There’s No Justification for Michigan’s Discriminatory Work Requirements

This post was co-authored with Eli Savit, an attorney and adjunct professor at the University of Michigan Law School.

As we highlighted yesterday in the New York Times, a Michigan bill to impose work requirements on Medicaid recipients would have severe racially discriminatory effects.  The bill exempts people from the work requirement if they live in a county with over 8.5% unemployment.  In Michigan, though, the only counties that meet that criterion are rural and overwhelmingly white.  By and large, the state’s black population—which is concentrated in high unemployment cities like Detroit, Flint, and Muskegon—will not qualify for work-requirement exemption.[1]

Could a court strike down Michigan’s Medicaid work requirements because of these discriminatory effects?  We think so.  Michigan needs a waiver from the Department of Health and Human Services (HHS) before it can move forward with its work requirements.  If HHS grants that waiver notwithstanding Michigan’s violation of HHS’s own disparate impact regulations, the waiver could well be struck down as arbitrary and capricious.  (We’ll have more to say about this potential route into court tomorrow.)

Yes, agencies get a lot of deference when applying their own regulations, including their Title VI regulations.  For a court to invalidate the work requirement, Michigan’s Title VI violation has to be so clear-cut that HHS’s decision to grant the waiver was “arbitrary.” But a challenge to Michigan’s work-requirement bill could probably meet even that demanding standard.

Courts and agenciesincluding HHS—use a three-part test to determine whether a facially neutral practice imposes an unlawful disparate racial impact.  First, does the practice have a racially disproportionate effect?  Second, if the practice has a racially disproportionate effect, is there nevertheless a “substantial legitimate justification” for the practice?  Third, even if there is a legitimate justification, is there an alternative that will achieve the same objective, but with less discriminatory effect?

Establishing a disproportionate racial effect should be a cinch.  As things currently stand, residents in 17 rural counties would be exempt from Medicaid work requirements.  In those 17 counties, African-Americans make up just 1.7% of the population, though African-Americans make up 14.2% of Michigan’s total population.  Indeed, the total number of black people living in the seventeen exempt counties is around 5400—approximately 0.3% of Michigan’s total black population.[2]

The exempt counties are small and sparsely populated. Still, they make up 3% of Michigan’s population as a whole.  That means the average Michigan resident is ten times more likely to be a beneficiary of the exemption than the average black Michigan resident.[3]

That takes us to the second element: whether there’s a “substantial legitimate justification” for exempting people at the county level, but not the city level, from work requirements.  The co-sponsor of the Michigan legislation is already trotting out an argument to defend the county-level exemption: namely, that people should be looking for work not just in their city, but all across their county. “I don’t know why anybody in Flint would say we need to be treated separate than Genesee County,” State Senator Mike Shirkey told reporters.  “I mean, is it too much of an expectation . . . if you happen to live in Flint, to look for a job in Genesee County? And the same argument applies to Detroit and Wayne County. How granular do you want to get?”

That kind of argument ignores the lived reality of Michigan’s urban communities. Low-income residents in Michigan’s cities are significantly less able to travel for work—even to another part of the county—than people in rural communities. That’s because public transit is virtually non-existent in Michigan. In the state that put the world on wheels, you need a car to get around. (For an illustrative story of how difficult it is to get around metro Detroit without a car, check out this profile of Detroiter James Robertson, who had to walk 21 miles a day to get to and from his job in the Detroit suburbs.)

The percentage of Michigan’s urban households who lack access to cars is sky-high: around 20% in Flint and Muskegon, and over 25% in Detroit.  By contrast, in counties that would be exempt from the Medicaid work requirements, virtually everyone has a car—in those counties, car ownership rates hover at around 95%. The disparity is due to the fact that Michigan has the highest auto insurance rates in the nation, with urban rates sometimes three times as high as they are elsewhere in the state.

If Michigan really wants to encourage people to seek jobs in every geographic location they can reasonably access, this bill gets it exactly wrong. It provides an exemption to rural residents who are likely to have cars and, as a result, can travel (even across county lines) for work.  At the same time, it punishes urban residents for their inability to travel to jobs they can’t possibly access.

That’s totally illogical.  And that illogic should sink the policy.  To show that a policy serves a “substantial legitimate interest,” the state must demonstrate a “manifest demonstrable relationship” between its purported objective and the challenged policy.  A policy that supposedly exists to encourage mobile people to travel for work—but actually punishes immobile people, while giving mobile people a break—doesn’t cut it.

As for the third element, it’s not at all difficult to come up with less discriminatory alternatives that achieve the same objective.  The bill’s sponsors could recognize that poorer people in its urban centers have at least as much trouble accessing jobs as those in hardscrabble rural areas, and extend the exemption to cities with high unemployment.  It could provide the exemption by ZIP code instead of by county.  Or it could come up with a more holistic method of determining when someone should be exempt from the work requirement: one that takes into account the jobs realistically available to a person given his or her education level, access to transportation, and so forth.

The Michigan legislature’s failure to grapple with these issues suggests that, at bottom, the county-level exemption is about politics, not policy.  The elected representatives from white, rural counties that would qualify for the exemption are Republicans who support the work-requirement bill.  Yet they recognize that work requirements will harm their own constituents—and so they’re exempting them, and not other Michiganders, from the dire consequences of losing health insurance.

That’s immoral and inequitable.  And, given the bill’s severe racial disparities, it’s also unlawful.

@EliNSavit & @nicholas_bagley


[1] In March 2018—the most recent month for which data is available—the unemployment rate in Flint was 10.4%, the unemployment rate in Muskegon was 9.2%, and the unemployment rate in Detroit was 8.7%.  All unemployment data cited in this post comes from the Michigan Bureau of Labor Market Information and Strategic Initiatives.  It can be accessed online at

[2] Figures are per the 2017 U.S. Census population estimates, available at

[3] In fact, the legislation, if enacted, will almost certainly exempt far more white rural residents than in those 17 counties.  That’s because the legislation says that people become exempt from the work requirement if the unemployment rate in their county, at any time, reaches 8.5%.  But thereafter, that person remains exempt so long as the unemployment rate in the county exceeds 5.0%.

In addition to the 17 counties that currently have unemployment rates over 8.5%, there are a number of predominantly white, rural counties where unemployment typically reaches 8.5% at some point during the year.  Residents of those counties, too, would qualify for the exemption, and would maintain that exemption going forward unless the unemployment rate in their county dipped below 5.0%.



Hidden information below


Email Address*