Presentazione di PowerPoint

1
Territorial Estimates and the Evaluation of
Public Investment Carla Carlucci Pietro
Cova Piacenza, 13 maggio 2005
I. Literature overview II. The model III.
Estimation IV. Data set V. Results VI.
Conclusions and ongoing work
2
Territorial Estimates and the Evaluation of
Public Investment
Goal Measurement of the Linkages between Local
Development and Public Investments Literature
Review Aschauer (1989, AER), Morrison and
Schwartz (1996, AER) Aggregate Production
Function and Aggregate Cost Function
approaches Rosen (1974, JPE) and Roback (1982,
JPE) Hedonic Prices and Compensating Variations
literature
3
Territorial Estimates and the Evaluation of
Public Investment
Theory Develop a spatial equilibrium model with
representative households and firms that interact
in fixed geographic areas, as in Gyourko and
Tracy (JPE 1989, 1991) and Haughwout (JPubE,
2002). Optimal individual behaviour (households
maximize utilities, firms maximize profits) leads
to equilibrium values for wages and rents given
binding individual and local constraints V(w,r
s) ? k C(w,rs) ? 1 Where s is a vector of local
variables amenities, fiscal attributes,
financial development, social capital, and public
investments
4
Territorial Estimates and the Evaluation of
Public Investment
Compensating variations and public investments A
graphical exposition
sgts
E.g. s Public Transportation
Case a improvements in public transportation are
nonproductive or the benefits to firms are of
second order importance Case b benefits to
firms are of first order importance
5
Territorial Estimates and the Evaluation of
Public Investment
Econometric Specification
Where HV housing value HQ vector of housing
unit structural traits HC vector of individual
human capital traits LF use and availability of
financial instruments (local financial
development) Z vector of community amenity,
fiscal attributes and public infrastructure
stocks
6
Territorial Estimates and the Evaluation of
Public Investment

• Econometric Estimation
• OLS ?N(0,?2?) and ?N(0,?2?)
• Random Effects ?ij ?j ?i where ? j N(0,?2?)
  and ?i N(0,?2?) and similarly for ? ij ?j ?i
  
• Under B we allow the error term to contain both
  an individual and a city component to capture
  local omitted group effects.

7
Territorial Estimates and the Evaluation of
Public Investment

• Construction of the Data Set Main Sources
• Survey of Households and Wealth (SHIW) contains
  detailed information on italian households and
  individuals (income, wealth, education, industry
  and occupation controls, etc.).
• ISTAT and Picci (2001) data on public investment
  stocks.
• ISTAT-Demos data set on social variables
  (health, culture, crime) for Italian provinces
  (NUTS 3 level).
• Guiso, Sapienza, and Zingales (2004) data set on
  social capital and local financial development
  (main source SHIW).
• Ragioneria Generale dello Stato data on local
  fiscal conditions.

8
Summary statistics for individual housing unit
structural traits 1989, 1991, 1993, 1995
Summary statistics for individual and individual
human capital traits 1989, 1991, 1993, 1995
9
Summary statistics for indicators of local
financial development 1989, 1991, 1993, 1995
Summary statistics for community amenity traits
1989, 1991, 1993, 1995
10
Summary statistics for fiscal attributes, and the
stock of public infrastructures (logs) 1989,
1991, 1993, 1995
11
Econometric Results Rents (endogenous variable)
12
Econometric Results Rents (endogenous variable)
Econometric Results Wages (endogenous variable)
13
Territorial Estimates and the Evaluation of
Public Investment
Some comments on the results

• Estimates have generally the expected signs
  beneficial local traits have positive signs,
  bads have negative signs
• Effects of public infrastructures are mainly
  reflected in rent prices (could also suggest
  problems associated to wage equation
  specification)
• Effects are robust to random effects
  specification
• Conservative estimates if benefits of local
  infrastructures may spill over into neighbouring
  areas

14
Territorial Estimates and the Evaluation of
Public Investment
Conclusions and ongoing work

• Want to associate public investment effects to
  individual local areas assessment of public
  infrastructures effects on local economic
  activity over time
• Introduce variables associated to effectiveness
  of local administrations (ongoing internal
  research project at UVER)
• Augment estimation by introducing spatial
  econometric techniques (e.g. check for spatial
  autocorrelation)
• Use estimation results to formulate a consistent
  dynamic local development model in the spirit of
  New Economic Geography literature (Fujita et al.,
  2002).