Belgian government investment, and specifically the part spent on infrastructure, is relatively low both in historical terms and compared to neighbouring countries. A simulation with the European Commission’s Quest III model suggests that increasing government investment permanently by 0.5% of GDP leads to a growth in GDP, private consumption and private investment. The impact of alternative financing mechanisms is compared. Finally, a budget neutral shift of investment in favour of infrastructure is found to yield significant benefits in terms of GDP and its main components already in the medium run.
In response to the long-lasting effects of the Great Recession on the potential growth of the Member States, the European Union considers the recovery in investment as a priority. The European Commission has operationalised this priority into the Investment Plan for Europe (IPE) aiming at fostering government and business investment in Member States. At the same time, pressures on the Member States have increased in the context of the European Semester with more Country Specific Recommendations (CSRs) related to investment. Belgium does not form an exception to this rule and the third recommendation addressed to Belgium by the Council of the European Union in July 2016 included: ‘(…). Address shortfalls in investment in transport infrastructure and energy generation capacity.’ The increase in government investment is however conditioned by the sustainability of public finances which imposes a sound path to the public deficit and debt-to-GDP ratios.
The Federal Planning Bureau has a long experience of simulating the impact of policy measure on the Belgian economy based on a demand-driven macro-econometric model, HERMES. In Bassilière et al. (2013), the impact of a permanent increase in public investment over an eight-year period was presented. However, in this simulation, the potential positive supply-side effect of public investment was not investigated. To take on board the impact of public investment on the production function of the private sector, another family of models should be mobilised.
This working paper examines the economic effects on the Belgian economy of a structural realistic increase in government investment, using the European Commission’s QUEST III model, slightly modified to better correspond to the current situation. It is found that beyond the short term positive impact on GDP through the demand channel, the increase in government investment also boosts the long-term growth through the positive impact on productivity and, to a lesser extent, on labour utilisation. The increase in government investment also crowds in private sector investment. The boost of potential output prevents overheating and inflation decreases after a limited, temporary swell. The fiscal impact in terms of public deficit rate is less than one-on-one, reflecting the positive effects of government investment increase on demand in the short run and on productivity in the long run. However, a tax is required to keep the debt-to-GDP ratio from rising. The increase in government investment is thus not fully self-financed (i.e. there is no fiscal free lunch).
The paper is organised as follows. After a first section devoted to the clarification of the different concepts usually used in the literature, the second section allows to underline the strong decrease in the Belgian government investment rate since the beginning of the seventies and the changes in the composition of this investment at the expense of infrastructure investment. In the third section, a brief survey of the relevant transmission channels identified by the literature is done. In the fourth section, a description of the QUEST model’s most important relationships is provided and calibration issues are discussed. The main results of a structural increase in the government investment by 0.5% of GDP are commented in the fifth section. Alternative simulations are also presented which explores the results of different financing modes and of a shift between different types of government investment. The sixth section presents the sensitivity analysis of the results in function of different values for important parameters of the model, such as the output elasticity of public capital, the elasticity of substitution of imports and domestic production or the proportion of liquidity constrained households. Finally, the last section concludes and discusses the analysis’ limitations as well as directions for further research.
The authors thank the teams Public finances and Sectoral and environmental accounts and analyses for their valuable help.