Five things governments can learn from complexity science in the fight against the current pandemic

complex systems in nature (tree)

By Medea Fux

Yaneer Bar-Yam is the founding president of the New England Complex Systems Institute (NECSI) and has been a longstanding and prominent advocate for a complexity-informed understanding of policy making. In a recent online conversation with more than 50 participants from across Switzerland we spoke with him about how complexity science can help governments understand pandemics and other wicked policy problems.

What do biodiversity, financial markets, ethnic conflict, and pandemics have in common? All of them pose major challenges for governments and all of them can be understood using the science of complex systems. While much has been said and written about the COVID-19 pandemic, some of the most insightful contributions have come from complexity scientists. This is why we wanted to speak with Yaneer Bar-Yam, president of the New England Complex Systems Institute (NECSI), frequent advisor to government entities and leading systems scientist.

Yaneer first warned of the dangers of COVID-19 in January 2020, well before most of us, and well before most governments, had fully realized what was coming our way. Almost anything that you can think of, which now is accepted as common sense but which was controversial at the time — mask-wearing, travel bans, the role of aerosols, etc. — Yaneer was there first. So we’ve got to ask: how is it that he has been so consistently ‘on the money’? The answer lies, in large part, in Yaneer’s background in complex systems science, which is a uniquely powerful way of thinking about pandemics and other wicked public policy problems.

Here is what we’ve learned:

When dealing with any system, complexity science tries to find the relevant control variables, Yaneer explained. Imagine, for example, trying to drive a car without knowing anything about cars. Knowing nothing about driving, you would want to first identify the car’s controlling parts, such as the gas pedal, the steering wheel and the brake pedal. Only once you know which parts of the car control its behaviour do you have any hope of being able to drive it.

In the current COVID-19 pandemic there are, according to Yaneer, two control variables: 

1) The reproduction number (called “R”): If R increases, the cases of disease increase exponentially. If R is decreasing the opposite, an exponential decrease , will happen. The role of R has, by now, been well understood.

2) Travel (in particular long-distance travel): Simulations show that long-distance travel “changes the game” in terms of what can happen with local disease outbreak. Even a small amount of long-distance travel increases the risk of a pandemic dramatically. This shift from local outbreaks to global pandemic does not happen as a slow and gradual process but as a phase-transition, i.e. gradually and then suddenly. 

On the basis of these variables Yaneer advocates for a so-called “zero COVID” strategy, an approach that has recently also been put forward by others. The aim of a “zero COVID” strategy is to eliminate all cases, not just to reduce them to an “acceptable” level.

To follow such a “zero COVID” strategy long-range transportation must be restricted and a “green area method” must be applied: severe restrictions must initially be put in place to reduce contact between people and cause R — and with it case numbers — to fall. Once a community, a city or a region is COVID-free restrictions can be eased. Travel must only be allowed between such “green zones” in order to avoid the reintroduction of the virus.

Such a community-based approach has proven to be successful during the 2014 Ebola outbreak in West Africa [6]. Research conducted by Yaneer has shown that such a strategy, when implemented with the required stringency, can eliminate COVID-19 cases — and therefore allow a return to some kind of “normal” —  within five to seven weeks.

Switzerland is well suited for such a “zero COVID” strategy, says Yaneer. The country is politically stable, government generally enjoys high levels of trust, and as a federal state there exist capable entities at the local level which could take the lead in such a community-based strategy.

What would complexity science tell us about the success factors of such a strategy?

Here are five things which, according to Yaneer, are key:

Local action is crucial
The disease can only go extinct by local action. We have to move from a race to the bottom between cantons (who has the least restrictions?) to a race to zero (who has the least cases?). An important incentive here is that local COVID-19-free zones could lift restrictions long before areas with more cases.

Travel restrictions, especially for long distance travel
Long-range transportation must be restricted, such as for example long-range flights. Moreover, transport between countries as well as between regions within the country must be restricted as well. The goal is to prevent mixing between areas, such that the number of cases can decrease locally.

Do not aim for perfection
The crisis management of the pandemic is not to be compared with a complicated watch, in which the placement of each tiny piece is crucial. Rough and directionally right decisions need to be taken. The aim of those ought to be getting rid of the virus rather than living with it, since living with the virus is not a sustainable strategy. Specifically this means, for example, that workers that need to cross borders daily would need to get travel permits. Necessary exceptions such as this one would, however, not significantly change direction of the strategy. The big decisions are what’s important, states Yaneer. 

Communication is key
The national government should not necessarily be doing a lot of implementation work, its most important task is communication. Information to citizens, to local authorities and institutions must be clear and transparent. Only if the knowledge about the situation and the information about the strategy is given, action can be taken. Furthermore, it is crucial to also explain the why. In other words the motivation for any kind of strategy needs to be clear.

Going all the way — the only way out
A “zero COVID” strategy only works if a country swings for the fences and goes “all in”. It cannot be done half-way. A “zero COVID” strategy is however, in Yaneer’s view, the only way out of the current situation. If applied stringently it would only take five to seven weeks to go to zero cases and allow COVID-free localities to lift restrictions.


The COVID-19 pandemic illustrates the novel and yet actionable perspective that a complex systems lens can bring to bear on wicked policy problems. All attention is currently and justifiably being placed on combatting the pandemic. We ought however, to learn for the future and make full use of the insights that complex systems science can bring for the many other challenges governments are faced with.