Comments about Rising et al (2022) - Challenges and innovations in the economic evaluation of the risks of climate change

Mainstream economic calculations and modelling are the foundation of the comments we hear in the news about how the economy is doing.

Unfortunately, the news we hear is distorted by the fact that the economic models underlying them are flawed.  This is because they fail to adequately address the impacts of human-driven climate change, described by Rising et al (2022) as “the dire impacts that most natural scientists project we could face from climate change”.
​
 
The first main problem with existing mainstream economic evaluation tools and methods -  completeness
 
For example, Rising et al say:
“At present, economic estimates of the impacts of future climate change are widely acknowledged to omit some of the biggest risks”.
https://www.lse.ac.uk/granthaminstitute/publication/challenges-and-innovations-in-the-economic-evaluation-of-the-risks-of-climate-change/
 
If we are to place reliance on future economic models (or improved models), they should cover, as Rising et al describe it:
“the full range of physical impacts and their associated market and non-market costs, considering the greater vulnerability of poor people and the challenges of adaptation.”
 
The core challenge of bringing economics up to speed with climate science is summarised by Rising et al as follows:
“Climate change is already having economic impacts around the world and will continue to do so for centuries, but our understanding of how it affects societies and economies is much less developed than our understanding of likely changes in the climate system”
 
The area of economics that needs to be improved the most is the understanding, and estimating, of the environmental, market and social impacts of climate change, in order to:
 “make [or recommend] sound policy decisions about the management of climate-related risks”, as Rising et al put it.
 
Mainstream tools used by economists currently include cost-benefit assessments of adaptation and mitigation actions, but such assessments have been incomplete and inaccurate to date because they aren’t comprehensive enough to include all impacts on markets and societies.  They exclude, for example, market and social impacts of changes in natural patterns of water distribution, and of changes in frequency or severity of extreme weather events.

Climate damages, in terms of impacts on markets and social welfare, should be one of the key metrics calculated as part of economic assessments.  However, this cannot be done properly unless comprehensive climate damage functions are constructed and included, covering all relevant damages and their costs.
 
In the language of economics, an important measure is total loss (or gain) in “social welfare”.  The term “social welfare” in this context means the wellbeing of all of us.  Not to be confused with the social welfare “systems” that governments put in place that focus on supporting the most vulnerable members of society.  “Social welfare” as an economics concept is about the social wellbeing of all of us (including, but not limited to, the most vulnerable).
 
Clearly, we’d like total social welfare (the economics concept) to increase (all other things being equal).  We’d also like there to be increasing economic success on other measured dimensions, but they should all ultimately feed into improving the value of the total social welfare function as the overall goal of society. 
 
Our policy decisions (or rather, those of the politicians we elect) can be evaluated by calculating their impact on this total social welfare function.  Positive impacts increase total social welfare, but negative ones, like most impacts of Anthropogenic Global Warming (“AGW”) decrease total social welfare.
 
The challenge is that mainstream economics has not been good at providing us with effective measurements of these impacts. 


So, what can be done about this? 
Part of the answer will lie in economists working more closely with people from other disciplines, for example; natural sciences, health, politics, psychology, anthropology and accountancy.  This might mean expanding/improving existing mainstream economic models or developing new economic models, and improving the communications between people from various disciplines working with such models.
 
Rising et al explain one of the most important challenges:
“There is… a diversification of [economic] models specialising in different features and insights, and both improvements in existing models, along with a generation of new models based on alternative principles. However, difficulties in communication between natural scientists, economists, and modellers remain, slowing the scientific process (Ciscar et al., 2019). Interdisciplinary groups remain rare.”

Rising et al give some pointers about the areas of research where improved understanding in recent years has not been reflected in the sorts of climate models mainstream economists use for their calculations:
“Much progress has been made in quantifying and monetising effects of climate change. Impacts on agriculture and forestry, water resources, coastal zones, energy consumption, air quality, tropical and extratropical storms, human health and mortality, physical performance, cognitive performance, crime, and social unrest have been widely studied and are, at least partially, quantified.”

Not all of these matters are yet capable of being incorporated in formal economic damage functions, for a range of reasons including lack of sufficient, reliable data and continuing uncertainties.  However, this highlights how far short of including a proper damage function most of the economic models are.

 
The second main problem with existing mainstream economic evaluation tools and methods - assessing risks
 
Classic cost-benefit analysis, the most significant mainstream tool used by economists, depends on the validity of a conventional “utility-maximizing” approach.  What this means is that, generally, the best policy choice among options analysed is assumed to be the option which maximises “utility” (for example, maximises the total social welfare function, projected forward into the future and added up for all the years in a specific timeframe, of, say 50 or 100 years).

However, Rising et al suggest that, because of the types of matters (especially damages from climate change) currently under-represented and under-estimated in mainstream economic models the resulting cost-benefit calculations play down or even ignore many important risks, resulting in misleading advice being given to policy-makers.

Rising et al recommend applying Precautionary Principles in this situation, and that policy makers make a conscious effort not to restrict themselves to just using classic cost-benefit analyses in their decision making:
“The existence of deep uncertainty in [some climate change impacts] provides a strong motivation for precautionary policy, as insurance against those disasters (Ackerman et al., 2009). It may be necessary to look to decision-making frameworks that extend beyond the conventional utility-maximizing approach, such as the tolerable windows approach, the safe landing approach, robust decision-making, decision-scaling, the smooth ambiguity model, info-gap decision theory or cost-effectiveness analysis, and ambiguity aversion”
 
They explain what they mean by the expression “deep uncertainty”:
“Deep uncertainty [means] situations in which conceptual models, probability distributions, and/or the value of various outcomes are unknown or cannot be agreed on.”
 
Another way of describing the situation is that risks associated with climate change and actions to address climate change are asymmetrical.  They do not generally show a bell-shaped probability distribution above and below the most likely values of outcomes. 

Comparing and contrasting high-emission scenarios and low-emission ones will illustrate this.  The risks of some of the most severe negative outcomes from some future climate scenarios represent high impact (from climate), low probability events.  On the other hand, the risks of most scenarios where emissions have been reduced to zero fast enough to stabilise at temps consistent with the Paris 1.5 target represent high probability, low impact (from climate) events. 

The mathematical “expected value” (probability multiplied by impact) from each of these types of scenarios might be similar, resulting in similar cost-benefit, which might lead policy makers to be indifferent between all these scenarios, because it is not clear that any of them have significantly better cost-benefit than any of the others.

In fact, it’s clear that we should not be pursuing the high-emissions scenarios because of the Precautionary Principle guiding us not to allow scenarios to unfold where there are high emissions and significant risks of crossing climate tipping points (something it has already been noted is, in any case, not well represented in mainstream econometric models and classic cost-benefit calculations).

 
Human (individual and societal) adaptation responses enter the fray

Another area that complicates, and even confounds, many modelling and cost-benefit approaches is adaptation response. 
 
As Rising et al explain:
“The complexity of the adaptation process represents a huge modelling challenge: humans will respond to both the realization and the anticipation of climate change, in ways which will ameliorate hypothetical impacts in some cases, aggravate them in others, and displace the impacts in yet others.”

This presents yet another uncertainty in how the future will unfold, and yet again supports a precautionary approach because of the vast potential uncertainties on downside risks.  A simple example could be a scenario where human responses to climate change make matters worse through selfishness, “othering” of out-groups and protection of in-group members at the expense of the rest of humanity and nature.
 
 
 
What’s the long-term optimum?

So far, I’ve talked about short term responses to AGW, by which I mean a time frame of years to decades.
But, consideration of social welfare functions and other aspects of economics has an important part to play in long-term decision making as well.

In the short term it helps identify policy actions to help avoid the worst-case climate scenarios from crystalising, especially when applying the Precautionary Principle.  Paris 1.5 and Global Net Zero by mid century are both good examples, and my comments so far have had that timescale in mind.

Let’s now consider long-term optimisation.
In the long term (by which I mean centuries to millennia) all of the tools and techniques above are also relevant to the following question:

“what is the optimal global average temperature for the flourishing of humanity and all other life on Earth?”
There is a variety of work by economists shedding some light on this.  However, they suffer from the same difficulties that I’ve pointed out in relation to cost-benefit calculations and so on.

While some economists have generally supported efforts to reduce the temperature anomaly to below 1.5 degrees C, there are a small number of outliers (eg Nordhaus suggesting somewhere between 3 and 4 degrees would be optimal for the planet).  In fact, given that one of the arguments for tackling AGW is that pre-industrial conditions appear to have been near-optimal for human civilisation for the last few thousand years and we are diverting dangerously away from that temperature at a very rapid rate, I think it’s important to consider if pre-industrial might, in fact, provide the optimum we are seeking now.  That translates into zero degrees above pre-industrial – ie well below even the Paris 1.5 target.  Work that suggest this has not had much attention, probably because, in the short-term, there is no practical prospect of getting to such temperature.  We are already at more than 1 degree above that level, and still heading upwards.  The policy priorities are currently (quite rightly) aimed at temperature and climate stabilisation in the first instance.  Once the global temperature anomaly has peaked (or nearly peaked), we can bend attention towards the longer-term optimum.  Until then, ie while temperature is strongly rising, it’s a moot point for future debate.

The main gist of my commentary, however, is that the economic tools we use to help us decide on our pathway forward need to improve, so that they are better at reflecting the nature and size of the risks we face with our current trajectory, and so that we are steered away from the worst-case outcomes.
 
As a footnote on change management, whatever the new long-term optimum temperature we eventually calculate, this doesn’t mean we should change the average temperature to achieve a new optimum (whether that be zero, 0.5 C above pre-industrial, 1.0, etc) as quickly as possible and at all costs.  There are tools, such as carbon pricing, that we can use to plot the most economically efficient and effective path to get there.  But that’s a discussion topic for another day.