Understanding the relationship between the judicial and legislative branches of government in the United States is fundamental to understanding not only judicial politics but policy making generally. Rather than the hierarchical relationship between court and legislature that judicial review might imply, many scholars posit an ongoing “dialogue” between the branches on constitutional and statutory issues (Dahl Reference Dahl1957; Fisher Reference Fisher1988; Graber Reference Graber1993; Katzmann Reference Katzmann1997; Rogers Reference Rogers2001; Lovell Reference Lovell2003; Barnes Reference Barnes2004; Miller Reference Miller2009; Hasen Reference Hasen2013). This idea of dialogue is consistent with Alexander Hamilton’s classic argument in Federalist no. 78 that the courts cannot act alone; they depend upon the responsiveness of the “political” branches. The strongest assertion that a court can make in this dialogue is to declare a statute unconstitutional using its powers of judicial review. When courts make such a statement, how do legislatures answer?
Scholarship has focused on Congress’s responses to US Supreme Court rulings that strike federal laws down on constitutional grounds (Pickerill Reference Pickerill2004; Devins and Whittington Reference Devins2005; Miller Reference Miller2009). J. Mitchell Pickerill, in particular, measured congressional awareness of court rulings by identifying all US Supreme Court rulings voiding a federal statute over more than 40 years and assessing whether Congress responded by amending the statute, repealing the law, passing a constitutional amendment, or simply ignoring the Court. He concluded that Congress rarely challenges the Court directly; more often, Congress amends or repeals the statute, but responses are highly variable depending on lawmakers’ commitment to the underlying policy.
Some previous work has partially addressed these issues at the state level. Langer and Brace (Reference Langer2005) argue that in abortion and death penalty legislation, state legislatures take into account the ideological preferences of their state supreme courts and refrain from lawmaking if the gap between their ideology and the Supreme Court’s is too great. Hoekstra (Reference Hoekstra2009) finds that the US Supreme Court’s preferences regarding minimum wage legislation for women in the early 20th century were influential on state legislatures, although the state supreme court’s preferences were less persuasive. Langer and Wilhelm (Reference Langer2005) surveyed state legislators and found that 65% of lawmakers said that they would rewrite or postpone policy adoption to avoid a veto from their state supreme court. Although each of these prior studies is valuable, this study has the benefit of examining actual legislative behavior through lawmaking across a much broader issue range than previously attempted.
Expanding on this scholarship, this article examines state lawmaking in response to state and federal court rulings of unconstitutionality. In three states, Michigan, Minnesota, and Wyoming, I identified all statutes declared unconstitutional by a state appellate court or federal court from 1940 to 2008 and tracked the legislative response. This study makes two chief contributions. First, it finds that the judicial and legislative branches do indeed communicate in some form about court findings, with responsiveness varying according to a number of factors, including the legislature involved, the policy subject, and the level of the court, among other considerations. Second, the study provides an initial set of data to spur hypothesis testing on legislature-court dynamics.
Methods
To keep the project manageable and to examine two important characteristics—professionalization of courts and legislatures and method of selecting judges—a case study was the best option. Michigan, Minnesota, and Wyoming exemplify varying degrees of legislative and judicial professionalization, according to prominent measures (Squire Reference Squire2007, Reference Squire2008). Greater professionalization should lead to a stronger legislative role in policy making (Rosenthal Reference Rosenthal1996; Squire Reference Squire2007; Miller et al. Reference Miller2015) and improved capacity to engage in dialogue with the courts, owing to greater staff resources to track court rulings and longer session lengths to respond. Michigan is near the top of all 50 states regarding the professionalization of legislature and courts. Minnesota is a midrange state, and Wyoming ranks near the bottom of the states both in the professionalization of its judiciary and its legislature (Squire Reference Squire2007, Reference Squire2008).
Judicial selection, which also varies among the three states, is important because different mechanisms imply different relationships involving courts, legislatures, and the public. In states with elected judges, one might expect greater contiguity between judicial and legislative preferences, with both bodies chosen by the public. Given Langer and Brace’s (Reference Langer2005) findings, one would then expect greater legislative willingness to respond to the courts, as ideological distance would be minimized. In states with unelected courts, ideological distance might be greater, leading to fewer legislative responses.
Michigan uses a form of partisan election, with party nominations for state supreme court, even though the general election is ostensibly nonpartisan. State appeals court judges are selected through nonpartisan election. Judicial politics in the state have long been contentious—in fact, the number of justices on the supreme court was reduced from eight to seven in the 1963 constitution to avoid divided rulings (Michigan Courts 2016). In recent years in Michigan, there have been multiple competitive, high-profile state supreme court races (Grand Rapids Press 2009). In Minnesota, elections for the supreme court and appeals court are nonpartisan, almost never competitive, or high profile. In Wyoming, judges were selected in nonpartisan elections before 1972, when a constitutional amendment mandated use of merit selection. As is typical for this system, retention elections in Wyoming are usually low profile (Kite Reference Kite2007).
To collect the data, I started by identifying all rulings by the US Supreme Court, US Circuit Courts of Appeals, US Federal District Courts (including US Bankruptcy Court), state supreme courts, and state courts of appeals (where present) that declared statutes in the three case study states unconstitutional during the study period. I included cases of federal preemption of a state statute because of the Supremacy Clause of Article VI. I limited the scope of the study to cases from 1940 to 2008 to capture time before and after the “reapportionment revolution” initiated when Baker v. Carr [369 U.S. 186 (1962)] transformed state legislatures into more representative and professional bodies (Moncrief and Thompson Reference Moncrief1993).
To identify the cases, I searched the LEXIS-NEXIS case law database for the state’s name and the keyword “unconstitutional.” For each result, I read the court opinion to determine whether the judges had declared a state statute legally void on its face or as applied to the litigants. I supplemented the LEXIS-NEXIS search with additional sources: the Congressional Research Service (2012), which has a list of all US Supreme Court cases declaring state statutes unconstitutional (although I identified a few that the service did not catch), and the Minnesota Office of the Revisor, a legislative office that issues biannual reports noting state laws declared unconstitutional. When a higher court reviewed a lower court’s ruling in depth (more than a summary order), I used the higher court decision, not the lower court decision, as data. Therefore, if a lower court voided a law, but a higher court reversed, the case is not included. These data also do not capture a ruling in which a court interprets a statute narrowly without declaring it unconstitutional. Nor do they encompass statutes that are likely unconstitutional but have not been ruled so by a court (e.g., the statute criminalizing flag desecration in Minnesota after Texas v. Johnson [491 U.S. 397 (1989)]).
I identified 451 cases in which a court held that a state statute was unconstitutional or preempted (188 in Michigan, 186 in Minnesota, and 77 in Wyoming). (In some cases a court declared multiple provisions of statutes unconstitutional; I treated each provision as an individual case for response purposes. In two cases a state and federal court declared the same statute unconstitutional almost simultaneously; I treated them as one case.) Then I examined the Minnesota, Michigan, and Wyoming Statutes Annotated, as well as the Minnesota Revisor’s Reports and other sources, to see whether the state legislature had acted regarding the provision/section after the court ruling, and when. I looked for legislative changes that specifically addressed the area of the court’s concern(s), not merely any change in the section. I recorded level of court, length of time that the statute had been on the books before the ruling, and length of time until the legislative response, if any. I differentiated among a repeal of the section/provision, an amendment to the section/provision, no response, and a constitutional amendment overturning the court ruling, a typology roughly following that of Pickerill (Reference Pickerill2004).
I divided the cases/statutes into 22 subject areas but allowed for multiple classifications (e.g., a First Amendment case that also dealt with the political process). The number of Minnesota, Michigan, and Wyoming laws declared unconstitutional varied by time period, with the key demarcation point being 1970 (see table A1 in the appendix). The number of laws declared unconstitutional increased significantly after 1970, mostly in expected categories. For example, abortion cases were rare before 1970 but more common afterward; the same is true for equal protection, criminal justice, and political process cases, all areas where Warren Court constitutional principles set new standards for state laws. As for time between enactment of a law and its judicial voiding, the results are similar to Pickerill’s (Reference Pickerill2004). Many laws were quickly declared unconstitutional, with the median time being 7 years (Minnesota and Wyoming) or 8 (Michigan) after enactment (see table 1). Pickerill’s data show a large number of federal laws voided soon after enactment, then a drop-off, then a spike again around 15 years after enactment. Minnesota had a similar upsurge around 15 years after enactment, but the other two states did not.
Timing of Court Rulings of Unconstitutionality after Passage of Statute

State courts issued about twice the declarations of unconstitutionality that federal courts issued, a pattern in all three states (see table 2). In the state court systems, most final declarations of unconstitutionality came from the supreme courts, unsurprisingly. The Michigan Court of Appeals began operating in 1965; the Minnesota Court of Appeals started operating in 1983; Wyoming has no intermediate court of appeals.
Statutes Declared Unconstitutional by Level of Court, 1940–2008 (Raw Numbers)

Legislative Responses
Now that we have seen the scope of the judicial declarations of unconstitutionality, we can move to the state legislatures’ responsiveness. As noted, I divided legislative responses into three broad categories, “repeal,” “amend,” or “no response” (disregarding the few cases in which state constitutional amendments nullified court rulings.)
First, on the timing of legislative responses, when legislatures acted, they did so relatively quickly, with nearly half the active responses (excluding the nonresponses) coming within 3 years of the court decisions (Wyoming’s legislature meets yearly, but nonbudget legislation is mostly confined to a 2-year cycle; see table 3). The median active response in all three states came within 4 years. In sum, 77% of the active responses happened within 10 years of the court declaration.
Timing of State Legislative Responses

Second, table 4 shows legislative responsiveness by issue area (these data are broken down by state in tables A2, A3, and A4 in the appendix). Some themes emerge from the responses (or nonresponses), as we will see. To generalize, legislatures were more likely to repeal laws when the federal government would not preempt them—that is, when the laws were more directly responsible for a purely state matter, and in cases where laws were based on biased assumptions regarding race, gender, and so forth, which were not worth trying to preserve by amendment. When a court declaration of unconstitutionality touched on a vital and ongoing area of state policy making such as health care, taxation, or election law, the legislatures were more likely to try to save the law by amendment. When the legislatures chose not to act because the issues were too contentious to resolve (as we will see in the abortion discussion), there was a practical workaround; the court ruling was either limited in time or application; federal action preempted the statute, reducing pressure to act; or the provision was not too important to lawmakers.
State Legislative Responsiveness in All Three States by Issue Area, 1940–2008

Repeals
One type involved laws that violated prohibitions on local or special legislation, particularly in Michigan, where the legislature seemed to want to treat Detroit differently from the rest of the state, only to capitulate after adverse court rulings. In Milk Marketing Board v. Johnson [295 N.W.2d 346 (MI 1940)], the Michigan Supreme Court faulted the delegation of power to set milk prices only in Detroit to private actors with a financial stake. In Klosinski v. Michigan State Board of Examiners [308 Mich. 70 (1944)], the high court voided a law setting hours and prices for barbers, in only cities with 500,000 residents or more (i.e., Detroit). In Avis Rent-a-Car v. City of Romulus [254 N.W.2d 555 (MI 1977)], the supreme court struck down the waiver of a tax exemption for rental car companies at airports, in only Wayne County (where the biggest city is Detroit). Such cases were not as common in Minnesota and Wyoming, but Board of County Commissioners for Sublette County v. Geringer [941 P.2d 742 (WY 1997)] illustrated the same pattern: the Wyoming legislature unequally funded local courthouses, the state supreme court rejected the allocation as local/special legislation, and legislators then deleted the provision.
Rulings against laws that violated equal protection often led to repeal as well. Laws that were struck down on the basis of suspect classifications (race, gender, age, etc.) were less common than laws voided for lack of a “rational basis,” but laws struck down over suspect classifications were more likely to be repealed than amended or simply left on the books. Sometimes these laws were based on biased gender assumptions, such as Michigan measures presuming women’s dependency (Day v. W. A. Foote [316 N.W.2d 712 (MI 1982)] on death benefits, and Pike v. City of Wyoming (MI) [433 N.W.2d 768 (MI 1988)] on workers’ compensation benefits). For child support, Michigan was prohibited from distinguishing children born to married couples from children born out of wedlock (Dones v. Thomas [534 N.W.2d 221 (Mich. App. 1995)]). In Minnesota, laws fell over gender limitation on the clients of barbers (Minnesota Board of Barber Examiners v. Laurance [218 N.W.2d 692 (MN 1974)]), preference for same-race adoption (Matter of Welfare of D.L. [486 N.W.2d 375 (MN 1992)]), and a requirement that pharmacists be US citizens (Wong v. Hohnstrom [405 F.Supp.727 (D. Minn. 1975)]). A Wyoming law concerning a retirement age for game wardens was voided by the US Supreme Court (EEOC v. Wyoming [460 U.S. 226 (1983)]). In all of these cases (although there was an initial attempt to rewrite the Minnesota adoption law), the lawmakers acquiesced and deleted the offending language rather than defend the differential treatment.
Amendments/Rewrites
When a court voided laws based not on the policy but on lack of adequate process, legislatures often tried to save the measures by amending or rewriting them. Thus, the Minnesota legislature added a pre-seizure hearing process for replevin cases (after Automotive Merchandise v. Smith [212 N.W.2d 678 (MN 1973)]) and rewrote civil commitment laws to insert process after Welsch v. Likins [550 F.2d 1122 (8th Cir. 1977)]. In Michigan, the legislature provided more criteria for waiving juveniles into adult court in light of People v. Fields [199 N.W.2d 217 (MI 1972)] and elaborated on the no-fault insurance laws after Shavers v. Kelley [267 N.W.2d 72 (MI 1978)]. In Wyoming, the state supreme court in Hoem v. State [756 P.2d 780 (WY 1988)] faulted the statutes regarding pretrial screening for medical malpractice cases. The legislature responded by creating a fuller process, although the court voided the new law as well.
When possible, legislators also tried to save laws regulating health care (though not laws on abortion, as we will see), likely out of recognition that health care—a core function of state government involving complex policy—is vital to large numbers of their constituents; lawmakers simply cannot abdicate their policy-making responsibility in these cases. Hoem above is an example. After a federal appeals court (Minnesota Association of Health Care Facilities v. Minnesota Department of Public Welfare [742 F.2d 442 (8th Cir. 1984)]) struck down a law limiting hospital rates charged to non-Medicaid patients and applying the limit retroactively to “excess” charges, lawmakers rewrote the law to add nuance and remove the retroactive portion. Minnesota legislators also rewrote law to clarify it and comply with federal statutes because of a court ruling over nursing-home charges for the value of patients’ unemployment benefits (New London Nursing Home v. Lindeman [382 N.W.2d 868 (Minn. App. 1986)]). After Blue Cross/Blue Shield v. Milliken [367 N.W.2d 1 (MI 1985)], the Michigan legislature rewrote a law to equalize treatment of health insurance providers selling policies to “administrative services only” consumers.
On another core function of state government, criminal justice, lawmakers frequently chose to revise statutes after unfavorable court decisions. The Minnesota legislature redefined “disorderly conduct” after Matter of the Welfare of S. L. J. [263 N.W.2d 412 (MN 1978)], although it left language in the statute about “anger, alarm, or resentment” (language that would figure in R. A. V. v. St. Paul [505 U.S. 377 (1992)]). The Michigan legislature revised the state ban on marijuana possession after People v. Sinclair [194 N.W.2d 878 (MI 1972)], and when the state supreme court voided a sentence of life without parole for possession of 650 grams of cocaine (People v. Bullock [485 N.W.2d 866 (MI 1992)]), amended the law to allow for parole for people convicted of major drug violations. In Wyoming, courts struck down a state law allowing the Motor Vehicles Division to override court-ordered suspensions of driver’s licenses of people convicted of driving while intoxicated (State Motor Vehicles Division v. Holtz [674 P.2d 732 (WY 1983)]), prompting the legislature to remove the ability to reverse the suspensions.
Also in Wyoming, the legislature and the courts engaged in dialogue over the relationship between the state’s vehicular homicide law and the prohibition against involuntary manslaughter. The courts originally held that the two statutes were reconcilable (State v. Cantrell [186 P.2d 539 (WY 1947)]), but in 1977, the Wyoming Supreme Court ruled that the subjects so overlapped that all vehicular homicides short of murder or voluntary manslaughter could be prosecuted only as involuntary manslaughter, rendering the vehicular homicide statute superfluous (Thomas v. State [562 P.2d 1287 (WY 1977)]). The legislature quickly responded by amending both statutes, and later adding a crime of “aggravated vehicular homicide” [Wyoming Session Laws chap. 139 (1979); Wyoming Session Laws chap. 75 (1982)]. Eventually, the revised statutes were challenged, with the key language in the vehicular homicide law being, “Whoever, except when the violation of law involves culpable neglect or criminal carelessness … but with a conscious disregard of the safety of others” [Wyoming Statutes 1977, 31-5-1117 (Cum. Supp. 1982)]. The supreme court ruled that the second clause of the sentence canceled the first, leading to unconstitutional ambiguity (State v. Sodergren [686 P.2d 521 (WY 1984)]). Swiftly, the legislature revised the law to substitute a more straightforward standard for vehicular homicide, that of criminal negligence in operating a vehicle and being the proximate cause of another’s death [Wyoming Session Laws chap. 44 (1984)]. The current vehicular homicide statute (found at W.S. 1977, 6-2-106) seems to have resolved the issue.
Election law is yet another matter for which law makers have an incentive to work with courts: the need to provide for stable rules. In Minnesota, when a federal judge voided the state’s 6-month residency requirement for voting (Keppel v. Donovan [326 F.Supp. 15 (D. Minn. 1970)]), the legislature did not eliminate the requirement, instead reducing it to 20 days. In Michigan, after the Socialist Workers’ Party successfully challenged the ballot access requirement of 100 signatures each in at least nine congressional districts (Socialist Workers Party v. Secretary of State [317 N.W.2d 1 (MI 1982)]), the legislature altered the provision to one hundred in at least half the districts. As for Wyoming, after a federal judge voided a host of state election laws in Blomquist v. Thomson [591 F. Supp. 768 (D. Wyo. 1984)], finding they improperly excluded nonmainstream political parties, the legislature replied in part by giving minor parties some of the benefits given to major parties, such as possible service as election judges and being allowed to file for special elections to fill legislative vacancies. Finally, in all three states, courts voided the legislature’s apportionments in response to the decennial census, leading to lawmakers passing new schemes, either by statute or, in Wyoming, by constitutional amendment to comply with the standard of “one person, one vote.”
Sometimes the judicial-legislative dialogue is implicit, as it was in Minnesota, where in 1988 a federal court barred enforcement of a statute limiting media exit polling during elections that stated: “No one, either inside a polling place or within 100 feet of the entrance to it, shall ask a voter how the voter intends to vote or has voted on any office of question on the ballot” (CBS, Inc., ABC, Inc., et al. v. Growe 15 Med.L.Rep. 2275 [D. Minn. 1988]). The law, the court ruled, violated the First Amendment. A year later, the legislature attempted to cure the constitutional problem by rewriting the law to bar anyone not a voter or poll worker from standing within 100 feet of a polling place. In 1993, the legislature clarified this statute to ensure that the restriction applied only to the area where polling was being conducted, not the entire building in which a polling place might be located (Minnesota Session Laws 1993, c. 223, Sec. 12). In 2008, however, the federal district court voided the new law, ruling that the revised statute was not narrowly tailored to preserve the First Amendment rights of media pollsters (ABC, Inc. et al. v. Ritchie et al., 2008 US Dist. LEXIS 83909; 36 Media L. Rep. 2601 [D. Minn. 2008]). Finally, in 2010, the legislature, although retaining the expanded definition of “polling place” as the entire building, gave into the courts by explicitly allowing media exit polling (Minnesota Session Laws 2010, c. 201, Sec. 35).
Taxation was another subject on which amendment and dialogue generally prevailed. In Michigan, the Headlee Amendment was a voter-approved constitutional provision that limited state taxes and prevented the legislature from shifting costs to local governments (Mich. Const., Art. 9, Sec. 29). This amendment has been frequently challenged, especially over education funding. In 1997, the Michigan Supreme Court ruled that the legislature had imposed unfunded mandates on school districts regarding special-education funding (Durant v. State (Durant I) [566 N.W.2d 272 (MI 1997]). The legislature responded by paying a settlement to the plaintiff districts and appropriating funds to nonsuing districts in return for an agreement to not sue the state (Michigan Public Act 142 of 1997 and Public Act 339 of 1998). The plaintiffs, though, filed suit again, claiming that the legislature had imposed improper conditions on the state money. The Michigan Court of Appeals agreed, and the state supreme court declined to hear an appeal (Durant v. State (Durant II), [605 N.W.2d 66 (Mich. App. 1999)]). In 2000, the legislature restructured the aid money (Michigan Public Act 97 of 2000), only to face a court challenge again; this time the new mechanism was upheld (Durant v. State (Durant III) [650 N.W. 2d 380 (Mich. App. 2002)]). Also see Adair v. State (Durant IV) [680 N.W. 2d 386 (MI 2004)]).
All of these areas such as health care, criminal justice, election law, and taxation, where the legislatures chose to amend instead of repealing or not responding, were vital and complicated enough that the legislature could not reasonably abdicate policy making.
Nonresponse
In 117 cases, state legislatures took no action after a statute was declared unconstitutional—likely because, in some instances, the law was not significant enough to lawmakers. In other cases, changes in the underlying facts made further action less likely. For example, in Northshor Experience v. City of Duluth [442 F.Supp.2d 713 (D. Minn. 2006)], the Minnesota legislature had passed a law regulating adult entertainment establishments, which a federal court found infringed on First Amendment rights. A few years later, the proprietor of the club at issue died, leading to the establishment’s closure and sale without reopening. Because there was no longer a controversy, the legislature was not under immediate pressure to act. A similar instance arose in the mid-1990s, when Hornell Brewing sold Crazy Horse Malt Liquor and ran afoul of the Minnesota legislature, which found the name offensive to Native Americans and passed a law restricting depictions of Native American leaders as part of liquor licensing. Hornell successfully challenged the law on First Amendment grounds in state court (Hornell Brewing v. Minnesota Department of Public Safety [553 N.W.2d 713 (MN App 1996)]); Congress got involved, passing a law against the brand that was also invalidated on constitutional grounds. The controversy waned in 2001 when Stroh’s Brewery Co. bought the rights to the brand and ended production of the beer.
In a few cases, legislative changes superseding (but not responding to) the court ruling likely explained the lack of response. In Michigan Injured Workers v. Blanchard [647 F.Supp. 571 (W.D.MI 1986)], workers disability compensation amendments denied plaintiffs access to courts in violation of due process, according to a federal court, but before the ruling, a new worker’s compensation system was already scheduled to take effect. Also in Michigan, in United Insurance v. Attorney General [300 Mich. 200 (1942)], the legislature passed a law on insurance regulation without an effective date, adjourned, and then upon reconvening passed a resolution making the law effective immediately, excluding the Governor. The Supreme Court of Michigan declared this process unconstitutional, but the law went into effect in less than 3 months anyway; apparently there was no dispute about the substance of the provisions. In Minnesota (Naftalin v. King [257 Minn. 498 (1960)]), the legislature redirected money away from a state building fund in violation of the state constitution, according to the state Supreme Court, but the justices refused to order that the money be returned to the fund (and so it was not).
Timing was also an issue in cases in which a court did not strike down text in a statute but ruled that retroactive application of the law was unconstitutional (see MONY Life Insurance v. Ericson [533 F.Supp.2d 921 (D.Minn.2008)] in Minnesota and Cloverdale Equipment Co. v. Manitowoc Engineering Co. [964 F.Supp. 1152 (E.D.Mich.1997)] and Wells v. Kent County Board [382 Mich. 112 (1969)] in Michigan). In Wells, for example, the law provided for an “incumbent” designation on the ballot for municipal or associate municipal judges running for the new office of district judge. The Michigan Supreme Court ruled that this designation misled voters and violated the “purity of elections” clause of the state constitution- but the offending statute applied only to the 1968 election, and not future elections.
On the application of statutes without regard to timing, there were 17 cases in which a court did not void statutory language but held that the statute was being applied unconstitutionally to the litigants. Since these determinations tended to be case-by-case, the legislatures did not take action to adjust the underlying statutory language. An example is People v. DeJonge [442 Mich. 266 (1993)], in which the Michigan legislature since 1921 had required all home-school teachers to be state-certified; the state supreme court in 1993 ruled that the law could not be constitutionally applied to parents with strong religious convictions against certification, an outcome that lawmakers did not contest. Another is U.S. v. Michigan [851 F.2d 803 (6th Cir. 1988)], in which a federal appeals court ruled that a state sales tax could not be applied to federal credit unions. Especially because federal credit unions were likely not a large source of state revenue, the legislature remained silent. Cases involving civil liability and the minimum contacts required for being sued in state court also fit into this very fact-intensive category (e.g., Uppgren v. Executive Aviation Services [304 F.Supp. 165 (D.Minn. 1969)], as well as City Suburban Agency v. Dade Helicopter Services [141 Mich.App. 241 (1985)] and Attorney’s Liability Insurance Society v. Fitzgerald [174 F.Supp.2d 619 (W.D.Mich. 2001)]).
Action at the federal level often provided a reason for the state legislatures to refrain from action. In 1977, the Minnesota legislature regulated chlorofluorocarbons (CFCs) in aerosol cans, and in 1978, a federal appeals court held that federal law preempted the measure (Cosmetic, Toiletry, and Fragrance Association v. Minnesota [575 F.2d 1256 (8th Cir. 1978)]—but later that year, the federal government banned CFCs, eliminating the need for the legislature to act. Nor did Minnesota lawmakers act after litigation over their requirement of a slow-moving vehicle tag for Amish buggies. The state supreme court (State v. Hershberger [462 N.W.2d 393 (MN 1990)]) ruled the measure violated religious freedom, yet not long afterward, the US Supreme Court held (Employment Division v. Smith [494 U.S. 872 (1990)]) that the state must clear a lower bar to survive a free exercise challenge. (The case then went back to the state court, which ruled that the statute violated the Minnesota Constitution.) US justices also favored the legislature over lower courts in in re Welfare of C.T.L. [722 N.W.2d 484 (Minn.App. 2006)] in Minnesota on DNA samples from arrestees (see Maryland v. King [133 S.Ct. 1958 (2013)] at the federal level) and Americans United for Separation of Church and State v. Grand Rapids School District [546 F.Supp. 1071 (W.D.Mich. 1982)] in Michigan, regarding public school teachers’ teaching of secular subjects in religious schools. In the latter case, the legislature left the statute in place despite the ruling and was vindicated 15 years later when the US Supreme Court upheld a similar provision in Agostini v. Felton [521 U.S. 203 (1997)].
Later court decisions putting a different spin on (but not overruling) previous court rulings of unconstitutionality were particularly prominent in Michigan. For example, in Brown v. Manistee County Road Commission [452 Mich. 354 (1996)], the Michigan Supreme Court ruled that different time limits for suing a county v. suing the state for poor road maintenance were irrational and violated equal protection principles. The legislature took no action in response, and 11 years later, the Court largely changed its mind in Rowland v. Washtenaw County Road Commission [477 Mich. 197 (2007)], strongly hinting that the unequal statutes of limitation could be legitimate. Also related to mixed signals from the courts, in Witzke v. Withrow (702 F.Supp. 1338 [W.D. Mich. 1988]) the Western District Court of Michigan ruled that revocation of parole for a convicted criminal required a parole board hearing to satisfy constitutional due process. The Eastern District Court of Michigan, however, ruled 3 years later that the plaintiff should have pursued a habeas corpus appeal (Caley v. Hudson [759 F.Supp. 378 (1991)]). The district courts divided, and the legislature did not act.
There would be less of a need for the legislature to deal with an unconstitutional statute if there were a practical work-around. For example, state courts might reinterpret a statute to make it constitutional in practice. In Detroit Metro Welfare Rights Organization v. Calahan (unpublished memorandum opinion of a special three-judge panel of the US District Court of the Eastern District of Michigan issued May 29, 1970 [Civil Action No. 34006)], a federal court cited the First Amendment in striking down a law against “(exciting) any disturbance or contention” at public meetings. When judges instructed juries, however, they left out the “or contention” language, leaving just “disturbance,” a solution proposed in People v. Mash [206 N.W. 2d 767 (Mich.App. 1973)] at the Michigan Court of Appeals; therefore, convictions would not be invalidated because of unconstitutionality. In State v. Cannady [727 N.W.2d 403 (MN 2007)], the Minnesota Supreme Court ruled that a provision in a child pornography law unconstitutionally shifted to the defendant the burden to prove that everyone in the offending material was 18 or older. Yet the court opined that the provision was severable from the rest of the statute; furthermore, the case involved “harmless error” despite the unconstitutionality, meaning that convictions were not overturned. Judicial workarounds also figure in cases regarding reapportionment. In Michigan, a federal court in the 1970s (Dunnell v. Austin [344 F.Supp. 210 (E.D.Mich. 1972)]) and the state supreme court in the 1980s (Anderson v. Oakland County Clerk [350 N.W.2d 232 (MI 1984)]) struck down the legislature’s maps. Lawmakers did not act after either case, and in both cases, the Michigan Supreme Court ended up drawing the maps.
Regarding nonjudicial workarounds, in Batinich v. Harvey [277 N.W.2d 355 (MN 1979)], the Minnesota Supreme Court ruled that a statutory provision allowing landowners to vacate or alter their plats violated due process for neighboring property owners—but the state attorney general had already opined that a different statute could apply without raising constitutional concerns. In Buback v. Romney [156 N.W.2d. 549 (MI 1968)], Michigan state law allowed the attorney general to conduct a hearing before a probate judge for removal of a county official; the state supreme court said that the judiciary was being involved unconstitutionally, yet there were alternate ways of removing county officials, such as gubernatorial action or popular recall, which would reduce pressure to alter the offending provision. Finally, in cases involving federal preemption of state law, the federal law would provide the fail-safe policy, lifting the burden on the legislature to fill a policy void. Data support this conclusion, as we will see, showing that in preemption cases from the three states, the legislatures were less likely to actively respond.
If no statutory work-around were available, the bluntest solution would be to ignore the unconstitutional law. In Wilkins v. Ann Arbor City Clerk [189 N.W.2d 423 (MI 1971)], the Michigan supreme court voided a measure prohibiting college students from voting at their school addresses, ruling that it violated the students’ due process rights. The legislature has taken no action to alter or remove the provision, but students regularly vote at their college address anyway, as long as it is their only address on file with election officials.
Abortion policy provides another dynamic of nonresponse: legislatures may decide to leave laws on the books not out of neglect but because they either substantively disagree with the court, or because they cannot agree on a response even after significant debate. After Roe v. Wade was decided in 1973 by the US Supreme Court, the Minnesota legislature quickly responded, passing in 1974 a measure imposing various abortion limitations based on “potential viability” of the fetus halfway through the pregnancy. A federal court interpreted viability as falling after 20 weeks, whereas in Roe the justices had discussed a viability mark of 24–28 weeks. The Eighth Circuit Court of Appeals in Hodgson v. Lawson [542 F.2d 1350 (8th Cir. 1976)] voided much of the Minnesota law, which required a physician certification of the necessity of abortion for maternal life or health after 20 weeks, as well as a rather vague provision stating that the abortion needed to be performed to reasonably assure, consistent with good medical practice, the live birth and survival of the fetus outside the mother. The voided provisions have remained on the books with no significant changes. In 1981, lawmakers chose another avenue for regulation, adopting a requirement of two-parent notification for a minor to have an abortion. The US Supreme Court ultimately ruled that the requirement infringed on the girl’s abortion rights (Hodgson v. Minnesota [497 U.S. 417 (1990)]). This law, too, has remained on the books, with no substantive changes.
The battles over abortion in the state have mostly shifted to the issue of public funding. In 1979, Minnesota abortion-rights opponents persuaded lawmakers to pass a law denying state funds to any group providing abortion services, with exceptions for hospitals and health maintenance organizations (HMOs). This law was primarily aimed at Planned Parenthood, which challenged the statute in federal court. Again the Eighth Circuit rebuffed the legislature, ruling (Planned Parenthood v. State [612 F.2d 359 (8th Cir. 1980)]) that the differential treatment of hospitals, HMOs, and Planned Parenthood lacked a “rational basis” and thus violated equal protection. As with the post-Roe statute, however, the legislature has left the exclusion on the books with no further action, even though the Minnesota Supreme Court in Doe v. Gomez [542 N.W.2d 17 (MN 1995)] overturned the 1978 law, ruling that the denial of funds violated the woman’s state constitutional right to privacy.
In each legislative session since 1995, statutes to restrict abortion funding and state constitutional amendments have been proposed to overturn Gomez. The legislative battles were fought to a draw, and no broad legislation resulted, but in 2005, with Republican Governor Tim Pawlenty in office, the legislature passed the Abortion Alternatives Act, providing state funds for groups supporting alternatives to abortion—excluding Planned Parenthood and similar organizations. In 2011 and 2017, abortion opponents in the legislature passed a ban on abortion funding (but not a constitutional amendment); as expected, Democratic Governor Mark Dayton vetoed both bills (Lopez Reference Lopez2017). Even though legislation was not enacted, the lack of a successful law should not be mistaken for apathy.
In Michigan lawmakers’ responses to court rulings on abortion, we can see more successful legislation but also some nonresponse. In 1978, Michigan lawmakers delegated to a state agency the authority to promulgate regulations for birth control/abortion clinics. These regulations were declared unconstitutional in Birth Control Centers v. Reizen [743 F.2d 352 (6th Cir. 1984)]. Although it took the legislature some time to respond, in 1999 it left the statutory provisions on the books but delegated to the agency authority to make exceptions to the clinic regulations to address the ruling. This is where the matter stayed until 2012, when the legislature passed a requirement that clinics possess outpatient surgical centers, a provision similar to the Texas measure that the US Supreme Court recently voided in Whole Women’s Health Clinic v. Hellerstedt [136 S.Ct. 2292 (2016)]. The legislature has not immediately responded to this ruling.
Like Minnesota, Michigan has debated issues of funding for abortions. In 1988, voters passed an initiative barring funds for abortions for victims of rape or incest. The Sixth Circuit in Planned Parenthood v. Engler [73 F.3d 634 (6th Cir. 1996)] ruled that this law was preempted as going even further than the Hyde Amendment, which was changed in 1993 to allow funding for victims of rape or incest. Lawmakers have left the law on the books, however, and 2 months after the Sixth Circuit ruling, they closed a loophole that allowed HMOs to refer patients to abortion providers for “family planning” services covered by Medicaid.
In Wyoming, there has been only one abortion-related case in which a state statute was voided: in Doe v. Burk [513 P.2d 643 (WY 1973)], the state supreme court declared Wyoming’s abortion law, which dated to 1890, was unconstitutional (the law was similar to the Texas law struck down in Roe). The legislature responded in 1977 by amending the statutory ban, but it still barred abortions after viability unless there was a substantial and imminent danger to maternal health. This law would likely fail the “undue burden” test of the US Supreme Court’s Planned Parenthood v. Casey [505 U.S. 833 (1992)] decision, but no appellate court has ruled as such, and the legislature has not further addressed the issue.
In the three states, lawmakers made deliberate choices not to act regarding abortion or tried to act but failed to achieve consensus. More common, though, were cases in which statutes were voided on their faces, but neither legislature nor court took any further action—a description fitting about 30 cases, or one-fourth of the total. In 20 other cases, the legislature amended the provision without addressing the court ruling. Consider Armour and Co. v. Ball [468 F.2d. 76 (6th Cir. 1972)]. The Michigan legislature had specified higher standards for meat content in sausage than provided for in federal law, and a US appeals court ruled in 1972 that the state standard was preempted. Since then, state lawmakers have amended the statutory provisions regarding definitions of meat and meat processing three times, but the language on sausage meat content has not been touched. Also consider Toll Northville Limited v. Township of Northville [743 N.W.2d 902 (MI 2008)]. The Michigan legislature had attempted to classify public service improvements as “additions” that would avoid state constitutional limits on taxation (the Headlee Amendment). Rejecting the legislature’s claim, the Michigan Supreme Court voided the provision. In 2014, the legislature addressed the general section of the law, even amending it in response to another court declaration of unconstitutionality (see WPW Acquisition Company v. City of Troy [643 N.W.2d 564 (MI 2002)]) but making no change to the language on public service improvements.
Then there were cases in which the legislature might have been either unaware of the court ruling on the untouched provisions or simply chose to not act. In a vivid example, the Wyoming Supreme Court (Local Union #415 of Int’l Brotherhood of Electrical Workers v. Hansen [400 P.2d 531 (WY 1965)]) voided a right-to-work statute. The court opined: “We are disposed to believe that had the legislature’s attention been called to the usurpation of power contained in § 5, it would have removed it from the act and enacted the remainder.” Despite that vote of confidence in lawmakers, the provision has remained for more than 50 years.
Similarities/Differences among the States
The least professionalized of the legislatures, Wyoming and Minnesota, were the most likely to respond to a court declaration of unconstitutionality (see table 5). Lawmakers repealed or amended unconstitutional legislation in 86% of the cases in Wyoming, 70% of cases in Minnesota, and 66% of the cases in Michigan, compared to Pickerill’s (Reference Pickerill2004) measure of 61% measurement for Congressional responsiveness.
State Legislative Responses to Court Rulings of Unconstitutionality

One might think that Michigan legislators would be more responsive to state court decisions of unconstitutionality because judges have to face the electorate in higher profile contests (essentially partisan elections) than Minnesota and Wyoming judicial elections. The data do not bear this assumption out, however, as Minnesota lawmakers were more responsive to judges elected in nonpartisan races, and Wyoming legislators were more responsive to judges selected through a Missouri Plan. (As mentioned, Wyoming changed its method of judicial selection in 1972 from nonpartisan election to the Missouri Plan. The legislative response rate before and after the change is virtually identical, however.)
What explains the unexpected pattern of response and nonresponse in these three states? One possibility involves mechanisms of communication between courts and legislatures (Katzmann Reference Katzmann1997), an aspect of legislative professionalism. These mechanisms vary widely among the states (for an in-depth look, see Abrahamson and Hughes [Reference Abrahamson1991]). Abrahamson and Hughes (Reference Abrahamson1991) focus on judicial-legislative communication in statutory interpretation cases (also see Widman Reference Widman2014), but their discussion is relevant to constitutional cases. In some states, the attorney general’s office takes the lead in relaying significant court opinions to the legislature; in others, the legislature has an advisory body of lawmakers and nonlawmakers that identify court opinions that recommend legislative reconsideration of an issue; and in others, the judiciary has a more formal role in communicating directly with the legislature through state-of-the-judiciary speeches and advisory opinions, among several means (Abrahamson and Hughes Reference Abrahamson1991).
In Minnesota, the Office of the Revisor, in addition to drafting all legislation, collects court opinions identifying possible legislative action, reports summaries of those opinions, and maintains a weblink identifying all extant statutory provisions declared unconstitutional (https://www.revisor.leg.state.mn.us/statutes/unconstitutional.php). In some years, the office recommended solutions to ambiguous wording but, due to legislative inaction on the recommendations, no longer suggests fixes (Minnesota Revisor of Statutes, “Report Concerning Certain Opinions of the Supreme Court,” 2 [Nov. 1988]). Occasionally the legislature will pass a “revisor’s bill,” making technical corrections in the law but also repealing unconstitutional legislation (e.g., Minnesota Legislature, Session Laws 2008, chap. 277).
In Michigan, according to Abrahamson and Hughes (Reference Abrahamson1991), communication about court opinions comes from at least two sources, the attorney general’s office and the Michigan Law Revision Commission. Composed of lawmakers and nonlawmakers, the commission does not perform the same bill drafting function as the Office of the Revisor; its annual reports focus on policy recommendations, with only notable court rulings mentioned for possible legislative action. The Commission describes itself as underfunded compared to similar bodies in other states, with just one permanent part-time staff member (Michigan Law Revision Commission, Annual Report 2006–7).
In Wyoming, Abrahamson and Hughes (Reference Abrahamson1991) describe legislative-judicial communications as informal; lawmakers and judges speak among themselves rather than communicate through an institutional mechanism. In recent years, Wyoming’s legislative research capacity has expanded, although it is still centered in the Legislative Service Office; legislators have no publicly paid staffers (http://legisweb.state.wy.us/LSOWEB/LegislativeServiceOffice.aspx). Also, although the office will do research for lawmakers, they issue no formal report about notable court rulings.
Given these institutional capacities for legislative-judicial communication, one might expect Wyoming to be the least responsive state and Minnesota the most responsive. In fact, Wyoming, like Minnesota, has a high response rate, whereas Michigan, with a medium level of formal communication, is at the bottom for responsiveness. In the most responsive state, Wyoming, the attorney general seems to regularly communicate with the legislature, through testimony to the Judiciary Committees, official opinions, or even newspaper articles (e.g., Baker Reference Baker2013). A legislative source said that after the US Supreme Court’s ruling in Miller v. Alabama [567 U.S. 10-9646 (2012)], Attorney General Greg Phillips worked with lawmakers to revise the state’s death penalty statute for those who committed their crime while a juvenile, even though Wyoming was not a formal party to the case. Additionally, according to this source, lawmakers have a close relationship with the Legislative Service Office, which they would trust to deliver news of any court ruling voiding a state law.
Perhaps informal communication, as in Wyoming, is an effective means of relaying judicial preferences to the legislature (for a discussion of informal communication, see Langer [Reference Langer2005]). Recently, US Senator Mike Enzi (R-WY) related his relationship with a retiring Tenth Circuit judge: “[Terence O’Brien] and I, when he was a Wyoming district court judge and I was in the Wyoming state Senate, used to have dinner together to solve the world’s problems” (Barron Reference Barron2013). Also, an anonymous Wyoming legislative interview source confirmed the almost-inevitable informal interactions between officials, whether legislative or judicial, in the small town environment of the capital, Cheyenne (population: 60,000).
In Minnesota, the Office of the Revisor’s role in communicating with lawmakers about unconstitutional statutes likely helps responsiveness, although it is not clear whether the state attorney general is as involved with the legislature as in Wyoming. In Michigan, the institutional actors, whether the state attorney general or the Law Revision Commission, seem to have less effect with the legislature. Michigan lawmakers have paid personal staff—multiple staffers in the case of senators. Staffers may not usually be tasked with tracking court decisions, though. Perhaps more professionalized legislatures with broader agendas are more distracted from responding to judicial decisions, many of which may be technical and unexciting to the legislative imagination. Still, it is remarkable that Wyoming, with only 60 official legislative days every 2 years and presumably many issues that must be speedily resolved, spends the time and resources it does responding to court decisions.
The data also show that legislatures were more responsive to state courts than federal courts after a finding of unconstitutionality. When a federal court decided a case, lawmakers repealed or amended the statute 56% of the time in Michigan and 60% of the time in Minnesota; when a state court decided a case, those figures were 74% and 80%. Wyoming was highly responsive to both levels of courts almost equally (see table 6).
Repeal/Amend/Constitutional Amendment Responses by Level of Court

The federal courts were much more likely to find federal preemption of state law than were the state courts (37 cases as opposed to 11). Legislatures in Michigan and especially Minnesota responded more tepidly to preemption findings than declarations of unconstitutionality, with only a 42% active response rate as opposed to the overall active response rate of 74%. Excluding the preemption cases, the overall response rate to federal court rulings of unconstitutionality was only 64%. Therefore, the legislatures in Minnesota and Michigan were substantially more likely to respond to decisions coming from their own state courts.
This surprising finding might be explained by more communication, formal or informal, between state courts and state lawmakers than between federal judges, who may be more removed, and state lawmakers. On the other hand, federal district court judges still live and work in the same state as the legislature. Also, one would imagine that a federal court ruling striking down a state statute as unconstitutional, an infrequent event, would be more likely than a state supreme court ruling to gain media attention.
Another possibility is that federal courts are handing down rulings that state courts are avoiding over the fear of local unpopularity. A significant body of state judicial research finds that elected judges decide politically charged cases (i.e., death penalty cases) differently in reelection years than in years when they are not on the ballot (Hall Reference Hall1992; Brace and Boyea Reference Brace2007; Canes-Wrone et al. Reference Canes-Wrone2014). In turn, these unpopular rulings would be less likely to spur legislative action to comply with the court’s decision; see, for example, the earlier discussion about legislators having left offending laws on the books after courts voided restrictions on abortion.
Conversely, the Minnesota Supreme Court was willing to find constitutional flaws in state procedures for sex offenders (State v. Shattuck [704 N.W.2d 131 (MN 2005)], the Michigan Supreme Court struck down the state’s legislative apportionment (Scholle v. Secretary of State [104 N.W.2d 63 (MI 1960)], and both the Wyoming and Michigan Supreme Courts declared their state’s education funding system unconstitutional (Governor v. State Treasurer (also known as Milliken v. Green), [203 N.W.2d 457 (MI 1972)]; Washakie Sch. Dist. #1 v. Herschler [606 P.2d 310 (WY 1980)]. These rulings cannot have been popular within their states, but the high courts were willing to issue them.
Perhaps the best explanations for greater state legislative willingness to respond to state courts combine lack of ability to respond to federal preemption findings and resentment against federal courts for their supervisory role (Whittington Reference Whittington2005).
Discussion
These data support the thesis that legislatures and courts are communicating with each other, with variation among states and between state and federal governments. In the case study states, we see legislative willingness to change statutes by amending or repealing them, often shortly after a court ruling of unconstitutionality. Lawmakers in Minnesota and Michigan created offices (the Minnesota Office of the Revisor and the Michigan Law Revision Commission) charged with identifying court rulings of unconstitutional statutes and communicating those to the legislatures. In some cases, mostly from Minnesota, the legislature was explicit about its goal of responding to a court ruling, including but not limited to, revisor bills. For example, in Hoene v. Jamieson [182 N.W.2d 834 (MN 1970)], a case involving sales taxes on state purchases for highways, the Minnesota legislature explicitly noted in its response repealing the offending provision that it was correcting the law, even 14 years after the state supreme court ruling (Minnesota Session Laws 1984, chap. 525, sec. 2.)
If we say that there is dialogue between the legislative and judicial branches, we nonetheless face a number of questions. When lawmakers negotiate with judges, is the bargaining different from the bargaining between Congress and the President (Cameron Reference Cameron2000)? Does it matter that the court’s goals are likely to be more process-oriented than the legislature’s (and executive’s), which might focus more on policy outcomes? Does the ability of the federal courts to veto-bargain differ from that of the state courts? If so, how? On which policies are legislatures more likely to be responsive to the courts (this article suggests health care, criminal justice, and election law), and on which are they less likely? A goal of this article is to contribute to the work that others (Pickerill Reference Pickerill2004; Langer and Brace Reference Langer2005; etc.) have begun in setting a baseline measure for legislative responsiveness. What causes some state legislatures to go “above” the baseline for responsiveness and others to go below? If legislative professionalization and method of judicial selection do not explain responsiveness, are there alternative explanations? Why are some legislatures more talkative than others? Further research will have to answer these questions.
Appendix
Statutes Declared Unconstitutional in All Three States by Subject Area and Decade, 1940–2008

Michigan Legislature’s Responses by Issue Area, 1940–2008

Minnesota Legislature’s Responses by Issue Area, 1940–2008

Wyoming Legislature’s Responses by Issue Area, 1940–2008











