Table 9.

Stability of Social Preferences in Surveys

Years	Variable	Correlation coefficient	Regression coefficient	# Obs.
2002 vs 2007	trust people in the world	0.064	0.068 (0.119) [0.585]	123
2007 vs 2009	trust people in the world	0.284** +	0.339* (0.176)	49
2002 vs 2007	trust people in the village	0.137	0.162* (0.096) [0.211]	123
2002 vs 2010	trust people in the village	0.440*** ++	0.425*** (0.132)++ [0.011]**	39
2007 vs 2009	trust people in the village	0.525*** +++	0.525*** (0.146)+++	49
2007 vs 2010	trust people in the village	0.254*** ++	0.206** (0.094)+ [0.052]*	119
2002 vs 2007	trust closest neighbors	0.273*** ++	0.275** (0.108)++ [0.013]**	123
2007 vs 2009	trust closest neighbors	0.463*** +++	0.545*** (0.159)+++	49
2002 vs 2007	bad to buy something you know is stolen	0.141	0.226* (0.123) [0.230]	123
2007 vs 2009	bad to buy something you know is stolen	0.372*** ++	0.353** (0.150)+	49
2002 vs 2007	would villagemates take advantage if had opportunity	0.251*** ++	0.167* (0.094) [0.108]	123
2007 vs 2009	would villagemates take advantage if had opportunity	0.355** ++	0.420** (0.190)+	49
2007 vs 2009	negative reciprocity	0.360** ++	0.236 (0.154)	49
2007 vs 2010	negative reciprocity	0.245*** ++	0.176* (0.099) [0.176]	119

Notes: Heteroskedasticity robust standard errors are in parentheses. Wild cluster bootrap p-values are in brackets.

Per-comparison significance: ***p <0.01,

^**p <0.05,

^*p <0.10.

FDR q-values: +++ q <0.01,

⁺⁺q <0.05,

⁺q <0.10 calculated for 14 hypotheses within table and column.

In regression column these are based on the heteroskedasticity robust standard errors. Controls in regressions include log income, sex, age, education, and village fixed effects.