BUILDING ON MICHIGAN'S AUTO INSURANCE REFORM LAW

INTRODUCTION

Michigan drivers have long paid the highest auto insurance

rates in the country. By mandating unlimited personal

injury protection (PIP) for all drivers and failing to regulate

reimbursement rates for medical providers, the state’s

average insurance rates had ballooned to over $3,000

per year by 2019, nearly $800 higher than the next closest

state.

While Michigan rates were high, rates in its biggest

city, Detroit, were astronomical, with drivers facing an

average rate of over $6,300, a function of the unlimited PIP

requirement coupled with insurance discrimination based on

geographic and other non-driving factors.

In the spring of

2019, the Michigan legislature passed sweeping reforms to

Michigan’s auto insurance laws, with the goal of driving down

these exorbitant rates. These reforms began to take effect in

the summer of 2020.

In this brief, we analyze the early outcomes of auto insurance

reform more than a year after the initial provisions of the law

went into effect. In a 2019 brief, which was widely cited in the

run-up to reform, we argued that high auto insurance rates

erected a significant barrier to economic mobility, forcing

those with low incomes to go without a car, dedicate a large

share of their limited income to insurance premiums, or drive

without insurance, exposing them to legal penalties. Here, we

evaluate the success of auto insurance reform through the

same lens, seeking to understand the extent to which recent

reforms have offered Michiganders a boost in their pursuit of

economic mobility.

We will provide some background on the 2019 law, use

preliminary data to describe the impact of the law more than

one year after taking effect, and discuss what still needs to be

done to ensure the cost of auto insurance does not continue

to stand as a barrier to economic mobility for Michigan

residents. Early data find the average cost of auto insurance

declined significantly in Michigan from 2019 to 2020—a steeper

decline than anywhere else in the country. However, the cost

of insurance remains unreasonably high for many Michigan

residents—and Detroit residents in particular. We will outline

how Michigan’s insurance laws can be improved to further

reduce rates statewide and in Detroit. In addition, we will

discuss some of the early, unintended consequences of the

reform law, such as reduced revenues for long-term care

facilities, and outline how careful policy reforms can assure

quality long-term care for vulnerable Michiganders, while not

driving up insurance rates across the system.

BUILDING ON MICHIGAN’S AUTO INSURANCE REFORM LAW

By Amanda Nothaft and Patrick Cooney

DECEMBER 2021

KEY FINDINGS

• The reforms instituted through the 2019 auto insurance

reform law made a difference. Early data suggest rates

across the state have fallen by nearly 20%.

• Even with these reductions, Michigan rates remain

the highest in the country, and average rates in Detroit

still eat up over 18% of the median household income in

that city.

• Insurance rates are still highly correlated with race,

and more can be done to eliminate the discriminatory

impact of using non-driving factors to calculate rates.

• The method used to cap medical fees may be

unnecessarily stringent and out of line with national

peers, causing a crisis in access to care for victims of

catastrophic accidents that occurred prior to reform.

AUTO INSURANCE IN MICHIGAN AND THE 2019

REFORM LAW

Owing to a unique mix of policies, Michigan has long held the

dubious distinction of having the highest auto insurance rates in

the country. Prior to the 2019 reform law, Michigan was the only

state in the country that required every driver to have unlimited

personal injury protection (PIP) to cover medical claims in the

case of severe injury, and auto insurance was almost always

considered the primary insurer in the event of auto accidents.

In addition, Michigan did not impose a medical fee schedule on

procedures covered by auto insurance. These two policies led to

high claims in Michigan, which drove up rates.

In addition, Michigan allowed the use of non-driving factors to

determine auto insurance rates, which led to rates in majority-

Black zip codes that were even higher than those in the rest

of the state. Under prior law, insurance companies could

take into account a range of non-driving factors to determine

auto insurance rates, including marital status, educational

attainment, homeownership status, credit scores, and the zip

code in which one lives.

Owing to historical and persistent

racial discrimination in housing and labor markets, as well

as structural racial inequities in our education system, many

of these non-driving factors highly correlate with race.

This

means that prior to reform, the share of Black residents in

a zip code was highly correlated with the average price of

auto insurance (correlation coefficient of 0.76, with 1 being the

strongest possible correlation), while other factors, such as a

zip code’s median income or poverty rate, were only weakly

correlated with the price of insurance.

Said another way, if

you knew the share of Black residents in a zip code, you could

offer a good guess as to the price of auto insurance. In 2019,

the average premium for all zip codes in the state was $3,100;

for the 37 zip codes in which more than 50% of residents were

Black, the average rate was $5,500.

Several elements of the law that took effect in 2020 are

designed to both rein in PIP payouts and reform discriminatory

rate setting practices. Under the new law, drivers are not

required to purchase unlimited PIP coverage, but they

can instead select PIP coverage that best fits their needs

and existing health insurance coverage. Consumers have

the choice of six different coverage levels, including an

opportunity to opt out of purchasing unlimited PIP coverage if

they have a health insurance policy that covers auto injuries

and has a deductible of $6,000 or less.

The law also provides

low-cost options for seniors and drivers with low incomes,

allowing seniors to completely opt out of PIP if they have

Medicare Part A and B, and Medicaid recipients to purchase

a policy with $50,000 in PIP coverage, so long as other

household members also have adequate coverage.

The new

law also seeks to reduce PIP costs by placing limits on the

fees health providers can charge auto insurance companies

for medical services, a change that should impact everyone

who carries PIP coverage.

The new law also prohibits the use of sex, marital status,

homeownership, education level, zip code, or credit score

when determining rates.

By removing these factors—most

of which are highly correlated with race and poverty—the

new law removes a set of elements that disproportionately

impact minority and low-income drivers and should have a

bigger impact on their rates compared to the average driver.

However, as we discuss below, it is unclear how this reform is

playing out in practice, as insurers can often use proxies for

these non-driving factors that may have an equally or perhaps

greater discriminatory effect.

HOW MUCH HAVE RATES CHANGED POST

REFORM?

The new law has been in effect for a little over a year, and

certain elements of the law—such as the reduction in

reimbursement rates for medical services covered by auto

insurance—only recently went into effect.

However, even

partial implementation of the law over a relatively brief time

period has led to a significant decline in rates across the state.

Relying on data from The Zebra, an auto insurance comparison

marketplace that collects rate information from public rate

filing and insurance rating platforms, we analyzed the changes

in estimated rates statewide and in Detroit.

Between 2019 and 2020, average estimated rates fell by 18%

statewide, the steepest decline of anywhere in the country

over that time period.

This is a decline of nearly $600 per

year for the average Michigan driver. Despite this dramatic

fall, Michigan’s auto insurance rates remain the highest in the

nation, after having risen by 40% between 2011 and 2019. The

high statewide average is partially driven by Detroit and metro

Detroit, where auto insurance rates are higher than the state

average. However, even when Detroit and the Detroit Metro

area are excluded from the calculations, Michigan’s average

rate—$2,336—is still slightly higher than any other state in

the country, and is 2.5 times the average rate in Ohio, which

at $926 has one of the lowest rates in the nation. Despite the

reform and a substantial decrease in rates, insurance rates

remain high and largely unaffordable in Michigan.

Insurance rates have always been substantially higher in

Detroit, partially due to greater numbers of PIP claims and

partially because insurers were allowed to use non-driving

factors to calculate rates.

In theory, the new law should lead

to larger declines in Detroit than the state overall because

the exclusion of non-driving factors in rate-setting should

have a larger impact in majority-Black areas most severely

impacted by discriminatory rate-setting practices. However,

post-reform, the rates in Detroit remain twice as high as the

rest of the state, falling at the same rate as the statewide

average (18%). While these data are preliminary and we may

see rates drop more dramatically in the next year, it appears

the new law does not go far enough to protect non-white and

low-income people from being discriminated against in the

insurance market.

ROOM FOR IMPROVEMENT?

While rates in Michigan remain the highest in the country

and $1,000 above the national average of $1,483, rates in

Detroit are an eye-popping $5,146 a year.

In New Orleans,

the city with the second highest rates in the country,

residents face an average rate of $3,564, nearly $1,500

less than Detroit.

Rates may decline further as actuarial

models adjust to the consequences of the law and new

insurers enter the Michigan insurance market, many of

whom did not previously offer policies in the state due to

the requirement of unlimited PIP coverage with no fee

limits.

However, the gap between rates in Detroit and the

rest of the state does not show any sign of closing despite

the removal of non-driving factors from rate calculations.

Additionally, as the law comes into effect, some individuals

injured under the unlimited PIP provisions in the prior law

are losing access to catastrophic care services that they

previously received, an unintended and unfortunate byproduct

of the reform law.

Understanding why rates may not be

falling at a faster rate for Detroit drivers and how the new cost

controls are impacting vulnerable auto accident victims can

give us insight into ways the law could be made stronger.

TAKING A CLOSER LOOK AT NON-DRIVING FACTORS

Although the new law specifically prohibits the use of zip

code and credit score in calculating rates, it still allows

insurance companies to group insurance risks by geographic

“territory” and use an insurance score, which has credit score

as a component, in determining rates.

Consequently, rates

remain as highly correlated with race as they were prior to

reform. The chart below shows the median price of auto

insurance in every zip code in the state plotted against the

percentage of residents in that zip code who are Black. Blue

dots represent 2019 rates, orange dots represent 2020 rates,

and we have added trend lines highlighting the relationship

between the racial makeup of a zip code and auto insurance

rates. As we can see by the trajectory of the trend lines,

while insurance reform has lowered rates across the board,

it has done nothing to change the relationship between the

percentage of Black residents in a zip code and the price of

auto insurance in that zip code.

FIGURE 1: AUTO INSURANCE RATES, 2011 TO 2020

$1,000

$2,000

$3,000

$4,000

$5,000

$6,000

$7,000

Detroit

Detroit Metro

Outstate

State Overall

2020201920182017201620152014201320122011

$4,501$4,501

$5,244

$4,346

$4,307

$4,175

$5,679

$6,029

$6,066

$6,314

$5,146

$2,398$2,398

$2,901

$2,356

$2,292

$2,211

$2,918

$3,180

$3,195

$3,477

$2,820

$2,068$2,068

$2,

587$2,587

$2,

093$2,093

$1,

961$1,961

$1,

895$1,895

$2,

419$2,419

$2,

636$2,636

$2,

656$2,656

$2,

857$2,857

$2,336

$2,220$2,220

$2,742

$2,222

$2,109

$2,038

$2,635

$2,869

$2,884

$3,106

$2,535

$1,000

$2,000

$3,000

$4,000

$5,000

$6,000

$7,000

Detroit

Detroit Metro

Outstate

State Overall

2020201920182017201620152014201320122011

$4,501$4,501

$5,244

$4,346

$4,307

$4,175

$5,679

$6,029

$6,066

$6,314

$5,146

$2,398$2,398

$2,901

$2,356

$2,292

$2,211

$2,918

$3,180

$3,195

$3,477

$2,820

$2,068$2,068

$2,

587$2,587

$2,

093$2,093

$1,

961$1,961

$1,

895$1,895

$2,

419$2,419

$2,

636$2,636

$2,

656$2,656

$2,

857$2,857

$2,336

$2,220$2,220

$2,742

$2,222

$2,109

$2,038

$2,635

$2,869

$2,884

$3,106

$2,535

LEGEND:

Geography can play a legitimate actuarial role in determining

rates—for example, driving in more congested urban areas

results in more accidents, driving faster in rural areas

may cause more fatal accidents, and medical costs may be

higher in one part of a state versus another.

However, using

geography to determine rates can also lead to insurance

“redlining,” in which residents of neighborhoods with high

concentrations of Black residents or residents with low

incomes may be charged higher rates based on perceived

risks that may not be borne out by the data.

To ensure geography is used in a way that balances increased

risk and geographical differences in medical costs without

discrimination, some states provide restrictions on how

territories can be determined, provide state-defined

territories, or require insurers to get approval for how they

define territories.

Michigan only restricts insurers from

using zip codes to determine rates and does not require

approval for, or transparency about, how insurance companies

define their territories. Using state-specified territories

or allowing insurers to develop their own territories based

on strict criteria, as is done in New Jersey, would provide

greater transparency and blunt discriminatory impacts.

Alternatively, the law could be amended to require insurers

to create territories that are at least 20 square miles, as is

done in California, potentially broadening the socioeconomic

characteristics of individuals in a particular area.

California’s

law also requires more weight be given to driving-related

factors—such as miles driven and vehicle characteristics—

than geography in developing rates.

In Connecticut, insurers

are required to moderate the impact of geography on rates by

weighting territory versus statewide experience, which uses

the lower statewide rate to temper the higher rates in the

urban areas.

Taking similar steps to ensure territories under

the new law do not have the same discriminatory impact as

zip codes under the old law could help close the gap between

rates in Detroit and the rest of the state.

Similarly, while the prohibition on using credit scores

to calculate risk was intended to diminish racial and

FIGURE 2: MEDIAN PRICE OF AUTO INSURANCE IN MICHIGAN ZIP CODES 2019 AND 2020 BY % BLACK IN ZIP CODE

0% 20% 40% 60% 80% 100%

1000

2000

3000

4000

5000

6000

7000

8000

Median price of

auto insurance 2019

Median price of

auto insurance 2020

Linear (Median price of

auto insurance 2019)

Linear (Median price of

auto insurance 2020)

0% 20% 40% 60% 80% 100%

1000

2000

3000

4000

5000

6000

7000

8000

Median price of

auto insurance 2019

Median price of

auto insurance 2020

Linear (Median price of

auto insurance 2019)

Linear (Median price of

auto insurance 2020)

LEGEND:

socioeconomic disparities in rates, the continued use of

insurance scores—which rely, in part, on credit scores—may

render this change irrelevant. The continued use of credit

scores, either explicitly or by proxy, will have a racially and

socioeconomically discriminatory effect, as the outcomes of

other forms of systemic discrimination are often visible on

credit reports.

Furthermore, analyses have found credit

scores are unrelated to driving risk.

To truly remove these

negative biases, insurance scores should be eliminated as a

factor in determining insurance rates, as California, Hawaii,

and Massachusetts have done.

REINING IN MEDICAL COSTS WHILE SUPPORTING

QUALITY CARE

For years prior to reform, the major driver of high rates across

the state was Michigan’s unique mix of policies that mandated

all drivers obtain unlimited PIP while at the same time placed

no limits on the cost of medical services covered by auto

insurance. Though the reform law ended the unlimited PIP

requirement—the likely cause of declining rates statewide—

and introduced a fee schedule on medical services that went

into effect this year, more can be done in both of these areas

to lessen the impact of medical costs on rates, while also

supporting quality long-term care for those who need it.

As reflected in their lower auto insurance rates, other states

have contained PIP costs better than Michigan. Post-reform,

Michigan still requires a minimum of $250,000 of PIP for

most drivers, and it remains the only state where unlimited

PIP coverage is an option. In other no-fault states, minimum

PIP coverage ranges from a low $3,000 in Utah to a high of

$50,000 in New York, where $50,000 is also the maximum

available.

Additional coverage is available in other no-fault

states, but these states set limits to coverage, such as $50,000

in New York or $250,000 in New Jersey. Though eliminating

mandatory unlimited PIP was a necessary step, Michigan's

still elevated rates may be in part due to its still relatively high

PIP requirements.

In states with lower minimums, health insurance becomes

the primary payer after PIP limits are reached, which helps

contain PIP costs. Some states, like New Jersey, have lowered

medical costs further by allowing drivers to designate their

health insurance as the primary insurer in a car accident.

Michigan’s new law allows auto insurers to offer a managed

care option, where auto insurers contract with health care

providers to cover the cost of care and medical deductibles for

drivers. However, insurers are not required to offer this option.

Ensuring consumers have the choice of selecting a managed

care option or a deductible on medical expenses could help

further reduce rates.

In addition to mandating unlimited PIP, the previous auto

insurance law did not impose a medical fee schedule on

service providers, resulting in inflated rates for medical

procedures covered by auto insurance as opposed to other

forms of health insurance. To control medical costs, the

reform law set rates at 200% of the Medicare reimbursement

rate, a level similar to those in other states that rely on

the Medicare charge list or the fees allowed by the state’s

workers’ compensation schedules.

For procedures not on the

Medicare master charge schedule, however, fees are capped

at roughly 50% to 75% of what they were in July 2019. This new

fee schedule was introduced in July 2021.

While the introduction of a medical fee schedule was

necessary to control costs and bring down rates for Michigan

drivers, the rate cuts for certain services subject to the lower

reimbursement rates, such as those provided by long-term

care facilities, may have been too dramatic. Already, reports

suggest these fee caps are causing problems for accident

victims who were injured prior to the passage of the reform

bill, as many providers face budget shortfalls and may be

forced to go out of business.

Indeed, other states have set rates for services not on the

Medicare schedule in a manner that seeks to reimburse

providers for the true cost of care. For example, New Jersey

creates a regional state fee schedule based on the “reasonable

and prevailing fees of 75% of practitioners in the region.”

For fees not on the schedule, providers justify their fees

based on fees billed to other payers. Insurers determine the

reasonableness of this fee relying on previous experience

with the provider and national databases of fees. Conflicts

between auto insurers and medical providers over non-fee

schedule charges are resolved through binding arbitration. In

Pennsylvania, fees not on the Medicare schedule are limited

to 80% of the providers’ usual and customary charge.

implementing a careful policy around reimbursement for

those services not on the Medicare schedule, Michigan can

both better contain costs and support the long-term care

facilities particularly burdened by the reform law.

In addition, we should bear in mind that Michigan’s

Catastrophic Claims Fund could be playing a much larger role

in ensuring long-term support for people catastrophically

injured in an auto accident. The fund, managed by the Michigan

Catastrophic Claims Association (MCAA), was worth $23

billion in 2020 and was designed under the old law to cover

long-term care that exceeded $600,000.

There is no reason

medical services paid for by this fund should be subject to a

low reimbursement rate—the fund was specifically designed

to cover the high costs of care that claimants with unlimited

PIP might incur. In addition, going forward, given that fewer

claimants will now have unlimited PIP, and therefore won’t

hit the $600,000 threshold, the fund could potentially be

repurposed to support long-term care providers by covering

more claimants at sustainable reimbursement rates.

CONCLUSION

One year after certain elements of the auto insurance

reform law went into effect, average insurance rates have

fallen considerably in Michigan, but remain the highest in

the country and unaffordable for nearly every jurisdiction

in the state.

To further lower rates, and ensure high auto

insurance prices don’t continue to stand as a barrier to

economic mobility, lawmakers could do the following:

OFFER GREATER CHOICE IN PERSONAL INJURY

PROTECTION COVERAGE.

While the changes to PIP made in the 2019 law clearly had

an impact on rates, Michigan is still an outlier in the amount

of PIP drivers have to purchase. More could be done to lower

rates by aligning with PIP practices in other states.

REQUIRE INSURANCE COMPANIES TO BUILD CERTAIN

FACTORS, AT CERTAIN WEIGHTS, INTO THEIR MODELS FOR

CALCULATING RATES.

In the 2019 law, the State barred insurance companies from

using a number of non-driving factors to calculate insurance

rates, only to see insurance companies use proxy measures

to replace the prohibited factors. This meant that while the

2019 law reduced rates for all Michigan drivers, it did nothing

to reduce discriminatory pricing by race. To reduce rates in

majority-Black zip codes, Michigan should look to California,

where insurance companies must give a certain weight to

three mandatory, driving-related factors in calculating their

rates: driving record, annual miles driven, and years of driving

experience.

By dictating what factors insurance companies

have to consider and the weight they must give those factors—

rather than the current practice of specifying what factors they

cannot consider—the state can limit the discriminatory impact

of non-driving factors.

REVISIT REIMBURSEMENT RATES FOR SERVICES NOT ON

THE MEDICARE SCHEDULE, AND USE THE CATASTROPHIC

CLAIMS FUND TO SUPPORT LONG-TERM CARE.

While containing the costs of medical fees is integral to

reducing auto insurance costs, the sharp reduction in

reimbursement rates for long-term care providers may drive

providers out of business, resulting in poor outcomes for

auto accident victims requiring long-term care. Lawmakers

should revisit reimbursement policies for services not

on the Medicare fee schedule, looking to models in other

states that have implemented a more nuanced approach. In

addition, lawmakers should seek to utilize the surplus in the

Catastrophic Claims Fund to restructure how that fund is used

to support long-term care facilities.

The 2019 reform law was an essential first step, but

lawmakers should not be content. When we first wrote about

Michigan’s high auto insurance rates in 2019, we emphasized

the ways in which high auto insurance rates create a barrier

to economic mobility. For many Michigan drivers, this barrier

has not been adequately diminished. More must be done to

eliminate discriminatory rate setting practices and further

curb the impact of PIP on rates. However, while we continue to

push for lower rates, we must carefully consider the impact on

those receiving long-term care and ensure that providers are

reimbursed for services in a way that enables those who have

been catastrophically injured in an auto accident to receive the

care they need.

The passage of the 2019 reform law was years in the making

and made meaningful changes that lowered rates for Michigan

drivers. Now, more than two years after the passage of the

law, we should celebrate what’s working right and fix what’s

not, to ensure Michigan’s insurance system does not stand as

a barrier to opportunity and also provides security and peace

of mind to Michigan drivers.

ABOUT THE AUTHORS

Amanda Nothaft is a senior data and evaluation manager at the

University of Michigan’s Poverty Solutions, a university-wide

initiative that aims to prevent and alleviate poverty through

action-based research.

Patrick Cooney is the assistant director of policy impact at

U-M’s Poverty Solutions.

ENDNOTES

1 The Zebra. “2021 State of Auto Insurance.” (2021). https://www.thezebra.

com/state-of-insurance/auto/2021/#introduction Some rates in this

brief, both for Michigan and Detroit, differ from rates cited in Poverty

Solutions' 2019 auto insurance policy brief. This is because the Zebra's

current year reports only use data from just over half of a calendar year.

In later reports, the Zebra revises prior year numbers after generating

an entire year's worth of quotes.

2 The Zebra. “2021 State of Auto Insurance.”; Patrick Cooney, Elizabeth

Phillips, and Joshua Rivera, “Auto Insurance and Economic Mobility in

Michigan,” (March 2019). https://poverty.umich.edu/files/2019/05/auto_

insurance_and_economic_mobility_in_michigan_2.pdf

3 Patrick Cooney, Elizabeth Phillips, and Joshua Rivera, “Auto Insurance

and Economic Mobility in Michigan,” (March 2019).

4 Patrick Cooney, Elizabeth Phillips, and Joshua Rivera, “Auto Insurance

and Economic Mobility in Michigan,” (March 2019).

5 See, for example: Richard Rothstein, The Color of Law. (Liveright, New York,

2017); Marianne Bertrand and Sendhil Mullainathan, “Are Emily and Greg

More Employable than Lakisha and Jamal? A Field Experiment on Labor

Market Discrimination,” National Bureau of Economic Research (July

2003); Sean F. Reardon, School Segregation and Racial Achievement Gaps.

The Russell Sage Foundation Journal, Vol. 2, No. 5, (September 2016).

6 We also conducted a regression analysis to understand the extent to

which the racial makeup of a zip code explains auto insurance rates in

that zip code relative to other factors, such as population density and

poverty. In all of the models, for both 2019 and 2020, race is the dominant

correlate with rates. While the explanatory power of the model increases

slightly when other factors are included, the impact of race is much

larger than any other factor.

7 Authors’ analysis of auto insurance data from The Zebra.

8 MCL, §500.3107c (2019).

9 MCL, §500.3107c (2019).

10 MCL, §500.2111(4) (2019), For a comprehensive summary of the law see:

https://www.michigan.gov/AutoInsurance

11 Patrick Cooney, Elizabeth Phillips, and Joshua Rivera, “Auto Insurance

and Economic Mobility in Michigan,” (March 2019).

12 MCL, §500.3157 (2019).

13 The Zebra. “2021 State of Auto Insurance.”

14 Patrick Cooney, Elizabeth Phillips, and Joshua Rivera, “Auto Insurance

and Economic Mobility in Michigan,” (March 2019).

15 The Zebra. “2021 State of Auto Insurance.”

16 The Zebra. “2021 State of Auto Insurance.”

17 JC Reindl, “New Michigan Auto Insurer Doesn't Use Credit Scores,

Charges Lower Rates,” Detroit Free Press, July 9, 2021. https://www.

freep.com/story/money/business/2021/07/09/michigan-cure-auto-

insurance-credit-scores/7880516002/

18 Anna Liz Nichols, “Car Crash Victims Could Lose Care Under New

Rules,” Detroit Free Press, May 29, 2021. https://www.detroitnews.com/

story/news/local/michigan/2021/05/29/car-crash-victims-lose-rehab-

coverage-michigan-insurance-reform/116678586/; Tracy Samilton,

“Some Insurance Companies Have Stopped Paying Healthcare Providers

for Auto Accident Survivors,” Michigan Radio, September 6, 2021. https://

www.michiganradio.org/law/2021-09-06/some-insurance-companies-

have-stopped-paying-healthcare-providers-for-auto-accident-survivors

19 MCL, §500.2151 (2019).

20 The correlation coefficient for both years is 0.76. Estimates for percent

Black in each zip code are from the authors’ analysis of 2019 American

Community Survey 5-year averages.

21 David E. Clark and Brad M. Cushing, “Rural and urban traffic fatalities,

vehicle miles, and population density,” Accident Analysis & Prevention,

Vol. 36:6, (November 2004):967-972. https://doi.org/ 10.1016/j.

aap.2003.10.006; Congressional Budget Office, Geographic Variation

in Health Care Spending. Publication No. CBO-2978, February 2008;

Insurance Information Institute, “2020 Insurance Fact Book,” (2021).

https://www.iii.org/search/node/2020%20Insurance%20Fact%20

Book; Janet L. Kaminsky, Territorial Rating for Auto Insurance. Office of

Legislative Research, State of Connecticut, September 2006. https://

www.cga.ct.gov/2006/rpt/2006-R-0542.htm

22 Jeff Larson, Julia Angwin, Lauren Kirchner, Surya Mattu, Dina Haner,

Michael Saccucci, Keith Newsom-Stewart, Andrew Cohen, and Martin

Romm, “How We Examined Racial Discrimination in Auto Insurance

Prices,“ ProPublica, April 5, 2017. https://www.propublica.org/article/

minority-neighborhoods-higher-car-insurance-premiums-methodology

23 Janet L. Kaminsky, Territorial Rating for Auto Insurance. Office of

Legislative Research, State of Connecticut, September 2006.

24 NJ Rev Stat § 17:29A-48 (2019).

25 Cal Ins Code § 11628(a)(2) (2012).

26 10 CA Code of Regs 2632.

27 Conn. Gen. Stat. § 38a-686 (2013).

28 Michelle Singletary “Credit scores are supposed to be race-neutral.

That’s impossible.” The Washington Post, October 16, 2020. https://www.

washingtonpost.com/business/2020/10/16/how-race-affects-your-

credit-score/

29 “The Truth About Car Insurance,” Consumer Reports, July 30, 2015.

https://www.consumerreports.org/cro/car-insurance/auto-insurance-

special-report/index.htm

30 “Personal Injury Protection Insurance (PIP)”, The Zebra, accessed

October 31, 2021. https://www.thezebra.com/auto-insurance/insurance-

guide/pip-insurance/

31 NJ Rev Stat § 39:6A-4.3 (2013).

32 NJAC § 11:3-29.4, Appendix, Exhibit 1 (2020), FL Stat § 627.736 (2012), NY

Ins Law § 28.5108 (2014), HI Rev Stat §431:10C-308.5 (2012), 31 PA Code

§69.21.

33 MCL §500.3157 (2019).

34 Carol Thompson, “I don’t think I’ve been this stressed’: No-fault

reforms upend life for Michigan’s crash victims,” The Detroit News,

July 27, 2021. https://www.detroitnews.com/story/news/local/

michigan/2021/07/28/no-fault-reforms-upend-life-michigan-crash-

victims-families/8043469002/

35 NJ Rev Stat §39.6A-4.6 (2013).

36 31 PA Code §69.21.

37 Carol Thompson, “’I don’t think I’ve been this stressed’: No-fault reforms

upend life for Michigan’s crash victims.”

38 The U.S. Treasury Department’s Federal Insurance Office deems auto

insurance “unaffordable” in areas where premiums exceed 2 percent

of a ZIP code’s median household income. See: U.S. Treasury Federal

Insurance Office, Study on the Affordability of Personal Automobile

Insurance, January 2017. https://www.treasury.gov/initiatives/fio/

reports-and-notices/Documents/FINAL%20Auto%20Affordability%20

Study_web.pdf

39 The California Code of Regulations, 10 CCR § 2632.5. h t t p s : / / g o v t . w e s t l a

w . c o m / c a l r e g s / D o c u m e n t / I 3 5 9 6 7 4 A 0 D 4 9 2 1 1 D E B C 0 2 8 3 1 C 6 D 6 C 1 0 8 E ? c o n

t e x t D a t a = ( s c . D e f a u l t ) & o r i g i n a t i o n C o n t e x t = d o c u m e n t t o c & t r a n s i t i o n T y p e =

D o c u m e n t I t e m & v i e w T y p e = F u l l T e x t