Insuring Takings Claims

Christopher Serkin

takings claims, however,

ABSTRACT—Local governments typically insure themselves against all kinds of losses, from property damage to legal liability. For small- and medium-sized governments, this usually means purchasing insurance from private insurers or participating in municipal risk pools. Insurance for regulatory is generally unavailable. This previously unnoticed gap in municipal insurance coverage could lead risk averse local governments to underregulate and underenforce existing regulations where property owners threaten to bring takings claims. This seemingly technical observation turns out to have profound implications for theoretical accounts of the Takings Clause that focus on government regulatory incentives. This Article explores the impact of insurance on land use regulations. In the process, it reveals important insights about public insurance more generally and offers a novel explanation for the burgeoning land use innovation in cities compared to the relative stagnation of land use in the suburbs. It concludes by suggesting new ways for promoting local land use regulations that risk generating takings claims.

AUTHOR—Professor of Law, Vanderbilt Law School. This piece benefitted from early discussions with Ed Cheng and Paul Edelman. I received helpful comments on earlier drafts from Tom Baker, David Dana, Lee Fennell, John Rappaport, and Kevin Stack. Thanks also to participants at the Georgetown Environmental Law & Policy Institute 2015 Regulatory Takings Conference, especially to John Echeverria, Nestor Davidson, and Carol Rose. Samantha Williams and Kacey Murphree provided research assistance. Special thanks to Dean Chris Guthrie for supporting this project.

Local governments face many different risks of financial loss, whether from flooding, workers’ compensation, embezzlement, property damage, or litigation, to name just a few. Litigation imposes a particularly broad set of risks. For example, victims of civil rights violations can sue under § 1983;1 people injured on public property, whether a playground or a street, might sue in tort;2 municipal employees can sue for discriminatory employment practices;3 and property owners burdened by land use regulations might sue under the Takings Clause.4 All of these risks expose local governments to the possibility of financial losses. But there is an unexpected difference between them. Civil rights violations, torts, contract claims, property damage, and so forth are almost always covered to a greater or lesser extent by municipal insurance.5 Regulatory takings claims, however, are not.6 Although previously unnoticed in the legal literature, insurance is by and large not available for regulatory takings litigation.7 Why is that? And,

more pressingly, does the absence of this municipal insurance make risk averse local governments reluctant to enact (or defend and enforce) socially beneficial regulations?

Answering these questions requires understanding the nature of regulatory takings risk and also the almost unexplored warrens of municipal risk management.8 The paucity of recent legal scholarship in the area is hardly surprising.9 Examining how municipalities deal with risk requires understanding technical aspects of insurance law and insurance markets, the political dynamics in local governments, and the nature and effects of municipal risk aversion. These are all topics that threaten to make eyes glaze and heads go woolly. But they turn out to be vitally important and in fact are crucial to understanding the operation of land use regulations on the ground.

If there is any doubt, protracted litigation in Half Moon Bay, California, should put it to rest. In 2007, property owners brought a successful regulatory takings claim against that municipality of 12,000 people and won a judgment of over $36 million.10 Crippled, the town eliminated its police and recreation services and then considered dissolving as an independent jurisdiction.11 The municipality was saved, however, when it successfully sued a former insurer for coverage under an “occurrence-based” policy that had lapsed more than twenty years earlier.12

It was an unusual victory for a local government based on an extremely unusual set of facts; in most cases, as surveyed below, the insurance company wins, and such regulatory takings are excluded from coverage.13 But the overall point is easy to see: the very survival of Half Moon Bay depended on its insurance coverage for regulatory takings.

focuses on competing accounts of

Municipal insurance implicates broader conceptual issues as well. Much of the important contemporary writing on land use and property regulation local government decisionmaking. Scholars have argued for years about the effects of compensation on governments’ regulatory incentives. Professor Richard Epstein, for example, has argued that compensation is necessary to force local officials to internalize the costs of their actions and thereby to induce efficient regulatory incentives.14 Public choice theorists, in contrast, have argued that governments do not internalize costs and benefits in this way, but are focused instead on maximizing their political capital.15 But neither of these sophisticated treatments has discussed the impact of insurance on government decisionmaking. Any account of local officials’ economic or political incentives surrounding environmental and land use regulation must contend with insurance, and yet no one to date has studied its impact on local officials.

Regulatory takings liability can be disastrous for a municipality, as the story of Half Moon Bay vividly demonstrates.16 Just the litigation costs alone of defending land use regulations can be exorbitant.17 But the most damaging effect may be on the ex ante incentives of local governments to avoid regulations that might trigger litigation, however frivolous the claims. In the absence of insurance, a risk averse government may choose not to enact beneficial land use regulations—or not to enforce existing

regulations—that create a risk of litigation, even if the likelihood of liability is remote, and even if the expected value of the regulation is strongly positive. And the more risk averse the government, the more cautious it will be in its regulatory enactments and enforcement decisions.

The absence of insurance for regulatory takings claims—both for the risk that a land use regulation is ultimately judged to be a taking, and also for the litigation costs regardless of outcome—is therefore likely to have unexpected and previously unnoticed distributional effects as between local governments. For larger cities that almost entirely self-insure against liability of all kinds, the absence of takings insurance will have little if any ex ante effect. Cities with large tax bases will be risk neutral in their regulatory incentives.18 Government decisionmakers in urban settings can, by and large, determine the expected value of regulatory decisions rationally and will not overweight the risk of liability (subject to dynamic political pressures, considered below). But the same is not likely to be true of smaller governments with fewer residents and a less diverse tax base. For those governments, the costs of takings litigation may be crippling, so decisionmakers will be averse to litigation risks.19

It may well be that this difference in relative risk aversion around land use regulations accounts for some of the dynamism in cities today and also for the relative stagnation of suburbs and exurbs.20 Because of their relative risk aversion, smaller local governments are less likely to be regulatory innovators and will seek instead to avoid enacting regulations that might generate regulatory takings claims. Larger cities, on the other hand, will be freer to push the regulatory envelope. This is not to suggest that holes in municipal liability insurance provide a complete or even predominant explanation for the recent rebirth of America’s urban core, but municipal risk management does play an unexpected role in reinforcing those dynamics.21

Because of the financial risks of takings litigation, and the resulting distributional consequences, to explore potential is mechanisms for allowing smaller local governments to offload at least some of the risk of takings and land use litigation more broadly. Having

important

it

identified a gap in coverage, there may be public or private insurance products that could step in to fill it. But this Article ultimately argues that the most appropriate and most likely solution comes from the state. Indeed, state intervention—through discretionary indemnification or the promise of direct legal representation—may prove a more effective and targeted subsidy for local land use regulation than anything currently in states’ arsenals.

This is not an invitation for local governments to ride roughshod over property rights. The problem for local governments is that every land use regulation comes with some risk of takings litigation and potential liability. Even if local officials are trying to regulate well within the bounds of what the Constitution permits, they still may be sued and may even lose because the regulatory takings standard is notoriously difficult to apply. Existing insurance mechanisms can help prevent indemnifying local governments for willful violations of the Takings Clause. Like most insurance, then, the goal is to design a product that will allow local officials to value the risk of takings claims rationally. This means neither being too cautious nor ignoring the risks altogether.

is an untheorized and underexplored field

The topic of risk management and liability insurance for land use litigation is not a technocratic backwater of local government operations local but governments’ ability and willingness to regulate in the first place. Studying this topic yields unexpected benefits along the way, like a deeper understanding of governmental risk and the motivations of public officials on the ground. And it reveals how insurance can be a valuable form of state subsidy for local officials.

that shapes

The argument proceeds as follows. Part I examines the nature of municipal risk and the ways in which local governments manage risk. Part II demonstrates that risk-spreading devices are not generally available for regulatory takings claims. Part II also argues that the absence of regulatory takings insurance can distort municipal regulatory incentives. Part III finally explores some private market solutions, and then ultimately proposes that states should offer a system of discretionary indemnification to encourage beneficial land use regulations.

This Article’s fundamental observation is that the absence of insurance for regulatory takings litigation has potentially damaging effects on local governments’ regulatory incentives. But that claim requires, first, exploring why local governments have insurance for other kinds of risks and the ways in which municipalities manage risk more broadly. Indeed,

the oddity here may not be the absence of insurance for takings claims but the presence of municipal insurance at all. This Part examines the relationship between risk and insurance generally, the reasons why insurance is valuable to local governments in particular, and the tools employed to address municipal risk management.

A. Municipal Risk and the Role of Insurance

In order to understand why the absence of regulatory takings insurance is significant, it is first necessary to understand, in broad strokes, why municipalities would have insurance against any kind of risk. Insurance, after all, is valuable because of risk aversion, which is a product of the diminishing marginal utility of money.22 Risk aversion makes decisionmakers unwilling to take risks that have a positive expected value because of an aversion to the possible loss.23 It appears at its strongest when a potential loss represents a substantial portion of someone’s total (or, at least, liquid) wealth.24 For individuals, the intuition is straightforward. A $100,000 loss is likely to be more valuable (costly) than a $100,000 gain.25 A $100,000 gain might allow you to buy a fancy car and pay down some student loans. But if you lose $100,000, you might have trouble buying food and paying rent. For most people, not being able to afford food and rent is a more harmful outcome than not being able to afford a fancy car. In short, people have a hierarchy of interests and are likely to be averse to risks that implicate more important ones.26

For individuals, how much someone would be willing to pay to avoid a risk of loss depends on her relative risk aversion.27 A completely risk neutral person would not pay to avoid a bet with an expected value of zero, no matter how high the stakes. But a risk averse person might pay a great

deal.28 She will prefer a definite small loss to a chance of a large one with the same expected value.

Why, then, would a local government official pay a premium to avoid some risk of loss? Insurance is valuable when it allows risk averse people or entities to transfer risk to less risk averse ones.29 But it is not obvious why governments and government actors would be risk averse in the first place. At least theoretically, a government always has the capacity to raise taxes. Therefore, a liability—no matter how large—cannot implicate the government’s ability to meet its higher priority expenses, such as schools or police, when it has limitless capacity to generate revenue. Indeed, most scholarly literature assumes that all governments are risk neutral.30

This theoretical claim, however, does not hold up in the real world. A local government cannot simply tax its way out of its liabilities. Those costs are ultimately borne by taxpayers who do not have unlimited access to revenue. In the absence of insurance, then, local officials are likely to be risk averse for two interrelated reasons: the risk aversion of local taxpayers who ultimately fund uninsured liabilities, and politicians’ own political self-interest.31 Both require some unpacking.

Consider taxpayers’ incentives first. To the extent local officials accurately represent their constituents’ preferences, the relevant inputs for governmental risk aversion are the risk preferences of local taxpayers.32 Taxpayers ultimately finance most government liabilities, and at the local level, this usually means property taxes.33 For property owners, then, the absence of municipal insurance raises the possibility of a large loss to the government that will appear on their tax bills. Risk averse taxpayers will therefore value, and be willing to pay for, municipal insurance to offload

some or all of the risk of a spike in their property taxes. They will, for example, prefer paying an extra $50 per year in taxes rather than risk a 4% chance of a $1000 spike in any given year, even though the expected cost of the latter is only $40.

Taxpayers’ risk aversion varies as a function of the municipality’s population and wealth. After all, it is the potential magnitude of taxpayers’ individual losses that will affect their aversion to risk. That, in turn, will depend on the per capita effect on property taxes, and not the total value of a municipal loss.34 Even a large judgment against a municipality does not necessarily translate into large per capita costs to local taxpayers. In effect, then, local governments spread risk through the number of taxpayers.35 Furthermore, whether that per capita amount is salient will depend on individual taxpayers’ relative risk aversion. Combining these two insights means that the larger and wealthier the tax base, the less risk averse municipal officials should be, if they are accurately representing their constituents’ preferences. This model therefore predicts that local officials in smaller and poorer municipalities will, indeed, be quite averse to risks that may measurably impact property taxes, while officials in larger and wealthier municipalities should be closer to risk neutral.36

Pause for a moment, though, and consider if this is necessarily true. After all, a municipality has an alternative mechanism for shielding taxpayers from unpredictable spikes in property taxes to fund municipal liabilities. Faced with a significant liability—whether a legal judgment or otherwise—a local government could float a bond to cover the cost.37 Debt will not eliminate nor even reduce the costs that the municipality might ultimately owe. Indeed, the costs of borrowing, reflected in the interest rates of municipal bonds, will mean that a municipality will end up paying

more in total when it funds liabilities through debt than when paying immediately. But selling bonds may well be less expensive than purchasing insurance ex ante, and they have largely the same effect.

some premium

Just as insurance transforms the risk of spiky losses into more predictable but certain ex ante costs, debt does the same thing ex post. If taxpayers prefer—and will pay for—avoiding unpredictability in tax rates, that can be accomplished simply by issuing debt as needed.38 Debt is, in a sense, a kind of ex post insurance mechanism, smoothing the impact of a sudden liability into predictable and smaller payments over time. This, now, is a more nuanced restatement of the question that opened this Section: Why would municipal insurance ever be valuable to a government that has other tools for controlling its own cash flow?

Here, again, the theoretical argument that local governments are risk neutral loses traction in the real world. There are, in fact, important limits on a municipality’s ability to borrow its way out of a sudden liability. For one, state law often caps the amount of debt that a municipality can incur.39 If a municipality is at or near its debt ceiling, it will not be able to spread the impact of a sudden liability into the future. For another, the due process, administrative, and political costs of incurring debt may be quite high. In many jurisdictions, issuing debt above a certain amount may trigger bond election requirements, requiring voters to approve the bonds.40 There is another subtler dynamic at work as well. The interest rate a municipality will pay for municipal bonds depends upon the municipality’s bond rating. That rating, in turn, depends on a number of factors, including

the size of the tax base and outstanding liabilities.41 If a municipality waits until after an adverse judgment, that liability will affect its ability to borrow on favorable terms. Relatedly, another key factor in municipal bond ratings is the municipality’s risk management practices. Rating agencies will look at how municipalities deal with risk ex ante in assigning a bond rating.42 For both of these reasons, a municipality that simply waits to issue debt until it is faced with large liabilities will end up paying much more than a municipality that has risk management strategies including insurance in place. In other words, credit markets will prevent governments from relying too heavily on ex post strategies for mitigating risk to taxpayers. Local officials that care about taxpayers’ risk preferences cannot simply rely on debt to smooth tax burdens over time.

There is also more to the impact of a takings judgment than the possibility of a higher one-time tax bill. The threat to taxpayers of an adverse takings judgment may not be fully captured by the incremental increase in their property taxes. The more serious problem arises when a judgment or settlement creates a competitive disadvantage vis-à-vis neighboring municipalities. Local governments, after all, are in competition with each other for mobile capital.43 The terms of the competition are taxes and services. Therefore, the absolute cost to a government may be less important than its impact relative to its neighbors. Even a small change in tax rates to fund litigation or an adverse legal judgment can have a big impact in the ability to attract the marginal business or affluent household. Qualitative empirical work supports the idea that local officials are in fact very sensitive and averse to risks that will create such a relative disadvantage.44 The one million dollars spent on takings litigation is, for

example, a foregone library expansion, reduced street cleaning, or an increased tax burden without offsetting benefits. Creating this kind of competitive disadvantage can reduce property values, which may be much more serious and salient to homeowners than a slightly higher tax bill.

All of this has so far ignored the political dynamics of legal risks—the second source of municipal risk aversion—and those are equally important. In addition to reflecting their constituents’ risk aversion, local officials also have their own risk preferences. Tax increases impose a political cost for local officials that outpace their hardship for individual taxpayers.45 For example, the risk of a $200 increase on local property tax bills is relatively insignificant. The average taxpayer would not be willing to pay much of a premium to avoid the risk of only a $200 loss. But the political costs of that increase may be significant, indeed.46 For local officials, then, managing risk through the normal budgeting process by purchasing insurance can immunize them from the political costs of the occasional but unpredictable tax hike that would otherwise occur. Insurance therefore has a value to local officials independent of its monetary value to the taxpayers ultimately footing the bill.

Notice, however, that the value of the insurance in this setting, too, will tend to vary inversely with the size of the government, just as it will if local officials are concerned primarily with taxpayers’ risk aversion. Cities can take advantage of the law of large numbers and build anticipated litigation costs directly into their budgets without the use of insurance.47 Moreover, local officials are likely to be less attentive to the desires and

concerns of individual taxpayers in large cities, which tend to be controlled by special interest groups and not local homeowner majorities.48

This is not a particularly flattering account of the role of municipal insurance. If a municipality is buying insurance that is valuable to local officials but not to voters, then this resembles a traditional agency malfunction with taxpayers paying for a product they do not value. Professors Baker and Griffith have powerfully criticized Directors and Officers (D&O) insurance on precisely these grounds.49 They argue that entity insurance does not benefit shareholders who can be made risk neutral through a diversified investment portfolio.50 D&O insurance is best explained by its value to corporate managers whose individual fortunes are tied to the corporation’s balance sheet and cannot be so easily diversified.

Municipal insurance, however, is critically different from D&O insurance because local taxpayers cannot diversify risk as easily as shareholders. In fact, for most homeowners, their houses represent their single largest investment.51 They are therefore likely to be extremely averse to risks that threaten property values, including adverse judgments that result in a competitive disadvantage vis-à-vis other local governments.52 Adding political costs and agency malfunction to the mix does raise the possibility that local officials’ risk preferences may diverge from those of their constituents, but at least they are likely to move in the same direction. This is not a merely hypothetical account. Empirical evidence of local officials’ risk aversion comes from the response to a municipal insurance crisis including municipalities’ increased exposure to liability as well as adverse market conditions and overly aggressive investment strategies by insurance companies, the private market for municipal insurance hardened in the mid1980s.53 Premiums skyrocketed, increasing sometimes 1500%.54 And in

to a combination of factors,

the 1980s. Due

in

some cases, insurance companies simply refused to write policies at any price.55 Municipalities responded by closing playgrounds and pools, and otherwise halting activities with liability exposure.56

Viewed in coldly political terms, this is a remarkable reaction. It demonstrates that local officials in fact prioritized future fiscal costs over short-term political ones. Closing pools and playgrounds is a politically toxic decision for a municipal government.57 And local officials had an alternative; they could have kept the facilities open and simply paid for any resulting tort liability directly, instead of through a private insurance function. The fact that they did not, and chose instead to incur the political cost, demonstrates how averse local officials were to the risk of uninsured financial losses.

The ultimate proof of ongoing local risk aversion can be found in local governments’ contemporary risk management practices. In reality, municipalities engage in varied and sometimes expensive responses to the risks of liability, demonstrating that local officials perceive some value— whether economic or political—to shielding taxpayers from the risk of loss. The next Section surveys this complex constellation of practices, largely invisible in the legal literature, by risk averse municipalities.

B. Municipal Risk Management Strategies and Practices

Municipal risk management is an unusual topic in law. It is ubiquitous, extremely important for understanding the application of the law on the ground, and yet is virtually unstudied.58 The purpose of this Section is to remedy that deficiency and to shed light on municipal risk management practices. In so doing, it also sets the stage for the central question animating this project: why risk averse municipalities retain the risk of regulatory takings liability while offloading other similar risks.

One reason for the paucity of scholarship in the area may be the difficulty in generalizing about risk management practices. Conversations with municipal risk managers undertaken for purposes of this research all began with a caution that risk management practices vary from government to government. That caution is well taken.59 Details undoubtedly vary in important respects between municipalities. Nevertheless, generalizations are possible with this caveat and can capture in broad form the range of approaches to managing municipal risk.

Today, local governments face the possibility of liability on many fronts. The most significant is workers’ compensation.60 Environmental issues loom large, as do car accidents, and claims arising out of the maintenance of roads, streets and sidewalks.61 Constitutional violations can also create legal liability, either directly or more commonly under § 1983.62 Section 1983 creates a private cause of action for constitutional violations, paradigmatically, when a municipality’s failure to supervise results in systemic police abuse, when local officials deny a license on discriminatory grounds or impermissibly suspend a business license, and so forth.63 Plaintiffs will often sue both the government and the individual public official whose conduct is at issue.64 The ultimate costs, though, are almost certainly borne by the government and not by the individual official because most governments indemnify their employees for any resulting

liability.65 Even when municipalities retain discretion to deny such that indemnification, indemnification is necessary to avoid inhibiting public officials or deterring people from becoming public officials in the first place.67 Section 1983 therefore imposes a substantial risk of liability on local governments.

is conventional wisdom

they rarely do.66 It

Municipal liability can be enormous. As of 2011, New Jersey municipalities, for example, were spending nearly $350 million per year in liability-related costs.68 New York City spent almost $2 million per month between 1994 and 1996 resolving lawsuits arising out of police misconduct alone.69 One recent study in New York State, based upon extremely conservative assumptions and excluding New York City, still found over $1 billion in legal liability for New York municipalities in a five-year period.70 A similar study in Louisiana found municipal liability of $100 million, reflecting significant growth in recent years.71 These studies employ different methodologies and it is not appropriate to compare them against each other. Taken together, however, they demonstrate in no uncertain terms that litigation against municipalities is expensive.

Risk management practices address these kinds of legal risks in two different ways. First and foremost, a municipal risk manager will seek to

implement policies and procedures that minimize the possibility of liability in the first place. A risk management office is typically involved in training municipal employees about their legal obligations, and offering advice and counsel when government actions run the risk of creating liability.72 Risk managers also help to develop and deploy procedures and practices that ensure compliance with the law.

Second, in addition to limiting liability and losses ex ante, municipal risk management addresses the risks that nevertheless remain, despite a municipality’s best efforts. Municipalities engage in three general strategies to address risk: retain it, insure it privately, or join a municipal insurance pool. These are considered in order.

1. Retention of Risk.—The most straightforward way for a municipality to handle risk is to retain the risk itself. This can happen through sophisticated funding mechanisms, such as the creation of a captive insurance company (effectively a subsidiary of the municipality that exists solely to insure the municipality).73 But a municipality can retain risk simply through the absence of insurance.74 Indeed, if a municipality makes no provision for the risk of a loss that occurs, it will have to bear the costs itself, which amounts to “noninsurance.”75 Colloquially, both phenomena are referred to as “self-insurance.”76 The phrase is a bit of a

misnomer but is nevertheless commonplace and usefully captures the approach.77

Self-insurance is not an all-or-nothing proposition. In many instances, a municipality will pay to offload some risk—perhaps the risk of catastrophic loss—but retain other risk, perhaps through high deductibles or large co-pays.78 In this way, a municipality can self-insure against part but not all of a risk. Self-insurance (including noninsurance) is therefore a ubiquitous part of comprehensive risk management. A municipality cannot fully insure itself against everything and so always retains some risk of loss through self-insurance, broadly defined.

As predicted in Section I.A, the extent of self-insurance tends to fluctuate with the size of the municipality. Larger cities retain more if not all risk of many different losses.79 This is partly because government officials in larger municipalities are likely to be less risk averse for the reasons described above.80 But cities also retain more risk because they can take advantage of the law of large numbers in ways that smaller municipalities cannot.81 For example, a large city can predict with greater accuracy what its overall § 1983 liability will be in any given year because the larger number of cases will tend to smooth out the outliers. There will certainly be fluctuations, but the unpredictability of those fluctuations will generally change inversely with city size. Large cities, in effect, are aggregators of risk just like insurance companies, and they do not need to outsource as much risk to take advantage of the law of large numbers and to make liability predictable.82

This dynamic also means that larger cities can build the anticipated costs of legal liability directly into their budgets. They do this in two

different ways. First, they typically maintain a legal department, which is a kind of self-insurance against the costs of legal services. Instead of being subject to the vagaries of hourly billing, city attorneys draw a predictable salary that can be funded through the city’s annual budget. And second, cities can predict the annual costs of losses like liability judgments against them and can include that anticipated expense as a regular line item in their normal budgeting process. This helps to insulate property owners from unpredictable spikes in their tax bills.

Where the city retains a separate account for paying out legal judgments—whether as a separate reserve pool or through a captive insurance company—it can specify the limits for payouts per claim, and even include exceptions for coverage.83 Liabilities not covered by the reserve pool—say, regulatory takings liability—must be paid from the general treasury, unless private insurance is available.84

2. Private Insurance.—The most familiar way of handling risk involves procuring private insurance. This is an important piece of many municipalities’ risk management strategies. Even cities like Nashville, Tennessee, which are large enough to self-insure against most losses, maintain private insurance against catastrophic property damage.85 The nature of the private insurance, however, responds to the particular political dynamics of public risk management.

Municipal insurance presents something of a puzzle: it represents a cost to the present government (premium payments) in order to avoid a cost in the future (defense costs and a possible adverse judgment). But this is not a payment that government actors are typically incentivized to make. Usually, political interests line up in favor of shifting risks and costs to the future.86 This puzzle largely disappears, however, with a more nuanced understanding of most municipal insurance today.

The hard insurance market of the 1980s (described above87) caused many municipalities to exit private insurance markets—voluntarily or involuntarily—and to form municipal risk pools (described below88). Private insurance markets eventually opened back up and again became an

important component of municipal risk management by the late 1980s and early 1990s. But when insurance companies started underwriting new policies, they began to take a new form: many changed from occurrencebased to claims-made policies.89 This technical-seeming change aligned private insurance markets with political incentives.

Prior to the hard market of the 1980s, most municipal insurance policies were occurrence-based, meaning that they covered all liability arising out of occurrences within the covered period, regardless of when a claim was actually filed.90 Occurrence-based coverage comes with a long tail of liability for insurance companies, and therefore exposes them to claims that are sometimes filed long after the policy period expired.91 In fact, in Half Moon Bay’s litigation, described in the Introduction, the primary insurance coverage came from a thirty-year-old occurrence-based policy that had been in effect in the 1980s, when Half Moon Bay transformed the plaintiff’s property into wetlands.92 The municipality argued, successfully, that those actions had caused the liability, and were covered by the insurance policy that had been in effect at that time.

Claims-made policies, in contrast, provide coverage only for claims actually made during the policy period regardless of when the underlying conduct occurred. Not only does this eliminate an insurance company’s exposure once the policy ends, it also allows insurance companies to assess the occurrences in the past that might generate claims during the policy period. That is, at least some of the claims covered by the policy will be based on conduct in the past, and the insurance company can therefore price the policy more accurately.93

While

insurance companies motivated

the change away from occurrence-based policies to avoid the liability tail, claims-made policies are also much better tailored to the political dynamics of public insurance.

According to a conventional description, public officials have a strong incentive to favor the present over the future.94 Immediate benefits are politically valuable, while future costs are likely to be borne by other officeholders. As a result, self-interested officials are likely to take risks that have significant upside potential in the short term so long as the downside costs can be shifted into the future. This dynamic is particularly familiar at the local level.95 State-imposed limits on municipal debt are a response to this very dynamic.96 The worry—with an unfortunate basis in history—is that local officials will incur excessive debts to engage in speculative ventures.97 If the speculation pays off, the government reaps the immediate rewards (the new baseball team, the revitalized neighborhood, the successful municipal facility, and so forth). But if it does not, the costs will be borne by future taxpayers who have to service the debt. At least some if not most of the costs would come due only after the public official leaves office.

From this perspective, the existence of municipal insurance is quite surprising. Insurance requires a government to pay up front in order to avoid a risk sometime in the future—exactly the opposite of the concern animating municipal debt limits and other limits on local officials.98

The development of claims-made policies, however,

realigns incentives. A government that buys a claims-made policy will be protected from costs that actually arise during the policy period. This is not coverage for the future costs of their conduct while in office, but instead insurance against the costs arising from conduct in the past (for which they or their predecessors may have been responsible). The local officials paying for the insurance coverage are much more likely to be the ones benefitting from it.99

Although most policies are claims-made, there is considerable variety in the specific forms that they take. Local governments typically purchase one or more of a number of different kinds of policies providing different

types of coverage.100 Property insurance covers losses from fire, flood, theft, and the like.101 Automobile insurance can be more or less comprehensive, and cover specific or comprehensive lists of damage to government-owned vehicles.102 A general liability policy covers personal injury claims and other losses resulting from tort liability, and from tortlike claims.103 An umbrella policy provides excess coverage above and beyond the policy limits in the other policies.104 Specialized policies provide workers’ compensation coverage, and sometimes coverage for electronic records and data recovery, marine insurance, “boiler and machinery” insurance for municipal hospitals and clinics, and others.105 These are all relatively straightforward and provide the kind of coverage one might reasonably expect from their names.

insurance, medical malpractice

Somewhat less familiar is an Errors and Omissions (E&O) policy, which is a form of professional malpractice insurance, analogous to Directors and Officers insurance for private companies. An E&O policy provides coverage for “wrongful acts” by municipal actors, whether school officials, police officers, EMTs, firefighters, or other municipal agents.106 A wrongful act is defined as “[a]ny actual or alleged error, omission, misstatement or misleading statement, act of neglect or breach of duty by an insured while acting within the scope of their duties as officials or employees of the covered organization.”107 These are the policies typically triggered by civil rights or other constitutional violations flowing from

decisions made by government officials. Examples include failure to supervise, employment actions, tax assessments, and land use decisions.108

There are important differences between general liability policies and E&O policies. General liability policies provide coverage for occurrences resulting in bodily injury or property damage. Occurrences in general liability policies, however, are limited to “fortuitous” events, like property damage resulting from a car accident or fire.109 E&O policies, on the other hand, provide coverage for intentional acts or omissions by municipal officials or employees acting within the scope of their duties and so are not limited to fortuitous events.110 Policy choices that adversely affect private interests will trigger E&O coverage. For example, if a municipality closes a road to construction vehicles, affecting costs to a developer of constructing a municipal building, the resulting contract claim would trigger coverage under the E&O policy.111 Most importantly for present purposes—and as described in more detail below—litigation arising out of land use decisions will typically implicate an E&O policy, even though such policies exclude coverage for regulatory takings claims.112

Each of these forms of insurance serves two—and sometimes three— important purposes. First and most obviously, the use of private insurance offloads the ultimate risk of litigation costs and liability. The insurance company will typically pay both legal fees and any adverse judgment, as specified by the policy limits.113 Insurance, then, smoothens potential spikes in liability, converting them into constant and predictable costs in the form of insurance premiums. This is particularly important for a local government because premiums can be included in the normal budgeting

process in a way that unpredictable, actual judgments in litigation often cannot be included.

Purchasing private insurance also then serves a second and closely related purpose: it allows a municipality to purchase the insurance company’s actuarial expertise. A municipality that self-insures and sets its own loss reserves must determine its likely annual liability.114 But that calculation can be complicated and can require both legal expertise and a sophisticated understanding of probability. While large municipalities often have a professional risk management practice in house and can set aside appropriate loss reserves in its budget, private insurance will effectively take over some of the more complicated aspects of that function for smaller municipalities.115 Insurance premiums can therefore be conceptualized, in part, as payment for that service.

The third and last role for municipal insurance applies only to general liability policies but is worth highlighting because of its important parallel to regulatory takings claims: the use of private insurance to spread costs within society.116 The existence of municipal liability insurance requires a special explanation because local governments are entitled to a narrow form of sovereign immunity called governmental immunity.117 This is not full-blown sovereign immunity because it is available exclusively for tort liability, and then only for actions undertaken in a governmental as opposed to a proprietary capacity.118 However, where governmental immunity applies, most states have waived local immunity to the extent of a municipality’s insurance coverage.119 Details vary state by state, but in

general a municipality purchasing insurance is actually exposing itself to liability that it is empowered to avoid.120 That is, by purchasing liability insurance, the local government is waiving the governmental immunity it would otherwise receive, at least up to the limits of the policy. Why would it do so?

Public officials evidently recognize that they are serving their constituents by providing some protection from municipal torts.121 In a sense, by consenting to liability, the municipality is providing a kind of insurance to its tort victims who might otherwise be without redress.122 General liability insurance for these kinds of claims therefore inverts the traditional relationship between risk and insurance. Municipalities are not securing insurance to offload risks that they do not want to bear themselves. Instead, municipalities are insuring themselves in order to accept risk. And they are doing so to reflect a distinct normative perspective on the appropriate allocation of costs in society.123 They are effectively operating as insurers themselves—insurers whose mandatory “premiums” are collected through tax revenue, but whose function is still to spread the risk of certain kinds of individual losses among all taxpayers.124 This loss spreading function of insurance characterizes third-party liability insurance regimes broadly, where costs are borne by the tortfeasor instead of by the victim (and the victim’s first-party insurance, if any).125

At this level of generality, the loss spreading function of insurance is remarkably similar to the underlying justification for regulatory takings law. As the Supreme Court has held, the purpose of the Takings Clause is to prevent some individual property owners from bearing “public burdens

which, in all fairness and justice, should be borne by the public as a whole.”126 It is, in effect, a kind of mandatory third-party insurance, covering the losses of affected property owners, just like liability insurance.127 Governmental liability—whether for takings or torts—reflects a substantive if implicit decision about how costs should be allocated in society. The difference is the lack of private insurance for takings claims, as described in Part II. But first, there is one more alternative form of risk spreading to consider.

last

time public sector

3. Municipal Risk Pools.—The

risk management received anything more than passing scholarly attention in the legal literature was during the sustained “insurance crisis” in the mid1980s.128 At that time, sweeping changes in the insurance industry transformed municipalities’ exposure to risk, although no one was sure exactly why.129 The effects, though, were painfully obvious. Premiums on some insurance products—particularly liability insurance—skyrocketed by “400%, 1000%, 1500%, and more.”130 Other insurance products simply disappeared from the market. As Professor George Priest described in a leading article from that time:

This crisis precipitated enduring changes in the insurance industry and in municipal insurance in particular (some of which have already been discussed above).132 In the absence of affordable insurance, or any private

insurance products at all in some cases, many municipalities turned to insurance pools to spread their risk. Groups of municipalities joined together, paying premiums into the pool and securing liability coverage in return. In this way, smaller municipalities could, together, take advantage of the law of large numbers, and also maintain coverage over risks that private insurers refused to provide.133

In most instances, municipal insurance pools are expressly authorized by the state. Typically, a private not-for-profit organization then operates as a plan sponsor and administers the pool.134 That not-for-profit is governed by a board of directors constituted by representatives of municipal participants in the pool.135 The plan sponsor is nominally an independent entity, but it operates in the service of the members of the insurance pool.

While municipal insurance pools gained real traction in the United States in the 1980s, they were hardly a new invention. In 1903, Municipal Mutual was founded in the United Kingdom to respond to a hardening insurance market there.136 In the United States, the first municipal insurance pool was formed in 1974 by the Texas Municipal League.137 By 2001, there were hundreds of formal and informal pools, covering approximately 35,000 public entities.138 More recent data is simply not available, but as of the early 2000s, it appears that most insurance pools include a professional staff.139 Most also offer more or less comprehensive risk management services, including actuarial expertise, legal advice, and other functions designed to minimize exposure to risk ex ante.140 In short, these risk pools provide broad risk management services in addition to pure insurance.

At first blush, the proliferation of municipal insurance pools is much less surprising than the fact that private insurance remains in the market at all. In theory, risk pools should be able to provide a better product for municipalities at less cost because they are not seeking to make a profit.141 That profit motive, however, can sometimes lead private insurers to outcompete municipal pools. Competitive insurance markets create market pressures for companies to lower costs by pricing risk more effectively, tightly managing administrative costs, or through other business innovations. To the extent municipal insurance pools operate in markets where private insurance is not readily available, their prices may actually be worse despite (or perhaps because of) the absence of a profit motive.

The eventual demise of England’s Municipal Mutual in 1992 is evidence that these concerns are not entirely speculative.142 After its formation in 1903, Municipal Mutual thrived in England for nearly ninety years.143 It was finally undone, however, by a combination of complacency, increasing exposure of municipalities imprudent investments.144 Municipalities in the United Kingdom were left seeking private insurance products, completing a great circle of risk management practices.145

liability, and

Ultimately, municipal risk management consists of some combination of self-insurance, private insurance, and risk pooling. Local governments can change which risks they retain and which are insured, and can adjust their risk management strategies according to their risk preferences in finegrained ways—but not for regulatory takings. Neither private insurance nor takings municipal litigation. The next Part explores the fact of the exclusion and its consequences, and Part III turns to an explanation and some proposed responses.

insurance pools provide coverage for regulatory

Local governments have a number of tools for dealing with risk, but the last Part demonstrated that smaller municipalities often seek to offload risk by purchasing insurance or participating in municipal risk pools. This makes sense. It is costly for risk averse local governments to bear too much risk. It is particularly surprising, then, that insurance is not available for

regulatory takings claims. This Part first identifies the fact that municipal insurance policies—everywhere except Minnesota, inexplicably—exclude coverage for regulatory takings claims. It then argues that this exclusion distorts local officials’ regulatory incentives. It concludes by arguing that this has distributional consequences, adversely affecting smaller local governments more than larger ones.

Part III explores the nature of regulatory takings risk in some detail, but it is nevertheless useful here to have a sense of the kinds of claims that are excluded from insurance coverage. The Fifth Amendment’s Takings Clause and its state analogues prohibit the government from taking property without paying just compensation.146 Where the government exercises its power of eminent domain to condemn property, there is no question that it has to pay. The more difficult question arises when a government has merely regulated property but nevertheless significantly restricts its usefulness and value. If the regulatory burden is substantial enough, it might rise to the level of a taking, even in the absence of the explicit use of eminent domain.147 Such regulatory takings can arise from overly restrictive zoning, the denial of subdivision permits or variances, onerous environmental regulations, and many other local regulatory actions.148 Most of these regulatory takings claims will be evaluated under the ad hoc Penn Central balancing test, which weighs—in relatively inscrutable fashion—the character of the regulation, the extent of the regulation’s interference with distinct (or perhaps reasonable) investmentbacked expectations, and the resulting diminution in value.149 As it turns out, municipal insurance excludes coverage for both condemnation and for regulatory takings.

A. The Inverse Condemnation Exclusion

It might seem difficult to generalize about the content of insurance coverage for takings claims. Insurance policies are not legislation; they do not apply statewide, for example, but are specific to each insurer and insured. As a practical matter, however, municipal insurance policies— whether private or governed by an insurance pool—tend to have very similar if not identical provisions.150 And courts follow other courts’

interpretations of similar policy language.151 It is therefore possible to look to case law to understand the substantive content of insurance coverage within a state. That is especially true of insurance for takings litigation, because the relevant policy language is remarkably consistent across policies, whether private or as part of a pool.

An insurance policy is, ultimately, a contract between the insured and the insurer (whether a private company or a municipal pool), and it defines the insurer’s obligation to cover specified losses. It therefore sets out both the losses that are covered and also certain exclusions, which are losses that would otherwise have been insured but are specifically carved out from coverage. For example, a homeowner’s policy might cover all losses to a house resulting from accident or natural occurrence, but exclude damage resulting from wind or a meteor strike.152

If a loss falls within an insurance policy’s coverage, the insurer not only covers the ultimate loss—up to the policy limit and minus any deductible—but also the costs of defending the litigation. These are two separate components of an insurer’s obligations: a duty to defend and a duty to indemnify. An insurer’s duty to defend is broader than its duty to indemnify, because the duty to defend arises simply from the face of the plaintiff’s allegations; if the allegations involve conduct that is ultimately covered by the policy, then the insurer must defend regardless of the merits of the claim or the likelihood of success.153 Texas courts colorfully call this the “eight corners” rule.154 The four corners of the complaint are judged against the four corners of the insurance policy, and coverage—the insurer’s duty to defend—is determined on that basis. The duty to

indemnify, in contrast, only arises from “proven, adjudicated facts establishing liability in the underlying suit.”155 In general, the insured bears the burden of demonstrating that the insurance policy covers the alleged conduct, but ambiguities are generally resolved against the insurer.156

Where a plaintiff brings a number of claims, the duty to defend will be triggered if the insurance policy covers any of the plaintiff’s claims.157 Frequently, plaintiffs suing municipalities will bring a variety of claims articulating multiple theories of liability.158 Some may be covered by a municipal insurance policy, some may not be covered losses, and still others may be subject to exclusions. Generally, insurance coverage of even one claim will obligate the insurance company to defend the litigation. Any ultimate liability may be apportioned among claims, and an insurer will only pay for the damages resulting from covered claims.159

The exclusion of coverage for regulatory takings claims can be understood against this backdrop. Almost every municipal insurance policy—whether private, or as part of a risk pool—contains some version of the following exclusion from coverage:160

Non-experts would be forgiven for missing the significance of this exclusion. But in fact, it excludes insurance coverage for all regulatory takings claims against a municipality. To see how and why this works, one

must understand the nature and scope of the inverse condemnation claims being excluded.

Inverse condemnation is the doctrinal mechanism for vindicating a regulatory takings claim in state court. Formally, inverse condemnation is an eminent domain proceeding triggered by the property owner instead of the government (it is the “inverse” of a traditional condemnation initiated by the government).162 It is a claim by the property owner that the government’s challenged regulation is, in effect, an exercise of eminent domain. Functionally, it is the state cause of action for alleging a regulatory taking and seeking just compensation.

This is not just a subset of regulatory takings claims, however. Inverse condemnation is, for all intents and purposes, the exclusive means of bringing regulatory takings claims against a municipality or the state. As the United States Supreme Court explained in Williamson County Regional Planning Commission v. Hamilton Bank of Johnson City,163 the Takings Clause does not prevent the government from taking property; it only prevents the government from taking property without just compensation. A local government therefore does not violate the Takings Clause until it has both taken property and denied compensation. The mechanism for seeking such compensation is through an inverse condemnation action.164 Therefore, before a property owner has a ripe regulatory takings claim under the Fifth Amendment directly, she must first pursue her inverse condemnation claim in state court.165 But that is not the only justiciability hurdle. In a notably parsimonious restriction of a federal forum, once the property owner has pursued her state inverse condemnation claim, issue preclusion will almost certainly prevent her from bringing her claim in

federal court.166 The result is that all regulatory takings claims against local governments begin and end with state inverse condemnation actions.167

The inverse condemnation exclusion is even broader than it seems because some courts have extended its reach to exclude coverage for land use litigation, even beyond regulatory takings.168 The typical policy exclusion refers to all claims “arising out of” or “in any way connected with” inverse condemnation, and so creative pleading by a property owner, or creative characterizations of those pleadings by a municipality, are unlikely to avoid the breadth of the exclusion.169 For example, in Transcontinental Insurance Co. v. City of San Bernardino,170 property owners had sued the city for flooding on their property, and had brought claims under both inverse condemnation and tort. If either had been covered by the City’s insurance, then the insurance company would at least have had to defend the City in the litigation.171 The Ninth Circuit, however, held that the tort claims were also subject to the inverse condemnation exclusion, reasoning that “inverse condemnation and tort are not distinct causes, but rather two ways to characterize the same cause of harm.”172 This holding is no outlier.173 Some courts have held the exclusion applies to due

process and equal protection claims,174 and even to Fair Housing Act claims, among others.175

interfered with

It is not at all obvious why the exclusion should apply so broadly because each of these different claims requires developing different factual records. The existence of a regulatory taking turns on the property owner’s expectations about the use of the property, and the extent to which the those expectations.176 Equal government’s actions protection, on the other hand, implicates the government’s justification for treating different property owners differently, and the Fair Housing Act focuses on the impact of a government action on protected classes.177 Even though the elements of these claims are so different, courts have nevertheless held that all of them are excluded from coverage as arising out of inverse condemnation. The inverse condemnation exclusion may thus go much further than precluding insurance for regulatory takings claims, and may preclude insurance coverage for any land use litigation. Indeed, a handbook addressed to local officials concludes: “As of this writing, the authors are not aware of any public agency[, joint powers agency,] or insurance carrier willing to provide express coverage for claims resulting from land use regulation.”178

While the breadth of the exclusion varies somewhat by state, it is also clear enough that, at a minimum, quintessential regulatory takings claims

are excluded from municipal insurance coverage in almost every state. The one major exception appears to be Minnesota, where the public insurance pool administered by the League of Minnesota Cities Insurance Trust (LMCIT) does, in fact, cover regulatory takings claims.179 The LMCIT is extremely unusual in providing comprehensive insurance for land use litigation of all forms, including claims exclusively for injunctive relief.180 Indeed, the LMCIT coverage is the exception that proves the rule. Its manual singles out for particular mention its coverage for regulatory takings claims because it is “frequently excluded under conventional liability insurance policies.”181 The Minnesota example also demonstrates that such coverage is both possible and valuable—both points taken up in Part III.

Why should we care? Uninsured risks impose some costs on risk averse municipal governments, but governing is full of risks. Is it really worth getting exercised about this kind of risk, especially when conventional wisdom holds that successful regulatory takings claims are vanishingly rare? Yes. The problem of uninsured risk of regulatory takings may lead to underregulation and may also have distributional consequences between local governments. Consider these in turn.

B. Takings Risk and Regulatory Incentives

There is little doubt that takings insurance could be very valuable to risk averse local governments. For some—like Half Moon Bay—it could even prove the difference between life and death.182 But the real value may not be in the ex post protection it provides for adverse judgments. After all, takings liability is actually quite rare. The primary value, instead, comes from the effect on ex ante regulatory incentives. Takings insurance is loses when municipalities valuable underregulate vis-à-vis their actual preferences because of the risk of takings liability.183 Or, to view this from a different baseline, the absence of takings insurance creates serious inefficiencies in governments’ regulatory incentives.

society generally, which

The absence of municipal insurance for regulatory takings claims is not just some narrow technical problem but can meaningfully affect local decisionmaking. Consider, for example, a municipal legislature deciding whether to downzone property in order to preserve open space, or a board of zoning appeals deciding whether to deny a subdivision permit for the same reason. Municipal officials will have to decide whether the action creates more benefits than costs.184 Here, the benefits might be preservation of open space and environmental protection, or less salutary goals such as exclusion or protectionism.185 On the cost side of the ledger, some are quantifiable, such as foregone property tax revenue if development does not occur. Some are less concrete, such as the development pressure that will increase on other property in the municipality if this particular land is downzoned. And some are speculative—or risky—such as the possibility of regulatory takings litigation and liability. At a sufficient level of generality, the decision facing the municipal officials will be whether the regulatory benefits outweigh the costs. It is here that municipal risk aversion looms large.

The possibility of takings litigation has an expected value for the municipality, computed by the cost of the liability discounted by its probability, plus litigation costs.186 Imagine that the government anticipates that a takings claim for the downzoning would cost the government $4 million, but that it appropriately assesses the likelihood of losing at only 10%. Leaving aside litigation costs for the sake of simplicity, it is easy enough to see how this should affect a risk neutral government’s decision. The expected cost of the regulation is $400,000. If the regulation is still worth it—if, in other words, the government would be willing to pay $400,000 to secure the regulatory benefits—then it should downzone the property. But for a risk averse municipal official, a 10% chance of a $4 million loss will weigh more heavily on the cost side of the ledger. Depending on the relative risk aversion, that same regulatory action may have an impact on decisionmaking that far outweighs its expected cost, perhaps $800,000, $1 million, or more. The effect may be that the government forgoes a regulation that, in fact, creates greater benefits than costs when rationally calculated.

At this level of generality, the effect of risk aversion on regulatory incentives is straightforward. And it can affect all kinds of government decisions—any, in fact, that potentially implicate the Takings Clause. The result can be to undervalue the net value of regulatory enactments, such as zoning changes, subdivision permits, environmental regulations, and so forth. But it can also lead government decisionmakers to underenforce existing rules by granting variances or failing to enforce land use regulations when an affected property owner threatens to sue.

Experienced practitioners in the area may well chafe at this claim. After all, conventional wisdom suggests that takings claims are extremely difficult for property owners to win and so successful claims are rare.187 But a developer (or any other adversely affected property owner) does not need to win a takings claim to impose substantial costs on a local government. The litigation costs associated with defending takings claims can also be very high, and so even the need to mount a defense can make local governments risk averse. To give a sense of the potential costs, land use litigation in Minnesota constitutes approximately 22% of all liability costs, 85% of which comes from litigation costs.188 To be sure, land use litigation is a broader category than taking litigation alone, but recall that in many states, the inverse condemnation exclusion applies to both.189 And no matter how careful a government may be, it has no significant control over whether a property owner will sue. Truly frivolous suits may come with Rule 11 sanctions and attorneys fees, but the ad hoc nature of takings litigation—discussed in detail below190—again makes it relatively easy for to plead a cognizable cause of action. Inverse property owners condemnation litigation is a risk that attends many regulatory actions. And the problem with municipal risk aversion is that even an unlikely

possibility of a successful takings claim can have a disproportionate impact on ex ante regulatory incentives.191

Whether this is a bug or a feature depends on the eye of the beholder. Some people are likely to have an intuition that governments should avoid doing anything that might conceivably violate the Takings Clause. If the threat of uninsured regulatory taking claims means that governments steer well clear of actions that might burden property rights, then that is all to the good. Indeed, some might object to this whole setup and argue that governments should, as a normative matter, internalize the full costs of their regulatory actions. Local officials should not discount the costs they impose by the likelihood of litigation success. The argument goes: whatever the vagaries of takings law, if a government is imposing $4 million worth of harm, it should fully internalize that cost. Any kind of insurance, in this view, would be perverse because it would allow governments to ignore regulatory burdens.192

This argument, however, relies on a base assumption that governments should compensate for every regulatory burden they impose. This is a principled view, closely associated with Richard Epstein among others, but it is most decidedly not the law—nor should it be.193 Professor Epstein notwithstanding, takings law and most takings theories recognize that there are many costs governments should not be forced to bear.194 If the regulatory burden does not go too far or does not interfere too much with an owner’s expectations, no compensation is due.195 As Justice Holmes

observed nearly a century ago, the government could hardly go on if forced to compensate for all the regulatory burdens it imposes.196

Regulatory takings law does not seek to require local governments to internalize all the costs of its land use regulations. Therefore, it is entirely rational and appropriate for a government to discount the regulatory burdens it imposes by the likelihood that it will be found to have triggered the compensation requirement. That discounting merely reflects the reality of the doctrine and the underlying normative perspective that most regulatory burdens should not require compensation. If risk aversion means that local governments exaggerate the prospective cost of regulatory actions, then governments may end up underregulating relative to the choices that a rational, risk neutral government actor would make. As argued in Section I.A, the risk of takings litigation could impact regulatory incentives whether one focuses on financial or political costs.197 Fiscal costs do not have to translate perfectly into political costs for risk aversion to affect government decisionmaking. So long as local politicians are averse to the political risks imposed by uninsured takings liability, they may choose to forego regulations that would, in fact, be beneficial to their community. In short, the threat of takings liability can lead to inefficient underregulation and underenforcement of land use and environmental regulations by risk averse governments.

C. Distributional Effects of the Inverse Condemnation Exclusion

In addition to concerns about the willingness of local governments to regulate, the absence of insurance may have important distributional consequences to consider as well. Because of the ways in which local governments spread risk, municipal risk aversion is likely to vary with the size and character of the government. Larger and wealthier local governments will be closer to risk neutral, while smaller and poorer ones more risk averse.198 Therefore, the same regulatory action, with the same expected value, is more likely to be adopted by a larger municipality than by a smaller one if it includes a risk of takings liability.

This can give risk neutral local governments a competitive advantage over those that fail to create valuable regulatory benefits because of their aversion to the risk of regulatory takings litigation and liability. A government that rationally calculates costs and benefits and enacts more

regulations that are net beneficial will—by definition—increase the community’s welfare more than those that forego such regulations. This still leaves plenty of room for disagreement about what counts as regulatory costs and benefits, and which regulations are actually welfare enhancing. But the point is simply this: whatever costs and benefits one wants to include—whether one favors a minimalist or maximalist government or something in between—risk aversion will lead smaller local governments to forego regulations with a risk of takings liability that risk neutral governments will enact.

These distributional effects may, in fact, be contributing to the recent revitalization and repopulation of cities and the urban core. The claim here is both modest and tentative. The factors leading to America’s reinvestment in cities are varied and complex.199 Decreases in crime, changes in racial attitudes, growing opportunities for agglomeration surplus, and many other factors have contributed to the rebirth of a number of American cities in the last decades.200 The overall demographic change has received enormous attention from political scientists, urban economists, planners, and land use practitioners.201 Each offers a different lens for viewing the phenomenon. But the role of aversion to the risk of legal liability has not been previously identified and offers a promising additional explanation.

Cities and suburbs have long been in competition with each other over residents and mobile capital.202 Until recently, suburbs have won round after round, coming close to knockouts in the 1970s and 1980s. They have offered more space and newer housing for less money because land values

are cheaper where supply is greater.203 They have attracted affluent residents by using zoning and other land use controls to maintain relative homogeneity in local property values—prohibiting or limiting inexpensive housing—thus minimizing the extent of redistribution through property taxes.204 And they have promoted and benefitted from substantial investments in roads and highways, allowing residents to commute to work by car.205 They have, in short, offered a product—a combination of services and taxes—that residents seemed to want. The result was a kind of death spiral for cities. They hemorrhaged jobs and capital, raising taxes just to maintain declining levels of services, which caused more capital flight.206 Add race and poverty to the mix, and it seemed for a time as though suburbs’ ascendance was unstoppable.207

Cities, however, fought back. And they did so partly through regulatory innovations. Responding to policy pressures and changing consumer demand, cities in recent years have, in particular, become land use innovators. The Standard Zoning Enabling Act (SZEA), reflected in traditional Euclidean zoning, was based on the premise of separating incompatible uses.208 Like a kind of ex ante nuisance prevention, this overarching goal propelled and favored development of the traditional suburb with its exclusive dedication to single-family residential uses and automobile dependence.209 That paradigm dominated land use for the better (or worse) part of the twentieth century. But cities have embraced new paradigms, such as mixed-use developments with ground-level commercial spaces and residences on higher levels, new urbanism that favors walkability and seeks to replicate traditional town centers on a sublocal scale, transit-oriented development with greater density around transit hubs

and density limits elsewhere, and form-based codes that focus on buildings’ form instead of permissible uses at all.210

As demand among housing consumers for city living has increased, cities have leveraged their land use authority to manage growth while extracting value from developers. Most generally, cities have slowly reformed their response to development pressures. Conventional wisdom holds that cities are growth machines, in the thrall and political control of development interests—builders, construction workers, realtors, bankers, and so forth.211 Today, however, cities have been more aggressively seeking to capture some of the value created by new development instead of leaving that value entirely in private hands.212 Exactions and required dedications are the most obvious examples. But incentive zoning that gives height or density bonuses for certain concessions has the same effect by encouraging private developers to provide public benefits.213 City zoning ordinances create many bargaining moments when municipalities offer regulatory concessions in exchange for public benefits, such as new parks, schools, infrastructure, and so forth.

More specifically, cities have found new ways to attract and retain mobile capital by protecting property values in affluent neighborhoods. Suburbs are long practiced in this art, using large-lot zoning and holding zones to restrict supply.214 Those same tools are not appropriate to the dense urban core so cities have been developing their own modern techniques aimed at protecting and enhancing local property values.215 Historic preservation, for example, has become an important tool for sublocal

neighborhoods to slow or stop new development and protect property values. That is not, of course, the purpose of historic preservation, but it is often its effect, as local property owners well understand.216 Some cities have begun authorizing “neighborhood conservation districts,” which do not even use the pretense of historic significance to protect certain neighborhoods from development pressure.217

Downzoning upscale neighborhoods, implementing noncumulative industrial zones, and imposing steep exactions all limit opportunities for new residential development. All of these techniques, and others, have the effect of limiting or eliminating the development potential of great swaths of urban land, thus reducing supply and increasing property values to the benefit of in-place property owners.218

Simultaneously, cities have found ways to incentivize higher end development, catering increasingly to development proposals that include luxury housing, high-end retail, and upscale commercial space. For developments like these, cities will grant upzonings, issue needed permits, assemble land using eminent domain, develop infrastructure, create transferable development rights, and even grant tax abatements. Cities have always sought to attract mobile capital, but they have new opportunities to do so and have used land use tools to gain an advantage over their suburban competitors, often over the objections of neighbors who face increased congestion and dislocation caused by gentrification.219

These have become well-honed tools in a municipal government’s regulatory toolkit. Some are in conflict with each other, and not all are available in all instances. Nevertheless, it is clear enough that the overall nature of land use regulations in the urban core has shifted quite dramatically from the static version of zoning reflected in the SZEA. And in the overall balance of power between private property and regulatory

control, these changes all tilt towards increased governmental power and intervention in land development decisions.220

All of these regulatory innovations also implicate the Takings Clause. Where regulatory burdens are too restrictive, or where governments seek to extract too much value, property owners and developers can allege a regulatory taking. Most of these programs and regulatory approaches should and will survive takings scrutiny, but they are at least open to challenge. They therefore represent some risk, however small, of takings liability. These kinds of regulatory approaches, that extract value from developers and that favor the public over private interests, are therefore more valuable to less risk averse governments. Relative to suburbs and rural towns, cities will be more willing to adopt regulations that burden private property rights. The lack of insurance for regulatory takings liability offers one explanation why cities have been at the forefront of these innovations.221

For people who decry these trends—who see cities as inappropriately bleeding off profits from private developers—the relative risk aversion of suburbs and towns may seem like welcome relief. Of course, developers themselves may feel this way and flee the cities if public demands grow too high. The point here, though, is not to defend the appropriateness or desirability of different regulatory approaches. The insight, instead, is that risk aversion will affect different local governments’ willingness to experiment with these and other land use regulations unevenly. It will impede suburban more than urban experimentation. Whatever one’s view of the appropriate level of land use innovation, the disparity in these effects of risk aversion on regulatory incentives seems entirely haphazard and unprincipled.

In theory, it should be possible to test this claim about relative levels of urban and suburban land use innovation empirically. Such a test would be an article unto itself, however, and must therefore be reserved for future work. Impressionistically, though, the intuition is borne out in practice: the larger the municipality, the more complex the regulatory apparatus is that controls development and property rights. And there is some empirical

support for this claim.222 Of course, it may be the case that cities regulate more because they have more property conflicts to manage, or because they tend to be politically more liberal than suburbs. Causation, in other words, is difficult to impute. Nevertheless, available empirical evidence is at least consistent with the idea that less risk neutral governments in general regulate more than risk averse ones.

If this is right, and relative risk aversion has given cities a competitive edge over suburbs and smaller governments, then the end result may be desirable if the return to cities generates positive economic, social, and environmental benefits. But to the extent that the risk of regulatory takings liability puts a thumb on the scale in government decisionmaking, its weight varies haphazardly and irrationally with the size of the municipality. The risk of uninsured regulatory takings liability is not evenly distributed, and that means that regulatory incentives may vary with the size and wealth of the government. For those who embrace cities’ resurgence, the absence of takings insurance is still a problem.223 There are more direct ways to give them a competitive advantage instead of hobbling smaller governments’ incentives to regulate rationally.

If the absence of municipal liability insurance imposes all of the costs identified above, why does takings insurance not already exist? If coverage were actually so valuable, municipalities would want it, and insurance companies should be offering it already. The fact that they do not means either that there is some impediment to its development in the market, or that it is not actually as valuable as this Article claims. This Part explores those possibilities and, after identifying and refuting some of the reasons for the absence of takings insurance, argues that states should subsidize certain kinds of regulatory actions by insuring against any resulting litigation.

A. Explaining the Inverse Condemnation Exclusion

Why is there no municipal insurance for regulatory takings claims? It cannot simply be that insurance is unavailable for the policy decisions of municipal officials because that is precisely what E&O insurance covers. Nor do regulatory takings claims arise exclusively out of broad regulatory enactments. Permit denials by individual zoning administrators or boards can generate takings claims, and they are difficult to distinguish from the kinds of employment decisions or misconduct by municipal officials that are routinely insurable. There are some likely reasons for the absence of takings insurance, but they are only satisfying in conjunction with each other; there is no single explanation.

One significant reason for the inverse condemnation exclusion may be, quite simply, a misunderstanding on the part of insurers and municipal officials alike about the nature of regulatory takings claims.224 Tellingly, the relevant policy language excludes insurance for both condemnation and inverse condemnation claims.225 The absence of insurance for the former is entirely straightforward. Condemnation—eminent domain—is not a “fortuitous” event, and therefore does not represent an insurable loss.226 The existence and extent of the loss are entirely within the control of the government. If a government wants to avoid paying compensation for condemnation, it can choose simply not to take the property. If it wants to pay less, it can take less expensive property. While a government incurs costs when it takes property by eminent domain, there is no risk and therefore no role for insurance.227

The same might seem to be true for regulatory takings as well. If the government wants to avoid paying compensation, it can choose not to regulate. But the analogy is inapt. Inverse condemnation is in many respects less like eminent domain than it is like a car crash or other accident. While the government undoubtedly intended to enact the relevant regulation, in most cases it will not have intended to violate the Takings Clause.228 The problem from a municipality’s perspective is that the standards for determining takings liability are notoriously vague.229 This is an area of law rife with legal uncertainty—precisely the conditions under which liability insurance is particularly valuable.230 As a result, land use regulations come with genuine risk for the government of both litigation and liability.

Nor can a municipality avoid takings litigation by choosing not to regulate. In many cases, property owners sue when a planning commission fails to grant a rezoning request.231 A planning commission or zoning board can therefore find itself in a particular bind. If it denies the rezoning request, the property owner might sue. But if it grants the rezoning request, neighbors might sue for a taking of their property, or under due process or other challenge to the favorable treatment of neighboring property. This catch-22 dynamic means that local governments are faced with genuine risk in their land use decisions. Just like you cannot drive a car without incurring some risk of an accident, a municipality cannot administer a land use regime without some risk of takings litigation and liability.232 Facile comparisons to the eminent domain exclusion in municipal insurance policies therefore do not explain the exclusion for inverse condemnation.

Importantly, too, the Constitution requires a damages remedy—i.e., an inverse condemnation action—for takings violations.233 In other words,

regulatory takings claims necessarily create a risk of financial loss for a municipality. A local government cannot control its exposure through sovereign immunity because local governments are not considered an arm of the state and are not entitled to sovereign immunity.234 And if they were entitled to sovereign immunity, the Takings Clause itself directly abrogates it.235 Moreover, even if a local government changes or withdraws a challenged regulation after a court finds it to be a taking, the government is still obligated to pay damages for the time that the regulation was in effect.236 Measuring those damages can be quite complex, but the overall standard is the diminution in value during the time the regulation was in effect (sometimes measured by fair rental value).237 If the regulatory harm is irrevocable or the regulation remains in place, permanent damages are the only remedy available, measured by the fair market value of the property taken.238 Regulatory takings therefore are not like eminent domain: municipalities face real risk with regards to liability, precisely the kind of risk that insurance could but does not cover. The inverse condemnation exclusion cannot be explained by the absence of genuine risk.

There is another constellation of explanations: moral hazard, adverse selection, and information asymmetries. The moral hazard problem is the most obvious and the most significant. The concern is that the presence of insurance may induce local governments to discount risks of regulatory takings litigation that they should, in fact, internalize. A zoning board or planning commission might risk coming closer to the constitutional line if it knows that any resulting litigation will be covered by insurance. This, in turn, makes the risk difficult to insure because the existence of insurance will make the insured more likely to incur the covered loss.

This is a serious concern. But this same problem is common to many different contexts implicating municipal liability where insurance already exists. There is no greater moral hazard problem associated with regulatory takings than with other insurable municipal risks, like police misconduct, employment discrimination, or automobile accidents. In these cases, too, a municipality might exercise too little care in the selection and training of police, in the imposition of employment criteria, or in controlling driving by municipal employees, and yet insurance is generally available for such risks.239

It is tempting to try to distinguish these contexts from regulatory takings. At least some of them—police misconduct for example—do not involve municipal policy but instead an agent for the municipality taking some rogue action.240 The analogue to regulatory takings is not whether an individual police officer will disregard the risk of civil liability because of her insurance coverage (the classic moral hazard problem),241 but whether municipalities will fail to implement policies and practices to deter police misconduct because of the municipality’s coverage. In this way, it may seem easier for insurers to monitor police officers and thereby link premiums to care.242 In other words, police policies are mediated through the conduct of individual officers in a way that seems distinguishable from zoning regulations. The moral hazard problem in these other contexts might seem less acute.243

But this cannot be right, or at least does not describe a categorical difference between these different contexts. After all, regulatory takings claims, like the downzoning of property, can arise directly from legislative decisions. But they can also arise from the denial of a subdivision permit by the Board of Zoning Appeals (BZA), the designation of property for

historic protection by a preservation committee, or other discretionary action by a zoning official.244 Such decisions also amount to a kind of mediated implementation of municipal policy, where there is a possible disconnect between the policymakers of the city council or other legislative body, and the people or agencies tasked with implementing it. The analogy therefore can be quite apt: instead of the police beating someone up, the BZA or other zoning official is roughing up someone’s property.

The distinction breaks down from the other direction too because other insured losses are difficult to distinguish from legislative zoning decisions. If a municipality, for example, adopts discriminatory hiring policies—like requiring credit checks that systematically discriminate against minorities, or physical tests that discriminate against women—any resulting litigation will almost certainly trigger the government’s E&O policy.245 An Oregon court, for example, upheld a school district’s claims against its insurance carrier, seeking indemnification for settlements it reached in a series of discrimination claims over its hiring practices.246 The court distinguished between discrimination claims alleging disparate impact and those alleging intentional discrimination. The insured’s E&O policy covered the former because any resulting liability could be considered the result of “negligent acts, errors or omissions under the policy.”247 It would not cover the latter, however, because an element of such claims is that the injury was intended.248

Regulatory takings do not require a showing of intentional harm. Rather, they are more closely analogous to disparate impact claims, where an otherwise permissible regulation imposes too significant a burden on an individual property owner. Intent is simply not relevant to the takings

inquiry.249 In short, there is no conceptual difference between the moral hazard problem that attaches to insurance for regulatory takings and for claims of employment discrimination, or many other sources of municipal liability.

Looking outside the municipal context reinforces the intuition. E&O policies bear a very close resemblance to Directors and Officer (D&O) policies, which shield corporate management “from almost all liabilityrelated costs arising out of any wrongful acts alleged to have been committed in the course of their duties.”250 Like municipal insurance, D&O policies appear to present a serious moral hazard problem.251 But that insurance product nevertheless exists and is, in fact, ubiquitous.

Insurers have developed sophisticated ways of mitigating moral hazard problems in these and other areas.252 As Professor John Rappaport has recently explained, insurers limit moral hazard for police misconduct— pricing premiums for the appropriate level of care—by shaping police department policies, engaging in education and trainings, and auditing police practices.253 The same can be done in the context of land use regulation. Indeed, the League of Minnesota Cities, which appears to be the one municipal insurance pool that insures takings litigation, provides a similar set of resources and training for local zoning and land use officials.254 It offers telephone consultations, written guidance, and online training.255 There is nothing inherent in the land use process that would prevent insurance companies from adopting the same kinds of monitoring and loss-prevention techniques that they apply to the police.

Insurance companies can also structure their policies to minimize moral hazard. They may write policies with large deductibles, or policies with both a low ceiling and tall floor of coverage, effectively insuring only

the middle tranche of liability.256 These terms mean that the insured has “skin in the game” and retains an incentive to avoid a loss.257 Insurers may also price their policies according to relatively fine-grained assessments of local decisionmaking, evaluating the risk management practices in place that are designed to mitigate losses ex ante.258 Here, though, municipal liability insurance for takings claims runs into additional problems: adverse selection and information asymmetries.259

Adverse selection occurs when, for example, the worst drivers or the sickest people are more likely to purchase the relevant insurance.260 This can force the safest drivers and healthiest people out of the insurance pool because they are, in effect, cross-subsidizing the higher risk policyholders. The result is an insurance death spiral as the average risk in the pool increases.261 The same is true of municipalities. It may be that only those governments that perceive a genuine risk of takings litigation would be likely to purchase the insurance.262 Their perception of risk would be based on some combination of the regulatory environment and the litigiousness of local developers.

Insurance companies can address both adverse selection and moral hazard problems with better information and more fine-grained risk ratings or experience ratings.263 Charging worse drivers more, for example, will reduce the extent of the cross-subsidy in the insurance pool. It will also

force municipalities to internalize some of the risk of their decisions, thus limiting moral hazard.264 For takings claims, however, the fact that a municipality has not been sued in the past may provide little information about the likelihood of takings litigation in the future.265 Municipal elections that result in (or reflect) changed attitudes toward land use regulation, different personnel on the planning commission or board of zoning appeals, or newly emerging development pressures, can all create discontinuities in the risks of takings litigation.

To some extent, these problems are mitigated by claims-made policies, where insurance covers only claims made during the policy period based on conduct in the past. But for many regulatory actions, both the underlying act and the resulting litigation will both occur during the policy period. An insurer cannot price the policy, then, by looking only at a municipality’s history of regulating. Moreover, the kinds of factors that make takings litigation more or less likely in a particular municipality are going information asymmetries between the municipality and insurer leave plenty of opportunities for adverse selection.266

to be relatively

insurer. The

inscrutable

to an

Nevertheless, the law of large numbers is a powerful tool. So long as the risk pool is large enough, an insurer’s actuarial expertise should be able informational local to overcome heterogeneous asymmetries, especially if regulatory takings is just one of many risks covered.267 And again, there is nothing unique in this regard about takings liability. A municipality that hires new police officers or adopts new toughon-crime policies may also create discontinuities in the risk of § 1983 litigation, but insurance markets have overcome those.

conditions

and

Ultimately, Minnesota’s unusual coverage demonstrates that it is possible to create insurance for regulatory takings claims.268 The structure of that insurance provides a real-world example of how an insurer can

address moral hazard and adverse selection problems for land use litigation more broadly. First, the insurance comes with substantial co-pays, and an aggregate limit of $1 million annually.269 Moreover, the co-pays in Minnesota are based on a sliding scale that includes whether or not municipal officials have participated in a land use training program designed to minimize the risk of legal liability.270 Finally, the cost of municipal insurance through the Minnesota insurance pool is experience rated, and land use litigation is factored heavily into that rating.271 Therefore, municipalities participating in the pool retain some of the risk of takings liability, even with the insurance that the pool provides.

It is therefore clear enough that municipal insurance for takings litigation is, in fact, available, and could be made available much more widely. The reality, of course, is that such insurance is note widely available outside of Minnesota. The gap, then, may come from the demand side instead of the supply side. Perhaps regulatory takings are not salient enough to justify seeking insurance.272

To the contrary, while the risk of takings litigation is low, and the risk of ultimate liability more remote still, the costs when such losses occur can be very, very high. These are the conditions under which insurance can thrive.273 And if costs are truly unlikely, then the insurance will be inexpensive to purchase. Of course, the infrequency of takings litigation may pose an additional challenge if local officials are generally unaware of the risk. They will not seek insurance for risks that they have not that rare, however.274 And contemplated. Takings developers and property owners threaten to sue more often than they actually do.275 An empirical study demonstrated that planners had fairly detailed knowledge of changes in Supreme Court jurisprudence in just one

litigation

is not

specific area of takings law.276 Presumably, general knowledge of the possibility of takings liability is much broader.

Why, then, is regulatory takings insurance still generally unavailable to most municipalities? The explanation likely involves a combination of moral hazard, adverse selection, and misunderstandings about the nature of regulatory takings liability. It may also be that insurance companies have little incentive to innovate in this space. Anecdotally at least, competition for municipal insurance policies often does not turn on the quality of the coverage provided. Instead, local officials tend to select the policy that is either cheapest or that is being sold by friends or political allies.277 As a result, insurance companies may invest less in identifying and plugging holes in municipal insurance than they would in other kinds of insurance products.278 All of this means that the absence of takings insurance does not undercut the claim that such insurance would be valuable. Instead, it may reflect specific impediments to the development of a market for such insurance products.

This could all be written off as an odd but inconsequential lacunae in municipal risk management if the stakes were not so high. Ultimately, the inverse condemnation exclusion is important not because of inefficiencies in insurance markets but because it profoundly affects local governments’ regulatory incentives. And this then places municipal risk management back squarely into the center of core regulatory takings doctrine and theory.

B. State Insurance as Municipal Subsidy

The argument so far has been that regulatory takings insurance could exist and the focus has been primarily on the problem of risk averse local governments underregulating because of the risk of takings litigation and liability. But as the discussion of moral hazard revealed, insurance can also allow governments to ignore risks that they should, in fact, internalize.279 That is, the expected value of regulatory takings claims can loom too large

in the absence of insurance, but can appear too small if insurance is available, leading to overregulation.

There may be a private solution to this Goldilocks problem. And if the absence of takings insurance is attributable simply to the market’s misunderstanding of the nature of regulatory takings claims, then this Article’s descriptive sections may be antidote enough. Once insurers understand the nature of the risk, or once local governments are aware of the need for insurance, the market may step in to fill this gap. Insurers’ sophisticated actuarial expertise might allow them to design a product that is priced appropriately, that minimizes moral hazard by keeping local governments’ skin in the game, and that produces more efficient regulatory incentives.280 But is simultaneously too arrogant and too modest—arrogant because it supposes that markets have so far failed to understand an important form of municipal liability, and modest because it then hands the problem off to insurers to solve.

is also a good chance

this hope

there

that

One can hope that pointing out the problem will lead to its solution. But in the likely event it does not—in case the combination of moral hazard and adverse selection make private insurance difficult to offer—it is important to consider public solutions as well, and specifically the extent to which states should assume the risk of regulatory takings litigation from municipalities.281 This would amount to a kind of subsidy for local governments, and it turns out to be an especially effective way for the state to encourage particular regulatory initiatives at the local level. It is both more targeted and more efficient than currently available tools, including even direct grants.282 And it responds to a specific and previously unidentified justification for state subsidies in the first place: local governments’ risk aversion.

States currently subsidize many local government activities. In fact, direct grants in aid have become the dominant source of local revenue, eclipsing even property taxes in many jurisdictions.283 The justifications vary in their details depending on the particular regulatory context, but they take familiar general forms. First, states subsidize some local activities that

generate positive externalities. Public schools, roads, environmental protection, and the like all create benefits that cannot be fully captured by individual municipalities. There is, therefore, a real concern that local governments will underinvest in those activities.284 State subsidies help to realign local incentives with the interests of the state as a whole.

Second, states subsidize some activities for redistributive reasons. There is a conventional account that local governments cannot engage in redistributive policies because of the risk of capital flight, and so redistribution must occur at the state level.285 And states do this routinely. Again, public schools are a leading example, but so are affordable housing and other social welfare programs, all of which are often supported by the state. Such redistributive policies operate at two different geographical levels.286 In the case of education in particular, state funding for local public schools serves in many states to redistribute money from wealthier municipalities to poorer ones.287 And intra-locally, it serves to protect poor and relatively powerless groups from the vicissitudes of local politics, which are often dominated by homeowner majorities.288 A community in which the most powerful constituents are not invested in public education can still not entirely remove funding from its schools.

Lastly, a state may subsidize regulatory efforts that require more expertise than local governments have on their own. Subsidies for land use planning, coastal management, and others have a long history and seek to combine the benefits of local knowledge with the state’s greater resources and expertise.289

This Article offers a fourth

justification for state subsidies: overcoming the risk aversion of local governments to takings litigation. Risk aversion can result in a local government underregulating or underenforcing its regulations when faced with the prospect or threat of

takings litigation.290 This justification for state intervention also suggests a new form of subsidy: insurance against takings litigation. In fact, this is likely to be a particularly effective and efficient form of subsidy.

Imagine a local government considering growth controls aimed at curbing sprawl. The new measures might include downzoning some undeveloped property, adopting new regulations limiting infrastructure expansions, enacting newly restrictive subdivision regulations, and so forth. The slate of regulatory changes creates some local costs and benefits, including the possibility—indeed, in this case the likelihood—of regulatory takings litigation by adversely affected property owners. There is no reason in the abstract to expect the local government should come down on one side or another, but assume here that a risk neutral government would rationally calculate the benefits as greater than the costs, while a risk averse government would not and so would not act.

The state, however, may also have an interest in curbing sprawl. It may recognize externalized benefits of preserving open space—both aesthetic and ecological—as well as other positive benefits, such as reducing driving and so reducing road maintenance costs and energy consumption, improving public health outcomes, facilitating public transportation, and so forth. State decisionmakers may also benefit from the advice of professional planners and policy experts who endorse this kind of development even if it remains unpopular among many housing producers and consumers. The state, then, may well want to weigh in to encourage local adoption of these pro-density local measures. If it wants to place a thumb on the scale, how should it do so?

The traditional tool is to provide direct financial assistance. That can take a number of likely forms in this example, such as grants to support public transportation that new denser development will require, subsidies for conservation initiatives, funding for planning assistance, and the like. These can be extremely valuable, and can tip the scales in favor of the local government’s action. But their value depends at least in part on the risk aversion of the local decisionmakers.

Faced with the prospect of uninsured regulatory takings liability, a risk averse local government will require a greater subsidy from the state to offset its risk aversion, and perhaps even a subsidy greater than the value of the positive externalities anticipated by the state.291 Alternatively, the state

could offer to indemnify local governments for takings litigation—what amounts to state insurance for regulatory takings.292 The advantage is that a state can, in effect, take advantage of its risk neutrality to subsidize local governments by assuming certain risks. The asymmetry in the value of risk means that the expected value of the contribution will always be higher to the municipality than to the state.293 Every $1 of risk that the state assumes will be worth more than $1 to a risk averse local government. A subsidy in this form expands the size of the pie.

There is an additional reason that takings liability might have a smaller expected loss for the state than for a local government. The willingness of adversely affected property owners to sue may diminish once they know that the state will be paying litigation expenses and ultimately footing the bill.294 Developer interests often have relatively deep pockets, and sometimes even deeper pockets than small local governments. By threatening to sue a local government, or actually filing a complaint, a developer can often force a local government to capitulate. The state, however, is not so easily bullied by the threat of litigation. Developers may actually sue less often if they know that the state will be picking up the tab because they can anticipate that the state will not capitulate so easily to the threat of litigation. Therefore, the risk of takings liability may decrease when the state steps in, regardless of relative risk aversion. Indeed, in other contexts, states (and the federal government) do provide insurance as a form of subsidy to private parties, and it can be a particularly valuable one.295

try

insurance product

It is clear enough that the state can provide a valuable subsidy for local regulatory actions by assuming to itself some or all of the risks of takings litigation and liability. But what form should this take? A state could, of course, to to provide an actual municipalities. In some stylized and idealized world, state insurance could be roughly calibrated to the risk aversion of each individual municipal government. The more risk averse the government, the greater the insurance protection that should be available. An ambitious state could attempt to approximate this outcome with coverage that varies in a more or less fine-grained way with the size and wealth of individual municipalities. If private markets cannot figure out how to provide this product, however—the assumption motivating this final discussion—it is difficult to see why states would fare much better. A more limited and therefore more promising approach does not seek to extend comprehensive regulatory takings insurance to all municipalities through complex risk-rating stratagems. Instead, it extends indemnification for regulatory takings claims as a kind of ad hoc state subsidy to incentivize particular kinds of local regulations.

This proposal could be implemented in many different ways, and there are undoubtedly state-level legal challenges that would require further work to evaluate. Nevertheless, in broad strokes, the idea is straightforward: states should offer takings liability protection in particular predefined regulatory settings. In so doing, states could (and already often do) impose significant limits on the form and content of local regulatory conduct they are willing to insure, thereby minimizing moral hazard concerns and exerting state policy pressure.296 To consider just one of innumerable examples, a state that wanted to encourage municipalities to plan for sea level rise could extend regulatory takings protection, but could also require oceanfront communities to include certain elements in their plans (e.g., soft-armoring, elevating buildings, infrastructure resiliency) but not others (e.g., managed retreat, rolling easements), depending on the state’s policy preferences. This is like an insurance company requiring an insured to keep working smoke detectors on her property. It may well be that the most important and valuable regulatory options are those that come with the greatest risk of takings liability. Where—as here—that is true, the state need not prohibit those options. That simply increases both the impact and the cost of the state’s subsidy.

In addition to extending takings insurance in specific contexts, a state could also create an even more ad hoc mechanism by which local governments petition the state for takings insurance for any regulatory action. The state would weigh the merits of the local decision at issue and decide whether it is the sort of regulatory action justifying a state subsidy, and whether the particular municipality is, in fact, risk averse. If the answer to both questions is yes, the state could subsidize the action by agreeing to defend regulatory takings claims and pay any resulting liability.

This could be extremely beneficial in the context of even routine development decisions. Imagine a developer with a controversial development proposal to which the local government objects. That developer may seek any number of regulatory approvals, from upzoning the property to subdivision permits. And the local government might be reluctant to deny those requests, even if it disapproves of the development, for fear of takings litigation. The proposal here would allow the local government to petition the state for takings insurance in denying the permits. From the state’s perspective, the decision whether to extend that protection to the local government will depend entirely on the state’s judgment of the value (or harm) of the development. If the state agrees with the local assessment that the development will be problematic—perhaps imposing worsening regional environmental costs, among others—the state could subsidize the local government’s refusal to allow the development. But, of course, if the state perceives to development as exclusionary, as protectionism for in-place property owners, or as reflecting some other problematic motive, the state could deny coverage.

infrastructure, and

traffic, burdening

local hostility

the

This proposal clearly raises the possibility of lobbying at the state level. In essence, giving the state discretionary power to protect local governments from takings liability simply upstreams the politics of development decisions to the state. That is true, but it cannot make local governments worse off than they are now. If a state bows to developer pressure and declines to insure a local government against regulatory takings claims, that local government is in precisely the same uninsured position it is in today. The discretionary regime proposed here may well result in too little protection for local governments—at least relative to the goal of incentivizing risk averse decisionmaking—but it is nevertheless a substantial improvement. And that should be enough, unless and until private insurers find a way to provide actuarially appropriate coverage that is more precisely tailored to address local risk aversion and moral hazard.

Thinking clearly about municipal risk opens up both new justifications for, and new forms of, state subsidies. Traditional justifications for state

intervention depend on the positive (or negative) externalities of local regulatory decisions, or on distributive concerns and problems of equity. But risk aversion provides a new reason why local governments may forego regulatory actions that would, in fact, make citizens of the state better off. In assuming some of that risk—by indemnifying local governments for the costs of regulatory takings litigation—states could improve local decisionmaking, and provide a targeted but valuable new subsidy for local governments.

The Takings Clause has given rise to important and high-stakes fights through the years and has generated a voluminous academic literature along the way. Important substantive issues implicate some of the deepest questions in the law: the relationship between the government and private rights; the distribution of regulatory burdens; and the nature of property. Those issues will undoubtedly continue to consume scholarly attention for decades to come. But sometimes more granular and practical concerns have important lessons for applying the theory on the ground. Municipal risk management practices are one such topic. It turns out to be impossible to predict the impact of different, contentious takings rules on local regulatory incentives without a clearer understanding of how local decisionmakers manage municipal risk. When it comes to takings liability, insurance mechanisms are surprisingly unavailable.

Identifying and explaining the inverse condemnation exclusion in municipal insurance policies may be enough to trigger change. Private insurers may realize that there are opportunities to sell insurance against this genuine risk. But in the likely event they do not, because moral hazard and other market impediments prove insurmountable, then this could prove an opportunity for developing an important new state subsidy. A state wishing to incentivize specific municipal land use controls could extend insurance against any resulting takings claims instead of, or in addition to, more direct grants in aid. And at the very least, recognizing the potential impact of risk aversion on local regulatory incentives suggests important reasons for caution before exposing municipalities to greater liability. The absence of insurance for risk averse governments means that the threat of takings liability may result in underregulation and underenforcement of important land use and environmental regulations.

--- FOOTNOTES ---

Copyright 2016 by Christopher Serkin

Printed in U.S.A. Vol. 111, No. 1

INTRODUCTION ................................................................................................................. 76 I. THE NATURE OF INSURANCE AND MUNICIPAL RISK .................................................... 80 A. Municipal Risk and the Role of Insurance ........................................................ 81 B. Municipal Risk Management Strategies and Practices .................................... 88 II. THE ABSENCE OF INSURANCE FOR REGULATORY TAKINGS ....................................... 102 A. The Inverse Condemnation Exclusion ............................................................ 103 B. Takings Risk and Regulatory Incentives ......................................................... 110 C. Distributional Effects of the Inverse Condemnation Exclusion ...................... 114 III. ADDRESSING THE INVERSE CONDEMNATION EXCLUSION .......................................... 120 A. Explaining the Inverse Condemnation Exclusion ........................................... 121 State Insurance as Municipal Subsidy ............................................................ 130 B. CONCLUSION .................................................................................................................. 137

INTRODUCTION

142 U.S.C. § 1983 (2012) (creating private right of action for constitutional violations). 2See, e.g., Johnson ex rel. Johnson v. City of Chicago, 807 N.E.2d 1100, 1102 (Ill. App. Ct. 2004)

(rejecting City’s claim of immunity when fence gate surrounding public library fell on plaintiff’s leg).

3See, e.g., Santiago v. City of Vineland, 107 F. Supp. 2d 512, 526 (D.N.J. 2000) (involving claim

by police officer that he was discharged because of his race).

4U.S. CONST. amend. V. See, e.g., Mayhew v. Town of Sunnyvale, 964 S.W.2d 922, 930 (Tex. 1998) (involving claims by property owners for violations of Takings Clause when town refused to rezone their property to permit dense development).

5See infra Section I.A. 6See infra Part II. 7See discussions infra Section II.A.

8Public risk management is occasionally discussed in other fields, such as political science. See, e.g., Yuhua Qiao, Public Risk Management: Development and Financing, 19 J. PUB. BUDGETING, ACCT. & FIN. MGMT. 33 (2007).

9For one recent and impressive exception that explores the impact of public insurance on police conduct, see John Rappaport, How Private Insurers Regulate Public Police, 130 HARV. L. REV. (forthcoming http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2733783&download=yes [https://perma.cc/QTB4-RZFH].

2017),

10Yamagiwa v. City of Half Moon Bay, 523 F. Supp. 2d 1036, 1087 (N.D. Cal. 2007) (ordering damages of $36,795,000). Plaintiffs brought an inverse condemnation claim, which is the primary form of bringing a regulatory takings claim against a state. For discussion of “inverse condemnation,” see infra text accompanying note 162.

11John Coté, Half Moon Bay Grapples with $36.8 Million Judgment Against It, SFGATE (Dec. 18, 2007, 4:00 AM), http://www.sfgate.com/bayarea/article/Half-Moon-Bay-grapples-with-36-8-million3234399.php [https://perma.cc/L2DM-XM7V]. As one County Supervisor put it:

One of the options, candidly, is . . . to dissolve . . . . That’s an extreme. But when you get a judgment of $36 million plus legal fees . . . even if you were able to finance it and stretch it out over a period of time, you would need significant reductions in your level of service to pay that off.

Id. (internal quotation marks omitted).

12See Aaron Kinney, Half Moon Bay: City Wins $10 Million in Arbitration, SAN JOSE MERCURY NEWS (Sept. 21, 2012, 4:13 PM), http://www.mercurynews.com/ci_21603337/half-moon-bay-citywins-10-million-arbitration [perma.cc/X3MA-UUQQ]; see also Mark Noak, City Seeks Payback for Beachwood Loss, HALF MOON BAY REV. (May 10, 2012), http://www.hmbreview.com/news/cityseeks-payback-for-beachwood-loss/article_4ab2f7e4-9acb-11e1-b7a5-0019bb2963f4.htm?mode=print

[perma.cc/6RFE-UQSB] (describing effort to pursue lapsed insurance policy); Press Release, City of Half Moon Bay, Half Moon Bay Mayor Announces Significant Legal Victory in Yamagiwa Insurance Claim (2012) [perma.cc/8976-PU9T] (describing details of award). For discussion on the difference between claims-made and occurrence-based policies, see infra text accompanying notes 90–93.

13See infra Part II. 14RICHARD A. EPSTEIN, BARGAINING WITH THE STATE 84–85 (1993); see also RICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW 73–74 (8th ed. 2011); Christopher Serkin, Passive Takings: The State’s Affirmative Duty to Protect Property, 113 MICH. L. REV. 345, 361–62 (2014) (describing this conventional economic account).

15See, e.g., Lawrence Blume & Daniel L. Rubinfeld, Compensation for Takings: An Economic Analysis, 72 CALIF. L. REV. 569 (1984); Daryl J. Levinson, Making Government Pay: Markets, Politics, and the Allocation of Constitutional Costs, 67 U. CHI. L. REV. 345 (2000).

16See supra text accompanying notes 10–12 (describing saga of Half Moon Bay). 17One indication of this—albeit indirect—is that up to 41% of insurance costs are “attributable to defense costs.” Ellen S. Pryor, The Tort Liability Regime and the Duty to Defend, 58 MD. L. REV. 1, 6 (1999).

18See infra Section I.A. 19For an earlier suggestion that risk aversion of municipalities tracks their size, see Christopher Serkin, Big Differences for Small Governments: Local Governments and the Takings Clause, 81 N.Y.U. L. REV. 1624, 1668 (2006) (“Government risk aversion therefore correlates more to the size than to the wealth of the tax base, and it is inversely related to the number of taxpayers over whom the risk is spread.”).

20See infra Section II.C. 21See, e.g., EDWARD GLAESER, TRIUMPH OF THE CITY (2011) (identifying rebirth of American

cities).

I. THE NATURE OF INSURANCE AND MUNICIPAL RISK

22See, e.g., Peter Siegelman, Adverse Selection in Insurance Markets: An Exaggerated Threat, 113 YALE L.J. 1223, 1267 (2004) (“[E]conomists use the term ‘risk aversion’ in a much narrower sense than does the general public. To an economist, risk aversion means only that the marginal utility of wealth declines as wealth increases.”).

23KENNETH J. ARROW, ESSAYS IN THE THEORY OF RISK-BEARING 90 (1971). 24STEVEN SHAVELL, FOUNDATIONS OF ECONOMIC ANALYSIS OF LAW 258 (2004). 25Tom Baker & Peter Siegelman, The Law & Economics of Liability Insurance, in RESEARCH HANDBOOK ON THE ECONOMICS OF TORT 169, app. at 194 (Jennifer Arlen ed., 2013) (summarizing conventional model of risk aversion).

26Cf. A.H. Maslow, A Theory of Human Motivation, 50 PSYCHOL. REV. 370, 386 (1943) (positing

human “hierarchy of basic needs”).

27See Siegelman, supra note 22, at 1265 (“Experience suggests that people differ in the extent to which they are averse to financial risks. Someone whose marginal utility of wealth falls off very rapidly as wealth increases is more risk-averse than someone whose marginal utility of wealth does not change much as she becomes wealthier.”).

28Risk aversion is idiosyncratic to a certain extent. Some people are simply more risk averse, or more risk seeking, than others. In general, risk aversion is inversely correlated with wealth. See, e.g., Blume & Rubinfeld, supra note 15, at 604 (“The general view held by economists is that absolute risk aversion declines with wealth.”); see also Daniel A. Farber, Economic Analysis and Just Compensation, 12 INT’L REV. L. & ECON. 125, 127 (1992) (“[T]he wealthy should be less willing to insure against losses of the same size than the poor, since a smaller portion of their total wealth is at risk.”).

29Cf. Siegelman, supra note 22, at 1265 (“The insured pays a small amount in good times, reducing wealth slightly at a time when wealth is plentiful and the money spent on insurance is relatively painless.”).

30See Serkin, supra note 19, at 1666 nn.163–64 (citing sources indicating government neutrality). 31For earlier work introducing the possibility of governmental risk aversion, see id. 32See Blume & Rubinfeld, supra note 15, at 616 (identifying that government officials may be risk averse if their constituents are risk averse); cf. Kenneth J. Arrow & Robert C. Lind, Uncertainty and the Evaluation of Public Investment Decisions, 60 AM. ECON. REV. 364, 370 (1970) (evaluating risk bearing by public entities and asserting that “a public investment can be considered an investment in which each individual taxpayer has a very small share.”).

33Serkin, supra note 19, at 1652–53 (discussing role of property taxes in financing local

governments).

34Arrow & Lind, supra note 32, at 373 (“If the size of the share [of risk] borne by each taxpayer is a negligible component of his income, the cost of risk-bearing associated with holding it will be small . . . . This situation will exist where the investment [or risk] is small with respect to the total wealth of the taxpayers.”).

35See Arrow & Lind, supra note 32, at 366 (“[G]overnment distributes the risk associated with any investment among a large number of people.”); Serkin, supra note 19, at 1668 (“It is as if governments diversify their exposure to risk through the sheer number of their constituents.”).

36Cf. George A. Warp, Tort Liability Problems of Small Municipalities, 9 L. & CONTEMP. PROB. 363, 367 (1942) (proposing capping liability for municipalities smaller than 10,000 people because of their particular susceptibility to the problem of large adverse judgments). That is especially true because larger local governments can often raise revenue through other forms of taxation, like sales or income taxes.

37Managers of insurance pools—discussed infra Section I.B.3—use debt in precisely this way. See Amy V. Puelz & Robert Puelz, Managerial Use of Debt to Fund Municipal Government Risks, 28 DECISION SCI. 745, 746 (1997) (“[A]mong [intergovernmental risk pools], the predominant alternative risk-financing strategy chosen by managers is the issuance of debt.”).

38If an example is useful, imagine a government facing a legal claim with an expected value of $1 million (perhaps a 25% chance of a $4 million loss). The government could fund that loss ex ante through insurance or ex post through debt. An insurance company might be willing to insure against that claim for $1,100,000 (converting the risk of a $4 million loss into a certain loss of only $1.1 million, where the extra $100,000 represents the risk premium being divided between the insurer and the insured). The alternative for the municipality is to wait and fund any liability through debt. If the claim fails, the government will owe nothing. But if there is a $4 million judgment, the government could fund this through a ten-year bond at 2.5% (a typical product in today’s markets). This will end up costing the government approximately $525,000 in interest. The cost-benefit calculus then should be whether the $100,000 insurance premium is preferable to the expected value of the interest it would owe if forced to float a bond, here a 25% chance that it will have to pay $525,000 (or $131,250). Of course, the actual calculus is even more complex because the insurance premiums are paid today, whereas the interest on debt is paid in the future and so is subject to further discounting.

39See Christopher Serkin, Public Entrenchment Through Private Law: Binding Local Governments, 78 U. CHI. L. REV. 879, 926 (2011) (discussing debt limits). However, there is reason to think that debt limits do not apply to raising money to satisfy regulatory takings claims. See F & L Farm Co. v. City Council, 77 Cal. Rptr. 2d. 360, 361 (Cal. Ct. App. 1998).

40See Serkin, supra note 39, at 926–27 (describing bond election requirements).

41See, e.g., MOODY’S INV’R SERV., RATING METHODOLOGY: US LOCAL GOVERNMENT GENERAL

OBLIGATION DEBT 3 (2014) (on file with author) (providing overview “scorecard” for bond rating).

42See, e.g., STANDARD & POOR’S, ABOUT CREDIT RATINGS (2012) (on file with author) (“In rating an issuer, such as a corporation or municipality, analysts conduct a review of the financial performance, policies, and risk management strategies of that issuer as well as of the business and economic environment in which the issuer operates.”); cf. TOM BAKER & SEAN J. GRIFFITH, ENSURING CORPORATE MISCONDUCT: HOW LIABILITY INSURANCE UNDERMINES SHAREHOLDER LITIGATION 64– 65 (2010) (describing increased borrowing costs for corporations without insurance).

43See Wallace E. Oates, The Effects of Property Taxes and Local Public Spending on Property Values: An Empirical Study of Tax Capitalization and the Tiebout Hypothesis, 77 J. POL. ECON. 957, 959 (1969); Charles M. Tiebout, A Pure Theory of Local Expenditures, 64 J. POL. ECON. 416, 420 (1956); see also David J. Barron & Gerald E. Frug, Defensive Localism: A View of the Field from the Field, 21 J.L. & POL. 261, 265 (2005) (citing PAUL PETERSON, CITY LIMITS (1981 (identifying cities’ singular interest in pursuing economic development to compete with other municipalities); Richard C. Schragger, Mobile Capital, Local Economic Regulation, and the Democratic City, 123 HARV. L. REV. 482, 484 (2009) (“Cities nonetheless have long sought to entice mobile capital.”).

44See Barron & Frug, supra note 43, at 282–83 (discussing results of interviews with local officials

identifying aversion to financial risks).

45This is related to the political salience of taxation, a topic that has received significant scholarly attention recently. See, e.g., Andrew T. Hayashi, The Legal Salience of Taxation, 81 U. CHI. L. REV. 1443, 1454 (2014) (“The political salience of a tax refers to the effect of its visibility or prominence on political decisions.” (footnote omitted . While this literature primarily examines and evaluates how different systems of taxation either minimize or exacerbate political salience, it also implicitly acknowledges that taxation can have outsized political costs. Cf. David Gamage & Darien Shanske, Three Essays on Tax Salience: Market Salience and Political Salience, 65 TAX L. REV. 19, 57 (2011) (“[C]onsumers do receive personal property tax bills, and, assuming these bills arrive before an election (as they must since assessed annually), personal property taxes would seem to have higher political salience . . . .”).

46See, e.g., Hal Dardick, Even with Emanuel Hike, City Homeowner Property Tax Rates Still Below Suburbs, CHI. TRIB. (Nov. 13, 2015, 11:59 PM), http://www.chicagotribune.com/news/ local/politics/ct-chicago-suburbs-tax-comparison-met-20151112-story.html [perma.cc/8MAZ-DSMH] (“Since at least the late 1980s, the property tax has been considered the third rail of Illinois politics.”); cf. HUMAN RIGHTS WATCH, SHIELDED FROM JUSTICE: POLICE BRUTALITY AND ACCOUNTABILITY IN THE UNITED STATES 108 (1998) (“[E]ven significant payouts in [civil rights] cases do not have much of an effect on the city’s operations, and only lead to change when they become an embarrassment.”).

47For discussion of the law of large numbers, see infra note 81.

48See WILLIAM A. FISCHEL, THE HOMEVOTER HYPOTHESIS: HOW HOME VALUES INFLUENCE

LOCAL GOVERNMENT TAXATION, SCHOOL FINANCE, AND LAND-USE POLICIES 39–57 (2001).

49BAKER & GRIFFITH, supra note 42. 50See id. 51FISCHEL, supra note 48, at 4 (“For the great majority of . . . homeowners, the equity in their home is the most important savings they have. . . . [M]edian housing equity is more than 11 times as large as median liquid assets among all homeowners . . . .” (quoting Gary V. Engelhardt & Christopher J. Mayer, Intergenerational Transfers, Borrowing Constraints, and Saving Behavior: Evidence from the Housing Market, 44 J. URB. ECON. 135, 136 (1998) (internal quotation marks omitted .

52See supra note 38. 53See, e.g., George L. Priest, The Current Insurance Crisis and Modern Tort Law, 96 YALE L. J. 1521, 1527 (1987) (discussing insurance crisis); Ralph A. Winter, The Liability Insurance Market, 5 J. ECON. PERSP. 115, 117–20 (1991) (discussing causes of market cycles, with focus on hardening of markets in 1980s).

54See Priest, supra note 53, at 1527.

AM),

55See id. 56Id. 57That political cost was recently on display when cities were pilloried for banning sledding on public hills due to liability concerns. See Melinda Wenner Moyer, Let Them Sled!, SLATE (Jan. 9, 2015, 10:56 http://www.slate.com/articles/double_x/the_kids/2015/01/sledding_bans_from_iowa _to_new_jersey_cities_are_outlawing_sledding_in_city.html [perma.cc/YKR9-4YXX]; see also Scott McFetridge, Liability Concerns Leave Cities Uneasy as Sledders Whoosh Down Snow-Covered Slopes, STAR TRIB. (Jan. 4, 2015, 3:00 PM), http://www.startribune.com/liability-concerns-prompt-some-citiesto-limit-sledding/287450011/ [perma.cc/C8NM-C9FM]. For a modern version, see Kevin Leininger, On Thin Ice: Residents Fuming over Park Skating Ban, Offer to Help Bring ‘Gem’ Back, FORT WAYNE NEWS-SENTINEL (Jan. 17, 2015, 12:01 AM), http://www.news-sentinel.com/news/local/leininger/Onthin-ice—Residents-fuming-over-park-skating-ban—offer-to-help-bring—gem—back [perma.cc/CK3L -XCB9] (discussing political costs of closing skating in city parks).

58But see Rappaport, supra note 9.

59See N.C. DEP’T OF STATE TREASURER, POLICY MANUAL FOR LOCAL GOVERNMENTS, SECTION 85: INSURANCE AND RISK MANAGEMENT 33 (2014) [hereinafter POLICY MANUAL] (“Errors and omissions policies are not standard policies and the scope of coverage varies greatly between insurers.”).

60Qiao, supra note 8, at 43 (listing the most expensive and commonly generated sources of

liability).

61Id. (identifying sources of liability and respective costs). 6242 U.S.C. § 1983 (2012). 63Robert L. Pratter & Joanne A. Baker, The Status of Personal Liability and Comprehensive General Liability Insurance Coverage of Civil Rights Damages, 48 INS. COUNS. J. 259, 259 (1981) (“Enacted in 1871, 42 U.S.C. § 1983 grants a right of recovery for money damages to any person who has been deprived by another acting under color of state law of any right, privilege or immunity granted by the Constitution or federal law.”). Originally, § 1983 suits were only available to challenge official government policies that violated constitutional rights. Over time, however, that limitation changed, and § 1983 was expanded to cover conduct of public officials simply acting under color of law, even if their actions were not specifically contemplated or permitted under state or local law or policy. See Monroe v. Pape, 365 U.S. 167 (1961); see also Martin J. Jaron, Jr., The Threat of Personal Liability Under the Federal Civil Rights Act: Does It Interfere with the Performance of State and Local Government?, 13 URB. LAW. 1, 5–8 (1981) (describing history of § 1983); Joanna C. Schwartz, Police Indemnification, 89 N.Y.U. L. REV. 885, 888 (2014) (describing current § 1983 doctrine).

64Cf. Monell v. Dep’t of Social Services, 436 U.S. 658, 663 (1978) (establishing that § 1983 liability can run against the government that employs a government actor); Jaron, supra note 63, at 8 (“After Monell, a cause of action can still be stated against an individual public official, but his municipal employer can now be joined as well.”).

65See, e.g., Richard Emery & Ilann Margalit Maazel, Why Civil Rights Lawsuits Do Not Deter Police Misconduct: The Conundrum of Indemnification and a Proposed Solution, 28 FORDHAM URB. L.J. 587, 587 (2000) (finding that New York City represents and indemnifies police officers in the overwhelming majority of civil rights litigation). This can be controversial. Some practitioners and scholars have worried that municipal indemnification for officials sued under § 1983 will underdeter wrongdoing. See, e.g., HUMAN RIGHTS WATCH, supra note 46, at 102 n.149; Donald J. Farley, Insurance Coverage in Civil Rights Cases, 20 IDAHO L. REV. 617, 634 (1984); Pratter & Baker, supra note 63, at 263. For a more comprehensive treatment of the relationship between indemnification and insurance, see Gary T. Schwartz, The Ethics and the Economics of Tort Liability Insurance, 75 CORNELL L. REV. 313 (1990).

66Cf. HUMAN RIGHTS WATCH, supra note 46, at 108 (“The individual officer who is the subject of a police misconduct lawsuit found in favor of the plaintiff is rarely forced to pay the victim.”). See generally Schwartz, Police Indemnification, supra note 63 (finding that indemnification is ubiquitous).

67Emery & Maazel, supra note 65, at 592 (“If our society is to encourage public service and attract qualified public servants, public officials cannot face financial ruin for every careless mistake that causes someone damage.”); Jaron, supra note 63, at 21.

68Marcus Rayner, Op-Ed: The High (and Hidden) Costs of Lawsuits Against Local Governments, N.J. SPOTLIGHT (May 26, 2011), http://www.njspotlight.com/stories/11/0525/2157/ [perma.cc/3KVZM4EZ].

69Emery & Maazel, supra note 65, at 590. 70SYDNEY CRESSWELL & MICHAEL LANDON-MURRAY, ASSESSING THE FISCAL IMPACT OF LAWSUITS ON NEW YORK STATE MUNICIPALITIES 9 (2011), www.nylawsuitreform.org/wp-content/ uploads/2011/06/Municipal-Lawsuit-Report-One.pdf [perma.cc/7UTX-4WF2].

71See LA. LAWSUIT ABUSE WATCH, DRINKING FROM THE TAXPAYER TROUGH: THE COST OF LAWSUITS AGAINST LOUISIANA MUNICIPALITIES 5–6 (2011), http://www.houmatoday.com/assets/pdf/ HC22040212.pdf [https://perma.cc/7JHJ-X88N] (estimating statewide costs of $94,271,472 over fouryear period).

72Risk management varies. For a sample of different approaches, see La. Mun. Ass’n, PowerPoint: Risk Management for Municipal Government, http://www.lma.org/Docs/Community%20Leadership/ Risk_Management_For_Municipal_Government2.pdf best practices to manage municipal risk); Risk Management, MICH. MUN. RISK MGMT. AUTHORITY, www.mmrma.org/risk-management/ [perma.cc/XV4W-H6QK] (describing risk management practices in Michigan, including model policies and training). For a more comprehensive discussion of risk management specifically in the context of policing, see Carol A. Archbold, Managing the Bottom Line: Risk Management in Policing, 28 POLICING: INT’L J. POLICE STRATEGIES & MGMT. 30 (2005).

[perma.cc/8AB2-EEUL]

(describing

73See George L. Priest, The Antitrust Suits and the Public Understanding of Insurance, 63 TUL. L. REV. 999, 1006–07 (1989) (discussing captive insurance as form of self-insurance); see also Thomas W. Rynard, The Local Government as Insured or Insurer: Some New Risk Management Alternatives, 20 URB. LAW. 103, 111 (1988) (“A true self-insurance plan contemplates the establishment of a fund based on projections of future losses and the identification and measurement of possible and actual claims against the self-insured entity so that money from the fund may be set aside to pay those claims if and when they come due.” (emphasis omitted .

74Some people distinguish between active and passive risk retention. They call the former selfinsurance. The latter simply amounts to ignoring the risk. See POLICY MANUAL, supra note 59, at 21 (“Passive risk retention is not a risk management tool; no risk management decision has been made. Active retention, commonly known as ‘self-insurance,’ is when the unit identifies the risk and consciously selects risk retention as the appropriate alternative to finance the risk.”); cf. Schwartz, Tort Liability Insurance, supra note 65, at 315 (calling “self-insured” an “odd euphemism for ‘uninsured’”).

75See POLICY MANUAL, supra note 59, at 21–22. 76See, e.g., id. at 21.

77Insurance, by definition, involves the transferring of risk to a third party, which does not occur

when the local government retains the risk and pays costs as they come due. See id. at 21.

78See, e.g., Schwartz, supra note 65, at 316 (“Almost every liability insurance policy written in this country has some cap or policy limits; above this cap the defendant is without insurance.” (footnote omitted .

79Cf. id. (“By 1988, the City of New York was wholly self-insured; apart from a limited number of specialized policies, so were the City of Los Angeles, the County of Los Angeles, and the State of California.” (footnotes omitted .

80See supra text accompanying note 36. 81See Priest, supra note 53, at 1540 (“Applied to insurance, the law of large numbers means that as one increases the number of insured persons possessing independent and identically-valued risks, one increases the accuracy of prediction of expected loss for each individual.”).

82While large cities still face risks of loss, those risks become increasingly predictable and therefore easier to price ex ante as the number of cases climbs. For the seminal discussion of the relationship between risk (which is calculable) and uncertainty (where the extent of risk is unknown and therefore difficult to calculate), see FRANK H. KNIGHT, RISK, UNCERTAINTY AND PROFIT 19–20 (1921). “‘[R]isk’ means in some cases a quantity susceptible of measurement . . . . We shall accordingly restrict the term ‘uncertainty’ to cases of the non-quantitative type.” Id.

83In this way, a city that self-insures can make itself amenable to tort suits in states that allow municipalities to waive sovereign immunity up to the extent of insurance coverage. See infra note 119 (discussing amenability to suit).

84Self-insurance can also operate ex post by issuing debt to cover losses that actually arise. See

supra text accompanying note 38 (discussing this strategy and its limitations).

85Telephone Interview with Tom Cross, Assoc. Dir., Nashville Law Dep’t (May 20, 2014). 86See infra notes 94–96. 87See supra text accompanying note 53. 88See discussions infra Section I.B.3.

89Priest, supra note 53, at 1526; see also Jeffrey P. Griffin, Note, The Inapplicability of the NoticePrejudice Rule to Pure Claims-Made Insurance Policies, 42 CONN. L. REV. 235, 242–46 (2009) (discussing rise of claims-made policies).

90See Griffin, supra note 89 at 242–46; see also Carolyn M. Frame, ‘Claims-Made’ Liability Insurance: Closing the Gaps with Retroactive Coverage, 60 TEMP. L. Q. 165 (1987) (discussing history of claims-made policies generally); Rebecca M. Bratspies, Splitting the Baby: Apportioning Environmental Liability Among Triggered Insurance Policies, 1999 B.Y.U. L. REV. 1215, 1217 n.13.

91See Frame, supra note 90, at 178 (“One of the major advantages of the claims-made policy to the

insurer is the exclusion of tail liability.”).

92See supra text accompanying notes 10–12 (describing Half Moon Bay litigation). 93See Kenneth S. Abraham, Efficiency and Fairness in Insurance Risk Classification, 71 VA. L. REV. 403, 413 n.24 (1985) (“Claims-made pricing requires much less prediction of the future because only the claims that will be filed during the forthcoming policy period need be predicted.”). See generally Steven P. Garmisa, Claims-Made Policies: Let the Lawyer Beware, 78 ILL. B.J. 292 (1990) (describing dynamics of claims-made policies).

94See, e.g., Serkin, supra note 39, at 935. 95See id. at 939 (discussing political dynamics of entrenchment). 96See id. at 906–07. 97In the mid-nineteenth century, for example, local officials throughout the country sought to influence the location of the railroads. They believed—quite correctly—that a station in town would generate substantial economic benefits. Many spent lavishly to try to attract the railroads, but of course only some succeeded. For those, the investment was a good one. But all the others that invested in failed campaigns were saddled with increased debt but had no new railroad revenue to repay them. The consequences were devastating, leading to a massive crisis in municipal finance. See id. (describing this history and citing sources).

98See id. at 925–26 (discussing justifications for municipal debt limits). 99This will only not be true of litigation that spans more than one election cycle.

100See, e.g., PA. GOVERNOR’S CTR. FOR LOCAL GOV’T SERVS., INSURANCE PRIMER FOR MUNICIPAL SECRETARIES 4 (4th ed. 2003), http://www.newpa.com/download/insurance-primer-formunicipal-secretaries-pdf/ [https://perma.cc/BUY8-85GZ] [hereinafter INSURANCE PRIMER] (describing “typical” insurance policy packages).

101See, e.g., id. 102See, e.g., POLICY MANUAL, supra note 59, at 28–29 (describing automobile coverage). 103Comprehensive general liability policies typically cover municipalities for liability resulting from occurrences that cause either bodily injury or property damage. See, e.g., Farley, supra note 65, at 625. The breadth of these policies is often contested, however, and extensive literature has examined, for example, whether they cover civil rights claims. See id.

104POLICY MANUAL, supra note 59, at 31. 105Id. at 27. For a list of types of coverage, see id. at 23–34; see also INSURANCE PRIMER, supra

note 100, at 6.

106See INSURANCE PRIMER, supra note 100, at 6. In many municipalities, police misconduct is covered by a specialized Police Professional Liability policy. See id. (recommending all municipalities that provide police services purchase Police Professional Liability coverage).

107POLICY MANUAL, supra note 59, at 32. In the words of one insurance manual for local officials: “This type of coverage is designed to cover the liability arising out of the business decisions of public officials and employees.” Id.

108See, e.g., INSURANCE PRIMER, supra note 100, at 6 (“Claims usually arise from decisions made by elected or appointed officials that allegedly cause loss of revenue, a loss of a property right, planning and zoning issues, licensing, free speech, privacy and alleged Constitutional violations.”).

109See, e.g., Cincinnati Ins. Co. v. Motorists Mut. Ins. Co., 306 S.W.3d 69, 74 (Ky. 2010) (“[T]he concept of fortuity is inherent in all liability policies, and . . . a loss [is] fortuitous if it was not intended.” (quoting Aetna Cas. & Sur. Co. v. Commonwealth of Ky., Nat. Res. & Envtl. Prot. Cabinet, 179 S.W.3d 830, 836 (Ky. 2005) (internal quotation marks omitted .

110POLICY MANUAL, supra note 59, at 32–33 (describing E&O policies). 111See id. at 32 (“An example of an error and omission covered claim would be a decision by a governing board that results in a financial loss to a contractor who then sues the governing board members for damages.”).

112A general liability policy may be implicated, too, depending on a particular state’s interpretation of the word “occurrence.” In some states, general liability policies have been extended even to civil rights claims, reasoning that even if a municipal agent acted intentionally, she may not have intended the resulting injury. Farley, supra note 65, at 625 (concluding that many courts will find “occurrence[s]” covered by a general liability policy include intentional acts so long as the resulting injury was not specifically intended).

113See infra note 159.

114See Rynard, supra note 73, at 109–12 (describing the “art” of pricing risk and the need of local

governments that self-insure to price risk effectively).

115See Peter C. Young & John Hood, Risk and the Outsourcing of Risk Management Services: The Case of Claims Management, 23 PUB. BUDGETING & FIN. 109, 109–10 (2003) (describing outsourcing of municipal risk management for municipalities without the capacity to manage risk themselves).

116This role of insurance is well established. See, e.g., Jeffrey W. Stempel, The Insurance Policy as Social Instrument and Social Institution, 51 WM. & MARY L. REV. 1489, 1507 (2010) (“[I]t can be said without exaggeration that liability insurance has played a major, perhaps dominant, role in the development of modern tort law.”); see also Joseph E. Stiglitz, Risk, Incentives and Insurance: The Pure Theory of Moral Hazard, 26 GENEVA PAPERS ON RISK AND INS. 4, 30 (1983) (discussing relationship between insurance and social policy).

117See, e.g., GA. CODE ANN. § 33-24-51(a)–(b) (2016); 42 PA. CONS. STAT. § 8542 (2016); TENN. CODE ANN. § 29-20-201 (2016); TEX. CIV. PRAC. & REM. CODE ANN. § 101.0215(a) (West, Westlaw through 2015 Reg. Sess.); UTAH CODE ANN. § 63G-7-201 (LexisNexis 2016); Meyer v. Walls, 489 S.E.2d 880, 884 (N.C. 1997); cf. MICH. COMP. LAWS § 691.1407 (2016) (requiring gross negligence for tort liability).

118Criticism of the distinction is longstanding. See, e.g., Gerald R. Gibbons, Liability Insurance

and the Tort Immunity of State and Local Government, 1959 DUKE L. J. 588 (1959).

119See, e.g., N.C. GEN. STAT. § 153A-435(a) (West, Westlaw through 2015 Reg. Sess.); § 33-24-

51(a)–(b).

120Cf. 18 EUGENE MCQUILLIN, THE LAW OF MUNICIPAL CORPORATIONS § 53:9 (3d ed. 2013)

(surveying approaches).

121Cf. James Fleming, Tort Liability of Governmental Units and Their Officers, 22 U. CHI. L. REV. 610, 614 (1955) (justifying encroachments on governmental immunity in part because “it is better to distribute [losses caused by government actors] widely among the beneficiaries of government than to let them rest on the individual victims”).

122See, e.g., Gibbons, supra note 118, at 590 (acknowledging that principal criticism of immunity “is its inconsistency with the modern socio-ethical notion that the risk of wrongful injury should not be borne by the individual upon whom the misadventure fortuitously falls, but by society as a whole”).

123The same normative underpinnings explain why municipalities indemnify public officials, such

as police officers for § 1983 liability. See, e.g., Farley, supra note 65, at 620.

124Cf. Gibbons, supra note 118, at 590 (“A means of equitably distributing the burden of the risk [of torts] has been available to the federal and state governments and even to large cities in the form of taxation.”).

125For one treatment of this distinction, see Winter, supra note 53, at 115–16; see also Roger C. Henderson, The Tort of Bad Faith in First-Party Insurance Transactions: Refining the Standard of Culpability and Reformulating the Remedies by Statute, 26 U. MICH. J.L. REFORM 1, 16–17 (1992) (“[A] real liability policy obligates the insurer to pay the third-party tort victim once the insured’s liability has been established by settlement or court action.”).

Municipalities and other governmental entities faced . . . extreme premium increases or the unavailability of market insurance coverage altogether. Some cities closed jails and suspended police patrols until insurance coverage was obtained. Parks and forest preserves were closed. Fourth of July celebrations were cancelled because of concerns over uninsured liability.131

126Armstrong v. United States, 364 U.S. 40, 49 (1960). 127The analogy between takings law and insurance is not new. See, e.g., Steve P. Calandrillo, Eminent Domain Economics: Should “Just Compensation” Be Abolished, and Would “Takings Insurance” Work Instead?, 64 OHIO ST. L.J. 451, 491 (2003). This analogy, however, has typically ended with viewing the government as the insurer. No one has considered whether the “insurance” function of takings law requires what amounts to reinsurance through private insurance companies to function effectively.

128See, e.g., Richard N. Clark et al., Sources of the Crisis in Liability Insurance: An Economic Analysis, 5 YALE J. ON REG. 367 (1988); Priest, The Antitrust Suits, supra note 73; see also James R. Hackney, A Proposal for State Funding of Municipal Tort Liability, 98 YALE L.J. 389, 389 nn.2–3 (1988) (collecting sources).

129See Priest, supra note 53, at 1523 (considering various theories, including collusion among

insurance companies, an adverse interest rate environment, and a rapid increase in corporate liability).

130Id. at 1527. 131Id. at 1521–22 (footnotes omitted). 132See supra note 89 (discussing change from occurrence-based to claims-made policies).

133Participation in a municipal risk pool is treated like private insurance for purposes of governmental immunity and amounts to a waiver of tort immunity up to the limits of the coverage. See, e.g., Lyles v. City of Charlotte, 477 S.E.2d 150 (N.C. 1996).

134See, e.g., CRESSWELL & LANDON-MURRAY, supra note 70, at 4 (describing New York’s pool sponsor as a nonprofit company); About CIRMA: Serving Connecticut’s Municipalities, Public Schools & Local Public Agencies, CONN. INTERLOCAL RISK MGMT. AGENCY, cirma.ccm-ct.org/Plugs/aboutcirma.aspx [perma.cc/VN8J-ACH4] (identifying Connecticut Interlocal Risk Management Agency as a not-for-profit association).

135See, e.g., MICH. MUN. LEAGUE LIAB. & PROP. POOL, INTERGOVERNMENTAL CONTRACT 3 (2014), www.mml.org/insurance/pool/pdf/p_igc_2014.pdf (describing membership in the pool); THE OHIO MUN. LEAGUE, THE OHIO MUNICIPAL LEAGUE AND ITS SERVICES (2016), http://www.omlohio.org/About.htm [perma.cc/L4BA-292P] (describing membership in Ohio Municipal League pool).

[perma.cc/Q46J-YZAG]

136MARTIN FONE & PETER C. YOUNG, PUBLIC SECTOR RISK MANAGEMENT 272 (2000). 137Qiao, supra note 8, at 37. 138Id. at 38 (estimating that insurance pools covered approximately 40% of the total market of

municipal insurance by the late 1990s).

139Id. (identifying between 250–285 pools operating with professional staffs). 140Telephone Interview with Ken Canning, Dir. of Risk Mgmt. Servs., Vt. League of Cities and

Towns (June 17, 2014).

II. THE ABSENCE OF INSURANCE FOR REGULATORY TAKINGS

141Cf. id. (identifying advantages of risk pools). 142FONE & YOUNG, supra note 136 (describing history of Municipal Mutual). 143Id. 144Id. 145See id.

146U.S. CONST. amend. V. 147See, e.g., Penn Cent. Transp. Co. v. New York City, 438 U.S. 104, 124 (1978). 148For consideration of a variety of claims, see infra text accompanying notes 173–75. 149See Penn Central, 438 U.S. at 124. For a thorough discussion of each of the elements, see Robert Meltz, Takings Law Today: A Primer for the Perplexed, 34 ECOLOGY L.Q. 307, 329–30 (2007). 150See, e.g., Michelle Boardman, Insuring Understanding: The Tested Language Defense, 95 IOWA L. REV. 1075, 1091 (2010); Farley, supra note 65, at 621 (“The coverages generally do not

vary in language from policy to policy, regardless of the issuing company.”); see also Christopher C. French, Insuring Landslides: America’s Uninsured Natural Catastrophes 10 n.39 (Penn State Law, Research Paper No. 4-2015, 2015), http://ssrn.com/abstract=2566238 [https://perma.cc/64T5-7DNT] (collecting sources). But see Daniel Schwarcz, Reevaluating Standardized Insurance Policies, 78 U. CHI. L. REV. 1263 (2011) (finding surprising variance in the terms of homeowners insurance policies compared to those of other forms of insurance).

151See, e.g., Nutmeg Ins. Co. v. Clear Lake City Water Auth., 229 F. Supp. 2d 668, 684 (S.D. Tex.

2002) (looking to prior case law to define the phrase “arising out of” in insurance policy).

152The policy also specifies how much of the losses the insurer will cover. It may, for example, include a reservation—like a co-pay—where the insured must pay either a percentage or fixed amount of the loss below or above a certain amount. For example, an insurance policy might provide coverage for all losses over $250,000, or 50% of all losses below $250,000, and 80% of all additional losses, and so forth. The permutations are endless.

153Nutmeg Ins. Co., 229 F. Supp. 2d at 675 (“The duty to defend and the duty to indemnify are distinct and separate.”); City of Collinsville v. Ill. Mun. League Risk Mgmt. Ass’n, 904 N.E.2d 70, 75 (Ill. App. Ct. 2008) (“[A]n insured contracts for and has a right to expect two separate and distinct duties from an insurer: (1) the duty to defend . . . and (2) the duty to indemnify . . . .” (citations omitted .

154Nutmeg Ins. Co., 229 F. Supp. 2d at 675.

155Id. 156Id. at 676; see also Kenneth S. Abraham, A Theory of Insurance Policy Interpretation, 95 MICH. L. REV. 531, 531 (1996) (“The first principle of insurance law is captured by the maxim contra proferentem, which directs that ambiguities in a contract be interpreted ‘against the drafter,’ who is almost always the insurer.” (footnotes omitted .

157St. Paul Ins. Co. v. Tex. Dep’t of Transp. 999 S.W.2d 881, 884 (Tex. App. 1999). 158See, e.g., Del Monte Dunes at Monterey, Ltd. v. City of Monterey, 920 F.2d 1496, 1500 (9th Cir. 1990) (“[Plaintiff’s] five substantive claims are based on the Fifth Amendment’s taking clause as incorporated in the Fourteenth Amendment, the equal protection and due process clauses of the Fourteenth Amendment, and common law principles of estoppel and unjust enrichment.”).

159Allan Windt, Determining Whether Grounds for Settlement Are Outside Policy Coverage, 32 INS. LITIG. REP. 407, 407 (2010) (“The insurer should have to reimburse the insured only to the extent that the settlement compromised claims that were covered by the policy.” (footnote omitted .

160See, e.g., CNA, Public Officials Liability Policy, OLD NAT’L INS. RISK PARTNERS (Nov. 1997),

of

N.D.

FUND

RESERVE

Coverage,

https://www.oldnationalins.com/pdfs/Insurance/JWF-Sample-Policies/Public-Officials-LiabilityPolicy.pdf [perma.cc/RPX2-ZCLK] (excluding coverage for “[a]ny ‘claim’ arising out of inverse condemnation, adverse possession, dedication by adverse use, [or] eminent domain”); Liability 2015), INS. Memorandum http://www.ndirf.com/image/cache/GL1001.pdf [https://perma.cc/L88M-D4WP] (excluding coverage for “[l]iability arising out of or in any way connected with any operation of the principles of eminent domain, condemnation proceedings, or ‘inverse condemnation’, by whatever name called”); Memorandum of Coverage, CAL. JOINT POWERS RISK MGMT. AUTH. 18 (July 1, 2011), http://www.cjprma.org/docs/library/memorandum-of-coverage-2011-2012.pdf [perma.cc/NU8B-8M26] (excluding coverage for “[c]laims arising out of or in connection with land use regulation, land use planning, the principles of eminent domain, condemnation proceedings or inverse condemnation by whatever name called”); Memorandum of Coverage, N.M. PUB. SCHOOLS INS. AUTH. 4 (July 1, 2013), (excluding https://nmpsia.com/PDFs/NMPSIA%20MOC%201.PDF coverage for “liability arising out of or in connection with the principles of eminent domain, condemnation or inverse condemnation, by whatever name called”); MO. PUB. ENTITY RISK MGMT.

[https://perma.cc/RTS3-2JH7]

(Aug.

EXCLUSIONS

This Memorandum [or policy] does not apply to:

. . . .

Claims arising out of or in connection with inverse condemnation caused by the construction of a public work or public improvement, land use regulation, land use planning, the principles of eminent domain, or condemnation proceedings by whatever name called . . . .161

2010),

INS. COMPANY

FUND, MEMORANDUM OF COVERAGE (on file with author) (excluding coverage for “proceedings to condemn property or inverse condemnation by whatever name”); Nat’l Cas. Co., Public Entity Policy: Public Officials Liablity Coverage Form Claims Made Coverage, EUCLID PUB. SECTOR 3, http://euclidps.com/2/wp-content/uploads/2014/05/Public_Officials_Liability_II_PE_PO_2_0802.pdf [perma.cc/3GXJ-GM98] (excluding coverage for “any damages arising out of land use planning or municipal zoning”); Public Entity Liability Policy, AM. (Jan. 1, 2008), http://www.acegroup.com/us-en/assets/public-officials-policy-pf-23536.pdf [perma.cc/PE6X-WESW] (excluding coverage for claims arising out of “eminent domain, condemnation, inverse condemnation, temporary or permanent taking, adverse possession or dedication by adverse use”); Public Officials Liability Policy with Employment Practices Liability Coverage, IRONSHORE SPECIALTY INS. COMPANY (May http://www.ironshore.com/pdfs/products/Public_Officials_Liability_Policy.pdf [perma.cc/H3KB-LVDN] (excluding coverage for claims arising out of “inverse condemnation, temporary or permanent taking, adverse possession or dedication by adverse use”); Sample Public INS., https://www.tridentinsurance.net/docs/ Officials’ Liability Coverage Form, TRIDENT SpecimenPolicyForms/PublicEntity/PublicOfclsCovDec.pdf (excluding coverage “resulting from proceedings in eminent domain or from inverse condemnation, by whatever name called”). But cf. Catherine A. Jones et al., “One of These Things Is Not Like the Others?” – How and Why a JPA is Not Like an Insurance Company, PUB. AGENCY RISK MGMT. ASS’N 1 (Feb. 9, 2014), http://parma.com/sites/default/files/files/pdf/c1_oneofthesethingsisnotliketheother.doc [https://perma. cc/KM3P-FFPK] (identifying statutory authorization allowing municipalities to purchase inverse condemnation insurance, but not demonstrating that such insurance exists). The challenges of examining even a meaningful sample of insurance policies makes such an empirical project prohibitive, especially since most insurance policies are not publicly available. For an example of a heroic effort to study the content of insurance policies empirically, see Schwarcz, supra note 150. It is therefore no surprise that “the last systematic attempt to examine the content of different insurance policies was a 1937 law review article.” Id. at 1274–75. Informal conversations conducted with insurance underwriters and municipal risk managers further confirmed that inverse condemnation is excluded from insurance coverage. There appears to be only one exception. See LEAGUE OF MINN. CITIES, LMCIT COVERAGE GUIDE http://www.lmc.org/media/document/1/lmcitliabilitycoverageguide.pdf?inline=true [perma.cc/D6E9-RSF3] [hereinafter LMCIT COVERAGE] (describing coverage for regulatory takings). The Minnesota coverage is discussed infra text accompanying note 179.

[perma.cc/6JYH-LZHB]

(2015),

161Pooled Liability Program, Memorandum of Coverage, VECTOR CONTROL JOINT POWERS http://www.vcjpa.org/DesktopModules/Bring2mind/DMX/Download.aspx?Entry [https://perma.cc/23UF-8BLL]

AGENCY Id=10743&Command=Core_Download&PortalId=13&TabId=572 (emphasis added).

12,

162See 11A EUGENE MCQUILLIN, THE LAW OF MUNICIPAL CORPORATIONS § 32.158 (3d ed. Supp. 2016) (“[I]n inverse condemnation proceedings, a landowner commences proceedings to recover just compensation for a taking or damaging of his property when formal condemnation proceedings have not been instituted.” (footnote omitted ; Frederic Bloom & Christopher Serkin, Suing Courts, 79 U. CHI. L. REV. 553, 605 (2012) (“[I]nverse-condemnation actions amount to a kind of eminent domain proceeding in reverse: they are initiated by the property owner rather than the government, but like eminent domain actions, they ask if property has been taken and the government should thus be forced to pay.”).

163473 U.S. 172, 194–95 (1985). 164See Bloom & Serkin, supra note 162, at 605 (describing inverse condemnation). 165Id.

166See San Remo Hotel, L.P. v City of San Francisco, 545 U.S. 323, 338, 347–48 (2005) (finding that issue preclusion applies to Fifth Amendment claim following inverse condemnation action in state court).

167See Bloom & Serkin, supra note 162, at 605 (“State courts thus get first bite at these actions

under Williamson County—and they get the only bite under San Remo.”).

168See, e.g., N. River Ins. Co. v. Town of Grand Island, No. 93-CV-100S, 1995 WL 250391, at *6 (W.D.N.Y Mar. 31, 1995) (applying inverse condemnation exclusion to municipality’s prohibition on disposal of clay when property owner sued under various theories). In a previous case involving Grand Island, River Oaks Marine, Inc. v. Town of Grand Island, 89-CV-1016S, 1992 WL 406813, (W.D.N.Y. June 19, 1992), the plaintiff alleged both that the government action imposed a significant economic burden and violated the Commerce Clause, among other challenges. N. River Ins. Co., 1995 WL 250391, at *2–3. The district court in North River held that the inverse condemnation exclusion applied to all of the plaintiff’s claims previously made out by the plaintiff in River Oaks. Id. at *8. Thus, the insurance company was not liable to provide coverage for the municipality. Id.

169See, e.g., Gen. Star Indem. Co. v. V.I. Port Auth., No. 2001-188, 2007 WL 185122, at *5 (V.I. Jan. 5, 2007) (“The ‘arising out of’ language suggests that ‘a claim need bear only an incidental relationship to the described conduct for the exclusion to apply.’” (quoting Nutmeg Ins. Co. v. Clear Lake City Water Auth., 229 F. Supp. 2d 668, 697 (S.D. Tex. 2002 ; Trumpeter Devs., LLC v. Pierce County, 681 N.W.2d 269, 271 (Wis. App. 2004) (“[W]e must look at the incident giving rise to [the property owner’s] claim, not the theory of liability.” (citation omitted ; see also Christopher Serkin, Strategic Land Use Litigation: Pleading Around Municipal Insurance, 43 B.C. ENVTL. AFF. L. REV. 463 (2016).

170No. 88-6590, 1990 WL 20819, at *1 (9th Cir. Mar. 7, 1990). 171See supra notes 153–55 (discussing duty to defend). 172Transcon. Ins. Co., 1990 WL 20819, at *1. 173Courts have been generous in applying inverse condemnation exclusion. See, e.g., Gen. Star Indem. Co., 2007 WL 185122, at *5 (adopting broad reading of exclusion); Trumpeter, 681 N.W.2d at 271–72 (same).

174See, e.g., S.C. Mun. Ins. & Risk Fund v. City of Myrtle Beach, 628 S.E.2d 276, 278–79 (S.C. App. 2006) (holding that inverse condemnation exclusion applied to equal protection and due process claims arising out of a rule requiring landlords to be secondarily liable for tenants’ water bills).

175See Nutmeg Ins. Co. v. Clear Lake City Water Auth., 229 F. Supp. 2d 668 (S.D. Tex. 2002); County of Boise v. Idaho Ctys. Risk Mgmt. Program, 265 P.3d 514, 518 (Idaho 2011); Village of Waterford v. Reliance Ins. Co., 640 N.Y.S.2d 671 (N.Y. App. Div. 1996). But see City of College Station v. Star Ins. Co., 735 F.3d 332 (5th Cir. 2013); Town of Cumberland v. R.I. Interlocal Risk Mgmt. Tr., Inc., 860 A.2d 1210 (R.I. 2004). Other courts have adopted even narrower readings, but these are unusual. See City of Collinsville v. Ill. Mun. League Risk Mgmt. Ass’n, 904 N.E.2d 70, 75–77 (Ill. App. Ct. 2008) (holding that inverse condemnation exclusion does not apply to § 1983 claims for takings of property); C.V. Perry & Co. v. Village of West Jefferson, 673 N.E.2d 613, 616 (Ohio Ct. App. 1996) (holding that property owner’s complaint that denial of building permits was a taking of property did not sound in inverse condemnation). For discussion of the impact of these different approaches of plaintiffs’ pleading decisions, see Serkin, supra note 169.

176See supra note 149 (discussing Penn Central test). 177See, e.g., 42 U.S.C. § 3604 (2012) (defining prohibited conduct under Fair Housing Act); Layne v. Zoning Bd. of Adjustment, 460 A.2d 1088 (Pa. 1983) (setting forth equal protection test for discriminatory line drawing).

TAKINGS

178DOUGLAS T. KENDALL ET AL., CMTY. RIGHTS COUNSEL, TAKINGS LITIGATION HANDBOOK: DEFENDING (2000), http://communityrights.org/LegalResources/Handbook/HBintro.php [perma.cc/D2A4-T7FX]; see also Thomas W. Kelty, Inverse Condemnation: Another Assault upon Municipal Regulation, ILL. MUN. REV., Aug. 1986, at 9, 11, http://www.lib.niu.edu/1986/im860809.html [perma.cc/A39Z-HP8R] (“[A]t this time it is very difficult for a city to obtain insurance covering inverse condemnation.”).

REGULATIONS

LAND USE

CHALLENGES

179See LMCIT COVERAGE GUIDE, supra note 160, at 3. 180Id. at 24. (“Compared to conventional liability insurance, a key difference of the LMCIT coverage is that litigation relating to these types of special litigation risks is covered regardless of whether the litigation includes a claim for damages.”).

181Id. at 26. The other category of coverage the LMCIT provides that conventional insurance usually does not extend to is “[a]wards of attorney’s fees in federal civil rights or state human rights actions.” Id.

182See supra text accompanying notes 10–12. 183See Serkin, supra note 14, at 397–98 (discussing local governments’ failure to regulate and

address sea level rise because of the concern of takings liability).

184In the context of land use, the due process standard requires some general cost–benefit calculus.

See ROBERT C. ELLICKSON ET AL., LAND USE CONTROLS (4th ed. 2013) (citing cases).

185Restricting development reduces the supply of new housing. All else being equal, this will tend to increase prices and thus reduce affordability. See, e.g., Christopher Serkin & Leslie Wellington, Putting Exclusionary Zoning in Its Place: Affordable Housing and Geographical Scale, 40 FORDHAM URB. L.J. 1667, 1667 (2013).

186Specifically, it will include litigation costs discounted by the likelihood of being sued, plus any

ultimate just compensation award again discounted by the probability of losing.

188Land Use,

OF MINN. CITIES,

187See generally James E. Krier & Stewart E. Sterk, Empirical Study of Implicit Takings, 58 WM. & MARY L. REV. (forthcoming 2016), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2779858## [https://perma.cc/8WL3-AGU6]. LEAGUE

http://www.lmc.org/page/1/land-use.jsp [https://perma.cc/RN7V-N236] (“LMCIT members spend about $2.5 million per year on land use claims—about 22 percent of all liability costs. The average cost of a land use claim is $35,000. Of those costs, more than 85 percent is spent to cover defense costs. In other words, these claims generally are not about paying damages to someone but rather about paying for legal defense of the city.”). An insurance pool in North Dakota has an inverse condemnation exclusion with an important and unusual addendum. It excludes coverage for: “Liability arising out of or in any way connected with any operation of the principles of eminent domain, condemnation proceedings, or ‘inverse condemnation,’ by whatever name called. However, this exclusion does not apply to defense, costs, or supplementary payments.” Liability Memorandum of Coverage, supra note 160, at 8 (emphasis added). At least the members of this pool have recognized the importance of covering litigation costs for regulatory takings claims.

189See supra text accompanying notes 168–75. 190See infra Section III.A.

191See, e.g., James L. Huffman, Why Liberating the Public Trust Doctrine Is Bad for the Public, 45 ENVTL. L 337, 359–60 (2015); Mary Christina Wood, Nature’s Trust: Reclaiming an Environmental Discourse, 25 VA. ENVTL. L.J. 243, 257 (2007) (“The shrill call of private property rights is heard in the halls of almost every agency, every day . . . . This private property rights rhetoric has cowered officials at every level of government . . . .”).

192This concern is reminiscent of debates over the availability of insurance for punitive damages or criminal activity. See, e.g., Tom Baker, Liability Insurance at the Tort-Crime Boundary, in FAULT LINES: TORT LAW AS CULTURAL PRACTICE (David M. Engel & Michael McCann eds., 2009); Catherine M. Sharkey, Revisiting the Costs of Accidents, 64 MD. L. REV. 409 (2005). Indeed, it is reminiscent of debates about whether liability insurance is ever appropriate. See Baker & Siegelman, supra note 25, at 5 (describing history of debates). In the tort context, as well, “many scholars share in the view that tort law’s deterrence objective is ‘severely, perhaps fatally undermined’ by the prevalence of insurance.” Schwartz, supra note 65, at 313 (quoting John G. Fleming, The Role of Negligence in Modern Tort Law, 53 VA. L. REV. 815, 823 (1967 .

193See EPSTEIN, supra note 14, at 84–85. 194See, e.g., Frank I. Michelman, Property, Utility, and Fairness: Comments on the Ethical Foundations of “Just Compensation” Law, 80 HARV. L. REV. 1165, 1214–16 (1967) (arguing that compensation should only be required when demoralization costs outweigh settlement costs).

195See Penn Cent. Transp. Co. v. New York City, 438 U.S. 104, 124 (1978).

196Pa. Coal Co. v. Mahon, 260 U.S. 393, 413 (1922) (“Government hardly could go on if to some extent values incident to property could not be diminished without paying for every such change in the general law.”).

197See supra Section I.A. 198See supra text accompanying notes 34–36.

199See, e.g., NICOLE STELLE GARNETT, ORDERING THE CITY (2010) (focusing on land use and policing in urban revitalization); EDWARD GLAESER, TRIUMPH OF THE CITY (2011) (describing economic reasons for revitalization of cities); WITOLD RYBCZYNSKI, MAKESHIFT METROPOLIS (2010) (describing role of land use in urban revitalization); see also Morgan Brennan, Downtowns: What’s Behind America’s Most Surprising Real Estate Boom, FORBES (Mar. 25, 2013, 12:02 PM), http://www.forbes.com/sites/morganbrennan/2013/03/25/emerging-downtowns-u-s-cities-revitalizingbusiness-districts-to-lure-young-professionals/ [perma.cc/6NUH-PD2Z] (discussing population growth among twenty-five to thirty-four year-old college-educated professionals into cities as “one of the reasons city planners have been plowing money and resources into revitalizing their core business districts”).

200See Paul L. Knox & Fang Wei, Neighborhood Change in Metropolitan America, 1990 to 2010,

50 URB. AFF. REV. 459, 462 (2013).

201See, e.g., STACEY A. SUTTON, URBAN REVITALIZATION IN THE UNITED STATES: POLICIES AND PRACTICES (2008), http://www.columbia.edu/cu/c2arl/pdf_files/USURRP_Phase_I_Final_Report.pdf [perma.cc/3D5V-AX7R] (discussing the various stakeholders interested in and active in urban revitalization).

202Cf. Oates, supra note 43, at 958–59 (discussing how the output of public services in the suburbs, such as quality of the local public schools, affects choice of residence); Tiebout, supra note 43, at 418 (discussing that local expenditures are variables that influence an individual’s choice to move to a certain municipality).

203See, e.g., Peter Mieszkowski & Edwin S. Mills, The Causes of Metropolitan Suburbanization,

7 J. ECON. PERSP. 135, 136 (1993).

204See generally Kenneth A. Stahl, The Suburb as a Legal Concept: The Problem of Organization

and the Fate of Municipalities in American Law, 29 CARDOZO L. REV. 1193, 1210 (2008).

205See Andrew F. Haughwout, The Paradox of Infrastructure Investment: Can a Productive Good

Reduce Productivity?, BROOKINGS REV., Summer 2000, at 40, 42.

206See James A. Kushner, Urban Neighborhood Regeneration and the Phases of Community Evolution After World War II in the United States, 41 IND. L. REV. 575, 581–82 (2008). But see FISCHEL, supra note 48, at 152–54 (challenging death spiral narrative).

207Kushner, supra note 206 at 583–84. 208See Village of Euclid v. Ambler Realty Co., 272 U.S. 365 (1926) (upholding zoning against facial due process challenge on grounds that separating incompatible uses is an appropriate use of the police power).

209Cf. Christopher Serkin & Gregg P. Macey, Post-Zoning: Alternative Forms of Public Land Use Controls, 78 BROOK. L. REV. 305 (2013) (describing original goal of zoning as separating residential districts from industrial and commercial ones).

210See generally Nestor M. Davidson, Leaps and Bounds, 108 MICH. L. REV. 957, 962 (2010); Nicole Stelle Garnett, Redeeming Transect Zoning?, 78 BROOK. L. REV. 571, 572–73 (2013) (describing adoption of new form-based codes).

211See, e.g., Vicki Been et al., Urban Land Use Regulation: Are Homevoters Overtaking the Growth Machine?, 11 J. EMPIRICAL LEGAL STUD. 227, 228–29 (2014) (characterizing traditional views of local governments).

212See Schragger, supra note 43, at 508–12 (describing the increasing “conditions on mobile capital” that cities use in their land use processes); see also Christopher Serkin, Penn Central Take Two, 92 NOTRE DAME L. REV. (forthcoming 2016), http://papers.ssrn.com/sol3/papers.cfm?abstract_id= 2728417 [https://perma.cc/Q7RV-KMD8] (describing recent decision by New York City to rezone property in exchange for public benefits, but thereby adversely affecting the value of private transferable development rights).

213See Serkin, supra note 212 (describing zoning ordinance in New York City requiring developer

to pay for certain infrastructure improvements in order to receive density bonus).

214See Serkin & Wellington, supra note 185, at 1671 (“Large-lot zoning was ubiquitous across suburbs that developed around major cities like New York and Saint Louis. In fact, in the late 1960s, Missouri had a four-year supply of one-third acre lots, but a stunning 350-year supply of one-acre lots.” (footnotes omitted .

215Id. at 1685–86 (contrasting new urban efforts to exclude the poor with more conventional

suburban ones).

216See David A. Lewis, Identifying and Avoiding Conflicts Between Historic Preservation and the Development of Renewable Energy, 22 N.Y.U. ENVTL. L.J. 274, 289 (2015) (“Preserved and restored historic resources can be a competitive advantage for cities and towns and rural areas alike. Examples abound of towns and cities using historic preservation as a strategy in developing local economic activity.” (footnotes omitted .

217William A. Fischel, Neighborhood Conservation Districts: The New Belt and Suspenders of

Municipal Zoning, 78 BROOK. L. REV. 339, 340 (2013) (describing the phenomenon).

218See FURMAN CTR. FOR REAL ESTATE & URBAN POLICY, HOW HAVE RECENT REZONINGS AFFECTED THE CITY’S ABILITY TO GROW? 7–10 (2010), http://furmancenter.org/files/publications/ Rezonings_Furman_Center_Policy_Brief_March_2010.pdf [perma.cc/QHB2-HAWH].

219See, e.g., Andy Newman, Clearing of Atlantic Yards’ Site Proceeds as Legal Thicket Grows Denser, N.Y. TIMES (Apr. 5, 2007), http://www.nytimes.com/2007/04/05/nyregion/05yards.html?_r=0 [https://perma.cc/V32W-PYB8] (describing opposition to Atlantic Yards development in Brooklyn).

220See Serkin & Macey, supra note 209, at 315–16 (describing expansion of permissible goals of

zoning and means of accomplishing them).

221There are, of course, exceptions. Some smaller local governments with particular political identities have been equally, if not, more active. Boulder, Colorado, comes immediately to mind. Moreover, there are other more immediate reasons why cities may have been more aggressive in adopting regulatory innovations. Because they have been faced with greater development pressures, and also face greater congestion of public resources, it is no surprise that cities perceive an acute interest in land use regulations. The claim here is simply that risk aversion supports these natural trends.

III. ADDRESSING THE INVERSE CONDEMNATION EXCLUSION

222Cf. Edward L. Glaeser et al., Why Is Manhattan So Expensive?: Regulation and the Rise in Housing Prices, 48 J.L. & ECON. 331, 359 (2005) (calculating zoning “tax” for a sample of municipalities across the country, and reporting in general that the largest cities have the highest “tax”). 223It is possible, however, that insurers would also find ways to constrain the regulatory innovations of insured municipalities. Cf. Rappaport, supra note 9, at 63 (discussing ways in which insurance companies limit the risks created by police departments in order to reduce the companies’ exposure to legal liability). If that is true, then insurance will not necessarily invite greater land use experimentation.

224Today, some firms—like Glatfelter and Alliant—specialize in public insurance. See, e.g., Public Entity Insurance Solutions, ALLIANT, http://www.alliant.com/Industry-Solutions/Public-Entity/Pages/ [http://perma.cc/8GRC-8BFH]; GLATFELTER, http://www.glatfelterpublicpractice.com default.aspx [https://perma.cc/V94K-VKZC]. However, insurance companies have, in the past, failed to understand the nature of municipal risk. See, e.g., Jaron, supra note 63, at 20 (“[I]nsurance companies do not understand state and local government. Few, if any, insurance firms treat government as a specialty; instead, they have been add-ons to other lines designed for the private sector.” (first citing NAT’L LEAGUE OF CITIES, THE NEW WORLD OF MUNICIPAL LIABILITY 4–5 (1978); and then citing OKLA. MUN. LEAGUE, MUCH ADO ABOUT MUNICIPAL LIABILITY 3–6 (1977 .

225See supra note 161 (quoting standard policy language). 226Schwartz, supra note 65, at 341 n.125. This is the justification for liability policies’ exclusions for intentional harms. See id. at 341 (“Insurance policies typically contain an exclusion for ‘intentional’ harms, and the standard explanation for this exclusion is that intentional harms are under the control of the insured in a way that makes insurance inadvisable.” (citing ROBERT E. KEETON, INSURANCE LAW: BASIC TEXT 286–87, 292–93, 297–98 (1971 .

227This statement is too strong in some jurisdictions. New York City, for example, allows a property owner to compel the government to complete a condemnation that it has initiated. This can create genuine risk because the government might file a petition for condemnation based on a faulty assessment of value. There is a chance that the condemnation will end up costing the government much more than it anticipated, and the government may not be allowed to reverse course and give up on the

condemnation. This is unusual, however, and the risk facing the government in eminent domain proceedings is therefore quite slight.

228This dynamic generally will not preclude insurance coverage. See, e.g., Calvert Ins. Co. v. W. Ins. Co., 874 F.2d 396, 399 (7th Cir. 1989) (“Even an intentional act will be covered under the policy language at issue here if it causes an unexpected and unintended result.”).

229See supra note 149 (discussing Penn Central test). 230See, e.g., Schwartz, supra note 65, at 345 (arguing that tort defendants purchase insurance

because they cannot necessarily avoid harm simply by taking care). As Schwartz argues:

The combination . . . of the costs of defending against unmeritorious suits, the costs of being subjected to liability when the jury errs, and the chance of bearing liability when there is genuine uncertainty in the negligence standard provides an explanation for why many defendants . . . purchase negligence liability insurance [instead of simply taking care to avoid negligence].

Id. at 346.

231See, e.g., Bd. of Cty. Comm’rs v. Snyder, 627 So. 2d 469, 471 (Fla. 1993). 232Thank you to Lee Fennell for this formulation. 233See First English Evangelical Lutheran Church of Glendale v. County of Los Angeles, 482 U.S.

304 (1987) (requiring compensation as remedy for takings violation).

234Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 701 (1978) (holding that municipalities can be liable for § 1983 violations). This has been defended on normative and doctrinal grounds. See Melvyn R. Durchslag, Should Political Subdivisions Be Accorded Eleventh Amendment Immunity?, 43 DEPAUL L. REV. 577 (1994).

235See First English, 482 U.S. at 316 n.9. For a view that sovereign immunity might extend to municipal governments, but that the Takings Clause nevertheless abrogates state sovereign immunity, see Eric Berger, The Collision of the Takings and State Sovereign Immunity Doctrines, 63 WASH & LEE L. REV. 493, 501 n.18 (2006). But see generally Richard H. Seamon, The Asymmetry of State Sovereign Immunity, 76 WASH. L. REV. 1067 (2001) (arguing that First English did not hold that the Takings Clause abrogates state sovereign immunity, and that it does so only when states are sued in state court). 236See First English, 482 U.S. at 321; see also Resource Investments, Inc. v. United States, 85 Fed.

Cl. 447, 480–85 (2009) (distinguishing between prospective and retrospective temporary takings).

237See Daniel L. Siegel & Robert Meltz, Temporary Takings: Settled Principles and Unresolved Questions, 11 VT. J. ENVTL. L. 479, 480 (2010); J. Margaret Tretbar, Comment, Calculating Compensation for Temporary Regulatory Takings, 42 U. KAN. L. REV. 201, 219 (1993).

238The Half Moon Bay litigation described in the introduction is one such case. See supra text

accompanying notes 10–12 (discussing the litigation).

239See, e.g., Durham City Bd. of Educ. v. Nat’l Union Fire Ins. Co. of Pittsburgh, 426 S.E.2d 451 (N.C. Ct. App.), rev. denied, 431 S.E.2d 22 (N.C. 1993) (requiring insurer to indemnify school board for negligent supervision arising out of alleged rape by basketball coach); Bd. of Pub. Educ. of Sch. Dist. of Pittsburgh v. Nat’l Union Fire Ins. Co. of Pittsburgh, 709 A.2d 910, 917 (Pa. Super. Ct. 1998) (finding insurer had duty to defend school district and related defendants in claim involving sexual misconduct by employee).

240For a discussion of the role of insurance in police misconduct cases, see Rappaport, supra note

9.

241See, e.g., Emery & Maazel, supra note 65, at 590 (describing lack of deterrent effect of civil judgments against police because of indemnification); see also Schwartz, supra note 63, at 912 (discussing practice of police indemnification).

242See Rappaport, supra note 9. 243Cf. BAKER & GRIFFITH, supra note 42, at 59 (“[T]he lower the level of which the loss-creating behavior occurs in a corporation, the less reactive that risk is to the presence . . . of insurance . . . . Lower level employees benefit less directly from the insurance, and the corporation has more ways to discipline them.”).

244See, e.g., Penn Cent. Transp. Co. v. New York City, 438 U.S. 104 (1978) (evaluating application of historic preservation law for takings violation); New Port Largo, Inc. v. Monroe County, 873 F. Supp. 633, 635 (S.D. Fla. 1994), aff’d, 95 F.3d 1084 (11th Cir. 1996) (takings claim arising out of rezoning property); Alachua Land Inv’rs, LLC v. City of Gainesville, 107 So. 3d 1154, 1156 (Fla. Dist. Ct. App. 2013) (takings claim arising out of denial of plat approval); City of Sherman v. Wayne, 266 S.W.3d 34, 40 (Tex. App. 2008) (takings claim arising out of enforcement of zoning requirement).

245See, e.g., Indep. Sch. Dist. No. 697 v. St. Paul Fire & Marine Ins. Co., 515 N.W.2d 576, 579 (Minn. 1994) (requiring insurer to indemnify school board for employment discrimination, and rejecting claim that E&O policy should cover only negligence); Bd. of Educ. of Twp. of Union v. N.J. Sch. Bds. Ass’n Ins. Grp., 719 A.2d 645, 649 (N.J. Super. Ct. App. Div. 1998) (holding that E&O policy “patently” covered claims for denial of educational services under Individuals with Disabilities Education Act).

246Sch. Dist. No. 1 v. Mission Ins. Co., 650 P.2d 929 (Or. Ct. App. 1982). 247Id. at 936. 248Id.

249Cf. Lingle v. Chevron U.S.A. Inc., 544 U.S. 528, 529 (2005) (holding that takings claims are concerned exclusively with the impact of the regulation on the plaintiff’s property rights and not with governmental purpose).

250BAKER & GRIFFITH, supra note 42, at 45. 251Id. at 60–61 (discussing moral hazard in D&O policies). 252See, e.g., Baker & Siegelman, supra note 25, at 169–70 (“[I]nsurance contracts contain numerous structural features designed to limit moral hazard. These features seem to work reasonably well.”).

253See Rappaport, supra note 9, at 36–50 (surveying loss prevention tools). 254Land Use, supra note 188 (“LMCIT’s team of land use attorneys works with members to provide customized information and training, and acts as a resource to elected and appointed city officials and to city attorneys. Land use loss control staff addresses topics including the different roles elected officials must play in making land use decisions . . . .”).

255See, e.g., Training and Services, LEAGUE OF MINN. CITIES, http://www.lmc.org/page/1/ LandUseServices.jsp [https://perma.cc/5YZB-JL3C] (describing range of land use resources available to “elected officials, city managers, attorneys, and zoning administrators”).

256See BAKER & GRIFFITH, supra note 42, at 61 (“As a result of . . . deductibles and limits,

insurance protection is incomplete, maintaining at least some incentive to prevent loss.”).

257See id.; see also Baker & Siegelman, supra note 25, at 169–70 (describing these responses to

moral hazard).

258Baker & Siegelman, supra note 25, at 178 (listing different ways of addressing moral hazard,

including underwriting, experience rating, coverage design, and loss control, among others).

259These are closely related concepts. See French, supra note 150, at 11 (Adverse selection may occur when “policyholders have an informational advantage over insurers, which enables the policyholders to use that informational advantage to allow people who know they are bad risks to buy more insurance than people who are good risks.” (footnote omitted .

260See, e.g., Siegelman, supra note 22, at 1223 (“The phrase ‘adverse selection’ was originally coined by insurers to describe the process by which insureds [sic] utilize private knowledge of their own riskiness when deciding to buy or forgo insurance.” (footnote omitted .

261See e.g., id. at 1224 (“As the good risks begin to exit, the average quality of those insureds [sic] remaining falls and prices rise in a vicious circle, ending in a so-called ‘death spiral’ where no one is covered.” (footnote omitted .

262At least one commentator has pointed to adverse selection as one reason why insurance markets hardened in the 1980s, exacerbating fluctuations in insurance markets. See Winter, supra note 53, at 130.

263The two concepts are distinct but very similar. Risk rating differentiates people based on ex ante assessments of risk categories. “Experience rating is the insurance term for charging different prices based on past experience. It is a form of risk classification because past experience is used to predict future risk.” TOM BAKER, CONTAINING THE PROMISE OF INSURANCE: ADVERSE SELECTION AND RISK CLASSIFICATION, in RISK & MORALITY 271 (Richard V. Ericson & Aaron Doyle eds., 2003).

264See Baker & Siegelman, supra note 25, at 179 (“Experience-rating has the potential to reduce

both ex ante and ex post moral hazard.”).

265Information is key to the adverse selection problem. See Siegelman, supra note 22, at 1223. 266See, e.g., Liran Einav & Amy Finkelstein, Selection in Insurance Markets: Theory and Empirics in Pictures, 25 J. ECON. PERSP. 115, 118 (2011) (“When the individual-specific loss probability (or expected cost) is private information to the individual, firms must offer a single price for pools of observationally identical but in fact heterogeneous individuals.”).

267Cf. Siegelman, supra note 22, at 1225 (“[W]hile adverse selection in insurance markets is clearly a possibility, it is often not the serious problem that it is taken to be.”). For an argument that bundling different coverage together minimizes adverse selection, see French, supra note 150, at 11 (“The risk of adverse selection diminishes . . . if all of the most common types of risk of loss are bundled together in the same policy.”).

268See LMCIT COVERAGE GUIDE, supra note 160, at 26.

269Id. at 5, 27 (setting out aggregate limits and co-pays). 270Telephone Interview with Jed Burkett, Loss Control, Minn. League of Cities (Nov. 19, 2015). 271Id. 272In one recent study, the broader category of “[p]ublic official liability” was near the bottom in

terms of both frequency and cost of claims. See Qiao, supra note 8, at 43.

273See Susan K. Laury et al., Insurance Decisions for Low-Probability Losses, 39 J. RISK & UNCERTAINTY 17, 18 (2009) (“[I]ndividuals are more likely to purchase insurance for the higherconsequence, lower-probability events.”); supra Section I.A (discussing loss aversion and the role of insurance). But see, e.g., Mark J. Browne et al., Behavioral Bias and the Demand for Bicycle and Flood Insurance, 50 J. RISK & UNCERTAINTY 141 (2015) (finding that consumers prefer to buy insurance for high probability but low salience events). Thank you to John Rappaport for pointing me to these sources.

274See Krier & Sterk, supra note 187 (analyzing takings claims over five-year period); see also supra note 188 (describing that land use litigation in Minnesota accounts for fully 22% of all municipal liability).

275See Wood, supra note 191 (describing developers’ threats of takings litigation).

276Ann E. Carlson & Daniel Pollak, Takings on the Ground: How the Supreme Court’s Takings Jurisprudence Affects Local Land Use Decisions, 35 U.C. DAVIS L. REV. 103, 116 (2001) (“The responses showed that California planners have a high awareness of the cases, particularly of Nollan and Dolan, the two cases that most directly affect their every day practice.”).

277Cf. Michelle Boardman, Insuring Understanding: The Tested Language Defense, 95 IOWA L. REV. 1072, 1094 (2010) (“If the consumer supposes that all insurers are pretty much the same and are all selling the same basic product, the question dwindles to ‘How much?’”).

278That does not explain why public insurance pools outside Minnesota have not stepped in to fill the gap. If regulatory takings claims really were insurable, and if there really were demand, then at least such insurance should be available through municipal insurance pools. But the inverse condemnation exclusion is ubiquitous there, too.

279See supra text accompanying note 192.

280For a similar suggestion in the context of corporate insurance, see BAKER & GRIFFITH, supra

note 42, at 222–23.

281This two-tiered private-then-public solution is not unique. See, e.g., French, supra note 150, at Part III (proposing both private mechanisms and then state insurance to address existing hole in insurance coverage).

282See infra text accompanying notes 283–89. 283ELLICKSON ET AL., supra note 184, at 649. See generally JOHN MARTINEZ, LOCAL GOVERNMENT LAW § 3:6 (2d ed. 2015); Christopher K. Odinet, Fairness, Equity, and a Level Playing Field: Development Goals for the Resilient City, 50 IDAHO L. REV. 217, 229 (2014).

284See, e.g., Jacques LeBoeuf, The Economics of Federalism and the Proper Scope of the Federal Commerce Power, 31 SAN DIEGO L. REV. 555, 568–69 (1994); cf. Serkin, supra note 19, at 1693 (describing desire to use state grants to create positive externalities).

285See, e.g., Clayton P. Gillette, Local Redistribution, Living Wage Ordinances, and Judicial Intervention, 101 NW. U. L. REV. 1057 (2007) (describing and criticizing conventional account, and suggesting that local governments can engage in redistribution).

286Serkin & Wellington, supra note 185, at 1668. 287See, e.g., Equal Educational Opportunity Act, 16 VT. STAT. ANN. tit. 16, § 4010 (2016) see also VT. AGENCY OF EDUC., ACT 60 OF 1997, (providing http://education.vermont.gov/laws/1997/act-60/fact-sheet [https://perma.cc/3T26-V7RL] (last updated Dec. 4, 2014) (describing Vermont education financing system).

redistribution);

for

288See Serkin, supra note 19, at 1646–52 (discussing FISCHEL, supra note 48, at 39–57). 289See, e.g., Dennis E. Gale, Eight State-Sponsored Growth Management Programs: A Comparative Analysis, 58 J. AM. PLANNING ASSOC. 425, 432 (1992) (describing different forms of state subsidies for local planning initiatives).

290See supra notes 189–97. 291Adding some stylized numbers to the example above demonstrates the point. A local government might value the benefits of the regulations at $3 million and costs of $3 million (perhaps in lost tax revenue or in the private harms to burdened property owners). However, the local government also anticipates a 10% chance of $10 million in regulatory takings liability and litigation costs. In this

scenario, the expected value of the regulation for the local government is a loss of $1 million. Now imagine that there are positive externalities that the state values at $2 million. How much will the state have to subsidize the regulatory action in order to induce the local government to act? If the local government behaved like a risk neutral rational economic actor, any subsidy over $1 million should induce the local government to act. But that changes with risk aversion. The impact of the 10% chance of $10 million in takings liability suddenly weighs more heavily than $1 million. Now the state will have to pay more, and potentially much more, in order to induce the local government to act.

292For a similar suggestion in a very different context, see Comment, Riot Insurance, 77 YALE L. J.

541, 555 (1968) (proposing that the federal government reinsure states for the costs of urban riots).

293Recall that public entities’ risk aversion varies inversely with their size, so states are necessarily

less risk averse than their municipalities. See supra Section I.A.

294In the context of private liability insurance, conventional wisdom is just the opposite: the presence of insurance may lead to more claims, including fraudulent ones. See Baker & Siegelman, supra note 25, at 181. The difference here is that the municipalities are not judgment proof, and so state insurance does not create access to an otherwise unavailable pool of money. See Kent D. Syverud, On the Demand for Liability Insurance, 72 TEX. L. REV. 1629 (1994) (suggesting that liability insurance might create demand by providing access to money that would not otherwise be available).

295See, e.g., Chad G. Marzen, Crop Insurance Bad Faith: Protection for America’s Farmers,

46 CREIGHTON L. REV. 619, 625–29 (2013) (describing federal crop insurance).

296A traditional response to moral hazard is to control the conduct of the insured. See BAKER &

GRIFFITH, supra note 42, at 66–67.

CONCLUSION