«Toward More Effective Endangered Species Regulation By Jacob P. Byl Dissertation Submitted to the Faculty of the Graduate School of Vanderbilt ...»
In this paper, I look at how landowners respond when faced with various choices about whether to conserve or develop the natural resources on their lands. Using tree harvest data gathered by federal agencies, I analyze whether landowners are systematically destroying habitat of an endangered bird, the red-cockaded woodpecker (RCW). This paper contributes to the literature by controlling for unobserved heterogeneity in forest plots with a difference-in-difference model and using recent data to estimate current behavior. I find evidence of preemptive habitat destruction with a 25% increase in the probability of harvest for land near endangered birds. This paper also contributes to the literature with the first measurements of the effectiveness of a federal safe-harbor program designed to mitigate the incentive to destroy habitat. I find mixed evidence on the program’s success at getting landowners who participate in the program to destroy less potential habitat and to manage land in ways that improve wildlife habitat.
The paper proceeds in Section II with a survey of the previous literature on perverse incentives of the ESA and attempts to measure effectiveness of ESA policies.
Section III presents a brief description of the legal background that creates perverse incentives to destroy habitat for endangered species. In Section IV, I provide background information on the red-cockaded woodpecker and the safe-harbor program designed to help protect these birds. Section V contains a conceptual model that provides predictions for how the risk of endangered species regulation will affect decisions of private landowners to enter safe-harbor agreements and/or harvest trees. Section VI introduces the data on forest plots, woodpeckers, and timber prices. Empirical models to test the predictions from the theoretical model are laid out in Section VII. Section VIII provides results from these models, which indicate an increase in harvests near woodpecker colonies. There is not strong evidence that the safe-harbor program has succeeded at preventing habitat destruction, although there are indications that the program has encouraged more landowners to improve wildlife habitat. In Section IX, I discuss how this better understanding of the way landowners interact with federal agencies can suggest ways to improve the effectiveness of endangered species regulation on private lands. Section X concludes.
II. Existing Literature Numerous scholars have noted that the Endangered Species Act may create perverse incentives. In the legal literature, Dana (1995) identifies the general incentive created by the ESA to engage in a “race to develop,” but does not explore how widespread this phenomenon may be. Rachlinski (1997) discusses how endangered species that face economic pressures from conflicting use of resources are less likely to see improvements in recovery status. Ruhl (1998) explores the middle-ground of regulation under the ESA, between the poles of strict land-use regulations and no regulations, in which agencies and landowners can cooperate to protect habitat.
In the economics literature, theorists and empiricists have looked at the possibility of preemptive habitat destruction. Shogren (1998) presents a theoretical model that predicts landowners will engage in purposeful habitat destruction if the threat of Fish & Wildlife Service (FWS) regulation is high enough. He argues that the FWS should actively consider the economic incentives created by its policies when creating recovery plans for species and engaging in enforcement actions. Polasky and Doremus (1998) model the role of asymmetric information between landowners and regulators to show that compensation may be required to get private landowners to cooperate with regulators, and even then the efficient outcome is not assured. Smith and Shogren (2002) use tools from mechanism design to explore how regulators may get landowners to better cooperate, and caution that under many conditions regulators will have to offer high incentives to private landowners to get them to protect endangered species habitat.
Lueck and Michael (2003) and Zhang (2004) both empirically test whether landowners are changing their timber harvest patterns in response to the threat of ESA regulations. Zhang uses a survey to collect information on landowners’ subjective beliefs about the presence of nearby RCWs and the threat of ESA regulation if those birds move onto their property. The landowners are also asked about their timber harvest patterns over the past ten years. Their responses indicate that landowners who have RCWs nearby and believe that the birds could move onto their property are more likely to use a shorter harvest rotation, thus preventing the birds from moving in.
Lueck and Michael attempt to answer a question similar to that asked by Zhang but using data on RCWs and timber harvests that come from federal agencies instead of from landowner surveys. Lueck and Michael find that a higher number of RCW colonies near a forest plot increases the probability that the plot will be harvested within a given period. This paper uses the Lueck and Michael analysis as a starting point for models of tree harvests near endangered species to test whether habitat destruction continues at a similar pace after the introduction of a safe-harbor program designed to help prevent preemptive habitat destruction.
Empirical tests of the impact of endangered species regulations can be difficult because it is hard to obtain good data on both outcomes and explanatory variables.
Despite these challenges, some researchers have been able to empirically measure the effectiveness of some FWS actions under the ESA. Ferraro, McIntosh, and Ospina (2007) use matching methods to measure the effectiveness of listing species as threatened or endangered under the ESA. They find that listing species can be detrimental to species, presumably because of the perverse incentives that listing triggers, unless the listing is accompanied with government funding to help conserve the species. Langpap and Kerkvliet (2012) use a discrete ordered-variable model and matching methods to evaluate the effectiveness of habitat conservation plans, which are plans the ESA requires to be in place for landowners to be able to get permits for the incidental take of species. Langpap and Kerkvliet find that species tend to do better when there are habitat conservation plans in place for them, especially when those plans cover relatively large geographic areas.
No study to date has measured the effectiveness of safe-harbor programs, although researchers have used theory and survey data to better understand how landowners behave in voluntary programs. Langpap and Wu (2004) use a model to assess under what conditions incentives like money and assurances of no additional regulations are more effective at getting landowners to protect habitat than traditional regulations.
They find that voluntary programs that offer assurances tend to outperform programs without assurances, although the resulting equilibrium is still sub-optimal. Langpap (2006) surveys forest owners to find that the most promising incentive program is a mix of both financial incentives and regulatory assurance for landowner cooperation.
Greenstone and Gayer (2009) discuss the challenge of non-random distribution of species when trying to measure impacts of the ESA on landowner behavior. With species clustered in prime habitat areas, there is likely to be unobserved heterogeneity across sites that have species compared with sites that do not. Greenstone and Gayer propose using indices of species rareness separate from status under the ESA to help measure the causal impact of ESA protections. As discussed in Section VII, I take a different approach and attempt to control for unobserved heterogeneity with difference-in-difference models when measuring landowner response to the perverse incentives in the ESA and the effectiveness of the FWS safe-harbor program.
III. Legal Background The ESA was passed in 1973 to conserve “ecosystems upon which endangered species... depend.” 16 U.S.C. § 1531(b). The power of the ESA has been prominently displayed in cases like Tennessee Valley Authority v. Hill, 437 U.S. 153 (1978), in which a dam-building project was halted because it would damage habitat for the snail darter, an endangered fish. The ESA gives extensive authority to the FWS, a federal agency, to promote the purpose of the ESA by restricting actions of the government and of private landowners to prevent the “take” of listed species. The ESA defines “take” as “to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or to attempt to engage in any such conduct.” § 1532(19). In Palila v. Hawaii, 852 F.2d 1106 (9th Cir. 1988), the Ninth Circuit held that it was within the FWS’s authority to force the state of Hawaii to eradicate feral animals because those feral animals were causing “harm” to listed species by destroying habitat. In Babbitt v. Sweet Home, 515 U.S. 687 (1995), the U.S Supreme Court upheld the FWS definition of “harm” that includes “significant habitat modification or degradation that actually kills or injures wildlife.” Id. at 798. As such, the FWS has jurisdiction to regulate land uses to prevent the destruction of habitat and can impose civil and criminal penalties on landowners who harm animals by disobeying the regulations. In Arizona Cattle Growers Association v. United States Fish & Wildlife, 273 F.3d 1229 (9th Cir. 2001), the Ninth Circuit struck down FWS restrictions on cattle grazing leases that were based on the historical presence of endangered species or the chance that endangered species would move onto the properties in question. Id. at 1243.
Instead, the agency had to show with “reasonable certainty” that actual harm to endangered species would occur from the habitat modifications. Id. In the same case, the court upheld restrictions on properties for which the FWS had documented the presence of endangered species. Id. at 1248.
Courts have consistently held that FWS regulation does not constitute a taking of property that would require just compensation under the Fifth Amendment of the U.S.
Constitution. See Seiber v. United States, F.3d 1356 (Fed. Cir. 2004). Actors who engage in activities that may cause the “take” of listed species can create a habitat conservation plan and apply to the FWS for an incidental take permit, but this is an expensive and lengthy process.
The stark contrast between the regulatory impacts of having established populations of endangered species or not has the potential to create three perverse incentives. First, landowners may be induced to destroy species or their habitat before the species are listed under the ESA. Second, landowners who have potential habitat of already listed species may destroy that habitat to prevent the protected species from moving onto their properties. Third, landowners who have listed species on their properties may illegally kill the species and hope that the FWS never finds out about it. In response to these concerns, the FWS has recently encouraged landowners to enter voluntary conservation agreements to obtain more regulatory certainty in exchange for the provision of quality habitat. To dampen the incentive to destroy species that are not yet listed under the ESA—candidate species—the FWS offers landowners a promise not to impose regulatory restrictions beyond those agreed to by landowners in negotiated agreements called candidate conservation agreements with assurances. To dampen the incentives to destroy habitat of listed species and illegally kill the species themselves, the FWS offers landowners safe-harbor agreements that promise not to increase regulatory burdens beyond those agreed to in these negotiated agreements. FWS currently has safeharbor programs for a range of species, including the woodpecker introduced in the next section that is used to help measure responses to perverse incentives in the ESA.
IV. Red-Cockaded Woodpecker The red-cockaded woodpecker (RCW) is a robin-sized bird that inhabits mature pine stands across the southeastern United States from North Carolina to Texas. RCWs are unique among North American woodpeckers in that they excavate their nesting cavities in live trees instead of dead trees. Excavating a hole in a live tree means that it takes a RCW several months to finish a cavity, but it allows the RCW to use a cavity for many years. The only trees that are suitable for these cavities are pines that are at least forty years old. RCWs make cavities in clusters called colonies. A colony typically consists of a mating pair, their immature young, and some helper adults. When young RCWs reach maturity and find a mate, they start a new colony in a stand of mature pine somewhere within ten miles of their parents’ colony. This means that property with mature pine within ten miles of existing RCW colonies is potential habitat for a new colony. Landowners of properties in this area with pine trees that are not yet mature may also pay attention to the birds because their lands may become attractive habitat for RCWs as the trees age, especially for landowners following traditional harvest rotations of forty to fifty years.
RCWs require open mid-story for effective foraging. Stands of southern pine are suitable for this because they have substantial space between trees, and undergrowth is discouraged by a thick bed of pine needles and periodic low-intensity forest fires. Stands of pine mixed with hardwood trees are not suitable for RCW foraging because the hardwoods add too many branches and leaves in the mid-story so the RCWs cannot effectively catch insects. The optimal foraging area for an RCW colony is 300 acres, although scholars have found colonies to survive with as little as 150 acres (U.S. FWS 1985).
RCWs have many characteristics that make them an attractive species to study for evaluating the impacts of the ESA on landowner behavior. First, RCWs are picky about their nesting and foraging sites, so it is relatively easy to isolate potential habitat. Second, RCWs are the only listed endangered animal to inhabit wide areas of the pine forests of North Carolina, so it is easier to isolate the impact of these birds as opposed to endangered species in general.1 Third, there is a clear conflict with maintaining habitat for RCWs and harvesting timber, which allows me to look at the impact of the species on an activity that has established economic models for landowner behavior.