Multivariate Adaptive Regression Splines (Mars) Method For Unemployment in OECD Countries

KOÇ H., DÜNDER E., KOÇ T.

Erciyes Üniversitesi Fen Bilimleri Enstitüsü Dergisi, cilt.35, sa.3, ss.46-51, 2019 (Hakemli Dergi) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 35 Sayı: 3
  • Basım Tarihi: 2019
  • Dergi Adı: Erciyes Üniversitesi Fen Bilimleri Enstitüsü Dergisi
  • Derginin Tarandığı İndeksler: TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.46-51
  • Ondokuz Mayıs Üniversitesi Adresli: Evet

Özet

Unemployment is one of the most important macroeconomic problemsin all countries and it is very important task for identification of the keydeterminants of it. Therefore, in recent years determining the factors affecting theunemployment is attracting the researcher. In this study, the factors affectingunemployment in Organization for Economic Co-operation and Development(OECD) countries were tried to be determined. In this context, data for the years2000-2017 were analyzed by using MARS method. For each year, we estimated theMultivariate Adaptive Regression Splines (MARS) models and we tracked theeffective predictors. According to our findings, the indicators Gross domesticproduct (Gdp), tax revenue rate, long term interest rate, saving rate and inflationusually have a significant impact on the unemployment rates. The annual growthrate of import, export and exchange rate do not influence the unemploymentratios. Besides these results, the industrial production, the industrial value addedand current account balance are influential for a few years.