Short-run and long-run implications of environmental regulation on financial performance.

I. INTRODUCTION

The Porter hypothesis asserts that properly designed environmental regulation motivates firms to innovate, which ultimately improves financial performance (Porter 1990, 1991; Porter and van der Linde 1995; van der Linde 1993). Many economists criticize this claim, arguing that firms voluntarily seek opportunities to improve financial performance regardless of regulation (Palmer, Oates, and Portney 1995). In particular, critics argue that environmental regulation undermines firms' abilities to pursue opportunities to improve financial performance. Although these arguments generally focus on the effect of regulation on long-run financial performance, they also apply to the effect of regulation on short-run financial performance. (1)

Porter and van der Linde (1995) introduce their argument by indirectly attributing competitiveness to financial performance in the context of lower costs and higher revenues:

Competitiveness at the industry level arises from superior productivity, either in terms of lower costs than rivals or the ability to offer products with superior value that justify a premium price, (pp. 97-98)

Jaffe and Palmer (1997) describe this focus on financial performance as the "strong" version of the Porter hypothesis:

The shock of a new regulation may therefore induce them [firms] to broaden their thinking and to find new products or processes that both comply with the regulation and increase profits, (p. 610)

Our analysis tests the "strong" version of the Porter hypothesis: tighter regulation improves the financial performance of individual firms. By-testing this hypothesis, we contribute to the economic literature that assesses the effects of environmental regulation on various aspects including firms' competitiveness, innovation activities, employment, productivity, investment, location decisions, and financial performance.

Previous empirical studies that examine consequences of environmental regulation generally do not examine both short-run and long-run consequences. In contrast, our study empirically examines both the short-run and long-run effects of environmental regulation on financial performance. In particular, we test the Porter hypothesis in terms of both short-run and long-run effects of Clean Water Act regulation (hereafter "clean water regulation") on the financial performance of publicly owned firms in the chemical manufacturing industries. To measure financial performance, we use return on sales (i.e., profits divided by sales), which captures profitability. As our measure of clean water regulation, we use permitted wastewater discharge limits, which are imposed on individual facilities. To strengthen our analysis, our study draws upon a panel data set. Thus, we are able to control more completely for heterogeneity across firms and exploit both interfirm and intrafirm variation.

Our results indicate that tighter clean water regulation (i.e., lower permitted discharge limits) improves financial performance in both the short run and long run with a stronger effect in the long run. In particular, return on sales increases by 1.5% in the short run and 4.5% in the long run.

II. RELATIONSHIP BETWEEN ENVIRONMENTAL REGULATION AND FINANCIAL PERFORMANCE

A. Theoretical Literature

To guide our empirical analysis, we first assess two conflicting theoretical arguments exploring the effect of environmental regulation on financial performance. Porter and van der Linde (1995) argue that properly designed and implemented environmental regulation removes organizational inertia and ultimately improves financial performance. Innovation and improved resource productivity are the mechanisms through which this relationship unfolds. As long as firms perceive their production processes and products as elements in a dynamic setting rather than a static setting, firms seize regulation as an opportunity to invest in technologies that not only minimize strains on the environment but also maximize the efficiency of production processes and/or improve the quality of products. The result is decreased production costs and/or increased revenues. Porter and van der Linde (1995) support their argument with a collection of case studies in which stringent environmental regulation improves polluting firms' financial performance. This argument has become known as the Porter hypothesis.

Porter and van der Linde (1995) do not specifically delineate their argument according to short-run and long-run financial performance. However, the authors consistently imply that the positive effect of regulation on financial performance is stronger in the long run than in the short run. In fact, Porter and van der Linde (1995, 100) claim that the innovation to be prompted by environmental regulation "cannot always offset the cost of compliance, especially in the short term before learning can reduce the cost of innovation-based solutions." Given that innovation is the crux of Porter and van der Linde's (1995) argument regarding the positive effect of regulation on financial performance, this statement reveals a long-run context for their argument. Thus, our analysis is less likely to reject the Porter hypothesis in the long run. (2)

Palmer, Oates, and Portney (1995) question the validity of the Porter hypothesis. In particular, they reject Porter and van der Linde's (1995) assertion that environmental regulation removes organizational inertia by providing firms with information and incentives that competitive markets somehow fail to provide. Instead, Palmer, Oates, and Portney (1995) posit that firms voluntarily seek profit-increasing opportunities regardless of regulation. Rather than a catalyst, environmental regulation serves only to constrain firms' abilities to pursue profit-increasing opportunities. As one specific consequence, firms facing more stringent regulation incur higher treatment costs. As the most general consequence, firms facing more stringent regulation are required to commit greater amounts of resources to uses that are neither productive nor profit-increasing. Hereafter, we identify this perspective as the "costly regulation hypothesis."

Although Palmer, Oates, and Portney (1995) imply that the constraining effect of regulation is more binding in the short run than in the long run because firms possess more time to accommodate the imposed regulation in the long run, the authors do not explicitly distinguish between the short-run and long-run effects of environmental regulation on financial performance. Given the long-run perspective of Porter and van der Linde's (1995) argument, Palmer, Oates, and Portney (1995) presumably do not make a distinction by virtue of their disagreement with Porter and van der Linde (1995).

For both Porter and van der Linde's (1995) study and Palmer, Oates, and Portney's (1995) study, a clear distinction between the short run and long run is important. Although innovation may reduce short-run and long-run costs, as long as innovation involves initial fixed costs, the cost savings are more likely to be realized in the long run than in the short run. On the other hand, variable costs may decrease through energy efficiency, reduced waste treatment, and higher resource productivity (Simpson and Bradford 1996). Thus, even in the short run, a decrease in variable costs may exceed the increase in fixed costs, causing total costs to decrease and financial performance to improve. Of course, as argued above, a decrease in total costs may not materialize until the long run. New cost-reducing technologies may require time for the firm to implement (Porter and van der Linde 1995). Likewise, new cost-reducing technologies may not be immediately available to a firm at the time the regulation is imposed. Thus, short-run financial performance depends not only on the degree to which new technologies reduce variable costs but also the speed with which new technologies are effectively incorporated.

B. Empirical Literature

In addition to the noted theoretical studies, other studies empirically examine the relationships described by the Porter hypothesis. These studies fall into five sets. One set assesses the effect of environmental regulation on firms' competitiveness, innovation activities, employment or productivity (Berman and Bui 2001; Greenstone 2002; Jaffe and Palmer 1997; Jaffe et al. 1995; Managi et al. 2005). Although evidence for employment and productivity is mixed, these studies generally find environmental regulation has had no effect on competitiveness and a positive effect on innovation activities. A second set generally finds that environmental regulation raises firms' costs (Gray 1987; Hazilla and Kopp 1990; Jorgenson and Wilcoxen 1990). Although they contribute to our understanding, Porter and van der Linde (1995) argue that these studies' results are subject to bias because they assume away innovation benefits. A third set of empirical studies examines the effect of environmental regulation on firms' location or investment decisions (Becker and Henderson 2000; Dean, Brown, and Stango 2000; Levinson 1996; List et al. 2003; McConnell and Schwab 1990). These studies generally find no evidence that environmental regulation affects location decisions but does have a negative effect on plant births and size.

A fourth set of empirical studies provides conflicting evidence regarding the effect of firms' environmental performance on financial performance (Barth and McNichols 1994; Fil-beck and Gorman 2004; Khanna and Damon 1999; Konar and Cohen 2001; Russo and Fouts 1997). Although these studies improve our understanding of the Porter hypothesis, they do not directly assess the effect of environmental regulation. With one exception--Khanna and Damon (1999)--the studies examine only short-run financial performance or only long-run financial performance. Nevertheless, we use the studies to identify meaningful factors to explain financial performance. (3)

In contrast to the fourth set of studies, the last set of empirical studies examines the effect of environmental regulation on financial performance. Brannlund, Fare, and Grosskopf (1995) use partially simulated data to study the effect of environmental regulation on profits for firms in the Swedish pulp and paper industry. Regulation is measured by the absolute amount of pollution a firm is permitted to discharge. The authors use a nonparametric, linear-programming approach to examine the ratio of short-run regulation-constrained profits to unconstrained short-run profits as a measure of the cost of regulation. Based on this simulation, the authors conclude that most firms in their sample are unaffected by environmental regulation, yet some experience lower profits with regulation. Alpay, Buccola, and Kerkvliet (2002) examine the effect of pollution regulations on the profitability and productivity growth of the Mexican and U.S. food industries. These authors find that U.S. pollution regulations have no impact on the profitability or productivity of U.S. food manufacturing, yet Mexico's rising environmental standards enhance food processors' productivity growth.

III. MEASURING AND EXPLAINING FINANCIAL PERFORMANCE

Our analysis draws upon the noted empirical studies to construct an econometric model that examines both the short-run and long-run effects of clean water regulation on financial performance. We use return on sales as a measure of financial performance. Return on sales is a good measure of financial performance because it indicates how effectively a firm is utilizing its resources to generate profits. By definition, return on sales, denoted ROS, is the ratio of a firm's operating profits, denoted [PI], to the firm's sales revenue, denoted S:

(1) ROS = [PI]/S.

Return on sales evaluates operational efficiency, reflecting how much operating profit is being earned per dollar of sales revenue. (4) Because profits equal sales less costs, denoted C, we are able to reformulate return on sales as follows:

(2) ROS = (S - C)/S = 1 - C/S.

Equation (2) shows that the return on sales captures a firm's ability to contain costs, conditional on a particular value of sales. In other words, given a fixed value of sales, any change in costs signals the effectiveness with which a firm turns sales into profits.

As our measure of clean water regulation, we use permitted wastewater discharge limits, which are particularly useful because they represent a performance-based standard rather than an input-based standard. The Porter hypothesis does not claim that any type of environmental regulation improves profitability. Instead, the Porter hypothesis enunciates that well-designed regulation improves profitability. Porter and van der Linde (1995) explicitly offer performance-based standards as examples of well-designed regulation. Thus, our assessment of a performance-based standard represents a proper test of the Porter hypothesis.

A. Short-Run and Long-Run Implications of Clean Water Regulation

To estimate return on sales, we use a linear specification. The presence of negative return on sales prevents the use of semi-log and log-linear specifications. Moreover, Khanna and Damon (1999) conclude that linear specifications are best for estimating rates of return. In the linear specification, the absolute level of the return on sales for firm f at time t, denoted RO[S.sub.(f, t)] is a function of these factors: (1) clean water regulation, denoted [R.sub.(f, t)], measured by permitted discharge limits; (2) variables other than clean water regulation, denoted [X.sub.(f, t)], including a firm's sales growth, capital intensity, age of assets, size, current ratio, research and development intensity, market share, and industry concentration; and (3) a stochastic error term, denoted [[epsilon].sub.(f, t)]:

(3) RO[S.sub.(f, t)] = [[alpha].sub.0] + [[alpha].sub.x] [X.sub.(f, t)] + [[alpha].sub.R] [R.sub.(f, t)] + [[epsilon].sub.(f, t)].

The contemporaneous clean water regulation measure captures the short-run effect of clean water regulation. In order to capture the long-run effect of clean water regulation on financial performance, we use both the contemporaneous regulation measure and a set of lagged measures (Lanoie, Patry, and Lajeunesse 2008). Specifically, in addition to the contemporaneous permitted discharge limit shown in Equation (3), we include the permitted discharge limit lagged by one quarter, two quarters, three quarters, four quarters, five quarters, and six quarters as follows:

(4) RO[S.sub.(f, t)] = [[alpha].sub.0] + [[alpha].sub.x] [X.sub.(f, t)] + [[alpha].sub.R] [R.sub.(f, t)] + [[alpha].sub.(R - 1)] [R.sub.(f, t - 1)] + [[alpha].sub.(R - 2)] [R.sub.(f, t - 2)] + [[alpha].sub.(R - 3)] [R.sub.(f, t - 3)] + [[alpha].sub.(R - 4)] [R.sub.(f, t - 4)] + [[alpha].sub.(R - 5)] [R.sub.(f, t - 5)] + [[alpha].sub.(R - 6)] [R.sub.(f, t - 6)] + [[alpha].sub. (f, t)].

We choose a six-quarter lag structure based on empirical relevance, which strikes a balance between too few quarters and too many quarters. The inclusion of too few quarters undermines our ability to capture the long-run effects of environmental regulation; the inclusion of too many quarters prompts the unnecessary loss of data points from the beginning of the sample period and further taxes the data's ability to discern a connection between permitted discharge limits and profitability. By considering the full lag structure, along with the contemporaneous limit, we are able to assess the short-run effect of clean water regulation and the ex post long-run effect of clean water regulation.

B. Firm-Specific Effects

To estimate Equations (3) and (4), we draw upon an unbalanced panel data set. Thus, the stochastic error term in Equations (3) and (4) can be further specified as follows with a firm-specific, time-invariant component, [[micro].sub.f], which may or may not be correlated with the independent variables, and an independent component, [[eta].sub.(f, t)], which is not correlated with the independent variables:

(5) [[epsilon].sub.(f, t)], = [[micro].sub.f] + [[eta].sub.(f, t)]

We consider three panel estimators: pooled ordinary least squares (OLS), fixed effects, and random effects. The pooled OLS estimator assumes that observations for a given firm are serially uncorrelated and that the errors are homoskedastic across firms and time (i.e., Var([[epsilon].sub.(f, t)]) = [[sigma].sup.2] ... [for all] f, t). The random effects estimator assumes [u.sub.f] is uncorrelated with the independent variables. The fixed effects estimator assumes Uf is correlated with the independent variables.

We use standard tests to assess the empirical strength of the panel models: the Breusch-Pagan Lagrange multiplier test, the Hausman test, and the F-test for fixed effects. The Breusch-Pagan test indicates whether there is an uncorrelated firm-specific component of variance (i.e., [H.sub.0]: Var([u.sub.f]) = 0 ... [for all] f). If not, the random effects estimator reduces to the pooled OLS estimator. Otherwise, the random effects estimator is more appropriate because the pooled OLS estimator fails to use information about the heteroskedasticity that results from using repeated observations on the firm. The Hausman test indicates whether the random effects estimates are consistent by assessing whether the covariance between the fixed effects estimates and the random effects estimates equals the variance of the random effects estimates (i.e., [H.sub.0]: Cov ([[[alpha].sup.FE].sub.X], [[[alpha].sup.FE].sub.X]) - Var ([[[alpha].sup.FE].sub.X]) = 0 ... [for all] X). The fixed effects estimator is consistent but not necessarily efficient. If [u.sub.F] is indeed orthogonal to the independent variables, the random effects estimator is consistent and efficient, while the fixed effects estimator is inefficient. Otherwise, the random effects estimator is inconsistent and the fixed effects estimator is both consistent and efficient. The F-test for fixed effects indicates whether the firm-specific intercepts are equal (i.e., [H.sub.0]: [u.sub.f] = 0 ... [for all] f). If the null hypothesis cannot be rejected, the fixed effects model is invalid.

The reported Breusch-Pagan test statistics reveal heteroskedasticity in the pooled OLS model for both Equations (3) and (4) so we do not report pooled OLS estimates for either specification. The reported Hausman test statistics indicate that the random effects estimates are not consistent for either Equation (3) or Equation (4) so we do not report the random effects estimates for either specification. The reported F-test for fixed effects statistics reveals significant firm-specific effects in the fixed effects model for both Equations (3) and (4). Consequently, we report only the fixed effects estimates.

In addition to empirical support for the fixed effects estimator, use of the fixed effects estimator proves helpful because this estimator identifies the effect of clean water regulation based exclusively on intrafirm variation. Thus, this estimator is less vulnerable to any concerns over the process that imposes permitted wastewater discharge limits and its variation across different firms. Also, the fixed effects estimator controls for the likelihood that each firm has its own way of managing wastewater discharges.

C. Interpretation of Coefficient Estimates

Both short-run and long-run implications are important for our assessment of the effect of clean water regulation on financial performance. Although a sizable portion of the costs of clean water regulation is likely to occur on an ongoing basis, many of the costs may occur only immediately after the regulation is imposed. Moreover, many of the benefits associated with the Porter hypothesis may only be realized in the long run.

Using parameters from Equation (4), we identify ten combinations of contemporaneous and lagged limit coefficients--categorized by sign--that help to discern short-run and long-run results. For this identification, we treat all six lagged limit coefficients as a whole, collectively denoted as [[alpha].sub.(R - 1)]

1 [[alpha].sub.R] = 0; [[alpha].sub.(R - 1)] < 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] < 0: Tighter clean water regulation (i.e., lower discharge limit) eventually improves return on sales but not immediately. This outcome is consistent with the Porter hypothesis because a firm may not be able to immediately exploit innovation offsets.

2 [[alpha].sub.R] < 0; [[alpha].sub.(R - 1)] < 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] < 0: Tighter clean water regulation immediately improves return on sales and even more so in the long run. This outcome is consistent with the Porter hypothesis because a firm is more able to exploit innovation offsets given more time.

3 [[alpha].sub.R] > 0; [[alpha].sub.(R - 1)] < 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] < 0: Tighter clean water regulation immediately undermines return on sales but prompts further adjustment that ultimately improves return on sales given more time to respond. This outcome is consistent with the Porter hypothesis because a firm may be less able to cope with regulation in the short run but more able to exploit innovation offsets given more time.

4 [[alpha].sub.R] < 0; [[alpha].sub.(R - 1)] = 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] < 0: Tighter clean water regulation immediately improves return on sales but prompts no further adjustment even given more time to respond. This outcome is not fully consistent with the Porter hypothesis because a firm should be more able to exploit innovation offsets given more time.

5 [[alpha].sub.R] < 0; [[alpha].sub.(R - 1)] > 0: [[alpha].sub.R] + [[alpha].sub.(R - 1)] < 0: Tighter clean water regulation immediately improves return on sales but prompts further adjustment that reduces the positive effect given more time to respond. Again, this outcome is not fully consistent with the Porter hypothesis because a firm should be more able to exploit innovation offsets given more time.

6 [[alpha].sub.R] > 0; [[alpha].sub.(R - 1)] = 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] > 0: Tighter clean water regulation immediately undermines return on sales but prompts no further adjustment even given more time to respond. This outcome is consistent with the costly regulation hypothesis because a firm is unable to cope with costly regulation in the short run.

7 [[alpha].sub.R] > 0; [[alpha].sub.(R - 1)] < 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] > 0: Tighter clean water regulation immediately undermines return on sales but prompts further adjustment that reduces the negative effect given more time to respond. This outcome is consistent with the costly regulation hypothesis because a firm is unable to cope with costly regulation in the short run but may be more able to cope given more time.

8 [[alpha].sub.R] = 0; [[alpha].sub.(R - 1)] > 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] > 0: Tighter clean water regulation eventually undermines return on sales but not immediately. This outcome is not fully consistent with the costly regulation hypothesis because a firm should be less able to cope with costly regulation in the short run.

9 [[alpha].sub.R] > 0; [[alpha].sub.(R - 1)] > 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] > 0: Tighter clean water regulation immediately undermines return on sales and even more so in the long run. Again, this outcome is not fully consistent with the costly regulation hypothesis because a firm should be less able to cope with costly regulation in the short run.

10 [[alpha].sub.R] < 0; [[alpha].sub.(R - 1)] > 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] > 0: Tighter clean water regulation immediately improves return on sales but prompts further adjustment that ultimately undermines return on sales given more time to respond. This outcome is not consistent with either hypothesis because a firm should be less able to cope with costly regulation in the short run and more able to exploit innovation offsets given more time.

IV. DATA

To estimate the noted functional relationships, we gather data on clean water regulation and financial performance for the chemical manufacturing industry group, which is a useful industry group for our examination for three reasons. First, it is a prominent source of wastewater discharges (Earnhart 2009). Second, the industries face performance-based standards, which Porter and van der Linde (1995) identify as prime examples of well-designed regulation. Third, the chemical manufacturing industry group includes several industries, which improves our ability to generalize conclusions, yet the industries are sufficiently similar to facilitate our ability to discern a meaningful relationship between regulatory stringency and financial performance. Of course, other industry groups also fit these criteria so our selection of the chemical manufacturing industry group is not critical to generate meaningful results.

A. Data on Environmental Regulation

As our measure of clean water regulation, we use the permitted wastewater discharge limits that are imposed by the Environmental Protection Agency (EPA) or authorized state regulatory agencies on major chemical manufacturing facilities regulated as point sources within the National Pollutant Discharge Elimination System (NPDES) Program. (5) The NPDES program is authorized by the Clean Water Act to control water pollution by issuing to facilities permits that specify discharge limits.

Permit Writing. Permit writers consider two standards to determine a permitted discharge limit: (1) the state water quality-based standard and (2) the Effluent Limitation Guideline standard. After a limit is determined under each standard, the more stringent limit is written into the permit. The state water quality-based standard is designed to ensure that the ambient water quality of the receiving water body meets state-based ambient quality standards. In other words, the discharge limit is set so that the facility's discharges do not cause the water body's ambient concentration of the relevant pollutant to exceed the acceptable level. Although state water quality standards do not differ within a state, the discharge limits identified by state water quality-based standards will differ across facilities and time because the background pollution of water bodies differs across time and space.

Effluent Limitation Guideline standards are designed to require a minimum level of wastewater treatment for a given industry (i.e., they establish a uniform upper bound on limits across the entire United States for a given industry). If no industry-specific Effluent Limitation Guideline applies to the particular facility, the permit writer uses his or her Best Professional Judgment, which draws upon all reasonably available and relevant data. In particular, the permit writer evaluates the effect of a permitted discharge limit on the environment. However, Best Professional Judgment is relevant for none of the sampled firms because all of the sampled facilities operate in industries covered by Effluent Limitation Guidelines. Regardless of the standard used to determine a limit, a facility is allowed to use any available technology to comply with the limit. (6)

Exogenous Determination of Permitted Discharge Limits. The use of permitted discharge limits faces two potential sources of endogenous determination. However, the process of permit writing to impose a permitted discharge limit provides assurance that permitted discharge limits are exogenously determined.

First, if permitted discharge limits depend on facilities' financial conditions, our use of permitted discharge limits as a measure of regulation may generate erroneous conclusions regarding the effect of environmental regulation on financial performance. This possible dependence of discharge limits on financial conditions can be explained by using a political economy theoretical framework in which regulators attempt to maximize political support for their policies. If critics of the Porter hypothesis are correct that tighter limits reduce profitability, imposing tighter limits lowers political support, especially when tighter limits prompt firms to close some of their facilities, which are highly visible effects of tighter regulation. Given the desire to maximize political support, regulators may impose discharge limits that are inversely related to a firm's financial performance. In this case, the relationship between discharge limits and financial performance represents a reverse causation: financial performance causes discharge limits rather than discharge limits causing financial performance. However, the process of determining permitted discharge limits indicates that permitted discharge limits are highly unlikely to depend on financial performance. Permitted discharge limit levels are determined by Effluent Limitation Guidelines, which apply uniformly across all facilities within a particular industry, or by state water quality standards, which clearly are not based on financial performance. Thus, neither of these two dimensions relate to an individual facility's financial conditions. More specifically, we are assured by conversations with EPA permit writers that financial performance is not considered when developing a permitted discharge limit.

Second, if permitted discharge limits depend on facility-level wastewater discharges or an individual facility's ability to control discharges (i.e., cost of compliance), we would misinterpret the effect of environmental regulation on financial performance because limits would not be exogenously determined. As long as a facility's ability to control pollution is correlated with financial performance to some degree, we would be incorrectly attributing influence to discharge limits rather than to a facility's ability to control pollution. Because limits depend on industrywide guidelines or ambient water quality concerns, facility-specific discharge limits cannot depend on facility-level wastewater discharges in the past, present, or future or facility-level costs of compliance.

Permitted Discharge Limits as a Regulatory Measure. We gather monthly data on permitted discharge limits from the EPA's Permit Compliance System (PCS) database for the sample period of January 1995 to June 2001. We use information on permitted discharge limits for two regulated pollutants--biochemical oxygen demand (BOD) and total suspended solids (TSS). These two pollutants provide a good generalization for other regulated pollutants for three reasons. First, they are both conventional pollutants, which receive the bulk of regulatory scrutiny. Second, all previous studies of wastewater discharges examine only BOD or both BOD and TSS (Earnhart 2004). Third, TSS is the most prevalent pollutant in our sample, and BOD is the second most prevalent.

Permitted discharge limits are based on the quantity or the concentration of pollutants. Quantity-based limits identify the absolute amount of pollutant discharges allowed (e.g., pounds per day). Concentration-based limits identify the amount of pollution allowed relative to the volume of treated wastewater (e.g., milligrams per liter). (7) A major chemical manufacturing facility may face one or both types of permitted discharge limit bases. Because we are not certain which permitted discharge limit basis is more restrictive, we retain information for both bases.

For each month, t, we assess potentially four different permitted discharge limits: TSS quantity-based limits, denoted [[L.sup.TQ].sub.t], TSS concentration-based limits, denoted [[L.sup.TC].sub.t], BOD quantity-based limits, denoted [[L.sup.BQ].sub.t], and BOD concentration-based limits, denoted [[L.sup.BC].sub.t]. In order to facilitate our focus on variation in clean water regulation generally rather than specific pollutants or specific discharge bases and in order to save degrees of freedom to employ a six-quarter lag structure to capture long-run regulatory effects, we combine the potential four permitted discharge limits into a single composite permitted discharge limit. We generate this composite limit for each facility in each month. First, we transform the two discharge bases--quantity and concentration--into a single permitted discharge limit for each pollutant--BOD and TSS. In particular, we divide each of the monthly discharge bases by the relevant subsample mean and then average the two scaled discharge bases. (8) Consider the case of TSS limits. We divide the TSS quantity-based permitted discharge limit, [[L.sup.TQ].sub.t], by the average of all quantity-based permitted discharge limits across all years and facilities, denoted [L.sup.TQ]. The resulting scaled TSS quantity-based limit is denoted [[L.sup.TQS].sub.t]: [[L.sup.TQS].sub.t] = [[L.sup.TQ].sub.t]/[L.sup.TQ].From similar calculations, the resulting scaled TSS concentration-based limit is denoted [[L.sup.TCS].sub.t]: [[L.sup.TCS].sub.t] = [[L.sup.TC].sub.t]/[[L.sup.TC].sub.t]. We then generate a single-scaled TSS discharge limit, denoted [[L.sup.TS].sub.t]: [[L.sup.TS].sub.t] = ([[L.sup.TQS].sub.t] + [[L.sup.TCS].sub.t])/2. Similarly, we generate a single-scaled BOD discharge limit, denoted Lfs. In order to assess clean water regulation generally, we generate a scaled composite limit measure, denoted [[L.sup.S].sub.t]: [[L.sup.S].sub.t] = ([[L.sup.TS].sub.t] + [[L.sup.BS].sub.t]/2. The resulting composite limit provides a unit-less measure of clean water regulation by comparing the stringency of limits at one facility relative to the stringency of limits at all other facilities in the sample. Thus, a lower "relative" permitted discharge limit indicates more stringent clean water regulation.

To justify the use of the composite limit, in lieu of four separate permitted discharge limits, to estimate financial performance, we conduct a Wald test based on the estimation of Equation (3) to assess the joint equality of the coefficients associated with the four individual scaled permitted discharge limits (i.e., [H.sub.0]: [[alpha].sup.TQS] = [[alpha].sup.TCS] = [[alpha].sup.BQS] = [[alpha].sup.BCS])] The Walt test statistic fails to reject the joint null hypothesis of equal coefficients [F (3, 675) = 1.15, p = 0.329]. Our use of the composite limit as the key regressor implicitly imposes this equality of coefficients because the composite limit represents the sum of the four underlying individual permitted discharge limits (divided by four). Use of this summed measure is only appropriate when the four individual coefficients equal a common magnitude. The coefficient on the composite limit is identical to this common magnitude after adjusting for the scalar of one-fourth used to calculate the average composite limit.

B. Data on Financial Performance and Related Explanatory Factors

In addition to the data on clean water regulation, we gather data on financial performance and factors that are shown by previous studies to explain financial performance. To examine financial performance, we must limit our analysis to firms with publicly available financial data, which is available from Standard & Poor's Compustat Research Insight 7.9[c]. Research Insight 7.9[c] contains annual and quarterly data from financial statements for publicly owned firms traded on stock exchanges in the United States. (9) We focus on quarterly data for two reasons. First, quarterly data are more likely than annual data to capture fluctuations in financial performance. Second, fluctuations captured by quarterly data may be important for linking to month-to-month variation in permitted discharge limits. We gather the quarterly data for the sample period: first quarter of 1995 to second quarter of 2001.

Our measure of financial performance is return on sales as defined in Equation (1). To construct return on sales, we use profits before interest and taxes (i.e., operating profits) and sales less returns and allowances (i.e., net sales) from firms' financial statements.

Given that return on sales reflects current profitability, we include as regressors the following factors that help to explain profitability: (1) a firm's sales growth, (2) a firm's capital intensity, (3) age of a firm's assets, (4) a firm's size, (5) a firm's current ratio, (6) a firm's research and development intensity, (7) a firm's market share, and (8) concentration of the industry in which a firm operates. (10)

Sales Growth. Although robust sales growth is generally indicative of a firm's ability to compete and shield itself from cyclical market variations (Perez-Quiros and Timmerman 2000), there may be an optimal point beyond which further sales growth impairs a firm's flexibility and adaptability, adversely affecting profitability. Thus, the ultimate effect of sales growth on profitability is ambiguous. However, empirical evidence consistently supports a positive relationship between profitability and sales growth (Capon, Farley, and Hoenig 1990; Khanna and Damon 1999; Konar and Cohen 2001; Russo and Fouts 1997). We capture sales growth as the quarter-over-quarter compounded percent change in a firm's sales over the previous 3 years.

Capital Intensity. Capital intensity is defined as the amount of fixed or real capital relative to other factors of production and is generally measured as capital investments or asset stocks relative to sales, output, or labor (Capon, Farley, and Hoenig 1990). Assuming that a firm takes its capital and other factors of production as given at a point in time, capital intensity serves as a proxy for capacity utilization. Higher capital intensity indicates lower capacity utilization. If a firm's assets are idle, the firm's profits are lower. Accordingly, empirical evidence generally supports a negative relationship between profitability and capital intensity at the firm level (Capon, Farley, and Hoenig 1990; Khanna and Damon 1999; Russo and Fouts 1997). We measure a firm's capital intensity as the ratio of the firm's capital investments to the firm's sales. Sales depend on both the output price and the output level, which is a function of both capital and noncapital inputs. Thus, our measure varies with the output price, whereas capacity utilization is not. Nevertheless, as the capital intensity ratio declines, the use of noncapital inputs rises, so that capital's utilization increases.

Age of Assets. A firm with older assets may be less operationally efficient than a firm with new assets, which frequently embed updated technologies that lead to greater productivity. Alternatively, new equipment may be more expensive than old equipment, which yields higher depreciation. Thus, the a priori effect of the age of assets on profitability is ambiguous (Khanna and Damon 1999). Existing empirical evidence finds a positive relationship between profitability and the age of assets (Khanna and Damon 1999; Konar and Cohen 2001). We capture a firm's age of assets as the ratio of net property, plant, and equipment to gross property, plant, and equipment, the difference between the two representing depreciation.

Size. Profitability is expected to vary with the scale of operations, yet both economies of scale and diseconomies of scale may influence profitability. Empirical results reflect this ambiguity (Capon, Farley, and Hoenig 1990; Konar and Cohen 2001; Russo and Fouts 1997). Thus, the a priori effect of firm size on profitability is ambiguous. We capture a firm's size as the natural log of total assets. (11)

Current Ratio. The current ratio is defined as the ratio of current assets (assets expected to be converted to cash within 1 year) to current liabilities (liabilities due within 1 year). Thus, the cuirent ratio indicates a firm's ability to meet creditors' demands within a 1-year period. If current liabilities exceed current assets (i.e., current ratio < 1), a firm may be unable to meet short-term obligations. Although empirical evidence generally finds that lower levels of debt are positively related to financial results (Capon, Farley, and Hoenig 1990), a high current ratio may signal that a firm is not efficiently investing cash for future growth and productivity. Thus, the a priori effect of the current ratio on profitability is ambiguous.

Research and Development Intensity. A firm that assumes the risks associated with research and development is expected to be innovative, agile, and a market leader and should reap returns from its investments. Consistent with these expectations, empirical evidence generally supports a positive relationship between profitability and research and development (Capon, Farley, and Hoenig 1990; Konar and Cohen 2001). We capture a firm's research and development intensity as the ratio of research and development to the firm's sales. (12)

Market Share. Standard economic theory asserts that monopoly power improves profitability. To capture monopoly power, empirical studies generally include a measure of market share. These studies support the predicted positive relationship (Capon, Farley, and Hoenig 1990; Konar and Cohen 2001). We capture a firm's market share as the ratio of the firm's sales to total industry sales by 4-digit SIC industry.

Industry Concentration. Industry concentration reflects the extent to which a small number of firms account for a large proportion of industrial activity. Accordingly, a firm in a high-concentration industry is expected to yield higher profitability than a firm in a low-concentration industry. In their extensive meta-analysis of the determinants of profitability. Capon, Farley, and Hoenig (1990) overwhelmingly conclude that higher profitability accrues to firms operating in high-concentration industries. We capture industry concentration as the four-firm sales ratio that is specific to the 4-digit SIC industry in which a firm operates.

C. Matching the PCS and Research Insight 7.9[c] Extracts

We next merge the data on dependent variables and the data on explanatory factors. In particular, we match the facility-level environmental data to the firm-level financial data by year using the firm-level name. Because the firm-level name is not initially available in the environmental data, we use the EPA's Toxic Release Inventory database to find the firm-level name by year for each facility in the environmental data set.

After we match the environmental and financial data, we address the potential level difference--facility and firm--and the temporal frequency difference--monthly and quarterly--between the two sets of data. The environmental data are recorded at the facility level with a monthly frequency for major chemical manufacturing facilities. The financial data are recorded at the firm level with a quarterly frequency for firms in all industries. To resolve the frequency difference, we average the environmental data across months to a quarterly frequency. This approach resolves the temporal frequency difference for all firms. To resolve the potential level difference, we average the environmental data across commonly owned facilities within a firm to the firm level. (13) This approach resolves the level difference for firms that own only chemical manufacturing facilities contained in the environmental data. However, this approach does not completely resolve potential level differences for firms that own facilities outside the chemical manufacturing industry. In these cases, the financial data reflect activity from all facilities, but the environmental data reflect activity only from major chemical manufacturing facilities. To address these potential level differences, we use the firm-level primary SIC code from Research Insight 7.9[c] to divide the sample into two subsamples: One subsample includes only firms with a primary 2-digit SIC code in the chemical manufacturing industries, while the other subsample includes firms with a primary 2-digit SIC code outside the chemical industries. A Chow test rejects the null hypothesis that the coefficient estimates are the same between the chemical and nonchemical samples. Based on these test results, we proceed with the subsample of firms with a primary 2-digit SIC code in the chemical industries. (14)

D. Summary Statistics and Correlation Coefficients

After matching the environmental and financial data and applying necessary restrictions, the sample includes an unbalanced panel of 815 observations, representing 53 firms. Table 1 contains descriptive statistics. Table 2 contains correlation coefficients among independent variables. Correlations among the independent variables are generally small, excluding the lagged permitted discharge limits. However, correlations among the contemporaneous permitted discharge limit, and the six lagged measures of permitted discharge limits are 0.97 and higher.

TABLE 1

Descriptive Statistics

                                        Standard
Variable                         Mean   Deviation

Return on sales (ratio)          0.066      0.110

Sales growth (percent)           0.042      0.126

Capital intensity (ratio)        0.079      0.070

Age of assets (ratio)            0.523      0.118

Firm size (U.S. dollars)        22.017      1.345

Market share (ratio)             0.148      0.206

Industry concentration (ratio)   0.680      0.200

Current ratio (ratio)            1.590      0.629

R&D intensity (ratio)            0.043      0.043

Discharge limit (ratio)          1.714      2.661

Number of observations             815

Number of firms                     53

TABLE 2

Correlation Coefficients for Independent Variables

                             (1)    (2)    (3)    (4)    (5)

Primary Factors

Sales growth                 1.00

Capital intensity           -0.19   1.00

Asset age                    0.15  -0.01   1.00

Firm size                    0.15   0.05   0.21   1.00

Market share                -0.01   0.01  -0.06   0.60   1.00

Industry concentration      -0.14   0.06  -0.28  -0.06   0.41

Current ratio               0.1 1  -0.02   0.08  -0.42  -0.25

R&D intensity                0.27  -0.05   0.16   0.53   0.02

Discharge limit              0.31  -0.04   0.19   0.39  -0.05

Contemporaneous
Limit and Lagged Limits

Contemporaneous limit        1.00

One-quarter lagged limit     0.99   1.00

Two-quarter lagged limit     0.99   0.99   1.00

Three-quarter lagged limit   0.99   0.99   0.99   1.00

Four-quarter lagged limit    0.99   0.99   0.99   0.99   1.00

Five-quarter lagged limit    0.98   0.99   0.99   0.99   0.99

Six-quarter lagged limit     0.97   0.98   0.98   0.98   0.99

                             (6)    (7)   (8)   (9)

Primary Factors

Sales growth

Capital intensity

Asset age

Firm size

Market share

Industry concentration       1.00

Current ratio               -0.05   1.00

R&D intensity               -0.37  -0.16  1.00

Discharge limit             -0.40  -0.13  0.61  1.00

Contemporaneous

Limit and Lagged Limits

Contemporaneous limit

One-quarter lagged limit

Two-quarter lagged limit

Three-quarter lagged limit

Four-quarter lagged limit

Five-quarter lagged limit    1.00

Six-quarter lagged limit     0.99   1.00

V. RESULTS

Our analysis seeks to estimate the short-run and long-run effects of clean water regulation, as measured by permitted discharge limits, on financial performance, as measured by the return on sales. Tables 3 and 4 display results from the fixed effects estimation of Equations (3) and (4), respectively. The displayed p-values are derived from standard errors that are clustered on individual firms to correct for any unknown form of heteroskedasticity and/or serial correlation. (15) In addition to p-values and standard errors, we report variance inflation factors (VIFs), which assess the degree of interdependence among the independent variables. (16)

A. General Exploration

We estimate two specifications of financial performance. One specification reflects Equation (3), which includes only the contemporaneous permitted discharge limit and evaluates only the short-run regulatory effect (short-run specification). The other specification reflects Equation (4), which includes both the contemporaneous permitted discharge limit and the set of six lagged permitted discharge limits and evaluates both the short-run and long-run regulatory effects (short-run/long-run specification). For both specifications, the coefficient estimates for the traditional control variables are generally consistent with our expectations. In the short-run specification, return on sales is positively related to the current ratio and negatively related to firm size and market share. At the 10% level, coefficient estimates for sales growth, age of assets, industry concentration, capital intensity, and research and development intensity are insignificant. In the short-run/long-run specification, return on sales is positively related to capital intensity and the current ratio. At the 10% level, coefficient estimates for firm size, market share, sales growth, age of assets, industry concentration, and research and development intensity are insignificant.

We next evaluate the effects of clean water regulation. Based on the short-run specification of Equation (3), we interpret the coefficient associated with the contemporaneous permitted discharge limit. Based on the short-run/long-run specification of Equation (4), we interpret the coefficients associated with the contemporaneous permitted discharge limit and the six lagged permitted discharge limits. For the short-run specification, the coefficient estimate for the contemporaneous permitted discharge limit is negative and significant at the 10% level. (17) For the short-run/long-run specification, the coefficient estimate for the contemporaneous permitted discharge limit is negative and significant at the 5% level. Among the six lagged permitted discharge limits, the four-quarter-lagged permitted discharge limit coefficient is negative and significant at the 10% level, while the other lagged permitted discharge limit coefficients are not significant. These results provide only weak evidence for lagged discharge limit effects. However, we also test the joint significance of the six lagged permitted discharge limit coefficients (i.e., [H.sub.0]: [[alpha].sub.(R - 1)] = [[alpha].sub.(R - 2)] = [[alpha].sub.(R - 3)] = [[alpha].sub.(R - 4)] = [[alpha].sub.(R - 5)] = [[alpha].sub.(R - 6)] = 0) The resulting test statistic indicates that the lagged permitted discharge limit coefficients are jointly significant at the 1% level. Thus, despite the presence of only a single individually significant lagged permitted discharge limit coefficient, our estimates reveal that firms respond to limits in periods beyond the current period. Moreover, our results appear to demonstrate that it takes a full year, plus the current period, to reach the long run because the lagged limit coefficients after four quarters are statistically zero. (18)

These results reflect combination (2) from Section III.C: [[alpha].sub.R] < 0; [[alpha].sub.(R - 1)] < 0; [[alpha].sub.R] + [[alpha].sub.(R - 1)] < 0. Recall that lower values of the permitted discharge limit indicate more stringent regulation. Accordingly, a negative coefficient estimate for the permitted discharge limit indicates that more stringent regulation improves financial performance, while a positive coefficient estimate indicates that more stringent regulation hinders financial performance. Given this interpretation, the reported coefficients linking clean water regulation to financial performance are consistent with the Porter hypothesis, while apparently rejecting the costly regulation hypothesis. More important, the pattern of coefficients is consistent with our a priori expectations regarding the distinction between the short-run and long-run effects of clean water regulation on financial performance. Specifically, a firm is more able to exploit innovation offsets given more time. (19)

B. Alternative Interpretation: Costs Relative to Sales

The reformulation of return on sales shown in Equation (2) reveals that return on sales reflects a firm's costs relative to sales. The estimated effects of the permitted discharge limit--both contemporaneous and lagged--on return on sales indicate that a tighter permitted discharge limit appears to reduce costs relative to sales in both the short run and the long run. Thus, tighter regulation seems to enhance a firm's ability to contain costs, conditional upon a particular level of sales.

C. Short-Run and Long-Run Economic Consequences of Clean Water Regulation

We next assess the short-run and long-run economic consequences of our results from the short-run/long-run specification. The average percent change in permitted discharge limits from one quarter to the next across all firms and quarters in our sample is approximately 2%, where we calculate the percent change for firm f at time t as follows:

(6) % [DELTA] L = ([L.sub.(f, t)] - [L.sub.(f, t - 1)])/[L.sub.(f, t - 1)].

We use the average percent change to calculate how a negative change in the permitted discharge limit affects return on sales in the short run and the long run. (20)

With respect to short-run economic consequences, a 2% reduction in the average permitted discharge limit, denoted R, causes the return on sales to increase according to the following equation:

(7) [DELTA] ROS = [[alpha].sub.R] x -0.020 x R.

As shown in Table 1, the average permitted discharge limit equals 1.714. Thus, a 2% reduction in the average permitted discharge limit causes the return on sales to increase by a factor of 0.001 (i.e., [DELTA] ROS = -0.033 x -0.020 x 1.714) in the short run. Relative to the sample average return on sales of 0.066 shown in Table 1, this factor represents a 1.5% increase. When evaluating the long-run economic consequences, we use only statistically significant coefficients in our calculation. A 2% reduction in the average permitted discharge limit causes the return on sales to increase according to the following equation:

(8) [DELTA] ROS = ([[alpha].sub.R] + [[alpha].sub.(R - 4]]) x -0.020 x R.

Thus, a 2% reduction in the average permitted discharge limit causes the return on sales to increase in the long run by a factor of 0.003 (i.e., [DELTA] ROS = (-0.033 - 0.042) x -0.020 x 1.714). Relative to the sample average return on sales, this factor represents a 4.5% increase.

Thus, the short-run and long-run net benefits realized by firms are both statistically meaningful and economically meaningful. However, the short-run benefits are less than half of the long-run benefits. Consistent with the Porter hypothesis, the majority of benefits are realized in the long run. These results suggest that tighter permitted discharge limits spur improvements in current and future operations. Put differently, some net benefits appear to be realized during the short-run transition to comply with a tighter permitted discharge limit, with additional benefits accruing to the firm in the long run when the firm has more time to innovate.

VI. CONCLUSIONS

In this article, we study the short-run and long-run effects of clean water regulation on the financial performance of publicly owned firms in the chemical manufacturing industries. As our measure of financial performance, we focus on return on sales, which indicates how effectively a firm is utilizing its resources to generate profits. To capture the short-run and long-run effects, we estimate return on sales as a function of both the contemporaneous permitted discharge limit and a six-quarter lag structure of permitted discharge limits.

Our empirical analysis identifies a positive relationship between tighter clean water regulation and financial performance in both the short run and long run, with a stronger effect in the long run. By reformulating return on sales, the relevant estimates indicate that tighter clean water regulation decreases short-run and long-run costs, conditional on a given level of sales.

Our study represents an important step toward resolving the debate regarding the effect of environmental regulation on financial performance. The Porter hypothesis appears to focus more on the long-run effect of regulation on financial performance, but other perspectives suggest that there may also be an important short-run effect of regulation on financial performance during the transition to future innovative solutions. Our short-run and long-run results are consistent with both points. If our results are believable, then perhaps economic theory should revisit its treatment of the link from environmental regulation to financial performance with a greater scrutiny between the short run and long run. As part of this revisitation, economic theory may explore whether the motivation to innovate prompted by more stringent regulation increases over time or the feasibility of innovation improves over time.

Lastly, we recognize that our results are specific to the performance-based regulation of two wastewater pollutants discharged by publicly owned firms in the chemical manufacturing industries. To assess whether these results generalize to other forms of regulation, other pollutants, other media, and other industries, future research should include studies that examine the effects of other types of regulation, such as market-based tradable permits, and studies that examine other wastewater pollutants, pollution in other environmental media, and other polluting industry groups.

APPENDIX: ESTIMATION OF SALES GROWTH

The Porter hypothesis asserts that properly designed environmental regulation motivates firms to innovate, which ultimately improves financial performance. Our analysis identifies a positive effect of tighter clean water regulation on return on sales in both the short run and long run. We conclude that our results are consistent with the Porter hypothesis. However, the analysis does not directly measure the effect of tighter clean water regulation on innovation. To address this point, this appendix considered the possibility that tighter clean water regulation induces a positive effect on return on sales without the presence of innovation.

To this end, we assume that clean water regulation restricts discharges that cannot be abated and that these discharges stem from the use of an essential input for which no substitute exists. Although the latter condition is possible, the former condition is remote in our sample because many good technologies exist for abating BOD and TSS discharges within the chemical manufacturing industries (Earnhart et al. 2006). For simplicity, we also assume that the regulation applies to all facilities operating across the nation. Although these conditions could only apply to a few of the industries in our sample, the connection could alter our overall conclusions. Because firms are unable to abate pollution or substitute away the essential input industry output would decrease, output prices would increase, and unit costs would remain unchanged. Under these conditions, the effect of tighter clean water regulation on return on sales would be positive even without the presence of innovation. In addition, because revenues are determined in part by prices, the effect of tighter clean water regulation on revenues under the aforementioned conditions could also be positive.

Given these possible connections, we assess the robustness of our results and strive to support our overall conclusions by examining the effect of permitted discharge limits on sales growth. If permitted discharge limits have no effect on sales growth, the result lends support to our overall conclusions. If permitted discharge limits have a positive or negative effect on sales growth, then our overall conclusions are subject to mixed evidence. To facilitate our assessment, we perform a fixed effects regression of sales growth for firm f at time t, denoted [g.sub.(f, t)], on the same factors used in Equation (4) to estimate ROS:

(A1) [g.sub.(f, t)] = [[alpha].sub.0] + [[alpha].sub.x] [X.sub.(f, t)] + [[alpha].sub.R] [R.sub.(f, t)] + [[alpha].sub.(R - 1)] [R.sub.(f, t - 1)] + [[alpha].sub.(R - 2)] [R.sub.(f, t - 2)] + [[alpha].sub.(R - 3)] [R.sub.(f, t - 3)] + [[alpha].sub.(R - 4)] [R.sub.(f, t - 4)] + [[alpha].sub.(R - 5)] [R.sub.(f, t - 5)] + [[alpha].sub.(R - 6)] [R.sub.(f, t - 6)] + [[epsilon].sub.(f, t)]

Our results indicate no statistically significant effect of any of the permitted discharge limit variables on sales growth. Thus, our conclusions remain unchanged.

REFERENCES

Alpay, Ebru, Steven Buccola, and Joe Kerkvliet. "Productivity Growth and Environmental Regulation in Mexican and U.S. Food Manufacturing." American Journal of Agricultural Economics, 84(4), 2002, 887-901.

Ambec, Stefan, and Philippe Barla. "A Theoretical Foundation of the Porter Hypothesis." Economics Letters, 75(3), 2002, 355-60.

Bandyopadhyay, Sushenjit, and John Horowitz. "Do Plants Over-comply with Water Pollution Regulations? The Role of Discharge Variability." The B.E. Journals in Economic Analysis and Policy, 6(1), 2006, Article 4.

Barth, Mary E., and Maureen F. McNichols. "Estimation and Market Valuation of Environmental Liabilities Relating to Superfund Sites." Journal of Accounting Research, 32, 1994, 177-209.

Becker, Randy, and Vernon Henderson. "Effects of Air Quality Regulations on Polluting Industries." Journal of Political Economy, 108(2), 2000, 379-421.

Berman, Eli, and Linda T. M. Bui. "Environmental Regulation and Productivity: Evidence from Oil Refineries." Review of Economics and Statistics, 83(3), 2001, 498-510.

Brannlund, Runnar, Rolf Fare, and Shawna Grosskopf. "Environmental Regulation and Profitability: An Application to Swedish Pulp and Paper Mills." Environmental and Resource Economics, 6(1), 1995, 23-36.

Business and Company Resource Center[C]. 2004. Thomson Gale, a part of the Thomson Corporation.

Capon, Noel, John U. Farley, and Scott Hoenig. "Determinants of Financial Performance: A Meta-Analysis." Management Science, 36(10), 1990, 1143-59.

Dean. Thomas J., Robert L. Brown, and Victor Stango. "Environmental Regulation as a Barrier to the Formation of Small Manufacturing Establishments: A Longitudinal Examination." Journal of Environmental Economics and Management, 40(1), 2000, 56-75.

Earnhart, Dietrich. "Panel Data Analysis of Regulatory Factors Shaping Environmental Performance." Review of Economics and Statistics, 86(1), 2004, 391-401.

__. "Effects of Permitted Effluent Limits on Environmental Compliance Levels." Ecological Economics, 61(1), 2007, 178-93.

__. "The Influence of Facility Characteristics and Permit Conditions on the Effects of Environmental Regulatory Deterrence." Journal of Regulatory Economics, 36(3), 2009, 247-73.

Earnhart, Dietrich, Donald Haider-Markel, Tatsui Ebihara, and Robert Glicksman. March 11, 2006. "Shaping Corporate Environmental Behavior and Performance: The Impact of Enforcement and Non-Enforcement Tools." Environmental Protection Agency report, 2006.

Filbeck, Greg, and Raymond F. Gorman. "The Relationship between the Environmental and Financial Performance of Public Utilities." Environmental and Resource Economics, 29(2), 2004, 137-57.

Gray, Wayne B. "The Cost of Regulation: OSHA, EPA, and the Productivity Slowdown/' American Economic Review, 77(5), 1987, 998-1006.

Greenstone, Michael. "The Impacts of Environmental Regulations on Industrial Activity: Evidence from the 1970 and 1977 Clean Air Act Amendments and the Census of Manufactures." Journal of Political Economy 110(6), 2002, 1175-219.

Hazilla, Michael, and Raymond J. Kopp. "Social Cost of Environmental Quality Regulations: A General Equilibrium Analysis." Journal of Political Economy, 98(4), 1990, 853-73.

Jaffe, Adam B., and Karen Palmer. "Environmental Regulation and Innovation: A Panel Data Study." Review' of Economics and Statistics, 79(4), 1997, 610-19.

Jaffe, Adam B., Steve R. Peterson. Paul R. Portney, and Robert N. Stavins. "Environmental Regulation and the Competitiveness of U.S. Manufacturing: What Does the Evidence Tell Us?" Journal of Economic Literature 33(1), 1995, 132-63.

Jorgenson, Dale W., and Peter J. Wilcoxen. "Environmental Regulation and U.S. Economic Growth." Rand Journal of Economics, 21(2), 1990, 314-40.

Khanna, Madhu, and Lisa A. Damon. "EPA's Voluntary 33/50 Program: Impact on Toxic Releases and Economic Performance of Firms." Journal of Environmental Economics and Management, 37(1), 1999, 1-25.

King, Andrew. "Retrieving and Transferring Embodied Data: Implications for Management of Interdependence within Organizations." Management Science, 45(7), 1999, 918-35.

__. "Organizational Response to Environmental Regulation: Punctuated Change or Autogenesis?" Business Strategy and the Environment, 4(9), 2000, 224-38.

King, Andrew, and Michael Lenox. "Exploring the Locus of Profitable Pollution Reduction." Management Science, 48(2), 2002, 289-99.

Konar, Shameek, and Mark A. Cohen. "Does the Market Value Environmental Performance?" Review of Economics and Statistics, 83(2), 2001, 281-89.

Lanoie, Paul, Michel, Patry, and Richard, Lajeunesse, "Environmental Regulation and Productivity: Testing the Porter Hypothesis." Journal of Productivity Analysis, 30, 2008, 121-28.

Levinson, Arik. "Environmental Regulations and Manufac turers' Location Choices: Evidence from the Census of Manufactures." Journal of Public Economics, 62(1 -2), 1996. 5-29.

List, John A., Daniel Millimet, Per Fredricksson, and Warren McHone. "Effects of Environmental Regulations on Manufacturing Plant Births: Evidence from a Propensity Score Matching Estimator." Review of Economics and Statistics 85(4), 2003, 944-52.

Managi. Shunsuke, James J. Opaluch. Di Jin, and Thomas A. Grigalunas. "Environmental Regulations and Technological Change in the Offshore Oil and Gas Industry." Land Economics, 81(2), 2005, 303-19.

McConnell, Virginia D., and Robert M. Schwab, "The Impact of Environmental Regulation on Industry Location Decisions: The Motor Vehicle Industry." Land Economics 66(1), 1990. 67-81.

Mohr, Robert. "Technical Change, External Economies, and the Porter Hypothesis." Journal of Environmental Economics and Management, 43(1), 2002, 158-68.

Palmer, Karen L., Wallace E. Oates, and Paul R. Portney. "Tightening Environmental Standards: The Benefit-Cost or the No-Cost Paradigm?" Journal of Economic Perspectives, 9(4), 1995, 119-32.

Perez-Quiros, Gabriel, and Allan Timmerman. "Firm Size and Cyclical Variations in Stock Returns." Journal of Finance, 55, 2000, 1229-62.

Porter, Michael E. The Competitive Advantage of Nations. New York: Free Press, 1990.

__. "Green Competitiveness." Scientific American, 264, 1991, 168.

Porter, Michael E., and Claas van der Linde. "Toward a New Conception of the Environment-Competitiveness Relationship." The Journal of Economic Perspectives, 9(4), 1995, 97-118.

Research Insight 7.9[c], 1993-2003. Standard and Poor's, A Division of The McGraw-Hill Companies, Inc., All rights reserved.

Russo, Michael V., and Paul A. Fouts. "A Resource-Based Perspective on Corporate Environmental Performance and Profitability." The Academy of Management Journal, 40(3), 1997, 534-59.

Simpson, David R., and Robert L. Bradford III. "Taxing Variable Cost: Environmental Regulation as Industrial Policy." Journal of Environmental Economics and Management, 30(3), 1996, 282-300.

van der Linde, Claas. "The Micro-Economic Implications of Environmental Regulation: A Preliminary Frame-work." Environmental Policies and Industrial Competitiveness. Organization of Economic Cooperation and Development, Paris, 1993, 69-77.

Vogan, Christine R. "Pollution Abatement and Control Expenditures, 1972-94." Survey of Current Business, 76(9), 1996, 48-67.

ABBREVIATIONS

BOD: Biochemical Oxygen Demand 

EPA: Environmental Protection Agency

NPDES: National Pollutant Discharge Elimination System

OLS: Ordinary Least Squares PCS: Permit Compliance System

TSS: Total Suspended Solids

VIF: Variance Inflation Factor 

(1.) This controversy proves important for our economy. As a share of U.S. gross domestic product, total expenditures for pollution abatement and control were approximately 1.8% from the mid-1970s to the mid-1990s, according to the most recently published pollution abatement and control expenditure estimates (Vogan 1996).

(2.) Other theoretical studies address particular aspects of the organizational behavior underlying the Porter hypothesis (Ambec and Barla 2002; King 1999, 2000; King and Lenox 2002; Mohr 2002).

(3.) These points notwithstanding, these studies need not lend substantial evidence regarding the effect of environmental regulation on profitability. Most important, environmental performance is an endogenous variable that reflects firms' decisions. In addition, the level of environmental performance serves as a weak proxy for environmental regulation because the measures of environmental performance generally include unregulated pollutant emissions (e.g., Toxic Release Inventory emissions).

(4.) The reliability of return on sales is sometimes criticized as subject to accounting manipulation. Although a firm is granted some flexibility in the classification of operating expenses as either costs of goods sold or selling, general, and administrative costs, performance measures that are calculated after the full consideration of operating expenses (e.g., return on sales) are not subject to accounting manipulation.

(5.) Major chemical manufacturing facilities are distinguished from minor facilities by the EPA as those with a significant effect on receiving water or a major rating code equal to or greater than 80. A major rating code is the total numeric score of ranking points assigned to a facility from the NPDES Permit Ranking Worksheet. Only minimal information on permitted discharge limits imposed on minor facilities is available from the EPA.

(6.) Permitted discharge limits may vary from month to month for three reasons. First, permits are generally issued on a 5-year cycle. However, these cycles are not tied to calendar years. Thus, the switch from one permit to another may occur mid-year. Second, permits impose three types of limits: initial, interim, and final. The limit levels frequently vary across three limit types. Again, the transition from one limit type to the next may occur mid-year because they do not apply to particular calendar years. Third and most important, permits may impose seasonal limits, which vary within a calendar year.

(7.) Some previous empirical studies examine the effect of permitted effluent limits on wastewater discharges. Some studies examine exclusively quantity-based limits (e.g., Brannlund, Fare, and Grosskopf 1995). Other studies examine exclusively concentration-based limits (e.g., Bandyopad-hyay and Horowitz 2006: Farnhart 2007).

(8.) The means of the subsamples for the TSS quantity-based limits and concentration-based limits are 1,199 pounds per day and 38.6 mg/L, respectively. The means of the subsamples for the BOD quantity-based limits and concentration-based limits are 805 pounds per day and 27.8 mg/L, respectively.

(9.) Because one might expect that the relationship between environmental regulation and financial performance depends on ownership structure, we do not claim that our results generalize to privately owned firms.

(10.) Consistent with previous studies that estimate profitability, including return on sales (Capon, Farley, and Hoenig 1990), a firm's sales value is reflected in both leftside and right-side variables of Equations (3) and (4) because sales value captures the competitive outcome of a firm in a way that no other measure captures. As its most important contribution, the use of sales value enables us to calculate profits in the numerator of return on sales. Likewise, use of sales value in the denominator of return on sales allows us to adjust for firm size. Although other scaling factors exist, the use of sales value enables us to reformulate return on sales in Equation (2) to facilitate interpretation of our results. On the right-side of Equations (3) and (4), a firm's sales value serves as the most effective scaling factor available in our data for capital intensity and research and development intensity. In addition, sales value captures market share and industry concentration, which cannot be captured by other information on the firm. Given our prevalent use of sales information, we assess the possibility of multicollinearity among the right-side variables by computing variance inflation factors (see Section V and Tables 3 and 4) and correlation coefficients (see Section IV.D and Table 2),

(11.) In addition to firm size, we control for the average size of all facilities owned by a firm by including as a proxy the average flow of wastewater that facilities are designed to manage (How design capacity), which measures the amount of water a facility is designed to handle, not the amount of pollution discharged by a facility. These data are available from the EPA's PCS database. We average across all facilities owned by a particular firm Similar to firm size, facility size may affect profitability due to the presence of economies of scale or diseconomies of scale at the facility level. Also, facility size and quantity-based limits are most likely correlated. Inclusion of facility size as a regressor avoids any potential omitted variable bias associated with this correlation. Without this inclusion, we might incorrectly attribute the influence of facility size on profitability as part of the effect of the permitted discharge limit on profitability. These points notwithstanding, we neither report nor interpret the coefficient estimates for facility size because they prove insignificant and because results are robust to the exclusion of this factor.

(12.) Quarterly research and development expenditures are poorly populated in Research Insight 7.9[c]. so we use annual research and development expenditures. We generate quarterly variation by including quarterly sales in the denominator.

(13.) This approach implicitly assumes that firms possess the ability to tailor their pollution control efforts to fit their individual facilities (i.e., firms are able to lower wastewater discharges more strongly at their facilities that are facing tighter permitted discharge limits, while lowering wastewater discharges less strongly at their facilities that are facing looser permitted discharge limits). Although this assumption is plausible, firms may not possess the noted ability. In this case, the minimum, not the average, discharge limit across all facilities owned by a firm may dictate the firm's chosen level of pollution control.

(14.) The Chow test statistic indicates that the firms focusing on chemical manufacturing respond differently to environmental regulation than do firms producing a mix of products. Thus, our results do not appear to generalize to all firms that produce at least chemicals. Of course, this difference may stem solely from our inability to examine the permitted discharge limits for facilities that do not manufacture chemicals, yet are owned by firms that do manufacture chemicals at other facilities. This difference need not surprise us. Firms that focus on chemical manufacturing may be better able to innovate in response to tighter limits because they focus on a smaller set of products. On the other hand, firms that manufacture a broader mix of products may be better able to innovate in response to tighter limits because their breadth of products facilitates a broader perspective on pollution control techniques (i.e. effective pollution control techniques for one product may transfer to the pollution control applied to another disparate product).

(15.) Our data contain an unbalanced panel of 53 firms with a maximum of 26 observations possible per firm. Given the number of observations provided for some firms, we consider the possibility of autocorrelation among the observations. In this case, the independent error component of Equation (5) can be specified as follows: [[eta].sub.(f, t)] = [rho] [[eta].sub.(f, t - 1)] + [[upsilon].sub.(f, t)]. For this assessment, we calculate a Baltagi-Wu LBI statistic for autocorrelation ([H.sub.0]: [rho] = 0). The statistic is 1.97 and 2.04 for the estimations of Equations (3) and (4), respectively. A Baltagi-Wu LBI statistic below 1.5 rejects the null hypothesis. Thus, autocorrelation does not appear to be a concern.

(16.) Guidelines of VIF < 5 or VIF < 10 are generally used to conclude that multicollinearity is not a concern and coefficient estimates are stable. None of the reported VIFs exceed 2.5.

(17.) In order to explore the short-run effect more fully, we generate firm-specific measures of the contemporaneous permitted discharge limit by interacting firm-specific intercepts with the contemporaneous permitted discharge limit. We then test the joint null hypothesis of equal firm-specific limit coefficients (i.e., [H.sub.0]: [[alpha].sub.(R,1)] = ... = [[alpha].sub.(R,53)]). The test statistic fails 10 reject the joint null hypothesis [F(31,721) = 0.600, p = 0.9591.

(18.) We explore the sensitivity of these reported conclusions based on alternative formulations of the short-run/long-run specification. Specifically, we explore a four-quarter lag structure, a five-quarter lag structure, and an eight-quarter lag structure. We do not consider smaller lag structures because they would not effectively assess the long-run effect of clean water regulation. Estimation of these alternatively formulated specifications generates conclusions identical to those reported. In particular, the sign and significance of the contemporaneous permitted discharge limit coefficient and the presence of joint significance among the lagged permitted discharge limit coefficients are strongly robust to the number of lags. Moreover, the inclusion of additional lagged permitted discharge limit measures identifies no significant effect beyond four quarters in the lag structure.

(19.) See the Appendix for further support of the effect of environmental regulation on innovation.

(20.) The limit regressor reflects an average of potentially four different permitted discharge limits. For expositional ease, we assume that all four permitted discharge limits change by 2%. In other words, we simulate an across-the-board reduction in permitted discharge limits.

TABLE 3

Fixed Effects Results to Estimate Only the Short-Run Effect of
Environmental Regulation on Return on Sales

Independent     Coefficient       Standard                    Variance
Variable          Estimate           Error         p-Value   Inflation
  (a)                                                 (b)      Factor

Sales                0.050               0.046       0.278      1.263
growth

Capital              0.134               0.082       0.109      1.486
intensity

Age                 -0.072               0.058       0.217      1.151
of assets

Firm                -0.044               0.021       0.038      1.382
size

Market              -0.228               0.126       0.077      1.962

Industry            -0.043               0.131       0.741      1.170
concentration

Current              0.031               0.011       0.005      1.044
ratio

R&D                 -0.080               0.383       0.836      1.081
intensity

Discharge           -0.009               0.005       0.073      1.018
limit

Number of                                  815
observations

Number                                      53
of firms

Within                                   0.050
[R.sup.2]

F-test                      F(52, 752) = 2.86,
of fixed                       p-value = .000
firm-specific
intercepts

Hausman test:                  [X.sup.2](10) =
random                                  31.07,
effects vs.                     p-value = .001
fixed effects

Breusch-Pagan                   [X.sup.2](l) =
test: random                            32.15,
effects                          p-value = 000
vs. pooled OLS

(a) The regression also includes an intercept term
and a factor capturing the average size of chemical
facilities owned by a firm (as measured by flow volume)
(b) p-values are derived from standard errors
that are clustered on individual firms.

TABLE 4

Fixed Effects Results to Estimate Both the Short-Run and Long-Run
Effects of Environmental Regulation on Return on Sales

Independent      Coefficient      Standard    p-Value   Variance
Variable           Estimate         Error      (b)     Inflation
  (a)                                                   Factor

Sales                  0.033          0.052    0.527      1.278
growth

Capital                0.182          0.080    0.027      1.769
intensity

Age                   -0.119          0.110    0.285      1.216
of assets

Firm                  -0.026          0.028    0.362      1.395
size

Market                -0.215          0.133    0.113      2.210
share

Industry              -0.025          0.119    0.832      1.261
concentration

Current                0.031          0.014    0.026      1.069
ratio

R&D                    0.370          0.542    0.499      1.093
intensity

Contemporaneous       -0.033          0.012    0.011      1.846
limit

One-quarter            0.004          0.012    0.726      2.326
lagged limit

Two-quarter            0.088          0.063    0.171      2.382
lagged limit

Three-quarter         -0.017          0.018    0.349      1.858
lagged limit

Four-quarter          -0.042          0.025    0.099      1.969
lagged limit

Five-quarter          -0.006          0.016    0.729      2.459
lagged limit

Six-quarter           -0.017          0.016    0.311      2.074
lagged limit

Number of                               584
observations

Number of                                47
firms

Within                                0.097
[R.sup.2]

F-test                           F(46, 521)
of fixed                             = 2.07
firm-specific                       p-value
intercepts                           = .000

Hausman                           [x.sup.2]
test:                         (16) = 20.07,
random                              p-value
effects vs.                          = .217
fixed effects

Brcusch-Pagan                  [x.sup.2](l)
test:                              = 19.46,
random                              p-value
effects vs                           = .000
pooled OLS

(a) The regression also includes an intercept term and a factor
capturing the average size of chemical facilities owned
by a firm (as measured by flow volume).
(b) p-values are derived from standard errors that are
clustered on individual firms.

doi: 10.1111/j.l465-7287.2010.00237.x

DYLAN G. RASSIER and DIETRICH EARHART *

* This article was developed under STAR Research Assistance Agreement No. R-82882801-0 awarded by the U.S. EPA. The EPA does not endorse any products or commercial services mentioned in this article. The article and the analysis herein were developed for the author's dissertation prior to and independent of the author's employment with the Bureau of Economic Analysis (BEA). The views expressed in this article are solely those of the authors and not necessarily those of the U.S. Department of Commerce or the BEA.

Rassier: Research Economist, Bureau of Economic Analysis, U.S. Department of Commerce, BE-40. Washington. DC 20230. Phone [phone], Fax [phone], E-mail [e-mail]

Eamhart: Professor. Department of Economics, University of Kansas, 435 Snow Hall, Lawrence, KS 66045. Phone [phone], Fax [phone], E-mail [e-mail]