Grantee Research Project Results
Final Report: An Experimental Economics Examination of Incentive Mechanisms for Reducing Ambient Water Pollution Levels from Agricultural Non-Point Source Pollution
EPA Grant Number: R830989Title: An Experimental Economics Examination of Incentive Mechanisms for Reducing Ambient Water Pollution Levels from Agricultural Non-Point Source Pollution
Investigators: Poe, Gregory , Schulze, William
Institution: Cornell University
EPA Project Officer: Hahn, Intaek
Project Period: June 1, 2003 through May 31, 2006 (Extended to May 31, 2008)
Project Amount: $279,999
RFA: Market Mechanisms and Incentives for Environmental Management (2002) RFA Text | Recipients Lists
Research Category: Environmental Justice
Objective:
Nonpoint source pollution, particularly from agriculture, is the largest remaining reason that the United States is not meeting the “fishable/swimmable” objectives of the 1972 Clean Water Act. Whereas regulatory and incentive-based approaches are well established in the theoretical environmental economics literature and in actual policy initiatives, extending these ideas to nonpoint source pollution is problematic: firm-specific emissions are difficult to measure, the relationship between emissions and ambient water pollution levels is stochastic and difficult to model, and there are adverse selection and moral hazard problems. Nevertheless, recent efforts to develop Total Maximum Daily Load (TMDL) approaches that cap combined contributions of point and nonpoint source pollution on a watershed-by-watershed basis underscores the immediate need to address these theoretical and practical problems. The critical economic policy issue remains how to create incentives for nonpoint source to cost effectively meet water quality standards.
An innovative paper by Segerson (1988) draws on the principal-agent literature to suggest that water quality standards can be achieved by charging or subsidizing polluting sources based on ambient concentrations measured at a discrete point such as a dam or a public water supply intake. Motivated by the TMDL provisions that water quality objectives may be implemented through the use of “incentive-based, non-regulatory or regulatory approaches” (US EPA 2007) this research uses experimental economic techniques to evaluate ambient-based water quality taxes and subsidies.
In Segerson’s model, ambient pollution,‘A’, is a function of firm production and abatement levels, and Ā represents a targeted ambient water quality standard. Because of uncertain environmental factors (e.g., rainfall and diffusion), the relationship between firm emissions and ambient levels necessarily contains a random component. Three categories of incentives are considered: a tax (or subsidy), ti, on each unit of pollution that exceeds (or falls short of) the targeted standard; a fixed penalty, ki, to be imposed if the ambient pollution level exceeds Ā, and a combination of ti and ki. This mechanism, T(A), can be formulated as follows.
- T(A) = ti*(A – Ā) + ki if A > Ā (i.e., each firm is taxed)
- ti*(A – Ā) if Ā < A (i.e., each firm is subsidized),
wherein the subscript ‘i’ indicates that the tax or subsidy is levied on each individual pollution source in a watershed. Note further that the per unit pollution Tax/Subsidy is constant, so that if actual pollution is three units less than the target level, each firm would receive a subsidy of ti *3. If instead, pollution exceeds the ambient water quality standards by five units, then each firm would pay a tax of ti*5, plus perhaps a fixed fine ki. In this manner ti is referred to as a linear tax, or, in the technical literature, a constant marginal tax. Appropriate values for ti and ki can be determined with knowledge of aggregate abatement cost functions, the degree of uncertainty in the relationship between emissions and ambient pollution levels, and the relevant number of polluting firms. Within this broad framework, the specific objectives and our research design were specified on Pages 13-16 of the original proposal. We paraphrase these here.
- Test, and extend, the linear Tax/Subsidy approach to evaluate the impact of allowing communication between experimental subjects and assess the responsiveness of abatement levels to changes experimental design parameters.
- Compare the performance of Segerson linear tax functions with damage–based tax functions that vary with the level of emissions.
- As an alternative to an immediate tax penalty, explore “voluntary-threat” mechanisms in which polluters are given the opportunity to meet water quality objectives voluntarily. If the voluntary approach fails, an ambient tax will be implemented in subsequent periods.
- Using the most effective mechanisms developed in #1-3 above, run parallel experiments with students and with NY dairy farm operators.
Summary/Accomplishments (Outputs/Outcomes):
The motivation for using laboratory experiments as a test-bed for ambient nonpoint pollution mechanisms is simple: there are no real world data on how firms respond to the incentives underlying these mechanisms. At a minimum, appropriately designed experiments reveal whether people identify and respond to incentives in a way consistent with economic theory. If subjects in a controlled setting fail to respond in a theoretically predicted manner, then this casts doubt on the ability of the theory to predict behavior in a naturally occurring setting. Experiments may also shed light on the appropriateness of the assumptions underlying theoretical models and given observed behavior in the laboratory.
Our research results are relevant to both economic-theoretic research and the design of water quality policy. As an example, the determination of the central role of communication in the performance of the ambient Tax/Subsidy and Voluntary Threat mechanisms suggest the need to incorporate this factor in future economic-theoretic research. Accounting for the level of communication is equally germane to actual policy design. As a caveat we note that the policy relevance of our research is tempered by the current regulatory structure in which such Tax/ Subsidy programs may be difficult to implement. Further, this laboratory-based research is most appropriately regarded as the first in a series of steps that would need to be undertaken in converting theory to actual policy. As such, we identify mechanisms that would be unlikely to work in actual field settings as well promising mechanisms that merit future research in field settings. With these caveats we organize our findings along the above four objectives.
- The economic-theoretic developments by Segerson and others are predicated on non-cooperative game theory in which each individual acts independently and seeks to maximize his/her own net returns without cooperative interactions with others. Theoretically such Tax/Subsidy mechanisms provide an incentive for each polluter to abate optimally. While consistent with the standard mechanism design theory, in actual watershed settings there are incentives and opportunities for polluters to form coalitions and agree on abatement strategies to reduce their expected tax burden.
Our research demonstrates that, holding ti and ki constant across split sample experiments, communication fundamentally affects the performance of ambient Tax/Subsidy mechanisms. When communication between experimental subjects is not allowed, appropriately designed marginal Tax/Subsidy mechanisms provide ambient levels of pollution that do not significantly differs from the target level Ā. In contrast, Fixed Penalties (i.e., by relying only on ki and setting ti=0) result in excess pollution. Combined approaches, in which both ki and ti are positive, also result in excess pollution. We attribute these latter outcomes for the Group Fine and the Combined Approach to the fact that, unlike the linear tax mechanism, firms do not face the correct marginal incentives when ambient pollution is far from the pollution target. In contrast, when communication is allowed between polluters, coordination of abatement levels serve to render the linear Tax /Subsidy instrument ineffective while increasing the efficacy of the Group Fine and Combined mechanisms. Specifically, the linear Tax/Subsidy instrument results in near zero emissions (i.e., abatement is extremely excessive, falling far below Ā) and thus requires large government subsidies. Hence, in terms of the basic Segerson Tax/Subsidy framework there is no dominant mechanism that is robust to changes in communication levels.
Seeking a mechanism that is less sensitive to the level of communication, we explored a tax-only variant of the Segerson model. This mechanism is asymmetric in the sense that linear taxes are charged if ambient levels exceed AT < Ā, but no subsidies are provided if the ambient level falls below the tax threshold AT. Overall, our results indicate that the most efficient ambient mechanism for achieving Ā, allowing for both conditions of communication and non-communication, is to set the threshold slightly less than Ā, with the degree of deviation between AT and Ā depending on the level of uncertainty in the Ā function. Suter et al. (2008) demonstrate that communication can effectively be harnessed in that encouraging communication can achieve Ā at lower marginal tax rates. This would reduce the burden of taxation, and would also lend a coordinating role for extension and outreach in pollution policy design.
Beyond this key result, research reported in Poe et al. (2004) and Suter et al. (2008) demonstrate that participants react systematically and in expected directions to changes in experimental parameters, lending credence to employing a laboratory experimental approach to test these mechanisms.
In most of the experiments we conducted, polluters were homogeneous in the sense that their production, endowments and abatement cost structures were identical. We also conducted tax-only experiments in settings where critical parameters varied across subjects. In the aggregate, our above results hold, in terms of achieving the desired standard Ā in heterogeneous situations. However, in heterogeneous setting our results also suggest that adopting a uniform marginal tax has potential impact on industry structure. Notably, small firms in heterogeneous settings designed to mimic New York’s dairy industry were driven to bankruptcy, a concern for policy design.
- In Suter et al. (2008) we compare a Tax-only version of the Segerson tax mechanism identified above in which ti is constant with an alternative “damage based” tax suggested in Hansen (1998) that varies with the level of realized A. The intuitive notion for this type of tax is that an additional unit of ambient pollution when existing levels of pollution are low creates a different amount of additional damages than an additional unit of pollution when existing level of pollution are high. Hence polluters pay taxes consistent with damages created by their emissions. Our results suggest that for most intents and purposes the constant marginal tax and the damage–based tax provide similar levels of abatement. Appropriately designed, each could achieve Ā, and hence there is not a preferred method between the constant marginal and damage based tax functions.
- Imposing ambient pollution taxes on agriculture will likely meet with substantial political resistance. Recognizing this impediment Segerson and Wu (2006) proposed a voluntary-threat approach to nonpoint source water pollution control. In such a mechanism a group of polluters is given the opportunity to meet a pollution standard (such as Ā) voluntarily, under threat of a mandatory tax policy to be implemented in the case of voluntary noncompliance. In the Segerson and Wu paper the threatened tax policy was a priori, or “exogenously”, specified. In Suter et al. (2009) we introduce an alternative “Endogenous Tax” threat in which the threatened tax policy is conditioned on the level of noncompliance in the voluntary stage. Experimental results indicate that neither the exogenous or endogenous voluntary-threat approaches typically lead to voluntary compliance unless communication occurs or the incentives for compliance are high. Despite this negative result, the efficiency levels of the Voluntary-Threat mechanisms are relatively high, enough so that we feel that such mechanisms merit further consideration in future research.
- Building on the Tax Only experiments reported above, in Poe et al. 2009, we conduct split sample experiments on separate participant pools composed of undergraduate students and NY dairy farmers. The experimental design incorporated firm-level heterogeneity calibrated to reflect small, medium and large dairy herds in NY. The results from these experimental sessions show that group level outcomes are consistent with Ā across treatments. There were, were however, differences in how different herd size groups behaved across subjects. Overall, while some differences are observed, our research suggests that actual NY dairy farmers react in a like manner to incentive programs in experimental economic laboratory settings as do students. As in #3 above, we identify such additional laboratory and field testing as an area of productive future research should policy interest be strong enough in this category of mechanisms.
References:
Journal Articles on this Report : 4 Displayed | Download in RIS Format
Other project views: | All 16 publications | 4 publications in selected types | All 4 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Poe GL, Schulze WD, Segerson K, Suter JF, Vossler CA. Exploring the performance of ambient-based policy instruments when nonpoint source polluters can cooperate. American Journal of Agricultural Economics 2004;86(5):1203-1210. |
R830989 (2003) R830989 (2006) R830989 (Final) |
Exit |
|
Suter JF, Vossler CA, Poe GL, Segerson K. Experiments on damage-based ambient taxes for nonpoint source polluters. American Journal of Agricultural Economics 2008;90(1):86-102. |
R830989 (2006) R830989 (Final) |
Exit |
|
Suter JF, Vossler CA, Poe GL. Ambient-based pollution mechanisms: a comparison of homogeneous and heterogeneous groups of emitters. Ecological Economics 2009;68(6):1883-1892. |
R830989 (Final) |
Exit Exit Exit |
|
Vossler CA, Poe GL, Schulze WD, Segerson K. Communication and incentive mechanisms based on group performance: an experimental study of nonpoint pollution control. Economic Inquiry 2006;44(4):599-613. (Awarded Editor's Choice Article-Economic Inquiry 2006). |
R830989 (2003) R830989 (2006) R830989 (Final) |
Exit |
Supplemental Keywords:
public policy, decision making, watersheds, agriculture, effluent, social science, NY, RFA, ENVIRONMENTAL PROTECTION AGENCY, Economic, Social, & Behavioral Science Research Program, Scientific Discipline, Reinvention, Economics, decision-making, Ecological Risk Assessment, Ecology and Ecosystems, Market mechanisms, Social Science, Economics & Decision Making, compliance behavior, effects of policy instruments, market-based mechanisms, nonpoint source pollution trading, nonpoint source pollution, impact of federal policy instruments, policy making, economic research, policy incentives, watershed, decision making, Clean Water Act, incentives, economic incentives, environmental Compliance, environmental policy, water pollution, agriculture, nonpoint source pollution (NPSP), voluntary programsProgress and Final Reports:
Original AbstractThe 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.