F520

Adam Eckerd is assistant professor in

the Center for Public Administration and

Policy at Virginia Tech, where he conducts

research on the complex relationship

between government decisions and social

outcomes, particularly with respect to

environmental justice, public participation,

and nonprofi t organizations. He holds a

doctorate from the John Glenn School of

Public Affairs at The Ohio State University.

E-mail: [email protected]

Roy L. Heidelberg is assistant professor

in the Public Administration Institute at

Louisiana State University. His research

focuses on power and accountability in

administrative contexts. He holds a doctor-

ate from the John Glenn School of Public

Affairs at The Ohio State University.

E-mail: [email protected]

252 Public Administration Review • March | April 2015

Public Administration Review,

Vol. 75, Iss. 2, pp. 252–261. © 2014 by

The American Society for Public Administration.

DOI: 10.1111/puar.12305.

Adam Eckerd Virginia Tech

Roy L. Heidelberg Louisiana State University

Public policies are often shaped by prominent public management frameworks or designed with particular management frameworks in mind (May and Winter 2009). In the 1980s and 1990s, the focus of environmental policy shifted to concentrate less on regulation and enforcement and more on incentives and legal protection (Jaff e et al. 1995). Th e timing of this change coincided (perhaps not coincidentally) with the rise of the New Public Management (NPM) (Hood 1991), and the various eff orts to motivate private involvement with con- taminated site remediation and redevelopment align closely with NPM principles. Essential to NPM is the use of incentives as a policy instrument to engage pri- vate sector fi rms as partners, a focus that functionally replaces regulation (DeLeon and Denhardt 2000).

Th e federal eff orts to address contaminated sites through liability reform are an excellent example: regulation is restricted so that incentives can be made more eff ective and the power of the market can be lev- eraged. Our primary interest is in the extent to which

various liability reform and incentive policies spur private sector involvement in site remediation. In line with NPM expectations, developers have indicated that their involvement in cleanup activities would increase with protections from liability (Wernstedt, Meyer, and Alberini 2006), similar to those imple- mented through amendments to the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA) in 2002. Th e 2002 amendments are situated under a diff erent managerial logic than the 1980 law. While the regulatory focus of the 1980 law implies a managerial approach of direct implementation to balance various priorities, the market-oriented focus of the 2002 amendments emphasizes getting incentives right and a more indi- rect approach to fi nal outcomes.

We address three main research questions through an evaluation of data from the Brownfi eld Assessment Grant Program of the U.S. Environmental Protection Agency (EPA). A hallmark of NPM is a results-oriented focus and engagement with the

Public Incentives, Market Motivations, and Contaminated Properties: New Public Management and Brownfi eld Liability Reform

Abstract: Brownfi elds pose challenges to both communities and policy makers. Public funds are insuffi cient to remedi- ate these contaminated sites, but, given the uncertainty of contamination and the complexity of liability, private interests are reluctant to become involved for fear of future litigation. From a New Public Management perspective, market incentives can be used to encourage private sector remediation of sites. However, this change implies a shift in administrative function from regulation to “getting the incentives right.” In this research, the authors investigate whether state and federal reforms aimed at increasing private sector involvement have actually done so, and they con- sider the implications for other goals of brownfi eld remediation, such as providing economic development assistance in communities where such change is needed. Findings show that developers respond to insurance and tax incentives, but the authors question whether public incentives are making unattractive redevelopment opportunities worth investing in or simply making profi table redevelopment opportunities more profi table.

Practitioner Points • Practitioners involved in brownfi eld remediation must try to distinguish between incentivizing already

attractive projects and encouraging remediation of projects that would otherwise be neglected. • A hands-off approach to steering brownfi eld remediation may not adequately address the uncertainty intrin-

sic to remediation, even with promises of liability protection. • Focusing resources on spurring private sector involvement will necessarily prioritize economic development

over health and equity considerations. • Governments may need to take a more balanced approach to brownfi eld remediation by taking ownership of

and planning for redevelopment in communities where private sector involvement is less likely.

Public Incentives, Market Motivations, and Contaminated Properties: New Public Management and Brownfi eld Liability Reform 253

that ownership presents the risk of unforeseen costs for cleanup (Wernstedt, Meyer, and Alberini 2006), and can also be held liable for the eff ects of the contamination (Howland 2000).

While there are some private fi rms that focus on purchasing and cleaning sites (Meyer and Lyons 2000), signifi cant impediments to broader private sector involvement exist. Th us, few sites are remedi- ated (Eckerd and Keeler 2012), and those that are tend to be located on valuable land (Daley and Layton 2004; McCarthy 2009), which is contrary to the goal of using brownfi eld remediation to bring needed development to communities (DePass 2006).

Developers have stated that reforming liability rules and off ering liability protection will promote more private sector involvement in projects (Wernstedt, Meyer, and Alberini 2006). Th e policy response was to spur (rather than compel) private sector involvement, shift- ing administrative functions from enforcement to incentivizing the redevelopment of brownfi eld sites. We see this response and the very framing of the issue as a refl ection of the infl uence of NPM thinking (Box 1999; DeLeon and Denhardt 2000). Indeed, over the past 15 years, states have addressed liability issues, set up funding mechanisms, created tax incentives for remediation, and reduced regulatory barriers. One purpose of this article is to explore whether developers’ stated preferences (Wernstedt, Meyer, and Alberini 2006) align with their actual investment choices and whether state policies designed to address barriers to private sector involvement have actually increased the prevalence of hazardous site cleanup.

NPM and the New Paradigm of Environmental Management Structuring the market to achieve a public purpose is in fact the opposite

of leaving matters to the “free market”—it is a form of intervention in the market . . . Structuring the market is also the opposite of creating

publicly administered bureaucracies to deliver services. It is a third way . . . a way of using public leverage to shape private decisions to achieve

collective goals. It is a classic method of entrepreneurial governance.

—Osborne and Gaebler (1992, 238–84)

A shift from regulation to market reform implies a diff erent admin- istrative approach and an emphasis on a diff erent set of results. Private sector involvement is one of the goals of brownfi eld reme-

diation policy, but as a means to the ultimate ends—reducing the health risks posed to communities and putting land back to pro- ductive use (DePass 2006; Eckerd 2013). Th e latter use is particularly important in eco- nomically depressed communities, and most brownfi elds are located in poor communities. Brownfi eld remediation and redevelopment off er communities economic development assistance that they might not otherwise receive while, at the same time, removing a community eyesore and potential health

threat. If policy can encourage the market to meet these needs, then a substantive regulatory apparatus is unnecessary.

Indeed, as NPM gained infl uence, the approach to environmental policy changed. Th e 1990 Clean Air Act and 2003 Clean Water

private sector (Box 1999), and so we fi rst evaluate whether there is a statistical diff erence in private sector involvement after liability reform. Second, we explore the eff ects of other market-oriented instruments at the state level, and third, we consider the trade-off made by emphasizing incentives over regulation. A key rationale of brownfi eld remediation policy is that it is a driver of redevel- opment in economically depressed areas (DePass 2006), but an incentive and liability reform approach can achieve this, by design, only indirectly. Our third research question attempts to determine whether the post-2002 reforms led to an increase in remediation in such areas.

The Problem with Contaminated Sites In the United States, there are more than 450,000, and potentially many more, known brownfi elds—properties on which the potential presence of hazardous materials impedes redevelopment, reuse, or expansion.1 Brownfi elds are diverse,2 ranging from shuttered dry cleaners or gas stations to large-scale former chemical manufactur- ers or coal yards (Gayer, Hamilton, and Viscusi 2000); most are located in urban areas. Th e sites pose several problems. Pollution can seep into groundwater or the air, and, at minimum, the site can be an eyesore. Pollution can have health eff ects both for the nearby residential population and, depending on the severity of the pollution, for an entire jurisdiction. Common pollutants at brownfi eld sites include asbestos, lead, petroleum, polychlorinated biphenyls (PCBs), and volatile organic compounds (VOCs) such as formaldehyde or benzene. Properties often become brownfi elds when the business operating on the site fails and, given the complex pollution remediation needs and potential liability issues, other entities have little incentive to purchase these properties (McCarthy 2002). Consequently, the sites often remain abandoned, generating neither productive use for the community nor property taxes for the local jurisdiction. Residential land nearby is likely to depreci- ate in value as those who can aff ord to move away do so (Hite et al. 2001). Th ese housing market eff ects can amplify poverty-related problems such as crime and joblessness (Leyden, Goldberg, and Michelbach 2011).

Given the vast number of brownfi elds, few sites are remedi- ated only with public resources. CERCLA (1980) was the fi rst eff ort to use federal funding to facilitate cleanup. CERCLA was a regulatory law. It dealt primarily with Superfund sites, with federal assistance focused on locating potentially responsible parties (PRPs) and coordinating state and local cleanup fund- ing. Public funds are usually insuffi cient to cover cleanup costs, so without identifying a PRP, sites are unlikely to be remediated (Fitzgerald and Leigh 2002). Th e adminis- trative focus of CERCLA, therefore, was to identify the private party that could be held responsible and compel that party to provide the resources needed for remediation.

Ideally, the PRP is the polluter and can be held liable, but this is practically diffi cult for many brownfi elds. Th e fi rm that caused the pollution is often no longer in existence. Th us, courts have determined that current owners can be held responsible for the cleanup costs of contaminants that they did not create, meaning

Brownfi eld remediation and redevelopment off er commu- nities economic development assistance that they might not otherwise receive while, at the same time, removing a com- munity eyesore and potential

health threat.

254 Public Administration Review • March | April 2015

Incentives may only provide more profi t potential (Meyer and Lyons 2000).

Th ere are a variety of market mechanisms that have been used to spur brownfi eld remediation. Table 1 reports the number of states that off er some of the common incentives and liability reforms.

Developers have stated that they will be more likely to invest in brownfi elds if there are suffi cient incentives and particularly struc- tures to mitigate developer risk (Wernstedt, Meyer, and Alberini 2006). However, it is not clear whether these stated preferences have resulted in consonant behavior, and if so, whether private interests target properties that are located in thriving or expected-to-thrive communities (Sherman 2003). If this is the case, then incentives could be benefi ting the developers who are motivated by favorable economic conditions (Howland 2000, 2010), which suggests that there could be a disconnect between the instrument being used (incentives) and the objective of the policy (Daley and Layton 2004).

Private Sector Involvement In this article, we evaluate whether and how a set of market-oriented reforms has spurred private sector involvement in brownfi eld sites. Our data provide some insight into whether these incentives have met their intended output and whether revealed behaviors match the stated preferences of developers reported in Wernstedt, Meyer, and Alberini (2006). Th is informs our fi rst research questions.

Research question 1: Is there a statistically signifi cant change in private sector involvement after federal liability reform?

Research question 2a: Does the stated preference for liability reform translate into a statistically diff erent propensity for private sector involvement, controlling for other policy instru- ments?

Research question 2b: Do policies instituted at the state level providing further incentives for developers produce a statisti-

cally diff erent level of involvement by private developers?

We also consider the extent to which these market incentives meet one of the key managerial goals of brownfi eld remediation: equity. We assess whether the characteristics of the community matter for redevelopment

strategies, and, following Sherman (2003), we model the eff ects of particular characteristics that are associated with strong develop- ment potential.

Act used trading markets to shift the burden of pollution costs to address acid rain and point source pollution in waterways. Rather than banning household pollutants that depleted ozone, excise taxes were used to reduce consumer demand, and the EPA established fuel effi ciency targets for auto manufacturers without specifying how those standards should be met (Gayer and Horowitz 2006). As described later, the 2002 brownfi elds law fi ts in the tradition of using policy to “steer” rather than “control” brownfi eld remediation.

Market-based incentives have often been assessed in a policy framework—typically aimed at minimizing management costs relative to environmental improvement (Stavins 2001). Under these terms, market-oriented approaches will likely appear to be the better option. However, the change from a command and control system to hands-off administrative function shifts the role of regula- tors from directing environmental change to guiding it. Th e onus shifts from targeting polluters to determining the appropriate set of incentives to appeal to developers. Th is also implies a subtle but important shift in stakeholder focus: where the public is the key cli- ent of a command and control system, an NPM approach removes administrators from this direct role with the public (Box 1999). Where historically administrators may have targeted and sought to remediate the most problematic sites (Hird 1993), either in terms of pollution or the eff ect of that pollution on nearby populations, an incentive focus leaves the market to determine the sites most in need of remediation (Daley and Layton 2004). Administrators are assessed on cost-eff ectiveness rather than responsiveness to public needs or equity issues (Box 1999). Here, we assess remediation on both terms—whether policy has increased private sector involve- ment and the eff ectiveness of various market-oriented incentives in doing so, and also whether improvements are occurring in those communities most in need.

Economic Incentives and Public Goals To spur private sector involvement, Congress amended CERCLA in 2002 to provide limited liability relief for PRPs that did not cause the pollution, an act that refl ects a belief in the essential importance of private fi rms in redevelopment eff orts (42 U.S.C. §§ 9601–9675). Th e law was modeled on state programs that emerged to encour- age brownfi eld remediation through regulatory relief, liability relief, tax-related incentives, expedited permitting opportunities, support- ive public infrastructure investments, and relief from the discovery of previous contamination or from future changes in regulatory standards.

However, such incentives will likely work only when a private developer expects profi tability. While governments may have an interest in ensuring that the most problematic of brownfi eld sites (in terms of overall risk or of undue burden on certain populations) are remediated, Sherman (2003) argues that fi nancial incen- tives do not make fi nancially unattractive projects attractive but rather make attractive projects more attractive. Th e goal of a market approach is to make remediation more attrac- tive to the market by making redevelopment more profi table, but private developers are not inclined to take on a particularly complex remediation project unless the land has high potential value and quality infrastructure (Howland 2000, 2010).

Table 1 Number of States with Brownfi eld Incentives

Incentive Number of States (percent)

Cleanup grants 32 (63%) Cleanup bonds 4 (8%) Cleanup loans 37 (73%) Tax abatement 12 (24%) Tax credits 27 (53%) Environmental insurance 6 (12%) Cleanup liability protection 6 (12%) Third-party liability protection 6 (12%) Citizen group grants 4 (8%)

Source: EPA (2009). Note: Total number possible for each incentive is 51, as the District of Columbia is included.

Th e goal of a market approach is to make remediation more attractive to the market by

making redevelopment more profi table.

Public Incentives, Market Motivations, and Contaminated Properties: New Public Management and Brownfi eld Liability Reform 255

In addition, considering that these sites off er our best glimpse into private sector prioritization in an aggregate manner (albeit within a subset of public sector priorities), we suggest that understanding the trends in remediation for these sites can contribute to our under- standing of both fi rm behavior and public administration.

Outcome Variables Given the large number of sites in our data set and the diff erent kinds of involvement of the private sector, we assess private sector involvement in several ways. First, private sector involvement can be indicated through the process of applying for (and receiving) an EPA grant. Because all of the projects in our data set are grant recipients, sites that are owned by nongovernmental, private entities that receive cleanup assistance clearly have private sector involve- ment. Th erefore, our fi rst outcome variable is an indicator variable of private sector (nongovernmental) ownership; approximately two- thirds of the sites in our data set are privately owned.

Private sector involvement need not be limited to ownership, however. For a variety of reasons, government entities may pos- sess these sites but intend them for future private use. A future use plan suggests that there is some proposed use of the land after the cleanup has been completed, and given that few sites are remedi- ated strictly with public money, the presence of a future use plan at the time of the grant application is an indicator that, regardless of ownership, there is likely some coordination of resources for the site after cleanup.

However, because the presence of a future use plan is not, in and of itself, indicative of private interest in the site, we made several exclusions in developing this variable. First, we only categorized those sites for which a plan was in place at the time of the grant application and not subsequently added. We did so to ensure that private sector interest was antecedent to federal funds being received. Further, while we do not have data on specifi c future use plans, we do know whether the site will be used for residen- tial, commercial, industrial, government, or greenspace purposes. Th erefore, we only categorized those sites with residential, com- mercial, or industrial plans. Our assumption is that such uses are unlikely to be undertaken solely by local governments without at least some level of private developer involvement (Alberini et al. 2005). After these adjustments, 8 percent of the sites included future use plans.

Finally, as the policy goal is to restore productive use to lands, we also consider the conditions under which sites are more (or less) likely to be fully remediated, as indicated by an offi cially sanctioned designation of no further action required (NFA). NFA letters are issued by state authorities and reported by the grant recipient. A site can be deemed NFA either as a result of a cleanup or a fi nding of no signifi cant environmental contamination. About 18 percent of the sites have achieved this status.

We appear to have mixed trends regarding our measures of private sector involvement. Figure 1 suggests that private sector ownership has not changed much over time; excluding a sharp drop in the data for 1999, private ownership has been relatively stable over time. However, as seen in fi gure 2, future use planning has been on a clear upward trend and a steep upward trend from 2004 to 2009.

Research question 3: Do incentives and liability reform pri- marily result in remediation of sites in comparatively higher- status communities?

Research questions 1 and 2 follow the evaluation criteria tradition- ally associated with NPM, which concerns eff orts to “get the incen- tives right.” Eff orts to promote private developers’ interests require aligning incentives in conjunction with insulating them from unreasonable risks. However, promoting private developers’ interests is a limited approach to governance. Th e recent guidelines of “good governance” show a growing trend toward a richer consideration of the quality of this involvement of the private sector, and research question 3 captures this analytical goal. As Grindle comments on governance, the desire to design locally attuned, context-sensitive policies “fall[s] short of what scholars, practitioners, politicians, citizens, and donors most wish to advance . . . parsimonious truths rather than complicated options” (2011, 416). But this should not prevent us from asking such questions.

Data and Method Th e primary data for this analysis come from the EPA, which provides a set of relatively low-dollar-value brownfi eld grants under its Brownfi eld Assessment Grant Program. Th ese grants are not intended to cover the costs of remediation but to assist in the envi- ronmental assessment of (potentially) contaminated sites. Recipients of these grants provide the EPA with initial and ongoing progress reports during the cleanup of the sites (Dull and Wernstedt 2010), provided that the cleanup moves forward after the assessment is completed. Th e data set consists of self-reported data on 6,932 brownfi eld sites distributed across the 50 states and the District of Columbia from the current site owner(s) and/or PRPs for each site. Th e data include sites that received grants between 1993 and 2009.

As Dull and Wernstedt (2010) note, sites that receive federal grants are not likely generalizable to the full population of brownfi eld sites, which number more than 1 million nationwide. We have a likely selection bias problem in that the sites for which we have data are already prioritized over the vast majority of brownfi eld sites by having received these grants and, moreover, by having organizations interested enough in site remediation to apply for the grant in the fi rst place. Th erefore, we are careful to note the limitations of our analysis up front—we caution against presuming any predictability or representativeness generated by our results.

Nevertheless, this set of brownfi elds is important to study for several reasons. First, as noted, sites are unlikely to be remediated without both public and private funding (Alberini et al. 2005), and assess- ment grants off er one of the most basic types of incentives. Second, our primary interest is in considering the relationship between pol- icy reform and managerial frameworks, so it is natural to consider sites that administrators are aware of, and this data set is the most comprehensive of such sites. Th ird, the EPA tends to favor cleanups in which private sector resources are involved (Dull and Wernstedt 2010; Eckerd and Keeler 2012), and these data should provide a glimpse into the revealed priorities of the private sector. Finally, these sites are the most salient sites to federal administrators, having received some amount of assistance, and thus indicate the sites that are administrative priorities. For this reason alone, understanding these sites is important for implementation of the brownfi elds law.

256 Public Administration Review • March | April 2015

off er full third-party liability protection. California, Missouri, and Massachusetts off er third-party liability protection under certain circumstances, while a handful of other states provide protection only for certain third parties, such as government owners or banks that own foreclosed properties.4

Cleanup or retroactive liability protects current owners from liability in case laws regarding cleanup change in the future. Provided that the cleanup meets appropriate standards at the time of the cleanup, the owner is protected regardless of future changes in cleanup stand- ards. We also include indicators for states that off er remediation bonds programs, grants, loans, tax abatement, tax credits, environ- mental insurance; several states also provide grants to citizen groups to clean up their communities. States vary with respect to how they balance incentives for private sector involvement with public goals of remediation (McCarthy 2002); we include indicator variables for those states that mandate public comment periods and public meetings and those that specify a procedure for state monitoring of remediated sites after cleanup. For each of these variables, we include indicator variables only for sites that applied for a grant after the policies were implemented in the state.

To account for important social goals of brownfi eld cleanup, such as targeting communities in need and addressing environ- mental justice (Daley and Layton 2004; Eckerd and Keeler 2012; Howland 2000), we include several census indicators to account for the nature of the neighborhood, the housing market, and the redevelopment prospects for the site. All of the following variables are derived from the census-tract level, from either the 1990 or 2000 census, depending on whether the site received a grant before or after 2000: median home values (logged); the proportion of homes that were vacant; the proportion of homes that were owner occupied; the proportion of residents in the census tract that were Hispanic, that were African American, and that received welfare benefi ts (Eckerd 2011; Ellen and O’Regan 2011; Hamnett 1991; Wyly and Hammel 1999).

Control Variables While we do not know the specifi c extent of contamination on the sites, we do know the type of pollution and the specifi c pollutants. To account for this, in the absence of full information, we use a simple additive index for each type of pollutant (e.g., lead, VOCs, PCBs, etc.) and each type of pollution present at the site (e.g., air, soil, groundwater, etc.).5 Finally, there are a number of site-level characteristics, including an indicator variable for those sites that are in industrial zones, which may be comparatively more favorable remediation targets given the lessened probability of future liability issues (Howland 2000), and the size of the site measured in acres (logged because of the presence of a few very large sites). Summary statistics for all variables are provided in table 2.

Method Owing to the largely exploratory and evaluative nature of our research questions, we opted to run a variety of analyses, both at the individual site level as well as in aggregate at the state level. Th e analysis is basically an interrupted time-series analysis conducted at these two levels. Th e fi rst set of analyses consists of three logistic regressions at the site level. As all sites in the analysis have received federal grants, the models predict the odds

Independent Variables As we noted earlier, reforms have occurred both at the federal level, mostly in terms of liability, and in a variety of forms within the states, including liability protection, tax incentives, and grants. At the national level, we include an indicator variable to account for those sites that began remediation before or after the 2002 federal brownfi elds law took eff ect (on January 1, 2003). To account for the diff erences among state remediation incentives, we include a variety of indicator variables regarding the incentive structures that states had in place at the time the site began the remediation process (see table 1). Th ese data were gathered from EPA documentation3 and verifi ed in each state’s code; they consist of protections such as cleanup (retroactive) liability and third-party liability, in addition to funding incentives and public involvement stipulations.

Th e 2002 federal law attempted to clarify a liability process and confi ne liability strictly to PRPs but also acknowledged that many states to go beyond this standard and off er further liability protec- tion. Th ird-party liability protection excludes current owners from any liability for site pollution (provided these parties were not the polluters). Only Connecticut, Georgia, and South Carolina

Figure 1 Proportion of Privately Owned Sites by Year

Figure 2 Proportion of Sites with Future Use Plans by Year

Public Incentives, Market Motivations, and Contaminated Properties: New Public Management and Brownfi eld Liability Reform 257

assessing the pace of remediation. Here, in addition to the number of years since policy was reformed, we include a measure of the amount of time, in days, since the site entered the remediation proc- ess until it received NFA designation, censored at December 31, 2009. Once again, this model includes fi xed eff ects for both the year the site received the grant and the state in which the site is located.

Th e third set of analyses is conducted at the state level to get an aggregate view. Similar to the site level, we ran three models predict- ing the proportion of a state’s sites, in any given year, that were privately owned and received grants, had future use plans, and were fully remediated (and also including the years since the reform was implemented). Th is aggregation left a panel of state-level character- istics per year, with a potential data set of 16 years of observations for 51 states (including Washington, DC), or 816 observations. However, many (smaller) states did not have any grant recipients in several of the years, particularly in the early 1990s, so we have a total set of 324 observations, with each state represented in at least one year. Again, there was no indication of autocorrelation, but there was heteroscedasticity, so we ran these panel regressions with robust standard errors.7

Results Results for the site-level analyses are provided in table 3; in the case of both an odds ratio and a hazard ratio, values near 1 are indica- tive of no relationship between the independent variable and the dependent variable, while values above 1 indicate an increase in the odds of the outcome in question, and values between 0 and 1 are indicative of decreased odds. We consider the fi rst two columns of table 3 indicators of private sector involvement in brownfi eld sites, and thus we interpret results that are consistent across both col- umns. Th e fi nal two columns are indicative of the ultimate implica- tions of the remediation eff ort.

Broadly speaking, neither federal reform nor state reforms have much of a time-dependent relationship with private involvement— neither of our interrupted time-series indicators provide much indi- cation of altering the fundamental relationship between brownfi eld redevelopment and private sector involvement. Nevertheless, there do appear to be some consistent trends with respect to the specifi c incentive mechanisms that states have used. First, direct funding mechanisms, such as cleanup bonds, grants, and loans, at best do not appear to foster private sector involvement and may actually discourage it. Th is counterintuitive result may suggest that the underlying logic of reform policies is disconnected from the reality of improving environmental conditions or, at the very least, that the direct funds provided are not enough to make projects profi t- able. Conversely, both tax credits and environmental insurance, more indirect methods, appear to foster private sector involvement, which may suggest that legal or regulatory certainty is valued over resources. Similar to the direct funding policies, and in contrast to the fi ndings of Wernstedt, Meyer, and Alberini (2006), liability protection at best does not foster private sector involvement and may discourage it.

Th e fi nal two columns in table 3 are the likelihood of a site being cleaned up and how quickly a cleanup takes place, respectively. Here, it does appear that federal reform may have altered both the probability and rate of cleanup. Sites that received grants after

of sites that received grants (1) being privately owned, (2) having a future use plan in place, and (3) being fully remediated (receiv- ing an NFA designation). Because all of the sites in our data set have received a remediation grant, the fi rst model estimates the probability of a privately owned site receiving a grant relative to a government-owned site. Th e second model is indicative of private sector involvement, regardless of ownership, and the third model assesses whether private ownership matters in terms of the policy goal of site remediation.6 Because major policy changes occurred at diff erent times in each state, we include a variable counting the number of years elapsed since the major policy change in the state and the year during which the site received a grant; this variable enables us to determine the extent to which the policy intervention altered the nature of private sector involvement over the long term.

Each of these models was assessed for the presence of autocorrela- tion, multicollinearity, heteroscedasticity, and the appropriateness of fi xed- or random-eff ects assumptions. In each case, controlling for unobserved fi xed eff ects of the state in which the site was located was necessary, as was controlling for the year in which the site received the grant and using robust standard errors to account for heteroscedastic- ity, although there was no indication of autocorrelation by year. Th us, each of these three models includes indicator variables for the year the grant was awarded (1993–2009) and the state. To conserve space, these fi xed-eff ect results are excluded from the tables.

For the second analysis, we turn our attention to the time it takes to remediate a site and estimate a Cox proportional hazard model

Table 2 Summary Statistics

Site Level State Level

Mean SD Mean SD

Dependent variables Privately owned site .660 .474 .576 .326 Future use plan .087 .282 .126 .229 NFA .179 .384 .340 .336 Average time to NFA (days) 838 753 — — Independent variables Cleanup grants .524 .499 .534 .500 Cleanup bonds .063 .243 .065 .247 Cleanup loans .806 .395 .691 .463 Tax abatement .216 .412 .219 .414 Tax credits .543 .498 .488 .501 Environmental insurance .146 .353 .117 .322 Retroactive liability protection .173 .378 .111 .315 Third-party liability protection .168 .374 .086 .281 Public comment periods .655 .475 .664 .473 Citizen grants .161 .367 .086 .281 Public hearings .671 .470 .590 .493 Post-cleanup monitoring .521 .500 .503 .501 Post–federal reform (cleanup after 2002) .975 .154 .861 .346 Average years since state reform 7.84 4.52 6.72 4.57 Control variables Industrial site .136 .343 — — Additive pollution index 1.31 2.08 — — Size of site (acres, logged) –.014 1.87 — — African American proportion of tract .270 .339 .248 .264 Hispanic proportion of tract .133 .191 .100 .141 Proportion on public assistance .154 .102 .140 .065 Median home value (logged) 11.12 1.34 11.02 1.21 Vacancy rate .121 .094 .123 .065 Owner occupancy rate .444 .217 .469 .145

Notes: n = 6,932 sites for all site-level variables; n = 324 state-years for all state- level variables.

258 Public Administration Review • March | April 2015

federal reform were about 64 percent more likely to be cleaned and were cleaned more quickly—moving forward in months from the time remediation began, a postreform site was more than twice as likely to be completed in a given month than a prereform site. Th e length of time since state reforms were implemented decreases the probability of cleanup but increases the rate of cleanup—however, in both of these cases, the hazard ratios, while signifi cantly diff erent from 1, are practically of minimal magnitude. Grants and bonds are

associated with a higher probability of cleanup and with the rapidity of cleanup along with tax incentives. Retroactive liability protec- tion is associated with higher probability of cleanup, but third- party liability protection is associated with decreased probability of cleanup and a slower pace getting to that point.

State-level results are shown in table 4. At this aggregate level, we fi nd little substantive relationship between policy instruments and

Table 4 State-Level Panel Regression Results

Dependent Variable

Private Ownership Future Use Plan NFA

a SE a SE a SE

Independent variables Cleanup grants –.104* .049 .003 .029 .066 .041 Cleanup bonds –.257* .090 –.074 .057 .092 .115 Cleanup loans –.023 .059 –.002 .040 .042 .047 Tax abatement –.014 .058 .001 .036 .047 .042 Tax credits .133* .055 .016 .038 –.046 .051 Environmental insurance –.021 .066 .156* .042 –.019 .074 Retroactive liability protection –.149* .072 .091 .050 .115 .074 Third-party liability protection .003 .087 –.050 .034 –.027 .058 Public comment periods .096 .058 –.076 .035 .065 .046 Citizen grants .065 .076 –.065 .056 –.023 .095 Public hearings –.027 .048 .029 .027 –.167* .046 Post-cleanup monitoring .043 .046 .012 .031 –.097* .041 Post–federal reform (cleanup after 2002) –.009 .076 –.002 .045 –.155* .064 Average years since state reform .012 .007 .001 .004 –.017* .006 African American proportion of tract .060 .122 .029 .084 –.210* .091 Hispanic proportion of tract –.175 .229 –.015 .076 –.447* .153 Proportion on public assistance .497 .490 –.712* .321 .196 .489 Median home value (logged) .025 .021 –.007 .012 .029 .018 Vacancy rate –.888* .341 .083 .201 –.470 .334 Owner occupancy rate –.457* .206 .078 .185 –.270 .183 Future use plan — — — — .209 .117 Private ownership — — — — –.247* .070

Notes: n = 324; * p < .05; Wald tests for all models p < .05.

Table 3 Site-Level Logistic Regression Results

Dependent Variable Private Ownership Future Use Plan NFA Time to NFA in Months (Cox model)

Odds ratio SE Odds ratio SE Odds ratio SE Hazard ratio SE Independent variables Cleanup grants .642* .051 .920 .147 1.61* .187 1.41* .129 Cleanup bonds .170* .033 .238* .097 3.29* .876 1.76* .400 Cleanup loans .768* .077 1.25 .248 1.11 .156 .848 .093 Tax abatement .687* .059 1.77* .238 1.15 .140 1.27* .129 Tax credits 1.85* .162 2.01* .322 .918 .115 1.27* .126 Environmental insurance 1.23* .137 3.22* .501 .876 .142 .701* .093 Retroactive liability protection .631* .094 1.42 .363 2.89* .555 1.11 .158 Third-party liability protection .943 .109 .365* .091 .423* .084 .585* .091 Public comment periods .935 .096 .301* .062 1.21 .181 1.06 .134 Citizen grants 1.04 .170 .999 .291 .340* .077 1.11 .192 Public hearings 1.03 .099 2.22* .425 .500* .071 .779* .092 Post-cleanup monitoring 1.35* .079 .714* .092 .996 .105 .860* .069 Post–federal reform (cleanup after 2002) 1.18 .227 1.25 .406 1.64* .116 2.01* .387 Average years since state reform .972* .008 1.00 .015 .958* .011 1.02* .010 Industrial site 1.13 .097 3.16* .344 1.02 .118 1.39* .139 Additive pollution index .859* .012 1.21* .023 1.69* .034 1.32* .017 Size of site (acres, logged) .910* .015 .936* .027 1.04 .025 .962* .019 African American proportion of tract .586* .072 .366* .085 .640* .128 .819 .128 Hispanic proportion of tract 2.43* .467 .486* .174 .420 .114 .581* .121 Proportion on public assistance .346* .146 .285 .200 .130* .089 .206* .103 Median home value (logged) 1.04* .023 .918* .032 .901* .030 .949* .020 Vacancy rate .239* .078 4.48* 2.25 .369* .193 1.26 .495 Owner occupancy rate .704 .137 1.59* .531 .865 .259 .987 .241 Future use plan — — — — 2.25* .290 1.69* .176 Private ownership — — — — .620* .051 .848* .057

Notes: n = 6,932; *p < .05; Wald tests for all models p < .05; fi xed effects by state and year included in all models but excluded from table

Public Incentives, Market Motivations, and Contaminated Properties: New Public Management and Brownfi eld Liability Reform 259

right, which means determining how best to institutionalize risk. A public–private partnership with government absorbing the liability risk may be the most eff ective way to do so if the objective is to promote private sector involvement.

Turning to the broader public goals of remediation beyond pri- vate sector involvement, we see the redevelopment for these sites as being in line with the propositions made by Sherman (2003) and McCarthy (2002, 2009), who argue that private interests will

become involved when they perceive eco- nomic promise regardless of public incentives. Private involvement is generally more likely in communities with higher home ownership, a large supply of vacant houses, comparatively lower poverty, and smaller minority popula- tions (particularly African Americans)—all conditions that have been found to be associ- ated with gentrifi cation (Hamnett 1991). Developers are not typically the fi rst investors in a neighborhood (Ley 1986); they tend to follow trends that have already begun and

look for areas with high potential value (Smith 1987). High vacancy rates along with high ownership rates may indicate transitional com- munities—the very sort of area that may be appealing as a poten- tially profi table investment regardless of public incentives.

Part of the rationale for the 2002 federal law was to encourage rede- velopment in those communities most in need (Dull and Wernstedt 2010); reforms were intended to provide incentives to redevelop poor and minority communities. Th at redevelopment plans in such communities are rare is not surprising (Daley and Layton 2004), but it is concerning given the policy goal of redeveloping areas in need. What we see in our results is not a broad-based collabora- tion between the public sector and private developers to develop areas that need redevelopment but rather a set of incentives that provide marginal support to the redevelopment of sites that were likely prone to redevelopment anyway, and little evidence of a focus on remediation for communities where economic development is needed.

However, in concluding our interpretation of the results, we must discuss three data limitations. First, on the aggregate scale of our data, we know very little about local socioeconomic conditions in communities. Brownfi elds are local, and assessing economic conditions at higher geographic units would be a mismatch with the neighborhood focus of sites; however, tract-level economic conditions are only assessed decennially, are not standardized geographically or temporally, and do not capture the nuances that give insight into the relative desirability of a project. We believe that local economic conditions are among the key drivers, and although we believe we have captured relevant proxy measures, it would be preferable to have more localized (and frequent) assessments of local market conditions.

Second, we know little about the political environment in these communities beyond broad (and potentially fl awed) generalizations that can be drawn from demographic characteristics. We also do not know the extent of local incentives for redevelopment or the more informal mechanisms and collaborations that occur. Th e extent to

broad levels of either private sector involvement or cleanup. Grants, bonds, and liability protection are associated with a low proportion of sites that are privately owned in a state, suggesting that at best these incentives may not have their intended eff ect, while tax credits are associated with higher private ownership, and environmental insurance is associated with future use planning. Public hearings and monitoring after cleanup are associated with lower proportions of sites ultimately being cleaned, and market reforms are not associated with the probability of cleanup, although this result should be con- sidered in the context that the number of sites receiving grants has increased over time.

Site-level results in table 3 indicate that future use is much more likely, and cleanup is faster, when a site is located in an industrial zone, perhaps suggesting that private interests may be more involved when reuse of an industrial site is the goal rather than redevelopment for other uses (Howland 2010). Private owner- ship is less likely when more pollutants are found on a site, but both future use plans and cleanup are more likely when more pollutants are found on a site. At both the state and site levels, private involvement and site cleanup become less likely as the proportion of African Americans in the neighborhood increases. Further, although private ownership of sites increases with the proportion of Hispanic residents, future use plans decrease, and the time to cleanup increases. Private owner- ship and site cleanup are also less likely and cleanup takes longer as poverty in the neighborhood increases.

Public Incentives and Market Motivations Our aim in this research was twofold: to assess the eff ectiveness of liability reform and incentives on the extent to which develop- ers’ revealed actions regarding brownfi eld site liability reform were consonant with their stated preferences (Wernstedt, Meyer, and Alberini 2006) and to consider whether the managerial focus under- lying these reforms enabled administrators to achieve broad policy goals. With the set of brownfi elds evaluated in this study, there does appear to be some evidence of congruence between stated and revealed preferences of private developers, at least in terms of certain types of monetary incentives. It is diffi cult for us to parse the actual motivation of the developers, but we can safely say that there is an overall trend of more sites entering the remediation process, more future use planning, and at least some evidence of increased private sector interest.

Th is trend could be partly attributable to market reforms, as sug- gested by our results; however, we see relatively limited overall impact. Tax incentives and environmental insurance appear to be important, particularly with respect to future use planning, but neither liability reform nor fi nancial assistance is associated with a higher likelihood of private sector interest; in fact, these incentives may even work against private involvement, but at the very least, the incentives have not achieved their intended goal. Th e fi nding that developers appear swayed by insurance but not further liability pro- tection may suggest that they remain risk averse to liability issues. Th ey may perceive government ownership and insurance as less risky options than reliance on courts to interpret liability statutes. Part of getting the incentives right is also getting the disincentives

Private involvement is gener- ally more likely in communities with higher home ownership, a large supply of vacant houses, comparatively lower poverty, and smaller minority popula-

tions (particularly African Americans).

260 Public Administration Review • March | April 2015

we see much room for improvement. Our fi ndings underscore the growing attention to a more multifaceted approach to adminis- tration and policy that conceives of governance as collaboration between government and the public sphere (Bryson, Crosby, and Bloomberg 2014; Denhardt and Denhardt 2003). Th e implica- tions of shifting attention from steering to serving, as Denhardt and Denhardt note, will have profound eff ects on designing policies to develop brownfi elds.

Notes 1. Th e U.S. Government Accountability Offi ce estimated in 2004 that there were

between 450,000 and 1,000,000 brownfi eld sites nationwide, although this is probably a conservative estimate.

2. Th e term “brownfi eld site” is defi ned as “real property, the expansion, redevelop- ment, or reuse of which may be complicated by the presence of potential pres- ence of a hazardous substance, pollutant, or contaminant” (42 U.S.C. § 9601).

3. See EPA (2011). 4. For more information, see the Brownfi elds Center at the Environmental Law

Institute at http://www.brownfi eldscenter.org. 5. We operationally defi ned this variable in a variety of ways, including in a similar

manner as Eckerd and Keeler (2012), through individual indicator variables, and through exploratory factor analysis. However, without specifi c knowledge of the full extent of the pollution, we opted for the simple operational defi nition used here. Substantive results regardless of how we included the pollution variable(s) were the same.

6. We opted not to run these analyses as two-stage (or Heckman) models for two reasons: fi rst, because we consider the models to be separate attempts to under- stand a concept that is diffi cult to measure, and second, because Hausman tests did not indicate any diff erence in coeffi cients if the models were run separately or as part of a system of equations.

7. We also estimated models using seemingly unrelated probit analysis to account for potential endogeneity with respect to error structures given common covari- ates. Hausman tests did not indicate a change in coeffi cients for any models whether run independently or as a system of equations, and therefore we ran each model independently. Further, given the low proportion of sites that had future use plans in place, we also estimated a rare events probit (scobit) for model 2. Given the similarity of the results of both models, we opted to report the logistic regression results for consistency.

References Alberini, Anna, Alberto Longo, Stefania Tonin, Francesco Trombetta, and Margherita

Turvani. 2005. Th e Role of Liability, Regulation and Economic Incentives in Brownfi eld Remediation and Redevelopment: Evidence from Surveys of Developers. Regional Science and Urban Economics 35(4): 327–51.

Box, Richard C. 1999. Running Government Like a Business: Implications for Public Administration Th eory and Practice. American Review of Public Administration 29(1): 19–43.

Bryson, John M., Barbara C. Crosby, and Laura Bloomberg. 2014. Public Value Governance: Moving beyond Traditional Public Administration and the New Public Management. Public Administration Review 74(4): 445–56.

Daley, Dorothy M., and David F. Layton. 2004. Policy Implementation and the Environmental Protection Agency: What Factors Infl uence Remediation at Superfund Sites? Policy Studies Journal 32(3): 375–92.

DeLeon, Linda, and Robert B. Denhardt. 2000. Th e Political Th eory of Reinvention. Public Administration Review 60(2): 89–97.

Denhardt, Janet V., and Robert B. Denhardt. 2003. Th e New Public Service: Serving, Not Steering. 3rd ed. Armonk, NY: M. E. Sharpe.

DePass, Michelle. 2006. Brownfi elds as a Tool for the Rejuvenation of Land and Community. Local Environment 11(5): 601–6.

which local governments are supportive of redevelopment eff orts and have the resources and political capital to collaborate with the private sector is unknown. Finally, we know very little about the private interests involved. While one might presume that rational developers are aware of and take advantage of all the incentives available, there is no evidence that this is the case. Nor is there any clear indica- tion as to what these fi rms do. Th e best information available, in an aggregate sense, is the Wernstedt, Meyer, and Alberini (2006) sur- vey—and they are careful to note the nonrepresentativeness of their sample. If we take a strict rationality view of developers, there is little reason to think that incentives would encourage a fi rm to redevelop a site that it otherwise would not have (given market conditions) (Sherman 2003), unless those incentives were substantive enough to make an unprofi table redevelopment profi table. We are skeptical that government incentives can reach this threshold.

Conclusion Following the 2002 federal legislation amending CERCLA to provide more liability relief to owners of brownfi eld sites, there does appear to have been an increase in private sector involvement. However, the reason for this involvement is unclear. Th e policy instruments of liability relief and redevelopment incentives align with the stated preferences of developers and the NPM philosophy. Indeed, liability relief in the form of insurance seems to be related to an increase in private sector involvement. Other incentives, such as tax credits, also appear to be related to an increase in private sector involvement, but at the state level, further liability relief is an insig- nifi cant factor in the likelihood of such involvement.

It may be the case that a developer prefers liability relief—after all, it reduces uncertainty and allows for a more accurate pricing of the development—but this relief may simply make attractive projects more attractive and do little to promote activity in underdeveloped sites. Th is may help explain the increase in private sector involve- ment, which raises a reasonable question about the misalignment of policy goals and management. Liability relief was thought to promote development in those areas that had been neglected by developers in the past because, it was assumed, the factor prohibit- ing development was an insuffi cient array of market incentives. Since 2002, there has been an increase in private sector involve- ment, but that increase has been concentrated where we would anticipate development anyway. Th is raises the question of whether those sites would have been developed without the liability reform, or at least what the marginal eff ect of liability reform actually is and where that margin is most pertinent.

Th e suggestion from this fi nding is that liability relief has the eff ect of benefi ting developers while doing little to benefi t underdeveloped communities. Th e interdependent nature of policies and man- agement frameworks suggests that the broader public interest of reducing risk and providing economic development to communities where such development is needed is less managerially salient than aligning market incentives with private sector interests. Liability reform does indeed lead to an increase in private sector involvement, but the extent and quality of that involvement is not, perhaps, what was intended by the policy instrument. We acknowledge that increased private sector involvement can appropriately be considered a good outcome, but if we care about the equity of environmental improvement and about what priorities drive administrative focus,

Public Incentives, Market Motivations, and Contaminated Properties: New Public Management and Brownfi eld Liability Reform 261

Leyden, Kevin M., Abraham Goldberg, and Philip Michelbach. 2011. Understanding the Pursuit of Happiness in Ten Major Cities. Urban Aff airs Review 47(6): 861–88.

May, Peter J., and Søren C. Winter. 2009. Politicians, Managers, and Street- Level Bureaucrats: Infl uences on Policy Implementation. Journal of Public Administration Research and Th eory 19(3): 453–76.

McCarthy, Linda. 2002. Th e Brownfi eld Dual-Use Policy Challenge: Reducing Barriers to Private Redevelopment While Connecting Reuse to Broader Community Goals. Land Use Policy 19(4): 287–96.

———. 2009. Off the Mark? Economic Effi ciency in Targeting the Most Marketable Sites Rather than Spatial and Social Equity in Public Support for Brownfi eld Redevelopment. Economic Development Quarterly 23(3): 211–28.

Meyer, Peter B., and Th omas S. Lyons. 2000. Lessons from Private Sector Brownfi eld Redevelopers: Planning Public Support for Urban Regeneration. Journal of the American Planning Association 66(1): 46–57.

Osborne, David, and Ted Gaebler. 1992. Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector. Reading, MA: Addison-Wesley.

Sherman, Scott. 2003. Government Tax and Financial Incentives in Brownfi elds Redevelopment: Inside the Developer’s Pro Forma. New York University Environmental Law Journal 11(2): 317–71.

Smith, Neil. 1987. Gentrifi cation and the Rent Gap. Annals of the Association of American Geographers 77(3): 462–65.

Stavins, Robert N. 2001. Lessons from the American Experiment with Market-Based Environmental Policies. Discussion Paper no. 01-53, Resources for the Future. http://www.rff .org/Documents/RFF-DP-01–53.pdf [accessed November 14, 2014].

U.S. Environmental Protection Agency (EPA). 2009. State Brownfi elds and Voluntary Response Programs: An Update from the States. Washington, DC: U.S. Government Printing Offi ce. EPA-560-F-09-522.

———. 2011. State Brownfi elds and Voluntary Response Programs: An Update from the States. Washington, DC: U.S. Government Printing Offi ce. EPA-560-R-11-001. http://epa.gov/brownfi elds/state_tribal/update2011/bf_states_report_2011.pdf [accessed November 17, 2014].

U.S. Government Accountability Offi ce. 2004. Brownfi eld Redevelopment: Stakeholders Report Th at EPA’s Program Helps to Redevelop Site, but Additional Measure Could Complement Agency Eff orts. Washington, DC: U.S. Government Printing Offi ce. GAO-05-94.

Wernstedt, Kris, Peter B. Meyer, and Anna Alberini. 2006. Attracting Private Investment to Contaminated Properties: Th e Value of Public Interventions. Journal of Policy Analysis and Management 25(2): 347–69.

Wyly, Elvin K., and Daniel J. Hammel. 1999. Islands of Decay in Seas of Renewal: Housing Policy and the Resurgence of Gentrifi cation. Housing Policy Debate 10(4): 711–71.

Dull, Matthew, and Kris Wernstedt. 2010. Land Recycling, Community Revitalization, and Distributive Politics: An Analysis of EPA Brownfi elds Program Support. Policy Studies Journal 38(1): 119–41.

Eckerd, Adam. 2011. Cleaning Up without Clearing Out? An Assessment of Environmental Gentrifi cation. Urban Aff airs Review 47(1): 31–59.

———. 2013. Policy Alternatives in Adaptive Communities: Simulating the Environmental Justice Consequences of Hazardous Site Remediation Strategies. Review of Policy Research 30(3): 281–301.

Eckerd, Adam, and Andrew G. Keeler. 2012. Going Green Together? Brownfi eld Remediation and Environmental Justice. Policy Sciences 45(4): 293–314.

Ellen, Ingrid G., and Katherine M. O’Regan. 2011. How Low Income Neighborhoods Change: Entry, Exit, and Enhancement. Regional Science and Urban Economics 41(2): 89–97.

Fitzgerald, Joan, and Nancey G. Leigh. 2002. Th e Brownfi eld Redevelopment Challenge. In Economic Revitalization: Cases and Strategies for City and Suburb, edited by Joan Fitzgerald and Nancey G. Leigh, 69–102. Th ousand Oaks, CA: Sage Publications.

Gayer, Ted, James T. Hamilton, and W. Kip Viscusi. 2000. Private Values of Risk Tradeoff s at Superfund Sites: Housing Market Evidence on Learning about Risk. Review of Economics and Statistics 82(3): 439–51.

Gayer, Ted, and John K. Horowitz. 2006. Market-Based Approaches to Environmental Regulation. Foundations and Trends in Microeconomics 1(13): 1–128.

Grindle, Merilee S. 2011. Governance Reform: Th e New Analytics of Next Steps. Governance 24(3): 415–18.

Hamnett, Chris. 1991. Th e Blind Men and the Elephant: Th e Explanation of Gentrifi cation. Transactions of the Institute of British Geographers 16(2): 173–89.

Hird, John. 1993. Environmental Policy and Equity: Th e Case of Superfund. Journal of Policy Analysis and Management 12(2): 323–43.

Hite, Diane, Wen Chern, Fred Hitzhusen, and Alan Randall. 2001. Property-Value Impacts of an Environmental Disamenity: Th e Case of Landfi lls. Journal of Real Estate Finance and Economics 22(2–3): 185–202.

Hood, Christopher. 1991. A Public Management for All Seasons? Public Administration 69(1): 3–19.

Howland, Marie. 2000. Th e Impact of Contamination on the Canton/Southeast Baltimore Land Market. Journal of the American Planning Association 66(4): 411–20.

———. 2010. Th e Private Market for Brownfi eld Properties. Cityscape 12(3): 37–54. Jaff e, Adam, Steven R. Peterson, Paul R. Portney, and Robert N. Stavins. 1995.

Environmental Regulation and Competitiveness of U.S. Manufacturing: What Does the Evidence Tell Us? Journal of Economic Literature 33(1): 132–63.

Ley, David. 1986. Alternative Explanations for Inner-City Gentrifi cation: A Canadian Assessment. Annals of the Association of American Geographers 76(4): 521–35.

Copyright of Public Administration Review is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.