Township Tax Shelters: How Michigan’s Constitutional Property Tax Restrictions Subsidize Sprawl and Segregation

Jacob Whiton

Abstract: Over the last 40 years, the state of Michigan has adopted two new constitutional restrictions on local governments’ use of the property tax. Together, these two additional restrictions have made the burden of local property taxes in the state more regressive both within and across local jurisdictions. In the Detroit metro area, these policies have drawn middle and upper class residents from high tax cities to low tax suburban and exurban townships, widening property tax regressivity across the region. Together with procyclical cuts to the state’s statutory revenue sharing program, regional property tax regressivity has imposed a severe austerity on the city of Detroit and other high tax cities where the region’s Black working class remain segregated. The state’s cap on assessment growth in particular limited the city’s ability to raise revenue once home prices began to recover from the housing market crash of 2007. To offset this constraint, the city overassessed properties in its most devalued neighborhoods by an estimated $600 million between 2010 and 2016, resulting in the dispossession of thousands of working class homeowners through property tax foreclosure. Given the inequity inherent in Michigan’s constitutional property tax restrictions, I conclude by calling for a significant expansion of the state’s statutory revenue sharing program funded by an income surtax on millionaires. This policy reform would help to equalize fiscal capacity across local governments, reduce regressivity in the state’s tax burden, and enable low income cities to make urgently needed investments in vital public goods and services.

I. Introduction

Still grappling with persistently high unemployment and depressed home values, homeowners in the city of Detroit were collectively overcharged by an estimated $600 million in property taxes between 2010 and 2016 (MacDonald 2020a). Michigan’s state constitution caps property tax assessments at half a property’s true market value, but assessments in Detroit were not adjusted to fully reflect the devaluation of the housing stock after 2007 and many struggled to keep up with their inflated bills (Atuahene and Hodge 2018). As a result, Wayne County, in which Detroit is located, foreclosed on over 100,000 tax delinquent properties in these years, over a quarter of the city’s pre-crisis housing stock (Atuahene and Berry 2019). The burden of overassessment and foreclosure was not borne equally by all of the city’s homeowners. The city’s most valuable properties were assessed at well below the state’s constitutional limit, but among the bottom 70 percent of homes by sale price, the average ratio of assessed value to sales price exceeded the state constitutional limit. Among the bottom fifth of homes, average assessed value was nearly five times higher than the average sale price (Atuahene and Berry 2019).

As part of Detroit’s restructuring in the wake of its 2013 bankruptcy, the largest municipal bankruptcy in U.S. history, the State Tax Commission took control of the city assessor’s office and oversaw a citywide reappraisal that took four years to complete (MacDonald 2020a). While the reappraisal did reduce the share of homeowners paying inflated property taxes, subsequent research has found many homeowners are still being overcharged and that assessment regressivity has gotten worse (CMF 2020). The city continues to dispute these and earlier findings on the extent of overassessment and maintains that it has neither the legal authority nor the fiscal capacity to reimburse homeowners affected by inflated property tax bills (MacDonald 2020b; Stein 2022).

The existing literature on property tax inequity in the city of Detroit has primarily focused on the problem of overassessment and its role in exacerbating the city’s foreclosure crisis, which dispossessed thousands of residents of their homes and left the city with a significantly diminished tax base. In this paper, I situate the city’s protracted fiscal distress in a broader regional context and argue that Detroit’s inflated, regressive assessments are a symptom of state-level restrictions on local governments’ use of the property tax. Michigan is one of only thirteen states to cap local property tax rates, revenue growth, and assessment growth, and all three limits are enshrined in the state constitution (CRCM 2021). Over the last 40 years, these restrictions have widened regressivity in property tax rates across jurisdictions in the Detroit metropolitan region, encouraging affluent homeowners to flee high tax cities for low tax townships. Assessment limits have also widened regressivity within jurisdictions by holding down effective tax rates in neighborhoods with relatively fast-appreciating properties.

Both forms of regressivity have created a structural inequality between localities in their ability to raise revenue from local property and income taxes. Exacerbated by steep cuts to redistributive state transfers, this inequality has in turn compelled shrinking high poverty cities to slash spending, assume unsustainable debts, raise revenue from regressive fines and fees, and inflate residents’ property tax assessments (Atuahene 2020; Scharff 2021). Because metro Detroit is highly segregated by race (Menendian, Gambhir, and Gailes 2021), the burden of municipal austerity and tax regressivity has fallen most heavily on the region’s Black working class. I conclude by recommending that the state fully fund its statutory revenue sharing program with an income surtax on millionaires, which would permanently expand local fiscal capacity and reduce regressivity in the state’s overall tax burden.

II. Sources and Implications of Property Tax Regressivity

Property taxes are administered by applying the statutory flat tax rate for a class of property to a parcel’s taxable value, which is the share of a property’s market value that can be legally taxed. Because a property’s market value is only observable when it is sold, assessors must make interim approximations using comparable sales (Urban Institute 2021). A property tax is regressive when taxes paid on higher value properties capture a smaller share of their market value than taxes paid on lower value properties. At the metropolitan scale, this can either be because differences in properties’ assessed value do not reflect differences in their market value or because higher value properties are located in jurisdictions that levy lower tax rates. The extent of assessment regressivity will depend on how frequently properties are reassessed, legal limits on the growth of assessed value, the accuracy of valuation models and their sensitivity to neighborhood characteristics, and biases in the assessment and appeals process (Avenancio-León and Howard 2020; Amornsiripanitch 2022).

When housing supply is perfectly elastic, regional variation in home prices reflects variation in demand for public services across households and the spatial distribution of fixed place-based amenities and dis-amenities. The same holds for the market in commercial property. Any demand shifts that increase the price of real estate in a locality, whatever the cause, will draw sufficient investment in new supply to eventually bring prices back down to their initial level (Sirmans, Gatzlaff, and Macpherson 2008). Under these conditions, localities are only able to generate additional revenue by raising tax rates, and any regional property tax regressivity would imply that demand for real estate is greater where tax rates are lower and public services more limited. The reality of pervasive physical, legal, and informal constraints on the supply of housing and commercial real estate means that property tax rates do become capitalized into local property values (Sirmans, Gatzlaff, and Macpherson 2008). Excess demand keeps property values elevated, and because rate reductions do not result in a one-to-one loss in revenue, supply-constrained low tax jurisdictions are able to provide public services at a lower cost to residents.

III. Property Tax Restrictions in Michigan: The Headlee Amendment

Local governments’ use of the property tax has long been heavily regulated in Michigan. Each piece of legislation authorizing the formation of municipal and general purpose governments going back to the Home Rule City Act of 1909 has included a cap on property tax rates, and since 1932, the state constitution has limited aggregate millage rates in unchartered counties and townships and in school districts (CRCM 2021). These restrictions, though, proved insufficient to shield Michigan homeowners from the “pocketbook squeeze” of the late-1960s and 1970s (Mound 2020). As inflation began to accelerate in these years, assessments continued to increase with rising nominal home values, swelling homeowners’ property tax bills at the same time as income growth was beginning to slow. Nationwide agitation and political mobilization against high property taxes peaked in the late 1970s, and in the election of 1978, Michigan voters approved the Headlee Amendment, the first new constitutional restriction on the property tax in the state in nearly half a century.

Headlee is structured as a limit on revenue. If total assessed value, net of additions or losses to the stock of taxable property, increases faster than the national rate of inflation, the maximum millage rate a locality is authorized to levy is adjusted downward to ensure that the real value of total revenue collected on existing properties remains the same (CRCM 2021). Local governments already levying a millage rate below their legal maximum need not make any adjustments in response to real assessment growth, and while rollbacks are automatic, localities can still seek voter approval to restore rates to their initial authorized level. Restraining local property tax revenues in this way has had the effect of mechanically ratcheting up regressivity in statutory rates in metropolitan areas experiencing home price divergence between low and high tax jurisdictions. The simplified model presented in Table 1 illustrates this dynamic in a hypothetical region consisting of one city and one township that initially differ only in their property tax rates.

Between years one and two, both places see property values appreciate faster than inflation. A new property, Township Property B, is constructed, though its contribution to total assessed value is deducted for the purpose of calculating the township’s Millage Reduction Fraction (MRF) (CRCM 2021). Because the value of existing property appreciated faster in the township than the city, its already low millage rate is rolled back by 2.9 percent and the city’s millage rate by one percent. Property tax rates in the region have fallen and are now regressive, the owners of relatively high value properties in the township paying a lower statutory millage rate than the owner of the less valuable City Property A. This model also shows how Headlee rollbacks advantage localities able to attract new construction and development. In the city, property tax revenue grows at the rate of inflation, but in the township, the addition of a new taxable property doubles total revenue. By year two, the township is collecting more revenue than the city, despite its larger rollback.

IV. Property Tax Restrictions in Michigan: Proposal A

Under Headlee, MRFs are calculated using the average assessment growth rate across a whole jurisdiction (CRCM 2021). This aspect of Headlee kept effective property tax rates flat within jurisdictions by shifting the total burden onto property owners with faster appreciating assets. Rising inflation-adjusted property tax bills remained a lingering source of frustration among Michigan homeowners, especially those in cities with high neighborhood-level variation in home price growth. Their concern was finally addressed as part of a broader effort to overhaul public education funding.

Aiming to reduce local property tax burdens and equalize per pupil funding across school districts, the state legislature voted in 1993 to ban districts’ use of the property tax to fund general operating expenses. The default statutory plan would have primarily made up for this $7 billion in lost revenue by raising the state income tax, but the following year Michigan voters approved an alternative ballot initiative, Proposal A, raising state sales and cigarette taxes instead (Michigan Department of Treasury 2002). Both plans included a new state property tax earmarked for education spending. In addition to leaning more heavily on regressive consumption taxes, Prop A also enshrined a new constitutional limit on property tax assessments that capped growth in any individual parcel’s taxable value at the rate of inflation, taxable value reverting to assessed value on resale. By 2020 the statewide gap between taxable and assessed value had grown to over $100 billion, introducing a new source of intra-jurisdictional regressivity to the state’s local property tax structure (CRCM 2021). Tables 2 and 3 return to the same hypothetical region from Table 1 to demonstrate Prop A and Headlee’s joint effect on property tax regressivity.

In this scenario, property values continue to diverge between city and township, increasing twice as fast in the township. With Prop A now in effect, a gap has opened in both jurisdictions between taxable and assessed value, benefiting the owner of fast appreciating Township Property A the most. Township Property B, which was resold between years two and three, saw its taxable value increase by more than the rate of inflation, boosting total taxable value in the township by three percent. Because of Headlee, the township must still roll back its millage rate to hold revenue growth at two percent, further widening regional regressivity in millage rates. Table 3 shows that most of the township’s revenue growth comes from the sale of Property B. The effective millage rate on Township Property A, the region’s most valuable property, is just $14.81 per $1,000 in assessed value, down from $16.00 in Year 1 and now the lowest in the region. Its owner’s total tax bill has increased by three percent, slower than the rate of inflation. The owner of City Property A, worth less than Township Property A, pays an effective rate of $19.62 and has seen a 4.1 percent increase in their total tax bill over the three year period.

The main beneficiaries of the way Headlee and Prop A restrict local property taxes are longtime property owners in low tax jurisdictions. Housing supply constraints also fully or partially capitalize these places’ low millage rates into home values, precluding working class households from affording homeownership (Sirmans, Gatzlaff, and Macpherson 2008; Lens and Monkkonen 2016). Expensive low tax enclaves are able to keep millage rates low and still meet residents’ demand for municipal services, while high tax cities are effectively blocked from generating additional property tax revenue, risking erosion of their tax base through outmigration and divestment.

V. Case Study: Property Tax Regressivity in Metro Detroit

The simplified model helps clarify how Michigan’s constitutional property tax restrictions interact with each other and with local housing supply restrictions to worsen regional tax regressivity, dynamics on full display in metro metro Detroit,[1] the state’s largest metropolitan area. These constraints have encouraged higher income white residents’ flight to suburban and exurban townships and forced shrinking cities to either borrow or wring additional revenue from a shrinking tax base. Thirty years of mounting pressure on municipal finances culminated most spectacularly in the city of Detroit’s 2013 bankruptcy filing, the largest municipal bankruptcy in U.S. history (Bomey and Gallagher 2013).

In 1977, Detroit property owners paid a total millage rate that was 1.3 times the average rate elsewhere in the region, due to the city’s considerably higher operating millage. At this time, Detroit Public Schools (DPS) still levied a lower millage rate than the average paid in other school districts. After Headlee, property tax rates became even more regressive, as the city and DPS raised property taxes by five and 14 mills, respectively, to make up for falling revenues. Rising school property taxes regionwide drove support for expanded state education funding in the early 1990s, but because school property taxes made up a larger share of residents’ total tax bill outside of Detroit, suburban property owners disproportionately benefited from subsequent cuts. On the eve of the Great Recession, Detroiters were paying total millage rates more than double the average millage rate across other jurisdictions, and they continue to do so, even though the average parcel in the city is worth a fifth of the value of the average parcel in the rest of the region. In 2019, Detroiters’ total property tax rates were higher than they were in 1977. Residents of other jurisdictions in the region saw an average tax cut of 30 percent over the same period.

In these four decades, growth in the metro region as a whole was anemic, the total population hovering at roughly 4.3 million residents between 1980 and 2019. Even as overall demand for housing stayed flat, the post-Headlee period was marked by a dramatic population shift from high tax cities to low tax townships. The population of Detroit nearly halved, and as other larger cities in Wayne, Oakland, and Macomb counties shrank too, the population living in townships and small suburban cities doubled. Reflecting this demographic inversion, most of the net increase in the region’s housing stock occurred in townships. Despite having the lowest average millage rates in the region, investment in new housing construction generated a significant revenue windfall for the latter.

Figures 1 and 2 present average annual percent change in revenue by source for municipal and township governments in metro Detroit. From 1977 to 1997, townships’ total property tax receipts grew by 9.3 percent per year, much faster than the five percent national rate of inflation, while Detroit and other shrinking cities like Highland Park and Pontiac saw large real reductions in property tax revenue. Detroit Public Schools (DPS) raised its operating millage by nearly 50 percent in the 15 years after Headlee, yet also failed to generate real revenue growth (Pierson, Hand, and Thompson 2015). The recessions of the early 1980s hit the auto industry and the broader regional economy hard, but only in Wayne County did job losses prove permanent (Bomey and Gallagher 2013). By 1997, there were 204,000 fewer jobs in Wayne County than there were in 1978, even though the total number of jobs regionwide had grown by 265,000 (U.S. BEA 2021). In the late 1990s, Detroit again began to see real growth in property tax receipts, buoyed by rising home prices, but still lagged well behind other cities and townships.

The period over which Michigan adopted its constitutional property tax restrictions was exceptionally austere in Detroit, the scope and quality of municipal goods and services declining significantly relative to other cities in the region. Inflation consistently outpaced total general revenue growth in the city, unlike in flush townships. Declining enrollment forced DPS to close dozens of schools and the physical condition of those that remained open began to deteriorate (Grover and van der Velde 2016). Poor academic performance, administrative corruption, recurring labor strife over low teacher pay, and state legislation authorizing charter schools and inter-district transfers all hastened students’ flight from the city’s public schools and were the grounds on which the state dissolved DPS’s elected school board in 1999. The district has been subject to state fiscal oversight ever since, and from 1999 to 2006 and 2009 to 2020 was controlled directly by appointed emergency leadership (CRCM 2019). Headlee and Prop A made all general purpose governments seek alternative sources of revenue, either to make up for a shrinking tax base or to serve a growing population. Regionwide, property taxes’ share of total revenue held constant at 28 percent in the first 20 years after Headlee, and in the city of Detroit, slid from 19 percent to 12 percent (Pierson, Hand, and Thompson 2015). This decline, in Detroit and elsewhere, went into reverse with the onset of the housing market boom, but also reflected significant changes in the state’s statutory revenue sharing program.

VI. Case Study: State Redistribution and Austerity

For more than 80 years Michigan has set aside a portion of state revenues for local governments’ unrestricted use (MML 2019). Its revenue sharing obligations are both constitutional and statutory. While the former has always been tied to the state sales tax and allocated to localities on a per capita basis, statutory revenue sharing programs have drawn on different sources of state revenue and distributed grants according to different formulas over the years. In the early 1970s, Michigan revised its allocation formula for state income and business tax sharing by weighting localities’ population share by their average millage rate relative to the statewide average rate. The “relative tax effort” (RTE) formula was meant to redirect general state budget support to cities with relatively high demand for municipal services and a greater diversity of infrastructural needs (CRCM 2015). Figure 1 shows that adoption of RTE did in fact help to mitigate post-Headlee revenue losses in Detroit, and between 1973 and 1990, state transfers’ share of city revenue rose from 11 percent to 33 percent (Pierson, Hand, and Thompson 2015). Low tax townships also saw a large influx of state support in this time, though largely from constitutionally mandated revenue sharing tied to population growth.

Statutory revenue sharing underwent several major changes in the 1990s, starting with the phase out of the state intangibles tax in 1991 and the substitution of state income and business tax sharing with expanded sales tax sharing five years later (MML 2019). This shift toward funding from more regressive sources of state revenue was shortly followed by the abandonment of the RTE formula in 1998, to be replaced with a new multi-formula system for allocating statutory revenue sharing. Anticipating that the new formulas and ongoing population loss would reduce its share of state transfers, Detroit’s municipal leadership and state representatives agreed to cut city income taxes in exchange for freezing state transfers at their 1998 level for the next eight years (CRCM 2015).

Before the new program was fully implemented, the bottom fell out of the state labor market. Total employment in Michigan contracted every year between 2000 and 2010, and more than a quarter of the 845,000 jobs lost statewide were in Wayne County (U.S. BEA 2021). Forced by balanced budget requirements added to the state constitution in 1963 (Rueben, Randall, and Boddupalli 2018), governors and legislators from both parties pursued procyclical austerity and slashed statutory revenue sharing (CRCM 2015). Chronic underfunding cumulatively diverted $8.6 billion in revenue from local governments between 2001 and 2018 (MML 2019). Figures 2 and 3 show the widespread impact of these cuts and their impact on Detroit in particular.

State-imposed austerity through constitutional property tax restrictions and cuts to statutory revenue sharing have made Detroit uniquely reliant on its local income tax and a wagering tax on casinos adopted in 1999 (Bomey and Gallagher 2013). These taxes together account for nearly two-fifths of the city’s own-source revenue (Pierson, Hand, and Thompson 2015). They are also closely tied to overall conditions in the regional economy and labor market, making the city’s fiscal capacity even more vulnerable to business cycle fluctuations. Figure 3 depicts the coincidence of state austerity and collapsing income tax receipts during the 2000s depression in Detroit. For the first few years of the decade, rising home values and new wagering tax revenue were sufficient to sustain city budgets and enable continued borrowing, but after 2005, property tax revenues fell by $236 million and unemployment climbed to 20 percent (Pierson, Hand, and Thompson 2015; U.S. Bureau of Labor Statistics 2021). From 2010 to 2013, the state slashed total intergovernmental transfers to Detroit by another $207 million (Pierson, Hand, and Thompson 2015). Plunged into a spiral of increasing borrowing costs and sharply falling revenues, the city was finally forced to declare bankruptcy in the summer of 2013 (Bomey and Gallagher 2013).

Table 6 shows that this difference in revenue mix between Detroit and other localities puts the city at a disadvantage even when the region isn’t in recession. Wayne County residents are employed in jobs that pay lower average wages than jobs held by commuters, and as a result, millions in earnings flow out of the county every year. Suburban and exurban jurisdictions in Livingston, St. Clair, and Lapeer counties benefit the most from intraregional net earnings flows. Detroit and a handful of other cities in the region do levy an income tax on non-resident workers, but it is typically set at half the income tax rate on residents. More than a taxable income stream, lost commuter earnings also represent purchasing power that would have otherwise stimulated local economic activity and put upward pressure on property values. Since the start of the COVID-19 pandemic, revenue from non-resident income taxes has been heavily impacted by the rise of remote work and Republicans in the state legislature have introduced legislation to prohibit their use statewide (Oosting 2021).

VII. Case Study: Regional Segregation by Race and Class

These commuting flows largely reflect the movement of high wage white workers. In metro Detroit, regional segregation by race long predates the Headlee Amendment. During and after the second World War, the region’s Black population swelled with migrants from the rural South seeking high wage industrial employment and refuge from the oppression of Jim Crow (Sugrue 2014). Until the 1968 Fair Housing Act, federal underwriting standards explicitly discriminated against segregated Black neighborhoods, channeling the dramatic postwar expansion in home mortgage credit to greenfield commuter suburbs with large and stable white majorities and few low income residents (Rothstein 2017). Despite high demand, these jurisdictions were zoned for low density single-family dwellings, which constrained the supply of new housing and inflated property values (Sahn 2021). White residents were able and willing to pay this premium for high quality public goods at low property tax rates, while most Black residents could not (Boustan 2017). Instead, Black families who could afford homeownership began acquiring property in urban neighborhoods vacated by white flight, leaving the most impoverished Black residents confined to the deteriorating housing stock in Detroit’s historic ghetto (Boustan and Margo, 2013; Sugrue 2014).

Over the last fifty years, legislation and litigation have done much to curtail racial redlining and other transparently discriminatory policies and real estate practices. And yet metro Detroit remains one of the most segregated regions in the country (Menendian, Gambhir, and Gailes 2021). Density restrictions and organized homeowner opposition continue to block the construction of affordable housing outside of large cities, confining the region’s Black working class to segregated urban neighborhoods (Lens and Monkkonen 2016; Trounstine 2020). The recent movement of middle class Black homeowners into suburban cities largely reflects the same pattern of concentric outmigration that characterized the era of postwar suburbanization, as these newcomers replace white residents decamping for even farther flung exurbs in Livingston, St. Clair, and Lapeer counties. In townships, home values average $264,750 and rental units, less than a fifth of the township housing stock, cost on average $340 more per month than rentals in Detroit. By widening regional property tax regressivity, Headlee and Prop A have reinforced longstanding patterns of regional segregation and made suburban housing into a tax-advantaged asset almost exclusively accessible to the white middle and upper class.

Eighty percent of those living in townships are native-born and identify as non-Hispanic white. Average household income across these jurisdictions is over $100,000 a year. In the city of Detroit, more than three-quarters of residents are Black and the average household makes $44,730. A third of city residents fall below the federal poverty threshold, which in 2019 was $25,570 for a family of four. Seventy percent of the region’s Black residents live in one of nine jurisdictions that is majority Black, in all of which no fewer than 10 percent of residents live in poverty. This segregated social geography is reflected in the region’s schools as well. In the 2018-19 school year, 70 percent of Black public school students in metro Detroit attended a majority Black school and more than half attended a school that was over 80 percent Black (U.S. Department of Education 2021). Eighty-three percent of students attending a majority Black school qualified for free or reduced price meals.

Because low income households struggle to afford unsubsidized market goods and services, their spatial concentration both increases demand for municipal services and makes it harder for local governments to raise the revenue necessary to meet that demand. This fiscal trap is worse in regions where race and class segregation intersect, as the devaluation of majority Black neighborhoods further depresses the local property tax base (Perry, Rothwell, and Harshbarger 2018). When home prices began to fall in late 2006, segregated neighborhoods targeted by predatory subprime mortgage lenders experienced a spike in defaults and foreclosures (Rugh and Massey 2010). This sudden increase in neighborhood housing supply at fire sale prices put additional downward pressure on home values, but because state law requires assessors only use “arm’s length transactions” as comps to determine a property’s true cash value, these sales were not fully reflected in occupied properties’ annual assessment revisions (Hodge et al. 2017; Atuahene and Hodge 2018). As unemployment increased and incomes fell, many homeowners became delinquent on their inflated tax bills and some ultimately dispossessed through property tax foreclosure (Atuahene and Berry 2019).

Most Detroit homeowners did manage to hold on to their property through the crash and recession, and when the housing market finally began to recover, Prop A helped to insulate them from rising real property tax bills. The 2019 statutory total millage rate in Detroit was $73.80 per $1,000 in taxable value, but the average property owner paid just $47.94 for every $1,000 in assessed value. This large gap between taxable and assessed value represents potentially tens of millions of dollars in revenue that the city would have collected were Prop A not in effect. By depressing property tax revenue in the aftermath of the housing market crash, Prop A created a powerful incentive for the city of Detroit to continue using inflated assessments for properties in devalued neighborhoods.

Overassessment amounted to a $600 million tax on Detroit’s working class homeowners, nearly all Black, shared between the city, school district, county, and state to help make up for recession-induced budget shortfalls and service their large debts (CRCM 2019; Ishag-Osman 2021). That Detroit’s tax base is only 65 percent of assessed value is a consequence of both overassessment and the age of the housing stock in Detroit. Nine percent of the city’s housing stock was built after 1980, while 24 percent and 53 percent, respectively, of housing units in other cities and in townships were built after Headlee. Taxable and assessed value are only set equal when an existing property is sold or if the property is new construction, so the gap between the two is predictably smaller, around 80 percent, in growing parts of the region where demand for housing is high. Given that assessments in Detroit are still inflated for low value properties (CMF 2020), taxable value’s share of assessed value would also likely be closer to the average elsewhere in the region if more accurate estimates of market value were used. For the owners of overassessed properties, effective tax rates are potentially even higher than the city’s statutory millage rate (Hodge et al. 2017).

Overassessment nevertheless fails to offset the structural inequality racial and class segregation creates between localities in their ability to raise revenue from property taxes. Township property owners pay an average total millage rate of $29.25 per $1,000 in taxable value but generate $3,035 per parcel for use by local governments and school districts because their property is so highly valued. In Detroit, the average parcel yields only $1,222 in property tax revenue even though property owners pay a tax rate that is two and a half times higher. Other majority Black cities in the region with many low income residents and devalued properties face the same constraint.

Across the country, segregation and state restrictions on local taxation have made devalued, high poverty cities reliant on regressive, and increasingly predatory, means of extracting revenue from their most vulnerable residents (Atuahene and Hodge 2018; Atuahene 2020; Scharff 2021). Forty years ago, the state of Missouri adopted its own constitutional property tax restrictions, introducing Headlee-style rollbacks and requiring voter approval for local tax increases (Kevin-Myers and Hembree 2012; Johnson 2015). Under these similarly austere conditions, Ferguson, a majority Black city in metro St. Louis, directed local law enforcement to prioritize revenue maximization in their collection of civil and criminal fines and fees to the point of actively encouraging abusive and unconstitutional policing against the city’s Black working class (Johnson 2015; Atuahene and Hodge 2018). In the summer of 2014, mounting tensions over this discriminatory and extractive system boiled over into weeks of mass protest and civil unrest when a Ferguson police officer shot and killed 18-year old resident, Michael Brown. Infamous as Ferguson’s predatory practices have become, it is representative, and like Detroit, illustrates well how municipal austerity disproportionately harms Black lives and livelihoods.

VIII. A Just Fiscal Federalism: The Case for a State Millionaire Surtax

The structural nature of this problem demands a structural solution. Detroit homeowners should continue to fight for the city to acknowledge and rectify the problem of overassessment, but as this paper has argued, overassessment is symptomatic of the fiscal regime Michigan has constructed over the last 40 years. Regressivity is inherent to the state’s constitutional property tax restrictions, which have compelled revenue-starved cities to overtax their own residents even while cutting spending (Atuahene and Hodge 2017; Atuahene 2020). A decade later, real per capita city spending in Detroit was still down from 2007 (Pierson, Hand, and Thompson 2015). And yet supporters of a fairer and more progressive state and local tax structure should not necessarily aim for the total elimination of Headlee and Prop A, were that even a realistic political prospect. Property taxes were regressive before Headlee and the national tax revolt of 1970s was animated as much by the discontent of working class urban homeowners as more affluent suburbanites (Mound 2020). In regions riven by racial and class segregation, fiscal devolution tends to reinforce inequalities in locally-provisioned public goods, which was the impetus for reducing school districts’ reliance on the property tax in the early 1990s.

Advocates for reform should instead prioritize fully funding statutory revenue sharing by eliminating the state’s constitutional restriction on graduated income taxes. Michigan has one of the country’s more regressive state and local tax structures: In 2018, the lowest income 80 percent of families paid average tax rates over 9 percent, while the richest 1 percent of families paid an average rate of 6.2 percent (Wiehe et al. 2018). A 3.75 percent surtax on incomes in excess of $1 million would affect the richest 0.6 percent of taxpayers and generate roughly $600 million in revenue, sufficient on its own to fully fund a new statutory revenue sharing program that meets the state’s commitments from the 1998 reform process (MML 2019; Internal Revenue Service 2021). This change would enable the state to either cut its regressive sales tax or reallocate revenue currently set aside for revenue sharing to other uses. In anticipation of likely arguments against such a change, there is little evidence that state income tax increases slow economic growth nor that they encourage any significant outmigration of high income households (Young et al. 2016; Tharpe 2019). Even with this proposed millionaire surtax, Michigan’s income tax rate on top earners would still be less than in most of the other 42 states that levy their own income taxes (Loughead 2021).

Constitutional revenue sharing is allocated based on population and funded from a portion of state sales taxes, disproportionately benefiting growing jurisdictions with low property tax rates. To ensure state transfers reach communities disadvantaged by this formula, statutory revenue sharing should be based on a formula that adjusts localities’ population share by both relative tax effort and the inverse of assessed value per parcel relative to the average parcel value statewide. The latter would have the added benefit of disincentivizing overassessment by reducing revenue sharing payments to cities with inflated assessments. Even low tax jurisdictions would stand to see a relatively large infusion of new state grant funding using this formula, given how underfunded the existing program is already.

As evidence of the growing appetite for such a policy change, state and local officials from both parties have publicly conceded that the fiscal crises of the 2000s made clear the need to rethink how the state funds local government (Oosting 2019). Voters nearly had the chance in 2020 to amend the state constitution and levy a graduated income tax tied to school and infrastructure spending, but the onset of the COVID-19 pandemic forced organizers to suspend the campaign (Oosting 2020). Arizona voters faced a similar ballot initiative and ultimately backed a 3.5 percent surcharge on high earners earmarked for increased school funding and teacher pay. While the state’s Republican-controlled legislature, Republican governor, and Republican-appointed supreme court have all worked to thwart the law’s implementation, voters will be asked again in November 2022 to either reaffirm or reject the progressive rate structure they supported two years earlier (Bushman 2021). Whether or not the Arizona Republican Party succeeds in nullifying the voter-approved tax increase, the state’s experience shows that progressive income taxes tied to urgently needed local spending can win broad support among a more conservative electorate despite intense institutional resistance.

Over the last 40 years, fiscal devolution and state restrictions on local taxation have contributed to growing inequality, underinvestment in public goods and services, and the criminalization of poverty and rise of mass incarceration (Scharff 2021). That voters have been receptive to reducing regressivity in state and local tax collection and boosting public spending on local services reflects a growing recognition of the toll austerity has taken on some of the country’s most vulnerable and disempowered communities. With its long history of state and local revenue sharing, Michigan is well positioned to lead the nation in transitioning to a more just fiscal federalism, in which localities have the means to meet residents’ diverse needs and to make urgent investments in education, housing, transportation, and public health that have been too long neglected.