Takings, Compensation and Endangered Species Protection on Private Lands

by Robert Innes, Stephen Polasky, John Tschirhart
Takings, Compensation and Endangered Species Protection on Private Lands
Robert Innes, Stephen Polasky, John Tschirhart
The Journal of Economic Perspectives
Start Page: 
End Page: 
Select license: 
Select License


Takings, Compensation and Endangered Species Protection on Private Lands

Robert Innes, Stephen Polasky and John Tschirhart

onsenation of species on private land and the potential conflicts with

private property rights and economic activity are at the forefront of the

current debate on endangered species policy in the United States (An

derson, 1993,1994;Mann and Plummer, 1995).For example, as much as 75 percent

of the prospective habitat for the endangered red-cockaded woodpecker is in pri

vate pine forests in the southeastern United States (Welch, 1995). Habitat protec

tions for the northern spotted owl may restrict logging on hundreds of thousands

of acres of private forest in the Pacific Northwest (Sugg, 1994). Protecting salmon

and steelhead trout may impose restrictions on development and water use from

the Pacific northwest to California's Central Valley (Jewett and Brinkman, 1998).

California farmland is habitat to the endangered tipton kangaroo rat, the SanJoa

quin kit fox, and the blunt-nosed leopard lizard (Johnson, 1995). Prime private

development land in southern California is habitat for the California gnatcatcher

and other endangered species (Stevens, 1997).Indeed, more than half of the listed

endangered species have at least 80 percent of their habitat on private land (US.

Fish and Wildlife Service, 1997).In a study of endangered specieshotspots, Dopson

et al. (1997) found endangered species clusters in a number of areas in which

competing private land uses are very valuable, including Hawaii, coastal California,

and southern and central Florida.

The prospective costs of preserving the habitat of species on private land can

Robert Innes is Professor of Agricultural and Resource Economics, University of Arizona, Tucson, Arizona. Stephen Polasky is Associate Professor of Agn'cultural and Resource Ece nomics, Oregon State University, Cmallis, Oregon.John Tschirhart is Pfofessor of Economics, University of Wyoming, Laramie, Wyoming. Their e-mail addresses are <innes@ag. arizona.edu>, <[email protected]>, and <[email protected]>, respectively.

be high. Annualized estimates for preserving Florida panther habitat are $6 and $24 per acre for maintenance and acquisition, respectively, on 1,000,000 acres (Main and Roka, 1997). A preliminary habitat conservation plan in San Diego County would require purchase of 27,000 acres at an estimated cost of $260- $360 million (Stevens, 1997).

Such costs will exist regardless of whether government pays for the land di- rectly, or whether the costs are imposed on private landowners. Much of the polit- ical debate about how these costs should be imposed has centered on the two issues of "takings," where "taking" has an entirely different meaning in each case.

The first issue of "taking" arises under Section 9 of the Endangered Species Act, which makes it unlawful for private citizens to "take" any threatened or en- dangered species on their property. The Fish and Wildlife Service, which together with the National Marine Fisheries Service are responsible for listing species as endangered, has interpreted "take" to mean any action which injures or kills an endangered creature, or significantly adversely modifies the habitat of an endan- gered species.' The only exception to this prohibition occurs when an "incidental take permit" is granted. To obtain an incidental take permit, a landowner must first submit a habitat conservation plan that demonstrates that proposed activities "will not appreciably reduce the likelihood of the survival and recovery of the spe- cies" (16 U.S.C. 5 1539(a) (2) (b) (iv)). With an approved habitat conservation plan, activities can be allowed even if some harm to a listed species occurs. Interior Sec- retary Babbitt has focused on habitat conservation plans and the incidental take provision as the preferred means of regulating private land under the Endangered Species Act.

The second issue of "taking" arises under the Fifth Amendment to the U.S. Constitution, which ends with the words, "nor shall private property be taken for public use, without just compensation." Many scholars have debated the appropri- ate limits of this "takings" clause: Under what circumstances (and for what types of government actions) must compensation be paid? How much compensation must be paid? As it stands, the Endangered Species Act does not require compen- sation for private property owners who suffer financial losses from habitat protec- tions on their land. Current jurisprudence suggests that losses due to endangered species protections do not come under the Constitution's takings clause (Meltz, 1994). However, both law and jurisprudence are in a state of flux. A variety of legislative initiatives in recent years, some of which passed the full House of Rep- resentatives, have aimed at requiring compensation for government actions, in- cluding the Endangered Species Act, that diminished the value of a private property by a minimal percentage, typically ranging from 20-33 percent.2

' This interpretation was recently upheld by the Supreme Court in Babbitt v. Sweet Home Chapter of Com- munitiesfor a Great Oregon, et al., 1995.

Specifically, the House takings bill, H.R. 925, passed the House of Representatives in March 1995, and would have required compensation when property values are diminished by at least 20 percent. The Senate bill, S. 605, while much more sweeping than H.R. 925 in other regards, set a 33 percent

In this paper, we focus on what light economics has to shed on the compen- sation debate. In regulating the "taking" of a species on private land, does govern- ment cross into a "taking" of property that should include compensation? Econ- omists often begin this debate with Coasian logic: compensation obligations depend on who has what property rights. Does the government have the right to take the property freely for public uses, or does the private owner have the right to use the property as he sees fit? If the government has the property right, a private owner will have to compensate the government for the lost public use value if the owner wishes to retain private use rights. If the private owner has the property right, the government will have to compensate the owner for the lost private use value if it wishes to obtain public use rights. Either way, the property will be put to its highest valued use. Ergo, the compensation debate affects distribution of who pays, but it is of no consequence to economic efficiency.

The contemporary economic thinking on takings and regulation begins by looking at pitfalls in the simple Coasian logic. There are problems of time and sunk costs: important long-lived decisions on private land use are made before public use decisions are made, and before negotiations between the government and a private owner may take place. There are problems of information, both about land use values and about actions that may affect the values. Striking the Coasian bargain is thus not straightforward. We will be concerned with a variety of incentive effects of compensation and regulatory policy, especially in settings where there may be sunk investments in private land and informational asymmetries between landown- ers and the government.

For the economic analyst, the compensation debate raises three central ques- tions: 1) Should "takings" compensation be paid? 2) If so, what form should the compensation take? 3)When is there a need forjudicial (or constitutional) restraint to ensure that the government has an incentive to choose efficient takings and compensation policies? We will develop answers to these questions and use them to discuss how to design endangered species regulation.

Compensation and Landowner Incentives

Perhaps the simplest approach to providing correct investment incentives to the landowner is to align private land-use incentives with societal economic inter- ests. In principle, this can be done through Pigouvian compensation, where land- owners are paid the public use value of their land whenever it is devoted to public

diminution-of-value standard. H.R. 2275, which passed the House Resources Committee in October of 1995, was aimed especially at the Endangered Species Act, and required compensation when an action under that act diminished private property value by 20 percent or more. More recently, in February 1998, a bill moved through the Senate Judiciar). Committee which would make it easier for landowners to challenge environmental and zoning laws in federal court. The bill is opposed by environmental groups, the National Governors Association, the National League of Cities and 38 state attorneys general.

use, or Pigouvian taxation, where landowners are taxed for the lost public use value of their land when they engage in private use (Hermalin, 1995; Innes, 1997a). Under either approach, landowners face the same net payoffs from their actions as society does, resulting in an efficient choice.

In the endangered species context, however, simple Pigouvian solutions are problematic. Their implementation requires assessing both the effect of private land uses on the survival probability of a listed species, and the existence value of the species, both of which are difficult to measure (Polasky and Doremus, 1998). Pigouvian compensation may require large budgetary costs. Moreover, in a context where there are sunk costs of investing in land improvements, and a variety of land parcels that might be chosen to support endangered species, and where the in- vestments may affect the value of the land as habitat for endangered species, and where information is imperfect or asymmetrically distributed between landowners and government, unsophisticated attempts at Pigovian taxation and compensation can readily become part of a complex game between landowners and government, in which endangered species are not especially likely to be the winners. Let us turn to these incentive problems.

A Model of Land Investment Incentives

As a starting point, let us consider a model which is a building block for much of the takings and compensation literature. This two-date analysis of a single parcel of land is due to Blume, Rubinfeld and Shapiro (1984). At the initial time, a forward-looking private owner makes a long-lived investment on the land. At the later time, the government discovers the value of the property in a public use; although this actual value was unknown at the initial time, there was at the initial time a known probability distribution for the possible public value of the land in the future. At the later time, the land can either be used privately or publicly, but not both. If the land remains in private use, the owner reaps a reward that rises with the owner's level of initial investment; otherwise, the private investment is rendered worthless. The government, striving to maximize society's economic wel- fare, takes the land for the public at the later period if and only if the public value is greater than the private value.

In this setting, the presence or absence of compensation will affect incentives for the private owner to invest in the property. If the landowner is not compensated when land is taken, then the return on investment will be an average of the return if the land stays private, and zero if there is a public taking. As a result, the owner will circumscribe investment to account for its potential loss in the event of a taking. On the other hand, if the landowner is fully compensated for a taking, the owner will always receive the private use value of the property, whether directly (the in- vestment pays off) or in compensation (when a taking occurs). In this case, the private returns to investment are higher than the true economic returns-since for purposes of optimal investment, the landowner should take into account the prob ability of the land being taken for public use. Full compensation thereby prompts excessive private investments. Compensation need not prompt overinvestment if the compensation is "lumpsum," which in this case means that the compensation does not vary according to the property's investmentdriven value and is determined only by Pigovian considerations. However, the practical design of this kind of lump sum compensation poses difficult issues of its own.

Habitat and Species Protection Incentives

The land-use and investment decisions of owners affect not only the potential private value of their property; their decisions can also affect potential conservation values. A variety of production practices or development activities are inconsistent with habitat preservation for endangered species. If landowners are not compen- sated when their land is taken for species preservation, they do not have an incentive to protect the species before the land it taken. Indeed, landowners may have in- centives to take actions that harm endangered species ifdoing so reduces the like- lihood that their land will be valuable as species habitat. For example, a landowner may cull old growth trees that can otherwise serve as home to endangered birds (Welch, 1995), or at an extreme, landowners can "shoot, shovel and shut-up" (Lambert and Smith, 1994) to assure that they won't be bothered by government protection of endangered species on their land.

When investment may affect both the private and conservation values of the land, and where no compensation is paid, landowners may invest inefficiently for two reasons. First, they have an incentive to reduce the risk of a taking, and the associated risk of financial loss, by increasing the private use value, or decreasing the potential conservation value, or both (Polasky, 1998). Second, since landowners do not reap any personal benefit from improving the conservation value, they lack any incentive to invest on behalf of such benefits.

However, designing compensation and/or tax policies to restore efficient land-use incentives involves some important subtleties (Innes, 1997b). Paying compensation equal to private value is inefficient because it fails to reward land- owners for improvements in the conservation value, as well as suffering from the overinvestment problem found in Blume, Rubinfeld and Shapiro (1984). On the other side, Pigouvian taxation-charging landowners for their land's op- portunity cost in public use when they retain the property in private use-only makes matters worse. With Pigouvian taxes, landowners will want to reduce the public value of the land for endangered species, and thereby reduce their ex- pected tax bills.

In principle, Pigouvian takings compensation-paying landowners the conser- vation value of their land-fully aligns private land-use incentives with societal ben- efits and costs. In practice, Pigouvian takings compensation requires an objective measure of the conservation value of the land. Developing estimates of such values are problematic on both biological and economic grounds (Polasky and Doremus, 1998). Moreover, when the government raises funds to pay compensation, there are deadweight economic losses due to both administrative costs of tax collection and economic costs of resulting distortions in labor, consumption, and investment decisions. Estimates of these costs range between 7 and 25 cents per dollar oftax revenue (Fullerton, 1991). In the presence of these deadweight costs, economic welfare will be increased if the incentive objectives of a compensation policy are achieved with low government budgetary exposure.

Efficient landowner behavior can be elicited at minimum compensation cost using a "negligence compensation" rule in the style of Miceli and Segerson (1994), under which compensation equal to the private land use value with efficient in- vestment is paid only when the landowner behaves efficiently (Innes, 199713). Such rules offer the promise that they can potentially elicit efficient choices at lower government budgetary cost than may otherwise be required. However, the potential scope for "negligence compensation" rules is limited by difficulties in observing all of the relevant decisions.

Development Incentives in a Multi-Parcel Model

Governments typically confront choices which are more complicated than sug- gested by the "take or not take" decision posed in the single-parcel model. Gen- erally, they face a set of land parcels on which heterogeneous private-use decisions have been made, and they decide not only how many parcels to take, but also which specific properties to take among those which can deliver comparable public ser- vices. Such multi-parcel considerations have important implications for the incen- tive effects of compensation (Innes, 1997a).

Consider another two-period model, this time of a region in which there is a set of homogeneous land parcels which can either be developed or not developed at either of two times. Because of increasing demand for private development, not all land is developed initially. Therefore, at the later time, the government faces two types of land that it can "take" for public use: that which was developed in the initial period, and that which was not. In choosing which land to take for public use, the government should not merely flip coins; rather, it should take the less valuable undeveloped land first.

With this optimal takings policy, however, landowners have a motive to de- velop in the initial time period if there is no takings compensation: by devel- oping early, they reduce the risk of losing their land to the government. With no compensation, landowners will therefore develop more land in the initial time period than is efficient. In the endangered species context, landowners who develop early reduce their risk of a taking both by making their property more costly to take, and by reducing the prospective public value of their prop- erty as species habitat.

Compensation for a taking can restore efficient development incentives by reducing the "use it or (maybe) lose it" motivation for overinvestment. However, if landowners are offered exactly full compensation for their lost private property value, then excessive development persists for the reason already developed in the Blume, Rubinfeld and Shapiro (1984) model. Instead of developing early to avoid a taking, landowners now overinvest because they are guaranteed full compensation for the developed value of their property; they have no incentive to take into ac- count in their investment decisions the probability that the land may turn out to be more valuable in a public use.%uch overdevelopment incentives can be coun- tered if owners of undeveloped land are offered more-than-full compensation for a taking, and owners of developed land are afforded exactly full compensation for a taking. By increasing the relative return to ownership of undeveloped land, this compensation policy reduces early development incentives.

Of course, such a compensation solution is very costly to the government- indeed, more costly than necessary to achieve efficiency. Efficient development incentives can be restored at zero government cost with a policy of tradeable de- velopment rights, under which each landowner is issued a fractional (and tradable) development permit at the initial time. To avoid a taking-and retain the right to hold developed land-a property owner must buy up tradeable development rights from other landowners. Those who sell their permits, and thereby give up their property for public use, are implicitly compensated by those who buy them. Early developers no longer reap excessive benefits from their reduced likelihood of a government taking, because by paying for tradeable development rights they share in the cost of a taking, even when their property remains in private use.

Risk Aversion and Takings Compensation as Insurance

The problem facing landowners who must consider the possibility of having land taken can be viewed as a question of risk and insurance. Government com- pensation implicitly provides landowners with free government insurance against the risk of loss from a taking. If private markets are not capable of providing insur- ance against government takings, and if the landowner is risk averse, the govern- ment is risk neutral, then risk-sharing-and hence, economic welfare-can be im- proved by the implicit insurance.

Private insurance markets against the risk of government takings may break down for several reasons (Blume and Rubinfeld, 1984). First, landowners may know more about the risk of a taking than an insurance company, leading to an adverse selection problem where only those property owners who are likely to suffer a taking decide to purchase such insurance. Second, private takings insurance would reduce the landowners incentives to lobby for government compensation, thus leading to a low level of compensation and high insurance premiums that would deter land- owners from purchasing policies.

However, there are powerful arguments why these reasons may not be sufficient to prevent a private market from functioning. For example, Farber (1992) and Kaplow (1986) dismiss the argument that landowners may know more about the risk of a taking primarily because it strains credulity to think that an insurance

" In fact, exactly full compensation also prompts excessive development even when there is no chance that developed land will be taken. There are cases in which the public use value of land is not high enough to justify the taking of any developed land, yet is more than high enough to justify the taking of all undeveloped land. For these cases, marginal development (at time zero) bears the societal cost of reducing the amount of land that \+ill be optimally appropriated for public use at time 1. Moreover, with full compensation, landowners do not circumscribe their investment to account for this cost (Innes, 1997b).

company with numerous policies at stake involving various takings situations would be less able to monitor the political situation than an individual resident. Moreover, as Kaplow and Farber point out, insurance companies may be at least as effective at political lobbying for government compensation as landowners, in which case political considerations would reinforce economic motivations for private takings insurance.

Information Acquisition and Regulation

Thinking of takings and compensation as an issue of insurance leads naturally to a consideration of the information requirements involved in takings and com- pensation. To enforce the prohibition against taking endangered species on private lands, regulators must be able to show that a listed species is present and will be harmed by a proposed land-use. By preventing access to the property, landowners may prevent regulators from gathering such information. Indeed, landowners have a strong incentive to prevent information collection unless the government com- pensates them when use of their property is restricted to conserve species.

Polasky and Doremus (1998) analyze a model with an information collection stage prior to the regulation of land use. Under the Endangered Species Act, where regulators must prove harm prior to restricting development, but do not pay com- pensation, landowners will not cooperate with information collection. Without co- operation, conservation will be difficult to mandate, and excessive development will occur. Two approaches can be used to restore incentives for landowner coopera- tion: paying compensation to landowners for lost development values, or instituting a permit scheme similar to habitat conservation plans that requires the landowner to prove that development will not harm a species before being allowed to develop. Neither approach alone generates incentives to make efficient information collec- tion and land use decision in all circumstances. However, Polasky and Doremus (1998) show that combining elements of both compensation and permitting can guarantee an efficient solution. The scheme gives property rights to landowners when the private value exceeds the expected conservation value and requires com- pensation if conservation is mandated. On the other hand, when the expected conservation value exceeds the private value, landowners must buy a permit whose price is set equal to the expected conservation value, before they are allowed to develop. This scheme works with imperfect information, but may generate an in- efficient solution when information is asymmetric.

Regulation Under Asymmetric Information

Asymmetric information between landowners and government further com- plicates the design of compensation policies to protect endangered species. To introduce the basic concept, consider an exam~le.~

Cities and counties have intro-

Smith and Shogren (forthcoming, p. 11) introduce a more complex version of the following example wherein species survival is greatly enhanced by preserving contiguous parcels from multiple private properties. The regulator designs contracts to retire acres and allows bonuses to promote agglomeration across landowners.

duced taxes whose sole purpose is to provide funding for presening undeveloped land; in Colorado alone, for example, at least ten cities and counties have open space sales taxes. Preservation is accomplished through purchasing either the land or the development rights to the land. In the latter case, the landowner retains the land but cannot develop it further. For instance, a rancher can continue working and living on the land, but the land must be maintained as a ranch.

For simplicity, suppose there are two landowners each with acres of land, any portion of which can be sold for development. Each acre has social value associated with preservation and development value which are common knowledge, and the social value exceeds the development value. The government can purchase the development value on a portion of land at a price above the development value but below the social value, and society would then enjoy gains. However, raising funds through taxes to pay landowners may be politically unpopular, as well as causing deadweight losses. Recognizing this, the government would like to pay the land- owners even less than the development rights, if possible. Since the landowners enjoy some value from keeping the land as it is and not developing, they are willing to take less than the development value per acre. But how much less they are willing to take is private information, and each landowner has an incentive to claim that they derive no pleasure from the undeveloped land at all. The government is as- sumed to know that one type of landowner, referred to as the preserver, places a higher total and marginal value on each acre retired from development than does the other type of landowner, referred to as the developer; but because these values are private information, the government does not know who is a preserver and who is a developer.

This situation can be viewed as a principal/agent problem, in which a principal must design an incentive instrument to elicit behavior from an agent, given that the principal and agent have different objectives and the agent has private infor- mation not available to the principal. In an excellent survey, Lewis (1996) extends the theory of principal/agent problems into environmental economics, offering a model of an environmental protection agency (principal) and polluting firms (agents) who have private information about their production and abatement costs, a model that could be converted easily to government and landowners who have private information about habitat protection costs.

In the case of this example, the government designs and offers a menu of contracts to the landowners, where each contract specifies a per acre payment and a number of undeveloped acres. The problem is subject to two types of constraints. First, the landowners must be willing to participate by selling development rights at the government's offered payments. Second, the contract must be incentive com- patible; that is, each landowner must prefer that person's own contract to the con- tract designed for the other type of landowner. Assuming a benevolent govern- ment,5 its objective is to maximize the sum of benefits from undeveloped and de-

'Laffont and Tirole (1991) examine a case where a government agency can be bribed by special interests, and this possibility limits the discretion granted to the agency.

veloped acres, minus the cost of payments to the landowners, plus the receipts of the landowners. The last two dollar amounts do not cancel because of the distor- tionary effects of raising public funds for the payments."

To appreciate the results of this model, consider a benchmark case in which raising public funds is costless and there is full information. In this case, the gov- ernment would retire all acreage by paying each landowner at least that landowner's reservation payment. The exact amount paid is immaterial, because if funds cost nothing to raise, then the payments are just transfers without any efficiency impli- cations. If funds are costly to raise, but full information remains in effect, then efficiency has the government purchasing development rights on each parcel such that the social benefit of preserving one more acre, net of the foregone benefit of development, equals the public cost of making good on the landowner's loss of not converting the acre to development. The government gets by with paying the land- owners less than what could be realized from development, because the landowners place value on nondevelopment. In either of these cases, the developer receives a higher payment per acre than does the preserver, because the personal loss of not converting is higher for the developer, and the government takes this information into account in the payments it offers.

Now return to the case where information is private. In this case, the preserver will claim to be a developer, so as to collect the higher payments the developer is to receive. Therefore, the government takes two actions: 1)it makes the developer's contract less attractive to the preserver by lowering both the payment and the re- tired acreage for the developer; and 2) it makes the preserver's contract more attractive by raising the payment while maintaining the same retired acreage. These actions leave the preserver retiring the efficient number of acres as under full in- formation, but the developer retiring an inefficiently small number of acres. Some efficiency loss here is unavoidable, given the asymmetric information. However, this sort of plan reduces the preserver's rents associated with privately held information, while purchasing the efficient amount of the relatively cheap land of the preserver, and still paying the developer what is necessary.


In terms of landowner incentives, some form of takings compensation has a variety of potential merits. Takings compensation can improve risk-sharing if private takings insurance markets fail (admittedly an arguable "if '). It can curb incentives for excessive early development as a hedge against the risk of a taking. It reduces landowners' incentives to avoid takings by impairing the prospective conservation value of their land, perhaps by cutting down trees to eliminate potential habitat or surreptitiously killing endangered creatures. It reduces the incentives of landowners to hinder government efforts to gather information about the potential conserva-

"n fact, this deadweight loss plays an important role in principle/agent models of regulation. Without it the government could acquire private information through costless transfers. See LafFont and Tirole (1993).

tion value of their property (Polasky and Doremus, 1998). It encourages landowners to reveal truthfully private information about the value they place on the nondev- elopment uses of their property.

However, we also argue that compensation should not take the form generally envisioned by its proponents, which is that the landowner should receive the full private use value foregone when a property is "taken." Instead, such compensation might be tied to the land's public use value (that is, Pigouvian compensation), or be rendered "lump-sum" in some way. Moreover, the deadweight costs of govern- ment taxation argue for compensation solutions that have lower budgetary costs. Lowering budgetary costs is particularly important under asymmetric information, where the government must design policies to limit rents. Two possible policies include "negligence compensation"-under which takings compensation is only afforded to those who behave efficiently in their choices of land investment, pres- ervation measures, and information provision (Innes, 199713) ;or a policy of tradable development rights, in which owners of "taken" land are implicitly compensated by other landowners who must purchase development rights from them, at no direct cost to the government. Further research is needed to determine the effectiveness of these policies under asymmetric information.

Government Incentives

To this point, we have assumed that government attempts only to maximize social welfare. This "public interest" assumption is analytically useful, and it is consistent with the theory of regulation in which government actions are an attempt to mitigate market failures. However, there is a competing "interest group" view of government which maintains that government's objective is to maximize political support from interest groups, while interest groups attempt to capture rents made available through government policies (for example, see Stigler, 1971). The central case for constitutional limits on government takings begins from this perspective; it is to protect property owners from possible government excesses.

The competing theories of government behavior provide different modeling paradigms that have important implications for the takings debate and the design of endangered species regulation. However, the "public interest" and "interest group" theories are not always at odds; in fact, they may sometimes yield the same predictions because public welfare and the welfare of some interest groups are likely to be positively correlated over the government's choice set (Becker, 1983).~ Moreover, as No11 (1989) argues, a "sophisticated public interest theory" may view com- petition between interest groups in the arena of government as a method of making

For applications of these theories in the environmental context, Crone and Tschirhart (1998)introduce a method for disentangling the predictions of the theories and apply it to the National Forest Service's land use decisions. Ando (forthcoming) applies interest group theory to Fish and Wildlife Service de- cisions about listing endangered species.

and enforcing Coase-style bargains. This will be especially true in situations char- acterized by imperfect information and high transactions costs owing to many po- tentially affected parties-characteristics allied with endangered species preserva- tion. We consider interest group models of government behavior, and their impli- cations for both government behavior and efficient constraints on this behavior. In general, we find that the arguments for some form of landowner compensation need not translate into arguments for constitutionally or judicially mandated compensation.

Fiscal Illusion and Free Riding

The "interest group" view of government can be used to argue either that endangered species may be underprotected or overprotected, depending on which group is most influential.

Suppose for example, that there are two interest groups: landowners whose property may contain critical habitat, and a general public comprised of nonlan- downers, landowners without critical habitat and environmentalists. Suppose fur- ther that the government must decide whether to list species as endangered and that listings result in takings. Because compensation is not required under the Endangered Species Act, landowners have an incentive to form interest groups and to lobby against listings. However, political power may nonetheless lie with the numerical majority, because environmental groups are successful in convincing policymakers that endangered species protection is a high profile issue among the general public. Then, the government, suffering from fiscal illusion, may be tempted to underweight costs to the minority. Stroup (1995) echoes this line of thought:

When a northern spotted owl, red-cockaded woodpecker, or other species listed as endangered or threatened is found on private property, the owners are required to meet the demands of the Fish and Wildlife Service biologists. Yet the biologists have no economic incentive to limit their demands. Since they have no requirement to compensate the owners of the land they control, other people's land has no budgetary cost to them; it is available free of charge.

Fischel and Shapiro (1989) argue that the U.S. Constitution's takings clause is de- signed to prevent a majority-rule government from exploiting minority interests. That is, at that time before anyone knows whether he or she will be in the majority, all will agree on a constitutional rule that requires some compensation and forces a majoritarian government to consider the private costs of its actions.

Alternatively, suppose that while species preservation is important to the gen- eral public and very valuable in the aggregate, individuals may feel that they have little personally at stake, and thus will not bother to lobby policymakers. Instead, most people will act as free riders on the efforts of environmental groups. However, landowners continue to have a strong incentive to lobby, and if they already have existing political or economic organizations such as cattlemen's associations or farm co-operatives, then their cost of organizing is low. In this setting, the concentrated special interest may prevail over the public interest, and the government will list and protect species less frequently than is efficient (assuming it cannot pay compensation).

Which of these theoretical scenarios is relevant in practice? The U.S. Fish and Wildlife Service, an agency responsible for listing endangered species, often chooses not to list a species-only to be reversed in the courts following suits filed by en- vironmental groups. For examples, in 1997 federal district courts ordered the Ser- vice to list the jaguar in Arizona, New Mexico and Texas, and to reconsider its decision not to list the lynx, the Barton Springs salamander and the northern gos- hawk. A U.S. Circuit Court of Appeals extended a ban on logging and grazing in the southwest so judges could consider environmental group claims that the U.S. Forest Service approved activities that do not conform to its own guidelines for protecting the Mexican spotted owl.

This apparent government reluctance to list species and protect habitats that are judged by the courts to be required under the Endangered Species Act is consistent with the view that successful lobbying by landowners is leading to underprotection of endangered species. However, it may also indicate govern- ment efforts to take economic costs and benefits into account, in the face of an Endangered Species Act which requires that species listing decisions not be based on the consideration of such costs and benefits. In either case, there is no need for a constitutional or judicial compensation requirement. A government that is beholden to landowner interests surely has the incentive-without judicial prompting-to compensate for private losses when the public interest is served by "taking" private land.

Government-Chosen Compensation

Until this point, we have presumed that government makes its takings decisions given a predetermined compensation rule. Innes (1997a) models a government that is unfettered by judicial restraints and free to choose both the takings and compensation rules given the interests of landowners and the general taxpaying public.8 If the government's welfare depends upon the net benefits its policies provide the interest groups, it wishes to design policies that maximize the sum of the net benefits, regardless of how much it may want to exploit one group and reward another. For landowner incentive problems that are contemporaneous with government takings and tax/compensation decisions, this logic suggests that the government's self-interest can be relied upon to prompt policies that elicit efficient behavior. In the endangered species context, many of the relevant landowner de- cisions discussed earlier are contemporaneous with relevant government decisions, including information collection, cooperation, and revelation, and a variety of ac-

'Innes (199713) shows that the model is robust to adding environmental groups as lobbyists

tions that may affect the prospective conservation value of a property. In such cases, there is no need for a judicial compensation obligation to provide the government with incentives for efficient regulation (although, of course, distributional issues will remain).

Indeed, a compensation requirement can actually distort government incen- tives, prompting inefficient underregulation. Consider, for example, a government that is deciding how much private land to set aside for endangered species habitat, given that it must purchase the land at fair market value from property owners who are in a political minority-and hence, about whose welfare the government cares little (if at all). Without required compensation, the government freely exploits the landowners, and pursues efficient takings policies to maximize the rewards it hands out to the majority. However, when it must buy up land for habitat (by paying compensation), each additional acre that it buys reduces the amount of land avail- able for private use, raising land prices (given a downward-sloping demand for land), and thereby raising the rewards it must deliver to the political minority, the landowners. A political cost of marginal takings is thus incurred. In essence, the government becomes a monopsonistic demander of land for habitat, and demands too little land as a result.

However, there are some problems with unfettered government behavior. One is that interest group support of government may not reflect the net benefits to the group, owing to free riding by group members."n this case, government welfare need not be aligned with net benefits and government will behave inefficiently. For example, substantial payments to landowners may be very unpopular with the public whose support of government understates the benefits from habitat protection, so government reacts by protecting too little habitat. Asecond problem arises when there is a temporal separation between relevant landowner and government deci- sions, with landowners making long-lived choices that affect the economic value of their property. To avoid distorting decisions of private landowners about develop- ment it will be important that the private benefits which landowners receive from developing or not developing their property reflect the social benefits. Otherwise, owners of undeveloped land may well end up bearing an excessive share of the private losses from government takings, while owners of developed land may be subject to political opportunism that taxes away any return to their successful in- vestments. However, this is a statement about relative payoffs, not absolute ones. If there is an argument for constitutionally or judicially mandated compensation, it is not so much that no landowners (or person) should be made worse off in absolute terms, but that the marginal private benefits of developing or not developing land should reflect the social benefits, and that compensation necessary to this aim might be mandated. A policy of tradable development rights satisfies this sort of "equal treatment" restriction, just as well as does an appropriate compensation policy that

" Nowell and Tschirhart (1993) demonstrate how free riding behavior within interest groups is necessary to distinguish behveen "public interest" or "interest group" theov predictions, and they relate the predictions to the efficiency of government decisions.

is much more costly to the government.10 For example, in a region where environ- mentally pristine land (land "eligible" for habitat) coexists with land that has been developed (and is "ineligible" for habitat), judicial "equal treatment" protection may be called for.

A few qualifications to this line of thought are worth noting. First, "the gov- ernment" is not the monolithic enterprise envisioned here; rather, it is comprised of multiple tiers of politicians and bureaucrats. Even if "top-tier" politicians aim to maximize the size of the economic pie, lower-tier bureaucrats-precisely those whom Stroup (1995) cites as the potential source of caprice in endangered species regulation-may have other agendas. While such realities clearly merit more thought, they do not argue for judicial restraint as much as they argue for legislative restraint on bureaucracies. Second, compensation requirements can affect both the incentives to lobby and, therefore, the economic resources wasted in this endeavor (Epstein, 1992; Fischel, 1992). However, casual empiricism suggests that any savings in lobbying expenditures are generally likely to be very small relative to the eco- nomic benefits at stake in the policies that lobbying is intended to affect, though this subject is worth further exploration. Third, compensation may make sense for reasons of landowner incentives, even if judicially or constitutionally mandated compensation is not likely to be the cure for perceived ills in government regulatory practice.


Compensation as a constitutional requirement for government policy has a number of drawbacks. If such payments are politically unpopular, the requirement may discourage protecting endangered species. Moreover, such a constitutional requirement seems unnecessary for the sake of efficiency, because a government with full powers of taxation and compensation has an incentive to design an effi- cient "takings" policy without anyjudicial restraint. Judicial restraint may be called for when forward-looking landowners make long-lived decisions, because an unfet- tered government may distribute economic largesse to meet political ends and not to reflect the relative societal benefits that landowners' earlier decisions deliver; even in this case, however, the required restraint is not one of strict compensation, but rather one that protects relative private payoffs at potentially zero government cost.

Final Thoughts:Habitat Conservation Plans

Economic welfare may therefore be enhanced by reforms to the Endangered Species Act that cleverly compensate those whose land is taken for critical habitat in ways that keep government budgetary costs to a minimum and, yet, provide

"'Innes (1997a) develops this argument fully.

landowners with incentives to protect their land's prospective conservation value, to cooperate in information gathering, and to make efficiently circumscribed pri- vate development and investment decisions. However, the argument that such com- pensation should be a generic requirement, enforced by the constitution or the judiciary, has less merit.

Rather than pushing the compensation requirement, the Clinton administra- tion has pushed "habitat conservation plans," which arguably reflect some progress on several dimensions. Although habitat conservation plans were authorized in 1982, only 14 of them existed before 1993. Now 211 exist, with more than 200 more on the way (Margolis, 1997). Habitat conservation plans allow a landowner to "take" a species if doing so does not threaten the survival or recovery of a species. In practice, this means that the landowner must provide adequate conservation safeguards, such as setting aside land for nondevelopment either on the owner's property or elsewhere, in exchange for the right to develop. A habitat conservation plan allows a landowner to develop without the threat of legal action under the Endangered Species Act, but does not provide compensation. In cases where con- senration value is known to be high, placing the burden on landowners is desirable. With conservation value already known, investment and information concerns are minimal and not paying compensation limits the deadweight burden of government spending.

The Clinton administration's approach to future conservation requirements is, however, quite different. The administration is backing a "no surprises" policy under which a landowner with an approved habitat conservation plan would be exempt from future restrictions should other endangered species be discovered in new locations." The no surprises policy ensures that government picks up the tab for the cost of new protections. Surprises are exactly the type of case where land- owners can invest in ways harmful to species, rush to develop prior to imposition of regulation, have private information, and be unwilling to cooperate with infor- mation collection. Given these incentive problems, paying compensation may be the best course despite the deadweight burden from increased government spending.

However, such compensation plans are not without their difficulties, either. Environmentalist critics contend that habitat conservation plans are likely to be susceptible to political vagaries and compromises that ignore biology, especially as landowners rush to take advantage of them by dealing with already overburdened government agencies. This fear is amplified where the agencies are concerned

'I In late 1997, Senators Dirk Kempthorne, John Chafee, Harry Reid and Max Baucus finalized a reau- thorization of the Endangered Species Act with a "no surprises" clause and with limits on requiring other federal agencies to consult with the Fish and Wildlife Service regarding potential takings. The bill is backed by Interior Secretary Babbitt. The bill is opposed by most environmental groups who favor instead Representative George Miller's reauthorization bill that would submit "no surprise" clauses to independent scientific peer review, and require landowners to insure themselves against future discov- eries of endangered species by securing a bond.

about pleasing landowner interests. In short, critics fear that such plans may be overused as a method of avoiding conservation. In the spirit of our earlier discus- sion, however, it's worth noting that a plan without compensation may lead to less conservation than ideal. Perhaps it would make sense to seek a compromise in which compensation would be paid to entice more landowners into habitat con- servation plans, while clarifying and tightening the rules for such plans in a way that would be responsive to environmentalist concerns.

w We are grateful for the diligence of the editors and for comments ?om participants at the University of Wyoming conference on Social Order and Endangered Species Preservation,

Laramie, Wyoming, Apil1997.

Anderson, T., "A Snail Retreat," Wall Street Journal, December 27, 1993, 6. Anderson, T., "Enraging Species Act," Wall Street Journal, April 19, 1994, A20.

Ando, Amy Whritenow, "Implications of Interest-Group Behavior for the ESA." In Ja- son Shogren and John Tschirhart, eds., Social Order and Endangered Species Presmation,forthcoming.

Becker, Gary, "A Theory of Competition Among Pressure Groups for Political Influence," Quarterly Journal of Economics, August 1983, 97, 371-400.

Blume, Lawrence, and Daniel Rubinfeld,

"Compensation for Takings: An Economic Anal- ysis," Cahjmnia Law Review, 1984, LXXIZ, 569


Blume, Lawrence, Daniel Rubinfeld, and Peny Shapiro, "The Taking of Land: When Should Compensation Be Paid?" Quarterly Journal ofEce nomics, 1984, 991, 71-92.

Brinckrnan, J., "$3 Billion Later, Columbia River Salmon Dwindle," The Oregonian,July 27, 1997a, A-1.

Brinclrman, J., "High Court Allows Logging- Restriction Suit," The Oregonian,April 12, 1997b, A-1.

Crone, Lisa,and John Tschirhart, "Separating Economic from Political Influences on Govern- ment Decisions," Journal ofEconomic Behavior and Organization, 1998, 1041, 1-22.

Dopson, A., J. Rodriguez, W. Roberts, and D. Wdcove, "Geographic Distribution of Endan-

gered Species in the United States," Science,January 24, 1997, 275,550-53.

Epstein, R., "Yee v. City of Escondido: The Su- preme Court Strikes Out Again," Loyola of Los Angela Law Review, 1992, XXVT, 3-22,

Farber, D., "Economic Analysis and Just Com- pensation," Zntaational Review of Law and Ece nomics, 1992, XU,125-38.

Fichel, W., "Exploring the Kozinsky Paradox: Why is More Efficient Regulation a Taking of Property?" ChicapKent Law Review, 1991, LXVZE3,865-912.

Fichel, W., and P. Shapiro, "A Constitutional Choice Model of Compensation for Takings," Intaational Review ofLaw and Economics, 1989,ZX 115-28.

Fullerton, D., "Reconciling Recent Estimates of the Marginal Welfare Cost of Taxation," American Economic Review, 1991, 81,302-08.

Hermalin, B., "An Economic Analysis of Tak- ings," Journal ofLaw, Economics and Organization, 1995, XI, 64-86.

Innes, Robert, "Takings, Compensation and Equal Treatment for Owners of Developed and Undeveloped Property," Journal of Law and Ece nomics, October 1997a, 4@2,403-32.

Innes, Robert, "The Economics of Takings in a Multi-Parcel Model With a Powerful Govern- ment." In Jason Shogren and John Tschirhart, eds., Social Order and Endangered Species Preserua- tion, forthcoming.

Innes, Robert, "The Economics of Takings and Compensation When Land and Its Public Use Value Are in Private Hands," Working Pa- per, University of Arizona, 1997b.

Jewett, Joan Laatz, and Jonathan Brinkman,

"U.S. Proposes Huge Fish-Rescue Effort," The Oregonian, February 27, 1998, A-I, A-20. Johnson, C., "Cornered in California," Farm Journal, Mid-March 1995, 12-14,

Kaplow, L., "An Ecoilomic Analysis of Legal Transitions," Hurunrd Law Review, 1986, XCIX, 509-617.

Laffont, Jean-Jacques, and Jean Tirole, "The Politics of Government Decision-Making: A The- ory of Regulatory Capture," Qtia7twly Journal of Economics, 1991, 106,1089-1127.

Laffont, Jean-Jacques, and Jean Tirole, A The-or)' ofIncentives in Procurement and Regulation, Cambridge, ha: MIT Press, 1993.

Lambert, T., and R. Smith, "The Endangered Species Act: Time for a Change," Center for the Study of American Business, Policy Study 119, 1994.

Lewis, Tracy R., "Protecting the Environment Then Costs and Benefits are Privately Known," R~L\DJournal ofEconomics, Winter 1996, 27, 819


Main, Martin B., and Fritz Roka, "Cost of Pub- lic Land Ownership to Preserve Florida Panther Habitat," presented at Conference on Flexible Incentives for the Adoption of Environmental Technologies in Agriculture, June 1997, Gainesville, Florida.

Mann, C., and M. Plummer, Noah's Choice: The Future of Endangered Species. Kew York: Alfred A. Knopf, 1995.

Margolis, Jon, "Critics Say 'No Surprises'

Means No Protection," High Country Naus, Au

gust 4, 1997, Paonia, Colorado.

Meltz, R., "The Endangered Species Act and

Constitutional Takings," paper presented at

Conference on Regulatoiy Takings and Re

sources, Natural Resources Law Center, Univer-

sity of Colorado School of Law, June 13-15,


Miceli, Thomas, and Kathleen Segerson, "Reg

ulatory Takings: U'hen Should Compensation Be

Paid?" Jou,nal of Legal Studies, 1994, XXIII, 749

76. Noll, Roger, "Economic Perspectives on the Politics of Regulation." In R. Schmalensee and

R. Willig, eds., Handbook of Industrial Organi- zation, North Holland: New York, 1989, 1253-


Nowell, Clifford, and John Tschirhart, "Testing Theories of Regulatory Behavior," Rmiau of Industrial Organization, 1993, 8, 653-68.

Pi,S., G. Russell, J. Gittleman, and T. Brooks, "The Future of Biodiversity," Science, July 21, 1995, 269, 347-50.

Polasky, Stephen, and Holly Doremus, "When the Truth Hurts: Endangered Species Policy on Private Land with Imperfect Information," Journal of Environmtal Economics and Management,

1998, 35:1,22-47.

Polasky, Stephen, Holly Doremus, and Bruce Rettig, "Endangered Species Conservation on Private Land," Contempora7y EconomicPoliq, 1997, 15, 66-76.

Polasky, Stephen, "Investment, Information Collection and Endangered Species Conserva- tion on Private Land." In Jason Shogren and John Tschirhart, eds., Social OrderandEndangered Spenes Preservation, forthcoming.

Smith, Rodney, and Jason Shogren, "Endangered Species Protection and the Design of Vol- untav Incentive Schemes." In Jason Shogren and John Tschirhart, eds., Social Order andEndan- gered Species Preservation, forthcoming.

Stevens, W., "Conservation Plan for Southern

California Could Be Model for Nation," The New

York Times, Februav 16, 1997, A-12.

Stigler, George, "The Economic Theory of Regulation," BellJournal ofEconomics, 1971, 5, 3-


Stroup, R., "The Endangered Species Act: Making Innocent Species the Enemy," PERC Policy Series # PS3, Montana State University, April, 1995.

Sugg, I., "Worried About That Owl On Your Land? Here's Good News," Wall Street Journal, April 6, 1994, A21.

United States Fish and Wildlife Service, News release, June 6, 1997.

Welch, L., "Property Rights Conflicts Under the Endangered Species Act: Protection of the Red-Cockaded Woodpecker." in B. Yandle, ed., The Takings Debate, 1995.

  • Recommend Us