Yothin Jinjarak, Rashad Ahmed, Sameer Nair-Desai, Weining Xin, Joshua Aizenman 20 May 2020
Key factors in modelling a pandemic and guiding policy-making include infection rates; mortality rates associated with infections; the ability and effectiveness of policies, medical systems, and societies to adapt to the changing dynamics of a pandemic; and other structural factors (Verity et al. 2020). Understanding these factors is needed to remobilise workers without risking a medical overload (Baldwin 2020). In Jinjarak et al. (2020), we take stock of the data gathered during the first three months of the COVID-19 pandemic, tracing the associations between COVID-19 mortality and pandemic policy interventions while accounting for global pandemic diffusion patterns. Pandemic policy interventions in our consideration refer to containment and closure policies that aim to limit social contact. Our empirical specification controls for these considerations, subject to the limited data available on key factors. Specifically, the scarcity of COVID-19 testing and the limited information on the precision of available tests, implies a vast underestimation of the infection rates per capita, possibly by a factor of two digits.1 The undercount of COVID-19 population mortality rates is also prevalent but by an order of magnitude below the errors associated with infection rates.2 Therefore, we focus mostly on accounting for the COVID-19 population mortality rates per capita during the first phase of the pandemic, controlling for policy and structural factors subject to data availability and quality. We plan to revisit these issues with better quality and longer-term data in the coming quarters.
A fair share of countries reached a local peak of COVID-19 population daily new mortality rate curve during the sample period (see Figure 1). Applying various techniques, we study the factors accounting for the empirical shape of the mortality curve from the onset of the pandemic to the local peak, with a focus on the impact of policy intensity interacting with structural variables. Like most similar studies, the results should be regarded with healthy scepticism. First, data quality and availability are a major limitation, as each country has its challenges with data collection, aggregation, and reporting. Second, ‘better performance’ in the first mitigation phase of a pandemic does not guarantee superior future performance, as the dynamics of a new viral pandemic are yet unknown. By design, flattening the pandemic curve shifts some mortality incidence forward.3
Figure 1 Sample countries and new mortality curves, 23 January 2020 – 28 April 2020
Note: Seven-day rolling average new mortality rate by country. Y-axis normalised to have all countries fit the same scale. Period: 23 January – 28 April 2020. Special case countries omitted from the above plots: China, Singapore, and Vietnam.
Our study relies on daily COVID-19 policy and case data reported by Oxford and John Hopkins University, as well as Apple mobility data and various controls. Our baseline estimation study examines OECD and Emerging Market (EM) sub-samples based on data from 23 January 2020 to 28 April 2020, or the first 97 days of the pandemic.
First, we investigate the evolution of weekly mortality growth rates over time and across countries. Applying local projections (Jorda 2005), the panel evidence suggests that administering more stringent pandemic policies were associated with significantly lower future mortality growth rates during the first pandemic phase.4
Taking slow-moving country fundamentals from the period pre-COVID-19 as exogeneous, we find that countries in the 75th percentile in the proportion of the elderly (people 65 or older) saw a much stronger reduction in mortality growth rates from the same ten unit rise in policy stringency index, SI, compared to countries with relatively low proportions of the elderly (25th percentile).5 In countries further away from the equator, SI measures had a stronger impact on mortality growth than countries closer to the equator. This heterogeneity may be consistent with what some describe as a temperature risk factor associated with many flu viruses (see Figure 2). Countries with higher proportions of the elderly or cooler temperatures over the January-April period may be at higher risk with regard to COVID-19, increasing the effectiveness of stringency measures for these countries. Greater policy stringency is also more strongly associated with lower mortality growth during the first phase of the pandemic in countries with greater population density; a greater proportion of employees in vulnerable occupations; and greater democratic freedom (measured with the EIU Democracy Index), but the economic significance is not as stark. While population density and employees in vulnerable occupations are intuitive risk factors for a pandemic like COVID-19, the role of democratic freedom is an ongoing topic of debate. Our results are consistent with the view that greater individual rights may be detrimental in this situation, making it more difficult for the government to place strict quarantines in place and have citizens abide by them.
Figure 2 Mortality impacts: government response, demographics, geography, and development level
Note: Red squares (blue circles) represent the local projection impact from a ten-unit higher stringency index on mortality growth for countries in the 75th percentile (25th percentile) of the country characteristic.
Next, we turn to cross-country regression results. Dependent variables include the logged peak mortality rate (calculated as the cumulative deaths out of the population at the peak of daily new mortality, by country); the logged peak new mortality rate (calculated as the new deaths out of the population at the peak of daily new mortality, by country); and the ratio of ‘peak new mortality rate’-to-‘pandemic duration to first peak’ measured in days (a proxy flatness/steepness of the mortality rate curve) (see Figure 3). Countries with higher early mortality (cumulative mortality within the first week following the first death) tend to have higher new mortality peaks but flatter mortality curves in the first pandemic phase. Also, countries with more aggressive policy interventions in place before the first death (Early SI) tend to exhibit lower cumulative and new mortality at the peak, and flatter mortality curves.6 Countries with greater elderly populations tend to have higher peak mortality rates. We also find some evidence suggesting that countries with higher mortality growth rates at the outbreak also had higher peak mortality rates. Overall, the evidence suggests (but does not necessarily assert) that policy stringency directly reduced peak mortality rates and flattened the mortality curve, and that other forces were also at play (e.g. demographics, initial pandemic conditions).
Figure 3 Characteristics of the quasi-bell mortality curve
Not only do mortality rates during the first pandemic phase differ across countries, but there is also considerable variation in how long new deaths continued to climb (measured in days). We term this the ‘pandemic duration to the first peak’ (PD). One should be careful when interpreting the effects of covariates on the PD in terms of altering the shape of mortality curves, as a longer PD could be accompanied by a higher peak mortality rate and thus a steeper curve, or a lower peak mortality rate and thus a flatter curve. Fitting a Kaplan-Meier curve for the PD over all countries in the sample, by number of days, suggests that countries with stricter policy interventions early on (Early SI > 19) had significantly lower PDs on the way to the first local peak of the daily mortality curve; reached the peak, with a probability of 75%, within around 40 days compared to countries without such interventions (Early SI < 19); and took ten more days to reach the same peaking probability (see Figure 4).
Figure 4 Time-to-peak duration analysis of mortality.
Note: Y-axis indicates the probability that the peak mortality/case is ‘yet to come’. The higher y-axis implies a lower probability of peaking. X-axis reflects the number of days since the first mortality was realised. Shaded areas represent 95% confidence intervals.
To better understand the cross-country variation in PD under a Cox proportional hazards model, we report that across most (but not all) specifications, stricter policy interventions early on are associated with shorter durations of the PD. Higher mortality rates early on are associated with shorter pandemic durations to the peak, while countries realising higher mortality peaks tend to have longer pandemic durations to the peak. Additionally, countries with greater elderly populations, higher population density, and greater shares of vulnerable employment tend to exhibit shorter pandemic durations to the peak. Moreover, under certain specifications, the level of democratic freedom appears to be a highly significant determinant in pandemic duration to the peak. As such, countries that are considered more ‘democratically free’ saw longer pandemic durations to the peak. Countries further away from the equator also tended to experience longer PD. While at this stage we are reporting suggestive statistical associations, more data and research are needed to provide fuller identifications of all these factors.
We conclude by cautioning that our results are subject to data limitations, including undercounts of COVID-19 infections and mortality. ‘Better or worse performance’ of a country in the first phase of the pandemic does not guarantee similar future outcomes. Flattening the mortality and infection curves may shift mortality and painful adjustment forwards. Premature opening of the economy without proper testing, contact-tracing, and selective quarantines of vulnerable or impacted segments of the population may induce future acceleration of the pandemic (Acemoglu et al. 2020). More medical research and advances toward better treatment and possible vaccinations, the quality of local and global public policies, and adjustment capabilities of countries will determine future dynamics of the pandemic (Lipsitch et al. 2020).
Acemoglu D, V Chernozhukov, I Werning and M D Whinston (2020), “A Multi-Risk SIR Model with Optimally Targeted Lockdown”, NBER Working Paper No. 27102.
Lipsitch, M, D L Swerdlow and L Finelli (2020), “Defining the epidemiology of COVID-19—studies needed”, New England Journal of Medicine 382.13: 1194-1196.
Jinjarak Y, R Ahmed, S Nair-Desai, W Xin, and J Aizenman (2020), “Accounting for Global COVID-19 Diffusion Patterns, January-April”, NBER Working Paper No. 27185.
Jordà, Ò (2005), “Estimation and inference of impulse responses by local projections”, American Economic Review, 95.1: 161-182.
Verity R, LC Okell, I Dorigatti et al. (2020), “Estimates of the severity of coronavirus disease 2019: a model-based analysis”, Lancet Infectious Diseases, 30 March. https://doi.org/10.1016/S1473-3099(20)30243-7
1AAAS Science of 21 April 2020 reports a vast undercount of COVID-19 infection rates. A Stanford University study by Bhattacharya and Bendavid estimated that for each positive COVID-19 test result in Santa Clara County, California, there are more than 50 times more infected people. Similar results were found in Los Angeles county, and in several studies in Europe. While the debate about the methodologies and the veracity of these studies is ongoing, the results probably reflect the strong testing selectivity: testing targeted mostly sick patients, at more advanced stages of possible infection than is medically optimal, thereby missing large population shares of patients with mild or asymptomatic COVID-19 symptoms.
2 A Financial Times study on 26 April 2020 reported that mortality statistics show 122,000 deaths in excess of normal levels across 14 countries, concluding that the global coronavirus death toll could be 60% higher than reported. This undercount reflects the scarcity of COVID-19 tests, underreported deaths at senior homes and assisted living centres, misdiagnoses, limited administrative capacities, and the like.
3 The susceptibility to secondary waves of infection remains a looming threat. Policies adopted in the second quarter of 2020, and the realised pandemic infections, containment, and treatment will explain the future performance of each country. Furthermore, only time and more medical research will show the degree to which infected persons that recovered gained immunity for a long enough period to allow smooth convergence to ‘herd immunity.’
4 Countries with a Stringency Index (SI) ten units higher than average had, two weeks later, mortality growth rates that were on average -3% lower. The reduction in mortality growth increases to roughly -6% by the third week and then stabilises. While the reduction in growth rates seems quite large, it is important to put these numbers in perspective. Given the exponential nature of disease spread, often times weekly mortality growth rates can be anywhere from +50% to +100% or greater.
5 Oxford’s SI is normalised between zero and 100, where 100 = strictest response. Countries in the 75th (25th) percentile saw mortality growth rates about -9% (-3.5%) after two weeks. Countries with a greater proportion of the elderly are unconditionally more susceptible to the pandemic, but for this same reason, are likely to benefit more under stringent policies.
6 We find that a one unit increase in Early SI is associated with peak cumulative and new mortality rates of -3% and -2.5% lower, on average.