The Role of State Attorneys General In Improving Prescription Drug Affordability

Impact litigation initiated by state attorneys general has played an important role in advancing public health goals in contexts as diverse as tobacco control, opioids, and healthcare antitrust. State attorneys general also play a critical role in helping governors and legislatures advance health policies by giving input into their drafting and defending them against legal challenges. State attorneys general have entered the prescription drug affordability arena in both these ways—for example, by initiating lawsuits relating to price fixing by generic drug manufacturers and defending state laws requiring disclosures of pharmaceutical prices. Yet the scope of their collective efforts is not well understood, and little is known about factors that facilitate and hinder them in their pursuit of policy objectives relating to drug affordability. In this Article, we report findings from an empirical study of state attorney general activities relating to pharmaceutical pricing. Drawing from key informant interviews with attorneys working on drug pricing issues as well as a scoping review, we report on how state attorneys general are working to address the problem of drug affordability, how they make decisions about resource investments in this area, what positions state attorneys general to be effective change agents in this space, and what challenges they confront in this work. We situate our results within the broader literature on state attorneys general as policy actors, and we suggest measures that could extend their capacity to successfully tackle the complex issues that give rise to unaffordable drugs.


Prescription drug affordability has risen to the very top of the policy agenda for lawmakers and the public. Even during the peak of the COVID-19 pandemic in January 2021, a national poll ranked taking action to lower prescription drug prices as Americans’ second-highest domestic policy priority, with eighty-seven percent of participants rating the issue “extremely important.”[1] Nearly one in four Americans reports difficulty affording their prescription medications, and patients may respond by delaying filling their prescriptions, cutting pills in half, or skipping doses.[2] Despite partisan divisions among voters in many areas of health policy, a bipartisan supermajority of Americans—nearly four in five—believes that prescription drug costs are “unreasonable.”[3]

Notwithstanding both this bipartisan support and pledges by leaders of both parties to take action on the issue following the 2016 election,[4] few substantive reforms have become law at the federal level in recent years, either through legislation or regulation.[5] Rather, states have become the locus of policy action on prescription drug costs. States’ interest in this area arises not only because high drug prices harm consumers, but also because state budgets are directly affected when drug costs for the Medicaid program or state employee health plans rise. Unlike the federal government, states face special financial pressures due to requirements that they balance their budgets.[6] Since 2017, 166 prescription drug pricing bills have become law in forty-eight states, tackling everything from price transparency to regulation of pharmacy benefit managers (PBM) to drug affordability boards.[7] State legislatures are even taking inspiration from bold proposals being introduced at the federal level by considering bills that would adopt international reference pricing approaches within their own states.[8] Despite this energetic activity, state legislative interventions adopted to date have had only modest impacts on drug pricing.[9]

There has been little scholarly attention paid to the drug affordability activities of another state-level policy actor with an important role to play: offices of state attorneys general (AGs).[10] State AGs are important players both in using affirmative tools, including litigation, to advance important policy goals and in defending states against the inevitable legal challenges to state legislative efforts on prescription drug affordability.[11] Past examples of state AG involvement in critical public health issues, including tobacco control and the opioid epidemic,[12] suggest that state AG litigation can be a powerful force in combating business practices that cause health harms. In many states, AGs also have a critical role to play in providing advice to legislatures trying to make progress on drug pricing amidst ever-present threats of lawsuits challenging their enactments. Building on their past experiences, as well as a successful history of addressing high drug prices using healthcare fraud-and-abuse statutes, state AG offices have recently expanded their activities in the drug affordability space to include a variety of interesting, creative approaches leveraging state and federal antitrust laws, state consumer protection laws, emergency price-gouging statutes, and other authorities.

In this Article, we report findings from an empirical study of state AG activities relating to pharmaceutical pricing. Few empirical analyses of state AGs’ activities have been conducted,[13] and to our knowledge, this is the first to focus on drug pricing. Through key informant interviews with attorneys working on drug pricing issues within state AG offices as well as a scoping review, we sought to learn more about how state AGs are addressing the problem of prescription drug affordability, how they make decisions about resource investments in this area, what positions state AGs to be effective change agents in this space, and what challenges they confront in this work. We situate our results within the broader literature on state AGs as policy actors, and we suggest measures that could extend their capacity to successfully tackle the complex issues that give rise to overpriced drugs.

Part I of this Article surveys the existing literature on the role of state AGs, considering key elements of their institutional role and exploring how those elements were instantiated in specific activities in the health context. Part II presents the methods we used in conducting this project. Part III presents the results of our analysis, including the major activities of state AGs in the prescription drug affordability space and key findings relating to how they carry out this work and the factors that empower and constrain them. Part IV reflects on these results, and Part V offers recommendations for strengthening state AGs’ role in this important area.

          [1].      Politico & Harv. T.H. Chan Sch. of Pub. Health, The American Public’s Priorities for the New President and Congress, Politico & Harv. T.H. Chan Sch. of Pub. Health 2 (2021).

          [2].      Ashley Kirzinger, Lunna Lopes, Bryan Wu, & Mollyann Brodie, KFF Health Tracking Poll—February 2019: Prescription Drugs, Kaiser Fam. Found. (Mar. 1, 2019), [].

          [3].      Id.

          [4].      Stacie B. Dusetzina & Michelle M. Mello, Drug Pricing Reform in 2021—Going Big or Going Bipartisan?, JAMA Health F. (July 8, 2021),
fullarticle/2781947 [].

          [5].      Rachel E. Sachs, The Rhetorical Transformations and Policy Failures of Prescription Drug Pricing Reform Under the Trump Administration, 46 J. Health Pol., Pol’y & L. 1053 (2021). Some small reforms that would attempt to eliminate anticompetitive abuses, such as the Creating and Restoring Equal Access to Equivalent Samples (CREATES) Act, did become law. See Further Consolidated Appropriations Act, 2020 § 610, 21 U.S.C. § 355-2; U.S. Food & Drug Admin., Access to Product Samples: The CREATES Act (2020),‌drugs/‌guidance-compliance-regulatory-infor
mation/access-product-samples-creates-act [].

          [6].      See, e.g., David Gamage, Preventing State Budget Crises: Managing the Fiscal Volatility Problem, 98 Calif. L. Rev. 749, 755 (2010).

          [7].      Trish Riley, Celebrating Five Years of State Action to Lower Drug Prices, Nat’l Acad. for St. Health Pol’y (May 18, 2021),
r-drug-prices [].

          [8].      Lev Facher, States Still Can’t Import Drugs from Canada. Now, Many are Seeking to Import Canadian Prices, STAT (Feb. 18, 2021),
es [].

          [9].      Michelle M. Mello & Trish Riley, To Address Drug Affordability, Grab the Low-Hanging Fruit, JAMA Health F. (Feb. 25, 2021),
icle/2777036 [] (summarizing data on recent price increases). For a discussion of legal limitations on what states can regulate in the drug affordability space, see infra note 237 and accompanying text.

        [10].      Hereinafter, we use the term “AGs” to refer both to the state’s top attorney and to the attorneys who staff the AG’s office.

        [11].      See, e.g., Ass’n for Accessible Meds. v. Frosh, 887 F.3d 664, 666 (4th Cir. 2018) (invalidating a Maryland statute regulating excessive price increases for essential off-patent and generic drugs).

        [12].      See infra Part I for discussions of these specific examples.

        [13].      See infra Part I for discussions of these specific examples. Danielle Keats Citron, The Privacy Policymaking of State Attorneys General, 92 Notre Dame L. Rev. 747, 751–52 (2016) (analyzing interview data and documentary evidence obtained from Freedom of Information Act requests to study how AGs are working on data privacy issues). Most empirical work has analyzed predictors of AGs’ participation in multistate litigation. Colin Provost, An Integrated Model of U.S. State Attorney General Behavior in Multi-State Litigation, 10 St. Pol. & Pol’y Q. 1, 2–3 (2010) [hereinafter Provost, Integrated Model]; Colin Provost, The Politics of Consumer Protection: Explaining State Attorney General Participation in Multi-State Lawsuits, 59 Pol. Res. Q. 609, 612 (2006) [hereinafter Provost, Politics of Consumer Protection]; Colin Provost, State Attorneys General, Entrepreneurship, and Consumer Protection in the New Federalism, 33 Publius 37, 43–44 (2003) [hereinafter Provost, Entrepreneurship]; Joseph F. Zimmerman, Interstate Cooperation: The Roles of the State Attorneys General, 28 Publius 71, 72–73 (1998); Thomas A. Schmeling, Stag Hunting with the State AG: Anti-Tobacco Litigation and the Emergence of Cooperation Among State Attorneys General, 25 Law & Pol’y 429, 430–31 (2003). A few studies have modeled predictors of states’ decisions to join amicus briefs. See Shane A. Gleason, The Dynamics of Legal Networks: State Attorney General Amicus Brief Coalition Formation, 39 Just. Sys. J. 253, 254 (2018); Margaret H. Lemos & Kevin M. Quinn, Litigating State Interests: Attorneys General as Amici, 90 N.Y.U. L. Rev. 1229, 1242–43 (2015); Cornell W. Clayton, Law, Politics and the New Federalism: State Attorneys General as National Policymakers, 56 Rev. Pol. 525, 544–48 (1994); Thomas R. Morris, States Before the U.S. Supreme Court: State Attorneys General as Amicus Curiae, 70 Judicature 298, 301–02 (1987). Other studies have analyzed characteristics of cases in which states have enforced federal consumer protection statutes, see, e.g., Amy Widman & Prentiss Cox, State Attorneys General’s Use of Concurrent Public Enforcement Authority in Federal Consumer Protection Laws, 33 Cardozo L. Rev. 53, 54–55 (2011); state AGs’ decisions to bring nonpharmaceutical healthcare antitrust suits in the 1970s and early 1980s, see, e.g., Ronald C. Lippincott, Redressing the Imbalanced Political Market for Health Policy: A Role for the State Attorney General?, 9 J. Health Pol., Pol’y & L. 389, 391 (1984); strategies employed in consumer-protection lawsuits, see, e.g., Prentiss Cox, Amy Widman & Mark Totten, Strategies of Public UDAP Enforcement, 55 Harv. J. on Legis. 37, 79–85 (2018); and changes in the types of lawsuits AGs have brought over time, see e.g., Paul Nolette & Colin Provost, Change and Continuity in the Role of State Attorneys General in the Obama and Trump Administrations, 48 Publius 469, 471–73 (2018).

* Mello is a Professor of Law at Stanford Law School; Professor in the Department of Health Policy at Stanford University School of Medicine; and a faculty affiliate of the Freeman Spogli Institute for International Studies, Stanford University; Ph.D., University of North Carolina at Chapel Hill; J.D., Yale Law School; M.Phil., University of Oxford; A.B., Stanford University. Grant funding from the Arnold Ventures is gratefully acknowledged. The funder had no role in the design of the study or the drafting or revision of the manuscript. Mello discloses expert witness fees received from law firms representing generic drug company plaintiffs in a pharmaceutical antitrust lawsuit; Mello and Sachs disclose receiving consulting fees from the National Academy for State Health Policy for authoring reports on policy strategies to regulate prescription drug prices; Mello reports serving as a consultant to CVS Caremark and, after the period in which work reported in this Article was conducted, as an expert witness on behalf of a health insurer suing a pharmaceutical company over alleged off-label promotion (no fees yet received for either engagement); and Sachs reports receiving honoraria for her participation on the Institute for Clinical and Economic Review’s independent Midwest Comparative Effectiveness Public Advisory Council and West Health’s Council for Informed Drug Spending Analysis. The authors extend their thanks to the attorneys who participated in this study and to Katie Siler, Taryn Marks, Heather Joy, Camila Strassle, Quinn Walker, and Corissa Paris for research assistance. For their thoughtful comments and suggestions in developing this Article, the authors would like to thank Margaret Lemos and the scholars who participated in workshops at Washington University in St. Louis School of Law and Johns Hopkins Bloomberg School of Public Health.

           †      Riley is a former Executive Director of the National Academy for State Health Policy; M.S., University of Maine.

           ‡      Sachs is the Treiman Professor of Law, Washington University in St. Louis School of Law; J.D., Harvard Law School; M.P.H., Harvard School of Public Health; A.B., Princeton University.

How the States Can Tax Shifted Corporate Profits: An Application of Strategic Conformity by Darien Shanske

Article | Tax Law
How The States Can Tax Shifted Corporate Profits: An Application of Strategic Conformity
by Darien Shanske*

From Vol. 94, No. 2
94 S. Cal. L. Rev. 251 (2021)

Keywords: Tax Law, Corporate Law, State Law


The combination of pandemic, recession and federal dysfunction has put severe fiscal strain on the states. Given the scale of the crisis and the essential nature of the services now being cut, it would be reasonable for states to contemplate inefficient—and even regressive—revenue-raising measures. Yet surely they should not start with such measures. They should start with making the efficient and progressive improvements to their revenue systems that they should have made anyway.

Improving the taxation of the profits of multinational corporations—the topic of this Article—represents a reform that would be efficient, progressive, and relatively straightforward to administer. Not only would such a reform thus represent good tax policy, but it would also raise significant revenue. And, if substantial revenue, efficiency, progressivity and administrability are not sufficiently motivating, then I will also add that it would be particularly appropriate to make these changes during the pandemic so as to raise revenue from those best able to pay during the current crisis.

To be sure, the argument that states can and should tax multinational corporations more has the whiff of paradox. After all, there is general consensus that no nation-state is currently taxing multinational corporations very effectively and, further, that subnational governments are in an even worse position to do so. This is because multinational corporations can exploit the mobility of capital even more easily between parts of the same country. Nevertheless, I will argue that the American states find themselves in a particularly strong position to do better at taxing multinational corporations and this is in part precisely because of the missteps made at the federal level.

The Tax Cuts and Jobs Act (“TCJA”), passed in December 2017, contained several provisions, including rules concerning Global Intangible Low-Taxed Income (or “GILTI”), that were meant to combat income stripping. The GILTI provision identifies foreign income likely to have been shifted out of the United States and subjects it to U.S. tax.

In this Article, I argue that the states should and can tax GILTI income. The basic policy argument is simple: states should not miss a chance to protect their corporate tax bases. The amount of revenue at stake is not trivial; it could be as high as $15 billion per year for the states as a whole or the equivalent of a 30% boost in corporate tax collections.

The basic legal argument is also simple: it cannot be the case—and it is not the case—that states need to take corporations at their word as to where their income is earned. If the states can make a reasonable argument that nominally foreign income has in fact been shifted out of the United States, then their choices as to their tax system should be respected.

This Article makes several other core arguments. First, the Article argues that returning to mandatory worldwide combination as a complete alternative to GILTI conformity would be preferable to GILTI conformity alone. Second, the Article argues that offering taxpayers a choice between GILTI conformity and worldwide combination is preferable to GILTI conformity alone.

Finally, this Article places all these issues in a larger framework of strategic conformity. As with GILTI, the states should look for other opportunities where they can take advantage of federal miscues while also advancing sound tax policy.

* Professor, UC Davis School of Law. Many thanks to audiences at the Association of Mid- Career Tax Professors, the NorCal Tax Roundtable, the University of Minnesota Law School Perspective on Taxation Lecture Series and to Eric Allen, Revuen Avi-Yonah, Kimberly Clausing, Steven Dean, Peter Enrich, Michael Fatale, David Gamage, Mark Gergen, Kristen Hickman, Ken Levinson, Michael Mazerov, Amy Monahan, Susie Morse, Michael Simkovic and Adam Thimmesch. I am particularly grateful to David Gamage who coauthored some shorter pieces on which this Article is based. All opinions and mistakes are my own.


View Full PDF

A Criminal Law Based on Harm Alone: The Story Of California Criminal Justice Reform by Joshua Kleinfeld & Thomas Hoyt

Article | Criminal Law
A Criminal Law Based On Harm Alone: The Story of California Criminal Justice Reform

by Joshua Kleinfeld* & Thomas Hoyt

From Vol. 94, No. 1
94 S. Cal. L. Rev. 35 (2020)

Keywords: Criminal Law, California Law, State Law


For many criminal justice reformers, the Holy Grail of change would be a criminal system that ends the war on drugs; punishes minor property and public order offenses without incarceration (or does not handle them criminally at all); and reserves prison mainly for violent offenders. What few appreciate is that California over the last nine years has done exactly that, and the results are breathtaking in their magnitude and suddenness: from 2011 to 2019, California released 55,000 people convicted mostly of nonviolent offenses (a quarter to a third of all California prisoners) and has been declining imprisonment—which often means declining arrest and prosecution altogether—for tens of thousands more who likely would have been imprisoned a decade ago. The changes happened piecemeal; this Article is the first to put the whole picture together. But we are now in a position to describe and evaluate the whole.

We come to three conclusions. First, California criminal justice reform reduced incarceration without increasing violence, but in so doing increased property crime, public drug use, street-level disorder, and likely homelessness to such an extent as to change the texture of everyday life in some California cities, including Los Angeles and San Francisco. Second, these changes alter the relationship between individual and state substantially enough to constitute a new social contract: California has gone farther than any other American state toward a society based on John Stuart Mill’s harm principle.

Third, this array of costs and benefits is complex and nuanced enough that it is not irrational or otherwise normatively illegitimate for someone to think them either justice-enhancing or -diminishing, good for human welfare or bad for it. But what unequivocally redeems California’s new policies for California are their democratic credentials: they were accomplished through a series of elections over multiple years at multiple levels of government with a high degree of public deliberation. Criminal justice democratizers and strong proponents of federalism should endorse what California has done as a matter of political self-determination. But they might rationally not want the same thing for their own states.

*. Professor of Law and (by courtesy) Philosophy, Northwestern University. †. JD Candidate, Northwestern University Pritzker School of Law.


View Full PDF

Walling Out: Rules and Standards in the Beach Access Context by Timothy M. Mulvaney

Article | State Law
Walling Out: Rules and Standards in the Beach Access Context
by Timothy M. Mulvaney*

From Vol. 94, No. 1
93 S. Cal. L. Rev. 1 (2020)

Keywords: State Law, Beach Access, Public Policy

The overwhelming majority of U.S. states facially allocate exclusionary rights and access privileges to beaches by categorically deciding whom to wall in and whom to wall out. In the conventional terms of the longstanding debate surrounding the design of legal directives, such “rules” are considered substantively determinant ex ante and, in application, analogically transparent across similarly situated cases. Only a small number of jurisdictions have adopted “standards” in the beach access context, which—again, on the conventional account—sacrifice both determinacy and transparency for the ability to accommodate ex post the complexities of individual cases. This Article contends that beach access policy illustrates the familiar limitations of this conventional rules-versus- standards account in two elucidating ways. First, the implementation of contemporary beach access law suggests that the gap between rules and standards with respect to the virtue of determinacy is not nearly as wide as the conventional account allows. In short, beach access rules are not and cannot in actuality be divorced from context, while beach access standards take shape through applications that reveal core archetypes. Second, while beach access rules reflect the virtue of transparency in the sense that they minimize some forms of arbitrariness, standards offer their own, robust version of transparency, which is grounded in promoting dialogue and demanding accountability. The Article offers these contentions not to press the view that standards are necessarily superior to rules en masse, but, instead, to prompt reflection on the nearly uniform and seemingly impulsive rule fetishism that has held sway in the beach access context.

*. Professor of Law and Associate Dean for Faculty Research, Texas A&M University School of Law. Thank you to Gregory Alexander, Vanessa Casado Pérez, Hanoch Dagan, Nestor Davidson, Eric Freyfogle, John Lovett, Nadav Shoked, Joseph Singer, Laura Underkuffler, Brian Weeks, and Katrina Wyman for commenting on earlier drafts of this Article, and to Cole Watson for diligent research assistance. Thanks, too, to Donna, Eugene, Patrick, Jacob, and Blake Mulvaney, Ashley, Tristan, and Ryan Hedrick, Christopher, Arthur, and Amy McCann, Kevin Tray, Michael and Robert Lowe, Paul D’Elia, Edward, Mary, and Allison Norcia, Andrew Martin, Bryan Wallach, Mark Lindquist, Doug Forken, Ryan Morra, Eleanor Hish, David Manzo, James Courtney, Christopher Lilien, Robert Leichte, Chris Seiler, Marc Buttacovoli, Mary Spanburgh, Anthony Casale, Matthew Smith, Maria Dunlap, Benjamin Greer, Craig Irrgang, Maikel O’Hanlon, Matthew Duffy, John Bramlette, Richard Billings, Kevin Gregory, Matthew Popowsky, Michael Newshell, Ashley Reichelman, Michael Hazlet, and Brian Gardner, and the many others along the way who participated in and contributed to the Habitat Paddle to Build project that served as inspiration for this Article. I benefitted from the opportunity to present various iterations and components of this manuscript at Cornell Law School, Harvard Law School, Maastricht University, the University of Cambridge, the University of Edinburgh, and the University of Michigan Law School.

Stategraft – Article by Bernadette Atuahene & Timothy R. Hodge

From Volume 91, Number 2 (January 2018)



Bernadette Atuahene[*] & Timothy R. Hodge[†]

Although sometimes difficult to detect, governmental power abuses can have detrimental impacts. Property tax assessments provide an effective lens to examine this phenomenon because, given the complexity of calculating property tax assessments, it is difficult for citizens to know when local government has exceeded its legitimate taxing authority and crossed into the realm of illegal extraction. Michigan is an ideal case study because it protects property owners by making assessment-related power abuses more visible through a unique state constitutional provision: property tax assessments cannot exceed 50 percent of a property’s market value. Abuses have persisted nevertheless. Between 2011 and 2015, one in four properties in Detroit were subject to property tax foreclosure, and inflated property tax assessments that violate the Michigan Constitution are the unseen thread in this complex tapestry of foreclosure.

Against this backdrop, this Article makes three primary contributions. First, no other article has argued and proven that property tax assessments in Detroit are illegal. Using assessment and sales data from 2009–2015 for the entire City of Detroit, we find that property tax assessments are substantially in excess of the state constitutional limit, and this illegality is most pronounced for lower-valued properties. Second, to remedy inflated assessments, in 2014 and 2015 Detroit’s assessor implemented assessment decreases ranging from 5 percent to 20 percent for select districts, but we find that systemic assessment inequity persisted for lower valued properties despite these reductions. Third, this Article uses the case of illegal property tax assessments in Detroit to develop a new theoretical concept called “stategraft,” which is when state agents transfer property from residents to the state in violation of the state’s own laws and to the detriment of a vulnerable group. Although the concept was developed using the Detroit case, stategraft applies beyond Detroit to many other cases, including the discriminatory fines imposed and enforced by the police and courts in Ferguson, Missouri; broken treaties with Native Americans; and abuses of civil forfeiture laws.


Local governments must determine the taxable value of properties in their jurisdictions, and these assessments are the basis of property tax calculations. Property taxes are a substantial source of revenue for local governments, so to protect property owners from paying more than their fair share, most jurisdictions require assessments to be uniform and/or equitable.[1] But they are often neither.[2] As a result, property tax assessments provide a unique and effective lens to study the larger phenomenon of governmental power abuses that are difficult for citizens to detect or defend against.[3] This Article uses Detroit as a case study because the Michigan State Constitution goes beyond the equitable and uniform mandate, which is inexact and vulnerable to varying interpretations. Instead, the Michigan Constitution has an explicit cap on assessments that allows property owners to more easily detect assessment-related abuses. More specifically, the Michigan Constitution prohibits property assessments that exceed 50 percent of a property’s market value: “The legislature shall provide for the determination of true cash value of such property; the proportion of true cash value at which such property shall be uniformly assessed, which shall not, after January 1, 1966, exceed 50 percent . . . .”[4] Although there have been several articles about property tax assessments in Detroit,[5] no other article has proven that assessment practices in Detroit violate the Michigan Constitution.

More importantly, this Article uses the case of illegal property tax assessments in Detroit to develop a new concept called “stategraft,” which is when state agents transfer property from residents to the state in violation of the state’s own laws and to the detriment of a vulnerable group. Stategraft is a phenomenon that also applies beyond Detroit to various other cases, including discriminatory fines and fees levied by police and enforced by courts in Ferguson, Missouri;[6] broken treaties with Native Americans; and abuse of civil forfeiture statutes.

 Determining whether property tax assessments in Detroit are unconstitutional is an important research question for three primary reasons. First, in a typical property tax bill, the assessed value of each property is multiplied by the property tax rate, subject to various exemptions.[7] Detroit residents endure the highest property tax rates in Michigan and some of the highest in the country.[8] In addition, this Article shows that the majority of residential properties are unconstitutionally assessed at over 50 percent of the property’s market value. As a result, Detroit residents are being hit twice: not only are their assessed values illegally high, they are then multiplied by one of the nation’s highest rates. Many Detroiters have not been able to afford this double hit, and nonpayment has resulted in widespread tax foreclosure.[9]

From 2011–2015, the Wayne County treasurer foreclosed on 100,116 Detroit properties for unpaid property taxes, although there are only about 384,675 properties in the city.[10] This means that about one in four Detroit properties were subject to property tax foreclosure during this five-year period.[11] In 2015 alone, the treasurer foreclosed on 28,158 Detroit properties.[12] That is, in 2015, Detroit had 3,949 property tax foreclosures per 100,000 people, which is drastically higher than other cities (e.g., New York City: 52; San Francisco: 48; Los Angeles County: 8; Erie County (Buffalo), New York: 62; St. Louis County, Missouri: 197).[13]

Since the primary economic asset of many Americans is their home, property tax foreclosure can have adverse financial impacts, but it can also have injurious emotional, social, political, and cultural consequences as well.[14] Like many other material objects, homes exist in tandem with emotional landscapes, which suffuse them with significant intangible value. An individual’s childhood home, for instance, is often filled with memories and emotional investments such that when the home is foreclosed upon, a family loses more than an economic asset; their personhood is also impacted.[15] Also, a house is often the basis of social belonging—a gateway to the benefits and burdens of community membership.[16] As neighbors interact, valuable social bonds and networks are formed, and thus foreclosure often leads to social disruption.[17] Politically, property serves as a bulwark against state encroachment on individual autonomy.[18] Although people cannot remodel public buildings to reflect their individual tastes, walk around naked in federal buildings, or exclude others from a public parks at will, all of these activities are allowed in privately owned homes, where citizens have greater autonomy to live the kinds of lives they have reason to value. Foreclosure curtails this important source of autonomy. Foreclosure can also adversely affect cultural and identity interests. Individual and group identities are often closely tied to a particular home, neighborhood, or city, and a home, in particular, can reflect a person’s unique personality and become intertwined with her sense of self.[19] In sum, when the state confiscates a home for nonpayment of property taxes, there are both monetary and non-monetary consequences.

 Second, on average, Detroit’s residents are more economically and socially insecure than other Americans (see Table 1). Unconstitutional assessments inflate property tax bills, which can be a significant financial blow to this already vulnerable population—regardless of whether or not it causes them to lose their homes. In addition, less affluent owners are less likely to challenge property tax bills.[20] Thus, when poorer homeowners are subject to unconstitutional assessments, these assessments are more likely to stand uncorrected. That is, invisible abuses of state power stand uncorrected. Consequently, poor homeowners—who are most unable to afford illegal and inequitable assessments—are the population that is most often subject to them.

 Third, there is preliminary evidence suggesting that the unconstitutional assessments in Detroit are racially discriminatory.[21] On July 13, 2016, the American Civil Liberties Union (ACLU) of Michigan, along with the NAACP Legal Defense Fund and the law firm of Covington & Burling, filed a lawsuit alleging that the unconstitutional property tax assessments and the resulting foreclosures in Wayne County (the taxing jurisdiction where Detroit is located) disparately impact African Americans, who constitute 82.7% of Detroit’s population.[22] The lawsuit claims that Detroit’s assessor is unconstitutionally assessing property owners, while the assessors in Wayne County’s other jurisdictions—which are not majority African American—are not engaging in this unconstitutional behavior. Consequently, African-American homeowners “lose their homes through tax foreclosure at a higher rate than non-African-American homeowners in Wayne County,”[23] violating the Fair Housing Act of 1968.[24]

If the claims made in the lawsuit prove true,[25] then unconstitutional assessments in Detroit are part of a long and sordid history of racially discriminatory property tax administration in the United States.[26] In the Jim Crow South, officials routinely assessed property owned by African Americans at a higher proportion of its market value than they did white-owned property, leading to high property tax payments—paradoxically used to fund institutions and services that the state prohibited African Americans from accessing.[27] Also, when their real estate became valuable or officials wanted to punish them for civil rights protest actions, one conventional mechanism for dispossessing African-Americans was to inflate property tax assessments.[28] Unfortunately, these abuses of governmental power did not perish along with the Jim Crow South. A 2004 study of property tax assessments in New Haven, Connecticut found that in communities where African Americans and Latinos constituted the majority of residents, assessments were on average 40 percent higher than the market value of the homes.[29] In contrast, for neighborhoods occupied predominately by white residents, assessments were on average 20 percent less than the market value of the homes.[30]

This Article proceeds in five parts. In Part I, we examine the legal framework for assessments in Michigan. We then review the literature and discuss our methodology in Part II. In Part III, we analyze the data. Using assessment and sales data from 2009–2015 for the entire City of Detroit, we provide a dynamic view of how assessment inequity has changed during the life cycle of the housing crisis and the mayor-led, across-the-board cuts in property tax assessments, which occurred in 2014 and 2015. We find strong evidence of unconstitutional assessments across all the years studied. But assessments for lower-valued homes are substantially higher than 50 percent of their market value, while assessments for higher-valued homes are closer to this state constitutional standard.[31] Although we offer some explanatory hypotheses in Part IV, we leave it to future research to determine the economic and political forces that brought about this pervasive unconstitutionality and why it has persisted for so long. In the short term, we recommend that Detroit place a moratorium on tax foreclosures for homes that are owner-occupied until it can ensure that it is in compliance with the Michigan Constitution. The City of Detroit must also ensure that property taxes owed by tax-delinquent homeowners are accurate. Finally, in Part V, based on the Detroit case, we develop a new concept called “stategraft.” More importantly, we demonstrate how this concept applies beyond Detroit.

I.  Legal Framework for Assessments

A.  The Michigan Constitution’s Assessment Provision:

The Legislative and Judicial Interpretation

By law, all taxable properties in the City of Detroit must be assessed on an annual basis by the city’s assessment division.[32] In 1963, Michigan approved a new constitution, which drastically changed the procedure for determining the assessed value of properties. According to Article IX, Section 3 of the state constitution:

The legislature shall provide for the uniform general ad valorem taxation of real and tangible personal property not exempt by law except for taxes levied for school operating purposes. The legislature shall provide for the determination of true cash value of such property; the proportion of true cash value at which such property shall be uniformly assessed, which shall not, after January 1, 1966, exceed 50 percent; and for a system of equalization of assessments.[33]

This constitutional provision goes on to say that property tax assessments may be even lower than 50 percent of the property’s true cash value because assessments can increase by no more than 5 percent per year so long as the property is not sold.[34] In addition, the 1978 Headlee Amendment placed additional limits on local taxing powers, including voter approval for increases in property tax rates (also known as millage rates) that go beyond the Amendment’s guidelines.[35]

The Michigan legislature codified Article IX, Section 3 of the state constitution in Section 211.27a(1) of the Michigan Compiled Laws: “[e]xcept as otherwise provided in this section, property shall be assessed at 50% of its true cash value.”[36] The legislature has defined true cash value as “the usual selling price at the place where the property to which the term is applied is at the time of assessment, being the price that could be obtained for the property at private sale . . . .”[37] The Michigan Supreme Court has declared that true cash value and fair market value are synonymous, and this is now a well-settled principle.[38] Consequently, if Detroit properties are assessed at more than 50 percent of their fair market value, there is a direct breach of the state constitution and supporting legislation.

It is also well established that assessors are allowed to use diverse appraisal methods to determine a property’s market value.[39] The legislature did not designate a specific valuation method for determining a property’s true cash value, and so the task of determining which approach is most appropriate for any given context has fallen to the Michigan courts.[40] As the Michigan Supreme Court explains, “[a]ny method for determination of true cash value which is recognized as accurate and reasonably related to fair market valuation will fill the statutory prescription and is an acceptable indicator of true cash value.”[41]

The courts have recognized three standard approaches: (1) the cost-less-depreciation approach, (2) the sales-comparison or market approach, and (3) the capitalization-of-income approach.[42] More importantly, the courts have made clear that whichever approach is used, “[t]he ultimate goal of the valuation process is a well-supported conclusion that reflects the study of all factors that influence the market value of the subject property.”[43]

For appraisals of residential housing, the industry standard is the market approach, which determines a property’s true cash value by analyzing recent sales of comparable properties.[44] This approach requires assessors to adjust the sale price of comparable properties, taking into consideration factors such as the property’s size, age, condition, location, existing use, zoning, natural assets, and present economic income.[45] One potential drawback of the market approach is that, in certain instances, the sale price might not be the “usual selling price” due to factors personal to the parties.[46] For instance, it is not uncommon for parents to transfer property to their children at discounted rates. Nevertheless, when the assessor considers the sale prices of numerous comparable properties, she reduces the likelihood that factors extrinsic to the properties will impact price calculations.[47]

In addition to having a robust sample of properties, the law requires assessors to include only arm’s length transactions—which is when there is a willing buyer and a willing seller and, thus, the sale price reflects the demand and supply for property in the market.[48] There is, however, an exception for instances where auctions are the “common method of acquisition” for properties in the area.[49] Although our data show that auctions have become a common method of acquisition, our estimates rely only on arm’s length transactions so that they are as conservative as possible. Consequently, we use arm’s length residential sales transactions alongside assessment data to determine whether Detroit is complying with the state constitutional mandate to assess each property at no more than 50 percent of its market value.[50]

B.  The Michigan Constitution’s Assessment Provision:

The Historical Origins

Several state constitutions require property tax assessments to be uniform and equitable.[51] But, by requiring that assessed value/market value ratio not exceed 0.5, Michigan’s constitution is one of the few to use an explicit assessment ratio to constrain its local governments. Our review of every state constitution shows that Louisiana, Mississippi, Oregon, New Mexico, Oklahoma, and Tennessee are the only other states to constitutionally require localities to assess properties at a specified fraction of their market value.[52] Given that these types of constitutional provisions are uncommon, how did this cap on assessed values end up in the Michigan constitution?

The historical evidence suggests Michigan’s constitutional drafters feared that without a specific constitutional provision, localities would abuse their powers and inequitably assess properties. One delegate from the 1962 constitutional convention—which was charged with drafting Michigan’s current state constitution, enacted in 1963—wrote, “[w]ell, there was very widespread distrust of the Legislature and people said, ‘leave it up to the Legislature and nothing will be done or it will not be done right. Let’s put it right in the Constitution and then we know the matter will be taken care of.’”[53] Thus, the very purpose of the Michigan Constitution’s explicit mandate that property tax assessments cannot exceed 50 percent of a property’s market value was to prevent local governments from abusing their discretionary powers.[54]

Michigan’s prior (1908) constitution required localities to assess all properties at full cash value,[55] but according to Barlowe:

Wide variations exist in the extent to which assessors abide by this provision. Assessment data studies conducted in recent years show that some properties are assessed at as little as two per cent of their market values while others are sometimes assessed at more than their full market sale values. Township, city, and county equalized assessments frequently represent levels of between 30 and 50 per cent of cash value.[56]

The issue of assessment inequity was so important that the governor, John Swainson, filed a report at the 1962 constitutional convention, detailing the unanimous findings of his Special Commission on Industrial Development Legislation: assessments varied widely and resulted in non-uniform property taxes that ultimately deterred investment.[57]

To address the well-documented assessment inequities, the Committee on Finance and Taxation introduced an amendment at the convention.[58] According to the Committee, the state’s 1908 constitutional provision allowed the court to “simply state that since the former taxpayer is not assessed in excess of 100%, the Constitution affords no basis for relief. This unavailability of judicial relief has made the cash value standard a positive impediment to the achievement of uniform assessment.”[59] Consequently, the Committee drafted an amendment explicitly stating that property assessments were not to exceed 50 percent of the property’s market value.[60] The committee suggested that 50 percent was the appropriate limit because it was “currently used by the State Tax Commission,” and it said that the standard was not unduly inflexible because the legislature would still have the ability to “change the standard to reflect changes in the general price structure.”[61] The committee set the start date for this provision in 1966, as an acknowledgment that some jurisdictions currently relied on assessments above 50 percent.[62]

A minority faction disagreed with the 50 percent limit and crafted a counter-amendment to delete it. [63] The faction’s critique of the 50 percent limit was based on issues of sufficient revenue for local governments, the proper role of the legislature, and the effectiveness of the 50 percent limit in actually providing relief to taxpayers. First, the minority report predicted “devastating immediate effect on the fiscal affairs of some communities.”[64] Second, the minority report argued that setting a fixed limit to assessment levels was not the role of the constitutional convention: “The inclusion of an arbitrary 50% limitation, statutory in nature, does not have the dignity of a mandate of the people, as does the sales tax rate limitation, expressed in a referendum.”[65] Third, the report argued that the 50 percent level did not make practical sense. Its authors argued that just because the State Tax Commission had used 50 percent as the proper assessment level when equalizing taxes among units of government did not mean that assessments should now be limited to 50 percent.[66] The report’s authors also argued that while using a 50 percent limit would reduce taxes for some individuals and industries, it would increase taxes for others because all assessments would be brought closer to 50 percent in an attempt to make up for the lost revenue.[67] The counter-amendment, however, was handily defeated 84 to 43.[68]

Today, Detroit cannot continue to ignore the clear and unequivocal limit on assessments, which withstood vigorous democratic debate to earn its place in the Michigan Constitution.

II.  Literature Review and Methodology

The Minnesota Center for Fiscal Excellence estimates that Detroit properties are subject to effective tax rates of 3.82%, more than double the national urban average of 1.5%.[69] Given Detroit’s notoriety for high property tax rates, there have been several studies examining this.[70] We, however, specifically build upon studies exploring assessment inequity in Detroit because our goal is to determine whether the city is complying with its state constitutional mandate to assess properties at no more than 50 percent of their market values.[71] For example, several economists have examined assessment inequity among Detroit residents using assessment and market data from 2010.[72] They observed severe over-assessment and regressivity at levels not previously observed in the United States.[73] In addition, Dewar et al. compared assessed values and sales prices in one Detroit neighborhood, Morningside, from 2008 to 2013.[74] They found that although market values declined over this period, assessments did not keep pace.[75] They also found that lower-valued homes were more likely than higher-valued homes to be significantly over-assessed during this period.[76]

Unlike the existing literature, our study provides a dynamic view of property tax inequity across all Detroit neighborhoods over a seven-year period, and we discuss the legality of the inequities we observe. In doing so, we make three unique contributions to the existing literature. First, we examine how assessment inequity has changed through time for the entire City of Detroit. Second, in response to prior evidence of over-assessment,[77] Detroit’s assessment office tried to remedy this by implementing assessment decreases for two consecutive years, resulting in planned assessment reductions of 5 to 20 percent for select districts each year.[78] We examine whether systemic assessment inequity persisted despite these reductions. Third, we examine how Detroit’s inflated assessments contravene the Michigan Constitution.

The International Association of Assessing Officers (“IAAO”)[79] has outlined the appropriate methods for analyzing assessment accuracy, and there are two major components: level and uniformity.[80] Level refers to the overall assessment ratio, defined as a property’s assessed value divided by its market value.[81] Two key statistics for analyzing assessment ratios are the mean and median.[82] Given Michigan’s constitutional requirement that assessments should not be greater than 50 percent of market value, the mean and median levels of assessment are expected to be no higher than 0.5.[83] If the ratios derived are consistently higher than 0.5, then this is reliable evidence that assessments in Detroit are systematically illegal, and the assessing officials must take corrective action.

In contrast, uniformity (also known as variability) refers to the degree to which assessment ratios have achieved horizontal and vertical equity. Horizontal equity means that properties similarly situated in terms of market value and type of neighborhood have similar assessment ratios, while vertical equity means that higher and lower priced properties have similar assessment ratios.[84] Horizontal and vertical inequity occur when assessment ratios vary widely. There are two methods for analyzing assessment uniformity: the Coefficient of Dispersion (“COD”) and the Price Related Differential (“PRD”).[85] The COD measures horizontal equity by quantifying the degree to which the assessment ratios deviate from the median.[86] The IAAO considers assessments uniform if the COD is between 5 and 15 for single-family residential properties.[87] The PRD is a measure of vertical equity. If high-value properties have lower ratios than low-value properties, the vertical inequity is termed regressive; if the opposite occurs, it is called progressive. Assessments should be neither regressive nor progressive because—although a degree of variation between ratios is expected as well as accepted—large differences result in the non-uniform and inequitable application of property tax assessments.[88] The PRD is calculated by dividing the mean ratio by the weighted mean ratio (the total assessed value of all property divided by the total sale price).[89] According to the IAAO, PRDs between 0.98 and 1.03 are vertically equitable while values greater than 1.03 are regressive, and values less than 0.98 are progressive.[90]

Beyond the IAAO’s standard measures, a simple approach for examining horizontal and vertical equity is to estimate the mean, median, minimum, and maximum assessment ratios across sale price quintiles.[91] This allows further analysis within a quintile (horizontal equity) and an examination of how ratios vary across quintiles (vertical equity). In Michigan, if the mean and median are less than or equal to 0.5 within a quintile, then horizontal equity is present and the assessments are systematically constitutional. If mean and median assessment ratios are no higher than 0.5 across quintiles, then the assessments are vertically equitable and constitutional.[92]

Our proposed method of analysis is called an assessment ratio study (also known as ratio study or sales ratio study), which is the primary mechanism that assessors, taxpayers, appeal boards, and taxing authorities use to determine if assessments meet the legal requirements of a jurisdiction.[93] IAAO defines an assessment ratio study as “a form of applied statistics, because the analyst draws conclusions about the appraisal of the population (the entire jurisdiction) of properties based only on those that have sold during a given time period.”[94] If the unsold parcels are appraised in the same manner as the sold ones, then it is valid to use the statistics derived from the sales ratio study to infer appraisal performance for unsold parcels. Our evidence shows that homes selling during the period under study are comparable to homes that did not sell. For example, the average age of all improved, taxable, residential property was 81.46 years, and the average age of our trimmed 2015 sample was 81.61 years. 

III.  Data Analysis

Since assessments in Detroit are calculated annually and are based on property values from the previous year, we divided assessed values with prior year sales information to produce assessment ratios. That is, we match assessment data from each year (20092015) with sales data from the prior year (20082014). For properties that were sold multiple times in the same calendar year, we took the last sale in a year as the determinant of the property’s value. As a result, the full dataset includes 123,400 residential property transactions (we exclude empty lots and non-taxable properties). The breakdown of data by year can be viewed in column (1) of Table 2. Data Driven Detroit provided parcel-level information on assessed values for all Detroit properties from 2009 to 2014.[95] We used the City of Detroit’s Open Data Portal to secure data on all property sales in Detroit from 2008 through 2014, as well as assessed values for 2015.[96]

For several reasons, we excluded 117,214 observations from the sample, leaving a total of 6,186 arm’s length property transactions in our analysis (column (2) of Table 2). First, our dataset included twenty-two different sales terms defined by the assessor’s office, and we excluded properties that were not arm’s length transactions.[97] Although the law states that only arm’s length transactions should be included, there is an exception if non-arm’s length transactions (distressed sales) “have become a common method of acquisition in the jurisdiction for the class of property being valued.”[98] Given that only about 5 percent of sales from 20082014 were arm’s length transactions (see Table 2), there is a strong argument that distressed sales are, in fact, the common method of acquisition for residential properties in Detroit. Nevertheless, we only include arm’s length transactions so that our estimates provide the most conservative measure of unconstitutionality. Scholars who include distressed sales in their analysis will only find unconstitutionality that is markedly more pronounced because these transactions involve smaller sums.[99]

Second, we excluded properties that were bundled and sold as a single transaction because it is impossible to determine the price of any single property within the bundle. Bundled properties are different parcels with identical sellers, sale dates, and sale prices.[100] Third, we excluded properties that had a sale price of zero because it is unlikely that these were arm’s length transactions; we also excluded properties with an assessed value equal to zero because these are likely to be non-taxable properties.

Although excluding 95 percent of the total observations may seem extreme, this is a result of the forced auctions and other non-arm’s length transactions that have proliferated in Detroit’s distressed real estate market. In addition, we further trimmed the data in accordance with IAAO’s nationally recognized standards, which recommend trimming the sample of statistical outliers because very low or high ratios can severely distort the analysis.[101] We show the total observations after trimming in column (3) of Table 2, and the percentage of trimmed observations in column (4). The number of observations we excluded from our sample due to outlier trimming is below the IAAO’s recommended standard of 10 percent.[102]

Table 3 shows the mean and median assessment ratios; the COD and PRD, which measure the uniformity and variability of the ratios; and the average sale prices for each year, highlighting changing market conditions. We also include the percentage of assessment ratios that are above the 0.5 constitutional limit to evaluate the quantity of unconstitutionally assessed properties. Table 3 provides substantial evidence that Detroit assessors are unconstitutionally assessing properties in a systematic fashion. First, for properties sold between 2008 and 2014, the majority of assessments violated the Michigan Constitution: 2009 (65%); 2010 (84%); 2011 (53%); 2012 (73%); 2013 (78%); 2014 (83%); 2015 (65%). Second, data from the assessment ratio study—which is required for making inferences about properties that were not sold during the period—show that the mean and median assessment ratios are greater than the constitutionally permitted ratio of 0.5.[103] While the 2011 and 2012 ratios are closer to 0.5, they remain statistically larger than 0.5 when the margin of error is set at the 95% confidence level. Third, the COD for each year is much higher than the recommended maximum of 15, which is evidence that assessments are not uniform. In addition, the PRD for each year is above 1.03, which is evidence of regressivity (i.e., higher-valued properties have lower ratios than lower-valued properties).[104] Again, we see the COD and PRD for 2011 and 2012 are closer to the recommended levels, yet remain above it.

Upon examining Table 3, one question remains: Why are 2011 and 2012 closer to the constitutionally permitted ratio of 0.5 while in the other observed years the ratios suggest even greater unconstitutionality? We provide two potential explanations. For one, the assessor classified fewer sales as arm’s length transactions in 2011 and 2012: it classified 1.7% and 3.6% of the total number of transactions as arm’s length for 2011 and 2012, respectively, whereas it classified between 2.6% and 12.4% of transactions as arm’s length in the other years. Second, median sale prices derived from the assessor’s data are much higher in 2011 and 2012 compared with the surrounding years. More importantly, the assessor’s sales data from 2011 and 2012 are not consistent with sales information from Zillow market data in these years, despite consistency between the two data sets in the other years.[105] To check our two explanations, we did a separate analysis that included sales labeled as “Review Needed,” a sales term used for transactions that the assessor did not have enough information to classify, and the number of observations and the average sale prices became consistent with the surrounding years; in addition, the average sales prices for 2011 and 2012 became consistent with Zillow market data. With the stated correction, the average assessment ratios would increase to 3.65 and 4.90 for 2011 and 2012, respectively, suggesting there was substantial unconstitutionality in 2011 and 2012.[106]

 Table 3 also provides mixed results concerning the across-the-board assessment reductions discussed by Mayor Duggan.[107] There is no evidence of cuts in 2014 because the mean ratio increased by 10 percent. Furthermore, vertical inequity worsened as the PRD increased in 2014, while the COD declined only slightly. However, overall assessment ratios declined in 2015 by approximately 26 percent, resulting in reductions that were larger, on average, than the 5 to 20 percent reduction the mayor expected.[108] Despite the large reductions, vertical inequity persisted as the PRD remained above 1.03 (at 2.45), and horizontal inequity actually worsened as the COD increased to 125.36 (up from 107.34).

To further explore assessment practices in Detroit, Table 4 provides simple snapshots of how assessments vary within and across sale price quintiles by showing the mean, median, minimum, and maximum assessment ratios. We highlight the average sale price of each quintile in the row labeled “Mean SP,” and the percent assessed above the constitutional 0.5 limit is in the row labeled “%>0.5”. The most important finding shown in Table 4 is that average assessment ratios decline as property values increase. That is, Detroit’s assessors are unconstitutionally assessing lower-valued properties by a substantial margin, while the assessment ratios for higher-valued property are at or even below the constitutionally permitted limit of 0.5. Specifically, the lowest-valued properties in 2009 were, on average, assessed at levels almost eighteen times larger than the constitutionally permitted 0.5 limit (8.87); middle-valued properties were assessed three times more than the constitutionally permitted limit (1.54); and the highest-valued properties were assessed below the constitutionally permitted limit (0.4). In addition to this stark evidence of vertical inequity, assessment ratios are less uniform for lower-valued property (ranging from 0.25 to 36.29) compared with higher valued property (ranging from 0.04 to 1.6), providing evidence of regressive horizontal inequity. These results are generally consistent across all years; however, for reasons mentioned earlier, 2011 and 2012 ratios are closer to 0.5.[109]

Table 4 further explores how Detroit Mayor Mike Duggan’s across-the-board reductions in property assessments—ranging from 5 percent to 20 percent, depending on the district—affected the ratios. Table 4 provides some evidence of reductions in 2014, with mean assessment ratios decreasing in quintiles 1, 2, and 5, while reductions were higher than the mayor projected for 2015. Specifically, Table 4 shows reductions ranging from 12% to 47% in 2015, depending on the quintile viewed.[110] More importantly, Table 4 shows that even after the reductions, the vast majority of lower-valued properties had an average assessment ratio equal to 3.29, not even close to the 0.5 level, while higher-valued properties—which already had ratios closer to the constitutionally permitted limit of 0.5—received reductions that brought assessments to an average ratio of 0.29. As a result, for quintile 1, assessed values were, on average, $18,507 above the constitutional limit, a pattern repeated in quintiles 2 ($19,900), 3 ($13,796), and 4 ($4,890). But quintile 5 was, on average, assessed below the constitutional limit by $10,881.

We use two specific properties to better illustrate how owners of lower-valued properties are bearing the undue burden of unconstitutional assessments. In the first example, the public record shows that a property located at 15455 Artesian Street sold in 2009 for $10,000; the assessor marked this sale as an arm’s length transaction. Nevertheless, the assessed value of that property in 2010 was not $5,000 (50 percent of the property’s market value), but rather $36,094, which is 7.2 times the constitutional limit. In a contrasting example, the public record shows that, in 2008, a property located at 4127 Buckingham Avenue sold for $115,000 in an arm’s length transaction. The assessed value in 2009 was $41,369, which is $16,131 below the constitutional limit.

As a final mechanism for examining assessment trends in Detroit, Table 5 highlights the changes in average assessment ratios across districts. Because it is district specific, Table 5 is the one place we see evidence of Mayor Duggan’s 2014 assessment reductions: District 2 (25%), District 3 (12%), District 7 (20%), District 9 (35%), and District 10 (24%). The remaining districts experienced persistent over-assessment with no correction. In 2015, assessments decreased between 10% and 69% for all districts excluding District 8 (40% increase).[111]

In sum, we find that the Detroit assessor is systematically assessing Detroit homeowners at levels that violate the Michigan Constitution. Even though we argue that the 2011 and 2012 sales data were inconsistent,[112] we nevertheless find substantial unconstitutionality using this faulty data.


The lingering question is: why has the Detroit assessor’s office systematically violated the Michigan Constitution? Although this question is beyond this Article’s scope, there are four possible answers. The first hypothesis is that the assessor’s office lacked the capacity to conduct the legally required market-based annual assessments,[113] and thus new assessments were based on non-market related, incremental adjustments to existing assessments.[114] This makeshift method based on conjecture became completely unworkable in 2008 when property values plummeted in Detroit (See Figure 1) and the chasm between incrementally adjusted home prices and actual home prices widened sharply. While the assessor relied on Detroit residents to appeal incorrect assessments, research shows that poor people are less likely to appeal their property tax assessments; and when they do, they have lower success rates than wealthier people.[115]

The Auditor General’s Performance Audit shows that the Detroit Assessor’s office lacked capacity to do its job properly. The report’s authors state:

As a result of our audit, we have concluded that the overall operation of the Assessments Division falls short of their goals and objectives. Moreover, guided by the State Tax Commission’s 14-Point Local Unit Review (of assessing units), we found that the Division’s assessing operations are inefficient, ineffective, and lacking in some areas of its assessing activities.[116]

The report’s findings are jaw dropping. For instance, auditors found that about 1,700 properties were still classified as tax exempt although they were no longer owned by the city.[117] Also, according to a representative from the Buildings, Safety Engineering and Environmental Department, over 70 percent of the tickets this department issued for blight violations were dismissed because the property ownership information it received from the assessor was incorrect.[118] In sum, the Auditor General’s report paints the Detroit Assessor’s office as a highly incompetent, inefficient, and broken bureaucracy, lacking the capacity to fulfill its mission.

The second hypothesis is that although there are various community organizations working on the property tax foreclosure crisis, community activists were unable to hold the Detroit assessor accountable because they were unaware of the state constitutional provision or they did not have access to the data and empirical skills necessary to prove systemic unconstitutionality.

The third hypothesis is that it is easier for Detroit to raise revenues by illegally assessing residents than by legally raising property tax rates. A Michigan constitutional amendment—the 1978 Headlee Amendment—requires voter approval for increases in property tax rates that go beyond the Amendment’s guidelines.[119] Given the property tax rate is already one of the highest in the nation, it may be challenging to get voters to approve a hike. In contrast, Wayne County reimburses Detroit for any property tax revenue that it fails to collect; in exchange, the county receives the right to collect the revenues (with penalties and interests) and to confiscate the home if payment is not forthcoming after three years of delinquency.[120] Without this arrangement, the Detroit officials would be incentivized to legally assess residents, which would decrease revenues but increase the probability that people would be able to afford their property taxes. With this arrangement, the city has a short-term incentive to make assessments high and extract as much money as possible to fill budget shortfalls, even though this is not in the city’s long-term interest because tax foreclosure devastates neighborhoods.

A fourth hypothesis is that even when the county does not make much money from the foreclosure auction sales, the county makes significant sums of money in late fees and interest, which is supposed to go into the Delinquent Tax Revolving Fund.[121] The Fund should be used solely for foreclosures, but Wayne County has been using the Fund to fill its chronic budget shortfalls and to recover from its recent financial emergency.[122] Consequently, the financially distressed county has come to rely on property tax foreclosures in Detroit to stay afloat.

These four hypotheses attempt to explain why Detroit has been unconstitutionally assessing its residents, but they have not yet been empirically confirmed or refuted. We leave this for other scholars to investigate.

V.  Stategraft

The findings of the study are clear: The City of Detroit is assessing homeowners in violation of the Michigan Constitution. More significantly, city and county coffers have benefitted greatly from this theft. But, in many ways, Detroit’s illicit actions are not unique. It is not unprecedented for state agents to increase state revenues using illegal means, and we create the term stategraft to describe this phenomenon.[123] Stategraft is when state agents transfer property from residents to the state in violation of the state’s own laws and to the detriment of a vulnerable group. The term intentionally combines the words statecraft and graft or corruption. A widely agreed-upon definition of corruption is “[t]he abuse of an entrusted power for private gain,”[124] but stategraft is different because there is no private gain. Instead, the abuse of state power primarily benefits the state itself. Statecraft is the art of conducting state affairs,[125] but stategraft highlights instances when state agents advance the state’s financial interests by stealing from those under its authority.

The term stategraft was born from the case of property tax injustice in Detroit. Detroit assessors are state agents who imposed inflated property tax assessments on Detroit property owners, and this act has transferred millions of dollars from property owners to city and county coffers in violation of Michigan’s state constitution. There are, however, several other poignant examples of stategraft. To initiate the conversation, we will discuss three.

The first example is lands taken from Native Americans in direct violation of a valid treaty. For instance, the Oneidas once occupied about six million acres in what is now New York.[126] Likewise, the Black Hills of South Dakota once belonged exclusively to the Lakota Sioux who considered it sacred land.[127] In both cases, U.S. state agents entered into legally binding treaties with the tribes but then later reneged and commandeered native lands, illicitly increasing their nascent nation’s territory. That is, the state itself (not private actors) benefitted from property illicitly taken from a vulnerable group. Prior to the introduction of the term stategraft, there was no vocabulary to discuss this phenomenon.

Second, the Department of Justice (DOJ) has sued the City of Ferguson for what amounts to stategraft.[128] The DOJ argues that Ferguson’s police have engaged in unconstitutional policing that targets African Americans and unfairly imposes civil and criminal fines on them.[129] Instead of reining in the unlawful police conduct, the courts issued arrest warrants when the fines went unpaid, advancing the city’s financial interest and foregoing their role as neutral arbiter.[130] It is the City of Ferguson itself, rather than individual police officers or judges, which financially benefits from the unconstitutional policing of African Americans: a classic case of stategraft.

The third and final example of stategraft is the abuse of civil forfeiture laws. These laws are meant to allow law enforcement to confiscate property obtained unlawfully and to use the proceeds in the fight against crime.[131] Nevertheless, in many cities, civil forfeiture has been abused, and police officers illegally seize the property of people never charged or convicted of any crimes.[132] When the opaque legal process that individuals must traverse to regain their property is more costly than the property taken, the police department often keeps the property.[133] For instance, in 2013, the Washington, D.C. public defender service filed a lawsuit against the district on behalf of 375 car owners who shared one thing in common: the police confiscated their vehicles but did not charge them with crimes.[134] The district court granted a preliminary injunction in the case, finding that the main plaintiff was likely to prevail on his claim under the Due Process Clause in the US Constitution.[135] Given the constitutional violations brought to light by the suit, the parties settled the case after the City agreed to modify its forfeiture laws.[136] More significantly, this is a classic case of stategraft because police officers abused civil forfeiture laws and confiscated cars from innocent people whose poverty rendered them unable to recover their vehicles. Most importantly, it was the police department itself (rather than individual officers) that financially benefitted from this illicit action.

This list of examples is not meant to be exhaustive; instead, the goal is to highlight the pervasiveness of stategraft, which has five principle elements: (1) state agents, (2) transferring property, (3) from residents to the state, (4) in violation of the state’s own laws, (5) to the detriment of a vulnerable group.

 State agents. “State agents” are individuals or groups who use the state’s police powers to compel people to act or refrain from taking certain actions.[137] When working in their official capacity, state agents either have the power to make laws, policies, and administrative rules, or are acting upon the explicit directions of those with the power to do so. But in their unofficial capacity, state agents promote the state’s financial interests without explicit permission from lawmakers, who either look the other way or support their actions without directly authorizing them. In Ferguson, for example, the targeting of African Americans for fees and fines was not an official written policy, but it was a pervasive practice which used the police and courts to enlarge city coffers.[138]

Transferring property. Property includes tangible property such as money, financial instruments, real property, and personal property. It also includes intangible property like entitlements, licenses, and intellectual property. Property transfer can involve taking away the right to use, exclude, or transfer property, or taking the property itself. In the three examples discussed above, the American government took land from native Americans;[139] the city of Ferguson took money away from its African-American citizens using discriminatory fees and fines;[140] through civil forfeiture, many cities have taken all types of real and personal property away from people not charged with a crime;[141] and illegal property tax assessments have led to inflated property tax bills in Detroit.[142]

From residents to the state. For these purposes, a “resident” is anyone who is subject to the state’s police powers, even if the person is not physically within the state’s geographic boundaries.[143] To qualify, property must be transferred from residents directly or indirectly to state accounts. For instance, land usurped from Native Americans went to U.S. federal and state governments;[144] fees and fines charged by Ferguson enlarged that city’s coffers; and property taken through civil forfeiture benefits the law enforcement agencies that confiscate the property.[145]

 In violation of the state’s own laws. Laws include legislation, judicial decisions, administrative rules, and policies that have the power to bind residents and regulate their actions.[146] Only the law in effect at the time the property was taken is relevant. In the examples above, police officers abuse civil forfeiture laws to illegally confiscate property, the US government abrogated legally binding treaties with Native peoples; the DOJ has argued that the City of Ferguson contravened the Equal Protection and Due Process clauses of the US constitution; and the Detroit assessor violated the Michigan Constitution’s property tax assessment limits.

 To the detriment of a vulnerable group. If the state breaks a law, then ideally citizens can go to the courts, police, or other state functionaries to correct the injustice. A vulnerable group is one that occupies a subordinate position within the polity, so they are often unable to identify or resist stategraft due to limited financial resources, lack of access to justice, and inadequate information. That is, vulnerability makes groups more likely to experience stategraft and also less able to combat it when it does occur. Tax injustice in Detroit is the perfect example. Our data show that homeowners with lower-valued homes experienced unconstitutional property tax assessments at a far greater rate than homeowners with higher-valued homes.[147] This is because people with more information and resources had the ability to hire lawyers and other intermediaries to assist with appealing their property tax assessments and averting stategraft. But due to lack of information and resources, Detroit’s established appeals process was unavailable to the most vulnerable populations.

Stategraft is a valuable concept because existing categories do not well describe this phenomenon. The most common definitions of corruption are predicated on private gain, and there is currently no concept to describe state-led theft that primarily benefits the state.[148] Constitutional takings is a category ill-equipped to describe the phenomenon because while sometimes instances of stategraft qualify as a takings, other times they do not. For instance, while courts have ruled that certain broken treaties with Native Americans violated the Fifth Amendment’s Takings Clause,[149] illegal property tax assessments in Detroit would most likely not qualify because courts have ruled that the state’s tax-and-spend powers do not infringe upon the Takings Clause.[150] Courts have also ruled that civil forfeiture does not run afoul of the taking clause.[151] Lastly, because stategraft, by its definition, involves an illegal transfer of property, the due process clauses of the Fifth and Fourteenth Amendments will always apply.[152] Stategraft, however, is a unique type of due process violation that deserves focused analysis because when a state steals from its own citizens, this has dire consequences for a nation’s democratic foundations. Consequently, stategraft is a theoretical framework that fills an important gap in the literature. We have provided the theoretical scaffolding for the concept of stategraft, upon which other scholars can build.


Property tax assessments provide a penetrating lens into governmental abuses of power that elude detection. Given the complexity of calculating property tax assessments, it is difficult for citizens to know when local government has exceeded its legitimate taxing authority and crossed into the realm of illegal extraction. By assessing properties in violation of the Michigan Constitution, Detroit has crossed this line and is engaging in stategraft. The illegality is systemic and has persisted since 2009, which is when property values in Detroit declined precipitously. In 2014, the City of Detroit recognized the need to comply with its state constitution by correctly aligning assessments with property values. Consequently, in 2014 and 2015 Mayor Duggan announced across-the-board cuts to property assessments. Nevertheless, lower-valued properties are still assessed far in excess of 50 percent of their market value, while the assessed values for higher-valued properties have fallen below this constitutional limit.

The illegal assessments in Detroit have severe consequences, the most dire being that it has unleashed a property tax foreclosure epidemic.[153] Based on our analysis, we recommend that Detroit place a moratorium on property tax foreclosures of owner occupied homes until it can ensure that it is in compliance with Michigan’s constitutional and statutory laws. At the very least, Detroit must ensure that delinquent taxpayers—on the verge of foreclosure—are not subject to unconstitutional tax assessments and inflated property tax bills.[154] More significantly, we encourage the City of Detroit to begin a dialogue with its citizens about how to heal from the social, economic, and psychological consequences of stategraft.


[*] *. Professor of Law, Chicago-Kent College of Law; Research Professor, American Bar Foundation.

[†] †. Assistant Professor of Economics, Oakland University. The authors give their sincere thanks to the following people who gave this Article a close read and provided helpful comments: Robert Ellickson, Carol Rose, Lee Anne Fennell, Joseph Singer, Evelyn Brody, Tracy Gordon, and Gary Sands. The authors are grateful for the opportunity to have presented this paper at the Property Works in Progress (PWIP) conference at Boston University and the Empirical Critical Race Theory Conference at Yale Law School, as well as in faculty workshops at the University of Iowa College of Law, Vanderbilt University Law School, University of Michigan Law School, University of Windsor Faculty of Law, University of Georgia School of Law, and the American Bar Foundation. A heartfelt thanks to Data Driven Detroit and Loveland Technologies, which were instrumental in acquiring the data used in this Article. Last but not least, we would like to thank our research assistants, Roisin Duffy-Gideon and Ayla Syed, for their exceptional work. The National Science Foundation funded this research.

 [1]. See Glenn Fisher, The Worst Tax?: A History of the Property Tax in America 152, 188 (1996). See, e.g., Pa. Const. art. VIII, § 1 (“All taxes shall be uniform, upon the same class of subjects, within the territorial limits of the authority levying the tax, and shall be levied and collected under general laws.”); Utah Const. art. XIII, § 2 (“[A]ll tangible property in the State that is not exempt under the laws of the United States or under this Constitution shall be: (a) assessed at a uniform and equal rate in proportion to its fair market value, to be ascertained as provided by law . . . .”).

 [2]. See Lee Harris, Assessing Discrimination: The Influence of Race in Residential Property Tax Assessments, 20 J. Land Use & Envtl. L. 1, 3–4 (2004). See generally Olha Krupa, Housing Crisis and Vertical Equity of the Property Tax in a Market Value-Based Assessment System, 42 Pub. Fin. Rev. 555 (2014) (discussing increases in vertical inequity in Indiana as a result of corrupt property tax assessments); Carmela Quintos, Improving Assessment Equity in Mass Appraisal Models, 11 J. Prop. Tax Assessment & Admin. 53 (2014) (affirming that knowledge-based, methodological problems, as well as logistical, administrative problems associated with tax assessments, frequently compromise both vertical and horizontal equity).

 [3]. See, e.g., Itai Beeri & Doron Navot, Local Political Corruption: Potential Structural Malfunctions at the Central-Local, Local-Local, and Intra-Local Levels, 15 Pub. Mgmt. Rev. 712, 720–21, 729, 735 (2013); Brian Rapp & Frank Patitucci, Improving the Performance of City Government: A Third Alternative, 6 Publius 63, 76–77, 89 (1976); Shelley Ross Saxer, When Local Government Misbehaves, 2016 Utah L. Rev. 105, 115–18 (2016).

 [4]. Mich. Const. art. IX, § 3. See also Mich. Comp. Laws § 211.27(a)(1) (2013).

 [5]. E.g., Gary Sands & Mark Skidmore, Lincoln Inst. of Land Policy, Detroit and the Property Tax: Strategies to Improve Equity and Enhance Revenue 3 (2015) (proposing strategies to increase tax revenue and mitigate inequity in property tax assessments); Timothy R. Hodge et al., Assessment Growth Limits and Mobility: Evidence from Home Sale Data in Detroit, Michigan, 68 Nat’l Tax. J. 573, 590–91, 595–96 (2015) (concluding that Michigan’s assessment growth cap makes property sales less likely); Timothy R. Hodge et al., Assessment Inequity in a Declining Housing Market: The Case of Detroit, 45 Real Est. Econ. 237, 254 (2017) [hereinafter Hodge et al., Assessment Inequity] (concluding that assessment practices have created horizontal and vertical inequity); Timothy R. Hodge et al., Tax Base Erosion and Inequity from Michigan’s Assessment Growth Limit: The Case of Detroit, 43 Pub. Fin. Rev. 636, 648–55 (2015) (stating that Michigan’s assessment growth cap creates horizontal and vertical inequity); Mark Skidmore et al., Property Value Assessment Growth Limits and Redistribution of Property Tax Payments: Evidence from Michigan, 63 Nat’l Tax. J. 509, 515 & n.17 (2010) (finding that Michigan’s assessment growth cap allows homeowners to see a tax reduction over time).

 [6]. See U.S. Dep’t of Justice Civil Rights Div., Investigation of the Ferguson Police Department 12–15 (2015).

 [7]. See Mich. Legislature, Michigan Taxpayer’s Guide 1–2 (2016) (explaining the state assessment process).

 [8]. See Mich. Dep’t of Treasury, 2015 Ad Valorem Property Tax Report 2 (2016), (showing Wayne County with the state’s highest average property tax rate, at $55.11 per $1,000 of assessed value).

 [9]. To understand the property tax foreclosure process in Wayne County, see Real Property Tax Forfeiture and Foreclosures, Mich. Taxes,,4676,7-238-43535_
55601—,00.html (last visited Jan. 19, 2018). See also James Alm et al., Detroit Property Tax Delinquency: Social Contract in Crisis, 14 Pub. Fin. & Mgmt. 280, 280–81 (2014) (concluding that properties with certain characteristics are more likely to be tax delinquent); Joshua M. Akers, Making Markets: Think Tank Legislation and Private Property in Detroit, 34 Urb. Geography 1070, 1070–93 (2013) (analyzing Detroit’s tax foreclosure crisis and the legislative process leading to it); Margaret Dewar et al., Disinvesting in the City: The Role of Tax Foreclosure in Detroit, 51 Urb. Aff. Rev. 587, 607 (2015) [hereinafter Dewar et al., Disinvesting]; Margaret E. Dewar, Selling Tax-Reverted Land: Lessons from Cleveland and Detroit, 72 J. Am. Plan. Ass’n 167, 169–71 (2006) [hereinafter Dewar, Lessons] (comparing Cleveland’s land-bank system for disposing of abandoned land with Detroit’s method); Ellen Kirtner, Interrupting the Blight Cycle: Managing the Future of Properties in Tax Foreclosure Sales Through Pre- and Post-Sale Interventions, 66 Case W. Res. L. Rev. 1083, 1097­–1100 (2016) (offering legal suggestions for local governments to put foreclosed properties to productive use).

 [10]. Detroit, Loveland, (last visited Jan. 19, 2018); Alex Alsup, A Recent History of Tax Foreclosure, Loveland (Nov. 9, 2015), See also Archival Tax Foreclosures in Detroit, 2002–2013, Data Driven Detroit, 9438afd734d348a694c42a28c4103731_0 (last visited Jan. 19, 2018).

 [11]. Alsup, supra note 10.

 [12]. Id.

 [13]. These numbers were calculated by dividing the number of tax-foreclosed properties listed in foreclosure sales in 2015 by each region’s 2010 census population and multiplying by 100,000. New York City had a 2010 population of 8,175,133; San Francisco, 805,704; Los Angeles County, 9,818,605; Erie County, 919,040; and St. Louis County, 998,954. Community Facts, U.S. Census Bureau: Am. Fact Finder, (last visited Jan. 19, 2018). In New York City, 4,228 property tax liens were sold in 2015. Number of Properties on City’s 90-Day Lien Sales List I              s Falling, Real Deal (Mar. 16, 2016, 10:32 A.M.), In San Francisco, 389 tax foreclosed properties were auctioned in 2015. City & Cty. of S.F., Public Auction Sales of Tax-Defaulted Property (2015), files/San%20Francisco%202015%20Tax%20Auction%20Post-Sale%20Report.pdf. In Los Angeles County, 781 tax foreclosed properties were auctioned in 2015. Cty. of L.A. Treasurer & Tax Collector Secured Prop. Div., 2015A Follow-up Online Auction (2015),; Cty. of L.A. Treasurer & Tax Collector Secured Prop. Div., 2015A Public Auction (2015),
proptax/docs/2015APublicSoldParcelsFinallist.pdf; Cty. of L.A. Treasurer & Tax Collector Secured Prop. Div., 2015B Online Auction (2015),
2015BSoldParcelsFinallist.pdf. In Erie County, 569 properties went to the 2015 tax foreclosure auction. Auction & Foreclosure Information, Real Prop. Tax Services,
20150816075239/ (last visited Jan. 19, 2018). In St. Louis, 1,964 properties were listed in 2015 tax sales. St. Louis Cty. Dep’t of Revenue, 8/24/15 1st 2nd Sale (2015),

 [14]. See Bernadette Atuahene, Takings as a Sociolegal Concept: An Interdisciplinary Examination of Involuntary Property Loss, 12 Ann. Rev. L. & Soc. Sci. 171, 172 (2016); Christopher E. Herbert et al., Is Homeownership Still an Effective Means of Building Wealth for Low-Income and Minority Households? (Was it Ever?) 2, 4, 27 (Harvard Univ. Joint Ctr. for Hous. Studies, Working Paper No. HBTL-06, 2013), (finding that home ownership continues to contribute significantly to household wealth and reporting that homeownership accounted for 22% of average wealth among homeowners in 2010).

 [15]. Margaret Jane Radin, Property and Personhood, 34 Stan. L. Rev. 957, 957 (1982) (“[T]he relationship between property and personhood . . . has commonly been both ignored and taken for granted in legal thought. The premise underlying the personhood perspective is that to achieve proper self-development—to be a person—an individual needs some control over resources in the external environment. The necessary assurances of control take the form of property rights.”).

 [16]. See Joseph William Singer, Property and Social Relations: From Title to Entitlement, in Property and Values: Striking an Equitable Balance of Public and Private Interests 3, 13 (Charles Geisler & Gail Daneker eds., 2000) (“Property law helps to structure and shape the contours of social relationships. Choices of property rules ineluctably entail choices about the quality and character of human relationships and myriad choices about the kind of society we will collectively create.”). 

 [17]. Through in-depth interviews with people who lost their homes and entire communities as a result of urban renewal during the 1950s and 1960s, Mindy Fullilove—a board-certified psychiatrist and public health professor at Columbia University—found that these displaced populations suffered from what she calls “root shock,” “the traumatic stress reaction to the destruction of all or part of one’s emotional ecosystem.” Mindy Thompson Fullilove, Root Shock: How Tearing Up City Neighborhoods Hurts America, and What We Can Do About It 11 (2004).

 [18]. See Charles A. Reich, The New Property, 73 Yale L.J. 733, 771, 787 (1964) (“[P]roperty performs the function of maintaining independence, dignity and pluralism in society by creating zones within which the majority has to yield to the owner. Whim, caprice, irrationality and ‘antisocial’ activities are given the protection of law; the owner may do what all or most of his neighbors decry. The Bill of Rights also serves this function, but while the Bill of Rights comes into play only at extraordinary moments of conflict or crisis, property affords day-to-day protection in the ordinary affairs of life.”).

 [19]. Radin, supra note 15.

 [20]. William M. Doerner & Keith R. Ihlanfeldt, An Empirical Analysis of the Property Tax Appeals Process, 10 J. Prop. Tax Assessment & Admin. 5, 16–17, 26 (2014) (discussing the fact that richer property holders have disproportionately more success with appeals than poorer and minority populations.); William M. Doerner & Keith R. Ihlanfeldt, The Role of Representative Agents in the Property Tax Appeals Process, 68 Nat’l Tax J. 59, 60–61, 89 (2015) [hereinafter Doerner & Ihlanfeldt, Representative Agents] (discussing the central role that unregulated appeals experts play, their focus on more affluent homeowners, and the resulting effects on existing disparities in property tax burdens).

 [21]. See generally Plaintiffs’ Expert Report, Morningside Cmty. Org. v. Sabree, No. 16-008807-CH (Mich. 3d Cir. Ct. 2016).

 [22]. Complaint at 2–3, Morningside Cmty. Org., No. 16-008807-CH; Race and Hispanic or Latino Origin 2010, supra Table 1 note g. In a follow-up paper, Bernadette Atuahene will test whether the assessors in Wayne County’s majority-white jurisdictions are illegally assessing residents to the same extent as the assessors in the county’s majority-African-American jurisdictions.

 [23]. Complaint at 46, Morningside Cmty. Org., No. 16-008807-CH.

 [24]. See 42 U.S.C. §§ 3601–3619 (2012). See generally Bernadette Atuahene, Assessing Racial Discrimination: Property Tax Assessments and the Fair Housing Act, 112 Nw. L. Rev. (forthcoming 2018).

 [25]. Plaintiffs sued both the City of Detroit and Wayne County. Litigation against Detroit is ongoing. But, the lower courts dismissed the litigation against Wayne County because they claimed that the case should have been brought before the Michigan Tax Tribunal. Plaintiffs have filed an interlocutor appeal to the Michigan Supreme Court. See Plaintiffs’ Application to File for Leave to Appeal, Morningside Cmty. Org v. Sabree, No. 156707 (Mich. Nov. 1, 2017).

 [26]. See Kenneth K. Baar, Property Tax Assessment Discrimination Against Low-Income Neighborhoods, 13 Urb. Law. 333, 338–56 (1981) (providing an overview of discriminatory property assessments); Harris, supra note 2, at 4 (finding that “residents of majority-minority neighborhoods are assessed at higher effective rates than residents of majority-white neighborhoods”); William S. Hendon, Discrimination Against Negro Homeowners in Property Tax Assessment, 27 Am. J. Econ. & Soc. 125, 125–32 (1968) (concluding that property assessments are higher for black homeowners than for white homeowners). See generally Andrew W. Kahrl, The Power to Destroy: Discriminatory Property Assessments and the Struggle for Tax Justice in Mississippi, 82 J. Southern Hist. 579 (2016) (discussing the history of discriminatory tax assessments in the American South through the late nineteenth and early twentieth centuries).

 [27]. Kahrl, supra note 26, at 585, 587–89 (“In 1916 the average assessment on white-owned farmland in Virginia, for example, was 33.1 percent of its market value; for black-owned land, it was 45.3 percent of market value.”).

 [28]. Id.

 [29]. See Lee Harris, Assessing Discrimination: The Influence of Race in Residential Property Tax Assessments, 20 J. Land Use & Envtl. L. 1, 3–4 (2004).

 [30]. Id.

 [31]. In January 2017, the City of Detroit completed a long overdue citywide re-assessment of all residential properties. Nevertheless, our analysis shows that 90 percent or more of properties valued less than $18,500 were still assessed in violation of the Michigan Constitution. See Bernadette Atuahene, Detroit’s Homeowners Deserve Better, Detroit News (Jan. 31, 2017),

 [32]. See Michigan Taxpayer’s Guide, supra note 7, at 1; William T. Dust, Manual of the Common Council and of the Municipal Government of the City of Detroit 29 (1886).

 [33]. Mich. Const. art. IX, § 3.

 [34]. Id. (“For taxes levied in 1995 and each year thereafter, the legislature shall provide that the taxable value of each parcel of property adjusted for additions and losses, shall not increase each year by more than the increase in the immediately preceding year in the general price level, as defined in section 33 of this article, or 5 percent, whichever is less until ownership of the parcel of property is transferred. When ownership of the parcel of property is transferred as defined by law, the parcel shall be assessed at the applicable proportion of current true cash value.”).

 [35]. Mich. Const. art. IX, §§ 26–31. See also Kevin C. Kennedy, The First Twenty Years of the Headlee Amendment, 76 U. Det. Mercy L. Rev. 1031, 1031–32 (1999) (evaluating the Headlee Amendment’s four core provisions: (1) that state, property, and other local taxes could not increase beyond the guidelines set out in the Amendment without voter approval; (2) that the state could not require new expenditures by local governments without providing financing; (3) that the state could not reduce the proportion of state spending used to assist local governments; and (4) that the state could not shift the tax burden to local governments).

 [36]. Mich. Comp. Laws § 211.27(a)(1) (2017).

 [37]. Id. § 211.27.

 [38]. See Meadowlanes Ltd. Dividend Hous. Ass’n v. City of Holland, 473 N.W.2d 636, 642 n.17 (Mich. 1991); C.A.F. Inv. Co. v. Mich. State Tax Comm’n, 221 N.W.2d 588, 592 (Mich. 1974). See also Great Lakes Div. of Nat’l Steel Corp. v. City of Ecorse, 576 N.W.2d 667, 672 (Mich. Ct. App. 1998) (“True cash value is synonymous with fair market value.”); Samonek v. Norvall Twp., 527 N.W.2d 24, 26 (Mich. Ct. App. 1994); Jones & Laughlin Steel Corp. v. City of Warren, 483 N.W.2d 416, 419 (Mich. Ct. App. 1992).

 [39]. See Meadowlanes, 473 N.W.2d at 642; Cty. of Washtenaw v. State Tax Comm’n, 373 N.W.2d 697, 705 (Mich. 1985); Antisdale v. City of Galesburg, 362 N.W.2d 632, 636–37 (Mich. 1984); President Inn Props., LLC v. City of Grand Rapids, 806 N.W.2d 342, 347 (Mich. Ct. App. 2011); Great Lakes Div., 576 N.W.2d at 678; Samonek, 527 N.W.2d at 26–27; Jones & Laughlin, 483 N.W.2d at 419; Teledyne Cont’l Motors v. Muskegon Twp., 378 N.W.2d 590, 593 (Mich. Ct. App. 1985); Northwood Apartments v. City of Royal Oak, 295 N.W.2d 639, 642–43 (Mich. Ct. App. 1980); Safran Printing Co. v. City of Detroit, 276 N.W.2d 602, 603 (Mich. Ct. App. 1979); Consumers Power Co. v. Big Prairie Twp., 265 N.W.2d 182, 188 (Mich. Ct. App. 1978).

 [40]. Antisdale, 362 N.W.2d at 636.

 [41]. C.A.F. Inv., 221 N.W.2d at 592 n.2.

 [42]. Meadowlanes, 473 N.W.2d at 642; Antisdale, 362 N.W.2d at 636–37 (citations omitted) (“Generally, there presently are three methods of valuation which are acceptable to the Michigan Tax Tribunal and the courts. They are the cost-less-depreciation approach, the capitalization-of-income approach, and the market approach.”); Great Lakes Div., 576 N.W.2d at 673; Samonek, 527 N.W.2d at 26; Jones & Laughlin, 483 N.W.2d at 419; Wolverine Tower Assocs. v. City of Ann Arbor, 293 N.W.2d 669, 671 (Mich. Ct. App. 1980).

 [43]. Meadowlanes, 473 N.W.2d at 643. See also Jones & Laughlin, 483 N.W.2d at 419.

 [44]. Int’l Assoc. of Assessing Officers, Standard on Mass Appraisal of Real Property 9 (2013).

 [45]. See Meadowlanes, 473 N.W.2d at 642, 651. See also Mich. Comp. Laws § 211.27(1) (2017); Great Lakes Div., 576 N.W.2d at 674, 678­–79. 

 [46]. Antisdale, 362 N.W.2d at 638 (“The most obvious deficiency in using the sales price of a piece of property as conclusive evidence of its value is that the ultimate sales price of the property, as a result of many factors, personal to the parties or otherwise, might not be its ‘usual’ price.”). See also Cleveland-Cliffs Iron Co. v. Republic Twp., 163 N.W. 90, 93 (Mich. 1917); Great Lakes Div., 576 N.W.2d at 679; Samonek, 527 N.W.2d at 26.

 [47]. Antisdale, 362 N.W.2d at 638 (“The market approach to value has the capacity to cure this deficiency because evidence of the sales prices of a number of comparable properties, if sufficiently similar, supports the conclusion that factors extrinsic to the properties have not entered into the value placed on the properties by the parties.”).

 [48]. Mich. Comp. Laws § 211.27(1); Antisdale, 362 N.W.2d at 637; Jones & Laughlin, 483 N.W.2d at 419 (“The market approach is the only valuation method that directly reflects the balance of supply and demand for property in marketplace trading.”); Teledyne Cont’l Motors v. Muskegon Twp., 378 N.W.2d 590, 593. See also Olson v. United States, 292 U.S. 246, 257 (1934) (citation omitted) (“[T]he market value must be estimated. In respect of each item of property that value may be deemed to be the sum which, considering all the circumstances, could have been obtained for it; that is, the amount that in all probability would have been arrived at by fair negotiations between an owner willing to sell and a purchaser desiring to buy. In making that estimate there should be taken into account all considerations that fairly might be brought forward and reasonably be given substantial weight in such bargaining.”).

 [49]. Mich. Comp. Laws § 211.27(1) (“As used in this act, ‘true cash value’ means the usual selling price at the place where the property to which the term is applied is at the time of assessment, being the price that could be obtained for the property at private sale, and not at auction sale except as otherwise provided in this section, or at forced sale. The usual selling price may include sales at public auction held by a nongovernmental agency or person if those sales have become a common method of acquisition in the jurisdiction for the class of property being valued.”); Samonek, 527 N.W.2d at 27.

 [50]. See Jones & Laughlin, 483 N.W.2d at 419 (“[E]vidence of the price at which an item of property actually sold is most certainly relevant evidence of its value at an earlier time . . . .”).

 [51]. See supra note 1.

 [52]. See La. Const. art. VII, § 18B (“The classifications of property subject to ad valorem taxation and the percentage of fair market value applicable to each classification for the purpose of determining assessed valuation are as follows: 1. Land 10%; 2. Improvements for residential purposes 10%; 3. Electric cooperative properties, excluding land 15%; 4. Public service properties; excluding land 25%; 5. Other property 15%.”); Miss. Const. art. IV, § 112 (The assessed value of property shall be a percentage of its true value, which shall be known as its assessment ratio. The assessment ratio on each class of property as defined herein shall be uniform throughout the state upon the same class of property, provided that the assessment ratio of any one (1) class of property shall not be more than three (3) times the assessment ratio on any other class of property. For purposes of assessment for ad valorem taxes, taxable property shall be . . . assessed at a percentage of its true value as follows: Class I. Single-family, owner-occupied, residential real property, at ten percent (10%) of true value; Class II. All other real property, except for real property included in Class I or IV, at fifteen percent (15%) of true value; Class III. Personal property, except for motor vehicles and for personal property included in Class IV, at fifteen percent (15%) of true value; Class IV. Public utility property, which is property owned or used by public service corporations required by general laws to be appraised and assessed by the state or the county, excluding railroad and airline property and motor vehicles, at thirty percent (30%) of true value . . . .”); N.M. Const. art. VIII, § 1A (“Except as provided in Subsection B of this section, taxes levied upon tangible property shall be in proportion to the value thereof, and taxes shall be equal and uniform upon subjects of taxation of the same class. Different methods may be provided by law to determine value of different kinds of property, but the percentage of value against which tax rates are assessed shall not exceed thirty-three and one-third percent.”); Okla. Const. art. X, § 8 (“Real property shall not be assessed for ad valorem taxation at a value less than eleven percent (11%) nor greater than thirteen and one-half percent (13.5%) of its fair cash value for the highest and best use for which such property was actually used . . . .”); Or. Const. art. XI, § 11(1) (“For the tax year beginning July 1, 1997, each unit of property in this state shall have a maximum assessed value for ad valorem property tax purposes that does not exceed the property’s real market value for the tax year beginning July 1, 1995, reduced by 10 percent.”); Pa. Const. art. VIII, § 1 (“All taxes shall be uniform, upon the same class of subjects, within the territorial limits of the authority levying the tax, and shall be levied and collected under general laws.”); Tenn. Const. art. II, § 28 (“Real property shall be classified into four (4) subclassifications and assessed as follows: (a) Public Utility Property, to be assessed at fifty-five (55%) percent of its value; (b) Industrial and Commercial Property, to be assessed at forty (40%) percent of its value; (c) Residential Property, to be assessed at twenty-five (25%) percent of its value, provided that residential property containing two (2) or more rental units is hereby defined as industrial and commercial property; and (d) Farm Property, to be assessed at twenty-five (25%) percent of its value.”); Utah Const. art. XIII, § 2 (“[A]ll tangible property in the State that is not exempt under the laws of the United States or under this Constitution shall be: (a) assessed at a uniform and equal rate in proportion to its fair market value, to be ascertained as provided by law . . . .”).

 [53]. James K. Pollock, Making Michigan’s New Constitution, 1961–1962, at 49 (1962).

 [54]. Mich. Const. art. IX, § 3.

 [55]. Mich. Const. of 1908, art. X, § 7 (“All assessments hereafter authorized shall be on property at its cash value.”). According to a preparatory report, “[t]his provision was carried over from Section 12 of Article XIV of the Constitution of 1850.” Raleigh Barlowe, Constitutional Convention Preparatory Comm’n, Taxation and Fiscal Policy in the Michigan Constitution 5 (1961).

 [56]. Barlowe, supra note 55, at 5.

 [57]. Letter from John B. Swainson, Governor of Mich., to Stephen S. Nisbit, President, Mich. Constitutional Convention (Jan. 31, 1962), in State of Mich., Journal of the Constitutional Convention, No. 70, at 442, 442 (1962) (“It was the unanimous finding of the Committee that an unhealthy lack of uniformity in property taxation exists in Michigan and the primary basis for this condition is the wide variation of assessments. . . . [I]t is impossible under these conditions to assure an industry seeking to settle in this state that it will be treated equitably in the matter of property taxes. The training, technique, and skill of the assessor are so varied in this state that in some instances it has been a deterrent to new industry. Dr. Harvey E. Brazer, of the University of Michigan, an eminent authority in this field, has stated that a study of a typical Michigan county revealed that some properties were assessed at less than 2% of market value, and others were assessed as high as 175% of their sales price. In one of our large metropolitan counties we have witnessed the practice of assessing inventories at 80% of book value, machinery and equipment at 100% of depreciated value, and real property assessed at 50% of current value.”).

 [58]. The 50 percent limit was introduced at the convention by the Committee on Finance and Taxation on January 31, 1962. Committee Proposal No. 51, State of Mich., Journal of the Constitutional Convention, No. 69, at 404, 404 (1962).

 [59]. Id. at 405.

 [60]. Id. at 404 (“The Legislature shall provide by law a uniform rule governing the general ad valorem taxation of real property and tangible personal property. The Legislature shall provide by law for the determination of true cash value of such property and shall specify the proportion of true cash value at which such property shall be uniformly assessed, which shall not, after January 1, 1966, exceed 50 per cent, and shall provide by law for a system of equalization of assessments.”).

 [61]. Id. at 405.

 [62]. Id.

 [63]. Minority Report on Committee Proposal No. 51, State of Mich., Journal of the Constitutional Convention, No. 73, at 523, 523 (1962).

 [64]. Id.

 [65]. Id.

 [66]. Id. at 522–23.

 [67]. Id. at 523.

 [68]. On February 8, 1962, the convention as a whole voted to send the provision to the Committee on Style and Drafting. Delegate Chard Austin unsuccessfully proposed to remove the 50 percent language on that day: “Amend page 1, line 13, after ‘assessed’ by striking out the comma and ‘which shall not, after January 1, 1966, exceed 50 per cent.’” Amendment No. 2, State of Mich., Journal of the Constitutional Convention, No. 75, at 557, 557 (1962).

 [69]. Lincoln Inst. of Land Policy & Minn. Ctr. for Fiscal Excellence, 50-State Property Tax Comparison Study for Taxes Paid in 2016, at 2, 18 fig.2 (2017),

 [70]. See, e.g., Gary Sands et al., Implementing Tax Abatements in Michigan: A Study of Best Practices, 20 Econ. Dev. Q. 44, 46–48 (2006); Gary Sands & Mark Skidmore, Making Ends Meet: Options for Property Tax Reform in Detroit, 36 J. Urb. Aff. 682, 682–83, 691–98 (2013) (offering reform options for current property tax policies that do not raise sufficient revenues for the City of Detroit); Michael Simoni, Tuning up the Motor City: The Viability of Restructuring Detroit’s Oppressive Property Tax System Within the Boundaries of Michigan’s Constitution, 51 Wayne L. Rev. 1309, 1320–23 (2005) (proposing a new reading of Proposal A, a ballot initiative to cap the maximum assessment). See generally Samuel J. McKim III, Is Michigan’s Ad Valorem Property Tax Becoming Obsolete?, 77 U. Det. Mercy L. Rev. 655 (2000); Timothy R. Hodge et al., The Land Value Gradient in a (Nearly) Collapsed Urban Real Estate Market (Lincoln Inst. of Land Policy, Working Paper No. WP15TH1, 2015) (arguing that peripheral neighborhoods are important to the property tax base).

 [71]. See supra note 5 and accompanying text.

 [72]. See id. Even when examining the same years as Hodge et al., however, we do not expect our numbers to be identical because Hodge et al. transform their collected assessed values by multiplying each value by two. Therefore, they expected their ratios to equal 1 while we expect ours to equal 0.5. See Hodge et al., Assessment Inequity, supra note 5, at 245 n.11.

 [73]. The PRD for Hodge et al.’s full sample was 2.68, exceeding the International Association of Assessing Officers’ (“IAAO”) upper limit of 1.03 and indicating regressivity. The COD was 109.55, again exceeding the IAAO’s acceptable range of 5 to 15 and indicating very low uniformity. Each quintile displayed horizontal inequity with mean values greater than 0.5, as well as regressivity and vertical inequity as lower value properties were more over-assessed (on average) than higher value properties. Hodge et al., Assessment Inequity, supra note 5, at 247 n.13. See also infra note 85 and accompanying text.

 [74]. See Stabilizing Morningside 47–49 & fig.3.8 (2015), sites/default/files/files/mup/capstones/2015_Stabilizing-MorningSide.pdf.

 [75]. Id.

 [76]. Id at 48–49 & figs.3.9 & 3.10. The study does not include any of the IAAO recommended statistics such as PRD and COD.

 [77]. After initial studies from Hodge et al. and the media attention their work received, the State Tax Commission intervened in 2013 and issued a sample of reassessments in Detroit. See Christine MacDonald, Detroit’s Property Tax System Plagued by Mistakes, Waste, Detroit News, Feb. 22, 2013, at A1; Christine MacDonald, Michigan’s Tax Board to Investigate Whether Detroit is Overtaxing Property Owners, Detroit News, Apr. 8, 2013, at A1. Shortly after these articles went to print, the City of Detroit announced it would reassess all property. The reassessments were completed in 2017. Christine Ferretti, Property Taxes Going down for Over Half of Detroiters, Detroit News (Jan. 23, 2017, 11:05 A.M.), (“The Duggan administration unveiled the proposed 2017 property assessments on the heels of the first parcel-by-parcel reappraisal of the city’s nearly 255,000 residential properties in 60 years.”).

 [78]. Matt Helms, Detroiters to See Property Assessment up to 20% Lower, Detroit Free Press (Jan. 28, 2015, 11:39 A.M.),
detroit-property-tax-assessments-reduced/22464209; Christine MacDonald & Christine Ferretti, Detroit Assessments to Fall 5–20 Percent, Detroit News (Jan. 28, 2015, 12:38 A.M.), (“Detroit Mayor Mike Duggan announced Wednesday that residential property assessment citywide will decline 5 to 20 percent, the second consecutive year he’s cut taxes. The change is in addition to last year’s assessment reductions of up to 20 percent.”).

 [79]. The IAAO is a nonprofit, educational, and research association that is comprised of government assessment officials and others interested in the administration of the property tax. The IAAO’s main role is to promote standardized practices for assessing property and monitoring assessment performance. About Home, IAAO, Content/About_Home.aspx (last visited Jan. 21, 2018).

 [80]. See Int’l Ass’n of Assessing Officers, Standard on Ratio Study 7 (2013) [hereinafter IAAO].

 [81]. Id.

 [82]. The mean is calculated by adding all assessment ratios together and dividing by the total number of ratios. The median is the middle ratio in an uneven number of ratios ordered by magnitude, or the average of the two central ratios in an even number of ordered ratios. The median is less affected by extreme outliers than the mean. Id. at 13.

 [83]. Mich. Const. art. IX, § 3.

 [84]. See IAAO, supra note 80, at 13–14, 28–29; Hodge et al., Assessment Inequity, supra note 5, at 241.

 [85]. We recognize other IAAO-recommended measures for analyzing assessments exist, including price-related bias, weighted mean, median absolute deviation, and regression analysis. See IAAO, supra note 80, at 12–14. However, we have presented the traditional assessment performance statistics highlighted in the IAAO’s publication, id., and used in other papers analyzing assessment ratios. See, e.g., Hodge et al, Assessment Inequity, supra note 5, at 241; Krupa, supra note 2, at 565–69; Daniel P. McMillen, Assessment Regressivity: A Tale of Two Illinois Counties, Land Lines, January 2011, at 10–12.

 [86]. As outlined in the IAAO’s publication for analyzing assessment accuracy, the COD is calculated as follows: “1. subtract the median from each ratio; 2. take the absolute value of the calculated differences; 3. sum the absolute differences; 4. divide by the number of ratios to obtain the average absolute deviation; 5. divide by the median; and 6. multiply by 100.” IAAO, supra note 80, at 13 (emphasis omitted).

 [87]. Id. at 17 tbl.1-3.

 [88]. Id. at 14–15.

 [89]. While the median and mean give equal weight to each parcel in the sample, the weighted mean gives equal weight to each dollar. Id. at 13.

 [90]. Id. at 14, 17 tbl.1-3.

 [91]. Quintiles are determined by dividing the data into five equal groups according to a variable (e.g. sale prices). See Hodge et al, Assessing Inequity, supra note 5, at 247 n.13.

 [92]. We also recognize non-descriptive techniques to analyze vertical and horizontal inequity exist. Particularly, regression analysis has been the standard method for deriving causal inference with the inclusion of various control variables. For studies using alternate regression specifications for examining vertical inequity, see generally Earl J. Bell, Administrative Inequity and Property Assessment: The Case for the Traditional Approach, 3 Prop. Tax J. 123 (1984); Pao Lun Cheng, Property Taxation, Assessment Performance, and Its Measurement, 29 Pub. Fin. 269 (1974); John M. Clapp, A New Test for Equitable Real Estate Tax Assessment, 3 J. Real Est. Fin. & Econ. 233 (1990); Levis A. Kochin & Richard W. Parks, Vertical Equity in Real Estate Assessment: A Fair Appraisal, 20 Econ. Inquiry 511 (1982); Morton Paglin & Michael Fogarty, Equity and the Property Tax: A New Conceptual Focus, 25 Nat’l Tax J. 557 (1972); Ronald W. Spahr & Mark A. Sunderman, Property Tax Inequities on Ranch and Farm Properties, 74 Land Econ. 374 (1998). For studies using alternate regressions with emphasis on horizontal inequity, see generally Marcus T. Allen & William H. Dare, Identifying Determinants of Horizontal Property Tax Inequity: Evidence from Florida, 24 J. Real Est. Res. 153 (2002); Brian J.L. Berry & Robert S. Bednarz, A Hedonic Model of Prices and Assessments for Single-Family Homes: Does the Assessor Follow the Market or the Market Follow the Assessor?, 51 Land Econ. 21 (1975); William C. Goolsby, Assessment Error in the Valuation of Owner-Occupied Housing, 13 J. Real Est. Res. 33 (1997). Although these regression techniques are not the focus of this paper, since we examine assessments through the standards set forth by the IAAO, we have analyzed each technique using various functional forms. Results are available upon request. Our findings are consistent with the results presented in this paper: vertical and horizontal inequity is present as lower-value properties are consistently over-assessed.

 [93]. IAAO, supra note 80, at 7.

 [94]. Id. at 8.

 [95]. Data Driven Detroit (D3) is a Low-Profit Limited Liability Company (L3C) focused on providing access to information about and analyses for Detroit and the surrounding region. See About, Data Driven Detroit, (last visited Jan. 21, 2018).

 [96]. Detroit’s Open Data Portal provides access to public data and information about city governments and service delivery. See About Detroit Open Data, City Detroit, (last visited Jan. 21, 2018). The Portal’s data tables are continually updated; data as used in this Article is on file with the authors.

 [97]. Non-arm’s length transactions excluded from our study comprise the following sales terms: “bank sale,” “city sale,” “county land bank sale,” “court order,” “government transfer,” “HUD,” “land bank,” “land contract,” “create trust,” “multiple parcels,” “NEZ,” “NQ,” “death cert/estate,” “deed in lieu of FCL,” “deed of personal rep,” “family sale,” “review needed,” “sheriff/tax deed,” “investor sale,” “to bank,” and “to investor.”

 [98]. Mich. Comp. Laws § 211.27(1) (2017).

 [99]. The estimates of illegality increase substantially when non-arm’s length transactions are included. For instance, the median assessment ratio when only arm’s length transactions are included versus when non-arm’s length transactions are included: 2009 (0.79 vs. 2.69), 2010 (1.95 vs. 4.13), 2011 (0.52 vs. 2.79), 2012 (0.69 vs. 3.85), 2013 (0.91 vs. 5.41), 2014 (1.04 vs. 2.88), 2015 (.69 vs. 2.32). Mean assessment ratios are even more severe when non-arm’s length transactions are included, ranging from 2,100 to 4,300 vs. 0.53 to 3.44. See infra Table 3.

 [100]. It may seem reasonable to assign a portion of the total price to each property based on the parcel’s size and characteristics; however, unobserved factors may prevent this split from accurately representing each property. For example, a parcel with a larger house and lot may be expected to have a higher value, but the house may be in worse condition than a smaller property it is bundled with, and thus should not automatically be granted a higher price.

 [101]. The IAAO defines an outlier as an assessment ratio below the first quartile or above the third quartile by 1.5 times the interquartile range (IQR), where the IQR is the difference between the first and third quartiles. The first quartile is the median value of the lower half of the data and the third quartile is the median value of the upper half of the data (the second quartile is the median of the entire dataset). The IAAO also recognizes that ratio distributions are often skewed to the right (i.e., a greater number of high ratios may be present), so to prevent dropping a disproportionate number of high ratios, the IAAO suggests taking the logarithm of each assessment ratio prior to trimming the outliers. IAAO, supra note 80, at 12, 53. The steps for trimming outliers are: (1) locate the first quartile (i.e., median value of the lower half of the data); (2) locate the third quartile (i.e., median value of the upper half of the data); (3) compute the interquartile range (IQR) and multiply by 1.5 = (Step 2 – Step 1) x 1.5; (4) establish the lower boundary = (Step 1 – Step 3); (5) establish the upper boundary = (Step 2 + Step 3). Id. Again, our results represent conservative estimates because they only increase when we include non-trimmed transactions. Here are the mean assessment ratios when only trimmed arm’s length transactions are included vs. when non-trimmed arm’s length transactions are included: 2009 (2.52 vs. 3.52), 2010 (3.44 vs. 5.59), 2011 (0.53 vs. 40.20), 2013 (1.57 vs. 1.65), 2014 (1.73 vs. 4.18), 2015 (1.28 vs. 60.74).

 [102]. See id. at 54.

 [103]. See supra Part II.

 [104]. See supra notes 86 & 89 and accompanying text (explaining the process for calculating COD and PRD).

 [105]. See infra Figure 1.

 [106]. Other years experienced increases in mean assessment ratios when “Review Needed” sales were included. However, they were not as drastic as the increases in 2011 and 2012.

 [107]. See supra notes 7778 and accompanying text (regarding Detroit’s tax cuts).

 [108]. The reduction is calculated by using the percentage change formula: [(new value–old value)/old value]*100. In this case, the old value is 1.73 and the new value is 1.28. Therefore, [(1.28–1.73)/1.73]*100 = [-0.45/1.73]*100 = [-0.26*100] = 26% decrease.

 [109]. See supra note 106 and accompanying text (suggesting that Detroit assessor’s office likely miscategorized several arm’s length transactions in 2011 and 2012 as “Review Needed”). Only in 2012 are the highest-valued properties (quintile 5) above 0.5.

 [110]. The decrease is more than the 20 percent decrease that Duggan authorized, in part due to our sample, which captures only properties sold (a much smaller subset of the total number of residential properties in each district).

 [111]. See Mike Duggan, Mayor of Detroit, Address at a City of Detroit Press Conference (Jan. 28, 2015) (“As I said when I was campaigning, I felt like the assessments in this city were higher than the actual sales price that people could sell their house for. We made a commitment to have honest assessments. Last year, you remember most Detroiters got assessment cuts between 10 and 20 percent and we have embarked on a parcel by parcel assessment—the first one in fifty years—and that’s going to be done at the end of 2016. We are assessing each individual home. We didn’t want to wait for that parcel to parcel assessment to be done to continue to try to do the best we can to make sure our assessments are accurate. And so we are again doing significant assessment cuts for 2015.”).

 [112]. See supra note 106 and accompanying text.

 [113]. Among the stated goals of the assessor for the 2010–2011 fiscal year was to “assess at statutory level of 50% all properties within the city of Detroit, . . . [c]onduct site reviews of the required 30% of all property annually; [and] field review, capture, and correctly value all real and personal property within the jurisdiction . . . .” City of Detroit, Office of the Auditor Gen., Performance Audit of the Finance Department Assessments Division July 2008–June 2011, at 6–7 (2012),

 [114]. See id. at 9 (“In the City’s electronic assessing system known as ‘Equalizer,’ a property is in an ‘override status’ when its assessed value is input as a total amount, versus the system method of calculating a value based on physical property attributes and other assessment criteria. The property’s assessed value is ‘disconnected’ in the system. Assessed values in Equalizer are historical aggregate amounts, which were transferred from the previous assessing system known as ‘IPDS’ (Integrated Physical Data Systems): Of the 42 properties audited, 28 (66.7%) remain in override status; A representative in the Assessments Division estimated that 92% of the City’s parcels 387,000 [sic] remain in override status in Equalizer.”).

 [115]. See Doerner and Ihlanfeldt, Representative Agents, supra note 20, at 60–61, 89. See also Andrew T. Hayashi, The Legal Salience of Taxation, 81 U. Chi. L. Rev. 1443, 1447 (2014) (“I find that reducing property tax salience makes homeowners less likely to appeal their property-value assessments, which makes it more likely that those homeowners will remain overassessed and overtaxed. These overtaxed homeowners never perceive their injury and consequently never seek the legal relief to which they might be entitled. I also find that property taxes are less salient for recent homeowners with higher-priced mortgages, who are more likely to be racial minorities, immigrants, and working families with children.”); Andrew T. Hayashi, Property Taxes and Their Limits: Evidence from New York City, 25 Stan. L. & Pol’y Rev. 33, 46 & tbl.1 (2014).


 [116]. City of Detroit, supra note 113, at 3.

 [117]. See id. at 18.

 [118]. Id. at 11.

 [119]. See supra note 35 and accompanying text.

 [120]. See Mich. Comp. Laws §§ 211.78(a), 211.87(b) (2017).

 [121]. Under Michigan’s Public Act 246 of 2003, the City of Detroit began participating in the Delinquent Tax Revolving Fund (“DTRF”) in the 2003 tax year. Major Activities and Descriptions, Wayne County Treasurer, (last visited Jan. 21, 2018). Fees and interest are deposited in the DTRF, but surpluses are under the control of the county board of commissions. See Mich. Comp. Laws § 211.87(b); Richard L. Baldermann, Mich. Dep’t of Treasury Numbered Ltr. 2001-5, Delinquent Tax Revolving Funds Revisions to Accounting After Public Act 123 of 1999 (Feb. 8, 2001),,1607,7-121-1751_2194-6024–,00.html. See also Mich. Ass’n of Cty. Treasurers, The Delinquent Tax Revolving Fund (2013),

 [122]. See, e.g., Charles E. Ramirez, Treasurer Could Add $82M to Wayne Co.’s General Fund, Detroit News (Oct. 13, 2015, 10:36 P.M.), (reporting that in 2015, Wayne County’s treasurer recommended transferring $82 million from the DTRF to its general fund); Robert Snell, Wayne County Poised to End Financial Emergency, Evans Says, Crain’s Detroit Bus. (Mar. 1, 2016), (“Wayne County could emerge from a financial emergency later this year after eliminating health care liabilities and an $82 million accumulated deficit, county Executive Warren Evans said Tuesday. . . . Evans said he will ask the state treasurer soon to release Wayne County from a consent agreement reached last year amid the county’s financial emergency.”).

 [123]. Searching Google Scholar and JSTOR, we found only three authors using this term in the last hundred years; none defined or developed it. See Thomas P.M. Barnett, Big-War Thinking in a Small-War Era: The Rise of the AirSea Battle Concept, 6 China Security 3, 4 (2010) (“And to the extent that America eschews such responsibilities, other rising powers seeking to protect their expanding network of economic interests will inevitably step into that void—albeit with less militarized delivery systems. China may do so, but, as is now becoming apparent, it prefers stategraft to nation-building, paying upfront from its sizeable cash coffers.”); Stephen S. Wise, The Return of Roosevelt, 191 N. Am. Rev. 738, 74142 (1910) (“When, for example, the corrupt followers of the Quay-Stone régime in the Keystone State desperately sought to avert the crushing defeat that even apathetic Pennsylvania seemed ready to inflict upon its betrayers, these masters of Stategraft rather than Statecraft brazenly utilized the name and prestige of Roosevelt in order to stave off the day of wrath.”); Art Buchwald, The Fine Points of Stategraft, er, Statecraft, Wash. Post, Feb. 25, 2003, at C3 (discussing U.S. foreign policy and using “stategraft” in title only).

 [124]. E.g., Susan Rose-Ackerman, Corruption and Government: Causes, Consequences, and Reform 91 (1999) (defining corruption as the “misuse of public power for private gain”); Transparency Int’l, The Anti-Corruption Plain Language Guide 14 (2009),; United Nations Handbook on Practical Anti-Corruption Measures for Prosecutors and Investigators 23 (2004), (“There is no comprehensive, and universally accepted definition of corruption. The origin of the word is from the Latin corruptus (spoiled) and corrumpere (to ruin; to break into pieces). The working definitions presently in vogue are variations of ‘the misuse of a public or private position for direct or indirect personal gain.’”); World Bank, Helping Countries Combat Corruption: The Role of the World Bank 89 (1997), (“We settled on a straightforward definition—the abuse of public office for private gain. Public office is abused for private gain when an official accepts, solicits, or extorts a bribe. It is also abused when private agents actively offer bribes to circumvent public policies and processes for competitive advantage and profit. Public office can also be abused for personal benefit even if no bribery occurs, through patronage and nepotism, the theft of state assets, or the diversion of state revenues.”); Oguzhan Dincer & Michael Johnston, Measuring Illegal and Legal Corruption in American States: Some Results from the Corruption in America Survey, Edmond J. Safra Ctr. for Ethics Harv U. (Dec. 1, 2014), (emphasis omitted) (“We define illegal corruption as the private gains in the form of cash or gifts by a government official, in exchange for providing specific benefits to private individuals or groups. It is the form of corruption that attracts a great deal of public attention. A second form of corruption, however, is becoming more and more common in the U.S.: legal corruption. We define legal corruption as the political gains in the form of campaign contributions or endorsements by a government official, in exchange for providing specific benefits to private individuals or groups, be it by explicit or implicit understanding.”).

 [125]. See David A. Baldwin, Economic Statecraft 8 (1985) (“Statecraft has traditionally been defined as the art of conducting state affairs.”); Roy Coleman, Steve Tombs & Dave Whyte, Capital, Crime Control and Statecraft in the Entrepreneurial City, 42 Urb. Stud. 2511, 2512 (2005) (“The definitional struggles that determine the trajectory of the governing process take place within the alliances between locally powerful agencies that—through the partnerships between city centre managers, chief executives, developers, local authorities, senior police and growth spokespeople—herald a form of neo-liberal statecraft.”); Robbie Waters Robichau, The Mosaic of Governance: Creating a Picture with Definitions, Theories, and Debates, 39 Pol’y Stud. J. 113, 115 (2011) (“Statecraft can be characterized as the ‘exercise of distinctively governmental responsibilities’ and as ‘the art of acting according to duty, justice, and reason on behalf of a community of citizens.’”).

 [126]. See Timothy Egan, The Nation; Mending a Trail of Broken Treaties, N.Y. Times: Wk. Rev. (June 25, 2000), (“The Oneidas, who once occupied six million acres in New York, treated with President Washington himself. But over the next century, state officials bought and sold their land, in violation of laws and court cases requiring federal approval of Indian land transfers. By 1919, the tribal land had shrunk to 32 acres. With recent gambling revenues, the Oneidas bought back 11,000 acres, and hired top legal talent to regain the 250,000 acres lost from the reservation. The 20,000 non-Indians who own some of that land say they should not have to pay the price for a failure by the state to follow the law—nor should they have to fight their own government to hold onto their cabins, farms and houses.”).

 [127]. See id. (“[T]he Black Hills of South Dakota . . . is the longest unresolved Indian land claim in the country. Promised to the Lakota Sioux in a treaty from 1868, the land was opened to white settlement after gold was discovered, leading to the Battle of the Little Big Horn. More than a century later, the Lakota won a settlement that would have given them $106 million in 1980. But for 20 years, the Indians living in the poorest county in America have refused it, holding out for the return of their sacred hills.”).

 [128]. See U.S. Dep’t of Justice Civil Rights Div., supra note 6, at 3 (“The municipal court does not act as a neutral arbiter of the law or a check on unlawful police conduct. Instead, the court primarily uses its judicial authority as the means to compel the payment of fines and fees that advance the City’s financial interests. This has led to court practices that violate the Fourteenth Amendment’s due process and equal protection requirements. The court’s practices also impose unnecessary harm, overwhelmingly on African-American individuals, and run counter to public safety.”); id. at 15 (“Ferguson’s strategy of revenue generation through policing has fostered practices in the two central parts of Ferguson’s law enforcement system—policing and the courts—that are themselves unconstitutional or that contribute to constitutional violations. In both parts of the system, these practices disproportionately harm African Americans. Further, the evidence indicates that this harm to African Americans stems, at least in part, from racial bias, including racial stereotyping. Ultimately, unlawful and harmful practices in policing and in the municipal court system erode police legitimacy and community trust, making policing in Ferguson less fair, less effective at promoting public safety, and less safe.”).

 [129]. See id. at 2–3 (“Police supervisors and leadership do too little to ensure that officers act in accordance with law and policy, and rarely respond meaningfully to civilian complaints of officer misconduct. The result is a pattern of stops without reasonable suspicion and arrests without probable cause in violation of the Fourth Amendment; infringement on free expression, as well as retaliation for protected expression, in violation of the First Amendment; and excessive force in violation of the Fourth Amendment.”).

 [130]. See id. at 3 (“Most strikingly, the court issues municipal arrest warrants not on the basis of public safety needs, but rather as a routine response to missed court appearances and required fine payments. In 2013 alone, the court issued over 9,000 warrants on cases stemming in large part from minor violations such as parking infractions, traffic tickets, or housing code violations. Jail time would be considered far too harsh a penalty for the great majority of these code violations, yet Ferguson’s municipal court routinely issues warrants for people to be arrested and incarcerated for failing to timely pay related fines and fees.”).

 [131]. See Sarah Stillman, Taken, New Yorker (Aug. 12 & 19, 2013),; Laura Sullivan, Police Can Seize and Sell Assets Even When the Owner Broke No Law, NPR (Nov. 10, 2014, 5:46 P.M.),
sections/thetwo-way/2014/11/10/363102433/police-can-seize-and-sell-assets-even-when-the-owner-broke-no-law (“It is legal for law enforcement agencies to take property from people who haven’t been convicted of a crime.”).

 [132]. See, e.g., Mobley v. City of Detroit, 938 F. Supp. 2d 669, 674–75, 686 (E.D. Mich. 2012) (holding that a 2008 police raid of the Contemporary Art Institute of Detroit, in which officers detained patrons’ cars despite having no probable cause, was unconstitutional); First Amended Complaint, Encarnacion v. City of New York, No. 1:16-CV-00156 (S.D.N.Y. June 3, 2016) (initiating a class action alleging that New York City’s civil forfeiture practices are unconstitutional); Michael Sallah, Feds Open Probe of Bal Harbour Police Money Laundering, Miami Herald (Oct. 25, 2015, 5:31 P.M.), (“In an earlier investigation three years ago, federal agents found Bal Harbour misspent money from seized cars and cash to pay for police salaries, but investigators never examined the money raked in during the laundering sting—at least $71.5 million—until now.”); Stillman, supra note 131 (telling the stories of several people who have had their property taken from them via civil forfeiture and noting, inter alia, that “[t]he Public Defender Service for the District of Columbia won the release of Moreira’s car last summer, and in May filed a lawsuit against the city on behalf of approximately three hundred and seventy-five car owners like Moreira. Describing the policy as ‘devastating for hundreds of families who depend on their cars for many of the urgent and important tasks of daily life,’ it called for higher standards of proof and the end of penal-sum fees.”).

 [133]. It can be extremely expensive to secure the return of property confiscated through civil forfeiture laws. See Stillman, supra note 131 (“Most of those detained had to pay more than a thousand dollars for the return of their cars; if the payment wasn’t made promptly, the car would become city property.”). See also Robert O’Harrow Jr. & Steven Rich, D.C. Police Plan for Future Seizure Proceeds Years in Advance in City Budget Documents, Wash. Post (Nov. 15, 2014),
1tVIzf5 (“Among other things, the Public Defender Service focused on a city requirement that vehicle owners post bonds of up to $2,500 before they were permitted to challenge seizures. In August 2013, all parties agreed to put the lawsuit on hold as the District worked to modify its forfeiture laws.”).

 [134]. Simms v. District of Columbia, 872 F. Supp. 2d 90, 105 (D.D.C. 2012) (citation omitted) (“[T]he District’s continued retention of his vehicle without the opportunity to be heard deprives him of his constitutionally-protected rights. ‘It is always in the public interest to prevent the violation of a party’s constitutional rights.’”).

 [135]. See id. at 91, 105–07.

 [136]. See O’Harrow & Rich, supra note 133 (“The council’s reform effort began last year after the Public Defender Service for the District filed a class action lawsuit against the city, alleging that police violated the constitutional rights of residents in the process of seizing their cars.”).

 [137]. See Steven Maynard-Moody & Michael Musheno, State Agent or Citizen Agent: Two Narratives of Discretion, 10 J. Pub. Admin. Res. & Theory 329, 337 (2000) (describing state agents as the state bureaucrats who are entrusted with actually enacting laws and policies). See also Claire O. Finkelstein, Changing Notions of State Agency in International Law: The Case of Paul Touvier, 30 Tex. Int’l L.J. 261, 270 (1995); Colin Yeo, Agents of the State: When Is an Official of the State an Agent of the State?, 14 Int’l J. Refugee L. 509, 517 (2002).

 [138]. See supra notes 12930 and accompanying text (describing the City of Ferguson’s discriminatory fees and fines).

 [139]. See supra note 126 and accompanying text.

 [140]. See supra notes 12930 and accompanying text.

 [141]. See supra notes 131–36 and accompanying text.

 [142]. See supra Part III.

 [143]. This includes both documented and undocumented citizens. The tax code’s definition of “resident” is also expansive. Harry A. Shannon, III, The General Definition of Residence Under United States Income Tax Treaties, 16 Intertax 204, 206 (1988) (“Under Article 4(1) [of the 1976 Treasury Department’s Model Income Tax Treaty], a person is a resident of a contracting state if he is subject to tax in that state on the basis of his domicile, residence, citizenship, place of management, place of incorporation or any other criterion of a similar nature. ‘Person’ is defined in Article 3(1)(a) to include an individual, a partnership, a company, an estate, a trust and any other body of persons. Both natural and juridical persons can be ‘residents’ for treaty purposes.”).

 [144]. See supra note 126.

 [145]. See Stillman, supra note 131 (“But civil-forfeiture statutes continued to proliferate, and at the state and local level controls have often been lax. Many states, facing fiscal crises, have expanded the reach of their forfeiture statutes, and made it easier for law enforcement to use the revenue however they see fit. In some Texas counties, nearly forty per cent of police budgets comes from forfeiture. (Only one state, North Carolina, bans the practice, requiring a criminal conviction before a person’s property can be seized.) Often, it’s hard for people to fight back. They are too poor; their immigration status is in question; they just can’t sustain the logistical burden of taking on unyielding bureaucracies.”).

 [146]. See Jeremy Waldron, The Concept and the Rule of Law, 43 Ga. L. Rev. 1, 6 (2008).

 [147]. See supra Part III.

 [148]. See supra text accompanying note 124.

 [149]. See United States v. Sioux Nation of Indians, 448 U.S. 371, 424 (1980) (“[T]he 1877 Act effected a taking of tribal property, property which had been set aside for the exclusive occupation of the Sioux by the Fort Laramie Treaty of 1868. That taking implied an obligation on the part of the Government to make just compensation to the Sioux Nation, and that obligation, including an award of interest, must now, at last, be paid.”).

 [150]. See I.R.C. § 7421(a) (2012) (prohibiting suits to restrain assessment or collection). See also Richard A. Epstein, Takings: Private Property and the Power of Eminent Domain 283 (1985) (arguing that taxes and all regulations should fall under the takings clause but acknowledging that “[t]he proposition that all taxes are subject to scrutiny under the eminent domain clause receives not a whisper of current support”); Lee Anne Fennel & Eduardo M. Peñalver, Exactions Creep, 2013 Sup. Ct. Rev. 287, 299 (“Courts and commentators alike have read Eastern Enterprises to mean that general obligations to pay money do not fall within the ambit of ‘private property’ protected by the Takings Clause.”); Eduardo M. Peñalver, Regulatory Taxings, 104 Colum. L. Rev. 2182, 2182 (2004) (“[T]he broad legal consensus is that legislatures effectively have unlimited authority to impose tax burdens.”).

 [151]. See, e.g., Bennis v. Michigan, 516 U.S. 442, 453 (1996) (citation omitted) (“We conclude today, as we concluded 75 years ago, that the cases authorizing actions of the kind at issue are ‘too firmly fixed in the punitive and remedial jurisprudence of the country to be now displaced.’ The State here sought to deter illegal activity that contributes to neighborhood deterioration and unsafe streets. The Bennis automobile, it is conceded, facilitated and was used in criminal activity. Both the trial court and the Michigan Supreme Court followed our longstanding practice, and the judgment of the Supreme Court of Michigan is therefore affirmed.”); AmeriSource Corp. v. United States, 525 F.3d 1149, 1150 (Fed. Cir. 2008) (holding that a government seizure of pharmaceutical drugs under its police power did not constitute a taking).

 [152]. U.S. Const. art. V (“[N]or shall any person . . . be deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation.”); id. amend. XIV (“[N]or shall any State deprive any person of life, liberty, or property, without due process of law . . . .”).

 [153]. See Alm et al., supra note 9, at 283 (concluding that properties with certain characteristics are more likely to be tax-delinquent); Dewar et al., Disinvesting, supra note 9, at 606–07 (arguing that tax foreclosure policies encouraged divestment, spread blight, and created negative externalities); Dewar, Lessons, supra note 9, at 168–73 (comparing Cleveland’s land bank system for disposing of abandoned land with Detroit’s method); Kirtner, supra note 9, at 1085 & n.9 (offering legal suggestions for local governments to put foreclosed properties to productive use).

 [154]. See generally Bernadette Atuahene & Christopher Berry, Taxed Out: Illegal Property Tax Assessments and the Epidemic of Tax Foreclosures in Detroit (Jan. 31, 2018) (unpublished manuscript) (on file with authors).