How much longer MCO should last? Part I

Analysis of determination of MCO under Covid pandemics

Dr. Wan M Hasni, Chief Data Scientist (MD Labs S/B, Malaysia)https://ibugroup.org
04-13-2020

Table of Contents


Pre-amble

Now we know that the Malaysian MCO will last until the 28th of April, and we have to wait at least for two more weeks before we know whether it will be ended then, or it will be extended further by (presumably) another two weeks. The process could repeat itself for some distant time in the future, or it could be ended as rapid as it started. The questions that linger in everyone’s mind are: How long will this MCO need to last? Will a further two-week extension (as announced recently) would suffice, or what we have is a two week by two weeks “game” of extension? What are the criteria of ending the MCO (after each period of extension), and how would these exercises impact the public’s response towards it, and how the response would be when they are ended?

The question at hand is not an easy one because on one hand, there is a high asymmetricity of the choices, and on the other hand, the consequences of any outcomes, for any given choice, are highly extreme (fat-tails) in terms of its potential impacts. Furthermore, with the added dynamics of the time element and other elements (such as global scenarios, Covid virus behaviors, availability of vaccine, etc.) would further aggravate the complexities of decision making (by the government and the public).

Let us observe a few pointers for the sake of understanding the problems. We have China, which is in many ways an autocratic country, hence decisions for the society is made centrally, at least for now seen to be among the countries with success to combat Covid’s pandemic; and on the other hand we have the United States, which (at the time of writing) has the largest number of case incidence, surpassing any other countries - where decisions are made in a fully democratic environment and process (of the federal government and local authorities). In between, we have many European countries (western-style democracies), regimes of monarchy in the middle east, and Southeast Asian democracies - all with varying conditions and situations (as of this writing).

OECD provides estimates that a month of lockdown will generate a loss of 2% of a country’s Gross Domestic Product (GDP)1, and if it continues more than a month, the mounting costs may be non-linear (i.e. maybe exponential). Expenditures to contain the disease, in terms of direct benefits to be provided to the public is high, for example in the United States it could be as high as USD2.4 trillion2, and in the case of countries like Malaysia, this direct cost could be anywhere between RM10 billion to RM20 billion per month, and could be spiraling as the period is extended much longer.3 In terms GDP, the losses for Malaysia for a month could be at RM25 billion and counting (as the MCO is extended).

At stake as well is the credibility of the government and government officials - since any wrong decisions made will be judged harshly by the population and the reputation of the country in the international community also hang in the balance. While the crisis is still on-going, lots of decisions may be tolerated by the public in the short-term, however in the medium and longer-term, post-crisis (which nobody can predict when), many of these (ex-post past) decisions will come under heavy scrutiny by all parties when more information and data becomes available.

Quoting Andrew Sheng on dealing with these questions: “We can only deploy “whole-of-government” and “whole-of-society” to fight the COVID war if and only if we are able to knock common sense into obstinate and self-important bureaucratic silos. The parts must work together to save the whole. The last but not least tough choice is how to re-open the economy, once the pandemic is under some semblance of control. The tough reality is that there is no formula, model or theory to guide us, given that the moral values, ideologies, information, resources and institutions are so different for different countries"4, reflect one clear thing: we can’t simply rely on government and government alone to act, we as the public need to perform our part as well if we want to increase the chance of success in dealing with the problems at hand.

Covid is a human crisis which requires we human to take hard thinking on many “old norms” and should force us to establish “new norms” as a human being living on this planet and at every nation as a society. The test might break us as a human beings and as a society living on brinks of disaster. As a human society, we would be required to make hard choices, many of which may require us to alter our norms and behaviors. To enable us to make such a decision, we need to use knowledge and science as our guide.

Under these backdrops, what we will do in this article is an attempt at modeling this highly complex decision-making process, using tools from social decision science. Our approach to doing this would be through various simulation exercises. We hope to be able to learn something from the simulations which may be used for guidance, at least from the public perspectives (assuming that we have no say on the government decision making process).

Social choice decision science

The subject of social decision science could be attributed back to Kenneth J. Arrow’s seminal work under Social Choice and Individual Values5, further expanded by the utility and probability setting based on the works of Von-Neumann and Morgenstern6, and established under equilibrium setting of game theory of Nash7. In short, the subject’s theoretical basis is the combination of Behavioral Game Theory and Social Choice decision making. These works are further contributed by Rawls8 and Sen9 into the areas of social justice.

A comprehensive survey of game-theoretic modeling of pandemics is covered by Chang, et. al.10, which emphasize that:“The scale and diversity of interaction patterns call for innovative modeling methodologies that can successfully and coherently accommodate three key elements: infectious disease dynamics, interaction patterns, and the decision-making of individuals” (and governments). The larger aspect of a game-theoretic social decision on pandemics will take even a larger center stage when economic factors are included in the process - that is when society at large is threatened economically with the lives and livelihood. The trade-offs between the fending of epidemics and economic activities could heighten tensions on international trades and trade wars, as countries scramble to get itself out of economic devastations of Covid’s pandemic. Using game-theoretic methods we also could perform measures of the systemic preparations (or lack of it) - which is an important learning process to prepare to tackle issues at hand as well lessons for the future11. The choices and decisions that we are facing are complex and multiplicative.

In any case, given the importance of the subject matter, we will attempt to use some of the commonly used models of game-theoretic and social choice theory, and apply them to the Malaysian settings with the hope that we could learn something which could be obvious from common logic, and yet obscured from methodical and concise argumentation point of views.

Modeling Covid epidemic structure

As time progressed from early days of Covid pandemic, lots of progress in terms of knowledge or learning that we have obtained - from virology and epidemiological perspectives (covid virus and spreading behavior, symptoms and mortality, and others), from public health perspectives (healthcare systems and intervention measures), and socio-economic perspectives (the social and economic costs of the pandemic). While there are many unknowns still exist, we have to caution that forecasting “too far ahead” is extremely hard and futile until more factors are known and confirmed based strong scientific evidence - which will take more time before any strong conclusions could be ascertained and relied upon.

In a recent report published by the Imperial College Covid-19 response team provides a good summary of the state of affairs for the European countries as of 30th March 202012, which suggest a semi-mechanistic Bayesian hierarchical model to measure effectiveness of public health intervention policies across 11 European countries. This is to back-calculate infections from observations.13

What they have documented is rather encouraging, that is based on various levels of measures (from closures of schools and ban on public events, self-isolation, social distancing, and lockdowns), at various stages, all the 11 countries studied show that by 28th of March, the Covid disease reproduction number \(R_o\) had been drastically reduced to credible interval near 1 (i.e. flatten the growth curve). The most important and decisive measure is a lockdown. What’s surprising is given China’s Wuhan experience of about 60 days (two months) lockdown is giving almost the same trending as the current European experience.14

Despite all the encouraging signs, it is clear that many countries would continue the lockdown into much longer period: Italy is planning to extend it to early May, Spain extended it until extended until 25 April, the UK has not announced its extension length as today, Germany has stated that the next review date will be on 19th April. And for Malaysia, the MCO is now extended until 28th April, subjected to another review to be done before that date.

There are few things worth noting though, for all these countries (China’s Wuhan and the 11 European countries), the lockdown measures are not the ONLY measures taken - namely, it is combined with aggressive testing and contact tracing.

Testing of Covid in the populations

From the plot above, we can say that there is much more to be desired in regards to testing for Malaysia (in terms of testing per-capita). For lockdown and post lokdown to be effective, it is agreed by most that testing and tracing the disease is critically important. While the data on testing is becoming more widely available (publicly), most countries (and Malaysia included) are still far from contact-tracing activities.

As the pressure is building to relax the lockdown measures for people to get back to work, contact-tracing is a key part of this because it can help authorities contain a potential resurgence of the virus as people resume regular activities. The good news is Apple and Google are both making some progress on this to make publicly available contact-tracing solutions for the mobile phones operating system. The solutions will be available by mid-May of this year.15

Flattening of Covid incidence

Next we run the model developed by the Imperial College Covid response team, using the R codes provided16, with the addition of Malaysian data and perform comparisons between Malaysia and some of the European country reported by the study. What’s interesting is the question of whether thus far, all the measures taken, in particular the MCO, had taken any effects in reducing the disease reproduction rates?

Let us revisit the reproduction rates curve for Malaysia (updated with the data upto 7th April 2020).

Let’s compare them with Italy, which is among the hardest hit country in Europe.

And Norway, which had the most aggressive testing program done to date.

And finally compare with Australia.

What we can observe is that while all countries cases incidence are flatenning, but the reality is also quite stark - that is a slow process of flatening. Another matters of concern though is the possibility of resurgence after the lockdown period. Based on similar approach of modeling, Chang et. al.17 estimates that (given there is no vaccination available in the shortest time), post lockdown and severe social distancing, in the case of Australia, the disease has a chance to peak again even after 100 days after the initial lockdown date, albiet a much lower numbers.

Conjectures for the Malaysian MCO

Based on all these studies covered thus far, what would be our guess for the lockdown period required for Malaysia? And what would be the scenario post-lockdown be?

One almost certain thing, the period of extension beyond the current date of 28th April is highly likely to be extended well beyond. As we could see that despite the reduction of the reproduction rates for many countries, and Malaysia is included, the presence of persistent incidence is highly likely. Wuhan did a lockdown for almost 60 days.

Comparison with Hubei (Wuhan) China:

If we take Hubei and the European countries mentioned before as benchmarks, and given that these countries (or places) are aggressive in their testing (and tracing) approach - and yet the results show that the lockdown or severe social distancing required is for a prolonged period - it would only be logical and probable that Malaysia’s case will not be too far different or is not any better.

Therefore, given the highly probable case of the MCO to be exercised over a much longer period, it would be pertinent for us to study the various scenarios of the society’s reactions using behavioral game-theoretic models of social choice as we have mooted earlier. This is what we will perform in the next article.


  1. https://www.oecd.org/newsroom/oecd-updates-g20-summit-on-outlook-for-global-economy.html

  2. https://www.mercatus.org/publications/covid-19-policy-brief-series/cost-covid-19-rough-estimate-2020-us-gdp-impact

  3. The number is deduced from a simple ratio of Malaysian GDP, of RM1.3 billion divided by 12 months multiplied by 10%-20% (10% x 1,300 billion/12)

  4. https://www.thejakartapost.com/academia/2020/04/12/four-hard-choices-in-the-covid-war.html

  5. Arrow, Kenneth J., Social Choice and Individual Values, John Wiley & Sons, 1951

  6. von Neumann, John F. and Oskar Morgenstern, Theory of Games and Economics Behavior, Princeton University Press, 1944

  7. Nash equilibrium, based on John F. Nash’s works on equilibrium in non-cooperative games

  8. Rawls, John.,A Theory of Justice, Oxford: Clarendon Press, 1971

  9. Sen, Amartya, Social Choice Theory: A Re-examination, Econometrica, Vol. 45, No. 1 (January 1977)

  10. Sheryl L. Chang, Mahendra Piraveenana, Philippa Pattisonc and Mikhail Prokopenkoa, Game theoretic modelling of infectious disease dynamics and intervention methods: a review, Journal of Biological Dynamics, 2020, Vol. 14, No. 1, 57–89 https://doi.org/10.1080/17513758.2020.1720322

  11. A good article on this is by Bartosz Maćkowiak, Mirko Wiederholt 19 March 2020 https://voxeu.org/article/lack-preparation-rare-events-and-policy-implications-time-covid-19

  12. Seth Flaxman, Swapnil Mishra, Axel Gandy et al., Estimating the number of infections and the impact of non- pharmaceutical interventions on COVID-19 in 11 European countries. Imperial College London (30-03-2020) doi: https://doi.org/10.25561/77731.

  13. The codes and data used is available at: https://github.com/ImperialCollegeLondon/covid19model/releases/tag/v1.0

  14. On casual observation of the parameters, it seems that the European average is following the Wuhan’s experience.

  15. The Silicon Valley rivals said on Friday that they are building the technology into their iOS and Android operating systems in two steps. In mid-May, the companies will add the ability for iPhones and Android phones to wirelessly exchange anonymous information via apps run by public health authorities. The companies will also release frameworks for public health apps to manage the functionality. Source: https://www.bloomberg.com/news/articles/2020-04-10/apple-google-bring-covid-19-contact-tracing-to-3-billion-people

  16. See note 13

  17. Sheryl L. Chang, Nathan Harding, Cameron Zachreson, Oliver M. Cliff, and Mikhail Prokopenko, Modelling transmission and control of the COVID-19 pandemic in Australia, 2 April 2020. https://arxiv.org/abs/2003.10218v2

Reuse

Text and figures are licensed under Creative Commons Attribution CC BY-SA 4.0. The figures that have been reused from other sources don't fall under this license and can be recognized by a note in their caption: "Figure from ...".