Impacts of Adjusting Environmental Regulations When Enforcement Authority Is Diffuse: Confined Animal Feeding Operations and Environmental Quality

Under the cooperative federalism incorporated in many U.S. environmental statutes, federal agencies commission states to monitor and enforce federal laws. As with the federal government, many state agencies administering environmental legislation are responsible for setting standards and monitoring compliance but lack the authority to prosecute violations. Instead, they must refer violations to the state attorney general (USEPA 1999). Models incorporating this diffusion of regulatory authority have tended to focus on either optimal enforcement given exogenous regulatory standards (Garvie and Keeler; Harford) or optimal standards given exogenous levels of enforcement (Viscusi and Zeckhauser). These models specify an objective function of the relevant regulatory agency consistent with the analytical framework in which the model is developed. There is, however, no consensus on the objective function from which regulatory decisions are actually derived.

Agency objective functions within the literature range from minimizing levels of noncompliance to maximizing social welfare, agency budgets or agency political support (Viscusi and Zeckhauser; Jones and Scotchmer; Niskanen; Peltzman). While the impetus for establishing regulatory parameters may be difficult to determine, there is no disputing that these parameters are routinely adjusted either through legislation or the issuance of rules.

This paper examines the potential repercussions of adjusting the parameters of regulations on environmental quality, irrespective of the agency's objective function. The derived results suggest that the direction of changes in environmental quality induced by adjustments in policy parameters may have counter-intuitive signs; namely, increasing the number of regulated firms could actually diminish environmental quality. These results are derived in the context of proposed changes to the regulations governing confined animal feeding operations (CAFOs).

In the sections that follow, the basic model is developed and used to explore how changes in regulatory parameters affect the environmental protection actions of a set of homogeneous firms. These effects are used to project changes in environmental quality induced by revised regulatory parameters. The analysis is extended to situations in which the agency responsible for prosecuting violations does not endorse changes to regulatory parameters. The paper concludes with a discussion of the policy implications of the results in the context of CAFO regulations.

Basic Model

Three fundamental questions help define the laws and regulations governing civil societies: “What?”, “Who?”, and “How?”. Within the context of this paper, “What?” refers to the particular action(s) within the purview of the statute; “Who?” refers to the actors—individuals or collectives—whose actions are subject to the statute; and “How?” refers to the administrative structure established to foster compliance. Over time, legislative bodies and regulatory agencies may adjust parameters that define the What, Who, and How of a statute, expanding or curtailing the purview of its authority. The model developed in this section is designed to account for changes in three regulatory parameters: (a) compliance standards; (b) the size, or scope, of the regulated population; and (c) the magnitude and diffusion of resources available to ensure compliance with the regulation.¹

Four entities are represented in the model: a group of n homogeneous firms, an environmental agency responsible for setting policy parameters and compliance monitoring, an agency responsible for prosecuting violations, and the courts. Firms seek to minimize the expected costs of complying with regulations set by the environmental agency. These costs are a function of effort exerted to comply with the regulation, C(x), and the expected value of sanctions incurred for noncompliance. C(x) is assumed to be increasing and convex in effort.

The expected value of sanctions is the product of the joint probability of being caught and sanctioned for noncompliance, P^{c, s}, and the value of the sanction issued, F. As in equation (1), the joint probability of getting caught and sanctioned is the product of the probability of getting caught, P^c, and the conditional probability of being sanctioned if caught, P^s|c:

(1)

The probability of getting caught is under the auspices of the environmental agency in charge of setting the regulatory standard (xs) and scope (n), and monitoring for noncompliance. P^c is a function of a firm's actual effort (x) and the amount of effort required to fully comply with the regulatory standard (xs). In particular, P^c depends on how far out of compliance the firm is (xs − x). P^c is also a function of the scope of the regulations, n, and the budget for noncompliance monitoring, M.

The conditional probability of getting sanctioned when caught is under the auspices of the prosecuting agency. P^s|c is a function of x, xs, and the enforcement budget, B. Equation (2) represents these relationships:

(2)

The value of the sanction, F, determined by the courts, is modeled in two ways. First, F is considered strictly a function of how far the convicted firm is out of compliance, F ((xs − x)), denoted F^xs. The value of the sanction is then allowed to be influenced by the level of effort exerted by the prosecuting agency F ((xs − x), B), denoted F^B. Regardless of the structure of the fine, the level of the sanction is nonincreasing in effort (F_x < 0).

The firm's objective is to choose the level of effort that minimizes the expected costs of the regulation. This decision is represented by equation (3),

(3)

where Fⁱ ∈ {F^xs, F^B} and all other variables are as defined above.

The following functional relationships are assumed for the various components of equation (3).² The probability of being fined for a violation is decreasing and convex in n, so that P^c,s_n = P^c_n * P^s|c ≤ 0; and P^c,s_n,n = P^c_n,n * P^s|c ≤ 0. The probability of being fined for a violation is increasing and concave in M, so that P^c,s_M = P^c_M * P^s|c ≥ 0; and P^c,s_{M, M} = P^c_{M, M} * P^s|c ≤ 0. The firm's minimization generates a first-order condition that leads to

(4)

whose solution is x*. We can rewrite equation (4) as an implicit function, f, to get

(5)

Environmental quality, Q, is modeled as the environmental benefits of the regulation, a function of each firm's equilibrium level of effort, x*, and the scope of the regulation, n. As such, Q(x*, n) is increasing in both x* and n (Q_x ≥ 0, Q_n ≥ 0) and displays strong essentiality with respect to both variables. That is, Q(0, 0) = Q(0, n) = Q(x*, 0) = 0. Further, the environmental benefits of the regulation are assumed to follow a traditional Sigmoid curve in which Q is convex at low levels of x* and n, and concave at high levels of the variables, as shown in figure 1. Here, the function represents the gain in environmental quality due to the regulation. Full compliance (x* = xs) with the regulation by all potential polluters (N) leads to maximum gain, Q-max. Deviations away from full compliance (x* < xs) shift the environmental quality function downward. Holding the level of compliance effort constant, changes in the number of polluters subject to the regulation represent movements along a given environmental quality curve.

A Pollution Problem

The failure of the United States to achieve the water quality goals delineated by the Clean Water Act of 1972 has led the public to seek ways to redress remaining pollution problems (Centner 2003; USEPA 2000a; Jackson; Lovell and Kuch; U.S. Department of Agriculture and U.S. Environmental Protection Agency 1999). Agriculture is thought to contribute pollutants to 59% of the impaired river and stream miles in the United States with animal feedlots contributing to about 16% (USEPA 2000b, pp. 62 & 65). The major concern involves excessive amounts of nitrogen and phosphorus in animal wastes (Gburek et al.; Heathwaite, Sharpley, and Gburek; Koelsch and Losoing; Meyer and Mullinax; Sharpley). The impairment of waters by animal feeding operations (AFOs) has led policy makers to conclude that more demanding restrictions are needed (Centner 2001; General Accounting Office; Innes; Metcalfe).

The Environmental Protection Agency (EPA) estimates there are 376,000 AFOs in the United States (USEPA 2001a, p. 2984). Approximately 12,700 of these qualify as concentrated AFOs (CAFOs) that are subject to point-source discharge requirements of the Clean Water Act (U.S. Environmental Protection Agency 2001a, p. 2984). The distinction is that CAFOs are required to secure National Pollutant Discharge Elimination System (NPDES) permits while all AFOs not classified as CAFOs are governed by federal nonpoint-source pollution regulations (U.S. Code Annotated; Code of Federal Regulations). The EPA concluded that only about 2,500 CAFOs had been issued permits in 1997 (USEPA 2001a, p. 3080), meaning approximately 10,000 CAFOs may not be in compliance with federal law.

A shortage of crop land around some AFOs means that excessive amounts of manure are applied to nearby fields with resulting nutrient contamination of streams and other waters (Bosch and Napit; Gollehon et al.; Parker; Taylor). The government calculated that AFOs produce 1.3 million tons of nitrogen and 0.7 million tons of phosphorous each year (U.S. Environmental Protection Agency 2001a, p. 2986). Moving manure to fields is costly (Araji, Abdo, and Joyce) so that 90% of manure does not leave the area where it was produced (Taylor). This may mean that concentrated operations place too many nutrients in a given locale (Letson and Gollehon). The potential threat of pollution from animal waste is greatest in the southeastern and mid-Atlantic states (Letson et al.).

Given the unacceptable levels of water pollution and the perception that AFOs are contributing to the problem, the EPA advanced in 2001, a proposed rule to address the impairment of waters by CAFOs (USEPA 2001a). The proposal called for doubling or tripling the number of CAFOs subject to point-source pollution regulations and more definitively stated that CAFOs include the land area where animal waste is applied (USEPA 2001a, pp. 2985, 3009–10).

In December 2002, the EPA adopted a final rule delineating new provisions governing CAFOs (U.S. Environmental Protection Agency 2003). The EPA retreated from its proposed rule so that only an estimated 23% of additional AFOs are classified as CAFOs under the final rule (USEPA 2002). The impacts are more pronounced for veal, heifer, and poultry producers. Whereas under the previous CAFO rules, no veal or heifer production facility was classified as a CAFO, 254 facilities raising such animals are expected to be CAFOs under the final rule (USEPA 2002). The final rule is expected to result in a 190%-increase in the number of large layer-hen facilities that must secure permits and a 901%-increase in the numbers of large facilities producing broilers that must be permitted (USEPA 2002).

At first blush, the purported goal of improving water quality by regulating more CAFOs appears reasonable. By regulating more producers under state NPDES permit programs, there should be more governmental oversight of activities that may be contributing to the impairment of U.S. waters. This assumes that the new regulation will result in more AFOs complying with the permit requirements and that the governmental proscriptions governing these AFOs will cut pollution. In the absence of greater monitoring and enforcement, governmental oversight of extended numbers of firms decreases the probability that a noncomplying firm will be inspected, which might encourage noncompliance.

Changes in Number of Firms Regulated and Monitoring Efforts

In light of the expected impact of the final CAFO rule on the number of regulated firms, this section applies the basic model to relate how equilibrium environmental quality may be affected by changes in regulatory scope. Changes in environmental quality induced by increasing monitoring effort are also analyzed. The response depends on the functional forms of P^c and P^s|c. The form these functions take depends on factors related to the ability of the monitoring agency to detect a violation. Four possibilities are explored: (a) clear violation with an increase of firms regulated; (b) clear violation with an increase in monitoring; (c) uncertain detection with an increase of firms regulated; and (d) uncertain detection with an increase in monitoring.

A clear violation represents situations in which violators that are monitored will be caught. Here, for firms not in full compliance [P^c = P^c(M, n; x < xs)], the probability of getting caught is independent of the deviation of the level of effort (x) from that required by the standard (xs). Failure of CAFOs to secure a required permit, and violations of certain permit conditions could be classified as clear violations. More generally, clear violations are likely to be associated with point-source pollution where standards are based on measurable discharges or subject to design standards. Uncertain detection represents situations in which the probability of getting caught is a function of the deviation of equilibrium effort from that required by the standard. Here, firm behavior that is further out of compliance is more likely to be detected by a monitor. A discharge by a CAFO in violation of permit specifications is an example of a violation characterized by uncertain detection. More generally, uncertain detection is likely to be associated with nonpoint source pollution and performance standards that rely on the implementation of management techniques.

Scenario 1: Clear Violation with an Increase of Firms Regulated

When a firm is clearly out of regulatory compliance, an increase in the number of firms regulated leads to an ambiguous change in environmental quality, Q. In other words, bringing more firms under the purview of a regulation could reduce or enhance environmental quality, or leave it unaffected. Under this condition, as the number of firms regulated increases, each firm's equilibrium effort, x*, would decrease, as shown in equation (6) as:

(6)

The ambiguity of the change in Q arises from the countervailing forces associated with an increase in firms being regulated. With an increase in n, there is a decrease in the probability of getting monitored. In turn, this decreases the expected cost of noncompliance, thereby reducing x*. Yet the increase in n increases Q for a given level of effort. As a result, the change in Q is ambiguous.³ This ambiguity can be seen in figure 1: the increase in n increases environmental quality by moving along a given curve while the reduction in compliance effort causes the environmental quality curve to shift downward. The new equilibrium level of environmental quality may be above, below, or the same as the original.

Scenario 2: Clear Violation with an Increase in Monitoring

When a firm is clearly out of compliance with a regulation, increased monitoring that discovers violations should lead to greater environmental quality:

(7)

With an increase in agency monitoring, firms will increase their equilibrium level of effort, x*. This increase is associated with the higher expected cost of noncompliance through the probability of getting caught.⁴ This also can be seen in reference to figure 1. The increase in x* induced by greater monitoring shifts the environmental quality curve upward. As a result, holding the number of regulated firms constant leads to an increase in environmental quality.

Scenario 3: Uncertain Detection with an Increase of Firms Regulated

When it is unclear whether a firm that is out of compliance will be detected if it is monitored, an increase in the number of firms regulated leads to an ambiguous change in environmental quality. The same reasoning as in Scenario 1 applies. As the number of firms regulated increases, each firm's equilibrium effort, x*, would decrease as

(8)

Due to the uncertain detection, however, the decline in equilibrium compliance effort will be greater here than under Scenario 1.⁵

Environmental quality will increase when the elasticity of environmental quality with respect to the number of regulated firms (ξ_{Q, n}) is greater than the product of the elasticity of environmental quality with respect to compliance effort (ξ_{Q, x*}) and the elasticity of compliance effort with respect to numbers of regulated firms (ξ_{x*, n}). That is, more regulated firms will lead to an increase in environmental quality if and only if equation (9) holds:

(9)

For the properties of Q(x*, n), ξ_{Q, n} > 1 when n is small, and ξ _{Q, n} is decreasing in n. Similarly, ξ_{Q, x*} > 1 when x* is small, and ξ_{Q, x*} is decreasing in x*. As a result, equation (9) need not always hold. An increase in the number of regulated firms could actually lead to a decrease in the environmental benefits of the regulation. This is most likely to happen for regulations covering large numbers of firms characterized by low equilibrium abatement effort—the situation that, arguably, characterizes current CAFO regulations.

Scenario 4: Uncertain Detection with an Increase in Monitoring

When it is unclear whether a firm out of compliance will be detected, an increase in monitoring effort should lead to greater environmental quality:

(10)

The same reasoning as in Scenario 2 applies. The increase of x* is associated with the increased expected value of noncompliance through the probability of getting caught.⁶

Impact of Diffuse Enforcement Authority

The diffusion of enforcement authority across state and federal agencies can result in significant coordination costs that impede the effective execution of an enforcement program and compromise the ability of regulations to enhance and preserve environmental quality. In addition to coordination costs, political and economic considerations may influence the willingness of the prosecuting agency to take action against certain violations (Riesel, Helland). In this section, the potential impact of a prosecuting agency's unwillingness to enforce expansion of the number of regulated firms (n) or an increase in the regulatory standard (xs) is explored within the context of the model.

The prosecuting agency's willingness to enforce new regulations is reflected in the conditional probability of being sanctioned, P^s|c. Expansion the number of firms within the purview of the regulation, n, affects the probability of being caught, but not the conditional probability of being sanctioned. As discussed above, increasing n will decrease the probability of being caught, leading to reduced equilibrium effort. However, if the prosecuting agency is unwilling to prosecute violations, the newly regulated firms, without fear of effective prosecution, will not expend any effort and environmental quality will diminish. Over time, the monitoring agency will likely cease monitoring the newly regulated firms and equilibrium effort and environmental quality would return to their original levels. That is, after a period of adjustment, increasing the number of regulated firms with an unwilling prosecutor should have no effect on long-run environmental quality.

If the environmental agency raises the regulatory standard, both the prosecuting agency's attitude toward the new standard and the manner in which fines are set by the court affect the impact on environmental quality. For regulations associated with clear violation, the probability of being caught will be unaffected by the increase in xs (P^c_xs = 0). Furthermore, if the prosecuting agency is unwilling to enforce violations of the new standard, the conditional probability of being sanctioned will also be unaffected (P^s|c_xs = 0). However, if the courts set fines with regard to how far out of compliance with the new standard a violation is, that is, if F_xs ≥ 0, then equilibrium effort and environmental quality will still increase:⁷

(11)

Greater effort results from an increase in the expected value of the sanction, leading to Q_xs > 0. However, if the fine is independent of how far out of compliance a violation is, then F_xs = 0 and, consequently, equilibrium effort and environmental quality will be unaffected by raising the standard.

When the regulation pertains to violations that are difficult to detect, increasing the standard should raise the probability of being caught, thereby increasing equilibrium effort and environmental quality irrespective of the attitude of the prosecuting agency or the manner in which fines are set. However, equilibrium effort and environmental quality will increase more when the prosecuting agency is willing to enforce the new regulations (P^s|c_xs > 0) and fines are set as a function of how far out of compliance a violation is (F_xs > 0).

Discussion

Governments want to control environmental pollutants from AFOs. To achieve this, both state and federal governments are expanding their regulatory proscriptions. The federal government has a new rule that expands the number of CAFOs required to secure state permits pursuant to delegated powers of the federal NPDES program. But can the new rule be expected to reduce contaminants entering our water bodies? This depends on the ability of a monitoring agency to detect noncompliance and the enforcement efforts related to sanctions. By analyzing different scenarios, it is shown that under some conditions, increasing the number of regulated firms might lead to greater pollution.

Scenarios 1 and 3 show increases in the number of CAFOs regulated may be expected to lead to an ambiguous change in environmental quality. Because increasing the number of regulated firms makes it less likely that an individual firm will be monitored, the expected cost of noncompliance decreases. In turn, under the new rule it may be expected that an increased number of CAFOs would secure permits. This should increase environmental quality. Taken together, these countervailing forces mean it is not clear that the new rule will enhance environmental quality.

Scenarios 2 and 4 show that increases in monitoring may be expected to lead to an improvement in environmental quality. Because of increased monitoring, CAFOs failing to comply with the regulations are more likely to get caught. The higher expected cost of noncompliance should lead CAFOs to increase their equilibrium effort to avoid being found out of compliance. This would enhance environmental quality. Thus, additional funding for monitoring, which is absent from the Final Rule, might be more effective in helping achieve water quality goals.

Data from the EPA suggest that nearly 80% of CAFOs have failed to timely secure required permits. Moreover, informal communications by states note that they lack the personnel and financial resources to monitor and prosecute violations of CAFO regulations (USEPA 2001b). These data suggest that existing monitoring efforts are inadequate. The EPA estimates the final rule will cost an additional $326 million annually for the newly regulated CAFOs (USEPA 2002). States are expected to incur an additional $9 million in regulating these CAFOs (USEPA 2002). The magnitude of these expected costs implies that the benefits accruing to CAFOs foregoing compliance are considerable.

Given the described scenarios, current enforcement conditions, and calculated compliance costs, it is unclear that the new rule constitutes an optimal response to the environmental problem of pollution by AFOs. The analyses show that, when the numbers of regulated firms expands under an existing broad regulation characterized by low compliance, environmental quality may decline. While the new rule would expand the number of regulated CAFOs, it may or may not enhance environmental quality given the compliance costs. In some instances it may be possible to estimate the expected change in environmental quality due to an expansion of the regulatory scope. Efforts to do this prior to setting new regulatory rules may lead to more optimal regulatory statutes.