Table 3. The effect of investments on percent deforestation at the fokontany-level.

	Full sample	Only investments mapped to Fokontany	Only Fokontany that received any investment	Full sample (omitting 2007 data)	Only investments mapped to Fokontany (omitting 2007 data)	Only Fokontany that ever received an investment (omitting 2007 data)
Investments measured as binary value (no lag)
Marginal effect	-0.086	-0.006	-0.088	-0.106*	-0.063	-0.114*
Std Error	0.068	0.061	0.071	0.059	0.058	0.063
Total Observations	704	594	488	616	545	427
Total Fokontany	88	88	61	88	88	61
Investments measured as 2014 USD (no lag)
Marginal effect	-0.003	-0.001	-0.002	-0.009	-0.007	-0.008
Std Error	0.007	0.007	0.007	0.005	0.006	0.008
Total Observations	648	545	448	567	501	392
Total Fokontany	81	81	56	81	81	56
Investments measured as 2014 USD (2-year lag)
Marginal effect	-0.043**	-0.035**	-0.049**	N/A	N/A	N/A
Std Error	0.017	0.016	0.022	N/A	N/A	N/A
Total Observations	486	457	336	N/A	N/A	N/A
Total Fokontany	81	81	56	N/A	N/A	N/A
Prob>F-value	0.039	0.094	0.090	N/A	N/A	N/A

Note: Marginal effects from linear fixed effects regressions. Percent deforestation and 2014 USD were both log-transformed. Level of statistical significance reported as:

* p<0.10;

** p<0.05;

*** p<0.01.

N/A indicates that output with or without 2007 data is the same because data is lagged. Specifically, when estimating Eq 3, percent deforestation for the year 2007 is not included in Y_it, because we did not have data on investments prior to 2007 (e.g., 2006 investments). Information on 2007 investments were included in Eq 3 as a lagged independent variable for 2008 outcomes (1-year lag) and 2009 outcomes (2-year lag).