Estimating Impact of Austerity Policies in COVID-19 Fatality Rates: Examining the Dynamics of Economic Policy and Case Fatality Rates (CFR) of COVID-19 in OECD Countries

The paper will attempt to estimate factors which determine the variability of case fatality rates of COVID-19 (Coronavirus) across OECD countries in the recent time. The objective of the paper is to estimate the impact of government health policies on fatality rates (Case fatality rates) of COVID-19 in OECD countries while controlling for other demographic and economic characteristics. The analysis is done using non-parametric regression method, i.e. Quantile regression. The result from quantile regression analysis shows that a policy of Austerity (health expenditure cuts) significantly increases the mortality rates of COVID-19 in OCED countries. The policy implication of the study is the need for a robust public-funded health system with wider accessibility to deal with major public health crisis like COVID-19 pandemic.


Objective of Research:
Review of existing literature on Covid-19 shows the dynamic interplay between the Covid 19 and the country-specific health policy is still missing. This paper attempts to fill this gap by highlighting the interrelationship between the long term structural health policies and the Covid 19 fatality rates among Organisation for Economic Co-operation and Development (OECD) countries 2 . Definition of Austerity policies is a widespread cut on government expenditure which is targeted to reduce government fiscal deficit and enhance economic growth (Konzelmann, 2014;). Such a significant reduction in government spending has a disproportionately negative impact on government social sector expenditure (Health, Education, social security etc.) The negative impact of austerity policies in terms of lowering employment, economic growth and increasing inequality is well studied (Blyth, 2013;Krugman, 2015;Stiglitz, 2012;UNCTAD, 2017). In the post.-2008 crisis period and under the impact of rising debts burdens, many countries in European counties imposed a policy of austerity in 2010. The most severe austerity policies were implemented in Greece, Hungary, Ireland, Latvia, Spain and Portugal (Leschke et al., 2015). Among the OCED group, there is variation in the extent of reduction in their health expenditure in pursuit of Austerity policies(fiscal consolidation) across countries (Van Gool & Pearson, 2014). The negative impact of such drastic funding cuts on access to health facilities and health indicators is well documented in many OCED countries (Antonakakis & Collins, 2014;Ayuso-Mateos et al., 2013;Ifanti et al., 2013;Kentikelenis et al., 2014Kentikelenis et al., , 2014Kentikelenis et al., , 2014Loopstra et al., 2016Loopstra et al., , 2016McKee et al., 2012McKee et al., , 2012Reeves et al., 2014Reeves et al., , 2014Ruckert & Labonté, 2017;Stuckler et al., 2017). So under the background of such drastic cuts in health expenditure, the papers will evaluate the impact of austerity policies (health expenditure cuts) on fatality rates of Covid-19 after controlling for other socio-demographic characteristics which have a significant impact on fatality rates of covid-19. The fatality rates are measured by crude Case fatality rates (CFC), which is the ratio of confirmed death to confirmed positive cases of covid-19 for each country.

Data source and Methodology:
The data for analysis is taken from different data sources. Following is the table providing a list of variable and their data source. The data for analysis is from thirty The analysis of the impact of austerity on covid regression after controlling for all other socio impact on Case fatality rate (Novel, 2020;Onder et al., 2020;Porcheddu et al., 2020;Wu & McGoogan, 2020). The advantage of quantile regression over norm regression (OLS) regression is that it gives a rich picture of the relationship between variables not only around mean value but across the distribution of variables 2001). It is distribution-free, robust distribution (Baum, 2013 Share of the population having a high blood pressure condition above 18 age.
om thirty-six countries 3 from the OECD group. shows that large part of infection and deaths from Co is concentrated in OECD regions only.
the impact of austerity on covid-19 fatality rates is done using regression after controlling for all other socio-demographic characteristics which have an (Novel, 2020;Onder et al., 2020;Porcheddu et al., 2020; Wu & The advantage of quantile regression over normal Ordinary least square regression is that it gives a rich picture of the relationship between variables not only around mean value but across the distribution of variables (Koenker & Hallock, free, robust to outliers, capable of modelling entire conditional (Baum, 2013;Cade & Noon, 2003;Yu et al., 2003). 19 fatality rates is done using Quantile demographic characteristics which have an (Novel, 2020;Onder et al., 2020;Porcheddu et al., 2020;Wu & Ordinary least square regression is that it gives a rich picture of the relationship between variables (Koenker & Hallock, outliers, capable of modelling entire conditional OECD 77%

Statistical Analysis:
The The average value of public spending in health to GDP variable, publichelathgdp, is 5.95 % and it has a minimum value from 0.9 per cent and the maximum value of 14.4 per cent. The existing clinical research shows that fatality rate of Covid-19 is influenced by the existence of pre-medical complication and the share of older adults in the population (Onder et al., 2020;Wu & McGoogan, 2020). The crucial demographic variable, population share above 65 years (popu65), has a mean value of 15 per cent and a standard deviation of 5.4. The share of the population having hypertension (above 18 years age) has a minimum value of 13 % and a maximum value of 38.2 %.
b. Quantile Regression: Table 3 shows the result of Quantile regression. The first model has a case fatality rate from March 30 as the dependent variable. In the second model, the dependent variable is the threeday median case fatality rate. The result from both models shows that the coefficient of the dummy variable for high fund cut has a positive impact on CFR and is significant at one per cent level of significance. The result shows that a country which has a history of drastic health fund cut is increasing the fatality rates from covid-19. Similarly, the coefficient of the variable of public health GDP is negative and significant at one per cent level of significance.
It shows that countries which have a higher share of the public-funded health system have a lower case fatality rate. The impact of good health infrastructure (measured by bed per 1000 population and doctor per 1000 population) on case fatality rates is negative. As expected in exiting literature, higher population in the older age group has higher fatality rates. Also, the higher share of pre-existing medical condition in the overall population, higher is the fatality rate from COVID-19.

Table 4: Model Specification Test: Link Test of Model-1
The presence of model specification error is done using the link test. If the regression model does not contain specification error, then the variable _hatsq will be statistically insignificant. Table 4 shows the result of the link test for Model -1. The P-value of variable _hatsq is 0.16, and hence it is statistically insignificant. So model-1 does not contain specification error. Similarly, the link test result of model-2 also shows that variable _hatsq is not statistically significant. Hence model-2 does not contain specification error.
In order to get an idea about coefficient of Quantile regression of independent variables across quantile of case fatality rates following two figure has been calculated using the Azevedo method (Azevedo, 2011). It shows how the impact of each independent variable varies across quantiles.   The coefficient of the public fund on health to GDP variable is negative( except for the first quantile) across the distribution of case fatality ratio. expenditure on health reduces fatality rates of Covid-19. The coefficient of doctor per 1000 population is negative across the distribution of case fatality ratio (except for the second quantile). The coefficient of hospital bed per 1000 population is negative across the distribution of case fatality ratio (except till the fourth quantile).

Conclusion:
The result from the Quantile regression analysis shows that countries which has pursued austerity policies has significantly higher fatality rates from COVID-19 after controlling for all other socio-demographic factors which influence case fatality rate of COVID-19. Higher public funding share, higher doctors per population, higher bed availability is associated with lower fatality rates from COVID-19. A higher share of the population with pre-medical conditions (diabetics, hypertension) and older age population increase fatality rates. So the policies of austerity (at least in terms of reduction in health expenditure) can significantly worsen health system's ability to fight pandemic live COVID-19 and can lead to a severe negative health outcome. The policy implication of the study is the need for a robust publicfunded health system with wider accessibility to deal with a major public health crisis like a covid-19 pandemic.

Limitation of Study:
The major limitation of preliminary study is that it uses case fatality rates as proxy of fatality of COVID-19 when epidemic is still not full blown or has not reached its peak in many countries which are selected for this study. Still results are indicative of impact of health fund cuts on fatality of COVID-19.

Conflict of Interest:
No funding from external agencies has been received to complete this study.