Technology Assumptions and Climate Policy: The Interrelated Effects of U.S. Electricity and Transport Policy.

• Author: Jaccard, Mark
• Position: Report

INTRODUCTION

To reduce greenhouse gas emissions from the energy sector, U.S. policy makers would benefit from a clearer understanding of the interaction of several uncertain factors. What are the prospects for cost reductions and market penetration of different technologies in different but related sectors, such as electric vehicles and solar electricity? What will be the preferences of firms and consumers in the face of technological change and climate policy? How will regulatory policies perform relative to price-based policies? What will be the cost and energy price effects of achieving a given emissions target?

To address these questions, energy-economy modelers apply different types of models. (1) Some are technologically-explicit, focusing in detail on the current stock of energy-using equipment and the characteristics of existing or emerging technologies. Some have little or no technological detail, focusing instead on the responsiveness of the economy as a whole and individual sectors to climate policies. Some try to bridge these two perspectives, albeit with some important limitations.

Conventional technology-explicit models have traditionally been referred to as bottom-up, engineering or technico-economic models. These represent the various energy end-uses in the economy and the technologies available to service them. Technology-explicit models are appropriate tools for examining the effect of new technological developments or policies that mandate a certain technological outcome, like a renewable portfolio standard that requires a rising share of renewable electricity, or a vehicle emissions standard that requires a growing market share for zero-emission vehicles.

However, their level of technological detail creates modeling challenges because they need detailed information on how firms and households make micro-economic decisions in choosing one technology over another. Most technology-explicit modelers use some kind of rule-of-thumb, suggesting for example that those technologies with the lowest financial cost should be preferred. Some compare this with an optimization algorithm that assumes an economy-wide cost-minimizing outcome in the selection of technologies. Unfortunately, this is unhelpful to policy-makers trying to assess how firms and households are most likely to respond to technology-focused policy, or even policy that combines an economy-wide emissions price with some technology-specific mandates. Finally, such models are usually partial equilibrium in that they fail to account for all feedbacks as the economy adjusts to a given policy, which can entail changes in the structure of the economy and its total level of output.

In contrast, models that are not technologically-explicit are typically referred to as top-down models. The standard approach is to apply computable general equilibrium (CGE) models that contain a set of elasticity parameters in order to simulate substitution between economic inputs and between final product demands.

Elasticities depict how a given increase in the price of an input like energy will lead to a given decrease in its market share relative to other inputs. This is an abstract way of representing a shift in the economy toward technologies that are more or less energy-intensive, or that use one form of energy instead of another. Since these models represent the economy at an aggregate level, it is not too difficult for them to encompass key indicators of macro-economic performance, including energy prices, capital costs, sectoral production costs, sectoral output, investment, and changes in economic output and welfare.

To the extent that their elasticities are estimated from past market responses, these models can provide useful information to policy makers. However, although these parameters may accurately depict past responses to price changes, they may be inaccurate in portraying the response in future periods, especially if the technology choice set is changing significantly. Historically-derived estimates for inter-fuel substitution for personal vehicles could be inaccurate for assessing the likely rate of uptake of electric cars since these were not a viable option for consumers during the historical period for which behavioral data are available. Likewise, elasticities for electricity generation options may not tell an accurate story when future options include renewable technologies that were less viable over the historical period.

Frustration with the deficiencies of these two conventional modeling approaches has led a growing number of modelers to develop hybrid variants that combine some aspects of the technology-explicit and the CGE models within the same framework. On one hand, a technology-explicit model might incorporate feedbacks between energy supply and demand and between sectoral production costs and sectoral output. On the other, a CGE model might incorporate detailed technological representation of electricity generation or personal vehicles in an ancillary sub-model that feeds into the general equilibrium solution.

CIMS-US is one of these hybrid models. It is technology-explicit, like a bottom-up model. But it simulates technology choices based on behavioral parameters estimated from discrete choice surveys of revealed and stated preferences, thus representing the micro-economic responsiveness of firms and households to changes in costs and regulations. (2) It also simulates shifts in service demand in response to changing service costs (for personal mobility for example) and shifts in industrial output for a change in production costs, thus representing in part the macro-economic responses to changing energy prices. It is, however, only partial equilibrium in that it lacks the full macro-economic linkages of a CGE model (Bataille et al., 2006).

In EMF 24, we applied CIMS-US to...