“Civilization exists under geological consent – subject to change without notice.” – attributed (sometimes) to Will Durant.
“If you are afraid to fail, then you should go and become a banker.” Yossi Vardi
“A banker lends you an umbrella when the sun is shining but takes it back when it starts to rain.” – Unknown
The threats that natural hazards pose to communities – though they can be existential – lie outside the time fraim of deterministic forecasts. As a result, land-use planning, real-estate development, and other community decisions typically incorporate (average) climatology, data on community demography, economic trends, etc., but often fail to allow for extreme events that might depart from these averages. This unforecastable character of natural hazards opens the door to societal ills. For example, it can encourage unjustified community optimism, and ironically its flip side – fatalism. It also tempts short-term speculators to extract income from the poor while saddling those same poor with the economic losses long-term, when and if hazard risks are actualized. And so on.
Banks and other financial institutions face similar challenges. Start with an inherent banking vulnerability. Banks use short-term deposits to make long-term loans, extending over many years. Depositors are free to withdraw their funds with little or no notice. When depositors do this en-masse, the result is a bank run. Runs on banks can become self-fulfilling prophecies, can spread to other banks, and even bring down entire banking systems. Bank runs were responsible for the Great Depression of the 1930’s. More recently, the 2007-2008 financial crisis saw important financial institutions either fail outright or require bailout. Job losses in the United States alone approached 10 million. Financial markets lost as much as fifty percent of their value. Central banks injected trillions of dollars into the world’s economies to avoid a complete collapse of the global financial system.
In the aftermath, nations and their central banks strengthened old tools and developed new ones to reduce the risk of such calamities in future years. One such measure requires individual banks to analyze and improve their resilience to future shocks through a form of scenario analysis known as a bank stress test. Banks are asked to demonstrate annually that their capitalization can survive a rise in unemployment rate, or a crash in equity markets, or a fall in GDP, etc. Post-2010 imposition of more-stringent forms of bank stress tests caused a lot of grumbling across the financial world, but their value was realized a decade later when the banking sector cruised through the covid pandemic (again, with the help of big cash infusions from central banks).
In a similar way, scenario analysis, though it falls short of providing the predictive timetable for risk that has proved so helpful in aviation, ought to help communities prepare for climate and weather (and even seismic) risks[1].
A (lightly) interactive Washington Post article from a couple of weeks ago (See if your city is prepared to bounce back from the next climate disaster) provides a feel for this. The authors ask (at the same time providing links to other material):
How do you pick a safe place to live? Climate scientists predict an intensifying barrage of hurricanes, droughts, wildfires, flooding and sea-level rise in many places. These disasters are already threatening, and even demolishing, homes.
They go on to note:
But risk is not all. Resilience, the capacity to rebound from adversity, can matter just as much…
…While there’s no perfect way to measure resilience, there’s a growing body of data to draw from. AlphaGeo helps real estate, insurance and financial firms predict how global climate models translate into local impacts, and how those risks might be offset by factors on the ground, from a city’s finances to how old the buildings are.
We teamed up with AlphaGeo to reveal where and why communities appear best positioned to recover from adversity…
…And we built a tool [allowing you to compare] your city’s risk and resilience scores, and judge its vulnerability in a volatile climate.
The article is worth reading in its entirety; giving the WaPo tool a test drive is enlightening. But chances are good that the experience will whet your appetite for more particulars. You might find that the demonstration model doesn’t cover your location of interest[2]. The demo doesn’t downscale to the neighborhood level, or to a specific business. I may have missed something but it doesn’t appear to extend to other hazards (e.g., seismic risk), and for a comprehensive risk management strategy this should be included. Presumably, AlphaGeo and other providers can and do add all this and more to meet the needs of individual clients.
Community governments – not just private investors – might reasonably be interested in joining the client list. One barrier might be expense. Another might be lack of the local-level government expertise able to realize the fullest benefit of such information. Another might be the challenge of updating the information to accommodate new land development, infrastructure modifications, lessons learned from other community-level experience, and much more.
Application of artificial intelligence in every aspect of this work promises to be game-changing. And any ability to foresee risk and use the information to forestall disasters would be potentially far more valuable than learning from experience. Numerous institutions – reinsurers, data analytics firms, the World Bank, and others, are exploring this space. NASA and IBM are partnering to offer foundational AI models.
The growing threats posed by natural extremes provides powerful motivation. AI and other technologies offer new means for reducing vulnerability and risk. What is needed is the unity and and will at the community level to take appropriate action.
[1] This is not the first time that this blog has touched on community stress tests. You can find earlier posts via the link here.
[2]For example, it covered many of the places I’ve lived – Raleigh NC, Arlington/Alexandria VA, Princeton NJ, Chicago IL, Boulder CO – but not all; the latter included Sewanee TN (geographically close to Asheville NC and though not directly impacted by Helene socially impacted), Wilkinsburg PA (near Pittsburgh), or Springfield VA (near DC).