0.175 ***
(0.029)
Agricultural Capital Index 0.019
(0.013)		 0.013
(0.011)
Business Capital Index 0.052 ***
(0.014)		 0.071 ***
(0.013)
Near Open Trade Area (=1) 0.000
(0.043)		 0.041
(0.037)
Pct. of Workers in Agriculture 0.001
(0.001)		 −0.001 *
(0.001)
Pct. of Migrant Workforce 0.004 ***
(0.001)		 0.004 ***
(0.001)
Jan-Mar Rainfall Shock 0.060
(0.047)		 0.009
(0.041)
Apr-Jun Rainfall Shock 0.212 **
(0.088)		 0.297 ***
(0.085)
Jul-Sep Rainfall Shock 0.306 **
(0.124)		 0.298 **
(0.118)
Oct-Dec Rainfall Shock −0.134
(0.085)		 −0.215 ***
(0.082)
No. Observations
Households (n): 375 375
Periods (T): 6 6
Total Observations (N): 2250 2250
R2 (within) 0.38 0.37
R2 (between) 0.27 0.39
R2 (overall) 0.35 0.37
*

Note: Significant at 10% level

**

Significant at 5% level

***

Significant at 1% level. Standard errors in parentheses. In the fixed effects model, standard errors have been adjusted according to the method proposed by Arellano (1987). This adjustment results in standard errors that are robust to the existence of heteroskedasticity and/or serial correlation in the disturbances. Standard errors in the random effects model have been adjusted using a variant of White’s heteroskedasticity-robust variance-covariance matrix, where the error variances are restricted to be the same for all observations derived from a particular household.