Publication: İktisadi büyümenin latent değişkenli panel veri modelleri ile analizi
Abstract
Bir ülkede üretilen çıktının toplam değerini ifade eden GSYH, bileşenleri ve hesaplanma yöntemleri açısından üzerinde tartışılan bir kavramdır. Ekonomik üretim ve örgütlenme biçimlerinin ülkeden ülkeye farklılık ihtiva etmesine karşın standardize edilmiş yöntemlerle hesaplanma yapılması, ülkelerin kendilerine has özelliklerinin gözden kaçırılmasına, kıyaslamalarda noksanlıkların oluşmasına ve yanlış politika üretimlerine sebep olmaktadır. Bu çalışmada, iktisadi büyümenin, GSYH’nin temel bileşenleri ve önemli belirleyicilerinde doğrudan gözlemlenemeyen faktörleri veya eğilimleri temsil eden ülkelere özgü latent değişken bağlamında incelenmesi amaçlanmıştır. Bu doğrultuda, harcamalar yöntemine göre GSYH büyüme oranını etkileyen faktörler, 68 ülkenin 1990-2021 dönemine ilişkin verileri kullanılarak geçmiş dönem dinamiklerini de göz önünde bulunduran dinamik panel veri modelleri kapsamında incelenmiştir. Büyüme oranını etkileyen faktörler için ilk olarak Arellano-Bond Genelleştirilmiş Momentler Tahmincisi ile tahminler elde edilmiştir. Ardından dinamik yapıda latent etkilerin incelenebilmesi amacıyla literatürde yer alan ve Arellano-Bond Tahmincisine alternatif olarak Allison, Williams ve Moral-Benito (2017), Williams, Allison ve Moral-Benito (2018) ve Moral-Benito, Allison ve Williams (2019) tarafından geliştirilen Maksimum Olabilirlik Dinamik Panel Veri modeliyle analiz gerçekleştirilmiştir. Modeller oluşturulurken ilk aşamada ülkelerin yer aldığı farklı gelir grupları ve bölgeler göz ardı edilerek, sabit yatırımlar, tüketim, kamu harcamaları, dış ticaret ve finansal gelişme değişkenlerinin büyüme üzerine etkileri irdelenmiştir. Daha sonra ise ülkelerin G20 üyelikleri, gelir grupları ve bulundukları coğrafi bölgeler, modellere kukla değişken olarak dâhil edilmiş, böylece bölgesel ve kurumsal ayrışmaların ve bu ayrışmalara neden olabilecek faktörler değerlendirilmiştir.
GDP, which expresses the total value of output produced in a country, is a concept discussed in terms of its components and calculation methods. While economic production and organizational forms vary from country to country, standardized methods of calculation can lead to the oversight of a country's unique characteristics, shortcomings in comparisons, and the formulation of incorrect policies. This study desires to analyze economic growth in the context of country-specific latent variables that represent factors or trends that cannot be directly observed in the main components and significant determinants of GDP. In this respect, the factors affecting the GDP growth rate according to the expenditures method were examined within the scope of dynamic panel data models that also take into account the dynamics of the past period, using the data of 68 countries for the period 1990-2021. First, estimates were obtained for the factors affecting the growth rate with the Arellano-Bond Generalized Moment Estimator. Then, to examine latent effects in the dynamic structure, an analysis was carried out with the Maximum Likelihood Dynamic Panel Data model developed by Allison, Williams and Moral-Benito (2017), Williams, Allison and Moral-Benito (2018), and Moral-Benito, Allison, and Williams (2019), which is an alternative to the Arellano-Bond Estimator. Analysis was carried out with the Maximum Likelihood Dynamic Panel Data model. In the initial stage of model creation, different income groups and regions in which countries are located were initially disregarded, and the effects of variables such as fixed investments, consumption, public expenditures, trade, and financial development on growth were discussed. Subsequently, the G20 memberships of countries, income groups, and geographical regions in which they are located were included in the models as dummy variables, thus evaluating regional and institutional differentiations and factors that could cause these differentiations.
GDP, which expresses the total value of output produced in a country, is a concept discussed in terms of its components and calculation methods. While economic production and organizational forms vary from country to country, standardized methods of calculation can lead to the oversight of a country's unique characteristics, shortcomings in comparisons, and the formulation of incorrect policies. This study desires to analyze economic growth in the context of country-specific latent variables that represent factors or trends that cannot be directly observed in the main components and significant determinants of GDP. In this respect, the factors affecting the GDP growth rate according to the expenditures method were examined within the scope of dynamic panel data models that also take into account the dynamics of the past period, using the data of 68 countries for the period 1990-2021. First, estimates were obtained for the factors affecting the growth rate with the Arellano-Bond Generalized Moment Estimator. Then, to examine latent effects in the dynamic structure, an analysis was carried out with the Maximum Likelihood Dynamic Panel Data model developed by Allison, Williams and Moral-Benito (2017), Williams, Allison and Moral-Benito (2018), and Moral-Benito, Allison, and Williams (2019), which is an alternative to the Arellano-Bond Estimator. Analysis was carried out with the Maximum Likelihood Dynamic Panel Data model. In the initial stage of model creation, different income groups and regions in which countries are located were initially disregarded, and the effects of variables such as fixed investments, consumption, public expenditures, trade, and financial development on growth were discussed. Subsequently, the G20 memberships of countries, income groups, and geographical regions in which they are located were included in the models as dummy variables, thus evaluating regional and institutional differentiations and factors that could cause these differentiations.