State Space Models

All state space models are written and estimated in the R programming language. The models are available here with instructions and R procedures for manipulating the models here here.

Tuesday, July 29, 2025

The Business-as-Usual Scenario (BAU): Models vs. Reality


The Intergovernmental Panel on Climate Climate Change (IPCC)  started publishing Emission Scenarios in 1990s and almost immediately abandoned using a Business-as-Usual (BAU) scenario because it was thought to be misleading. BAU implied that nothing was being done but governments and industries were adopting some sort of Climate policies (even if they were just Greenwashing). What Business-as-Usual really meant was becoming murky.

With the Fifth Assessment Report (AR5) in 2014, the IPCC began using an handful of Representative Concentration Pathways (RCPs, see graphic above for RCP1--RCP5). Unfortunately, what happened is that the Worst-Case Scenario (RCP8.5) became the BAU scenario. In 2020, Hausfather and Peters wrote an article in Nature and on the Carbon Brief Blog arguing the Worst-Case or BAU scenarios were unlikely and interfered with policy development.

First, the IPCC tackled a really difficult (unsolvable?) problem (the Future is Unknowable) in a very intelligent way (maybe the only way) and I have used the approach extensively in all my economic and environmental projections (see the Boiler Plate). But, I haven't given up on the BAU Scenario because it can be defined precisely with Systems Theory (not that a precise definition will reduce confusion but at least I can try).

A BAU Model df = A dynamic system model without inputs.

In reality, it is very rare to find closed systems. Everything is related to everything.  This is why scientists construct experiments, to control (isolate) system inputs. When we are study socio-economic-environmental systems, we cannot conduct experiments; we cannot control system inputs. But, we can construct dynamic models and we can control the inputs of the models.

And, in my experience, the mental models of most policy makers involve BAU systems, systems without inputs. For example, when politicians impose protective tariffs (a favorite policy choice of the Trump II Administration in the US), they seem to uniformly forget that their country is embedded in a World-System where there will be second-round effects and retaliation. If the BAU model is the most likely mental model used by Policy Makers, it needs to be involved as a comparison for different (provably better, at least within the models) policy measures.

Systems Theory also helps here by helping define the Policy Space as filled with alternative systems, many of which will have inputs. Starting with the BAU model helps us define the other models.

Notes

From Google AI:











No comments:

Post a Comment