No abstract

The coronavirus pandemic (COVID-19) has caused the biggest economic contraction in global economy since the Second World War. COVID-19 pandemic has forced governments to take unprecedented measures to prevent the spread and to protect their economies that presented a dilemma because of their conflicting outcomes. This paper investigates the presumption of healtheconomy trade-off due to COVID-19 by comparing the GDP declines and deaths in per million population in OECD countries and China. The empiric data shows the countries with the highest death rates have seen the largest economic downturns. The clustering analysis by using k-means algorithm finds that there are three partitions of countries for current account balances, GDP growth rate, and deaths in per million population. The countries with current account surpluses above 2.5% of GDP managed to limit their GDP decline below -15% and are in the same cluster. On the other hand, the countries with higher death rates and current account deficits group another cluster and saw GDP declines as above 15% except for USA and Brasil.

According to the New Keynesian economic theory, built on the assumption that prices and wages are sticky, the money is not neutral and the monetary expansion generates real results in the economy. Despite the aggregate data supporting the sticky prices assumption, microdata indicates that disaggregated prices change much faster than conventionally assumed. To explain the disagreement between macro and microdata, A FAVAR model (Bernanke, Boivin ve Eliasz, 2005) is developed and the common and sector-specific components of Turkish economy from 2005: 01 to 2014: 12 is computed. The analysis reveals that the most of the monthly changes in prices in disaggregated data are due to sector-specific factors. Moreover, the persistence of the TÜFE inflation is largely due to macroeconomic factors.

In the Vector Autoregressive (VAR) models, which are widely used in economic studies and developed by Sims (1980), impulse response functions can only be obtained from variables included only because of the infrequent use of information sets, and the dimensions of structural shocks can not be measured precisely. It is also not possible that for some variables to be represented by a single time series. The VAR estimation is insufficient for parsing operations involving large data sets. FAVAR (Factor Augmented Vector Autoregression) method was developed by Bernanke, Boivin and Eliasz (2005) and this method can use large data sets. In this study, FAVAR method is tried to be explained by comparing with VAR, and a literature search is being conducted in this subject.

The COVID-19 pandemic has caused significant economic contractions and employment vulnerabilities for the economies, including Turkey. The pandemic exacerbated structural challenges related to high unemployment, low labor force participation, and widespread informality. This study aims to analyze the differences in the labor market between the 2019 and 2020 years in Turkey. For this purpose, we used the clustering method. While applying the clustering method, we used education type, gender, and age group data. Moreover, the study also employed information from employed, unemployed, and not in labor force data. We implemented a Machine Learning method, K-modes analysis, using the Turkish Statistical Institute's employment statistics and Labor Force Statistics Micro Datasets for 2019 and 2020.