Main Page                       Teaching                             E-Mail

Work in Progress

"A Framework for Indirect Inference", 2008 (joint with Paul L. Fackler)

Abstract:

A general framework for simulation based indirect inference methodology is proposed. The framework highlights the distinction between the structural (economic) model that describes  the data generating process and the reduced form (statistical) model used for estimation. The framework is illustrated with two simple examples. A MATLAB implementation is described and illustrated with worked examples.

Download Simulated GMM toolbox here .

"Feedbacks between capital flows and growth, interest and real exchange rate in Turkey : a frequency domain approach", 2008 (joint with Nuri Yildirim)

Abstract:

In this paper we study the interactions and feedbacks between capital flows and three macroeconomic variables, growth, interest and real exchange rates in Turkish economy for the 1992-2007 period using frequency domain techniques, mainly, a new version of causality test of Geweke (1982) and Hosoya (1991) in the frequency domain developed by Breitung and Candelon (2006). Besides, we make use of other tools of spectral analysis such as cospectrum, squared coherency and phase spectrum to decompose total covariance between capital flows and macroeconomic variables across main frequency bands and capture lead/lag interactions between them. Some of our empirical findings are as follows: Not all subcategories of capital inflows have positive contribution to growth in the business cycle frequencies. Capital flows and interest rate differential have counter-cyclical co-movement and the feedback from capital flows to spread has negative sign. 30-40% of the variation in real exchange rate fluctuations in low frequencies is attributable to the capital flows shocks.

"On the Validitiy of Three Hypotheses on Public Expenditures in Turkey: 1950-2004". 2008 (joint with Turan Yay)

 Abstract:

The share of government expenditures in GDP has displayed steady increase in both developed and developing countries during the 20^th century. This observation has led economists to explore the reasons and the underlying mechanism both theoretically and empirically. Several hypotheses on the relationship of public expenditures with income growth, budget deficits and government revenues have been reexamined in the light of recent developments in econometric methods. This study presents results from testing three hypotheses, namely, the Wagner Hypothesis, the Buchanan-Wagner Hypothesis and the Tax-Expenditure Hypothesis, using data from Turkey for the period 1950-2004. In the empirical section we employed time series econometric techniques to analyze long run economic relationships. Several unit root and cointegration tests are utilized to see the robustness of results across different methods. Empirical results indicate that there exists a long run relationship among variables for one version of the Wagner’s Law with Granger-causality running from government expenditures to national income, contrary to Wagner hypothesis. Government expenditures and budget deficits are found to be cointegrated in the long run but the causation runs from government expenditures to budget deficits, contradicting the Buchanan-Wagner conjecture. Moreover, empirical results indicate that government expenditures and revenues are cointegrated in both level and per capita versions but the Granger-causality runs from government expenditures to revenues supporting expenditure-tax hypothesis rather than tax-expenditure hypothesis.