Introduction
In June 2017, the Supreme Court of Iowa ordered a new trial in Tina Haskenhoff’s employment discrimination lawsuit against her former employer, Homeland Energy Solutions, LLC.
Haskenhoff alleged sexual harassment and retaliation under the Iowa Civil Rights Act (ICRA).
The trial court had instructed the jury that, to prevail on her retaliation claim, Haskenhoff needed to prove that her complaints about sexual harassment “played a part” in Homeland’s adverse employment decision.
The jury returned a verdict for Haskenhoff, which the high court vacated.
The court held that while the “motivating factor” causation standard in the instruction may have been proper for status-based discrimination claims such as sex discrimination, retaliation claims required the higher “significant factor” standard.
Had the court treated the claim as the U.S. Supreme Court treats retaliation claims under Title VII, it would have applied yet another standard: but-for causation.
The facts surrounding Haskenhoff’s employment and trial illustrate why these competing causation standards matter. Haskenhoff’s supervisor, Kevin Howes, engaged repeatedly in grossly inappropriate behavior, including commenting on Haskenhoff’s body, speculating out loud what it would be like to have sex with her, simulating sexual behavior in front of her, and suggesting that she was getting married for money.
After this last incident, Haskenhoff called Howes expletives in front of her coworkers, went home early without permission, and then later sent Howes an email indicating her disgust.
Howes prepared a written warning for Haskenhoff for leaving work early and indicated that he wanted to terminate Haskenhoff for insubordination.
In the meantime, Haskenhoff complained of sexual harassment to the company’s management, which determined that Howes had made inappropriate comments in the workplace.
A day after Howes interrupted a meeting to present Haskenhoff with a draft of a performance-improvement plan, Haskenhoff resigned.
On the instruction that Haskenhoff need only prove that her complaints “played a part” in the adverse employment decision, the jury returned a verdict for Haskenhoff on the retaliation count.
But one can see why it might be more difficult to prevail under a higher causation standard: She left work early without permission, openly used expletives to refer to her supervisor, and was under investigation for further evidence of insubordination.
If this conduct turned out to be an independently sufficient reason to fire her, Haskenhoff would likely not prevail under a but-for causation standard.
The court in Haskenhoff tackled a question that now confronts many state courts: In the wake of a pair of recent Supreme Court decisions applying but-for causation to Title VII retaliation
and Age Discrimination in Employment Act (ADEA)
claims, how should state courts interpret similarly worded state statutes?
Like their federal analogues, most state employment discrimination statutes outlaw making certain employment decisions “because of” an individual’s protected characteristics or activities.
State courts looking to the U.S. Supreme Court for persuasive authority
now find two distinct interpretations of “because of” in the case law interpreting federal employment discrimination statutes.
State courts have exhibited a variety of responses, ranging from automatic application of the new federal standards
to completely independent construction of state statutes.
Recently, a growing number of state appellate and supreme courts have applied the but-for requirement to state statutes, with varying degrees of attentiveness to the question of exactly how much deference state courts interpreting state statutes owe to the Supreme Court’s interpretations of analogous federal statutes.
This Note surveys and analyzes these state decisions in the context of debates about parallel construction and what has sometimes been called the “new” judicial federalism.
Although state courts are not always explicit about all the considerations that enter into the decision of whether to follow the Supreme Court’s construction of a federal statute when construing a state analogue, the various arguments for either independent or parallel construction provide a framework for assessing these state-court decisions.
Part I of this Note provides an overview of current causation doctrine in federal employment discrimination law. Part II analyzes various state-court decisions addressing the question of whether to follow Gross v. FBL Financial Services, Inc. and University of Texas Southwestern Medical Center v. Nassar in applying but-for causation requirements to state employment discrimination statutes, including age discrimination and retaliation statutes analogous to those at issue in Gross and Nassar. From this survey, the Note concludes that, whether they decide to adhere to or depart from the Gross –Nassar standard, state courts rarely attend to the full range of policy considerations that underlie the judicial federalism debate and that ought to enter into their decisional calculus. Part III discusses the major arguments for and against parallel construction of similarly worded state and federal statutes. It concludes that the typical arguments for parallel construction—that it increases efficiency, better satisfies legislative preferences, and preserves the public’s perception of judicial legitimacy—carry little weight in deciding between competing causation standards. The benefits of independent construction, on the other hand—policy experimentation and fidelity to state policy preferences—are strong. Accordingly, this Note argues that state discrimination statutes should be interpreted independently of Gross and Nassar.
I. Causation in Federal Employment Discrimination Law
This Part provides a brief historical overview of the development of causation doctrine in federal employment discrimination law. This overview is crucial for understanding why different standards apply under different statutes, even when the statutes use equivalent language in describing the required causation.
Section I.A describes the classic burden-shifting framework for analyzing Title VII claims laid out by the Supreme Court in McDonnell Douglas Corp. v. Green
and notes the scholarly and judicial debates about the relationship between that framework and causation standards. Section I.B describes the Court’s application, in Price Waterhouse v. Hopkins,
of a motivating-factor analysis and the rejection of but-for causation by a plurality of the Court. Section I.C discusses the congressional adoption of the motivating-factor framework in the Civil Rights Act of 1991, which endorsed a version of this standard for the so-called “status-based discrimination” provisions of Title VII. Finally, section I.D discusses the recent and controversial Gross and Nassar cases, in which the Supreme Court held that the motivating-factor standard is unavailable to plaintiffs under the ADEA and the Title VII retaliation provision, in part because these statutes were not amended in 1991.
A. The Traditional Framework: The McDonnell Douglas Approach to Disparate Treatment Cases Under Title VII
Under Title VII of the Civil Rights Act of 1964, it is unlawful for an employer to discharge or refuse to hire an individual, or otherwise discriminate with respect to an individual’s terms of employment, “because of such individual’s race, color, religion, sex, or national origin.”
While this provision indicates a clear congressional denunciation of employment discrimination on the basis of the listed factors, the precise contours of the prohibited conduct are less self-evident. Exactly what kinds of decisions on the part of employers can fairly be said to be “because of” these factors?
Intimately tied to the causation question is the issue of evidence. If the plaintiff in an employment discrimination suit must show some sort of causal link between her protected status or activity and the adverse employment action she is challenging, what kinds of evidence does she need to bring to satisfy the causation requirement?
In light of these difficulties, courts have long recognized the need for robust tools for determining liability in employment discrimination cases, developing various doctrinal frameworks for assessing discrimination claims in the absence of direct evidence of discriminatory intent.
The most well-known of these is the burden-shifting framework laid out in McDonnell Douglas Corp. v. Green.
Under McDonnell Douglas, the plaintiff bears the initial burden of establishing a prima facie case of discrimination, which in McDonnell Douglas involved showing that: (1) the plaintiff belonged to a racial minority group; (2) he was qualified and applied for the position for which the employer was soliciting applicants; (3) he was rejected despite his qualifications; and (4) the position remained open after his rejection, and the employer continued to solicit applications from candidates with the individual’s qualifications.
Once the plaintiff establishes a prima facie case, the burden of production shifts to the defendant to articulate a “legitimate, nondiscriminatory reason” for the challenged employment action.
By carrying this burden, the defendant rebuts the presumption created by the prima facie case, and the plaintiff must then ultimately prove that the proffered legitimate reason was pretextual and that the real reason for the employment action was discrimination.
The burden of persuasion remains always with the plaintiff.
There is some scholarly disagreement about the role of causation in the McDonnell Douglas framework.
Neither McDonnell Douglas nor any of the major Supreme Court cases elaborating on the McDonnell Douglas framework explicitly mention a particular causation standard.
Nevertheless, the predominant view has equated McDonnell Douglas with but-for causation.
B. Rejecting the But-For Requirement: Price Waterhouse and the Mixed-Motive Framework
A frequent criticism of McDonnell Douglas is that it presumes employers have a single rationale for all their decisions, ignoring the complex realities of human decisionmaking and workplace dynamics.
This perceived shortcoming of the framework was one of the driving forces behind the development of a subsequent important federal framework, commonly known as “mixed motive.” In Price Waterhouse, the Court was faced with a sex discrimination claim by a female employee who had been passed over for partnership in an accounting firm.
Although the defendant had persuaded the district court judge that it legitimately refused to promote the plaintiff to partnership because of the plaintiff’s poor interpersonal skills, Justice Brennan’s plurality opinion noted that there were “clear signs” that “some of the partners reacted negatively to Hopkins’ personality because she was a woman.”
Crucially, the plurality in Price Waterhouse opined on the meaning of “because of” in Title VII: “We take these words to mean that [the protected characteristic] must be irrelevant to employment decisions.”
The plurality explicitly rejected Price Waterhouse’s argument that but-for causation should apply, arguing that Congress’s intent to prohibit employers from taking sex, race, religion, and national origin into account in making employment decisions “appears on the face of the statute.”
Invoking “common sense,” Justice Brennan’s opinion argued that it would strain credulity to conclude that Congress, with its use of the simple “because of” language, intended to require Title VII plaintiffs to disentangle the legitimate and illegitimate motivations for an employment decision and prove that the illegitimate motivations rose to the level of but-for causation.
In addition to this common-sense or plain-meaning argument, the plurality drew on the statements of various legislators before the passage of the Act to conclude that the purpose of Title VII was to prohibit employers from taking the enumerated characteristics (race, color, sex, etc.) into account when making employment decisions and to promote hiring based on job qualifications.
Because of this, the Court held that if a plaintiff showed that her gender played a motivating part in the challenged decision, as Hopkins had, then the defendant could only avoid liability by prevailing on a “same-decision” defense—that is, by proving that it would have made the same decision even had it not taken the protected characteristic into consideration.
Under McDonnell Douglas, Hopkins likely would have lost because she would not have been able to show that the employer’s proffered reason, her poor interpersonal skills, was pretext.
Indeed, the district court judge in Hopkins’s case specifically found that Price Waterhouse had not invented its reservations about Hopkins’s interpersonal skills as pretext for discrimination.
Moreover, the district judge intimated that Hopkins would not have satisfied a but-for causation requirement: “Because plaintiff had considerable problems dealing with staff and peers, the Court cannot say that she would have been elected to partnership if the Policy Board’s decision had not been tainted by sexually based evaluations.”
For that reason, the dissent would have directed judgment for Price Waterhouse.
C. Congress Adopts Mixed-Motive: The Civil Rights Act of 1991
Two years after Price Waterhouse, Congress passed the Civil Rights Act of 1991.
Among other things, the Act codified an altered version of the mixed-motive method of proof established in Price Waterhouse.
As in Price Waterhouse, the amended Title VII stated that “an unlawful employment practice is established when the complaining party demonstrates that race, color, religion, sex, or national origin was a motivating factor for any employment practice, even though other factors also motivated the practice.”
However, unlike in Price Waterhouse, the 1991 amendments did not allow for the employer to prevail through the same-decision defense; rather, even if the employer could prove that it would have made the same decision absent the impermissible motive, this would not absolve the defendant of liability but merely limit the plaintiff to declaratory and injunctive relief, as well as attorneys’ fees and costs.
For the five status-based discrimination claims listed in 42 U.S.C. § 2000e-2, then, the 1991 Act made Price Waterhouse obsolete. However, Congress did not make such an addition to various other federal employment discrimination statutes, including the Title VII antiretaliation provision
and the ADEA.
D. The Shifting Federal Tide: Gross, Nassar, and the Rise of But-For Causation
The Title VII amendments led to confusion about the reach of the Price Waterhouse decision.
Because Congress amended only the Title VII status-based provisions in 1991, where did that leave other federal statutes using the “because of” language to describe the required nexus between protected characteristics or activities and adverse employment actions?
The Supreme Court addressed this question head-on in Gross v. FBL Financial Services, Inc.
Noting that Congress neglected to add a motivating-factor provision to the ADEA in 1991 when it added such a provision to Title VII, even though it amended the ADEA in other ways, the Court concluded that Title VII jurisprudence did not control its interpretation of the ADEA.
With Price Waterhouse out of the picture, the Court turned to the dictionary to ascertain the meaning of “because” and concluded that it meant but-for causation.
It bears noting that, apart from the dictionary, which would presumably lead to the same interpretation of “because” no matter what statute it appears in, the only affirmative hint relied upon by the Court to discern the ADEA’s meaning was Congress’s failure to amend it in 1991. Though it is not clear precisely how much the holding depends on this fact, it is fair to say that the Gross Court placed less weight on the intent of the enacting legislature than the Price Waterhouse Court did.
Because the Court in Gross largely relied on the dictionary definition of “because,” it perhaps signaled that it would be willing to apply a similar reading to other similarly drafted statutes.
In 2013, in University of Texas Southwestern Medical Center v. Nassar, the Court did just that with respect to the Title VII antiretaliation provision.
Noting that Congress added the motivating-factor provision to the section of the statute dealing with status-based discrimination rather than a section dealing with all unlawful employment actions or with retaliation, the Court held again that but-for causation was the appropriate standard.
Citing Gross as persuasive authority for the meaning of “because of” and the significance of the 1991 amendments, the Court indicated its willingness to extend the Gross line of reasoning to other statutes.
And it went further by opening its analysis with the assertion that “[c]ausation in fact . . . is a standard requirement of any tort claim. This includes federal statutory claims of workplace discrimination.”
Moreover, it all but repudiated Price Waterhouse by noting that “there is no reason to think that the different balance [of a lessened causation requirement and an affirmative defense] articulated by Price Waterhouse somehow survived [the 1991 Act’s] passage.”
After Nassar, it is doubtful that a motivating-factor analysis is available under any federal statute except those covered by the 1991 amendments, regardless of any arguments about congressional intent.
The Court’s decisions in Gross and Nassar have been subject to vigorous critique among scholars and plaintiff-side practitioners.
One common criticism is that the Court committed precisely the error that the Price Waterhouse plurality warned against: Taking “because of” as a shorthand for but-for causation misunderstands the intent that the plurality found apparent on the face of Title VII—namely, that protected characteristics were to be irrelevant in employment decisions.
More pragmatically, the widespread perception was that Gross and Nassar made it significantly more difficult for plaintiffs to prevail in ADEA and Title VII retaliation suits.
Those who agree with the Price Waterhouse plurality that the motivating-factor standard is the common-sense interpretation of “because of” in employment discrimination statutes have reason to lament the developments of Gross and Nassar: In both cases a 5-4 majority repudiated that interpretation and inserted in its place the tort concept of but-for causation.
That is not to say that plaintiffs seeking the benefit of the mixed-motive framework are out of luck for all claims except the status-based discrimination claims covered by the 1991 amendments. A client alert published by a defense-side firm shortly after the Nassar decision warned that, because subnational laws might still be analyzed using the motivating-factor standard, the Supreme Court’s holding in Nassar “ultimately may not reduce the number of retaliation and harassment claims lodged against employers. Rather, [Nassar] may serve only to push plaintiffs away from Title VII and towards any applicable state and local equivalents . . . .”
Of course, this strategy remains available only to the extent that state courts retain the motivating-factor standard in interpreting their state statutes after Gross and Nassar. In confronting this question, state courts—whether explicitly or implicitly—must contend with their role within our system of dual sovereignty. State courts may afford differing levels of deference to federal law, and their reasons for doing so may vary. How they approach these issues has profound implications for the federal system. It is to these state-court decisions that this Note now turns.
II. Interpreting State Employment Discrimination Statutes
This Part examines state-court interpretations of state employment discrimination statutes in the wake of Gross and Nassar. Although these courts interpret different statutes and differ considerably in the weight that they give to federal precedent in interpreting those statutes, a relatively circumscribed set of jurisprudential and policy concerns underlies the decision in each case to follow or depart from federal doctrine. These concerns track broader debates about judicial federalism, balancing the values of state sovereignty, interjurisdictional dialogue, and experimentation against considerations of uniformity, efficiency, and perceptions of judicial legitimacy.
Nevertheless, as this Part will detail, these concerns rarely come to the surface. Rather, in deciding to follow or depart from federal law, state courts often give little explicit consideration to the weighty issues on both sides of the debate. This Part’s catalogue of different state-court approaches sets the stage for Part III, which assesses these approaches in light of the broader considerations at play.
Section II.A lays out the field by surveying the state and local statutory landscape, noting some common differences between subnational employment discrimination statutes and their national analogues. Section II.B analyzes the case law that has emerged from state courts, especially state courts of last resort, in the years since Gross and Nassar. It concludes that these state courts often give short shrift to the federalism concerns that lurk beneath their treatment of federal law.
A. The State Statutory Landscape
All fifty states have some form of employment discrimination statute.
While many of these statutes were modeled partly after Title VII,
none of them are completely identical to federal models in all significant details.
Unlike the federal model, most states have omnibus statutes that include all of the protected statuses under one statutory framework.
Some state and local discrimination statutes contain directives to the courts to construe the statutes liberally,
language that is absent from the federal statutes.
Although it is frequently said that most state employment discrimination statutes were modeled after federal statutes such as Title VII,
around half of the states had fair employment statutes at the time of Title VII’s enactment.
These statutes anticipated many of the features of Title VII, complicating the narrative that the state statutes were “modeled” on federal ones.
Additionally, in 1966 the Uniform Law Commissioners drafted a Model Anti-discrimination Act.
Although the Model Anti-discrimination Act uses the same “because of” language as Title VII to describe the practices that it prohibits,
at least one state supreme court has noted that portions of the state employment discrimination statute were modeled after the Model Act—rather than Title VII—in declining to follow federal Title VII doctrine.
In fact, these state and local statutes bear a variety of relationships to Title VII and federal employment discrimination policy. At one extreme of the spectrum, some state statutes explicitly endorse or indicate their subordination to the federal policy embodied in statutes like Title VII. For example, the Texas Commission on Human Rights Act notes that its purpose is to “provide for the execution of the policies of Title VII of the Civil Rights Act of 1964 and its subsequent amendments,” as well as those “embodied in Title I of the Americans with Disabilities Act of 1990 and its subsequent amendments.”
This kind of language seems to militate against independent construction, linking the state statute to federal policy.
A unique case on the other extreme is the New York City Human Rights Law (NYCHRL), which reads:
a. The provisions of this title shall be construed liberally for the accomplishment of the uniquely broad and remedial purposes thereof, regardless of whether federal or New York state civil and human rights laws, including those laws with provisions worded comparably to provisions of this title, have been so construed.
b. Exceptions to and exemptions from the provisions of this title shall be construed narrowly in order to maximize deterrence of discriminatory conduct.
c. Cases that have correctly understood and analyzed the liberal construction requirement of subdivision a of this section and that have developed legal doctrines accordingly that reflect the broad and remedial purposes of this title include Albunio v. City of New York, 16 N.Y.3d 472 (2011), Bennett v. Health Management Systems, Inc., 92 A.D.3d 29 (1st Dep’t 2011), and the majority opinion in Williams v. New York City Housing Authority, 61 A.D.3d 62 (1st Dep’t 2009).
The NYCHRL—and the case law that interprets it, now incorporated into the statute by amendment—presents perhaps the strongest legislative endorsement of independent construction of a subnational statute. One of the cases cited approvingly in the statute, Bennett v. Health Management Systems, Inc., held that because of the broad-construction provision of the NYCHRL, a defendant seeking summary judgment had to establish that no jury could find the defendant liable under either the traditional McDonnell Douglas analysis or a mixed-motive analysis encompassing the motivating-factor causation requirement.
This holding applied even though the wording of the status-based discrimination provision of the NYCHRL is substantially similar to that found in Title VII.
The NYCHRL is an extreme example of why the text of state or local statutes might compel state courts to interpret state discrimination statutes in a manner that diverges from the federal case law, even though the statutory language that suggests causation—“because of”—is substantially the same.
In spite of the common differences between federal, state, and local statutes, state courts often look to federal employment discrimination cases for guidance on questions of statutory interpretation, including which causation standard the statute implies. One possible reason for this is that federal courts are often tasked with interpreting state discrimination statutes because plaintiffs bring claims under both federal and state laws in federal court.
Federal courts often use similar analytical frameworks for analyzing state and federal claims brought together.
Moreover, state courts confronted with the issue of how to react to Gross and Nassar often face similar practical constraints on their ability to develop interpretations that depart from federal case law. As longtime chief judge of the New York Court of Appeals Judith Kaye has noted, a “crucial distinction between state courts and federal courts interpreting statutes is the quantity of the legislative history that is available.”
Chief Judge Kaye noted that, more often than federal courts, state courts have to infer legislative intent from the language of the statute itself.
Understanding this range of policy concerns and practical constraints that influence state-court decisions is a crucial prerequisite to evaluating state-court responses to Gross and Nassar.
B. State Courts Decide the Causation Question
In light of the opposing pulls of state autonomy and deference to federal law, it is perhaps not surprising that many state courts have deferred deciding whether Gross and Nassar should apply to claims under state statutes.
With increasing frequency, however, state courts—including some state supreme courts—are deciding whether the but-for causation standard laid out in Gross and Nassar applies to the interpretation of various state discrimination statutes. This section surveys that landscape, which is only now beginning to take shape. Section II.B.1 discusses the states that have applied but-for causation or another requirement higher than the motivating-factor standard. Section II.B.2 discusses the states that have rejected the application of Gross and Nassar to state statutes and held that motivating-factor standards apply.
1. State Courts Applying a But-For or Other Heightened Causation Requirement. — In the wake of Gross and Nassar, state courts applying a heightened causation requirement have generally done so in one of three ways: (1) deciding the issue on independent state law precedent and noting the congruence with recent Supreme Court cases as dicta; (2) establishing congruence with federal law as a precedential principle that outweighs state law precedent on the correct standard; or (3) automatically applying the federal standards without justification.
a. Deciding the Issue on Independent State Law Precedent. — State courts might, of course, independently decide that the state law dictates the same result as the federal case law. Perhaps the clearest example of this is Asbury University v. Powell, in which the Supreme Court of Kentucky held that a but-for causation standard applied to a retaliation claim under the Kentucky Civil Rights Act (KCRA).
The court explained that it was merely following state case law, which had always followed the but-for standard that the federal courts had only recently adopted in Nassar.
By “happenstance,” the Nassar Court “merely restated the law as it has always been interpreted in Kentucky.”
One might wonder, however, to what extent a state court applying but-for causation on state law grounds is influenced by the gravitational pull of the recent Supreme Court decisions.
These cases provide an aura of legitimacy to state-court decisions and may help tip the balance in favor of a heightened causation requirement when the weight of state law authority is not necessarily clear.
Nassar may have played this role in the majority opinion in Haskenhoff.
Noting that “the retaliation provision of the ICRA mirrors almost exactly the retaliation provision of Title VII”
and that the Iowa Supreme Court has “consistently employed federal analysis when interpreting the ICRA,”
the court supported its rationale for a higher causation standard with reference to the Supreme Court’s decision in Nassar.
Although the Iowa majority rejected allegations of blind adherence to federal law, it invoked the “long-standing practice of looking to federal decisions to interpret the same or equivalent statutory language.”
Nevertheless, the court explicitly relied on state law precedent to hold that the proper standard is “significant factor” and declined to apply the but-for standard of Nassar itself.
A vigorous dissent disagreed, arguing that “the Nassar case has no bearing in the interpretation of the ICRA” because Nassar relied upon the legislative history of Title VII, which was fundamentally different from the legislative history behind the ICRA.
Noting the similar causation language in the status-based discrimination and retaliation provisions of the ICRA, the dissent saw no reason to interpret the two provisions as requiring different standards, especially given the ICRA’s mandate to “broadly construe [its] provisions,” a directive absent from Title VII.
The dissent stressed that “the ICRA is not simply a knockoff of the Federal Civil Rights Act”
and critiqued the majority’s “mirror” theory by noting that both the ICRA and Title VII drew upon state statutes that preceded Title VII.
The dissent urged that interpreting the ICRA required “serious, provision-by-provision analysis, recognizing similarities [with Title VII] when they appear, but also honoring the differences.”
Thus, although by the explicit terms of its decision the Haskenhoff majority was following state law precedent,
it gave the Nassar decision considerable persuasive weight, citing Nassar at length for its policy concern about the proliferation of unfounded claims.
Additionally, the majority, contrary to its assertion that it was deciding based on state precedent, noted that the enactment of the ICRA after Title VII rendered “look[ing] to federal decisions for guidance” appropriate.
Thus, even when the state courts purport to apply heightened causation requirements based on state precedent, the Gross and Nassar decisions may gravitationally pull the court’s analysis in the direction of a certain result.
b. Establishing Congruence with Federal Law as a Binding Precedential Principle. — Perhaps the most common reason that state courts follow Gross and Nassar is that the case law has established some principle or practice of following analogous federal decisions. Recently, the Supreme Court of Florida authorized a new set of standard jury instructions in civil cases.
Although the main text of the jury instruction simply uses the “because of” language from the statute,
the instruction also reads: “(If necessary, clarify the causation standard further.).”
In the “Notes on Use” for the discrimination jury instruction, the court noted that decisions construing the ADEA apply to the Florida Civil Rights Act’s age discrimination provision.
The practice of looking to the ADEA was so entrenched in Florida law that the court felt the need to cite only a state appellate court to justify relying on Gross.
Similarly, the Court of Special Appeals of Maryland has applied a presumption of federal–state congruity. In White v. Parker, the court opined, “We see no reason why we would not adopt the but-for causation standard established in Nassar for [Maryland Fair Employment Practices Act] retaliation claims.”
The court wrote that absent specific language or policy considerations that support an independent construction, it would presumptively interpret the Maryland Fair Employment Practices Act according to federal standards.
This presumption is quite similar to the approach advocated by at least one proponent of parallel construction.
c. Automatic Application of the Federal Standards. — In some instances, state courts, even state high courts, cite Supreme Court precedent as if it is binding authority on the interpretation of state statutes, without any further inquiry into why the court should follow the Supreme Court’s reasoning. Occasionally, these cases note that their interpretations of the state statutes are guided by federal law, but they go on to cite federal decisions as if binding. Such cases may arise when the principle of following federal law is so entrenched that courts feel free to cite federal case law without acknowledging the distinctness of the state and federal regimes.
For example, in Robinson v. Board of Supervisors for the University of Louisiana System, the Louisiana Supreme Court cited a Louisiana state appellate court for the proposition that Gross applies to age discrimination claims under Louisiana’s Age Discrimination in Employment Act.
Although the court noted that “Louisiana courts have traditionally looked to federal case law for guidance” in interpreting the LADEA,
it conducted minimal independent analysis in applying Gross directly to the state statute.
In 2013, the Idaho Supreme Court automatically applied the but-for requirement of Gross to an age discrimination claim under the Idaho Human Rights Act.
Similarly, in Villiger v. Caterpillar, Inc., an Illinois appellate court merely asserted that but-for causation was the requirement in age discrimination cases, with a citation to Gross.
Given that state courts are the ultimate authorities in the construction of state statutes,
these cases may constitute an abdication of the state courts’ role. In the case of state high courts, such as the Idaho Supreme Court, the role of interpreting the state statute is even more important because federal courts may certify questions of state law to the state’s highest court.
Following federal standards lockstep would effectively collapse the analogous federal and state provisions into one.
2. State Courts Applying a Motivating-Factor or Substantial-Motivating-Factor Standard. — Other state courts have more vigorously asserted their prerogative to construe state statutes independently from case law interpreting federal analogues. Although these decisions are rarely as comprehensive as the Haskenhoff dissent, they often turn on the same factors identified by that opinion as weighing in favor of independent construction of a statutory provision.
Because these decisions typically identify particular features of the state statutes at hand that help justify independent construction, they are less susceptible to categorization than cases which establish lockstep-like principles. Sometimes relying on the textual differences between an omnibus statute and the fragmented federal statutory scheme, sometimes relying on a provision instructing state courts to construe the state statute liberally, and sometimes simply disagreeing with the reasoning of Gross and Nassar, state courts have applied the motivating-factor standard with varying degrees of sensitivity to the efficiency and legitimacy considerations that should enter into a court’s decisionmaking calculus.
In the most straightforward cases, state legislatures have enacted statutory language either explicitly or implicitly endorsing the motivating-factor standard—much in the same way that Congress did in 1991. In 2011, for instance, a Texas appellate court noted that the law was unsettled as to whether Gross applied to age discrimination claims under the Texas Commission on Human Rights Act (TCHRA).
Because the relevant provision of the TCHRA, like the 1991 Title VII amendments, contained the motivating-factor language, the court held that Gross did not apply.
This was in spite of the court’s acknowledgement that, “[b]ecause the TCHRA’s stated purpose is to ‘provide for the execution of the policies of Title VII of the Civil Rights Act of 1964 and its subsequent amendments,’ Texas courts apply analogous federal case law when interpreting the [TCHRA].”
A less explicit but no less certain endorsement of the motivating-factor standard appears in the NYCHRL, which gives legislative imprimatur to a decision holding that the motivating-factor standard applies under the NYCHRL.
Less straightforward are cases in which state courts focus on differences between state statutes and their federal analogues to justify departing from federal precedent. In the unreported decision of West Virginia American Water Co. v. Nagy, the Supreme Court of West Virginia upheld a mixed-motive jury instruction in a state law age discrimination case, emphasizing that the West Virginia Human Rights Act encompasses all forms of discrimination and disparate treatment within a single omnibus statutory scheme.
Because this was one of the main considerations that the court discussed in rejecting the Gross rule, it is likely that the critics of the new judicial federalism would find the “bootstrap reasoning” in West Virginia American Water Co. to be inadequate, given the prudential reasons for parallel construction.
The inclusion of age discrimination in the same statutory scheme as the other status-based discrimination claims may, to some critics, be a flimsy pretext for what is essentially a policy-based disagreement with the Gross opinion.
Courts may be somewhat less susceptible to these criticisms in states whose employment discrimination statutes contain broad-construction provisions. Faced with a race and disability discrimination claim under the Alaska Human Rights Act, the Alaska Supreme Court in 2010 noted that “while we look to federal discrimination law jurisprudence generally,” Alaska’s employment discrimination law was intended to be construed more broadly than its federal analogue.
Further noting that, unlike in the federal scheme, the Alaska statute included age discrimination and other discrimination claims in the same sentence, the court declined to apply Gross to the claims at issue.
Another case in which a state court distinguished the state statute at hand from federal analogues based on differing provisions, including a liberal-construction provision, came down in California in 2013.
The California Supreme Court rejected the argument that Gross controlled the interpretation of “because of” in the sex discrimination provision of California’s Fair Employment and Housing Act (FEHA).
Drawing on the FEHA’s stated purposes, as well as the provision that the statute should be “construed liberally for the accomplishment of [its] purposes,” the court applied on state law grounds a standard similar to the one that Congress enacted for status-based discrimination in 1991.
Specifically, if the plaintiff proves by a preponderance of the evidence that discrimination was a “substantial factor motivating” an adverse employment decision, then the employer will be held liable under the FEHA; nevertheless, if the employer proves by a preponderance of the evidence that it would have made the same decision absent the impermissible motive, then the plaintiff may not receive damages, back pay, or an order of reinstatement but is limited to declaratory and injunctive relief, as well as reasonable attorneys’ fees and costs.
Crucially, the court applied this interpretation to all of section 12940(a) of the FEHA, which encompasses not just the status-based discrimination covered by Title VII but also discrimination on the basis of age, sexual orientation, physical disability, mental disability, gender identity, and a host of other protected characteristics.
In construing the FEHA, the California Supreme Court noted that its goal was to “give effect to the Legislature’s intent” and that it had to “begin with the statutory text” to discern intent.
The plain meaning of the statute, however, was unavailing because, “[l]inguistically, the phrase ‘because of’ is susceptible to many possible meanings.”
Furthermore, the FEHA’s legislative history did not justify the same kind of negative inference that the Supreme Court had made in Gross. Thus, the court turned its attention to the enacting legislature’s statements about the FEHA’s aims.
The court relied heavily on the FEHA’s stated purpose of preventing and deterring unlawful discrimination in rejecting but-for causation.
Nevertheless, cautious about the imposition of liability for “mere thoughts or passing statements unrelated to the disputed employment decision,” the court declined to apply the motivating-factor standard, instead applying a higher “substantial motivating factor” standard.
Because this holding applied to all the protected characteristics listed in section 12940(a), the result is that California, as compared to federal law, imposes a higher causation standard for the status-based discrimination claims covered by the 1991 Title VII amendments but a lower standard for age discrimination.
From the foregoing survey, it is clear that state courts are not engaged in any sort of unified discourse on the relative merits of parallel and independent construction. Whether it is because they feel the inexorable gravitational pull of federal law, they are bound by stare decisis to decide the question in a certain way, or they simply do not think the issue worth discussing, state courts tend to give little attention to federalism considerations when deciding to follow or depart from federal law. But systematic application of these principles to the specific problems of interpretation that these courts face may help clarify how courts should weigh the competing concerns of autonomy and deference in this context.
III. Why Federalism Considerations Favor Independent Construction
This Part assesses the range of state-court approaches discussed in Part II in light of scholarly debates about the merits and drawbacks of parallel construction. It argues that while the benefits of parallel construction are weak when it comes to causation standards in employment discrimination, the countervailing benefits of independent construction are abundantly present. These include fidelity to state policy preferences and interjurisdictional dialogue. Section III.A provides background on the discourse of “new judicial federalism” and argues that these debates provide a useful frame for the systematic analysis of the problem at hand. Although the new judicial federalism is most commonly associated with the interpretation of state constitutional provisions independent of their federal constitutional analogues, this section builds on existing scholarly work to argue that a similar set of concerns underlies the employment discrimination context. Section III.B argues that the standard legitimacy and efficiency concerns counseling in favor of parallel construction are relatively absent in the context of causation in employment discrimination statutes. It also explores differences between statutory and constitutional law, positing that, although there may be less room for productive vertical (between state and federal) dialogue on this particular question of statutory interpretation than on questions of fundamental constitutional rights, there is still the possibility of horizontal dialogue. Ultimately, this section argues that the benefits of state independence here outweigh the costs. Concerns about efficiency, legislative preference, and legitimacy should not discourage state courts from fulfilling their independent function within our system of dual sovereignty.
A. The “New” Judicial Federalism
State supreme courts are the final, authoritative voices in interpreting state statutes and constitutions.
As one scholar has put it, “[I]n a situation where a state supreme court interprets a state constitutional provision—even one textually indistinguishable from the federal provision—the [U.S.] Supreme Court, far from being final, has nothing at all to say.”
Indeed, Justice Brennan famously urged state courts to be open to the possibility of interpreting state constitutions to provide stronger protections for individual rights than the U.S. Constitution.
Commentators sometimes use the term “new judicial federalism” to refer to the practice in which state high courts rule on questions of individual rights by relying on state constitutional provisions as adequate and independent grounds of decision.
Some scholars also use the phrase to refer to a similar dynamic in the statutory context.
Not all commentators celebrate this independent interpretation of state law, however. Some insist that, when it comes to state constitutional provisions, state courts should defer to the Supreme Court’s interpretation of analogous federal constitutional provisions.
Critics of the new judicial federalism often invoke concerns about efficiency,
legislative preference,
and legitimacy
to advocate for parallel construction. Parallel construction bolsters efficiency, the argument goes, because courts deciding cases arising under parallel federal and state constitutional and statutory provisions need only learn and apply one set of rules. As to legislative preference, the hypothesis is that legislatures generally prefer parallel construction with federal statutes.
If true, this hypothesis also raises a legislative - efficiency rationale, as erring on the side of parallel construction conserves the resources needed for the legislative monitoring and abrogation of undesirable judicial rules.
Independent construction raises a potential legitimacy concern if it is likely that divergent state-court opinions will be seen as capricious or less justifiable than opinions that follow federal case law in a “lockstep” fashion, a concern amplified in states where judiciaries are elected and therefore subject to more immediate political pressures than their federal counterparts.
The “new” judicial federalism has been frequently criticized for its “result-oriented cast.” For instance, Justice Brennan has been accused of advocating for state constitutionalism precisely to make up for the erosion of the Warren Court’s individual rights jurisprudence under Chief Justice Burger.
Critics inveigh against the “bootstrap reasoning” whereby state courts, disagreeing with the U.S. Supreme Court’s interpretation of a federal provision, focus on flimsy or trivial differences between the federal and state provisions to justify reaching a different result under state law.
Such departures from federal law may “call into question unnecessarily the wisdom or moral authority of the Supreme Court.”
In light of these perceived shortcomings of the new judicial federalism, Professor Alex B. Long proposed—years before the Gross and Nassar decisions—that state courts adopt a substantive canon of construction favoring parallel construction of similarly worded state and federal statutes unless courts can “identify some meaningful difference between state and federal law, some fundamental change in approach at the federal level, or some outright error on the part of the federal courts.”
This approach, Long argues, rejects “[b]lind adherence to federal precedent”
while also avoiding the harms to legislative efficiency, legislative preference, and judicial credibility that accrue from irresponsible independent construction.
Of course, under this canon, it is still incumbent on state courts to determine whether a federal case represents a “fundamental change in approach” or an “outright error.”
In any event, even the critics do not completely absolve state courts of the responsibility of evaluating federal decisions and determining whether to follow them; however, they envision for state supreme courts a much more modest and deferential role in the federal system than do the proponents of the new judicial federalism.
Between the two extremes of this debate—a focus on state sovereignty on the one hand and an emphasis on federal supremacy on the other—is a more recent body of writing on federalism that justifies the independence of state courts, not because states are sovereigns that are entitled to operate in blithe indifference to the national conversation, but because a proper allocation of power to state governments produces benefits for the federal system as a whole.
One of the most frequently commented-upon benefits is that of dialogue between federal and state actors about fundamental questions of governance.
As Justice Brandeis famously wrote, “a single courageous State may . . . serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.”
In at least some visions of federalism, this dialogue does not lead to unending plurality and uncertainty but rather helps to forge a national consensus.
But the dialogue need not lead to a grand telos in order to be beneficial: The accumulation of knowledge may lead to better legal rules even without the prospect of ultimate consensus. Lockstep approaches to parallel statutes may impoverish this process of knowledge accumulation, leading to stagnation.
Moreover, state-court independence likely results in greater fidelity to the policy preferences of state electorates. As Professor Scott Dodson has argued, “state courts eager to track nonpreemptive federal law may misinterpret state law, resulting in further distance between the preferences of state citizens and the laws that govern them.”
Greater variation among states may well result in interpretations that are, in the aggregate, better anchored in the preferences of each state’s citizenry.
Thus, independent interpretations may in fact improve perceptions of judicial legitimacy.
Just as importantly, as Professor Dodson argues, when states give into the gravitational force of federal law, they “lend[ ] credence to the position that states are just not as good at being sovereign as the federal government is.”
This degrades the role of the states within our system of dual sovereignty.
Moreover, states’ incorporation of federal law may be self-reinforcing: “If states routinely mimic federal law rather than innovate, then all eyes will train on the federal leader[,]” creating a feedback loop that only entrenches state actors’ habit of following.
B. Motivating-Factor Standards Do Not Raise the Typical Concerns About the New Judicial Federalism
As noted above, critics of the new judicial federalism sometimes invoke notions of legitimacy, legislative preference, and legislative efficiency to argue for parallel interpretation.
Moreover, although it is possible that states and localities may arrive at different causation standards as a matter of policy, the preeminence of Title VII and the other federal discrimination statutes may suggest that employment discrimination is, to paraphrase Chief Justice Rehnquist, “truly national” as opposed to “truly local.”
Unless there is reason to think that a more lenient causation standard is an appropriate expression of the state’s particular custom or policy preference, which is unlikely given the scarcity of legislative history,
it might seem sensible to adopt Professor Long’s presumption in favor of parallel construction.
Adopting such a presumption would also avoid the complication of presenting two sets of jury instructions and asking juries to consider two different causation standards simultaneously, which some commentators have worried would overwhelm “even the most competent juror.”
However, these concerns fail to justify such a renunciation of the state courts’ obligation to interpret state statutes within a federal system characterized by overlapping statutory protections. As a threshold matter, Title VII’s text itself limits its preemptive effect to state law “which purports to require or permit the doing of any act which would be an unlawful employment practice” under federal law.
Therefore, as a matter of statutory scheme, states are empowered to put in place higher protections than the federal law. In other words, Title VII itself seems to envision setting a federal floor for employee rights rather than a ceiling. This alone, however, does not justify independent construction: While Title VII clearly indicates that state legislatures can impose higher requirements, as the TCHRA and NYCHRL have,
it does not necessarily suggest what limitations are appropriate for state courts operating within a national system.
The typical concerns about efficiency, legislative preference, and judicial legitimacy, though they may justify going with the federal tide on some issues of statutory interpretation, have comparatively little purchase on the question of which causation standard the word “because” expresses. The legislative-preference and legislative-efficiency justifications for parallel construction rest on strong, and not necessarily defensible, presumptions about how enacting legislatures want their statutes interpreted: “The initial decision to borrow federal law . . . can reasonably be viewed as establishing a general preference in favor of future parallel construction of the state statutes, provided that those constructions are reasonable and generally consistent with the progression of federal law existing at the time of borrowing.”
An equally strong presumption, however, might hold in the opposite direction: State legislatures would have little reason to enact separate discrimination statutes if they expected or wanted their citizens’ rights under the statutes to be exactly duplicative of their rights under federal analogues.
There is just as good a reason to presume that the legislative preference would be for judicial rules that are foreseeable at the time of enactment.
Here, given the weight of commentary describing the causation standards applied under Gross and Nassar as “new” or surprising, one may conclude that the but-for requirement was not foreseeable at the time of enactment of many state discrimination statutes.
The point here is not that one presumption regarding legislative preference is more defensible than the other; the point is that appeals to legislative preference do not necessarily support parallel construction.
And, of course, explicit evidence of the enacting legislatures’ preferences for whether and how the interpretation of “because” would track federal law is likely nonexistent.
It is worth recalling that, faced with a similar absence of evidence as to Congress’s intent, the plurality in Price Waterhouse invoked “common sense” in finding that the congressional intent was to prohibit consideration of the protected characteristics or activities in employment decisions, necessitating a motivating-factor standard.
The much scarcer legislative history in the state context makes this sort of reasoning even more critical.
This is not to say that the motivating-factor standard is the only commonsensical interpretation of “because” in discrimination statutes; rather, the commonsensical appeal of the Price Waterhouse plurality’s interpretation could prevent courts from having to follow highly speculative paths about how legislatures foresaw the development of the interpretation of their enacted language.
Indeed, apart from state-specific considerations, there may be other good reasons for state courts to consider not following Gross and Nassar. Recent empirical research has suggested that the but-for test is far more restrictive than layperson understandings of “because of.”
A state court might also consider adopting the motivating-factor standard as what Professor Einer Elhauge calls a “preference-eliciting default rule.”
Elhauge’s general theory of statutory canons is that they are meant to maximize the satisfaction of “enactable preferences,” or the legal rules that would be enacted if they were on the legislative agenda.
One might think, then, that canons would tend to favor the powerful because legislative preferences more often track the preferences of the powerful. Elhauge’s insight, however, is that many canons that favor the politically powerless (for example, the rule of lenity, the canon favoring Indian tribes, and others) are preference-eliciting default rules.
These canons are more likely to elicit legislative reactions that help us understand which preferences are enactable (because the canons would result in rules unfavorable to the politically powerful, who have the resources to influence the legislative agenda).
This analysis may be especially attractive in the state context, in which legislative preferences are less clear (due in part to the paucity of legislative history), as well as in the context of discrimination causation standards, about which it seems unlikely that enacting legislatures have very coherent preferences. Therefore, a lower causation might be preferable for eliciting more frequent legislative reactions.
Additionally, the “common sense”–style reasoning of Price Waterhouse may be even more amenable to the institutional strengths of state courts as compared to federal courts. When courts apply causation standards to discrimination statutes, it is technically a question of statutory interpretation. But because of the indeterminacy of the statutory language to be interpreted, it involves courts acting more in their traditional common law capacity than many modern questions of statutory interpretation do.
Indeed, the enacting legislatures may have intended to delegate the authority to develop causation standards to the courts.
This is consistent with courts and commentators’ sense that causation in employment discrimination is more “tort-like” than in the average statutory question.
The more “tort-like” a statutory claim is, arguably, the less pressing the concerns about legislative preference and efficiency (because the enacting legislatures presumably knew that standards would need to be developed and elected to leave that task to the courts) and judicial legitimacy (because state courts are more in the business of common law reasoning than federal courts).
Although there is potentially some loss to judicial efficiency that results from the divergence of state and federal standards, this is no more than the general inefficiency involved in federal courts adjudicating state law claims that are outside their expertise. In fact, because federal courts must apply the motivating-factor analysis for Title VII status-based claims by statute,
they will be well versed and will not have to learn any separate state causation tests. As to juror confusion, jurors must already understand two causation standards when plaintiffs bring federal age and sex discrimination claims together, or when they bring race discrimination and retaliation claims together. There is no compelling reason to believe that having to consider the same type of discrimination under two different standards should be any more confusing for jurors than applying different standards to different types of discrimination.
Finally, as suggested above, too much state following of federal law poses legitimacy concerns at least as grave as those raised by too little following.
Parallel construction means that state courts presumptively follow “even abrupt and counterintuitive changes in federal law.”
If done uncritically, following federal law could misinterpret state law
and degrade states as quasi-independent sovereigns in our federal system.
As to the dialogic values that some federalism theorists espouse, it is possible that parallel state and federal statutes provide fewer opportunities for productive dialogue than parallel constitutional provisions. Unlike the establishment of a new fundamental right under a state or federal constitution, or a novel or unexpected statutory interpretation expanding the previously understood reach of the statute,
the applicable causation standard in various types of discrimination suits does not provide lower federal courts any room to engage in dialogue: They are bound by straightforward pronouncements by the Supreme Court and, if called upon to interpret the analogous state statutes, will simply follow the state supreme court. There is good reason to believe that disagreement among the state supreme courts as to the proper causation standard will have little direct effect on the interpretation of federal law.
In any event, as noted above, variation can be valuable even if it does not lead to a future consensus.
Interjurisdictional dialogue accumulates knowledge and may lead to better legal rules. Independent of any questions about the preferences of the enacting legislatures, variation may lead to rules that are better suited to local conditions and public preferences. This is all the more true because of the closeness of the question at the Supreme Court. Because a bare minority of Justices in Price Waterhouse interpreted “because” to mean motivating-factor causation
and a bare majority of Justices in Gross and Nassar found it to mean but-for causation,
it seems likely that some state electorates prefer the Price Waterhouse approach. Moreover, variation may improve accountability by disallowing state judges from hiding behind federal case law to avoid political backlash.
These benefits of independent construction outweigh the at-best-equivocal arguments in favor of parallel construction.
Conclusion
As federal employment discrimination and retaliation laws continue to be interpreted in more plaintiff-unfriendly ways, the role of state courts and state statutes in the broader employment discrimination project grows potentially larger. When the Supreme Court all but repudiated Price Waterhouse—which made a fair attempt to effectuate the congressional purpose regarding causation in employment discrimination litigation—in favor of a narrow textualist approach, it likely reduced the number of viable discrimination and retaliation claims in federal court. State courts interpreting analogous provisions in state antidiscrimination law face a dilemma as to how closely to adhere to federal doctrine. Ultimately, concerns about judicial federalism should not restrain courts from exercising their independent judgment, lest they cede their authority in our federal scheme to vindicate the rights of their state’s citizens.