Modeling
The ultimate value we gain from modeling depends on the dominant underlying values at play
So many of the decisions we make and the policies we create are dependent in part, directly and indirectly, on modeling. Examples range from our choice of what to wear influenced by weather forecasts, to the selection of a specific mechanical system type partially based on energy modeling, to updating coastal land use regulations based in part on climate modeling. Any resulting decisions we make, from the mundane to the profound, can be heavily influenced by the specific model input variables as well as how the resulting outputs are interpreted. Which means those who are making the decisions on what variables to include and those providing the interpretations have significant influence on any resulting decisions made.
on his podcast recently interviewed Dr. Erica Thompson about the use and misuse of mathematical modeling (with a focus on climate modeling). It’s well worth listening to, and I look forward to reading her book that much of the podcast conversation draws from: Escape from Model Land: How Mathematical Models Can Lead Us Astray and What We Can Do About It.I very much appreciate Dr. Thompson’s comprehensive view of modeling as a range of activities, from simple, single formulas, to complex mathematical formulations, to even storytelling where narratives, often complex, are formulated to predict future conditions (e.g., climate change fiction). One way stories can be superior to mathematical models is that they’re capable of providing potential outcomes often not found in mathematical models simply because the modelers don’t know how to estimate the probability of “X” being an outcome (and so it’s excluded). This often-necessary over-simplification that occurs within mathematical models can result in a failure of imagination compared to storytelling – it excludes possible future outcomes we may still need to consider. Of course, combining both has the potential to be more effective for understanding potential future outcomes as well as generating needed collective action.
Another important point made was the need to increase the diversity among those doing the modeling. This increases our ability to capture more of the relevant variation in inputs and incorporate a more nuanced assessment of the outputs, which in turn results in a more comprehensive view of potential outcomes and a more equitable risk or cost/benefit assessment. And since value judgements underlie model construction, choices of inputs, and analyses of outputs, another good reason for increasing the diversity among those doing the modeling is that it’s always easier to see value judgements that aren’t your own (that aren’t shaped by your own cultural background(s), formative years, and intellectual or disciplinary traditions).
Building on that, I think the most important statement made by Dr. Thompson, that underlies much of this conversation, is how important values are to all of this (including understanding who we are as a species and how we relate to the rest of the planet). She emphasized that we should be focusing less on the potential technological solutions per se and more on our varying values that shape the free-market economy and the individual and collective actions we take, including what technology we want to focus on versus reducing our collective consumption (in response to climate change).
For models (climate, building performance, etc.), values shape what goes into them (what we include and exclude), how we assign costs and benefits (sometimes arbitrarily), how we weight and interpret the results, as well as the actions “we” take as a result of the output. We should be having more conversations focused on values - making them explicit, honestly examining how they differ, and looking at how we equitably reconcile those differences.
These same ideas of diversity and values can be applied to environmental, social, and governance (ESG) frameworks and standards, which make use of modeling to varying degrees. When researching, compiling information, and assessing different ESG standards and frameworks, it’s also important to assess the underlying values driving their structure. What voices were included in their development? Are annual profit margins and shareholder value still the ultimate driver (directly or indirectly via limitations on such things as transparency and accountability)?
How are biodiversity, community resilience, diversity/inclusion, outdoor air quality impacts, shifting growing zones, climate refugees, etc., and their inequitable impacts integrated within these standards and frameworks? How are they weighted against annual profit margins and shareholder value? How prosocial are they actually? Is there an attempt to extend the time frame outwards, which tends to put the organization’s and society’s costs/benefits in greater alignment?
How much required transparency (both internal to the organization and external) is integrated within the standards? How is accountability incorporated? How much does an organization’s members have a say in all of this, including the selection of specific metrics, goals, and milestones - is this required in any way? Does the surrounding community have any influence? In essence how well are Ostrom’s principals integrated within the different ESG frameworks and standards? And is there any research on how effective different frameworks and standards have been with respect to all of this?
In terms of life cycle assessments (LCAs), is the equivalent tonnage of CO2 converted to dollars using some value of the social cost of carbon (SCC)? What’s included in that SCC value in terms of loss of biodiversity, years of life lost, the inequitable distributions of these impacts, etc.? Is there an attempt to localize that SCC in any manner, accounting for such things as inequitable environmental/health impacts by neighborhood? Is that localization attempt focused on all the relevant localities – all the potential varying locations for extraction, manufacturing, construction, and end of life?
Life cycle cost analyses (LCCAs) and building performance modeling (BPM) typically focus on first costs, utility and energy/water savings, and other aspects of building maintenance/replacement costs. By not consistently including quantified estimates of occupant health and productivity impacts, “we” are making value judgements that the occupants’ quality of life and success in their day-to-day is less important than energy consumption and other aspects of building performance (to the detriment of both the occupant and the organization they’re a part of). We reinforce these value judgements when we also ignore the stories of occupants by excluding ethnography during pre/post occupancy evaluations. Such stories can further inform modeling inputs and the interpretations of outputs.
All of this has a direct impact on the decisions made during planning, design, and construction – these value judgements shape the design, the eventual experiences of those using the facility, and the amount of greenhouse gases (GHGs) produced by that facility over the course of its life. They shape the ultimate value that modeling provides to an individual project, to communities, and to society overall.