Fiscal multipliers in Slovenia, Croatia and Serbia: A bucket approach vs. empirical estimation

The main objective of the paper is to estimate the size of fiscal multipliers in selected Western Balkan countries: Croatia, Slovenia and Serbia, using a narrative ‘bucket approach’ developed by the International Monetary Fund (IMF) (Batini et al., 2014). When estimating the size of the fiscal multipliers, one has to consider various structural characteristics of the economies which, directly or indirectly, affect the transmission from the government stimuli to the economic activity. Due to the fact that selected countries are small open economies we assume that shocks of foreign origination can have notable effects on its performance. The majority of papers dealing with dynamic effects of (discretionary) fiscal policy on economic activity are based on empirical estimations i.e. structural vector autoregressive (SVAR) models. For empirical research, the main obstacle in the research identification was the lack of data i.e. quarterly data are not available through the sufficient period to include more explanatory and control variables. That is the main reason why generally little is known about the size of the fiscal multipliers in developing and low-income economies. Because of data availability and empirical estimations limits, this paper is focused on a narrative “bucket approach”. Further we compare results with existing empirical estimations and offer some conclusion about pros and cons of both models given structural characteristics of selected countries. The results show that the government spending has a positive and statistically significant effect on output, and net (indirect) taxes have a negative and statistically significant effect on output, with Croatia having the highest spending multiplier and Slovenia the lowest one. Also, results show insignificant differences between bucket approach estimates compared with existing empirical estimates, suggesting that rather simple narrative approaches like the “bucket approach” can be used equally well for estimation of fiscal multipliers in small, open and transition economies.