The effectiveness of fiscal spending in Croatia, Slovenia and Serbia: the role of trade openness and public debt level

Abstract

Various structural characteristics of economies, directly or indirectly, affect the transmission from government stimuli to economic activity and determine the size of fiscal multipliers. In this article, we expand the standard Blanchard–Perotti fiscal SVAR model by incorporating the public debt and trade openness variables to assess the influence of these structural determinants on the effectiveness of fiscal spending in three selected former Yugoslav countries – Slovenia, Croatia and Serbia. The results confirmed the main hypotheses, which state that public debt level and trade openness significantly affect the effectiveness of fiscal spending through the means of reduction in size of fiscal effects in all countries analysed. When comparing internationally, this reduction tends to be more evident in countries with a higher degree of average public debt level and trade openness.

Keywords:

• Fiscal spending
• economic fluctuations
• structural VAR

JEL classification:

• E62
• C32

Notes

1. Data were unavailable for Bosnia and Herzegovina, FYR Macedonia, Kosovo and Montenegro.

2. For the pros and cons of empirical versus model-based estimates, see Batini, Eyraud, and Weber (2014).

3. For the literature review on the estimation of the size of fiscal multipliers, based on different methods and created for different countries, see Ramey (2011) and Spilimbergo, Symansky, and Schindler (2009). For detailed methodology using SVAR, see Caldara and Kamps (2012) and Ilzetzki, Mendoza, and Végh (2013). For existing estimations of the fiscal multipliers estimations in emerging market and low-income economies, see Batini et al. (2014).

4. Later, in Perotti (2002), this model is extended by adding short-term interest rates and price levels.

5. According to the Global Competitiveness Report, labour market efficiency is between 4 and 4.2 in these countries. Furthermore, the transition process from planned to market economy is the slowest when it comes to labour market issues. Experience from other CEE countries shows that labour market conditions tend to improve slowly as the EU accession process proceeds (Schreiner, 2008).

6. For PVAR estimates that include CEE countries, see Hory (2014) and Ilzetzki et al. (2013).

7. Some of the mentioned articles use other approaches along with the BP identification method. Muir and Weber (2013) use the IMF’s Global Integrated Monetary and Fiscal Model (GIMF). Boiciuc (2015) and Karagyozova-Markova et al. (2013) also use the approach- and time-varying parameter VAR model. Fiscal multiplier estimations do not vary among different methods. Also, in a recursive approach, Karagyozova-Markova et al. (2013) include foreign demand in the list of endogenous variables because Bulgaria is a small open economy and external shocks have a strong effect on domestic output.

8. After an estimation of all the models presented in this section, the authors conducted model stability (inverse AR) and adequacy tests (autocorrelation and heteroscedasticity) which show that all analysed models are stable, with no violations of non-autocorrelation and homoscedasticity assumptions. All structural models are also just-identified. However, due to the size of the results (a total of 48 tables), the results of these tests are not included in this article but are available upon request.

9. In former Yugoslav countries, capital markets are generally shallow, illiquid and underdeveloped. In such conditions assets are less liquid and prices more volatile. The behaviour of interest rates may be difficult to explain due to a large number of factors affecting yield curves (Aljinović, Marasović, & Škrabić, 2008; Zoričić & Orsag, 2013). Furthermore, hard pegs and high euroisation influence central banks’ interest rates, which were and remain non-referent. For example, in Croatia, the central bank’s money issuing function was reduced to an instrument of foreign exchange auctions, while open-market operations – as the main instrument of modern monetary policy – were and are of secondary importance (Ćorić, Šimović, & Deskar-Škrbić, 2015).

10. One standard error interval is often used in determining the significance of the effects of fiscal policy in the SVAR framework. For a detailed explanation, see Šimović and Deskar Škrbić (2013, p. 69).

11. In the literature relating to fiscal multipliers, government spending is seen as effective if a one unit increase of government spending increases GDP by more than one unit. The same logic can be applied here.

12. But, it indicates that the effectiveness would be substantially limited.