Skip to main content

Table 3 The effects of European Structural Funds on average 2008–13 growth, controlling for LLM time invariant characteristics

From: European structural funds during the crisis: evidence from Southern Italy

 

(1)

(2)

(3)

(4)

(5)

(6)

 

Average 2008–2013 growth in:

 

Employment

Population

House price per sqm

ln(cumulative pc payments 07–13)

−0.0001

−0.0035

0.0002

−0.0002

−0.0118**

−0.0093**

(0.0026)

(0.0026)

(0.0007)

(0.0007)

(0.0049)

(0.0044)

Controls selected by the double selection procedure:

ln(allocated per capita funds)

−0.0156

0.0003

0.0021

0.0004

0.0124

0.0072**

(0.0126)

(0.0021)

(0.0038)

(0.0006)

(0.0293)

(0.0032)

fraction of surface composed of municipalities on the coast

−0.0017

 

0.0004

0.0017***

0.0112**

 

(0.0022)

 

(0.0007)

(0.0005)

(0.0049)

 

ln(house price per sqm)2005

0.0541***

0.0051***

0.0001

 

0.0149

 

(0.0182)

(0.0016)

(0.0054)

 

(0.0309)

 

ln(employment)2006

0.1356

 

−0.0471

0.0005***

1.6895***

 
 

(0.2104)

 

(0.0549)

(0.0002)

(0.3555)

 

ln(house price per sqm)2006

0.0012

 

0.0012

 

−0.1604***

−0.0186***

(0.0158)

 

(0.0051)

 

(0.0334)

(0.0035)

ln(house price per sqm)2007

−0.0664**

 

0.0257***

0.0008

0.0241

 

(0.0284)

 

(0.0083)

(0.0005)

(0.0915)

 

population growth 2004-07

−0.3647

0.2179***

0.0946

0.2158***

−0.4706

 

(1.3937)

(0.0323)

(0.4192)

(0.0106)

(3.0035)

 

Additional controls

All remaining vars in f i

No additional controls

All remaining vars in f i

No additional controls

All remaining vars in f i

No additional controls

Obs

325

325

325

325

325

325

R2

0.5031

0.2719

0.8396

0.7771

0.4836

0.1145

  1. Note:
  2. The unit of observation is the LLM over time. See Table 1 for data sources. The regressions include a constant. Robust standard errors in parentheses. The average growth over 2008–13 is calculated as a geometric average. See Table 2 for the full list of covariates. Controls in columns (2), (4), (6) have been selected using the “double selection” of Belloni et al. (2014) and the code provided by the authors. Columns (1), (3), (5) include all f i ′, but only coefficients on those that are also “double selected” are shown (full regressions table are available from the authors)
  3. *p < .10 **p < .05 ***p < .01