The impact of public debt on economic growth in Serbia: An empirical study using the ARDL approach

The paper analyses the relationship between public debt and economic growth in Serbia. The empirical analysis was conducted using the autoregressive distributed lag (ARDL) and non-linear autoregressive distributed lag (NARDL) approaches for the period 2001-2023. The estimated long-run parameters in NARDL show that a 1% increase in the public debtto-GDP ratio is associated with a 0.25% decrease in real GDP per capita growth and that a 1% decrease in public debt is on average associated with a 0.7% increase in real GDP per capita growth, both with the first lag. This shows that real GDP per capita growth reacts asymmetrically to changes in public debt, such that the magnitude of the positive reaction to public debt reduction is greater than that of the negative reaction to an increase in public debt. The estimated short-term coefficients also confirm this asymmetry. The long-run coefficient of the public debt service parameter indicates the negative impact of debt service on economic growth, as a 1% increase in public debt service to GDP ratio is associated with a 0.17% decline in real GDP per capita growth. The long-run relationship between the current account deficit and economic growth is positive and shows that the increase in the current account deficit by 1% of GDP correlates with the increase in real GDP per capita growth by 0.26% with the first lag.