Spreading of Covid19 on Human Populace
Dr. Subrata Roy
Associate Professor
Department of Commerce
Mahatma Gandhi Central University
Motihari, East Champaran, Bihar
Mr. Avneesh Kumar
Assistant Professor
Department of Commerce
Mahatma Gandhi Central University
Motihari, East Champaran, Bihar
Abstract
The present study tries to examine the impact COVID19 pandemic on world population and thus, daily data on COVID19 cases, time, death, recovery and active cases are considered and obtained from the official websites of UN and WHO over a period from 22^{nd} January 2020 to 24^{th} May 2020. The study employs three nonlinear regression equations and the results are varied according to the model. It is observed that loglog regression equation is the appropriate model to capture the time series data based on R^{2}. The slope coefficients of the double model are positive and statistically significant. All the slope coefficients positively influence the spread of COVID19 in world population. But it is also found that if recovery rate is increased by 20% then COVID19 spreads is only 1.44% and it is possibly only when proper medical facility is provided.
Keywords: COVID19, NonLinear, Death, Recover, World Population, Pandemic
JEL Classification: C13, D8, D84
Introduction
Entire world is facing the dangerous effect of COVID19. The world population is combating against the corona virus for protection and trying to guard the human civilization from its evil effects. A large number of people is infected and in isolation with treatment. The spreads of COVID19 into the world populace is increasing at an exponential rate with the increase of time. The scientists all over the world are trying to develop vaccine by which the pandemic can be controlled at any cost. The anxiety is obvious to everyone and spending lives with worry. The origin of the epidemic of the new SARSCoV2 corona virus causing COVID19 pandemic is at Wuhan in China and spreads across the world with more than 0.55 million infected and over 0.35 million deaths (till May 24, 2020). According to Baud et al., 2020, opines that mortality rate varies between 5.5% and 5.9% but it may be reached to 20% in Wuhan where the epicenter of the pandemic takes place. In the mean time, the World Health Organization (WHO) declares various guidelines to combat with this pandemic and circulates this guideline to all the health authorities across the world as per protocols and informs the countrymen to take measures like constant washing of hands, hygiene habits, social distancing to detection tests, stay at home, using of masks, patient care inside the home, epidemiological supervision and many more precautions in socioeconomic context (www.who.int/es/emergencies/diseases/novelcoronavirus2019). But all the population groups around the world are at risk of astringent COVID19 and the situation is more vulnerable for those group who are above 60 years of old with chronic diseases such as cardiovascular disease, diabetes, unrelieved respiratory disease and suffering from kidney problem (Caramelo et al., 2020; Cheng et al., 2020; Fu et al., 2020; Jordan et al., 2020; Lippi & Henry, 2020). It’s true that COVID19 pandemic creates severe socioeconomic effects on happiness, quality of life, livelihoods, health, food and security of all populace around the globe. The judgment of human health risk must be understood as the susceptibility of a population and its ability to react to new conditions in order to alleviate possible effects. The vulnerability means the degree to which human being is exposed to deal with the adverse effect. Here, the vulnerability of a population depends on its local environment, the level of material resources, the success of governance and institutions, the quality of public health infrastructure and access to local information (Cutter, 1996). But the above aspects are not homogeneous and may differ according to geography, demographics and socioeconomic reasons. So, to protect health from COVID19, the factors which are related to the origin of the disease and those that affect the wellbeing of the population must be dealt with.
Although, it is true that outbreak of COVID19 and worldwide lockdown has negative impact on economy but the environment and everything attached with it (except human) get rid off from the huge anthropogenic force like massive emission of pollutants of diverse kinds (Ray et al., 2020; Beine et al., 2020; Huang et al., 2020; Wu et al., 2020; Mahato et al., 2020) and this lockdown brings a immense opportunity to judge the anthropogenic intervention on qualitative degradation of environmental elements at very local to global scales (Anjum, 2020; Becchetti, 2020; Cadotte, 2020; Das et al., 2020; Layard et al., 2020; Saadat et al., 2020).
It is also well known that complete lockdown is not the only way out to save ourselves from the deadly disease COVID19. The combine effect of COVID19 and lockdown makes the population at health jeopardy as well as economic downturn around the globe. All the sectors of the economy are affected sternly. The governments, health organizations and other authorities are constantly focusing on identifying the cases affected by COVID19 and trying to develop vaccine for preventing the populace from infection. It is recognized that health is the most expensive asset in our life. There are many factors like social, cultural, economic, psychological and environmental that influence health and wellbeing. So, to protect the population from serious health hazards, help the people live longer, offer healthier lives and improving the health of the poorest are the substantial challenges to us during this pandemic situation. The COVID19 pandemic strains the healthcare systems and causes economic recession unparallel in the recent history and force the countries to take proper decisions on how to best meet the needs of their people.
The present study tries to examine the impact of COVID19 on population when it spreads at exponential rate. More extensively, the study scrutinizes the influence of death, recovery, active cases and time on the expansion of COVID19 pandemic in human civilization in the world populace. The introduction section provides a detail about the nature of the study. Section two deals with literature review and research gap. Objectives of the study are given in section three. Section four highlights on data and study period. Methodology is described in section five. Section six explains the result and section seven deals with conclusion and recommendation.
Literature Review
Corona virus is not flu, because it is not due to an influenza virus but it is very similar to those types. But, there are two epidemics of corona virus that has created severe respiratory illness in the recent past. The first one is Severe Acute Respiratory Syndrome (SARS) that is caused by SARSCOV virus and spreads among 30 countries across the world in 200304 which infected more than 8500 people and killed 813 people among them. The second epidemic is Middle East Respiratory Disease Syndrome (MERS) in Arabia due to corona virus that affected 2500 people and killed 858 people which is examined by Sullivan in 2018. Epidemics and sometimes called pandemics bring misery to the human beings. Their evil effects spread out into the human body rapidly and kill large number of population. If we remember about bubonic plague that started in 1347 and spread in Eurasia that killed approx 50% population. In the early 1500s a new epidemic popularly known as smallpox in America killed many Americans and the mortality rate was more than 50%. Similarly, cholera (1881), Spanish flu influenza (1918), Asian flu influenza (1957), Hong Kong flu influenza (1968) and HIV (1981), all these pandemics came after a certain period of time and killed a lot of people around the world. Likewise, SARS (2003) and MERS (2012) extended their roots almost 37 countries and many people were died. Correspondingly, Swine flu influenza (2009) pandemic spread globally and a huge death was occurred approx 0.68 million due to this dangerous virus. Then, West Africa Ebola virus epidemic (2013) extended almost 10 countries and due to its severe effect lots of people lost their valuable lives. Similarly, Zika virus pandemic (2015) in the recent past extended its roots in a rapid way among 76 countries and badly affected many people. Finally, the entire world is now combating against COVID19 (2019).
COVID19 outbreak expands significantly into a global pandemic. It is the epidemic of a respiratory illness caused by a virus identified and popularly known as COVID19 (Corona virus). The virus is officially designated as “Severe Acute Respiratory Syndrome Corona virus 2 (SARSCOV2)” that is assigned by the International Committee on Taxonomy of Viruses. It is related genetically but unlike from acute respiratory syndrome corona viruses (SARSCoV) (WHO, 2020). The Pattern of symptoms is the same in the majority of patients with mild or levels of symptoms (JSA & AIPSN, 2020).
Although, the literature on COVID19 is very scanty but research is carrying on around the world on it. A study by Tomes (2010) observes that Spanish influenza arrived in the United States at a time when new forms of mass transportation, mass media, mass consumption and mass warfare had vastly expanded the public places in which communicable diseases could spread. So, public health authorities tried to implement social distancing measure to control over the situation of deadly crowd disease. The study reports that social distancing is an important measure that significantly can reduce the mortality rates affected by H1N1 influenza pandemic during 19181919. In 2012, Lewandowsky et al., says that widespread prevalence and persistence of misinformation in contemporary societies, such as the false belief that there is a link between childhood vaccinations and autism, is a matter of public concern. They speak about myths surrounding vaccinations which prompted some parents to refuse to give immunization from their children have led to a marked increase in vaccine preventable disease as well as unnecessary public expenditure on research and public information campaigns aimed at rectifying the situation. They examine how this misinformation is disseminated in society both unintentionally and purposely. Misinformation can originate from rumors but also from works of fiction, governments and politicians and vested interests. Moreover, changes in the media landscape including the arrival of internet have fundamentally influenced the ways in which information is communicated and spread and they also examine the misinformation at the level of the individual and review the cognitive factors that often render misinformation resistant to correction. Recently, we have seen misinformation spreading over the social media about COVID19 pandemic and it is very difficult to assess which is wrong and which right. This misinformation is spreading by the individual, political leaders, religious saints and their followers for their vested interest. So, it is dangerous for a country to control over the situation of COVID19 and thus, to control over the debunking of misinformation they provide specific recommendations pertain to the ways in which corrections should be designed, structured and applied in order to maximize their impact. Kim et al., (2016) once again memorizes the Ebola scare in 2014 where many people evinced strong fear and xenophobla and their study informed by the pathogenprevalence hypothesis, tested the influence of individualism and collectivism on xenophobic response to the threat of Ebola. Benjamin Hurlbut (2017) examines the relationship between political norms associated with governance of pandemic risk and argues that scientific regimes are laying claim to a kind of sovereignty by positioning the norms of scientific practice including a commitment to unfettered access to scientific information and to the authority of science to declare what needs to be known as essential to global governance. He reports that scientific authority occupies a constitutional position insofar as it figures centrally in the repertoire of imaginaries that shape how a global community is imagined. Another study by Lancet (2018) concludes that influenza virus causes death to children but fortunately the corona virus has not led to death in children but infected (CIA et al., 2020). Similarly, a study by Li et al., 2019 depicts that 71000 influenza associated excess respiratory deaths occurred in individuals aged 60 years or older. In 2019, Yang et al., says that COVID19 outbreaks in Wuhan, Hubei province and began spreading rapidly with more than 80000 cases confirmed and 3000 deaths in China. Many countries like Italy, Iran, Korea, Japan, USA and others have reported nearly 60000 cases and most of the COVID19 patients initially suffered from fever, cough, fatigue and breathing problem along with muscle pain, headache, chest pain and diarrhea similar to the symptoms observed after chemotherapy, targeted and immune therapy.
A study by Guan et al., in 2020 claims that around 14% per 1000 were died. In India, currently, 7000 persons have been tested since the epidemic began (Scroll, 2020). A joint mission on COVID19 conducted by WHO and China (2020) reports that COVID19 is transmitted via droplets and fomites during close unprotected contact between an infector and infectee, in China, human to human transmission of the COVID19 is largely occurring in families and China has a policy of meticulous case and contact identification for COVID19. Similarly, Ceukelaire et al., (2020) says that corona virus (COVID19) outbreak has already spread from China to the entire world in less than two months. Now, the time is to take stock and to assess the responses of different countries to the outbreak so far. In 2020, Suleman et al., describes the challenges of COVID19 in Pakistan and thus, they develop a transparent user friendly method to stimulate the outbreak data. They use GIS tools to identify and define the field of investigation that helps to the authorities to locate the highly effected area and can take proper measures to those areas. They also suggest that GIS technique can be used for more effective investigation in the vulnerable geographical areas in Pakistan. Another study by Hao Xu et al., (2020) examines the possible environmental effects caused by COVID19 in China and therefore, they have considered daily data of COVID19 confirmed cases, air quality and meteorological variables of 33 locations in China over a period from 29^{th} January 2020 to 15^{th} February 2020 by applying Poisson regression equation. They find that the effect air quality index (AQI) on confirmed cases associated with an increase in each unit of AQI is statistically significant in many cities. They also suggest that an enhanced impact of AQI on the COVID19 spread under low relative humidity (RH). Similarly, in 2020 Lorena et al., critically examines the risk of social, environmental and health factors in Mexican indigenous population in the context of COVID19 pandemic and proposes strategies to mitigate the impacts on these on communities. They suggest community based approach to address these social, environmental and health gaps in the communities. They also opine that health is the preeminent factor of cohesion for inserting development and progress proposals in indigenous communities during this COVID19 pandemic. Likewise, Mondal and Pal (2020), examines the impact of forced lockdown on environmental components in pre and during lockdown by identifying stone quarrying and crushing areas during COVID19 pandemic in middle catchment of Dwarka river basin of Eastern India. They observe that maximum Particulate matter 10 is 189 to 278 ug/m^{3} in pre lockdown period reduced to 50 to 60 ug/m^{3} after 18 days of lockdown. Similarly, land surface temperature dominated by the crusher areas reduced to below 65dBA after lockdown and the adjacent river water is qualitatively improved after declaration of lockdown. So, they opine that emergency lockdown shows a way out to restore environment and ecosystem in a rapid rate.
A 55 year old individual from Hubei province in China may have been the first person to have infected COVID19 in 17^{th} November 2019 (South Morning China Post). At that time, authorities suspected that the virus came from something sold at a wet market in the city. However, it is now confirmed that early in what is now a pandemic, some infected people had no connection to that seafood market that is reported by the researchers in 20^{th} January 2020 in a journal (The Lancet). Scientists now suspect that corona virus has been originated in a bat and somehow hopped to another animal, possibly the pangolin which then passed it into the human body. So, here there is a lot of anomaly about origin of COVID19. Someone claims it is artificially invented by the scientists and the other groups claim it has been passed from the animal body. Now, the disease is spreading among the people without any animal intermediary and the scientists are trying to trace the virus back to where it has been originated to learn more about its spread.
From the above studies it is clear that the researchers have tried to find out the impact of COVID19 on environment, health, economy and other related areas. Keeping in view of the above, the present study tries to examine the growth/expansion rate of COVID19 caused by death, recovery, time and active cases in the world population. The study is new one and there is scanty of literatures and surely adds value to the existing literature.
Objective of the study
The present study tries to examine the following objectives:
Data & Study period
The study uses five variables which are exclusively related to COVID19 pandemic. The variables are COVID19 cases (affected by corona virus), time, death caused by COVID19, recovery from COVID19 and active cases in the whole population. Here, COVID19 is considered as the dependent variable and the remaining are the independent variables. Here, daily data are considered i.e. from 22^{nd} January to 24^{th} May 2020 and collected from the official websites of United Nation and World Health Organization and also crossed checked with the data provided by various secondary sources.
Methodology
The methodology is the technique by which various problems can be solved systematically. In economics it is assumed that population relationships are linear and that’s why regression analysis takes the right form. But in reality linearity assumption is not follow always and therefore, we experience various problems like inflation accelerates as unemployment rates fall; elasticity of demand and supply usually change as price change; diminishing marginal utility; diminishing marginal returns; diminishing trend of stock return due to financial recession etc.
Therefore, critical errors may take place if a nonlinear situation is studied by considering a linear model.
Suppose, COVID19 (denoted by Y is a dependent variable) spreads exponentially when ‘t’ (times an independent variable) increases and this situation can be shown as under:
t 
Figure 1 COVID spreads 
Estimated Reg. Line 
Y 
It is observed from figure 1 that the estimated regression line only correctly represents the population line at two points. The estimated regression line moves towards right over time with large forecast errors. So, it is important to choose an appropriate functional form to study this situation. The above situation can be addressed by considering an exponentiallogarithmic function.
The properties of exponentiallogarithmic functions are as under:
(i). generally, exponential function looks like , > 0, where ‘ ’ represents the base (constant), represents exponent and ‘y’ is growing exponentially as long as .
(ii). Log is the inverse of exponential function that takes the form (a logarithm of y to a given base say ‘ ’ is the power to which ‘ ’ must be raised in order to arrive at ‘y’.
(iii). The most common base for exponential function is the constant ‘ ’, where
So, the most common logarithms are logs to the base ‘ ’ (natural logs) looks like to that can be shown as under:
Figure 2: Exponential Curve 
Figure 3: Logarithmic Curve 
(iv). both exponential and logarithmic functions are monotonic increasing transformations. Suppose ‘ ’ and ‘ ’ are two numbers where > , then
(v). log operations for any two positive numbers ‘ ’ and ‘ ’ as below:
(vi). exponential function operations for any two positive numbers ‘ ’ and ‘ ’ as under:
The study applies appropriate functions when necessary to allow nonlinear relationship among the variables. Similarly, derivative of ‘ ’ may also be introduced with respect to ‘ ’ to not be a constant but rather to vary as ‘ ’ changes. The properties of the derivatives of logarithmicexponential functions are as under:
(a). If
(b). If
©. If
(d). If
From the above, it is expected that equilibrium in economics must be maintained and thus, it may be identified the elasticity of ‘ ’ with respect to ‘ ’ for exponentiallogarithmic function as below:
Let, then the properties may be expressed as below:
(a). If
(b). If
After details reviewing the properties of logarithmicexponential functions, appropriate models may be developed those look like linear after transformation where ordinary least square (OLS) estimation is possible. With this view, the present study considers three logarithmicexponential functions.
(I). Let’s start with the linearlog function as under:
(1)
Where, ‘ ’ is the response variable, ‘ ’ is the predictor variable, is the disturbance term with zero mean and constant variance, ‘ ’ and are the coefficients to be estimated.
Equation 1 is appropriate when it is assumed that the underlying relationship between ‘ ’ and ‘ ’ looks like as below:
X 
Figure 4: Diminishing marginal return 
The above figure tells that ‘ ’ follows diminishing marginal returns when ‘ ’increases in the short run. But, if ‘ ’ variable is transformed into log then the above figure looks like as below:
Y 
Figure 5: Increasing marginal return 
Thus, by transforming one of the variables into log form we can follow the OLS technique to measure linear relationship and only then the OLS estimators is recognised as BLUE. Here, the estimator is no longer report the marginal effect of ‘ ’ upon ‘ ’ rather it reports the marginal effect of upon ‘ ’. So, first of all untransform the estimators generated by the regression equation and then interpret.
So, the marginal effect of ‘ ’ upon ‘ ’ may be shown as under:
(2)
It follows a diminishing marginal effect when ‘ ’ increases and the elasticity of ‘ ’ with respect to ‘ ’ may be written as below:
(3)
But, here the equilibrium changes as ‘ ’ changes which indicates average level of ‘ ’.
(II). The study also considers a loglinear (semilog) model that can be presented as below:
(4)
Consider equation 1 again (linearlog model), where log is used to transform a variable in the right hand side (independent variable) to allow unbend a concave line to straight line. In the same way, if log transformation is used in the left hand side (dependent variable) of a variable then it is possible to allow unbend a convex and exponential curve into a straight line.
Suppose, COVID19 (represented by ‘ ’) grows exponentially at a rate ‘ ’ at any point of time ‘ ’ and this relationship may be expressed as under:
(5)
Where, zero (0) represents any subjective starting time point. The spreading rate or growth rate (‘ ’) of COVID19 can be estimated only when the natural log of both sides is taken possible in equation 5 as under:
(6)
The above notation can be simplified as below:
Now, inserting an error term in equation 6 to each time period to allow disturbances then equation 6 looks like:
(7)
Now, if exponential rule is followed in equation 7 then it looks like as below:
(8)
So, it is possible to estimate the unknown spreading rate or growth rate of COVID19 (represented by ‘ ’) by regressing the natural log of ‘ ’ against time ( ) and then untransform the coefficients for their interpretation. Here, the regression slope coefficient measures the natural log of which is equal to the percentage change in ‘ ’ per one unit change in ‘ ’. The underlying spreading rate denoted by ‘ ’ may be estimated as below:
(9)
On the other hand, the marginal effect of ‘ ’ (time) upon ‘ ’rather than the spreading rate or growth rate can be estimated by applying the properties of the derivatives of logarithmic and exponential functions which is explained above:
(10)
and the equilibrium between ‘ ’ and ‘ ’ can be shown as under:
(11)
So, to forecast ‘ ’ after putting some expected value of ‘ ’ into the regression equation 8 it looks like as under:
(12)
If = 1, then the forecast is unbiased, consistent and efficient but if
(13)
then the forecast is biased, inconsistent and inefficient. This problem may be corrected to forecast ‘ ’ by using the following loglinear regression model as under:
(14)
Where, represents the sample variance of the disturbances.
(III). Finally, the study considers a loglog (double log) model to analyse the effect of COVID19 into the human life. Therefore, COVID19 spreading, death by COVID19, COVID19’s active cases and recover from COVID19 are taken into consideration.
Generally, a loglog model can be written as under:
(15)
The above equation can be linked with the CobbDouglas production function in a judicious way. According to CD production function, a firm produces ‘ ’ by using two inputs ‘ ’ (labour) and ‘ ’ (capital) and also subject to changes the technology measured by the parameter ‘ ’ and that can be presented in the following form:
(16)
After taking log on both sides and adding an error term in the equation it looks like as below:
(17)
Where, , ‘ ’ and ‘ ’are the coefficients to be estimated. But the expected coefficients are the unbiased estimators of the equilibrium of ‘ ’ with respect to the exogenous variables ‘ ’ and ‘ ’. However, the sum of the estimates gives a measure of returns to scale. If the sum is greater than one meaning that increasing returns to scale and viceversa and also if the sum is equal to one then indicates constant returns to scale.
Here, the marginal effect of ‘ ’ upon ‘ ’can be derived by applying the derivatives of logarithmicexponential functions as under:
(18)
So, based on the CD production function the COVID19 issue can be formulated as under:
(19)
Here, it is assumed that COVID19 affects human life in various ways. Due to COVID pandamic, a lot of people has already been died, many people is suffering till now and many of them have recovered subject to changes the medical treatment measured by the parameter ‘ ’.
Now, after taking log on both the sides and putting an error tem in equation 19 the converted equation looks like:
(20)
Where, ; represents the growth or transmission of COVID19 among the people; indicates the patients died in COVID19; tells about patients are suffering from COVID19 till now; signifies that how many patients are recovered from COVID19 after treatment; ‘ ’, ‘ ’ and ‘ ’are the coefficients to be estimated by OLS technique and ‘ ’ is the disturbances term with its usual assumptions.
Results & Analysis
The world scenario of COVID19 from 22^{nd} January to 24^{th} May, 2020 is given in figure 1. It is observed that the transmission of COVID19 to the human body during this short period of time is increasing sharply (exponential rate) only a little relief on 16^{th} May, 2020 and then again as before. But, the recovery rate after medical treatment is also increasing which is a good sign to us. Although the recovery rate is lower as compared to the transmission rate. It is also found that the graph of the active cases who are under medical treatment in between the graphs of transmission case (COVID19) and recovery case. Here, it is found that the direction of the active graph will follow a diminishing trend if recovery rate is higher and will cross the active graph very soon if medical treatment is improved. The death rate is also in increasing trend but the rate of growth is very low as compared to recovery and the death rate will be under control if proper medical treatment could be provided.
Figure 6: World Scenario of COVID (22^{nd} January to 24^{th} May, 2020) 
The coefficient of the estimated regression models are presented in table 1. It is found that the estimated coefficient of the linearlog model is positive and statistically significant. In this model time (independent variable) is denoted as and transformed into log form. So, to interpret the coefficient, it has to be divided by hundred (1243010.00/100) and the value is 12430.00 that means when time (independent) is increased by 1% then the dependent variable increases by about 12430.00 or in other words, for every 10% increase in time ( ), transmission of COVID19 to the human life increases by about [1243010.00*log(1.10)] 51451.52. So, it may be opined that human transmission by COVID19 is increased sharply according to the time.
Similarly, it is observed that the estimated coefficient of the loglinear model is also positive and statistically significant. In case of levellevel regression model, it is very easy to interpret the slope coefficient. But, in this case slope coefficient cannot be interpreted simply. So, to interpret the slope coefficient, first exponentiates and then subtracts one from this coefficient and then multiply by 100 and then the resultant value indicates the percent increase (or decrease) in the dependent variable for every oneunit increase in the independent variable. Here, the slope coefficient is 0.0600 that can be interpreted after converting the slope coefficient in this form [exp(0.0600) – 1)* 100 = 6.19]. So, it may be interpreted that for every one day increase in time (independent variable), the COVID19 spreading to the world population increases by about 6.19%.
Finally, the outcome of loglog model (double log) is presented in table 1. It is found that all the slope coefficients are positive and statistically significant. Before interpreting the death coefficients of the loglog model, first the coefficient is to be converted into this from (1.01^{0.3053}1)*100 that means for 1% increase (decrease) in the independent variable then what percent changes is occurred in the dependent variable. After converting the coefficient of death it is found 8.47% that means 1% increase in death by COVID19 in world population the spreading rate of COVID19 into human population is 8.47%. Similarly, the coefficient of active case is found 0.4815 and after converting it becomes 20.84% that means 1% increase in active case (till suffering from COVID19) or who are in treatment, the spread of COVID19 into the world population is 20.84%. In the same way, the slope coefficient of recovery rate is observed 0.0787 and after conversion it is found 0.60% that indicates that 1% increase in recovery rate the spreads of COVID19 into the population is 0.60%. If it is extended in this way that if 20% increase in recovery rate the spreads of COVID19 into the world population is found to be 1.44%. So, it may be said that if recovery rate is increased in a rapid way with proper treatment then spreading of COVID19 in the world populace may be controlled. Finally, the slope coefficient of time is 0.2016 and after converting, it is found 3.77% that means that 1% increase in time, the spreading of COVID19 to the population is 3.77% or almost four times that means with the enhancement of time, the spreading of COVID19 into the human life has increased harshly.
From the above results it may be determined which model is appropriate in this perspective. So, coefficient of determination (R^{2}) is used to identify the best model. It is observed that each model has produced R^{2} value (goodness of fit). It is assumed that model suitability depends on higher R^{2} value. It is observed that the R^{2 }value of the loglog model (Double log) is higher as compared to the linearlog model as well as loglinear model and thus, it may be said that loglog model is an appropriate model to fit the time series historical data on COVID19 of the world population.
Table 1: Regression estimates of the exponentiallogarithmic models
Model 
Beta Coefficients 
R^{2} 

β_{1} 
β_{2} 
β_{3} 
β_{4} 

LinearLog Model:

1243010.00* (10.1373) 0.0000 



0.4572 
LogLinear Model:

0.0600* (40.6788) 0.0000 



0.9313 
LogLog Model: ln(COVID19) = α_{0}+β_{1}ln(Death)+β_{2}ln(Active)+ β_{3}ln(Recover)+ β_{4}ln(Day)+ε 
0.3053* (10.81311) 0.0000 
0.4815* (18.5552) 0.0000 
0.0787* (3.9988) 0.0001 
0.2016* (4.1656) 0.0001 
0.9993 
*Significant at 5% level
Source: Authors’ own calculation
Conclusion & Recommendation
The present study examines the impact of COVID19 on human populace by considering three nonlinear regression models. It is found from the figure that spreads of COVID19 on population is increased at an exponential rate with respect of time. But the death curve is not hike so much like others. However, there are large differences on the results estimated by all the regression models but the coefficients are positive and statistically significant in all cases. It is also concluded that loglog model is an appropriate model to capture the entire situation. It is observed that all the independent variables are statistically significant and positively influence the spreading of COVID19 on world population. The COVID19 has spread over time almost four times but spreading of COVID19 is not so increased sharply according to the recovery rate. It is also found that when active cases is increasing then positive cases are also increased in a rapid way. But when positive cases are compared with the death rate then it is found that spreading rate is almost 8.5% where death rate is 1%. So, it may be opined that if health services are improved then positive cases may be controlled. In this situation more testing is required, proper isolation must be arranged, good treatment is required to control over the situation until vaccine is not developed. It may be recommended that global corona virus pandemic teaches us lesson like strong public health systems have the flexibility to address massive health threats with the collective responses they require. Unnecessary privatization of health services and individualization of risks might further weaken our ability to address this and future global pandemics. So, the Government will give more attention on health services, social safety and security, foods and financial stability for sustaining human civilization. More research is needed to find out solution from COVID19 spreading and how the whole human civilization will combat with it.
References