Structural Breaks in Long-Term Turkish Macroeconomic Data,1923-2003

. This study examines the time series properties of long-term Turkish macroeconomic data by using the sequential Dickey-Fuller type and the minimum Lagrange multiplier (LM) type endogenous break(s) unit root tests. Zivot and Andrews (1992) and Lumsdaine and Papel (1997) tests do not provide evidence against the unit root hypothesis, indicating the shocks are permanent. On the contrary, Lee and Strazicich (2003 and 2004) minimum LM unit root tests strongly reject the null hypothesis in favour of trend stationarity with two breaks in particular. Eleven macroeconomic variables, namely, real GNP, nominal GNP, real per capita GNP, employment, GNP deflator, consumer prices, money stock – M1 and M2, velocity, export, and import series are used in the study. The data used are annual observations extracted from the State Institute of Statistics (SIS) publication of Statistical Indicators: 1923 – 2002, except for the nominal, real, and per capita GNP data, which have been revised as outlined above. The sample period starts as early as in 1923 (when Republic of Turkey was founded) and ends in 2002. Classification-J