European structural funds during the crisis: evidence from Southern Italy

Table 8 The effects of European Structural Funds on annual growth, 2008–13. Heterogeneity with respect to slackness in the housing and labor markets

Annual 2008–13 growth in:	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
Annual 2008–13 growth in:	Employment			Population			House price per sqm
ln(annual pc payments)_t	0.0027	−0.0008	0.0010	−0.0006***	−0.0001	−0.0005**	0.0064**	0.0021	0.0035
ln(annual pc payments)_t	(0.0020)	(0.0020)	(0.0016)	(0.0002)	(0.0003)	(0.0002)	(0.0027)	(0.0022)	(0.0022)
ln(annual pc payments)_t × housing slack indicator	0.0012	0.0032	0.0020	0.0005*	0.0004	0.0007**	−0.0066*	−0.0022	−0.0064*
ln(annual pc payments)_t × housing slack indicator	(0.0025)	(0.0022)	(0.0021)	(0.0003)	(0.0004)	(0.0003)	(0.0036)	(0.0031)	(0.0034)
ln(annual pc payments)_t × labor mkt slack indicator	−0.0011	0.0049*	−0.0007	0.0002	−0.0002	0.0000	0.0100***	−0.0007	0.0076**
ln(annual pc payments)_t × labor mkt slack indicator	(0.0027)	(0.0028)	(0.0022)	(0.0003)	(0.0004)	(0.0004)	(0.0033)	(0.0027)	(0.0030)
Housing slack indicator		−0.0150	−0.0118		−0.0022	−0.0041**		0.0183	0.0381**
		(0.0100)	(0.0095)		(0.0018)	(0.0017)		(0.0142)	(0.0155)
Labor mkt slack indicator		−0.0214*	0.0057		0.0026	0.0006		0.0054	−0.0321**
		(0.0126)	(0.0099)		(0.0020)	(0.0018)		(0.0129)	(0.0142)
Additional controls	LLM FE	All f _i′;	Chosen by double selection	LLM FE	All f _i′;	Chosen by double selection	LLM FE	All f _i′;	Chosen by double x`selection
		f _i′ × t;			f _i′ × t;			f _i′ × t;
		f _i′ × t ²			f _i′ × t ²			f _i′ × t ²
Obs	1950	1950	1950	1950	1950	1950	1950	1950	1950
R2	0.2352	0.3415	0.2366	0.1550	0.6771	0.5973	0.3368	0.5191	0.3167

Note:
Standard errors clustered for LLM in parentheses. See Table 2 for other info. The housing slack indicator is a dummy for the lowest quintile of housing prices across all LLMs in 2007. The labor market slack is a dummy for the lowest quintile of employment rate in 2007. Columns (1), (4), (7) include only LLM FE, with no additional controls. To avoid introducing an additional source of variation, controls in columns (3), (6), (9) are the same selected for Table 2
*p < .10 **p < .05 ***p < .01