Pandemic Strategy: Health, Economic and Social Sustainability

Making Policy Decisions with Uncertain & Unreliable Data

We are at war with a mysterious, unseen, global and deadly pathogen with exponential human-to-human transmission rates. We have very little reliable data about the pathogen. This may, or may not be a serious epidemic that infects and kills a significant number of people. We are flying blind and need a prudent plan of action to kill or disrupt the pathogen as soon as possible.

Imagine you are the chief data scientist for your nation/state's president/prime-minister/governor or top leader. A nasty pathogen hits the world exponentially fast and furious and your boss needs reliable data and real-world (not academic) interpretation of that data to make optimal policy decisions for citizens. 

Human lives are at stake. Economic stability is at stake. Social well-being and cohesion is at stake. This is what you were trained for. This is your moment.

First order of business is to quickly build a team of subject matter experts (epidemiologists and other healthcare professionals, statisticians, lawyers, economists and financial market specialists), data scientists, data engineers and communication/visualization specialists. 

As chief data scientist and team leader your primary job is to synthesize and interpret data, and counsel policy makers / leaders to understand objective reality, conduct scenario planning, and provide policy recommendations. In a large data project with human lives at stake, it is usually prudent to form multiple teams of data scientists to interpret data independently to avoid groupthink and confirmation biases, and provide a wide diversity of viewpoints including out-of-the-box thinking. Your mind is open and calm - a data crunching high performance computer running top speed on all cylinders.

Second order of business is to find reliable data sources and narrowly target smart relevant data at a high level of veracity and quality. This is a huge data engineering job collecting, analyzing and storing massive amounts of data requiring multiple information flows to the right people at the right time.

Third order of business is to interpret the data using scientific techniques with high standards and present reasonable real-world scenarios to decision maker(s) to help them understand fluid (not static) reality and make optimal policy decisions. You apply probability theory to weigh upside and downside risks using high performance compute power crunching massive amounts of data.

In this case, you find uncertain and unreliable data. You assign specific tasks and goals to your data science teams to find, collect and classify both past and real-time data. Awesome BS detectors required when dealing with governments and organizations who do not share enlightened values and human rights.

Part of the data science teams should include machine learning algorithm (AI) design professionals. Massive amounts of data collected require machine learning techniques to find patterns and correlations for data scientists to examine and conduct targeted deep dives. Super compute power and time resources should be reallocated to help crunch the data within reasonable time frames.

First goal is to understand whether global distribution of the pathogen looks like an iceberg that may be one-tenth above water, or a structure where we can see the whole picture. Note that if we have uncertain and unreliable data we can only see a small part of the pathogen and thus flying blind. Not a good feeling.

Here, we notice strong evidence of exponential human-to-human transmission rate and thus may be in serious trouble in near future. Danger.

With uncertain and unreliable data our professional duty as data scientists is to help policy makers understand and AVOID MASSIVE DOWNSIDE RISKS. Building models with bad data and unreasonable assumptions that purport to show worst and best case scenarios is NOT helpful.

Therefore, you recommend a massive public collective social distancing / lockdown strategy to slow down exponential transmission rate of pathogen (flatten curve) and ramp up health care services. 

The FIRST PRIORITY GOAL is to cut the (unseen) pathogen tail risk off quickly to radically slow human-to-human transmission and create an environment for the healthcare system to prepare and execute services without being overwhelmed.

Time is of the essence: early effective action saves lives and significantly reduces economic and social costs. Adequate epidemic war preparation is like a low cost insurance policy. Once the epidemic hits, insurance rates skyrocket and you will likely be in a world of pain. 

Speed and competent execution is critical. Keeping calm yet with a sense of urgency is the trick. Promote confidence and esprit de corps within your teams. Gently coach them up when needed and always use positive reinforcement. Firm yet calm competence promotes confidence in leadership and mission.

Policy Recommendations

Phase I: Pathogen Suppression and Mitigation Strategy

Research strategies of other nations and learn what works and does not work. Taiwan, a stones throw from China, provides great evidence of optimal prevention and mitigation strategy against this pathogen. See: Response to COVID-19 in Taiwan - https://jamanetwork.com/journals/jama/fullarticle/2762689. Singapore and South Korea also offer success strategies. Track multiple strategies in real-time to perhaps learn what worked yesterday may no longer work today. 

This is a fast changing dynamic, not static, environment. Change and adapt continually. You are now an adrenaline junkie like a high end powder skier on a steep slope. Welcome to my world! You remain calm yet have just the right amount of anxiety for high performance. Embrace the pain to achieve massive gain.

While buying time flattening the curve, find and collect data to help ramp up healthcare services. Urgent production of ventilators, drugs, masks, protection clothing for healthcare workers and any other tools to help the infected. Managing healthcare delivery services to improve speed and lower death rates is critical and thus flattening the exponential transmission rate curve is an URGENT GOAL.

Expedited development and distribution of a vaccine or antiviral is mission critical. Design appropriate incentives to spark innovation in private sector.

Phase II: Decentralized / Centralized Data Collection and Real-time Data Analysis

Expedite design, development and execution of data collection with both centralized and decentralized information flows - to get timely information in the right hands in multiple places. 

Design and implement timely distribution of pathogen testing and collecting (reliable) data for near real-time analysis by the RIGHT SUBJECT MATTER PROFESSIONALS. Understand false positive and false negative margins of error and create processes to continually improve tests. Set high standards for engineering data veracity. Yet considering urgency don’t let the perfect be the enemy of the very good. Tricky balancing act.

Design information flow and execute knowledge processes. Understand what information policy makers need to make optimal decisions at both tactical and strategic levels, and at different time scale phases. Form competing data science teams with diverse experts to interpret data, create scenario plans and make policy suggestions.

Design and implement mobile phone data tracking systems for those infected to enforce legal and practical apprehension, and quarantine procedures. Develop contact tracing processes to isolate the contacts of infected from the public in a timely and humane manner. Work with lawyers to make legal and have solid data to prove in court with sceptical judges.

Prioritize identifying and isolating “super-spreaders” who possess high infection transmission risk IMMEDIATELY. An example is college age kids ignoring social distancing and going to the beach on spring break partying in large groups. Design effective, legal and humane apprehension and quarantine processes. Collect data from and study these people individually and groups of people collectively. 

If data shows large numbers of people infected and recovered -  test them repeatedly to determine whether they are still infectious and should be isolated or released to open society. Form competing teams to study human recovery from pathogen at both individual and group scale levels.

Create a system that identifies all positive tests (wristband or a ID card). Design and implement data collection and distribution systems to study human reaction to pathogen and share in near real-time with appropriate healthcare providers and researchers.

Design and execute learning algorithms to help find correlations and patterns in massive data sets. Reallocate time on research university and government sponsored supercomputers to crunch data and form teams of specialists to study and make recommendations. Time is of the essence so collaboration and coordination is important. Right smart data to right people in real-time.

Phase III: Legal Issues to Balance Public Safety vs. Data Privacy / Human Rights

Form a team of seasoned public, private and civil libertarian lawyers and judges to design and implement a judicial procedure to monitor mobile phone and personal health data collection to create humane processes and prevent / detect personal data privacy abuses.

Train new judges for a specialized court to review cases with the goal of balancing safety needs of collective society with personal human rights. The public is more likely to accept a fair system with clear rules adjudicated by an independent judiciary than by the government.

Phase IV: Economic and Social Cohesion Recovery

Pathogen suppression and mitigation phase will eventually give way to cost-benefit analysis: economic damage balanced against real-time harm and future risk harm of pathogen. 

This has tricky ethical and moral implications and the final decision should be made by top leaders accountable to the people. You are a professional and your duty is to provide data interpretation and counsel - NOT to make the final decision. 

The trillion dollar question is: will the economic costs of a national recession or global depression outweigh the real damage and potential risks of pathogen?

Your job is to interpret the right smart data and apply probability theory to answer this critical question.

On the one hand the severe social and economic trade-offs of unlimited quarantine is an important consideration that requires deep thought. On the other hand a four (4) week social distancing / lockdown strategy that extinguishes or severely cripples the pathogen would be worth it. Short term economic pain yet economy and financial markets should bounce back fast.

Yet if social distancing / lockdown goes past four (4) weeks, policy makers are going to have to make a very important and consequential decision for society. Past four (4) weeks the national and global economy will likely dip into a deep economic recession with severe consequences that may be much worse than damage caused by pathogen.

Moreover, many people in free societies will likely not put up with forced isolation and light martial law, and break lockdown with enforcement not practical.

Thus, after a four (4) week pathogen suppression and mitigation phase, a gentle transition should begin unless strong evidence suggests existential threat.

After social distancing and lockdown strategy reduces exponential transmission rate, the new goal is to develop a plan to transition from lockdown / light martial law to sustainable social distancing strategy while turning the economy back on. This is easier said than done and very tricky. Spend time and brainpower thinking this through. Trial and error with quick adjustments likely. Be prepared to improvise on the fly.

Form a team of public and private economists, scientists, statisticians and business leaders to design and execute a plan to get the economy back to normal as fast as possible. The trick is balancing pathogen risk with the right amount of economic and social activity. 

Design real-time data collection and analysis to calibrate and recalibrate as needed in different geographic areas (some heavily infected and others not). Get this right or risk a new wave of pathogen in a more deadly mutated and re-combinated form. Use Monte Carlo Simulations and probability theory to help policy makers understand upside and downside risks.

Develop and continually improve pathogen tests that are fast, cheap, reliable and ubiquitous. The goal is to screen everyone, regularly, so we can let most people return to normal or reasonable economic and social activities.

Reopen schools and places where people gather with reasonable restrictions when we can determine the healthy and isolate infected with high certainty. If we can be assured that people who congregate aren’t infectious, they can socialize within acceptable risk levels. 

Design and execute government policy of loans and handouts to vulnerable citizens and small businesses. Be careful of moral hazard.

Phase V: Public Communications Strategy

Trust in public institutions is critical to motivate mass collective action to fight pathogen at both collective scale and individual level.

Treat citizens as adults and provide timely and accurate information using multiple media platforms. Build trust and credibility with useful information. Level with folks when you don’t know but explain balance of risks for optimal policy acceptance - especially during lockdown. Get public concerned without panic. Project competence and confidence in safety and success at all times.

Let the public know lockdown strategy is temporary and provide a good faith estimate when normal or modified social and economic intercourse shall resume. This is important for psychological buy-in and reducing police enforcement.

Help the public participate in the plan to transition from lockdown to sustainable social distancing strategy. Reassure the public of policy urgency to restart the economy and allow limited social intercourse with specific guidelines and rules to be applied equally to all.

Repeat often policy goals to aggressively recalibrate an appropriate strategy to both fight pathogen and restart the economy. Tricky to get the right balance.

Phase VI: Future Epidemic Preparations

A long term important goal is to prepare for the next serious epidemic as a low cost insurance policy. Taiwan offers a very good epidemic preparation strategy learning from past epidemics. As a result in 2020 Taiwan significantly reduced infections, mortality rate and both economic and social damage. See: Response to COVID-19 in Taiwan - https://jamanetwork.com/journals/jama/fullarticle/2762689

Therefore a team of respected experts in different disciplines should be formed to learn from previous epidemics, and design and implement a prudent epidemic war strategy for the future.

Design processes for: scenario planning, vaccine research and healthcare worker training. Create a budget to present to lawmakers for consideration. Create functional databases of information for fast and easy future retrieval.

Understand we may get lucky with this recent (2020) epidemic and have the ability to learn from successes and failures at a collective (state and global) scale level. The next pathogen may be an existential event with little to no room for margins of error. Planning for a more deadly pathogen war is good public policy and our moral duty.

Wisdom of Sun Tzu holds that wars are won or lost before the first battle is fought. Data is the lifeblood of future wars and the correct interpretation of that data will be key to winning future wars against deadly pathogens.

Be prepared my friends!

Respectfully,

Michael A. Walker

Chief Data Scientist / Team Leader

Data Science Strategy Team