Explaining public investment in Western Europe

Abstract

Budgetary consolidations are considered the obvious explanation for the decline in public investment that most Western European countries experienced over the past three decades. However, regressions based on budgetary variables tend to overpredict public investment during the post-1990 period, i.e. when the budgetary stress eased. We supplement the budgetary consolidation approach to public investment with ideas from behavioural economics to explain why these investments do not increase when additional budgetary resources are available. We use the peak/end evaluation procedure to capture the frustration of voters as cuts in government consumption expenditures accumulate. This ‘memory-effect’ of budgetary consolidations implies that voters recall the previous peak in government consumption expenditures. They remain discontent as long as current expenditures are below the peak value. When the budgetary situation improves, policy makers will choose to increase government consumption because this is electorally more rewarding. Public investment will thus decline when budgetary consolidations are imposed and will remain constant when additional budgetary resources emerge. We test for a memory-effect by introducing expenditure gaps in public investment regressions. These gaps equal the difference between the highest previously observed primary government consumption to Gross Domestic Product (GDP) ratio and the current ratio. The regression results for most EU countries support our assumption.

Notes

¹ See, for example, Berndt and Hansson (1991), Ford and Poret (1991), Conrad and Seitz (1994), Evans and Karras (1994), Baltagi and Pinnoi (1995), Sturm and De Haan (1995, 1998), Vanhoudt et al. (2000), Clark et al. (2001) and Mehrotra and Välilä (2005).

² See, for example, Lybeck (1986), Lybeck and Henrekson (1988) and Gemmell (1993).

³ See, for example, International Monetary Fund (2004), Perée and Välilä (2005) or Välilä et al. (2005) for a general discussion of the evolution of public investment in different countries.

⁴ Välilä et al. (2005), p. 26.

⁵ See De Haan et al. (1996) for a tabular presentation of the econometric studies that was available in the mid-1990s. See also Lybeck (1986) and Gemmell (1993) for surveys of the theories of the growth of the government

⁶ See Lybeck and Henrekson (1988), Chapter 3, for explanations about how the studies relate to the theories explaining the growth of governments.

⁷ See also Sturm (1998), Chapter 3.

⁸ See also Sturm (1998), Chapter 2.

⁹ In Perée and Välilä (2005) only the regression results with significant coefficients are given; Mehrotra and Välilä (2005) report the complete results.

¹⁰ See summary table in Giudice et al. (2003), p. 7.

¹¹ One can link this view to the behavioural finance result that holds that investors will keep their stocks until previous losses are offset. This effect is known as the disposition-effect; investor are then said to suffer from ‘get-evenitis’ (Shefrin, 2000, Chapter 9).

¹² Note that anchoring is also an element of behavioural decision making.

¹³ For expository reasons we will not systematically repeat that the primary consumption expenditures considered exclude public investment. This budgetary item is eliminated so as to avoid spurious correlation when the expenditure gaps are introduced as explanatory variables in public investment regressions.

¹⁴ This is an approximation to the well-known utility function in behavioural finance that expresses prospect theory whereby the part that reflects the utility associated with gains is replaced by the abscissa.