California Systems Dynamics Economic Pandemic Recovery Model

Apr 19 2020

It is quite clear that the California Task Force on Business and Jobs Recovery chaired by Tom Steyer and Ann O’Leary should use Systems Dynamics Modeling to trade off policy positions that are being considered. This is because we are in a dynamic situation in which the Pandemic might return and it may be necessary to close down again after partially reopening. This dynamic situation in which there may be a resurgence in the viral load on the economy needs to be modeled in relation to the reopening strategies. Also, the goal should be future resilience and the model could help attempt to plan future strategies that will lead to greater resilience of the economy in the future against not just resurgence of Covid-19, but also against future pandemics.

So, the question needs to be asked what the form of such a California Systems Dynamics Economic Recovery Model should be like. How do we build such a model to support the work of the Task Force that will allow them to trade off different policy choices in relation to the dynamics of the possible pandemic resurgences that might cause the economy to retract again after opening? There clearly has to be a part of the model focused on the pandemic effects and possible contagion dynamics. On the other hand there needs to be a part of the model that focuses on different sectors of the economy. And another part of the model that is social and related to economic factors based on different behaviors of the public as they go back to work and start mixing again creating vulnerabilities for the continued spread of the virus. The right granularity of modeling must be picked based on the kinds of questions the task force has with regard to their policy recommendations to the Governor. Experts from various disciplines such as Economics, Sociology, Business Administration, Infectious Disease Experts and other disciplines need to have input into this model in order to make it robust enough. However, it cannot be too detailed to become unwieldy.

It is suggested that the place to start is with the Causal Loop diagrams that are fundamental to the way in which the dynamics in the economy and in the pandemic are understood. For us the rate of infection is unknown because of a lack of testing. Therefore, we only have a lagging indicator of the rate of hospitalization as the basis of our current modeling. Fortunately, we have been able to flatten the curve and to so far avoid the dire predictions that were first made with regard to the deaths from the pandemic. But the problem is that there is this fine line to be walked between opening up business and reinfection rates. And this is where good modeling could help policy makers attempt to understand different scenarios related to different policy choices in order to avoid unintended consequences. It is actually thrashing that has the most probability of causing harm to the economy. Thrashing means opening up then having to close down again quickly repeatedly. It is the dampening of this possible thrashing that Systems Dynamic Modeling might help us understand better before we make mistakes in our reopening policy recommendations.

My suggestion is that we divide up into teams that would create sub-models of the overall model. It is recommended that there be a generic economic sector model that is elaborated for each specific sector. And that there is a general economy model that represents that sector’s interaction with the whole economy. There needs to be a pandemic epidemiological sub-model hooked to various models of the pandemic itself and its dynamics. There needs to be a part of the model that represents government actions and then another part of the model that represents public response. The workforce needs to be represented and the relation between unemployment and reemployment. Companies and their supply chains also need to be represented in this model. In other words, it needs to be comprehensive with regard to the Pandemic situation we face but at a high enough level of abstraction so that it can actually help reflect the differences in policy choices that are considered by the task force. The model should be open to inspection by the public so that others might critique it and so we get maximum input into the model from experts in the various fields represented in the model. A public model would inspire more confidence in decisions based on the model than a private model. But this is of course a task force decision.

Here is the key to why such a model is useful in a situation like this. The model becomes a repository for your knowledge of the situation. It allows you to focus on what you need to know in order to make decisions which you can then build into the model. Then it gives you a place to plug in the relevant data as it is known in order to discover the impact as the situation evolves. It gives a way to focus on details as sub-systems within the overall model are considered by experts, but then there is an overall answer at the highest level of abstraction that can drive policy choice selection given different scenarios.

Systems Dynamics is a standard technique. There are many different experts in building these kinds of models who exist and can be recruited to help build it. It brings to the fore the dynamics one is concerned about which in this case is re-infection verses business functioning and the openness of the economy. This same approach could work for every state. There could be a shared model between states that is generic so that other states could benefit from the work on the California model.

I recommend that Systems Scientists in California and other experts who know how to build Systems Dynamics Models as well as disciplinary specialists cooperate to produce a model that Government and Business leaders can use to guide their policy recommendations and choices in these trying times.


Coda: To Sys-sci google group

Thanks to everyone for your comments.

I am reading this study with an eye to whether it provides a starting point for attempting causal loop diagrams.
What I think it is worth while trying to model is the possible thrashing of the system if there is a sequence of closed…open…closed…open… as a result of opening too early that looks likely in some places. When a machine starts thrashing that is when it can literally tear itself apart. And it is this possibility of thrashing that I think it is worth trying to model using Systems Dynamics in order to try to inform policy alternatives.
It is obviously a big job to produce a model like this. I am merely trying to raise awareness of the role Systems Dynamics could play in this policy development process in case the task force is unaware of that particular technique. A generic model could be created that could apply to any state not just California. The idea just occurred to me in the context of the discussion about the California task force. It is an opportunity to bring Systems Theory and Systems Engineering to the mind of people as something useful.
Kent Palmer

No responses yet

Comments are closed at this time.

Shelfari: Book reviews on your book blog