jrissman
commented
4 years ago Fully implemented in 929cb86. The multiplier adjusts:
- transportation travel demand
- building energy services demand
- industrial energy demand
- industrial process emissions
- electricity imports and exports
- non-electricity fuels imports, exports, and production
Production of electricity is influenced indirectly via the change in electricity demand from the transportation, buildings, and industry sectors. Most financial metrics should automatically reflect the impacts of the changes in energy demand. Remember that since AEBC is a control lever, not a policy lever, it affects both the BAU and Policy cases, and most financial outputs only report the difference between these cases due to the policy package, so for the most part, you won't see any "financial impacts caused by the recession" in the model outputs. Rather, it's just adjusting the BAU case, so that policy impacts on financial metrics correctly reflect the recession environment.
In the AEBC Excel file, this implementation includes an approach to estimating values for the U.S. by taking per-capita GDP and energy use data from two one-year periods (during and before the Great Recession of 2008-2009) to establish a relationship between recession-driven changes in per-capita GDP and energy use. As we use one-year periods, we don't need to worry about seasonal variance in energy use. We then apply this relationship to Goldman Sachs predictions of the likely severity of the recession caused by the coronavirus pandemic to estimate resulting changes in energy demand over the 2020-2025 timeframe, relative to the input dataset's pre-pandemic assumptions.
The EPS estimates the effects of policies and various other technical measures relative to a BAU case. It builds the BAU case largely from input data. Hence, there is normally only a single BAU case, and dozens of easily-made and easily-edited policy scenarios. Making alternate BAU cases requires modifying a variety of input data files (e.g. travel demand, building energy demand, industry energy demand and process emissions, etc.). This allows for high accuracy but can be laborious.
As a labor-saving measure, we propose adding a control lever that broadly and approximately represents economic recessions or booms (that is, increases in energy demand and spending relative to the standard BAU case) due to exogenous economic forces. As this would be a control lever, not a policy lever, it would affect both the BAU case and the policy case.
Users would still have the ability to modify data files on a granular level to create detailed, custom alternate BAU cases, but this control lever could help make it fast and easy for users to approximate the effects of recessions or booms without modifying input data files.
This control lever would be carefully woven into the calculations throughout Vensim as a multiplier that reduces or increases demand for various things (travel demand in the transportation sector, demand for industrial products in the industry sector, demand for energy-using building services in the buildings sector, etc.). In the Excel file governing this control lever, there would likely be a relationship between change in GDP and change in demand for these things, so the user would specify a single percentage change in demand (time series), and the Excel file would convert it to sector-specific demand multipliers. An economy-wide demand multiplier could be used if sector-specific multiplier data are unavailable.