Table 3 Tobit regressions for intervention impact on the pursuit of specific solar products

From: Behavioral savings sessions increase the pursuit of solar products among refugees in Uganda

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

 

Panel battery components

Lantern: no charging

Lantern: with charging

Solar home system

Weekly solar savings: VSLA

Weekly solar savings: Other

Weekly solar savings: Total

Treatment VSLA

0.03*

0.10***

0.09***

0.13***

1888***

1104***

2120***

 

(0.02)

(0.02)

(0.03)

(0.03)

(262.08)

(328.89)

(295.93)

Replacement

−0.02

−0.01

−0.09**

−0.06*

−740.6**

−700.9**

−942.1**

 

(0.02)

(0.03)

(0.04)

(0.03)

(326.14)

(347.97)

(384.02)

VSLA controls

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Individual control

Yes

Yes

Yes

Yes

Yes

Yes

Yes

SEs clustered VSLA

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Observations

1186

1186

1186

1186

1161

749

1184

  1. Each column presents the results of Tobit regressions for dependent variables on the binary treatment assignment variable. Dependent variables are binary in columns 1–4 and continuous in 5–6. Results are ITT estimates due to imperfect intervention compliance. All regressions control for VSLA-level controls (a settlement dummy, a dummy for VSLAs that meet at least weekly, the number of members in the VSLA, and the VSLA’s share price (normalized to weekly)), and individual-level controls (dummy for at least primary education). Robust standard errors clustered at the VSLA level are displayed below coefficients in parentheses. Asterisks denote a statistically significant difference at the 1% ***, 5% **, or 10% * levels.
Back to article page