Final Report: Introducing Markets for Green Products: Product Demand, Environmental Quality & Economic WelfareEPA Grant Number: R828626
Title: Introducing Markets for Green Products: Product Demand, Environmental Quality & Economic Welfare
Investigators: Moore, Michael R. , Kotchen, Matthew J.
Institution: University of Michigan
EPA Project Officer: Hahn, Intaek
Project Period: January 1, 2001 through December 1, 2002 (Extended to June 30, 2003)
Project Amount: $68,042
RFA: Market Mechanisms and Incentives for Environmental Management (2000) RFA Text | Recipients Lists
Research Category: Economics and Decision Sciences
The objectives of this research project were to: (1) construct a theoretical model that formalizes consumer behavior before and after introduction of a market for a green product; (2) test predictions from the model in an empirical application involving the introduction of a market for green electricity; (3) estimate the economic benefits from the green market, and assess the potential scale of the market using revealed-preference data; and (4) evaluate markets for green products, and green electricity in particular, as an information-based approach to environmental policy.
Markets for green products are forming to replace many pollution-generating products in the economy. During this research project, we conducted a theoretical and empirical investigation of consumer behavior in a market for a green product.
We developed a consumer-choice model that moves from a baseline of a market for a "conventional" product to a market that offers the opportunity to purchase a green version of the product at a higher price. Two types of individuals operate in the model: those with a taste for environmental quality ("conservationists"), and those without such a taste ("nonconservationists"). Two forms of behavior are considered. First, individuals who care about environmental quality may voluntarily restrain their consumption of goods that generate a negative externality. Second, individuals may choose to pay a voluntary price premium for goods that are more "environmentally friendly." The model generates four testable propositions on electricity consumption that investigate the relationship between voluntary restraint and a voluntary price premium.
The model uses the example of conventional electricity and green electricity, and begins from a setting in which only conventional electricity is available. Individuals who are nonconservationists derive utility from the consumption of conventional electricity and a numeraire consumption good. Conservationists, in contrast, have an additional term in their utility function that represents disutility from pollution generated by their use of conventional electricity. The first proposition derived from this setup is the following:
Proposition 1 (Voluntary Restraint). Holding income constant, conservationists will consume less conventional electricity than nonconservationists. The model incorporates the introduction of a green electricity program. Individuals participate in the program by volunteering to pay a higher price for green electricity, with the higher price applying to all of the household's electricity consumption. The higher price is a fixed premium above the price of conventional electricity. Green electricity and conventional electricity are perfect substitutes, except for the price premium and pollution emissions. As a result, conservationists may be willing to pay the price premium for green electricity to avoid the disutility from generating pollution through their use of conventional electricity.
Both conservationists and nonconservationists also may derive utility from joining the program. This participation benefit is related to psychological or social benefits such as "the purchase of moral satisfaction" or a "warm glow" from membership. This utility comes in the form of a lump-sum benefit.
Several propositions relate to the decision of whether to purchase green electricity:
Proposition 2. If a nonconservationist household participates in the green electricity program, then: (1) the household must enjoy a lump-sum benefit from participation; and (2) after participating, the household’s electricity consumption will decline by as much as it would if there had been an increase in the price of conventional electricity equal to the premium for green electricity.
Proposition 3. If a conservationist household participates in the green electricity program, then: (1) the household's electricity consumption may increase, decrease, or remain the same; and (2) if consumption decreases, the household must enjoy a lump-sum benefit from participation.
Proposition 4. If both conservationists and nonconservationists participate in the green electricity program, then: (1) holding income constant, households of both types will consume the same amount of green electricity; and (2) conservationists will reduce their electricity consumption by less (if at all) than nonconservationists.
Propositions 1-4 are tested in an empirical study of conventional electricity consumption with introduction of a green electricity program in Traverse City, MI. The municipal utility company, Traverse City Light and Power (TCL&P), provides electrical service to approximately 7,000 residential households. In 1994, TCL&P began soliciting households to voluntarily finance a centralized wind turbine that would replace electricity generated at the local coal-fired power plant. Based on the level of interest, TCL&P constructed a wind turbine and began operating the "Green Rate" program in 1996. To participate in the program, households agree to a 3-year commitment to purchase all of their electricity at a price premium of $1.58 cents per kilowatt hour. This translates into an average residential premium of $8.40 per month, or a 24 percent increase.
The data used in the empirical analysis come from the combination of: (1) an original household survey of participants and nonparticipants in the Green Rate program; and (2) monthly panel data on electricity consumption between 1994 and 2002. Using a mail survey of residential customers, we collected data on socioeconomic characteristics, physical attributes of each residence, and household behaviors related to conservation. The sample included households that were (at the time of the survey) participants in the Green Rate program, on the waiting list, or nonparticipants. Specifically, the sample was stratified to include all 122 households that were participants, all 32 households that were on the waiting list, and a random sample of 846 households that were nonparticipants with utility records dating back to 1994. The overall response rate was 70 percent, which is relatively high.
With these data, we estimated equations to explain household electricity consumption before and after the introduction of the green electricity program. Self-reported membership in an environmental organization is used to identify conservationist households. All of the results are consistent with the theoretical predictions and provide insight into Propositions 1-4. The key results of the econometric analysis are:
· We found evidence of voluntary restraint by conservationists (households with environmental memberships). These households consume 1.6 kilowatt hours per day; on average, this is less electricity than nonconservationists. This is a difference of 8.6 percent.
· Nonconservationist households participating in the green electricity program reduce electricity use by 0.89 kilowatt hours per day, or 4.8 percent. This translates into a "voluntary green" price elasticity of -0.22, which is well within the range of estimates in the literature for the price elasticity of conventional electricity demand. Therefore, we cannot reject the hypothesis that participating nonconservationists respond as if there had only been an increase in the price of conventional electricity. In the context of the theoretical model, this result also implies that participation is motivated by a lump-sum benefit, possibly due to social and psychological benefits of program membership.
· Conservationist households participating in the program decrease electricity use by an average of 0.464 kilowatt hours per day, or 2.7 percent. Thus, conservationists-who were already exhibiting voluntary restraint-reduce electricity consumption even further after participating in the Green Rate program. Referring back to the second part of Proposition 3, this result provides additional evidence that a lump-sum benefit is part of the motivation for participation. Thus, both conservationist and nonconservationist households appear to purchase some type of social or psychological satisfaction by joining the program.
· The econometric model also provides insight into the relationship between participating conservationists and nonconservationists. We found that, after joining the program, nonconservationists reduce their electricity consumption more than conservationists. The difference is 0.43 kilowatt hours per day. Lastly, we found that conservationists and nonconservationists are indistinguishable with respect to green electricity consumption. According to the theoretical model, this occurs because conservationists have no reason to voluntarily restrain their consumption of green electricity, as it generates no pollution.
We applied two approaches to estimating the economic benefits of the TCL&P green electricity program: (1) a revealed-preference approach based on the theory of public-good provision through a voluntary contribution mechanism; and (2) a benefits-transfer approach that relies on estimates of the marginal damage costs of the air pollution from generating conventional electricity.
The TCL&P green electricity program uses a voluntary contribution mechanism to finance the program. A result in the economics literature shows that the sum of individual contributions to a public good serves as a lower bound estimate of the good's economic benefits. We obtain this lower bound estimate by computing aggregate contributions by subscribers to the green electricity program. The computation uses the green price premium, average daily consumption of green electricity by subscribers, and the number of subscribers. Estimated aggregate contribution-the lower bound estimate of benefits-is $11,107 per year.
The benefits-transfer approach isolates the value of the air pollution reductions from generating green electricity. The wind-generated electricity displaces coal-fired electricity in the TCL&P case. We use a TCL&P estimate that this reduces carbon dioxide emissions by 2 million pounds per year. Using estimates of pollution emissions from coal-fired electricity, we compute related emission reductions of sulfur dioxide, particulates, nitrogen oxides, lead, methane, and nitrous oxide. Next, dollar estimates of marginal damage costs of the pollutants are transferred from two sources: (1) the literature on the environmental costs of electricity generation; and (2) the literature on climate change. Two estimates are computed for most of the pollutants: a low-range estimate based on pollution from a power plant in a rural setting, and a high-range estimate based on a power plant in a metropolitan-fringe setting. Summed across pollutants, the benefit estimates range from $2,863 to $5,188 per year.
To evaluate the potential scale of the green electricity market, we applied findings from Objective 2 on the difference between consumption of conventional and green electricity. Conservationists reduced electricity consumption by 2.7 percent after participating in the TCL&P program, while nonconservationists reduced consumption by 4.8 percent. Households made these reductions in response to the higher price for green electricity. The nonconservationists' response was similar in magnitude to estimates in the literature on household responsiveness to price increases in conventional electricity. These estimates of the "voluntary price" elasticity are important for making forecasts about the potential scale of markets for green electricity.
The final objective assessed implications of the research results for environmental policy. The research uncovered two motives for subscribing to a green electricity program: (1) avoid the disutility or guilt from household pollution emissions; and (2) receive the psychological or social benefits of program membership. Voluntary programs oriented to consumers, consequently, can achieve environmental-quality improvements without relying on mandatory regulations. In this way, green electricity programs can complement the Clean Air Act in reducing pollution emissions.
Two caveats, however, deserve mention concerning green electricity programs and environmental policy. First, participation rates in the programs are relatively low. In the TCL&P program, 154 households have joined the program as active subscribers or waiting list members, yet TCL&P serves more than 7,000 residential customers. Nonparticipants may be "free riding" on participants' contributions. Simultaneously, we found that price affects the subscription rate. In this case, the price premium combined with electricity used to establish an annual green contribution (in dollars per year) on a household basis. This is the measure of the household's price of subscribing to the program. The results showed that a $5.77 decrease in annual contribution increased the probability of subscribing by 0.04 percent. This is a relatively inelastic response to price. Nevertheless, it suggests that government subsidy of green electricity programs would increase participation rates if the subsidy translated into lower premiums for green electricity.
Second, cap-and-trade programs for individual pollutants may neutralize the impact of green electricity programs. For example, under the federal Clean Air Act, a cap is set on aggregate emissions of sulfur dioxide from electric utilities. This creates an unanticipated circumstance. When a utility offers a green electricity program to displace conventional electricity generation, it may simply sell the reductions in sulfur-dioxide emissions on the market. Consequently, a reduction in aggregate sulfur-dioxide emissions may not occur as a result of the green electricity program. Whether electric utilities are operating in this manner is an empirical question, yet it stands to reason that they would partake in profit-yielding opportunities. As green electricity programs continue to grow-and as cap-and-trade programs are developed for other pollutants-policymakers should consider addressing this apparent "loophole" in the law.
We developed and tested a set of hypotheses related to voluntary restraint in electricity use and its relation to a voluntary price premium for green electricity. Households identified as conservationists consume 9 percent less conventional electricity on average, than nonconservationist households. Upon participating in the program (and paying the premium for green electricity), both conservationists and nonconservationists reduced electricity consumption. On average, conservationists reduced consumption by 2.7 percent, and nonconservationists reduced consumption by 4.8 percent. We also found evidence of a lump-sum benefit from participating in the green electricity program. We interpreted this benefit as a psychological or social benefit that occurs independently of electricity consumption. Overall, the results are wholly consistent with the model of conservation behavior, as none of the theoretical predictions can be rejected.
We applied a revealed-preference approach and a benefits-transfer approach to estimating the economic benefits of the TCL&P green electricity program. The first approach yields a lower-bound estimate of $11,107 per year. The second approach yields estimates in the range of $2,863 to $5,188 per year. The two approaches do not establish an "apples-to-apples" comparison. Estimates from the benefits-transfer approach largely relate to health benefits. In contrast, the actual contributions may represent several sources of benefits, including health benefits, environmental benefits unrelated to health, and "warm-glow" benefits. This may explain the larger estimate from the revealed-preference approach.
Finally, although the voluntary emission reductions of green electricity programs may complement the Clean Air Act, participation rates in such programs are relatively low, as "free riding" may be pervasive. The participation rates also are sensitive to price, which creates a potential role for government subsidy to lower green price premiums. Cap-and-trade programs under the Clean Air Act create a final policy issue. For example, a utility may simply sell sulfur dioxide emission reductions from its green electricity program on the sulfur dioxide market, resulting in no net decrease in aggregate emissions of the pollutant. Thus, the relationship between green electricity programs and cap-and-trade programs needs further attention.