Pub Date : 2023-10-10DOI: 10.3390/econometrics11040023
Alecos Papadopoulos
We derive a new matrix statistic for the Hausman test for endogeneity in cross-sectional Instrumental Variables estimation, that incorporates heteroskedasticity in a natural way and does not use a generalized inverse. A Monte Carlo study examines the performance of the statistic for different heteroskedasticity-robust variance estimators and different skedastic situations. We find that the test statistic performs well as regards empirical size in almost all cases; however, as regards empirical power, how one corrects for heteroskedasticity matters. We also compare its performance with that of the Wald statistic from the augmented regression setup that is often used for the endogeneity test, and we find that the choice between them may depend on the desired significance level of the test.
{"title":"A New Matrix Statistic for the Hausman Endogeneity Test under Heteroskedasticity","authors":"Alecos Papadopoulos","doi":"10.3390/econometrics11040023","DOIUrl":"https://doi.org/10.3390/econometrics11040023","url":null,"abstract":"We derive a new matrix statistic for the Hausman test for endogeneity in cross-sectional Instrumental Variables estimation, that incorporates heteroskedasticity in a natural way and does not use a generalized inverse. A Monte Carlo study examines the performance of the statistic for different heteroskedasticity-robust variance estimators and different skedastic situations. We find that the test statistic performs well as regards empirical size in almost all cases; however, as regards empirical power, how one corrects for heteroskedasticity matters. We also compare its performance with that of the Wald statistic from the augmented regression setup that is often used for the endogeneity test, and we find that the choice between them may depend on the desired significance level of the test.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136356671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The article considers the problems of young people aged 20-24 on the labour market affected by unemployment in European Union countries. Unemployment is one of the most important economic and social problems, which at the same time constitutes one of the biggest measures characterising the condition of the economy. The diversity of the economic situation in EU countries directly affects young people, an individual group of people entering the labour market and have little or no professional experience. At the same time, they are ready to start work, facing great difficulties in entering the market, influenced by socio-economic as well as demographic factors which directly and indirectly affect employment. Considering the above premise, the aim of the article was to identify the determinants of unemployment of young people aged 20-24 in the EU. The study used data from two years: 2010 and 2020, and applied multiple regression. Statistical data were taken from Eurostat databases. The study allowed to examine the dependence of the influence of individual socioeconomic as well as demographic factors on youth unemployment. The study found that the multivariate regression showed that factors related to young people's participation in education and training (including the NEET rate) relative to labour market status, as well as social inclusion, had a significant impact on the unemployment studied. Over the decade, a decrease was seen in unemployment in most EU member states, in as many as 19 countries, while the remaining eight countries showed an increase.
{"title":"Evaluation of the Labour Market Situation of Young People in EU Countries – The Multiple Regression Approach","authors":"Magdalena Kawecka","doi":"10.15611/eada.2023.3.03","DOIUrl":"https://doi.org/10.15611/eada.2023.3.03","url":null,"abstract":"Abstract The article considers the problems of young people aged 20-24 on the labour market affected by unemployment in European Union countries. Unemployment is one of the most important economic and social problems, which at the same time constitutes one of the biggest measures characterising the condition of the economy. The diversity of the economic situation in EU countries directly affects young people, an individual group of people entering the labour market and have little or no professional experience. At the same time, they are ready to start work, facing great difficulties in entering the market, influenced by socio-economic as well as demographic factors which directly and indirectly affect employment. Considering the above premise, the aim of the article was to identify the determinants of unemployment of young people aged 20-24 in the EU. The study used data from two years: 2010 and 2020, and applied multiple regression. Statistical data were taken from Eurostat databases. The study allowed to examine the dependence of the influence of individual socioeconomic as well as demographic factors on youth unemployment. The study found that the multivariate regression showed that factors related to young people's participation in education and training (including the NEET rate) relative to labour market status, as well as social inclusion, had a significant impact on the unemployment studied. Over the decade, a decrease was seen in unemployment in most EU member states, in as many as 19 countries, while the remaining eight countries showed an increase.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139344384","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The primary goal of this research was to estimate the quantile of a conditional distribution using a semi-parametric approach in the presence of randomly missing data, where the predictor variable belongs to a semi-metric space. The authors assumed a single index structure to link the explanatory and response variable. First, a kernel estimator was proposed for the conditional distribution function, assuming that the data were selected from a stationary process with missing data at random (MAR). By imposing certain general conditions, the study established the model’s uniform almost complete consistencies with convergence rates.
{"title":"Single Functional Index Quantile Regression for Functional Data with Missing Data at Random","authors":"Nadia Kadiri, Sanaà Dounya Mekki, A. Rabhi","doi":"10.15611/eada.2023.3.01","DOIUrl":"https://doi.org/10.15611/eada.2023.3.01","url":null,"abstract":"Abstract The primary goal of this research was to estimate the quantile of a conditional distribution using a semi-parametric approach in the presence of randomly missing data, where the predictor variable belongs to a semi-metric space. The authors assumed a single index structure to link the explanatory and response variable. First, a kernel estimator was proposed for the conditional distribution function, assuming that the data were selected from a stationary process with missing data at random (MAR). By imposing certain general conditions, the study established the model’s uniform almost complete consistencies with convergence rates.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139344976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract The aim of this article was to analyse functional data of the number of hospitalised individuals, intensive care patients, positive COVID-19 tests, deaths and convalescents during the second wave of the COVID-19 pandemic in Poland. For this purpose, firstly the author convert data of sixteen voivodeships to smooth functions, and then used the principal component analysis and multiple function-on-function linear regression model to predict the number of hospitalised and intensive care patients due to the COVID-19 infection during the second wave of the pandemic. Finally, the results were compared with those previously obtained for the combined data of the second and third wave of the COVID-19 pandemic in Poland (Hęćka, 2023).
{"title":"The Second Wave of the COVID-19 Pandemic in Poland – Characterised Using FDA Methods","authors":"Patrycja Hęćka","doi":"10.15611/eada.2023.3.02","DOIUrl":"https://doi.org/10.15611/eada.2023.3.02","url":null,"abstract":"Abstract The aim of this article was to analyse functional data of the number of hospitalised individuals, intensive care patients, positive COVID-19 tests, deaths and convalescents during the second wave of the COVID-19 pandemic in Poland. For this purpose, firstly the author convert data of sixteen voivodeships to smooth functions, and then used the principal component analysis and multiple function-on-function linear regression model to predict the number of hospitalised and intensive care patients due to the COVID-19 infection during the second wave of the pandemic. Finally, the results were compared with those previously obtained for the combined data of the second and third wave of the COVID-19 pandemic in Poland (Hęćka, 2023).","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139345717","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-30DOI: 10.3390/econometrics11030022
Dean Fantazzini, Yufeng Xiao
Detecting pump-and-dump schemes involving cryptoassets with high-frequency data is challenging due to imbalanced datasets and the early occurrence of unusual trading volumes. To address these issues, we propose constructing synthetic balanced datasets using resampling methods and flagging a pump-and-dump from the moment of public announcement up to 60 min beforehand. We validated our proposals using data from Pumpolymp and the CryptoCurrency eXchange Trading Library to identify 351 pump signals relative to the Binance crypto exchange in 2021 and 2022. We found that the most effective approach was using the original imbalanced dataset with pump-and-dumps flagged 60 min in advance, together with a random forest model with data segmented into 30-s chunks and regressors computed with a moving window of 1 h. Our analysis revealed that a better balance between sensitivity and specificity could be achieved by simply selecting an appropriate probability threshold, such as setting the threshold close to the observed prevalence in the original dataset. Resampling methods were useful in some cases, but threshold-independent measures were not affected. Moreover, detecting pump-and-dumps in real-time involves high-dimensional data, and the use of resampling methods to build synthetic datasets can be time-consuming, making them less practical.
{"title":"Detecting Pump-and-Dumps with Crypto-Assets: Dealing with Imbalanced Datasets and Insiders’ Anticipated Purchases","authors":"Dean Fantazzini, Yufeng Xiao","doi":"10.3390/econometrics11030022","DOIUrl":"https://doi.org/10.3390/econometrics11030022","url":null,"abstract":"Detecting pump-and-dump schemes involving cryptoassets with high-frequency data is challenging due to imbalanced datasets and the early occurrence of unusual trading volumes. To address these issues, we propose constructing synthetic balanced datasets using resampling methods and flagging a pump-and-dump from the moment of public announcement up to 60 min beforehand. We validated our proposals using data from Pumpolymp and the CryptoCurrency eXchange Trading Library to identify 351 pump signals relative to the Binance crypto exchange in 2021 and 2022. We found that the most effective approach was using the original imbalanced dataset with pump-and-dumps flagged 60 min in advance, together with a random forest model with data segmented into 30-s chunks and regressors computed with a moving window of 1 h. Our analysis revealed that a better balance between sensitivity and specificity could be achieved by simply selecting an appropriate probability threshold, such as setting the threshold close to the observed prevalence in the original dataset. Resampling methods were useful in some cases, but threshold-independent measures were not affected. Moreover, detecting pump-and-dumps in real-time involves high-dimensional data, and the use of resampling methods to build synthetic datasets can be time-consuming, making them less practical.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42522175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-25DOI: 10.3390/econometrics11030021
Emil Palikot
I study the relationship between competition and innovation, focusing on the distinction between product and process innovations. By considering product innovation, I expand upon earlier research on the topic of the relationship between competition and innovation, which focused on process innovations. New products allow firms to differentiate themselves from one another. I demonstrate that the competition level that creates the most innovation incentive is higher for process innovation than product innovation. I also provide empirical evidence that supports these results. Using the community innovation survey, I first show that an inverted U-shape characterizes the relationship between competition and both process and product innovations. The optimal competition level for promoting innovation is higher for process innovation.
{"title":"Competition–Innovation Nexus: Product vs. Process, Does It Matter?","authors":"Emil Palikot","doi":"10.3390/econometrics11030021","DOIUrl":"https://doi.org/10.3390/econometrics11030021","url":null,"abstract":"I study the relationship between competition and innovation, focusing on the distinction between product and process innovations. By considering product innovation, I expand upon earlier research on the topic of the relationship between competition and innovation, which focused on process innovations. New products allow firms to differentiate themselves from one another. I demonstrate that the competition level that creates the most innovation incentive is higher for process innovation than product innovation. I also provide empirical evidence that supports these results. Using the community innovation survey, I first show that an inverted U-shape characterizes the relationship between competition and both process and product innovations. The optimal competition level for promoting innovation is higher for process innovation.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42479713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-18DOI: 10.3390/econometrics11030020
S. Kumbhakar, Jingfang Zhang, Gudbrand Lien
In this study, we leverage geographical coordinates and firm-level panel data to uncover variations in production across different locations. Our approach involves using a semiparametric proxy variable regression estimator, which allows us to define and estimate a customized production function for each firm and its corresponding location. By employing kernel methods, we estimate the nonparametric functions that determine the model’s parameters based on latitude and longitude. Furthermore, our model incorporates productivity components that consider various factors that influence production. Unlike spatially autoregressive-type production functions that assume a uniform technology across all locations, our approach estimates technology and productivity at both the firm and location levels, taking into account their specific characteristics. To handle endogenous regressors, we incorporate a proxy variable identification technique, distinguishing our method from geographically weighted semiparametric regressions. To investigate the heterogeneity in production technology and productivity among Norwegian grain farmers, we apply our model to a sample of farms using panel data spanning from 2001 to 2020. Through this analysis, we provide empirical evidence of regional variations in both technology and productivity among Norwegian grain farmers. Finally, we discuss the suitability of our approach for addressing the heterogeneity in this industry.
{"title":"Locationally Varying Production Technology and Productivity: The Case of Norwegian Farming","authors":"S. Kumbhakar, Jingfang Zhang, Gudbrand Lien","doi":"10.3390/econometrics11030020","DOIUrl":"https://doi.org/10.3390/econometrics11030020","url":null,"abstract":"In this study, we leverage geographical coordinates and firm-level panel data to uncover variations in production across different locations. Our approach involves using a semiparametric proxy variable regression estimator, which allows us to define and estimate a customized production function for each firm and its corresponding location. By employing kernel methods, we estimate the nonparametric functions that determine the model’s parameters based on latitude and longitude. Furthermore, our model incorporates productivity components that consider various factors that influence production. Unlike spatially autoregressive-type production functions that assume a uniform technology across all locations, our approach estimates technology and productivity at both the firm and location levels, taking into account their specific characteristics. To handle endogenous regressors, we incorporate a proxy variable identification technique, distinguishing our method from geographically weighted semiparametric regressions. To investigate the heterogeneity in production technology and productivity among Norwegian grain farmers, we apply our model to a sample of farms using panel data spanning from 2001 to 2020. Through this analysis, we provide empirical evidence of regional variations in both technology and productivity among Norwegian grain farmers. Finally, we discuss the suitability of our approach for addressing the heterogeneity in this industry.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43414369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-10DOI: 10.3390/econometrics11030019
Bilel Sanhaji, Julien Chevallier
Using the capital asset pricing model, this article critically assesses the relative importance of computing ‘realized’ betas from high-frequency returns for Bitcoin and Ethereum—the two major cryptocurrencies—against their classic counterparts using the 1-day and 5-day return-based betas. The sample includes intraday data from 15 May 2018 until 17 January 2023. The microstructure noise is present until 4 min in the BTC and ETH high-frequency data. Therefore, we opt for a conservative choice with a 60 min sampling frequency. Considering 250 trading days as a rolling-window size, we obtain rolling betas < 1 for Bitcoin and Ethereum with respect to the CRIX market index, which could enhance portfolio diversification (at the expense of maximizing returns). We flag the minimal tracking errors at the hourly and daily frequencies. The dispersion of rolling betas is higher for the weekly frequency and is concentrated towards values of β > 0.8 for BTC (β > 0.65 for ETH). The weekly frequency is thus revealed as being less precise for capturing the ‘pure’ systematic risk for Bitcoin and Ethereum. For Ethereum in particular, the availability of high-frequency data tends to produce, on average, a more reliable inference. In the age of financial data feed immediacy, our results strongly suggest to pension fund managers, hedge fund traders, and investment bankers to include ‘realized’ versions of CAPM betas in their dashboard of indicators for portfolio risk estimation. Sensitivity analyses cover jump detection in BTC/ETH high-frequency data (up to 25%). We also include several jump-robust estimators of realized volatility, where realized quadpower volatility prevails.
{"title":"Tracking ‘Pure’ Systematic Risk with Realized Betas for Bitcoin and Ethereum","authors":"Bilel Sanhaji, Julien Chevallier","doi":"10.3390/econometrics11030019","DOIUrl":"https://doi.org/10.3390/econometrics11030019","url":null,"abstract":"Using the capital asset pricing model, this article critically assesses the relative importance of computing ‘realized’ betas from high-frequency returns for Bitcoin and Ethereum—the two major cryptocurrencies—against their classic counterparts using the 1-day and 5-day return-based betas. The sample includes intraday data from 15 May 2018 until 17 January 2023. The microstructure noise is present until 4 min in the BTC and ETH high-frequency data. Therefore, we opt for a conservative choice with a 60 min sampling frequency. Considering 250 trading days as a rolling-window size, we obtain rolling betas < 1 for Bitcoin and Ethereum with respect to the CRIX market index, which could enhance portfolio diversification (at the expense of maximizing returns). We flag the minimal tracking errors at the hourly and daily frequencies. The dispersion of rolling betas is higher for the weekly frequency and is concentrated towards values of β > 0.8 for BTC (β > 0.65 for ETH). The weekly frequency is thus revealed as being less precise for capturing the ‘pure’ systematic risk for Bitcoin and Ethereum. For Ethereum in particular, the availability of high-frequency data tends to produce, on average, a more reliable inference. In the age of financial data feed immediacy, our results strongly suggest to pension fund managers, hedge fund traders, and investment bankers to include ‘realized’ versions of CAPM betas in their dashboard of indicators for portfolio risk estimation. Sensitivity analyses cover jump detection in BTC/ETH high-frequency data (up to 25%). We also include several jump-robust estimators of realized volatility, where realized quadpower volatility prevails.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-08-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44587270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-07-31DOI: 10.3390/econometrics11030018
Manabu Asai
Despite the growing interest in realized stochastic volatility models, their estimation techniques, such as simulated maximum likelihood (SML), are computationally intensive. Based on the realized volatility equation, this study demonstrates that, in a finite sample, the quasi-maximum likelihood estimator based on the Kalman filter is competitive with the two-step SML estimator, which is less efficient than the SML estimator. Regarding empirical results for the S&P 500 index, the quasi-likelihood ratio tests favored the two-factor realized asymmetric stochastic volatility model with the standardized t distribution among alternative specifications, and an analysis on out-of-sample forecasts prefers the realized stochastic volatility models, rejecting the model without the realized volatility measure. Furthermore, the forecasts of alternative RSV models are statistically equivalent for the data covering the global financial crisis.
{"title":"Estimation of Realized Asymmetric Stochastic Volatility Models Using Kalman Filter","authors":"Manabu Asai","doi":"10.3390/econometrics11030018","DOIUrl":"https://doi.org/10.3390/econometrics11030018","url":null,"abstract":"Despite the growing interest in realized stochastic volatility models, their estimation techniques, such as simulated maximum likelihood (SML), are computationally intensive. Based on the realized volatility equation, this study demonstrates that, in a finite sample, the quasi-maximum likelihood estimator based on the Kalman filter is competitive with the two-step SML estimator, which is less efficient than the SML estimator. Regarding empirical results for the S&P 500 index, the quasi-likelihood ratio tests favored the two-factor realized asymmetric stochastic volatility model with the standardized t distribution among alternative specifications, and an analysis on out-of-sample forecasts prefers the realized stochastic volatility models, rejecting the model without the realized volatility measure. Furthermore, the forecasts of alternative RSV models are statistically equivalent for the data covering the global financial crisis.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44567047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-20DOI: 10.3390/econometrics11020017
Mateusz Szysz, Andrzej Torój
In some NUTS 2 (Nomenclature of Territorial Units for Statistics) regions of Europe, the COVID-19 pandemic has triggered an increase in mortality by several dozen percent and only a few percent in others. Based on the data on 189 regions from 19 European countries, we identified factors responsible for these differences, both intra- and internationally. Due to the spatial nature of the virus diffusion and to account for unobservable country-level and sub-national characteristics, we used spatial econometric tools to estimate two types of models, explaining (i) the number of cases per 10,000 inhabitants and (ii) the percentage increase in the number of deaths compared to the 2016–2019 average in individual regions (mostly NUTS 2) in 2020. We used two weight matrices simultaneously, accounting for both types of spatial autocorrelation: linked to geographical proximity and adherence to the same country. For the feature selection, we used Bayesian Model Averaging. The number of reported cases is negatively correlated with the share of risk groups in the population (60+ years old, older people reporting chronic lower respiratory disease, and high blood pressure) and the level of society’s belief that the positive health effects of restrictions outweighed the economic losses. Furthermore, it positively correlated with GDP per capita (PPS) and the percentage of people employed in the industry. On the contrary, the mortality (per number of infections) has been limited through high-quality healthcare. Additionally, we noticed that the later the pandemic first hit a region, the lower the death toll there was, even controlling for the number of infections.
{"title":"Socio-Economic and Demographic Factors Associated with COVID-19 Mortality in European Regions: Spatial Econometric Analysis","authors":"Mateusz Szysz, Andrzej Torój","doi":"10.3390/econometrics11020017","DOIUrl":"https://doi.org/10.3390/econometrics11020017","url":null,"abstract":"In some NUTS 2 (Nomenclature of Territorial Units for Statistics) regions of Europe, the COVID-19 pandemic has triggered an increase in mortality by several dozen percent and only a few percent in others. Based on the data on 189 regions from 19 European countries, we identified factors responsible for these differences, both intra- and internationally. Due to the spatial nature of the virus diffusion and to account for unobservable country-level and sub-national characteristics, we used spatial econometric tools to estimate two types of models, explaining (i) the number of cases per 10,000 inhabitants and (ii) the percentage increase in the number of deaths compared to the 2016–2019 average in individual regions (mostly NUTS 2) in 2020. We used two weight matrices simultaneously, accounting for both types of spatial autocorrelation: linked to geographical proximity and adherence to the same country. For the feature selection, we used Bayesian Model Averaging. The number of reported cases is negatively correlated with the share of risk groups in the population (60+ years old, older people reporting chronic lower respiratory disease, and high blood pressure) and the level of society’s belief that the positive health effects of restrictions outweighed the economic losses. Furthermore, it positively correlated with GDP per capita (PPS) and the percentage of people employed in the industry. On the contrary, the mortality (per number of infections) has been limited through high-quality healthcare. Additionally, we noticed that the later the pandemic first hit a region, the lower the death toll there was, even controlling for the number of infections.","PeriodicalId":11499,"journal":{"name":"Econometrics","volume":null,"pages":null},"PeriodicalIF":1.5,"publicationDate":"2023-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48304319","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}