Jump to main content.

Research Project Search
 Enter Search Term:
   
 NCER Advanced Search

Final Report: Dynamic Adjustment to Incentive-Based Environmental Policy to Improve Efficiency and Performance

EPA Grant Number: R830990
Title: Dynamic Adjustment to Incentive-Based Environmental Policy to Improve Efficiency and Performance
Investigators: Burtraw, Dallas , Palmer, Karen
Institution: Resources for the Future
EPA Project Officer: Wheeler, William
Project Period: June 1, 2003 through May 31, 2005
Project Amount: $173,684
RFA: Market Mechanisms and Incentives for Environmental Management (2002)
Research Category: Economics and Decision Sciences

Description:

Objective:

Policymakers advance economic efficiency when they set policy goals at levels that equate the marginal costs of additional pollution controls with the marginal benefits of improvements in environmental quality. Theory and experience indicate that economic efficiency is advanced also when policymakers employ market-based approaches such as tradable permits to achieve these goals in a least cost manner. When attempting to set goals, policymakers face a great deal of uncertainty about the costs and benefits to society of achieving a particular goal and, in particular, how those costs and benefits are likely to change over time. Ideally policymakers would like to be able to update environmental targets specified in legislation or regulation as new information about the costs or benefits of achieving a particular target becomes available.

This project addresses the incorporation of new information about costs into market-based environmental policy. When policymakers adopt the use of market-based policies they put in place an institution that provides instantaneous information about the cost of compliance and this information might be used to update emission targets.

The objective of this research is to examine regulatory approaches that extend the use of market-based environmental policy to facilitate the updating of environmental targets to reflect new information about the cost of the policy. We consider several mechanisms that could be used to make regulation more flexible and able to adjust to new information. We also consider how investors respond to new information, and how policy can anticipate that. We evaluate the robustness of different approaches in different situations and in response to various sources of uncertainty. A secondary focus is to consider the incentives for continual improvement in environmental performance that such an approach may provide. To make the research concrete and relevant to contemporary policy, we focus on emissions of air pollutants from the electricity sector, although we aim to draw lessons that have more general applicability.

Methods

This project addresses the way that market-based policy can be designed to anticipate uncertainty in the implementation of the policy and especially to be able to anticipate how new information can be incorporated in an automatic fashion. Two distinct projects took shape under this research. We discuss the research methods pertaining to each before turning to a description of the research results, in part because some of the model development was relevant to both projects. The Haiku electricity market model solves for equilibria in 20 regions of the US linked by transmission capability. The model solves for capacity investment and retirement and system operation over an extended time horizon, accounting for three seasons and four time blocks. The model provides a framework for examining policy scenarios and provides internally consistent equilibrium forecasts of relative prices of fuel, allowance prices and wholesale and retail electricity. This model is combined with analytical efforts, and the use of other models in each of the projects described below.

A. Dynamic Adjustment and Symmetric Cost Management

One of the driving concerns in the design and implementation of market-based policy is uncertainty over the potential cost of the policy. The presence of uncertainty about both the costs and benefits of regulation affects the choice of policy instruments. Meaningful efforts to control emissions could prove much more costly than anticipated, or alternatively they could be less costly, and each eventuality would have important relevance to the efficiency of the policy. 

The flexibility given to individual firms that is inherent in a cap and trade program has as itsunderlying justificationthe variability in costs of the environmental policy at individual facilities. To other firms and the government, the variability in costs appears like uncertainty. Investors would be expected to take into account the distribution of potential outcomes, so the underlying variability along with uncertainty about various factor prices and technical issues may be thought of as a fundamental characteristic of the problem.

Concern about costs has motivated the proposal for a cap and trade program for carbon dioxide, with a provision called a “safety valve” as a component of some federal legislative proposals that would militate against spikes in the cost of emission reductions by introducing additional emission allowances into the market when marginal costs rise above the specified allowance price level. We investigate this policy approach from two perspectives.

First we examine the performance of previous cap and trade programs. We use an integrated assessment model called the Tracking and Analysis Framework to analyze price volatility in the SO2 trading program and the influence on benefits and costs of the program if various types of cost management policies had been implemented.

The second perspective is to analyze the incentives that cost management policies give to investors. The imposition of a one-sided safety valve will influence the market equilibrium and affect the decisions of investors, with unintended and potentially negative consequences that could undermine policy goals.

To illustrate this we consider the profit function for a single firm that offers nonemitting electricity generation. In a profit maximization model we show that an asymmetric high-side safety valve intentionally alters the distribution of the potential allowance price, so that the price cannot rise above the safety valve level (SV). The addition of the safety valve censors the potential distribution of allowance prices, reducing the expected value of allowance prices. A consequence of the change in the allowance price would be a change in the equilibrium in the electricity market, leading to a lower price under the safety valve.

The lower expected electricity price would lead to a reduction in its investment and output. The consequence is to reduce the expected profitability and therefore the level of investment in the nonemitting facility. One can conjecture that in the aggregate the policy leads to less investment in renewable technology or low-emitting technology that may suffer a price disadvantage when the external social costs of electricity generation are not included in electricity price. The cap and trade program serves as a mechanism to internalize into investment decisions the social cost of technology choices and “level the playing field,” as many observers have suggested. However, the single-sided safety valve would appear to provide an asymmetric influence that would tilt the playing field away from investments in nonemitting sources.

This formulation assumes the behavior of other investors or actors in the market does not respond to the change in expectations, and that aggregate quantity and characteristic of generation by other parties is unchanged due to the change in allowance price. We conjecture that a new market equilibrium would yield an expected electricity that is between the two measures that are illustrated in the algebraic example. To find the equilibrium we use simulation modeling.

In the simulation we explore underlying uncertainty about natural gas prices in the future. As the central case we adopt EIA (2006) forecasts reported in the Annual Energy Outlook. In the modeling exercise we freeze natural gas and coal prices at the assumed forecast values in each year and thus these fuel prices are not allowed to vary with the level of fuel used. We also freeze the level of electricity consumption in order to avoid second best issues in the welfare calculation that are associated with differences between price and marginal cost. For all of the pollution policies emission allowances are allocated to emitters on the basis of historic generation and additional allowances are purchased at the safety valve price.

In one set of results we identify equilibria when actors behave as though they have certain and perfect foresight – e.g. they know the future path of natural gas prices and respond accordingly. Investment decisions are made as though each actor knows for certain the future values of every variable, as well as the decisions of every other actor, so there is no uncertainty taken into account in the model solution. 

In the previous example, the finite differences are calculated using the model with perfect foresight. We explore an alternative with certain but imperfect foresight; wherein investment decisions made under one set of assumptions could prove imprudent were conditions to change unexpectedly. For example, if gas prices deviate from expectations after investment decisions have been made, then generators could experience large losses in profits and welfare could be negatively affected. Since the safety valve is a policy attempt to mitigate the welfare costs of surprises such as this one, we consider a case where investors’ expectations are incorrect, and use these data to calculate finite differences.

B. Incentives for Investment and Innovation

The solutions to important environmental problems are thought to require major levels of investment and to encourage technological change; the most prominent example of this notion is what will be required to address climate change. Another aspect of a number of environmental problems, especially climate change, is the potential multi-faceted policies that may be employed. For example, different proposals to reduce greenhouse gas emissions from the electricity sector suggest that U.S. climate policy to restrict emissions of CO2 is likely to include both incentive-based provisions (most likely cap and trade) and some type of technology standard. The likelihood of technology standards being part of a new climate regulatory landscape is increased by the fact that standards are a widely discussed mechanism at the state and regional level and by polling evidence that the U.S public prefers regulatory standards to approaches that impose a price on CO2.

In this project we examine the effects of combining a national CO2 cap and trade allowance market with a technology forcing performance standard. We model three different ways in which a technology standard could be used to control CO2 emissions from coal-fired power plants: 1) a New Source Performance Standard (NSPS) regulation that mandates maximum CO2 emission rate standard calibrated to require the installation of carbon capture and sequestration (CCS) equipment on all new coal-fired power plants (analogous to new source performance standards for sulfur dioxide), 2)  a flexible NSPS mechanism that allows the owner of a newly constructed facility to pay a carefully chosen emissions surcharge for new technologies that fail to meet the maximum CO2 emission rate standard, and 3) a policy that accumulates the revenue from the surcharge into an escrow fund and uses that revenue to fund investment in CCS technology.  

Table 1. Technology Policy Definitions

Policy 0

No technology policy

Policy 1

A strict technology policy that requires that CCS be installed on all new fossil generation facilities.

Policy 2

A flexible technology policy that requires CCS be installed on a new fossil generation facilities or that generators pay an emissions surcharge on emissions of CO2 for any new investment that does not include CCS.  Many different versions of Policy 2 were modeled by varying the CO2 emissions surcharge.

Policy 3

Policy 2 plus the creation of an escrow fund with the surcharge revenues that can be drawn upon by firms to help defray the costs of investment in CCS.

We examine the options for a merchant power plant investor in the Illinois coal basin region for construction of new fossil-fired generation facilities and opportunities for applying CCS either at the time of initial construction or as a post-construction retrofit. Analytical methods are used to generate hypotheses that are tested using numerical methods. The numerical simulation modeling combines three separate models. We use the Haiku electricity market model, which is a national capacity planning and dispatch model, to identify equilibria in electricity, fuel and allowance markets nationally and in the Illinois region. These results become parameters in a simulation model that uses stochastic dynamic optimization to forecast investment strategies for different regulatory regimes with associated CO2 price trajectories, and each of the three NSPS mechanisms. Data for the model draw on 2007 Carnegie Mellon University’s Integrated Environmental Control Model.

To characterize the uncertain environment in which policy takes shape and investment decisions must be made, we consider three possible natural gas price trajectories and four possible cap and trade regulatory scenarios for CO2. Each of these is considered uncertain and they determine the equilibrium value of other uncertain price variables. Over time, the investor gains new information about the future course of natural gas price and climate policy that determines the emission allowance price. These scenarios are listed in Table 2.

Table 2. Matrix of Scenario Characteristics (scenario numbers in cells)

Climate Cap and Trade Policy*

None

Weak

Moderate

Strong

Natural Gas Price Assumption**

Low

1

2

3

4

Mid

5

6

7

8

High

9

10

11

12

*The weak climate policy is a policy that imposes a cost of reducing CO2 emissions that is roughly
half the cost of the Lieberman-McCain bill under mid level natural gas prices.  The moderate policy
is the Lieberman McCain caps and the strong policy is one that imposes a CO2 price that is roughly
50% higher.
**The mid natural gas price path is based on AEO2007 forecasts.  The low price projection represents a
33% shift downward in the intercept of the natural gas supply curve based on AEO2007 forecasts. 
The high price projection assumes a 33% upward shift in the intercept of the natural gas supply curve
based on the AEO2007 forecasts.

We assume the investor can choose among several different options for new generation including sub-critical pulverized coal, supercritical pulverized coal, ultra-supercritical pulverized coal, integrated gasification combined cycle (IGCC) and natural gas combined cycle (NGCC). 

Summary/Accomplishments (Outputs/Outcomes):

We summarize our accomplishments and research results for each of the two major projects in turn.

A. Dynamic Adjustment and Symmetric Cost Management

We investigate the implications of a cost-management strategy from two perspectives. First, we examine the historical performance of other cap and trade programs. We find a significant problem with the proposal to use a one-sided safety valve approach, stemming from the asymmetry of an instrument that militates only against a price increase. The most important examples of price volatility in cap and trade programs have occurred not when prices spiked, but instead when allowance prices fell below their expected values. For example, in the case of SO2 emission trading, the inability of the trading program to adjust to the fall in allowance prices led to welfare losses of between $1.5 and $8 billion dollars per year, measured against congressional intent.

Second, we examine the incentives that a cost-management policy could have on investors. Using a Taylor series expansion around the mid case for uncertain future natural gas prices, we measure the distribution of key variables of interest in a detailed electricity market model, where we show that a high-side safety valve can be expected to increase emissions and decrease investment in nonemitting technologies, relative to the absence of a safety valve altogether. The high-side safety valve leads to the expectation of greater emissions than in the no safety valve case because with some probability the safety valve will be triggered, thereby placing extra allowances on the market. As a consequence the allowance price and electricity price are lower. All variables except welfare are normalized using the no safety valve case as a numeraire (the value is set equal to one). For welfare, the difference between the no safety valve case and the mid case in the deterministic model is normalized as a numeraire because only changes in welfare have economic relevance. A potentially important unintentional result is that the lower expected allowance price leads to lower expected payoffs to investment in renewable technologies. Consequently we see a decline in renewable generation.

Nonetheless, the one-sided safety valve improves welfare relative to the baseline. Welfare is calculated as the sum of changes in producer and consumer surplus, plus the change in environmental benefits associated with changes in emissions relative to the emission quantity target. The change in welfare for each case is measured relative to the mid gas case. The greatest improvement comes from adding a safety valve in the high gas price case.

In this analytical framework, we examine a symmetric safety valve, which is a price stabilization policy that addresses both unanticipated spikes or drops in the allowance price. When allowance price falls below the safety valve floor, the symmetric safety valve contracts the number of allowances issued in the market. In simulation analysis with uncertain natural gas prices, the symmetric safety valve returns the expected value for these and other key parameters to near their levels in the absence of a safety valve. In addition, although a high-side safety valve improves welfare, a symmetric safety valve improves welfare even further. In summary, we find a symmetric safety valve can improve the performance of allowance trading programs, improve welfare, and may help overcome political objections from environmental advocates who have opposed the use of a safety valve.

B. Incentives for Investment and Innovation

This project examines investment decisions in the context of a market problem with uncertainty with regard to market variables (natural gas prices), policy variables (climate policy as represented through the price of CO2 allowances), and numerous other variables whose value depends on the state of the world. We examine the interaction of technology policy (NSPS) and market-based policy (cap and trade), and the decisions facing an investor in CCS technology.

In this analysis we propose a flexible emissions technology standard for CO2 that can serve several purposes. The standard involves combining a new source performance standard with an emissions surcharge that may or may not be used to subsidize future retrofits of a fossil generating plant with CCS. Introducing such a policy as a prelude to federal action on cap and trade could help to provide early incentives for investment in clean coal technology or for the retrofit of new generators with CCS.  A flexible standard should be more acceptable to electricity generators as it provides a way to postpone the high costs of installing CCS.  Such a policy could help to minimize the potential negative effects of an inflexible climate policy on utility willingness to invest while helping to make sure that investment plans are more consistent with future need to control CO2 emissions.

Our results show that the strict technology policy can delay investment and can also affect the choice of generation technology.  In the cases of scenarios 1 and 5, which assume no CO2 cap and trade policy, investment in new technology does not occur and the incremental baseload power supply is assumed to come from the existing generators in the region, which are largely coal fired. There is an important limitation to this result, which makes the result preliminary. This analysis assumes the “nth of a kind” CCS cost is available immediately. Follow-on work that will be completed before a final manuscript is released is imposing a timeline over which CCS technology becomes available.

The modeling has highlighted limitations that we plan to continue to address through another revision to the modeling strategy and another round of simulations. However, several observations are possible based on our preliminary results. 

In summary, our findings indicate that a flexible standard would help to accelerate installation of new investment in general, compared to the implementation of a technology standard. The reason is that a technology standard raises the cost of investment and provides an incentive to extend the life of existing facilities. Further, we find a flexible standard helps accelerate the installation of CCS technology when uncontrolled coal technologies would have otherwise been chosen. Adding the option for an escrow fund can accelerate the timing of the CCS retrofit, but it can also lead to the choice of a different (controlled) generation technology when doing so increases profits. Uncertainty on the future carbon policies seems to delay investment, and this may be remedied by the technology standard but only when that standard includes flexibility.

Conclusions:

This research addresses the role of uncertainty as it interacts with the design of market-based regulations. Two major projects were organized.

One project looked at the role of dynamic decision rules in a cap-and-trade program, specifically, the role of cost management and how program goals can respond to new information about the cost of the program. Many observers have criticized the suggestion from the economics literature about the potential role of a safety valve due to the concern that it might undermine the environmental targets of the program. We obtain this result in analytical and simulation modeling. We find emissions are higher and investments in new technology are lower as a result of the high-side safety valve. The reduction in investments initiates a cascade of consequences, as there is less learning as a result of the decline in investment, so the costs of renewable technologies remain above their levels in the absence of the safety valve.

However, the unintended consequences are remedied when the cost management is implemented with a symmetric instrument. In this case, emissions fall back to virtually the same level as in the absence of a safety valve, and renewable investments increase to above their level in the absence of a safety valve. Not only do measures of interest to environmental advocates return to their intended levels, but welfare improves even further than in the case with only a high-side safety valve. Also, electricity price and allowance price return to nearly the same level as in the absence of the safety valve.

This ambitious research encountered a number of analytical and modeling challenges. Manuscripts have presented before various conference presentations and work continues with additional funding. Nonetheless, the work and various presentations have already had an influence on policy at the federal and state level. For example, a briefing to legislative staff related to these issues led to the introduction of a measure in proposed federal policy that would provide symmetric cost management. Also, the Regional Greenhouse Gas Initiative design for an auction to allocate emission allowances draws on this research in the decision to set a reserve price in the auction. The reserve price serves as a low-side cost management mechanism. If the reserve price is not met, a given lot of emission allowances would not be sold in that auction, thereby reducing the quantity of allowances and indirectly pushing up the allowance price.

The second project addresses the interaction of market-based and technology-forcing regulations, especially within the context of uncertainty and the need to promote new investment and technology innovation. We explore the introduction of flexibility in a technology-forcing policy that can help overcome the negative unintended consequences often associated with such a policy. The project combines three sophisticated modeling in another ambitious research effort that is still ongoing. We find that the introduction of flexibility can help accelerate installation of new investment in general, compared to the implementation of a technology standard. Further, we find a flexible standard helps accelerate the installation of CCS technology when uncontrolled coal technologies would have otherwise been chosen.

Journal Articles:

No journal articles submitted with this report: View all 7 publications for this project

Supplemental Keywords:

ENVIRONMENTAL PROTECTION AGENCY, Economic, Social, & Behavioral Science Research Program, Scientific Discipline, RFA, Social Science, decision-making, Economics & Decision Making, Reinvention, Ecology and Ecosystems, Economics, Market mechanisms, compliance costs, allowance allocation, decision analysis, incentives, market-based mechanisms, decision making, emissions trading, allowance market performance, cap and trade systems, market incentives, effects of policy instruments, environmental economics, compliance behavior, emission fees, incentive based environmental policy, environmental decision making

Progress and Final Reports:
2004 Progress Report
2005 Progress Report
Original Abstract

Top of page

The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.

Local Navigation

Jump to main content.