Role of Construction Sector in Economic Growth: New Evidence from Turkey

After the 2001 financial crisis, Turkey's economy moved full-speed ahead, except for a temporary reversal in 2009 during the global financial crisis. Throughout the years of accelerated growth in 2002-2007 and 2010-2011, construction output increased at a faster rate than the economy as a whole, but in the periods of stagnation in 2001 and 2008-2009, construction industry was the first to suffer. This paper investigates the causal relationship between construction investments and economic growth in Turkey from 1998Q1 to 2014Q4. Unlike the previous studies that use two variable – real GDP and construction industry growth – Granger causality tests, the present study additionally employs three-variable – real GDP growth, construction industry growth, and real interest rate – VAR models to investigate the causal relationships in a multivariate setting. The paper also employs Zivot-Andrews test for determining structural breaks in data and then extends the causality analysis by dividing the seventeen-year sample period into smaller sub-periods that are defined according to the location of breaks in data. The results for the entire sample of 1998-2014 indicate that economic growth in Turkey has preceded construction activities with two- to four-quarters lags, but not vice versa. Hence, unlike the widespread belief that the construction plays a crucial role in Turkey’s economic growth, construction industry is not a driver of GDP growth but a follower of fluctuations in the macro-economy. However, our sub-sample analysis reveals that the causal relationship between economic growth and construction investments varies noticeably across the sub-periods in the national economy. We find that expansion in construction sector caused GDP growth over the last five years. The low interest rate environment with the help of radical changes in urban legislation and city building boosted up the construction industry, which resulted in economic growth in sub-period 2010-2014. Hence, we conclude that the temporary effect of construction industry growth on the GDP growth in the sub-period 2010-2014 is not justified for the overall sample period. Provided that much of the cyclicality in construction investment stems from the sector’s sensitivity to interest rates, we also find that there exists a bidirectional relationship between construction activities and real interest rates both for the entire sample period and for the sub-period 2002-2014. Lastly, construction activities have short-lived effects on the economic growth and thus cannot offer permanent solutions for the economic troubles in Turkey.