key: cord-0832882-sva6h76l authors: Arshed, Noman; Meo, Muhammad Saeed; Farooq, Fatima title: Empirical assessment of government policies and flattening of the COVID19 curve date: 2020-08-27 journal: J Public Aff DOI: 10.1002/pa.2333 sha: 13746260879299a772a633ab2ca1f815a1f6e5e7 doc_id: 832882 cord_uid: sva6h76l The objective of the study is 2‐fold. First, it estimates the 2019 new coronavirus disease (COVID19) flattening curve using Panel Random Coefficient Model. This allows each country to have its trajectory while allowing for random error effects to transfer across countries. Second, it calculates the expected number of days to reach the flattening point of COVID19 curve and estimate the empirical effectiveness of government policies around the world using Poisson regression. This study avails global COVID19 incidence data for 190 countries between January 22, 2020 and May 11, 2020. In the absence of a vaccine or of more appropriate treatment options, non‐pharmaceutical approaches must be used to control the spread of the COVID19. This study proposed that the contact tracing, stay at home restrictions and international movement restrictions are most effective in controlling the spread and flattening the COIVD19 curve. At the same time, habits that hurt the immune system like smoking have a negative effect on the flattening of the curve. The government should integrate these policies in their lockdown plan to make it smart lockdown. Coronavirus SARS CoV-2 (COVID19) was detected on January 21st when the first of the Chinese health workers were infected in Wuhan (Nature, 2020; Sharif, Aloui, & Yarovaya, 2020) . While it was spreading in China, the developed county's health systems were of the perception that like previous scourges, that is, yellow fever, cholera, HIV/AIDS and Ebola, their systems will prevail against the COVID19. Practically past experiences were of little use, especially for the case of the US Health System (Chokshi & Katz, 2020) . This was the story for the developed economy; on the other hand, there are more than 63,000 cases in 53 countries of Africa. Their fragile system was not able to cope with the needs of the population (Divala, Burke, Ndeketa, Corbett, & MacPherson, 2020) . Furthermore, the fallout in terms of economic recession after COVID19 is predicted to overshadow the financial crises of the late 2010s. This is because this crisis is not only affecting the developed; rather, it is also changing developing economies. It will create not only supply short but also the demand shock leading to an estimated fall of GDP by 6% (Wren-Lewis, 2020). One study projected Indonesia's poverty to rise from 9.2 to 12.4% (Suryahadi, Al Izzati, & Suryadarma, 2020) . Kilbbin and Fernando (2020) provided several scenarios in which COVID19 could decrease GDP. According to estimates, the GDP loss could be up to −9.9% for Japan, −8.7% for Germany, and −8.4% for the rest of the Euro area. This shock had reduced the market capitalization by 30% at the global level (Siddiquei & Khan, 2020) . Each country handled the COVID19 differently, leading to different consequential forms of the pandemic. Countries like USA, Italy and the UK delayed government intervention in the expectation of achieving community-based immunization. In contrast, countries like India and New Zealand locked down international borders and restricted domestic movements. Lastly, there were countries which used the hybrid approach of allowing businesses to continue with special operating procedures (SOPs), social distancing and smart tracking of the COVID19 incidence. Figure 1 shows the exponential increase in the confirmed cases while the log series show slight flattening after 100 days from the first case of COVID19 in their country. Several studies have proposed the theory to handle the COVID19. One of the most popular is the flattening of the COIVD19 incidence curve. Figure 2 shows the quadratic fit of log COVID19 incidence against the count of days since the first case was reported. It is observed here that there is a flattening pattern emerging globally. Studies claim that COVID19 will eventually expose every one as it is invisible, and it tends to communicate via social interaction. The young ones who are not experiencing its symptoms will be the one spreading it to the masses. The social and medical way to evade this spread is discussed under the measures to flatten the curve. This approach may not eradicate the COVID19, but it will reduce the strain on the health systems and pressure on the development of cure (Anderson, Heesterbeek, Klinkenberg, & Hollingsworth, 2020; Giesecke, 2020; World Health Organization, 2020a) . The policies to manage the spread included lockdowns across the world (Koh, 2020) which are useful to contain the spread (Sharma, Talan, & Jain, 2020) . In the literature, various studies find various factors which contribute to flattening the COVID19 curve. For example, Rotondi, Andriano, Dowd, and Mills (2020) find that in Italy, comovement restriction and social distancing contribute in flattening COVID19 curve. Block et al. (2020) also find that social distancing strategy significantly affects on COVID19 curve and make it flatten. Breitenbach, Ngobeni, and Ayte (2020) ways. First, we include other important variables such size of the economy, smoking prevalence, population density, cancelling of events, stay at home restrictions, internal movement restrictions, contact tracing, information campaigns, international movement restrictions, workplace closure and school closure. The inclusion of these variables in the model provides a better understanding of the factors that affect COVID19 curve. Second, this study uses a large data set that has data from 190 countries, including developed and developing, a large data set proved more detailed outcomes. Third, this study estimates the Panel Random Coefficient Model (RCM); this model allows country-specific slope coefficients while constituting homogenous residuals. This specification helps in the determination of countryspecific COVID19 curve with random spillover effects. 1 Several other competitive models exist which can estimate this data set like panel data models (fixed and random effect) which does not allow the slope coefficients to vary across cross-sections. Fourth, this study provides various policies based on empirical findings. The rest of the paper is structured, as Section 2 discusses the theory of the COVID19 curve. Section 3 discusses the review of the literature. Section 4 consist of data and methodology. Section 5 discusses the outcomes and discussions. Finally, Section 6 discuss conclusions and policy implications. While discussing the curve, this paper will present the positive and negative forces for both regions (Figure 3 ). For the case of an increase This study proposes that each country have different positive and negative factors of increase and decrease in COVID19 cases. This study developed the Panel RCM model to estimate the country-specific COVID19 curve and country-specific days to reach the flattening stage. Further, this study will explore the effectiveness of each government policy instrument in reducing the number of days to flatten the COVID19 incidence curve. This assessment will be instrumental in the strategizing the smart lockdown approach whereby appropriation combination of restrictions are applied to limit the spread of the virus. A study by (Saez, Tobias, Varga, & Barceló, 2020) theoretically assessed the potency of different measures to flatten the epidemic curve as this situation is crucial for the case of weak economies who are facing a dilemma of saving the economy vs saving the lives. The spread of SARS CoV-2 has not even spared the people in prisons. Last few decades have seen a high number of incarcerations in the USA, adding to 2.2 million people in prisons and jails. Akiyama, Spaulding, and Rich (2020) asserted the importance of flattening the curve within prisons and within the community. Fineberg (2020) discussed that China was able to flatten the COVID19 curve successfully. They did it using clear command and control, aggressive testing, protecting medical personnel and forming cluster-based policy. African countries are looking for a lockdown, but there had been no studies evaluating the cost and benefits of government intervention for these countries (Divala et al., 2020) . Park, Choi, and Ko (2020) discussed the success of advanced tracking in tackling the COVID19 for the case of South Korea. In this strategy, multiple data sources were utilized like mobile services, immigration services, policy, credit card companies, public transit companies, government agencies and lastly health insurance agencies. All of them integrate their data to disclose ware bouts of population publically. Giordano et al. (2020) They are now experiencing falling in COVID19 cases (Cousins, 2020) . India was the first to lock down its international borders after their first case completely. However, it led to the severity of socioeconomic conditions within a 1.3 billion population. Now they are moving towards testing and tracing contacts and isolating patients (Lancet, 2020) . Similarly, the misinformation must also be dealt with at Government level, so that correct information regarding COVID19 can be disseminated among the masses (Donovan, 2020) . Further, since the youth are the major carriers of COVID19, the closure of the schools are the first and foremost strategy to limit the movement of virus (Couzin-Frankel, 2020). Cowling and Aiello (2020) proposed that to mitigate the pandemic, there is a need for temporary close schools, workplaces and discourage gatherings. The more people accept to stay at homes; the more chances are that it will flatten the curve. Poland (2020) stated that the positive COVID19 cases are doubling in every 3-4 days across the world. To handle this, there should be a collective and appropriate suspension of social interaction which provide impetus to SARS-CoV-2 transmission. The comorbidities of the society, like smoking, obesity and immunocompromised people have led to magnified death rates in several countries. The availability of large data sets helps the researchers in terms of making decisions to combat the spread of this pandemic (Callaghan, 2020) . Research in the domain of social sciences can help in managing the perceived risk of individuals and behavioural response to the epidemic until medical treatment or vaccination is developed (Betsch, 2020) . Bastos and Cajueiro (2020) use the data of Brazil to model the incidence of COVID19 and Atkeson (2020) did for the US. In South Asia, several strategies were adapted to counter the spread, which included contact tracing, travel restrictions, stay at home restrictions and awareness sessions (Sharma, Talan, Srivastava, Yadav, & Chopra, 2020 ). (2020) This study has availed the COVID19 data repository in R 3.6.3 developed by (Guidotti & Ardia, 2020) . This data constitutes the indicators of confirmed, recovered and dead cases of COVID19, qualitative indicators of government intervention and quantitative indicators of the economy. This daily data constitutes of 190 countries ranging between, January 22, 2020 to May 11, 2020. Following is the functional form of the model and description of the symbols used in the functional form for the selected variables is given in This study estimates the Panel Random Coefficient Model (RCM Model by, Swamy, 1970) with the natural log of confirmed cases and quadratic function of the number of days since the first case as an independent variable. RCM model will allow country-specific slope coefficients while constituting homogenous residuals. This specification helps in the determination of country-specific COVID19 curve with random spillover effects. 2 Several other competitive models exist which can estimate this data set like Panel data models which do not allow the slope coefficients to vary across cross-sections. Equation (1) is used to estimate the days to flatten COVID19 curve. It is expected that the COVID19 incidence curve will follow the law of diminishing returns, whereby α 2i > 0 and α 3i < 0. While equating first derivative equal to zero, this study will estimate the country-wise value of an ∂lCase it ∂Days it = α 2i −α 3i 2 Ã Days it = 0, This discrete data of the number of days to flatten is then used as a cross-sectional data, and the effectiveness of policy intervention are used as independent variables while controlling for the size of the incidence curve is heterogeneous for all countries. Here it can be seen that on the average, increase in 1 day after the first case, lead to an increase in total cases by 0.198%. This is because of several factors like its spread and increase in testing. However, the effect of days is not linear. Here for every increase in days, the marginal impact of days diminish by 0.001%. This is because of factors like increase in people recovered, an increase in the immunity levels and an increase in awareness and precautionary measures related to COVID19. In Table 4 , the days to flatten are estimated using Equation (2). The estimation was done using 110 valid samples. The significant value of LR Chi 2 shows that the overall model is fit, and the proposed independent variables explain 18.6% variation in the days to flatten. For the case of controlling factors, a 1% increase in the smoking prevalence, leads to a 1.6% increase in days to flatten the curve. This shows that people who smoke first are addicted, they have to visit shops often to buy cigarettes which increases chances of becoming (Heinen, 2020) . Hence out of all policies, the contact tracing had been the most effective strategy. Under this premise, several strategies were proposed by research studies like social distancing, school and work closure, national and international travel restrictions, cancelling of events and gatherings, information disbursement and contact tracing. But the issue at hand is that countries are not unified on the type of strategies to be used. This study firstly developed an indicator which represented the pace of flattening of COVID19 curve. Figure 5 presented the global distribution of days to flatten. This indicator is then empirically analyzed against different policy strategies indicators. The recent developments showed that the straight forward lockdown will not be an optimal strategy especially for the poor economies. It will effect more because of poverty and hunger than of COVID19. Hence there was a need to compare and share the most effective strategy to cater the spread of COVID19. The multipronged analysis using Panel RCM and Poisson Regression led to a final assessment of government strategies. Here, the most effective strategy was, contact tracing, international movement restriction, stay at home restrictions, work closure and information campaigns, arrange in decreasing order of effectiveness. This assessment rests all the debates related to the potency of the counter COVID19 strategies. WHO and local government needs to invest in the rigorous implementation of contact tracing mechanisms. This will help in minimizing the spread of possibly lethal viral infection. https://orcid.org/0000-0002-8940-5391 Muhammad Saeed Meo https://orcid.org/0000-0002-8340-0442 ENDNOTES Flattening the curve for incarcerated populations-Covid-19 in jails and prisons Evaluation of the effectiveness of movement control order to limit the spread of COVID-19 How will country-based mitigation measures influence the course of the COVID-19 epidemic? The Lancet Causal relationship between military expenditure, foreign aid and terrorism: An analysis of muslim countries Applied cross-sectional econometrics Education stock and its implication for income inequality: The case of Asian economies Can government expenditures deter crime? An empirical analysis across the district of Punjab What will be the economic impact of COVID-19 in the US? Rough estimates of disease scenarios (no. w26867) Modeling and forecasting the early evolution of the Covid-19 pandemic in Brazil How behavioural science data helps mitigate the COVID-19 crisis Social network-based distancing strategies to flatten the COVID-19 curve in a post-lockdown world The first 100 days of COVID-19 coronavirus-how efficient did country health systems perform to flatten the curve in the first wave? Differential impacts of contact tracing and lockdowns on outbreak size in COVID-19 model applied to China COVID-19 is a data science issue Emerging lessons from COVID-19 response in New York City New Zealand eliminates COVID-19 Does closing schools slow the spread of coronavirus? Past outbreaks provide clues Public health measures to slow community spread of coronavirus disease 2019 Africa faces difficult choices in responding to COVID-19 Social-media companies must flatten the curve of misinformation Cigarette smoking reduces the efficacy of intralesional vitamin D in the treatment of warts Ten weeks to crush the curve The effect of stay-at-home orders on COVID-19 infections in the United States The invisible pandemic. The Lancet Modelling the COVID-19 epidemic and implementation of population-wide interventions in Italy COVID-19 data hub On interaction of the energy: Human capital Kuznets curve? A case for technology innovation. Environment Editorial: Working from home will help flatten the curve. Opinion-Channel Domestic and foreign incomes and trade balance-a case of south Asian economies The economic impact of COVID-19 COVID-19 lockdowns throughout the world India under COVID-19 lockdown Coronavirus: The first three months as it happened Information technology-based tracing strategy in response to COVID-19 in South Korea-Privacy controversies SARS-CoV-2: A time for clear and immediate action. The Lancet Infectious Diseases Stop covid or save the economy? We can do both Early evidence that social distancing and public health interventions flatten the COVID-19 curve in Italy This interactive chart shows countries flattening their COVID-19 curves Effectiveness of the measures to flatten the epidemic curve of COVID-19. The case of Spain COVID-19 pandemic, oil prices, stock market, geopolitical risk and policy uncertainty nexus in the US economy: Fresh evidence from the wavelet-based approach Policy response to the economic challenge from COVID-19 in India: A qualitative enquiry A qualitative enquiry into strategic and operational responses to Covid-19 challenges in South Asia Economic implications of coronavirus The impact of COVID-19 outbreak on poverty: An estimation for Indonesia. SMERU Working Paper Swamy The impossibility of social distancing among the urban poor: The case of an Indian slum in the times of COVID-19 Coronavirus disease 2019 (COVID-19): Situation report Coronavirus disease 2019 (COVID-19): Situation report Do men and women lockdown differently? An examination of panamas COVID-19 sex-segregated social distancing policy The economic effects of a pandemic Role of energy consumption preferences on human development: A study of SAARC region Arshed is MSc Economics from the University of Edinburgh UK. His expertise are Development Economics and Econometrics Muhammad Saeed Meo is founder of Meo School of Research and currently working as a faculty member at The Superior College Lahore Pakistan Women work Participation, empowerment and poverty alleviation in Pakistan: Empirical evidence from southern Punjab". She also accomplished her Post Doctorate Research of one year from Putra Business School (UPM), Malaysia. She has more than 55 research publications in impact factor & Pakistan HEC recognized journals. Her work focuses on Macroeconomic problems, Globalization and Environmental related issues. How to cite this article: Arshed N, Meo MS, Farooq F. Empirical assessment of government policies and flattening of the COVID19 curve