Transaction Costs of Policies to Reduce Agricultural Phosphorous Pollution in the Minnesota River.(Statistical Data Included)

I. INTRODUCTION

Nonpoint source pollution control is a difficult problem in theory and in practice. The inability to observe emissions at reasonable cost means that emission taxes and standards are difficult if not impossible to implement. In the case of agricultural nonpoint source (NPS) pollution, we observe programs or policies based on education, cost sharing/technical assistance, land retirement, and conservation compliance while economic theory indicates that incentives such as input taxes are generally more efficient. Yet this conclusion is based on cost estimates that exclude transaction costs.

Inclusion of transaction costs in policy evaluation is important for three reasons:

1. transaction costs may affect which policy alternative attains an environmental goal at least cost;

2. they will reduce the amount of abatement that is optimal from the point of view of society; and

3. their inclusion may lead to the design of policies and institutional arrangements which lower transaction costs.

Transaction costs (including administrative costs) might be particularly high for nonpoint source pollution due to the high cost of monitoring emissions and the large number of polluters. These higher costs may be one reason why point sources (PS) have been emphasized in water quality legislation.

Severe water quality problems exist in the Minnesota River that make it unswimmable, unfishable, and even uncanoeable in the Twin Cities metropolitan area (MPCA 1994). Water quality problems in the Minnesota River include phosphorus (P) and nitrogen (N) enrichment, bacteria, suspended solids, and ammonia. The Environmental Protection Agency (EPA) requires a 40% reduction in biochemical oxygen demand (BOD) between Shakopee and the mouth of the Minnesota River. Models indicate that this would sustain aquatic life in the Metro area. lt is estimated that it would cost the waste treatment plants $400 million to comply with these regulations (MPCA 1994).

In the Minnesota River, P has been identified as the limiting nutrient for the algal growth which causes eutrophication (MPCA 1994). Phosphorous comes from point sources such as industry and waste treatment plants, as well as nonpoint sources such as private septic systems, pastures, erosion of farmland, and urban runoff. The amount of P loading from agricultural runoff is affected by a variety of factors including the P content of the soil and the amount of soil erosion. Estimated contribution of NPS has varied from 35% of P loading in a low rainfall year (1988), to 90% in a high rainfall year (1991) (MPCA 1994).

Various policies have been proposed to solve NPS water quality problems in the Minnesota River Basin. There have already been educational efforts to reduce agricultural pollution of the Minnesota River and an expanded initiative has been proposed. Conservation tillage is being promoted to reduce erosion and is being adopted by a number of farmers. Also, there has been a proposal to expand Reinvest in Minnesota (RIM), a state program that obtains permanent easements on land. While neither a tax on fertilizer inputs, nor a requirement for conservation tillage have been proposed, they were examined in this study for comparison purposes. This study estimates the magnitude of transaction costs associated with alternative policies to reduce agricultural nonpoint source phosphorus pollution in the Minnesota River in order to: (1) determine whether transaction costs help explain the prevalence of the policies actually observed, and (2) identify the underlying factors affecting transaction costs in the case of environmental policies. Federal budget deficits, aversion to new programs that would result in increased bureaucracy, and requirements for benefit-cost analysis make this a particularly relevant policy question.

II. LITERATURE REVIEW

The concept of transaction costs was first related to environmental policy in Coase's 1960 article "The Problem of Social Cost" and is now being considered more seriously in the evaluation of environmental programs (McCann 1997). One recent emphasis has been on determining the factors influencing the level of transaction costs. The magnitude of transaction costs involved with eliminating externalities is affected by the number and diversity of agents (Oates 1986; Williamson 1985; Milgrom and Roberts 1992; Kozloff, Taff, and Want 1992), available technology (North 1990), policy under consideration (Coase 1960; Easter 1993), and the amount of abatement or the size of the transaction (Stavins 1995b). The magnitude and type of transaction costs also depend on the institutional environment (Coase 1960; North 1990; Griffin 1991; Vatn and Bromley 1994) as do externalities themselves (Vatn and Bromley 1994).

Recent literature on transaction costs and environmental policy indicates that since it is difficult to know, a priori, whether the transaction costs of market solutions will be greater or less than non-market solutions, options need to be evaluated on a case-by-case basis (Griffin 1991; Easter 1993; Stavins 1995a, 1995b). In order to be incorporated in policy evaluation, transaction costs must be measured. Wallis and North (1986) tried to estimate the transaction "sector" for the United States by classifying certain jobs and industries as being primarily associated with transactions and found that public and private sector TC increased from one-quarter to one-half of GNP in this century. Colby (1990) looked at policy-induced transaction costs in western water markets and found that transaction costs averaged 6% of the price paid by the applicant to transfer water. Howitt (1994) reported that the overhead costs incurred by the State Department of Water Resources for the California Water Bank were about 8% of the water purchase cost. Hearne and Easter (1995) estimated the gains-from-trade from the sale of water rights and found that the transaction costs involved with water transfers in Chile represented 2-11% of the transaction price. McCann (1997) analyzing Natural Resource Conservation Service cost share and technical assistance data found that transaction costs represented 38% of the total conservation cost. Boggess (1995) cites a 1992 Rand study which estimated that 88% of Superfund payments were for transaction costs, not cleanup.

The literature suggests that transaction costs of environmental policies are likely to be significant, especially in the case of NPS pollution, so they should be included in policy evaluation. Most studies of transaction costs and environmental policy to date have either compared transaction costs qualitatively (Easter 1993), used the cost savings as an upper bound on transaction costs (O'Neil 1980), arbitrarily plugged a range of transaction costs into a model (Netusil and Braden 1995), assumed transaction costs to be some constant …