The impact of FDI inflows on real GDP in Estonia: Evidence from a cointegration approach and causality test

Theoretically, it is believed that FDI is a principal factor supporting and accelerating economic enhancement of an economy over the years. FDI boosts the productivity of host countries and promotes economic development, as FDI may not only provide direct capital financing, but may also create positive externalities via the adoption of foreign technology and know-how. In this paper we focus our attention on Estonia, as a (former) transition economy. Estonia stands out as one of the most successful ones in terms of economic growth and institutional changes among former command economies, as in less than two decades Estonia has been able to build functioning market-economy-institutions virtually from scratch and has joined the European Union (EU) and the Euro-zone. This paper, using cointegration methodology for the period of 1994:Q1-2013:Q2 focuses on the causal relationship between FDI and real GDP in Estonia. The real GDP data is obtained from Eurostat (2005 = 100) while the FDI data is obtained from the Bank of Estonia. Exploring the cointegration test is done through two methods: the residual-based test of Engle and Granger (1987) and Johansen’s methodology. The cointegration tests require that series to be integrated of order one. For this reason we first test for the existence of a unit root using Augmented Dickey Fuller unit root test (ADF, 1979, 1981), and Kwiatkowski et al. (KPSS, 1992). We further investigate the causal relationship using Granger Causality analysis and Vector Error Correction Model (VECM). The stationarity tests show that the variables are integrated of order 1, confirming that FDI and GDP are not stationary. The Engle-Granger method shows that FDI and real GDP are not cointegrated. Continuing with Johansen Cointegration Test, which is more generally applicable than the Engle-Granger method, the findings indicate that both the real GDP and FDI series are cointegrated, hence, the variables have long term relationship in that they will not deviate arbitrarily from each other and that their deviations from long run equilibrium path are corrected. Finally, Granger Causality test provides evidence that GDP does respond to changes in FDI in the long run but not in the short run, besides FDI does Granger cause real GDP. Our findings demonstrate that FDI can be very useful for Estonia, and it will be interesting to examine whether such findings are essentially due to large share of FDI in the banking sector and/or due to large share of FDI from high-income countries.