key: cord-300439-d86p43u1
authors: Bello, Ajide Kazeem; Ridwan, Lanre Ibrahim; Alimi, Yasiru Olorunfemi
title: Estimating the impacts of lockdown on Covid-19 cases in Nigeria
date: 2020-09-06
journal: nan
DOI: 10.1016/j.trip.2020.100217
sha: 
doc_id: 300439
cord_uid: d86p43u1

The study examines the extent to which lockdown measures impact on COVID-19 confirmed cases in Nigeria. Six indicators of lockdown entailing retail and recreation, grocery and pharmacy, parks, transit stations, workplaces, and residential, are considered. The empirical evidence is anchored on the Negative Binomial regression estimator, due to the count nature of the dataset on the daily cases of the virus. The study established the key following findings: First, retail and recreation, grocery and pharmacy, parks, transit stations, and workplaces are statistically significant and negatively signed as relevant predictors of the virus. Second, the impact of residential is positive and statistically significant at the conventional level. Lastly, the results are robust to an alternative estimator of Poisson Regression. The emanated policy message centres on the need to direct efforts toward ensuring total compliance to the lockdown rules as it holds the key to keeping the virus under check.

Against the foregoing backdrop, this study seeks to examine the impact of lockdown on COVID-19 cases in Nigeria. The present study, to the best of knowledge is the first empirical attempt at examining the effects of lockdown on COVID-19 for the country, specifically from an econometric perspective. There is no denying the fact that there have been a series of policy papers, opinions, and predictions about the pandemic for the country. However, most of these submissions can be best described as purely qualitative empirical exercise, building merely on hunches, perceptions or at best, intuitive logics. Thus, this study is quantitative based in nature, thus presenting the contribution to the stock of extant literature on COVID-19 at least from the country-specific stance. While the introductory takes up the first section, the stylized facts about the lockdown measures vis-à-vis the virus spread in Nigeria are presented in section two. Section three presents a brief review of the extant literature;

sections four and five dwell on the empirical strategy and estimation of the lockdown-COVID-19 nexus. Section six concludes with some policy implications and caveats.

This segment presents the stylized facts about coronavirus cases in Nigeria during the lockdown periods. For ease of comprehension, the lockdown period is partitioned into three phases: pre-lockdown, lockdown, and lockdown easing. This division will provide a deep understanding regarding COVID-19 cases during these identified phases.

J o u r n a l P r e -p r o o f

In the pre-lockdown phase, no palpable pattern can be discerned from the trend regarding the number of confirmed cases. The lockdown phase witnesses some dramatic changes in COVID-19 cases as compared to the pre-lock down period. However, in the latter part of the lockdown, there were several sporadic spikes with the highest reported cases being 238 in a single day. The easing period equally experiences some increases, as can be observed from the diagram. On the first day of the easing, 245 cases were recorded, while the subsequent days witness upward trends in the number of confirmed cases. Figure 2 presents the summarized version for the three periods in which the total aggregate of the easing surpasses 13,000 altogether. The same Figure 2 displays the state ranking with respect to the reported cases of COVID-19 in which the Lagos State surpasses other states, thus topping the list, and directly followed by FCT, Kano in that order.

J o u r n a l P r e -p r o o f Note: confirmed= total number of confirmed positive cases of COVID-19 based on clinical tests from the index to the most recent reported cases. Active= the number of COVID19 patients on admission and undergoing treatment at the various isolation centers. Recovered= the number of COVID-19 patients who have tested negatives after treatment and are certified free of the virus by medical personnel. Death=the number of people who lost their lives courtesy of the virus.

The literature on COVID-19 can, at best, be described as emerging or in its embryonic stage.

Thus far, the available studies on COVID-19 have only examined the prevalence and control measures (Ceylan, 2020 , Zhao et al., 2020 , governance, technology, and citizen behavior (Shaw et al., 2020) , socio-economic impacts (Tang et al., 2020) . Other strands had equally focused on respiratory syndrome (Al-Raddadi et al., 2020) , temperature (Briz-Redón & Serrano-Aroca, 2020), mortality rates (Ferdinand & Nasser, 2020; Wang et al., 2020) , and climate factor (Tosepu et al., 2020) , among others.

On the Nigerian front, studies have concentrated on the resurgence of Lassa fever amidst COVID-19 outbreak (Reuben et al., 2020) , Almajiris displacement (Akintunde 2020), comparative analysis of models and estimators (Ayinde et al., 2020) , hunger prevalence (Kalu, 2020) , online forecasting (Abdulmajeed et al., 2020) , impact on transportation (Mogaji 2020) , and economic crisis (Ozili, 2020) .

Relatedly, studies on social distancing and the spread of COVID-19 cases include De Vos, 2020; Friedman et al., 2020; Schueller et al., 2020; Musinguzi et al., 2020; Vinceti et al., 2020; and Zhang et al., 2020 . These studies only focused on countries from the developed and emerging nations, implying that little or nothing is on record concerning the African countries. This is in fact, surprising given the prevalent rate of the virus across the countries J o u r n a l P r e -p r o o f in the continent in general, and Sub-Saharan African economies like Nigeria in particular.

This study fills this lacuna. It is worth mentioning also that the large chunk of the extant studies has only adopted descriptive or at best, discussion methods (Akintunde et al., 2020; Al-Raddadi et al., 2020; Crossley, 2020; Mogaji, 2020; Ozili, 2020; Shaw et al., 2020) , while those embracing econometric approaches are scanty to date (Ayinde et al., 2020; Ceylan, 2020) .

This study employs a negative binomial regression to unravel the impact of lockdown on COVID-19 cases in Nigeria for at least two reasons; first, the dependent variable used is a count data that only covers discrete and nonnegative values. Thus, as a skewed discrete distribution, using ordinary least squares (OLS) estimates can only yield inefficient, inconsistent, and biased (Long, 1997) . Second, if this dependent variable fits equi-dispersion, then the Poisson regression model becomes inevitable. If otherwise, using negative binomial model remains a credible option. This estimator is often used when the variance is larger than its mean (over-dispersion). The robust standard errors are further clustered in order to produce standard errors that are robust to both heteroskedasticity and a general-type of serial correlation within the cross-sectional unit. More importantly, this estimator has found extensive application in studies such as accidents, conflicts, terrorism among others, given the count data nature of their data (Gassebner and Luechinger 2011; Krieger and Meierrieks 2011, Ajide and Raheem, 2020; Ajide, Adenuga and Raheem, 2020).

In a more general form, to specify a negative binomial regression, the mean of the outcome variable y is determined by the exposure time t and a set of k explanatory variables (the ' xs). Hence, the empirical expression relating to these quantities is specified as: J o u r n a l P r e -p r o o f

The regression coefficients are thus estimated using the method of maximum likelihood. (see, NCSS, 2017).

The study employs an all-inclusive daily situation report of COVID The data on lockdown is obtained from Google mobility data. 

The descriptive statistics are presented in Table 1 . From the Table, At the level of the states, Lagos has the highest recorded COVID cases (see Table 2 ), which is consistent with Figure 2 above. Table 4 presents the results of the negative binomial regression estimations of the lockdown effects on Covid-19 for Nigeria. The results for all the indicators of lockdown variables are statistically significant and negative except for the residential variable. These results consistent with the theoretical priors, suggesting the mitigating role of lockdown policies on coronavirus spread. By implication, as people comply with the "stay-at-home" order, and limit their visits to essential places, thus reduce their chances of being infected by COVID-19. Correspondingly, this also tends to reduce human-to-human contact, which is the main transmission channel of COVID-19. Intuitively, a 1% increase in compliance to the stay-athome order leads to a corresponding reduction by the magnitudes 0.026%, 0.019%, 0.035%, 0.020% and 0.020%. On the contrary, the impact of residential is positive and statistically relevant. This sounds plausible as people desert essential places of visits, they tend to increase their presence at home. In particular, the majority of infected persons usually have one or more of their family members or close relatives infected. This explains why residential remains a key predicting channel to contacting COVID-19 and such reasons can be advanced as why COVID increases during the lockdown. J o u r n a l P r e -p r o o f 

This study examines the extent to which lockdown measures impact on COVID-19 confirmed cases in Nigeria. Using the Negative Binomial regression estimator on the daily situation data, the following results are established. First, retail and recreation, grocery and pharmacy, parks, transit stations, and workplaces are negative and statistically significant across the models. Second, the impact of residential is positive and statistically relevant, thus running contrary to other lockdown measures with negative theoretical priors. Lastly, the obtained results are robust to an alternative estimator of Poisson Regression.

The study has some relevant policy implications. First, since the importance of lockdown policy has been quantitatively confirmed to be effective in combating the spread of COVID-19 cases, focus should be placed on residential houses, which act as a spur to the virus. This can be effectively achieved through public enlightening programs and general awareness on the need to comply with lockdown measures. More importantly, the government should guarantee and ensure constant supply of electricity to the people, their staying at home is largely predicated on enjoying uninterrupted supply of electricity. This sounds plausible in the Nigerian context where the supply of electricity has been erratic most times. This has often remained one of the reasons why people seek pleasure outside of their homes. This mostly takes the form of visits to relaxation centres like parks, recreation centres, restaurants, etc. Second, if the government has to ease the lockdown, at all, it must be gradual with all the necessary precautions dully enforced. Notwithstanding, this must be supported by sanctions to the defaulters. Third, lack of necessary welfare-oriented supports from the government serves as a reason people often advanced for not making them ''compliant agent(s)'' during the lockdown periods. Going forward, future research can be conducted using the state-level unit of observations for analysis in order to arrive at a more robust policy generalization.

What is more, since Nigeria shares similar socioeconomic and political characteristics with other African countries, the outcome of this research work could serve as useful research inputs for other countries in the region to extrapolate.

All the authors participated in the making of the manuscript. Specifically, Ajide Kazeem Bello conceived and designed the manuscript. Ridwan, Lanre Ibrahim provided empirical studies, data, and proofread the manuscript. Alimi, Olorunfemi Yasiru, handled the model estimation of and proofread the manuscript.

ONLINE FORECASTING OF COVID-19 CASES IN NIGERIA USING LIMITED DATA

Does Democracy Really Fuel Terrorism in Africa?

Natural resource rents, political regimes, and terrorism in Africa

Public health implication of displacement of Almajiri children in specific states of Northern Nigeria amidst COVID-19 pandemic

Burden of Middle East respiratory syndrome coronavirus infection in Saudi Arabia

Modeling Nigerian Covid-19 cases: A comparative analysis of models and estimators

A spatio-temporal analysis for exploring the effect of temperature on COVID-19 early evolution in Spain

Indirect virus transmission in cluster of COVID-19 cases

Wuhan lockdown unprecedented, shows commitment to contain virus: WHO representative in China

Estimation of COVID-19 prevalence in Italy, Spain, and France. Science of The Total Environment

COVID-19 in Nigeria: a disease of hunger. The Lancet

Impact of COVID-19 on transportation in Lagos, Nigeria. Transportation Research Interdisciplinary Perspectives

Nigeria Centre for Disease Control (NCDC, 2020)

Covid-19 pandemic and economic crisis: The Nigerian experience and structural causes

Is the COVID-19 pandemic masking the deadlier Lassa fever epidemic in Nigeria?

An updated estimation of the risk of transmission of the novel coronavirus (2019-nCov). Infectious disease modeling

Correlation between weather and Covid-19 pandemic in Jakarta

Real-time estimation and prediction of mortality caused by COVID-19 with patient information based algorithm

Preliminary estimation of the basic reproduction number of novel coronavirus (2019-nCoV) in China, from 2019 to 2020: A data-driven analysis in the early phase of the outbreak

J o u r n a l P r e -p r o o f

The authors have no conflict of interest and no fund was received in favor of this study.

 (-1) 0.013*** 0.013*** 0.010*** 0.014*** 0.014*** 0.013*** 0.013*** 0.010*** 0.012*** 0.013*** 0.010*** 0.014*** 0.009*** 0.012*** 0.009*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) % ∆ of Retail and recreation -0.014*** 0.000 0.011*** 0.032*** 0.042*** J o u r n a l P r e -p r o o f