Trans. Inst. Indian Geographers ISSN 0970-9851 Indexed in Scopus Factors affecting COVID-19 infection and deaths in the millionplus cities of India Purva Yadav* and Aisharya Bhattacharjee, New Delhi Abstract Big cities are highly complex spatial units and have diverse subpopulations and neighborhoods with different sociocultural needs and vulnerable groups concerning public health emergencies, such as COVID-19. When these cities face an epidemic that spreads rapidly within the urban community, then human density and the urban setting have proven to be a tremendous liability and a matter of concern. The present study in this context is a preliminary attempt to understand the COVID-19 pandemic in the Indian context. The paper presents a city-level study of 46 million-plus cities in India to identify major determining factors (demographic, economic, environmental, infrastructural, and institutional interventions) of the COVID-19 infection and related deaths in different lockdown and unlock phases. Using the multiple regression model we found that the impact of these factors does significantly explains the variation in the COVID-19 infections and related deaths. However, the role of individual indicators does seem to have a differential impact across phases of the lockdown strategy. Indicators, such as GDP, hospital-doctor ratio, public transport usage, and administrative status of the city have been found to be the most significant factors influencing COVID-19 cases and deaths. On the other hand, co-morbidities do not appear as consistent significant factors, while the much-debated density parameter plays an insignificant role in the Indian big cities. We would like to emphasize that the results are at best indicative in nature, and for an in-depth understanding of each of these factors and spatial complexity, we require further detailed analysis at a more disaggregated level. Keywords: COVID-19, million-plus cities, India, lockdown, regression models Introduction Big cities are highly complex spatial units and are not only dependent on each other (regionally and globally) but are also tied with the neighboring small towns and rural areas. Moreover, these urban areas have diverse subpopulations and neighborhoods with different socio-cultural needs and vulnerable groups concerning public health emergencies, such as COVID-19. When these cities face an epidemic that spreads rapidly within the urban community, then human density and the urban setting have proven to be a tremendous liability and a matter of concern. In developing countries like India with inherent diversity and stark disparity, a crisis such as COVID-19 highlights vulnerabilities of the cities which are intricately tied with the space-blind and uniform policies. The urban risk from extreme events is affected by the location, density, scale, and connectivity. For example, location and scale Transactions | Vol. 43, No. 2, 2021 | 227 (the positioning and city size) have linear effects, primarily affecting the population and the availability of resources. On the other hand, density and connectivity often have nonlinear impacts, both positive and negative. In addition, the urban risk is also affected by the feedback effects of interaction between the urbanization process and climate change (Siri et al., 2015). At this point, the vector and velocity of modern transport networks would play a key role in turning the epidemic into a pandemic. And in the case of COVID-19, it has been proven to be true. For example, a recent study by Yadav and Bhattacharjee (2020) has noted the impact of urbanization and connectivity on COVID-19 spread in India for the initial phases of the lockdown. The urban vulnerability may arise for many extreme events, and many times the factors that make cities attractive are also the ones that increase the risk. Cities are primarily high-density zones with heavy resource dependence and often resulting in their depletion. Land-use change is often witnessed in and around the cities. Hence, they are especially vulnerable to extreme events and often play major roles in aggravating the intensity of such events. Evidences suggest that civilizations (and, indeed, cities) that generate unsustainable demand are vulnerable to rapid collapse. If we track the spatial trajectory of the ongoing pandemic, almost 90 percent of COVID-19 cases are found in urban areas and the consequent impact is most damaging in cities. In an urban setting where risks of transmission and its multidimensional impacts, such as social, physical, and economic are much higher (primarily due to a greater concentration of people and activities as well as better linkages with other regions), data and investigation are crucial 228 | Transactions | Vol. 43, No. 2, 2021 for early outbreak detection and response (testing, diagnosis, isolation, contact tracing, and quarantine) to contain the transmission of the virus (UN, 2020). In this backdrop, it is important to understand the nuances of contagion from spatial-temporal perspectives. The current paper is a preliminary attempt in this direction to identify the factors affecting the transmission of this virus and the resultant deaths with regard to the million-plus cities in India. Theoretical context and purpose Multiple combinations of risk factors have been hypothesized to have an impact on COVID-19 cases and deaths. Yet not many studies have analyzed the potential confounding effects of these factors (Priyadarsini and Suresh, 2020). The present study with the help of multivariate analysis attempts to identify the factors affecting COVID-19 cases and death in the 46 millionplus cities of India (Figure 1) that have become the epicenters of the pandemic. The choice of the variables hypothesized as significant factors in the spread of COVID-19 has been guided by studies that have tried to portray the relationship of certain variables with COVID-19 infection and related death rates. For example, several studies have focused on the impact of density on COVID-19 infection and mortality rates (Arif and Sengupta, 2020; Hamidi, et al., 2020; Henderson, 2020; Rocklöv and Sjödin, 2020; Shoichet and Jones, 2020). Hughes (2020) has noted that the metropolitan region with a higher number of counties linked with one another has become more vulnerable to the pandemic. This is further supported by Papandreou (2020) who pointed out that the connectivity factors have a direct effect on the transmission of the pandemic in a Fig. 1: Million plus cities in India Transactions | Vol. 43, No. 2, 2021 | 229 local-global-local transportation pattern. Also, the role of public transport has been emphasized, drawing from a cluster case in Hunan (S. Chen, 2020; Meinsenzahi, 2020). Certain economic aspects have also affected the infection and death rate. For instance, a significant positive association between GDP at the prefectural level and confirmed cases has been identified in selected provinces of China by Zhang et al. (2020). Moreover, a positive association between the poverty level and COVID-19 cases is also traceable in some studies (Finch & Finch, 2020). Apart from economic aspects, infrastructural aspects like rigorous testing for COVID-19 (advocated for pandemic control) have been found to possess a positive association with confirmed cases and consequently the recovery rate (Engelberg, 2020; Beaubian, 2020). However, it has also been noted that the testing rates vary significantly across space. For instance, according to Monnat and Cheng (2020), testing rates are lower in areas with more black and poor residents in the United States. This inevitably draws attention to certain social indicators. Low levels of literacy have been found to pose a substantial barrier in checking the spread of COVID-19 (Lopes and McKay, 2020; Frieden, 2020; Paakkari and Okan, 2020). Another important aspect pertains to health facilities, particularly the strength of the hospital staff. However, given the proliferating number of cases, health infrastructure has become burdened, and hence a distinct relationship between them is far from being acknowledged (Agarwal, 2020; Sebastian, 2020; Cavallo et al., 2020; Singh et al., 2020). Moreover, institutional factors such as government effectiveness and the extent to which ruling parties are well established have also explained variations in COVID-19 transmission rates (Maor & Howlett, 2020). 230 | Transactions | Vol. 43, No. 2, 2021 Studies have also focused on environmental factors mainly because of “flu” like symptoms and have revealed higher epidemic rates of COVID-19 with lower average temperature (Alvarez-Ramirez and Meraz, 2020; Ficetola and Rubolini, 2020; Pirouz et al., 2020). Moreover, a positive association between atmospheric pollution levels and a high level of COVID-19 lethality has been noted, since high pollutant levels render the population more vulnerable to respiratory diseases (Conticini, et al., 2020; Setti, et al., 2020). In fact, the discourse on incumbent respiratory disorders and COVID-19 necessarily brings into context an in-depth study of how vulnerable groups are affected by COVID-19. WHO (2020) has identified vulnerable groups in urban areas that include the elderly population and persons with underlying medical conditions and several studies have also reflected a positive association between underlying medical conditions like chronic kidney disease, obesity, cardiac issues, diabetes mellitus to the COVID-19 infection rate and deaths (BMJ, 2020; Centre for Disease Control and Prevention, 2020). In this backdrop, the present study provides a holistic viewpoint to capture the causality of the multiple set of factors to COVID-19 cases and deaths in Indian cities. Database and Methodology To better understand the causal factors, a multivariate regression model is used. To deal with heteroscedasticity, we have used robust standard errors (see Yadav, 2020 for details). At the initial stage of the study, 22 indicators were selected, but due to the multicollinearity, three indicators namely, the share of the elderly population, the share of obese/overweight population, and the presence of international airport have been dropped. Finally, 19 indicators have been used in the model as outlined in Figure 2, along with the data sources. The dependent variables namely, COVID-19 cumulative cases and deaths reported in the cities have been collated from the public data platformhowindialives.com. Regression was run on the share of cumulative COVID-19 cases and deaths for the starting date of each of the four lockdowns and five phases. The government has announced nationwide lockdown phases and the same are mentioned below: y Lockdown Phase I-25th March-14th April 2020 y Lockdown Phase II-15th April-3rd May 2020 y y Lockdown Phase III-4th May-17th May 2020 th th Lockdown Phase IV-18 May-30 May 2020 These are followed by five unlock phases as given below: y Unlock phase I-1st June-30th June 2020 y Unlock phase II-1st July-31st July 2020 y Unlock phase III-1st August-31st August 2020 y Unlock phase IV-1st September-30th September 2020 Unlock phase V-1st October-30th October 2020 y Analysis and Discussion Results of the multivariate regression models (Tables 1 to 4) have reflected revealed that the set of multiple factors, such as demographic, economic, infrastructural, institutional, and environmental, explains about 75 percent to around 91 percent of the variation in COVID-19 cumulative cases reported in the 46 cities of India while it explains nearly 68 percent to 91 percent of the variation in the share of the reported cumulative deaths across the different lockdown and unlocks phases considered in the paper. The analysis is divided into sections-I and II presenting regression results of COVID-19 cases and related deaths respectively and is confined to the indicators that are statistically significant in explaining each of them. Population density, the share of the anemic population, road connectivity to National Highways, literacy rates, hospitals per lakh population, and ruling political party are the indicators that are statistically insignificant for explaining cases and deaths in different phases. Section I: Factors affecting COVID-19 infection in million-plus cities in India: The share of cases is significantly affected by the economic size of the city and is positively associated (Table 1 and 2). This is directly related to the fact that big cities act as economic giants with agglomeration effects and transport hubs that encourage higher flows and the associated risk of the virus spread. In line with a similar argument, we hypothesized that cities that are administrative capitals will report more infection. Results have lent support to this with a significant positive association for lockdown phase 1.0 and unlock phase 3.0, 4.0, and 5.0. Social infrastructure indicators, such as hospital beds per lakh population have significantly affected the cases at the onset of lockdown 3.0 and 4.0. This could primarily be because the rapid increase in the number of cases resulted in immense pressure on the already skewed public health infrastructure resources. In fact, at the onset of lockdown 3.0 and unlock phase 3.0, 4.0, Transactions | Vol. 43, No. 2, 2021 | 231 232 | Transactions | Vol. 43, No. 2, 2021 Fig. 2: Factors affecting COVID-19 Transmission and Deaths Table 1: Factors affecting COVID-19 transmission in Lockdown Phases - Multivariate Regression Variables Share of Cases (Dependent Variable) 25th March Coef. t stat Population Density in persons per sq. km 0.00 1.04 Share of anaemic population above 15 years 0.00 -0.01 Share of population above 15 years with high blood sugar level Share of population above 15 years with high state of hypertension GDP (Rs. Crore) -0.01 -0.11 0.22 0.45 0.00 3.44* HC Ratio -0.01 -1.08 Literacy Rate -0.02 -0.64 Hospitals per lakh population Hospital Beds per lakh population Hospital Doctors per lakh population 0.02 0.19 0.00 1.03 -0.02 -1.96 15th April 4th May Coef. t stat Coef. t stat 0.00 0.59 -0.01 -0.19 0.00 0.48 19th May Coef. t stat 0.00 0.57 0.03 0.42 0.03 0.58 0.12 0.74 0.19 0.84 -0.05 -0.27 1.43 1.86 1.38 1.7 1.42 1.86 0.00 4.35* 0.00 3.62* 0.00 4.16* 0.02 0.00 -0.18 0.00 0.17 -0.03 -0.44 -0.01 -0.09 0.05 0.72 0.14 0.91 0.17 1.5 0.01 2.74* 0.00 2.6* -0.01 -0.95 -0.04 -2.05* -0.03 -1.75 0.79 -0.03 -0.24 0.00 0.7 Public Transport Usage by Other Worker (%) 0.01 0.51 0.03 0.69 0.03 0.67 0.08 2.21* Road Connectivity to National Highways 0.35 1.48 0.10 0.3 0.18 0.47 -0.07 -0.2 Average Temperature (in C) 0.02 0.39 -0.17 -1.47 -0.24 -1.91 -0.16 -1.79 Annual average PM<= 10 conc. 0.00 -0.06 1.19 0.02 1.66 -1.37 Annual average NO2 conc. Time interval (in days) 0.01 1.56 0.01 0.00 -0.19 -0.05 -1.43 -0.06 -1.16 -0.05 -0.07 -1.35 -0.11 -1.12 -0.22 -1.94 -0.18 -2.02* Administrative Status of City 1.74 2.74* 2.01 1.98 2.72 1.81 2.02 1.71 Ruling Political Party in State 0.03 0.27 0.36 0.41 0.44 0.21 0.28 -1.14 -1.67 -1.53 -1.61 -0.85 -1.08 0.29 -3.39 -0.47 Number of Testing Centres 0.08 -0.65 -1.29 _cons 0.58 2 0.14 R = 0.74 *statistically significant at 5 percent; Coef. = Coefficients and 5.0, the association between hospital doctor ratio and the COVID-19 cases has emerged significantly with negative relation indicating that cases surged up tremendously owing to fewer resources in terms of health care professionals and the existing huge gap between the demand and supply of the medical staff. 3.70 2 0.46 R = 0.80 2.78 2 R = 0.75 2 R = 0.80 Source: Based on Authors’ calculations If we look at the transport aspect and its role in the spread of the contagion, it is interesting to note that with the start of the lockdown phase 4.0 and unlock phase 1.0, 2.0, and 3.0, usage of public transport by “other” workers category1 has emerged significantly with a positive association. The share of public transport usage is 1 Census of India (2011) provides data for the mode of travel to the top workplace by only “Other Workers” category i.e. other than cultivators, agricultural labourers and household industry workers. The defense forces and similar paramilitary personnel are not included. If the person was engaged in more than one economic activity during the last year, travel to the main economic activity is considered. Transactions | Vol. 43, No. 2, 2021 | 233 Table 2: Factors affecting COVID-19 transmission in Unlock Phases-Multivariate Regression Results Variables Share of Cases (Dependent Variable) 1st June Coef. t stat 1st July Coef. t stat 1st August Coef. t stat 1st September Coef. t stat 1st October Coef. t stat Population Density in persons per sq. km 0.00 0.63 0.00 1.23 0.00 1.69 0.00 1.58 0.00 1.38 Share of anaemic population above 15 years 0.05 0.69 0.00 0.05 -0.01 -0.25 0.00 0.03 0.00 0.15 Share of population above 15 years with high blood sugar level -0.17 -0.68 Share of population above 15 years with high state of hypertension 1.38 1.51 1.63 1.66 0.42 0.80 0.06 0.16 -0.06 -0.19 GDP (Rs. Crore) 0.00 3.52* 0.00 8.97* 0.00 6.62* 0.00 4.92* 0.00 4.84* -0.01 -0.20 0.01 0.43 -0.02 -1.61 Literacy Rate 0.07 0.93 0.10 1.42 0.02 0.52 0.00 0.04 0.00 0.00 Hospitals per lakh population 0.25 1.91 0.14 1.77 0.09 1.53 0.05 0.82 0.05 0.83 Hospital Beds per lakh population 0.00 0.97 0.00 -0.66 0.00 -0.74 0.00 -0.86 0.00 -1.30 Hospital Doctors per lakh population -0.03 -1.49 0.00 -0.02 PublicTransport Usage by Other Worker(%) 0.13 2.70* 0.13 3.61* 0.06 2.64* 0.02 0.94 0.00 0.07 Road Connectivity to National Highways -0.10 -0.23 -0.45 -1.16 0.00 -0.02 0.13 0.83 0.19 1.41 Average Temperature (in C) -0.15 -1.50 -0.04 -0.61 -0.01 -0.37 -0.02 -0.66 -0.03 -1.13 Annual average PM<= 10 conc. 0.02 1.49 0.01 1.55 -0.01 -1.43 -0.01 -2.61* Annual average NO2 conc. -0.06 -1.30 -0.03 -1.54 0.02 1.48 Time interval (in days) -0.23 -2.15* -0.07 -1.24 HC Ratio -0.47 -2.19* -0.33 -2.74* -0.01 -2.31* -0.08 -2.10* -0.20 -2.38* -0.02 -2.22* -0.02 -3.51* 0.02 2.30* -0.07 -2.24* -0.15 -2.11* -0.02 -2.12* -0.01 -3.61* -0.01 -2.93* 0.02 2.56* -0.06 -2.24* Administrative Status of City 2.38 1.61 1.06 1.42 1.05 2.36* 0.83 2.84* 0.67 2.77* Ruling Political Party in State 0.44 0.46 -0.19 -0.24 -0.10 -0.21 -0.04 -0.13 -0.01 -0.03 Number of Testing Centres -0.82 -0.86 0.11 0.17 0.02 0.05 0.09 0.40 0.11 0.57 _cons -5.87 -0.70 -8.56 -1.43 -0.30 -0.07 1.90 0.53 2.22 0.69 R2 = 0.79 R2 = 0.88 *statistically significant at 5 percent; Coef. = Coefficients 234 | Transactions | Vol. 43, No. 2, 2021 R2 = 0.91 R2 = 0.89 R2 = 0.89 Source: Based on Authors’ calculations Table 3: Factors affecting COVID-19 deaths in Lockdown Phases - Multivariate Regression Results Variables Share of Deaths (Dependent Variable) 15th April Coef. t stat 4th May Coef. t stat 19th May Coef. t stat Population Density in persons per sq. km 0.00 0.15 0.00 -0.20 0.00 -0.26 Share of anaemic population above 15 years 0.12 0.55 0.05 0.54 0.07 0.81 Share of population above 15 years with high blood sugar level Share of population above 15 years with high state of hypertension GDP (Rs. Crore) 0.83 1.26 0.52 1.73 0.60 2.18* 2.84 1.16 1.33 1.21 0.99 1.03 0.00 2.86* 0.00 2.74* 0.00 2.71* HC Ratio -0.08 -1.11 -0.03 -0.75 -0.01 -0.36 Literacy Rate -0.15 -0.72 -0.03 -0.30 0.01 0.10 0.40 0.95 0.19 0.99 0.23 1.31 0.02 3.39* 0.02 4.73* 0.01 4.96* Hospitals per lakh population Hospital Beds per lakh population Hospital Doctors per lakh population -0.17 -3.03* Public Transport Usage by Other Workers(%) -0.04 -0.29 0.01 0.15 0.03 0.69 1.55 1.47 0.54 1.10 0.39 0.89 -0.37 -2.14* -0.30 -1.99 Road Connectivity to National Highways Average Temperature (in C) Annual average PM <= 10 conc. -0.87 -2.29* -0.07 -2.99* -0.06 -2.87* 0.00 0.15 0.01 0.78 0.01 1.12 Annual average NO2 conc. -0.10 -0.70 -0.04 -0.71 -0.04 -0.70 Time interval (in days) -0.85 -2.53* Administrative Status of City Ruling Political Party in State 9.31 2.13* 1.21 0.48 Number of Testing Centres -6.75 -2.46* _cons 27.54 1.05 R2 = 0.68 *statistically significant at 5 percent; Coef. = Coefficients tremendously high for the metropolitan cities and the risk of COVID-19 infection has a direct connection to it, as triggered by inhaled aerosols, or tiny particles, breathed out by the infected passengers onboard (Chen, 2020). Also, the adherence of the passengers to the COVID-19 related protocols in public transport modes has remained a daunting task for the authorities to enforce. It is interesting that with the staggered easing of mobility restrictions since lockdown phase 4.0, there was a fear of the increase in the number -0.37 -2.41* -0.33 -2.43* 4.11 2.01* 0.67 0.57 0.66 0.64 -2.83 -2.19* -1.97 -1.68 0.28 0.03 6.91 0.58 R2 = 0.71 3.21 1.79 R2 = 0.74 Source: Based on Authors’ calculations of cases in these cities, and the same was supported by the regression results. Furthermore, the time gap between implementation of the lockdown by respective states and that of the Central government has emerged significantly with the negative association since the onset of lockdown phase 4.0 because the states which imposed a total lockdown, days preceding the national lockdown were better able to check the spread of the pandemic. Last but Transactions | Vol. 43, No. 2, 2021 | 235 Table 4: Factors affecting COVID-19 deaths in Unlock Phases - Multivariate Regression Results Variables 1st June t stat Coef. t stat 1st September Population Density in persons per sq. km 0.00 -0.16 0.00 0.40 0.00 1.07 0.00 1.09 0.00 0.75 Share of anaemic population above 15 years 0.08 1.07 0.08 1.13 0.04 0.76 0.02 0.71 0.03 0.92 Share of population above 15 years with high blood sugar level 0.34 1.30 -0.01 -0.03 -0.23 -1.44 -0.16 -1.50 -0.13 -1.39 Share of population above 15 years with high state of hypertension 0.95 1.06 1.00 1.29 0.72 1.22 0.34 0.80 0.01 0.03 GDP (Rs. Crore) 0.00 3.23* 0.00 4.18* 0.00 4.88* 0.00 4.47* 0.00 4.29* -0.01 -0.24 0.01 0.34 -0.01 -0.80 -0.02 -1.73 Literacy Rate 0.03 0.32 0.06 0.86 0.04 0.77 0.03 0.89 0.04 1.36 Hospitals per lakh population 0.29 1.88 0.21 1.61 0.14 1.39 0.12 1.39 0.18 2.63* Hospital Beds per lakh population 0.01 4.58* 0.00 1.03 0.00 0.72 0.00 0.31 0.00 -0.83 -0.06 -2.70* -0.03 -1.40 -0.03 -2.41* Coef. t stat 1st October Coef. Hospital Doctors per lakh population Coef. 1st August Share of Deaths (Dependent Variable) HC Ratio t stat 1st July -0.03 -3.21* Coef. t stat -0.03 -2.38* -0.02 -3.30* PublicTransport Usage by Other Worker (%) 0.09 1.75 0.11 2.53* 0.11 3.69* 0.07 2.95* 0.04 2.30* Road Connectivity to National Highways 0.32 0.73 -0.01 -0.02 0.17 0.65 0.24 1.27 0.21 1.25 -0.26 -2.04* -0.13 -1.31 -0.03 -0.42 -0.04 -0.78 Average Temperature (in C) -0.09 -2.18* Annual average PM <= 10 conc. 0.02 1.29 0.02 1.77 0.01 0.78 0.00 -0.40 0.00 -1.22 Annual average NO2 conc. -0.04 -0.83 -0.07 -1.35 -0.02 -0.78 0.00 0.17 0.01 0.65 Time interval (in days) -0.33 -2.61* -0.20 -1.84 -0.16 -2.45* -0.14 -3.03* -0.15 -3.66* Administrative Status of City 3.04 1.80 2.18 1.51 2.55 2.67* 1.89 3.03* 0.89 2.00* Ruling Political Party in State 0.78 0.79 0.65 0.74 0.58 0.97 0.54 1.27 0.47 1.28 Number of Testing Centres -1.80 -1.62 -0.88 -0.95 -0.99 -1.58 -0.66 -1.61 -0.06 -0.22 _cons -2.51 -0.26 -7.79 -0.96 -5.21 -0.93 -2.68 -0.63 -1.29 -0.35 R2 = 0.79 R2 = 0.83 *statistically significant at 5 percent; Coef. = Coefficients 236 | Transactions | Vol. 43, No. 2, 2021 R2 = 0.89 R2 = 0.90 R2 = 0.91 Source: Based on Authors’ calculations not the least, environmental factors like PM 10 concentration have been found to have a significant (negative association) while NO2 concentration has a significant positive association since the onset of unlocking phase 4.0 and 5.0. Although the reason behind the former is unclear, the latter is directly related to the incumbent risk of getting affected by respiratory diseases. Section II: Factors affecting COVID-19 deaths in million-plus cities in India Tables 3 and 4 relate to the factors affecting COVID-19 related deaths in the big cities of India. Economic size (positive association), hospital doctors ratio and time gap (negative association) have been found statistically significant in affecting the share of COVID-19 related deaths in different lockdown phases. At the onset of lockdown phase 4.0, a positive association between co-morbidity i.e. share of high blood pressure and death cases is evident. The same has been noted in research conducted by the Centre for Disease Control and Prevention (2020) where patients with high blood pressure were associated with an increased risk of COVID-19 deaths. Similar to COVID-19 infection, the share of public transport usage by the “other” worker category has emerged to have significantly affected the COVID-19 related deaths since the onset of the unlock phase 2.0. This could mainly be because workplaces have gradually opened up, particularly for those activities where work from home option is not viable or are part of essential services; hence to meet the mobility demand public transport started operating in a staggered manner, which has increased the contagion risk and the consequent deaths. Furthermore, the environmental factor i.e. average temperature has emerged significantly with a negative association in lockdown phase 2.0, 3.0, and unlock phase 1.0 and 5.0; however, requires further detailed analysis for a better understanding of this. Other factors that have been found to yield a significant effect on COVID-19 deaths, are capital city status (positive association) at the onset of lockdown phase 2.0, 3.0 and unlock phase 3.0, 4.0, and 5.0, and the number of testing centers (negative association) in phase 2.0 and 3.0 of the lockdown phase. Conclusions In the Indian context, the impact of the set of multiple factors, such as, demographic, economic, infrastructural, environmental, and institutional does significantly explains the variation in the COVID-19 infections and related deaths. However, the role of individual indicators does seem to have a differential impact on the dependent variables in different phases of the lockdown strategy. Indicators, such as GDP, hospital-doctor ratio, public transport usage, and administrative status of the city have been found to be the most significant factors influencing COVID-19 cases and deaths. On the other hand, for example, co-morbidities do not appear as consistently significant factors, while the much-debated density parameter plays an insignificant role in the Indian big cities. Among others, these are few interesting empirical observations based on our preliminary understanding of the COVID-19 pandemic in India from the spatial-temporal perspective. When we started working on this study, the COVID-19 situation was very new, and the availability of consistent comparable data across time did pose a major challenge, besides the lack of real-time comparable data on some of the parameters we have considered Transactions | Vol. 43, No. 2, 2021 | 237 in the paper. Hence the study is limited in its scope but can be a useful contribution to the COVID-19 literature. The results are at best indicative in nature, and for an indepth understanding of each of these factors and spatial complexity, we require further detailed analysis at a more disaggregated level. Since a pandemic may cause sudden, widespread morbidity and social, political, and economic disruption, studies reflecting on regional nuances are much useful though missing in contemporary research and in the policy framework. References: Agarwal, N. (2020, 12-June). Lack of hospital beds amid rising coronavirus COVID-19 cases in Delhi a cause of concern. Retrieved 2020, 12-July from https://zeenews.india. com/india/lack-of-hospital-beds-amidrising-coronavirus-covid-19-cases-in-delhia-cause-of-concern-2289431.html Alvarez-Ramirez, & Meraz. (2020, 23-March). medRxiv. Retrieved 2020, 6-May from Role of meteorological temperature and relative humidity in the January-February 2020 propagation of 2019-nCoV in Wuhan, China: https://www.medrxiv.org/ content/10.1101/2020.03.19.20039164v1 Arif, M., & Sengupta, S. (2020). Nexus between population density and COVID 19 pandemic in the south Indian states: A geo-statistical approach. Environment, Development and Sustainability, 23(3), 1-29. Retrieved 2020, 12-July from Science Daily: https://www.sciencedaily.com/ releases/2020/06/200601101308.htm Cavallo, J. J., Donoho, D. A., & Howard P. Forman. (2020). Hospital Capacity and Operations in the Coronavirus Disease 2019 (COVID-19) Pandemic—Planning for the Nth Patient. JAMA Health Forum, 1(3), 1-16. Centre for Disease Control and Prevention. (2020, 25-June). People with Certain Medical Conditions. Retrieved 2020, 12July from COVID-19: https://www.cdc. gov/coronavirus/2019-ncov/need-extraprecautions/people-with-medical-conditions. html Chen, S. (2020, 20-April). Coronavirus can travel twice as far as official ‘safe distance’. Retrieved 2020, 11-July from South China Morning Post: https://www.scmp.com/news/ china/science/article/3074351/coronaviruscan-travel-twice-far-official-safe-distanceand-stay Conticini, E., Frediani, B., & Caro, D. (2020). Can atmospheric pollution be considered a co-factor in extremely high level of SARS-CoV-2 lethality in Northern Italy? Environmental Pollution, 261. Engelberg, S. (2020, 23-March). The Coronavirus Testing Paradox. Retrieved 2020, 12-July from ProPublica: https://www.propublica. org/article/covid-19-coronavirus-testingparadox-united-states Ficetola, & Rubolini. (2020). Climate affects global patterns of COVID-19 early outbreak dynamics. Science of the total Environment, 761, 1-24. Beaubian, J. (2020, 12-March). Singapore Wins Praise For Its COVID-19 Strategy. The U.S. Does Not. Retrieved 2020, 12-July from The Coronavirus Crisis: https://npr.tumblr.com/ post/612394953510961152/singapore-winspraise-for-its-covid-19-strategy Finch, W. H., & Finch, M. E. (2020). Poverty and Covid-19: Rates of Incidence and Deaths in the United States during the First 10 Weeks of the Pandemic. Frontiers in Sociology, 5(47), 1-10. BMJ. (2020, 1-June). Underlying illness risk factors for severe COVID-19 or death. Frey, W. H. (2020, 2-May). Coronavirus is making some people rethink where they want to live. (CNN, Interviewer) Retrieved 238 | Transactions | Vol. 43, No. 2, 2021 2020, 9-July from CNN: https://www. wyff4.com/article/coronavirus-is-makingsome-people-rethink-where-they-want-tolive-1588513869/32355488# Frieden, J. (2020, 19-June). Lack of Health Literacy a Barrier to Grasping COVID-19. Retrieved 2020, 12-July from MedPage Today: https://www.medpagetoday.com/ infectiousdisease/covid19/87002 Hamidi, S., Sabouri, S., & Ewing, R. (2020). Does Density Aggravate the COVID-19 Pandemic? Journal of the American Planning Association, 86(4), 495-509. Henderson, E. (2020, 3-July). High population density in India associated with spread of COVID-19. Retrieved 2020, 7-July from News Medical: https://www.news-medical. net/news/20200703/High-populationdensity-in-India-associated-with-spread-ofCOVID-19.aspx Hughes, N. (2020, 18-June). Urban Density not linked to higher coronavirus infection rates. Retrieved 2020, 9-July from Science Daily: https://www.sciencedaily.com/ releases/2020/06/200618110953.htm Lopes, H., & McKay, V. (2020). Adult learning and education as a tool to contain pandemics: The COVID-19 experience. International Review of Education, 66, 575–602. Maor, M., & Howlett, M. (2020). Explaining variations in state COVID-19 responses: psychological, institutional, and strategic factors in governance and public policymaking. Policy Design and Practice, 3(3), 228-241. Meinsenzahi, M. (2020, 27-February). Photos show what it’s like to travel around the world by train, bus, boat, and plane in the age of coronavirus. Retrieved 2020, 11July from Business Insider: https://www. businessinsider.in/slideshows/miscellaneous/ photos-show-what-its-like-to-travel-aroundthe-world-by-train-bus-boat-and-plane-in- the-age-of-coronavirus/slidelist/74326701. cms Monnat, S. M., & Cheng, K. J. (2020, 1-April). COVID-19 Testing Rates are Lower in States with More Black and Poor Residents. Syracuse University. Lerner Center for Public Health Promotion. Office of the Registrar General & Census Commissioner, Ministry of Home Affairs, Government of India. (2011). District Census Handbook. Paakkari, L., & Okan, O. (2020). COVID-19: Health literacy is an underestimated problem. The Lancet Public Health, 5(5), 249-250. Papandreou, T. (2020, 27-March). Is the Coronavirus The Transportation Industry’s Opportunity? Retrieved 2020, 16-July from https://www.forbes.com/sites/ timothypapandreou/2020/03/27/is-thecoronavirus-the-transportation-industrysopportunity/?sh=5af228c4752b Pirouz, B., Golmohammadi, A., Masouleh, H. S., Delazzari, C., & Violini, G. (2020, 23July). Relationship between Average Daily Temperature and Average Cumulative Daily Rate of Confirmed Cases of COVID-19. Retrieved 2020, 1-October from MedRxiv: https://www.medrxiv.org/ content/10.1101/2020.04.10.20059337v3 Rocklöv, J., & Sjödin, H. (2020). High Population Densities Catalyse the Spread of COVID-19. Journal of Travel Medicine, 27(23), 186-192. Rodrigue, J. P. (2016). The geography of transport systems (5th ed.). Oxon: Routledge. Sebastian, M. (2020, 19-May). India Confirms Over 1 Lakh Covid-19 Cases: How Many ICU Beds And Ventilators Does The Country Have? Retrieved 2020, 12-July from HuffPost: https://www.huffpost.com/archive/in/entry/ india-covid-19-cases-hospital-beds-icuventilators_in_5ec375adc5b6e607c1990187 Transactions | Vol. 43, No. 2, 2021 | 239 Setti, L., Passarini, F., Gennaro, G. D., Baribieri, P., Perrone, M. G., Borelli, M., . . . Miani, A. (2020). SARS-Cov-2 RNA Found on Particulate Matter of Bergamo in Northern Italy: First Preliminary Evidence. Environmental Research, 188, 216-233. Singh, P., Ravi, S., & Chakraborty, S. (2020, 24-March). Is India’s health infrastructure equipped to handle an epidemic? Retrieved 2020, 12-July from Brookings: https://www. brookings.edu/blog/up-front/2020/03/24/ is-indias-health-infrastructure-equipped-tohandle-an-epidemic/ Siri, J. G., Newell, B., & Proust, K. (2015). Urbanization, extreme events, and health: the case for systems approaches in mitigation, management and response. Asia Pacific Journal of Public Health, 28(2), 15-27. World Health Organisation. (2020). Strengthening Preparedness for COVID-19 in Cities and Urban Settings. London, UK. Yadav, P. (2020). Globalization and India’s international trade: does distance still matter? GeoJournal, 86(1), 1927-1941. 240 | Transactions | Vol. 43, No. 2, 2021 Yadav, P., & Bhattacharjee, A. (2020). Impact of COVID-19 on mobility in India: A spatial approach. Radical Statistics(Special Coronavirus Issue 126), 56-66. Zhang, Y., Tian, H., Zhang, Y., & Chen, Y. (2020, 9-April). Is the epidemic spread related to GDP? Visualizing the distribution of COVID-19 in Chinese Mainland. arXiv. Retrieved 2020, 11-July from arXiv:2004.04387 (q-bio): https://arxiv.org/ abs/2004.04387 Purva Yadav* Assistant Professor and Aisharya Bhattacharjee Research Scholar, Centre for the Study of Regional Development, School of Social Sciences, Jawaharlal Nehru University, New Delhi, India *Author for correspondence E-mail: purvayadavjnu@gmail.com