Abstract:

Empirical differences about indirect effects of public investments on private investments (crowding-in or crowding-out effect) reflect different theoretical ideas. The potential reason of these differences is that studies applied traditional unit root and cointegration tests which ignore structural breaks. Since, the power of these tests might decrease in the presence of structural breaks. For this purpose, in this study which analyzed the relationship between public investment expenditures and private fixed investments over 1970-2013 period in Turkey, Unit Root Tests taking into account the structural breaks such as Zivot - Andrews Unit Root Test (One Break) and CMR Unit Root Test (Two Break) was applied. Consequently, the vital differences between findings of traditional tests and tests with breaks were detected. ARDL Bounds Test which to construct in terms of these findings indicates that there is no long-run relationship between public investment expenditures and private fixed investments contrary to general findings in literature. On the other hand, Classical Regression Analysis indicates that public investment expenditures have exhibited crowding-in effect on private fixed investments in short-run.

Keywords: