A social discount rate for Italy

Abstract

Because of the budgetary constraints imposed by the Stability and Growth Pact, the need for cost and benefit evaluation of public investment has become increasingly relevant. In a cost-benefit analysis framework, the definition of the social discount rate is key to the selection of projects and programmes on the basis of their socio-economic return. To this extent, the Italian Ministry of the Economy has passively adopted the 5% rate proposed by the European Commission to evaluate projects financed by Structural Funds. In this article, we estimate a social discount rate for Italy, finding that a 3.7–3.8 rate would be appropriate, thus 1.2–1.3% lower than the official one.

Acknowledgements

The author wishes to thank David Evans for very insightful comments on earlier drafts of the present article. Financial support from Bocconi University and CERTeT is also gratefully acknowledged.

Notes

¹ The reader might want to refer to Evans and Sezer (2002) for further analytical details.

² For a theoretical analysis of the aggregation of individual time preference rates, see Gollier and Zeckhauser (2005).

³ See the demographic statistics at the website demo.istat.it (Tavole di mortalità 2001 e 2002).

⁴ Notice that the ratio T/Y measures the average tax rate.

⁵ See the website www.oecd.org/taxation

⁶ Data are from the national statistics compiled by Istat, which provides also seasonally-adjusted time series (see the web site con.istat.it).

⁷ Notice that the White test rejects the presence of heteroskedasticity, in addition we have performed a Chow test to detect the eventual presence of changing tastes and structural breaks in the parameters of interest. Results (not reported in the table) do not yield significant breaks over the considered time period. Notice that this would imply constant income and price elasticities.

⁸ Although results are not reported in the table, it should be stated that the F-statistics of Ramsey test for the CEM omitting the dummy variable for 1994 is 48.629, a value rejecting the hypothesis of correct functional specification at 99%.

⁹ The ratio of food consumption to total consumption equals 0.171.

¹⁰ Notice that the 5% value is simply the rule-of-thumb discount rate proposed by the European Commission (2002) and does not have any background empirical analysis, nor has it any strong supportive argument.