IN THE SUPREME COURT OF TEXAS
════════════
No. 10-0245
════════════
Patrick O. Ojo, On Behalf of Himself and
All Others Similarly Situated, Appellant
v.
Farmers Group, Inc., Fire Underwriters Association,
Fire Insurance Exchange, Farmers Underwriters Association,
and Farmers Insurance Exchange, Appellees
════════════════════════════════════════════════════
On Certified Question from the United States
Court of Appeals for the Ninth Circuit
════════════════════════════════════════════════════
Argued October 14, 2010
Justice Willett, concurring in part.
The Court is right that today’s outcome is dictated by the Insurance Code as it is written. That being so, I wish the Court were more inhibited to do what we have prohibited—mine extratextual clues to illuminate an already-unambiguous statute. Text alone does not answer every question, but it answers many, including today’s, as the Court concedes. I accept a cautious (and non-villainous) role for extrinsic aids, including certain legislative history, where a nebulous statute is susceptible to varying interpretations,1 but our rule for unambiguous statutes is uncomplicated: “Where text is clear, text is determinative,”2 making any foray into extratextual aids not just inadvisable but, as we have repeatedly derided it, “inappropriate.”3
The Court nowhere states—or even suggests—the Insurance Code is ambiguous. But even assuming arguendo it is, “thus justifying cautious use of secondary construction aids,”4 the Court beckons some strange ones, including some we have consistently decried as patently unreliable (like failed bills in a subsequent Legislature). The Court’s detour may be well meaning, but it is not well supported, and I regret its “disparate impact” on our interpretive precedent. I would hold to our holdings—when the Legislature speaks plainly, the judiciary should as well. In other words, and applying a rule less prudish than prudent, if it is not necessary to look further, it is necessary not to look further. An unembellished interpretation of an unambiguous statute can be spare without being sparse. For these reasons, I agree with all but Part III.C of today’s opinion.
As for Chief Justice Jefferson’s concurrence, I confess it vexes me. It at once espouses methodological stare decisis, that our interpretive rules merit precedential effect, yet declines to embrace it. It reaffirms our concretized rule that “extrinsic aids are inappropriate to construe an unambiguous statute,” yet allows legislative history a “general background” and “non-interpretive” role.5 Boiled down, legislative materials cannot be used to construe but can be used to contextualize. This distinction strikes me as gossamer-thin (and is notably undrawn by the Court itself).6 Given that context is baked into construction, I cannot easily discern at what epochal point “contextualizing” (permitted) ends and “construing” (prohibited) begins. In my view, such opaque differentiations incautiously invite semantic gamesmanship (by litigants, legislators, and judges alike), and I am concerned that one judge’s context is another judge’s pretext. One mystifying (and hazardous) byproduct of the proposed context/construction distinction: Judges using legislative history to “contextualize” apparently have unfettered license to rely on materials (like subsequent failed bills) that precedent roundly condemns as untrustworthy. How can something we have discarded as absolutely unreliable suddenly rate absolute reliance? And how exactly are readers, absent an express disclaimer, to divine the true interpretive basis of a court’s decision? The concurrence depicts judges as “storytellers,”7 but given that context and construction are inextricably fused, how can one know when a court is telling a story versus selling a story?
I. The Court Rightly Holds the Statute is Unambiguous, Making it “Inappropriate” to Rummage Around in Legislative Minutiae.
The Court’s textual analysis is clear, careful, and convincing. The crux is whether scoring is based on race, not bears on race.
Beyond the Insurance Code’s use of “based on”8 and “because of”9 is its conspicuous non-use of disparate-impact language. I say conspicuous because at least two other Texas antidiscrimination laws predating these Insurance Code provisions—one in the Labor Code and one in the Government Code—expressly authorize disparate-impact liability in other areas.10 The Legislature knows well how to permit such claims, and the inclusion in those laws suggests something about its exclusion in this law: It was intentional. We should generally presume the Legislature “acts intentionally and purposely in the disparate inclusion or exclusion” of particular statutory language,11 and it would exceed our judicial function “to eliminate clearly expressed inconsistency of policy” when an unambiguous statute collides with policies adopted in related statutes.12 Unlike other Texas laws, the textual focus of the Insurance Code is solely motivative, not resultive, asking if credit scoring purposely draws racial lines or merely falls along such lines. In my view that ends the inquiry, textually and contextually.
Instead of calling it a day, however, the Court swerves into the flotsam and jetsam of legislative history. This roundabout detour is a little jarring given our forceful pronouncements over the years that unambiguous text equals dispositive text. By and large, this is a text-centric Court. Less than eight months ago, and 9-0, we cemented our commitment to determinacy in Texas Lottery Commission v. First State Bank of DeQueen, holding that extrinsic sources have no place when unambiguous text yields a non-absurd result.13 This point about ambiguity is mighty unambiguous, and mighty settled given our repeated reaffirmations in 2006,14 2007,15 2008,16 2009,17 2010,18 and 2011.19 As we held in DeQueen, materials outside the statute matter little—actually, not at all—when a statute itself decides the case. I pray DeQueen has not been dethroned.20
This Court has not adopted an overarching interpretive methodology to govern all statutory-interpretation cases, but we have agreed on one elemental rule: Definitive text equals determinative text, the singular index of legislative will. In other words, ambiguity is a prerequisite for wielding the extratextual tools listed in the Code Construction Act. A mere 84 days ago, we held that “[w]hen a statute’s language is nebulous and ‘[t]he point of disagreement lies between two plausible interpretations,’ we resort to additional construction aids to divine the Legislature’s intent”—and only then did we look to the Code Construction Act.21 The precedential takeaway is that while we sometimes consult the Act’s permissive tools, we do so only on tip-toe and only when faced with “unclear” language.22 The springboard for diving into legislative history is ambiguity.
Our cases steadfastly decline the interpretive free-for-all that occurs when courts peek behind plain language. Instead we have opted for a simpler and less-manipulable principle: Clarity means finality, and judges should read the laws that govern our lives in a manner faithful to what those laws actually say. In other words, the Code Construction Act may provide us with a buffet of interpretive options. But smartly, we have been picky eaters. The Act’s language is permissive: We may consider outside aids even absent ambiguity. The Court’s rule, however, has been mandatory: We shall not, and doing so is forbidden.
Regrettably, today’s departure from interpretive precedent is not limited to legislative history. The Court also considers an Insurance Commissioner report that while credit scoring, like all factors used to price insurance, might impose a disproportionate impact, it is not unfairly or intentionally discriminatory under the law. The Court claims to cite the Commissioner’s 2005 report “more for evidence of the Texas Legislature’s awareness of potential disparate impacts” than for the Commissioner’s reassuring construction that disparate-impact claims are not cognizable under the Insurance Code.23 Three brief responses: (1) the Court justifies relying on the report by citing the Code Construction Act, which invites consideration of “administrative construction of the statute” in order “to determine the Legislature’s intent”;24 (2) the Legislature was already mindful of disparate impact when it first authorized credit scoring two years earlier, according to the Court’s own reliance on a 2003 House Research Organization bill analysis and also the Legislature’s 2003 request for an interim study that would examine disparate impact in credit scoring;25 and in any event (3) any renewed legislative “awareness” in 2005 (when two bills failed to ban credit scoring outright) is wholly irrelevant to what an earlier Legislature with different members intended in 2003 when it authorized credit scoring. It seems to me the Court discusses the report “more” because the Commissioner’s no-liability construction confirms the Court’s no-liability construction, rather than as a multi-page prelude to a single sentence regarding two bills unenacted by a subsequent Legislature.
Our reluctance to wield the Code Construction Act’s extrinsic aids absent statutory ambiguity extends beyond legislative materials. The Act does invite judges to consider the “administrative construction of the statute” even if the statute by its terms is crystal clear. “Thanks but no thanks” has been our steadfast reply. This Court does not consider agency interpretations of unambiguous statutes. A generation ago, we articulated that “[i]f the [statute being construed is] plain and unambiguous, there is no need to resort to rules of construction, and it would be inappropriate to do so. If, on the other hand, the meaning of the provision be doubtful or ambiguous, the construction placed upon a statutory provision by the agency charged with its administration is entitled to weight.”26 Fast forward to just a few weeks ago, when we reaffirmed that principle, repeating that before we even consider the reasonableness of an agency’s interpretation, much less grant it deference, “the [statutory] language at issue must be ambiguous; an agency’s opinion cannot change plain language.”27 If there is no ambiguity, “that is the end of the inquiry.”28 Today, with whiplash-inducing speed, the Court says the opposite, that even absent ambiguity, it will consider an agency’s construction of the statute—and not merely as noninterpretive “background,” but rather, as the Court declares, “to determine the Legislature’s intent.”29
My respectful plea is for the Court to bring more predictability to its bread-and-butter work of statutory interpretation. As this case teaches, it is one thing to state consistent rules; it is another to stick to them. What are lower-court judges and litigants to do when early-2011’s interpretive consensus suddenly becomes mid-2011’s interpretive conflict? How are everyday Texans to order their affairs with certainty if the methodological goalposts shift to-and-fro from case to case? Interpretive clarity advances both stability and transparency, essential virtues that benefit legislators who draft statutes, agencies that implement them, courts that interpret them, and citizens who live under them.
II. In Today’s Age of “Legisprudence”—Judges Construing Statutory
Texts—Interpretive Consistency is Vital.
Methodology matters. The lion’s share of modern-day appellate judging is “legisprudence”—interpreting statutes. Day by day, the universe of free-form, common-law judging shrinks, meaning the bulk of this Court’s time is spent deciding what the Legislature’s words mean. Hence the tumultuous statutory-interpretation wars, since what judges decide is often shaped by how judges decide. I believe our legal system, which erects a framework for broader society, serves the public best when it provides clear guidance, consistently applied.
Predictability matters. Vacillation occasioned by the absence of set interpretive rules (or variance from them) breeds confusion. When we say we are taking the Legislature at its word, Texans are entitled to take us at ours. Are our oft-stated rules of statutory interpretation fixed or are they fluid? Discarding our simple-but-settled rule barring secondary aids absent ambiguity raises the question of when exactly does the Court deem it appropriate to do what we have repeatedly branded as “inappropriate.” Upon what basis will the Court determine that extratextual tools have a legitimate role even when the controlling statute is clear on its face? Honoring set interpretive rules recognizes that society operates “in the shadow of the law.” Texans embroiled in legal disputes (or trying to avoid them) doubtless appreciate interpretive determinacy, which lets everyone know where the Court stands—everyday Texans looking to conduct business, lawyers looking to advise them accurately, courts looking to decide cases consistently according to clear guidance, and lawmakers looking to draft laws with assurance of how they will be interpreted.
The Court holds, rightly, that the Insurance Code as written decides this case. That unambiguity pulls the plug on consulting legislative history, which is either impermissible (if used to affect meaning) or extraneous (if not used to affect meaning). Our precedent simply does not allow the Court to have it both ways.
III. Even if the Insurance Code Were Ambiguous, the Court Relies on Some
Legislative Materials We Have Rejected as Unreliable.
Even if one reads the Insurance Code as nebulous, thus inviting guarded consideration of extrinsic materials, some of what the Court considers is troubling. Legislative materials are not created equal; some are more authoritative than others. We have not adopted an overarching legislative-history hierarchy to help separate wheat from chaff and guide our wary and infrequent reliance on such materials.30 Even so, the Court errs in consulting two items that rank notoriously low on the reliability scale: (1) the fact that two bills banning credit scoring died in committee,31 and (2) a House Research Organization bill analysis summarizing the views of unidentified opponents of the credit-scoring bill.32
Most troubling is the Court’s “perilous”33 reliance on two failed bills in 2005 that would have banned credit scoring in pricing certain lines of insurance.34 The Court states that given the bills’ failure, it “can only conclude” that the Legislature did not intend to allow disparate-impact liability.35 That is not so. A review of precedent “can only conclude” that interpreting language passed in 2003 through the prism of language unpassed in 2005 enjoys scant legal support.
First, “subsequent legislative history is a ‘hazardous basis for inferring the intent of an earlier’ Congress.”36 The United States Supreme Court reiterated this wariness just three months ago: “Post-enactment legislative history (a contradiction in terms) is not a legitimate tool of statutory interpretation.”37 And such reliance is “particularly dangerous . . . when it concerns, as it does here, a proposal that does not become law.”38
The operative judicial inquiry is what did the Legislature in 2003 mean by the credit-scoring language it voted into law, not what legislators in 2005 thought their predecessors meant (or should have meant).39 Thus, contrary to the Court’s description, the two failed bills, as a factual matter, are undeniably not part of “[t]he legislative history of the credit scoring bill” from two years earlier.40 Rather, they are akin to post-hoc evidence seeking to color our interpretation of earlier-passed legislation. Importantly, we have never permitted such items to enter our analysis,41 even in what the concurrence would consider a “non-interpretive” and context-providing fashion. I would not second-guess the settled wisdom that allowing bills filed in a subsequent legislative session to affect the meaning of an earlier-passed statute “would set a dangerous precedent.”42
Second, the bills (identically worded companion bills) in no way proposed disparate-impact liability; rather they proposed to ban credit scoring outright. The bills never mentioned disparate impact one way or the other, much less expressly authorized such claims as do the Labor and Government Codes. Our unbroken precedent for 40-plus years (and United States Supreme Court precedent for just as long) rejects reliance on failed bills.43 As recently as 2009, we reaffirmed that “we attach no controlling significance to the Legislature’s failure to enact [legislation].”44 A generation earlier, we recognized an age-old legislative truism, equally true today, that bills often die “for reasons wholly unrelated to the Legislature’s view of what the original statute does or does not mean.”45 Legislative inaction, we thus held, “would afford a dubious distinction for drawing an inference of [l]egislative intent one way or the other.”46 Our unwavering position since then has been that divining legislative intent from dead bills requires naked “inference” involving “little more than conjecture.”47 For even longer, the United States Supreme Court has reiterated its “reluctan[ce] to draw inferences from Congress’ failure to act.”48 Indeed, because legislative inaction is susceptible of multiple interpretations, even High Court decisions that have mined other types of legislative history have, in the same opinion, frowned upon attaching importance to failed bills, reasoning that “unsuccessful attempts at legislation are not the best of guides to legislative intent.”49 Most recently, the Court warned “[i]t is always perilous to derive the meaning of an adopted provision from another provision deleted in the drafting process.”50 As a matter of law and logic, that peril seems even graver in this case, where the Court looks not to provisions deleted while the credit-scoring bill itself was being drafted, but rather two bills that failed years later. The Court offers no reason why such “evidence,” rejected as untrustworthy in 2009, is suddenly trustworthy in 2011.
Of this there can be no doubt: Bills die in the Texas Legislature for reasons both innumerable and inscrutable—“there are many reasons for saying no.”51 Lending any weight to the fact that two bills “died in committee” two years later contravenes our unbroken precedent and also potentially pivots on something perhaps nonpivotal. It may be much ado about nothing, or much ado about something, but that something can never be known. It is less legislative history than legislative mystery.
Also worrisome is the Court’s reliance on a House Research Organization bill analysis and its summary of the views of the credit-scoring bill’s unidentified detractors. HRO is a nonpartisan and independent department of the House with a well-earned reputation for impartiality and professionalism.52 It provides a wide range of information on policy issues facing state government, and its myriad reports are undoubtedly helpful to individual legislators and their staff. But as HRO itself acknowledges, its publications “are not an official part of the legislative process nor an official expression of the views of the Texas House of Representatives.”53 By contrast, bill analyses prepared by a legislative committee are attributable to that committee (though often authored by legislative staff or outside interests who are promoting the bill). The bill analyses prepared by HRO and by legislative committees differ in timing, authorship, audience, content, and purpose, and only committee bill analyses are considered part of the formal legislative process.
In any event, the statements of bill opponents, like failed bills, are discredited indicators of statutory meaning. “Courts almost never rely on representations about legislation by opponents, who have every incentive to misstate the bill’s effect.”54 Justice Frankfurter 63 years ago recognized the “common practice” (even then) by those “with their own axes to grind” manipulating the legislative record in order to assure “desired glosses upon innocent-looking legislation.”55 That same year, Justice Jackson similarly observed that “[i]t is a poor cause that cannot find some plausible support in legislative history.”56 That out-and-out untrustworthiness explains why, in the hierarchy of legislative material, “statements by opponents are among the least authoritative, as they are meant to defeat the bill in question and do not represent the considered and collective understanding of those . . . who passed the bill into law.”57 Bryan v. United States58 is instructive on this point. In that case involving an ambiguous statute, the United States Supreme Court consulted selected bits of legislative history but drew a firm line at considering statements by bill opponents. The Court explained, “[t]he fears and doubts of the opposition are no authoritative guide to the construction of legislation,”59 because, “[i]n their zeal to defeat a bill, they understandably tend to overstate its reach.”60 It is revealing that even when courts delve into the recesses of legislative history, most pay no attention to bill opponents given “the well-recognized phenomenon of deliberate manipulation of legislative history.”61 And of course the more judges rely on it, the less reliable it becomes, as “technocrats, lobbyists and attorneys have created a virtual cottage industry in fashioning legislative history so that the Congress will appear to embrace their particular view in a given statute.”62 With good reason the laws of evidence are wary of records that are prone to being corrupted in order to posture for eventual litigation—“so are legislative histories, which often have become seedbeds for interested parties to plant pre-litigation amicus briefs.”63
These doubts regarding authoritativeness are especially well-warranted when one understands the nature and preparation of HRO bill analyses, which, as HRO itself acknowledges, “are not an official part of the legislative process.”64 When a bill clears its House committee and heads for the floor, HRO prepares a bill analysis that, in part, summarizes the views of supporters and opponents. But importantly, when an HRO bill analysis speaks of “opponents,” it is referring principally to nonlegislative opponents, the outside lobbyists and other third-party advocacy interests—plus everyday civic-minded Texans—who have weighed in for, on, or against a bill. HRO staff may listen to audio (or watch video) of the committee hearing, but if time is short they may not. HRO will contact those who registered at the hearing and solicit their two-cents’ worth. But if a committee hearing is sparsely attended, HRO will reach out to a range of other activists and interest groups it surmises might have an interest in the bill, whether or not they ever formally weighed in. HRO will also gather online and other written advocacy materials from those who want to influence the bill’s fate.
Finally, the Court’s statement that the opponents’ views indicate “that the Texas Legislature was aware of the possibility of a disparate impact”65 is factually untrue. HRO is a House entity that prepares information for House members, just as the Senate Research Center provides information for senators. The Court indulges a naked inference that the Legislature as a whole had studied the HRO’s summary of what certain unidentified opponents, who may not have even attended the committee hearing, had to say.
Unsuccessful bills and unsuccessful bill opponents—“I can think of no better example of legislative history that is unedifying and unilluminating.”66 I side with Justice Jackson, who believed, consistent with our precedent, that resort to legislative history “is only justified where the face of the Act is inescapably ambiguous.”67 As he elegantly explained: “Laws are intended for all of our people to live by; and the people go to law offices to learn what their rights under those laws are.”68 And even if citizens could afford the “cost of repeatedly examining the whole congressional history,”69 they still “would not know any way of anticipating what would impress enough members of the Court to be controlling.”70 Thus, to allow legislative history to modify statutory meaning “is to make the law inaccessible to a large part of the country.”71 I would not inflict another layer of expense and complexity where the Legislature has already spoken plainly.
IV. It is Perplexing Why The Court Abruptly Changes Interpretive
Course in This Case.
Given our oft-professed allegiance to unambiguous language, it is difficult to understand the Court’s decision to peek behind ambiguity-free text without asking a simple question: Why? What is distinctive about this case that warrants departing from our firm—and recently reaffirmed—maxim that extrinsic aids are “inappropriate” where the statute itself decides the case?
It is true that this case involves a sensitive public-policy matter, but we have not to this point altered our ordinary interpretive rules for controversial cases. Indeed, few cases since I joined the Court have stoked more rancorous debate than Entergy Gulf States, Inc. v. Summers,72 our 2009 workers-compensation decision. There, Summers and several amici (including some legislators) implored the Court to embellish its analysis “by going beyond the statutory text and looking to extrinsic aides [sic] such as the Act’s legislative history.”73 Given the Act’s textual clarity, we declined, instead repeating our time-honored rule: “[W]e have been clear that we do not resort to such extrinsic aides [sic] unless the plain language is ambiguous.”74 As shown by Entergy, for all its sound and fury, sensitive cases have not desensitized us to our prior holdings. Even then, we chose methodological consistency over the view that certain controversies merit scrapping our settled rule that “[w]here text is clear, text is determinative,”75 meaning “the judge’s inquiry is at an end.”76
So I remain flummoxed as to why the Court, which declares this case can be decided under the Insurance Code alone, instead blazes a new interpretive trail. Whatever the reason—and I hope it is not subject-matter sensitivity, as even disparate-impact cases require non-disparate treatment—differentiated standards no doubt confound litigants embroiled in other disputes who expect their cases will be reviewed according to consistent rules, consistently applied. Every case that arrives at this Court is consequential; every case deserves our most painstaking study; and every litigant must have confidence that his legal questions are being answered under unvarying “rule of law” standards from case to case.
Jurists of goodwill have diverse views regarding legislative history—when to use it and how to use it. And while I side with the textualists, I well understand the counterarguments. My concern today is focused narrowly on cases where the Legislature’s words decide the case all by themselves. And my regret is the Court’s abandonment of its text-minded moorings absent any showing, or even assumption, of textual ambiguity. Because the Insurance Code is clear, particularly when juxtaposed against the Labor and Government Codes, I would not look beyond it. If this statute were unclear, I would consider a sensible legislative-history hierarchy that, in part, would reject sources we have repeatedly decried as unreliable.
V. The Concurrence Rests Upon a Meringue-Like Distinction, That Context
Can Be Divorced From Construction.
The concurrence labels as its “animating principle” the rejection of any interpretive role for legislative history when statutory text is clear.77 That principle, one I welcome, stands in tension with the Code Construction Act, which invites courts to consult legislative history “[i]n construing a statute” and to do so “whether or not the statute is considered ambiguous on its face.”78 As I read Chief Justice Jefferson’s concurrence, Texas courts should steadfastly decline the Act’s open-ended invitation to utilize extrinsic aids when construing unambiguous statutes. So far, so good. But then comes a potentially rule-swallowing exception: Such material can play a “non-interpretive” role, like providing context or “general background.” This is a puzzling distinction, and one that handily shoehorns forbidden material into the mix.
First, in another case decided today, we reaffirm the truism that context is inseparable from construction: “Language cannot be interpreted apart from context.”79 They are indivisible, an elemental point applicable not just to legal language but to all language. A statute’s words always reign supreme, but statutory meaning is not always found solely in strict literal parsing; it turns wholly on context. Modern textualism is not allergic to context, and I have dissented when I thought the Court was taking literalism too literally and adopting a wooden construction foreclosed by statutory context: “The import of language, plain or not, must be drawn from the surrounding context,” a self-evident rule rooted in common sense, Texas statutory law, and caselaw from both this Court and the United States Supreme Court.80 Indeed, even the same word can mean diametrically opposite things depending on the context.81 As Justice Scalia, textualism’s staunchest and most prominent proponent, puts it, “In textual interpretation, context is everything.”82 That being true, a judge reading a statute must always consider “the surrounding statutory landscape”83 and be mindful of context—but more specifically, interpretive context.
Granted, jurists of divergent philosophical stripes might disagree on the best elements of context and where they lead, but modern textualists agree that language cannot be isolated from its surrounding context. Indeed, even the word “context” must be read in context. It means one thing when detailing a case’s background facts and how it came before us, but something else when ascertaining textual meaning and explaining to readers how we reasoned our way to a conclusion. When construing even a facially clear statute that seems intuitively obvious, we may well look to related legislation plus other extra-statutory contextual cues to glean the text’s accepted semantic import (grammatical conventions, treatises, dictionaries, specialized legal or technical usage, colloquial nuances, judicial interpretations from other jurisdictions, and so forth).84 In other words, we tackle it much like the Court does today, minus the jolting digression into legislative history. That is, we seek interpretive context, a reading rooted in textual and structural evidence. So when we are construing statutes like the credit-scoring bill, we give precedence to semantic context (how a skilled user of words would read the statute); we do not consider or second-guess policy context (how well those words achieve the statute’s apparent policy goals).85 This approach, our precedent teaches, is more faithful to the Legislature’s policymaking supremacy and more respectful of the complex and grueling hurdles that bills (and the compromises embedded within them) must scale to become law. In this case, as the Court ably explains, semantic context points to a clear, and therefore determinative, meaning. I would stop there.
Second, the Court nowhere adopts the context/construction distinction posited by the concurrence. Chief Justice Jefferson assures readers that the Court “makes no attempt to construct the statute’s meaning by looking at its history,” and none of the cited tidbits are being used “to construe” the Insurance Code, “thus the Court follows our standard practice.”86 The Court itself, though, steers clear of the construction/context demarcation. Indeed, its foray into legislative history and other extrinsic material makes sense only as a method of construction, as an attempt to lend interpretive weight. The Court is aiming to give authoritative content to the meaning of Section 559.051, and it seems to concede as much. Most telling, it cites as authority for peeking at legislative history the tellingly titled Code Construction Act which speaks of “construing a statute . . .”87 not “contextualizing a statute.” (The concurrence parts company with the Court on this point, declaring that the Act’s extrinsic aids, including legislative history, have no interpretive role when a statute is unambiguous.) Also revealing, the Court, in “determining whether the statute gives rise to a disparate impact theory of liability”88—in other words, interpreting the statute—cites three disparate-impact decisions from the United States Supreme Court that consulted legislative history, explicitly for interpretive purposes. In those cases, the High Court, like this Court, was trying to fortify its interpretive analysis, using legislative materials to divine a statute’s meaning.89 The Court also attempts to justify its peek at legislative history “because the declared policy of the McCarran-Ferguson Act is to ensure that state legislatures are able to regulate the business of insurance without unintended federal interference.” __ S.W.3d at __. I fail to see the relevance of this. If anything, the desire for minimal federal intrusion would seem to welcome application, not frustration, of our ordinary interpretive regime. Such use is apparent in the final paragraph of Part II of today’s opinion: “Given the [legislative materials],” the Court holds, “we can only conclude that the Legislature did not intend to create a cause of action for disparate impact discrimination.”90 That phrasing sounds remarkably like a Court that is construing a statute, not contextualizing one.
Third, the context/construction locution invites peculiar consequences. Under the concurrence’s view, because the Court is merely using legislative history for background purposes, it is 100 percent free to rely on materials we have rejected as innately unreliable. The concurrence says we should not “blind ourselves to otherwise useful information.”91 That is precisely the point. We must be blind to material that cannot help us see. We have derided failed bills as the polar opposite of “useful”—useless to be precise. Similarly, the concurrence asks why we trust judges to make fair decisions after reciting compelling facts but not after looking at legislative history.92 The dilemma in this case is not untrustworthy judges but, as our cases explain, untrustworthy evidence such as failed bills and off-the-record comments from unidentified detractors bent on derailing legislation.
Even if the Court were using legislative materials merely to set the legal stage and nothing more, why would that permit reliance on a subset of material that we have spent decades condemning as untrustworthy? Sound construction demands a sound foundation. Surely a “dubious”93 and “perilous”94 foundation amounting to “little more than conjecture”95 provides a rickety basis for construction. Given that “context is everything” in textual interpretation,96 I would expect the Court to be as insistent on reliable context as it is on reliable construction.
Fourth, the concurrence avers textualism is difficult to implement.97 Far from it. Today’s case is not a difficult one; the Court is unanimous on all but Part III.C. The Insurance Code, given its fair interpretive context, yields a ready answer. I believe it is exceedingly (and unnecessarily) more grueling to slog through inconclusive minutiae “not always distinguished for candor or accuracy”98 and that often does more to confuse than to clarify.
When it comes to judicial opinion-writing, less is quite often more,99 especially less evidence that cannot be trusted. Analysis based on clear text alone may be shorter, but it is no less complete. And it is more convincing, by definition, than analysis based on material we have previously decried as unauthoritative and prone to contrivance. Plus it avoids sending the disquieting message that clear text, despite our categorical assurances to the contrary, can never be determinative—“that, presumably under penalty of malpractice liability, the oracles of legislative history, far into the dimmy past, must always be consulted.”100 As Justice Jackson first lamented 60 years ago, this imposes on ordinary citizens a high price, supplanting clarity with opacity and accessibility with indeterminacy.101
I agree with Chief Justice Jefferson that an appellate opinion “is not a mere recitation of legal standards and conclusions.”102 But I disagree with his suggestion that opinions provide factual background just to improve storytelling. The foremost reason we include relevant facts is to demonstrate we are exercising judicial power properly. Courts do not issue advisory opinions; we decide live, real-world disputes that arise from a given set of facts. We thus describe the underlying facts to show the nature of a live controversy, and in the common-law context to help future courts discern whether a legal principle fits the facts of other disputes.
* * *
The Court’s textual analysis is clear and incisive, and I join it unreservedly. The meaning of the Insurance Code is apparent from its language, read in context, especially as contrasted with the Labor and Government Codes, both of which explicitly allow disparate-impact liability. All in all, though, I wish the Court were more allegiant to our longstanding interpretive precedent. We should treat similar cases similarly, not disparately. Given the rise of state legisprudence, we owe interpretive clarity—and consistency—to the courts below us, the litigants before us, the citizens beside us, and the cases beyond us.
_______________________________________
Don R. Willett
Justice
OPINION DELIVERED: May 27, 2011
1 Alex Sheshunoff Mgmt. Servs., L.P. v. Johnson, 209 S.W.3d 644, 652 & n.4 (Tex. 2006) (“[W]here enacted language is nebulous, we may cautiously consult legislative history to help divine legislative intent,” but as a rule, “[i]f the text is unambiguous, we must take the Legislature at its word and not rummage around in legislative minutiae.”).
2 Entergy Gulf States, Inc. v. Summers, 282 S.W.3d 433, 437 (Tex. 2009) (citations omitted).
3 Molinet v. Kimbrell, 2011 WL 182230, at *6 (Tex. Jan. 21, 2011)(citations omitted); Tex. Lottery Comm’n v. First State Bank of DeQueen, 325 S.W.3d 628, 637 (Tex. 2010) (citation omitted); City of Rockwall v. Hughes, 246 S.W.3d 621, 626 (Tex. 2008) (citations omitted); Ex parte Roloff, 510 S.W.2d 913, 915 (Tex. 1974) (citation omitted); Fox v. Burgess, 302 S.W.2d 405, 409 (Tex. 1957).
4 Galbraith Eng’g Consultants, Inc. v. Pochucha, 290 S.W.3d 863, 868 n.5 (Tex. 2009).
5 __ S.W.3d at __ (internal quotation marks omitted).
6 It brings to mind Chico Marx’s classic quip, “I wasn’t kissing her, I was just whispering in her mouth.” The Oxford Dictionary of Quotations 451 (Angela Partington ed., 4th ed. 1992).
7 ___ S.W.3d at ___.
8 Tex. Ins. Code § 560.002(c)(3)(C).
9 Id. § 544.002(a).
10 Tex. Lab. Code § 21.122(a)(1); Tex. Gov’t Code § 419.103(b).
11 Russello v. United States, 464 U.S. 16, 23 (1983).
12 W. Va. Univ. Hosps., Inc. v. Casey, 499 U.S. 83, 100 (1991).
13 325 S.W.3d at 637 (“When a statute’s language is clear and unambiguous, it is inappropriate to resort to rules of construction or extrinsic aids to construe the language.” (emphasis added) (quoting Hughes, 246 S.W.3d at 626)).
14 Sheshunoff, 209 S.W.3d at 651–52 (“[W]hen a statute’s words are unambiguous and yield a single inescapable interpretation, the judge’s inquiry is at an end.” (citation omitted)).
15 In re Estate of Nash, 220 S.W.3d 914, 917 (Tex. 2007) (“If a statute is clear and unambiguous, we apply its words according to their common meaning without resort to rules of construction or extrinsic aids.” (citation omitted)).
16 Hughes, 246 S.W.3d at 626 (“When a statute’s language is clear and unambiguous, it is inappropriate to resort to rules of construction or extrinsic aids to construe the language.”).
17 Entergy, 282 S.W.3d at 437 (“When text is clear, text is determinative of [legislative] intent. This general rule applies unless enforcing the plain language of the statute as written would produce absurd results.” (citations omitted)).
18 DeQueen, 325 S.W.3d at 637 (“[B]ecause the Legislature expressly and unambiguously set out the method for resolving conflicts between the UCC and other statutes, it would be improper to go outside the language of the statute and use canons of construction to resolve the question.”).
19 Molinet v. Kimbrell, 2011 WL 182230, at *6 (Tex. Jan. 21, 2011) (“When a statute’s language is clear and unambiguous it is inappropriate to resort to the rules of construction or extrinsic aids to construe the language.” (quotation marks omitted)).
20 A needed rule bites the dust.
21 In re Smith, 333 S.W.3d 582, 588 (Tex. 2011) (citations omitted). Another recent case further cemented our view that, notwithstanding the Act’s open-ended language, we require an ambiguity predicate: “If we assume that both of [two] interpretations are reasonable . . . .” HCBeck, Ltd. v. Rice, 284 S.W.3d 349, 356 (Tex. 2009) (emphasis added).
22 Pochucha, 290 S.W.3d at 868 n.5 (“While the language in today’s statute is somewhat unclear, thus justifying cautious use of secondary construction aids, we recently reaffirmed that such aids ‘cannot override a statute’s plain words.’” (citation omitted)).
23 __ S.W.3d at __.
24 See Tex. Gov’t Code § 311.023(6) (emphasis added).
25 __ S.W.3d at __ (citing to House Research Org., Bill Analysis, Tex. S.B. 14, 78th Leg., R.S., 20 (May 21, 2003) and Act of June 2, 2003, 78th Leg., R.S., ch. 201, § 3.01, sec. 15(a), (b)(1), (b)(5)–(6), 2003 Tex. Gen. Laws 916, 920–21 (expired Mar. 1, 2005) (previously located at Tex. Ins. Code Ann. art. 21.49-2U, § 15 (West Supp. 2003))).
26 Calvert v. Kadane, 427 S.W.2d 605, 608 (Tex. 1968) (citations omitted).
27 R.R. Comm’n of Tex. v. Tex. Citizens for a Safe Future & Clean Water, 2011 WL 836827, at *4 (Tex. Mar. 11, 2011) (quoting Fiess v. State Farm Lloyds, 202 S.W.3d 744, 747 (Tex. 2006)).
28 Id. at *3 (citing Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984)).
29
__ S.W.3d at __. This approach seems flatly at odds with the concurrence’s embrace of our general rule that “extrinsic aids are inappropriate ‘to construe’ an unambiguous statute.” Id. at __. See also Citizens for a Safe Future & Clean Water, 2011 WL 836827, at *12 (Tex. Mar. 11, 2011) (Jefferson, C.J., concurring) (“We do not defer to agency interpretations of unambiguous statutes.”).
30 Once ambiguity opens the door to guarded use of legislative history, having a principled rank-order hierarchy strikes me as worthwhile. What materials are most authoritative? We have never tackled this issue comprehensively, but given the often contradictory content within legislative history, plus the propensity of lawyers to stress only those snippets that favor their side, it seems vital that courts set (and adhere to) standards that minimize contrivance and maximize reliability. Over the years a few judges and scholars have proposed various hierarchies, an exercise that depends at least in part upon a source’s accessibility, relevance, and reliability. See generally William N. Eskridge, Jr., Philip P. Frickey & Elizabeth Garrett, Legislation and Statutory Interpretation 295–307 (2000) [hereinafter “Eskridge, An Introduction to Statutory Interpretation”]; Abner J. Mikva & Eric Lane, An Introduction to Statutory Interpretation and the Legislative Process 36–41 (1997); Frank B. Cross, The Theory and Practice of Statutory Interpretation 64–67 (2009) [hereinafter “Cross, The Theory and Practice of Statutory Interpretation”]; Lars Noah, Divining Regulatory Intent: The Place for a “Legislative History” of Agency Rules, 51 Hastings L.J. 255, 274–77 (2000); Kenneth W. Starr, Observations About the Use of Legislative History, 1987 Duke L.J. 371 (1987). It is important to note that such rank-ordering proposals all focus on federal and not state legislation. And while the proposals share some points of agreement—the items they deem most probative and least probative—they diverge somewhat on the wide middle of the reliability spectrum. All observers rank congressional conference-committee reports (a joint House-Senate explanation of the final compromise bill) the highest in terms of persuasiveness. See, e.g., Cross, The Theory and Practice of Statutory Interpretation 64. However, there is no comparable document in Texas legislative practice. Our Legislature’s version of a conference-committee report merely indicates section-by-section which version, House or Senate, the conference committee adopted. Unlike a federal conference-committee report, there is no explanatory commentary of what the bill intends to do. Reliability falls off sharply with items like statements from committee and floor proceedings, when one begins to fret more about attempts “to manipulate the statute,” id. at 67, and suspect that material “might have been strategically planted in the record” to give the statute a favored gloss, Eskridge, An Introduction to Statutory Interpretation 304. Finally, most everyone agrees that one hits rock bottom on the reliability scale when one consider failed bills, the views of nonlegislator opponents/proponents, and post-enactment commentary.
31 __ S.W.3d at __. The Court also cites to the Legislature’s recodification of portions of the Insurance Code, including what is now Section 559.051, which, as the Court notes, “made no relevant changes to the Code.” Id. at __. Variations in enacted text can lend helpful interpretive context, and nobody should quarrel with examining how an enacted statute changes over time. Such variations lend helpful interpretive context as they boast the imprimatur of the Legislature as a whole. See Entergy, 282 S.W.3d at 443 (“We give weight to the deletion of [an enacted] phrase . . . since we presume that deletions are intentional and that lawmakers enact statutes with complete knowledge of existing law.”) (citation omitted). Put differently, this is the history of the legislation, not legislative history.
32 __ S.W.3d at __.
33 District of Columbia v. Heller, 554 U.S. 570, 590 (2008).
34 __ S.W.3d at __ (citing Tex. S.B. 167, 79th Leg., R.S. (2005) and Tex. H.B. 23, 79th Leg., R.S. (2005)).
35 __ S.W.3d at __.
36 Pension Benefit Guar. Corp. v. LTV Corp., 496 U.S. 633, 650 (1990) (quoting United States v. Price, 361 U.S. 304, 313 (1960)).
37 Bruesewitz v. Wyeth LLC, 131 S. Ct. 1068, 1081 (2011).
38 LTV Corp., 496 U.S. at 650.
39 See Fogarty v. United States, 340 U.S. 8, 14 (1950) (“If there is anything in these subsequent events at odds with our finding of the meaning of § 3, it would not supplant the contemporaneous intent of the Congress which enacted the Lucas Act.”).
40 __ S.W.3d at __.
41 E.g., Entergy, 282 S.W.3d at 443–44 (“Just as we decline to consider failed attempts to pass legislation, we likewise decline consideration of lawmakers’ post-hoc statements as to what statute means.”).
42 Bruesewitz, 131 S.Ct. at 1082.
43 See, e.g., Tex. Emp’t Comm’n v. Holberg, 440 S.W.2d 38, 42 (Tex. 1969); Brecht v. Abrahamson, 507 U.S. 619, 632 (1993); Schneidewind v. ANR Pipeline Co., 485 U.S. 293, 306 (1988); City of Milwaukee v. Illinois, 451 U.S. 304, 332 n.24 (1981); Transamerica Mortg. Advisors, Inc. v. Lewis, 444 U.S. 11, 33 (1979); American Trucking Ass’ns, Inc. v. Atchison, Topeka & Santa Fe Ry., 387 U.S. 397, 416-418 (1967).
44 Entergy, 282 S.W.3d at 443 (quoting Holberg, 440 S.W.2d at 42) (quotation marks omitted).
45 Holberg, 440 S.W.2d at 42.
46 Id.
47 See Dutcher v. Owens, 647 S.W.2d 948, 950 (Tex. 1983).
48 Brecht, 507 U.S. at 632 (quoting Schneidewind, 485 U.S. at 306).
49 Red Lion Broad. Co. v. F.C.C., 395 U.S. 367, 381 n.11 (1969).
50 Entergy, 282 S.W.3d at 443 (quoting Heller, 554 U.S. at 590) (quotation marks omitted).
51 Reed Dickerson, The Interpretation and Application of Statutes 160 (1975). A bill may boast strong bipartisan and bicameral support yet fail on a technical parliamentary point of order. It may be squashed by a contrary committee chair. It may die a mysterious death in the House Calendars Committee. It may fall prey to rigid timetable deadlines. It may be held hostage as leverage to affect other legislation. It may have 20 votes in the Senate but not the traditional 21. It may get caught in inter-member crossfire. All to say, a bill has many off-ramps.
52 House Research Organization, Texas House of Representatives, About the House Research Organization, available at http://www.hro.house.state.tx.us/about.aspx (last visited May 26, 2011) [hereinafter “House Research Organization, About the House Research Organization”].
53 Id.
54 Eskridge, An Introduction to Statutory Interpretation 304.
55 Shapiro v. United States, 335 U.S. 1, 48–49 (1948) (Frankfurter, J., dissenting).
56 Robert H. Jackson, Problems of Statutory Interpretation, 8 F.R.D. 121, 125 (1948).
57 Natural Res. Def. Council v. E.P.A., 526 F.3d 591, 605 (9th Cir. 2008) (citation and quotation marks omitted). See also Cross, The Theory and Practice of Statutory Interpretation 66 (“This legislative history is generally considered less reliable, because opponents may have an incentive to distort the bill to make it appear unreasonable, in order to gain allies for its rejection.”).
58 524 U.S. 184 (1998).
59 Id. at 196 (quoting Schwegmann Bros. v. Calvert Distillers Corp., 341 U.S. 384, 394 (1951)).
60 Id. (quoting NLRB v. Fruit & Vegetable Packers & Warehousemen, Local 760, 377 U.S. 58, 66 (1964)).
61 F.E.C v. Rose, 806 F.2d 1081, 1090 (D.C. Cir. 1986).
62 Starr, 1987 Duke L.J. at 377.
63 Note, Why Learned Hand Would Never Consult Legislative History Today, 105 Harv. L. Rev. 1005, 1019 (1992).
64 House Research Organization, About the House Research Organization.
65 __ S.W.3d at __.
66 Schwegmann Bros., 341 U.S. at 397 (Jackson, J., concurring).
67 Id. at 395.
68 Id. at 396.
69 Id.
70 Id.
71 Id. at 397.
72 282 S.W.3d 433 (Tex. 2009).
73 Id. at 442.
74 Id. (citing Nash, 220 S.W.3d at 917, and Sheshunoff, 209 S.W.3d at 652 n.4). The Court in Entergy did include a brief discussion of so-called legislative history but only after a threshold assumption of statutory ambiguity. And even then, we rejected outright any consideration of failed legislation (and of lawmakers’ post-hoc statements). Id. at 442–44. Instead we credited only how enacted language had changed over time, which obviously is not “legislative history” but “the legislation’s history.” Id. at 443.
75 Id. at 437 (citations omitted).
76 Sheshunoff, 209 S.W.3d at 652.
77 __ S.W.3d at __.
78 Tex. Gov’t Code § 311.023(3) (emphasis added).
79 TGS-NOPEC Geophysical Co. v. Combs, ___ S.W.3d ___, ___ (Tex. 2011).
80 Hughes, 246 S.W.3d at 632–33 & nn.6–9 (Willett, J., dissenting) (citations omitted).
81 Id. at 632 n.6 (noting that the word “cleave” can mean either “to adhere” or “to divide”).
82 Antonin Scalia, Common-Law Courts in a Civil Law System: The Role of United States Federal Courts in Interpreting the Constitution and Laws, in A Matter of Interpretation 3, 37 (Amy Gutmann ed., 1997) [hereinafter “Scalia, A Matter of Interpretation”]. The concurrence notes two occasions when Justice Scalia himself cited to legislative history. True enough, but both times reflect settled, if wary, use of legislative history by textualist judges who ordinarily eschew such material. Justice Scalia looked at legislative history (1) where literalism would inflict absurdity, Green v. Bock Laundry Mach. Co., 490 U.S. 504, 528 (1989) (Scalia, J., concurring), and (2) to resolve an ambiguity arising from a complex regime of interrelated statutes enacted over time—but importantly, he did so only after independently verifying a committee report’s accuracy and persuasiveness through other, non-legislative history sources, United States v. Fausto, 484 U.S. 439, 444–45 (1998). The concurrence also tries to draw support from Professor Manning’s observation that textualist judges, including Justice Scalia, sometimes consult extrinsic sources in their search for context. John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 702 (1997). Yes, but the article must be read, like all things, in context. The article was speaking here of using “non-legislative history” sources to assign meaning to codified terms of art. Nobody disputes that “if a word is obviously transplanted from another legal source, whether the common law or other legislation, it brings its soil with it.” Felix Frankfurter, Some Reflections on the Reading of Statutes, 47 Colum. L. Rev. 527, 537 (1947). But Professor Manning was referring to extrinsic sources beyond legislative history, like Justice Scalia’s use of legal and general-usage dictionaries, treatises, and long-repealed statutes to construe a term embedded with technical meaning, see Moskal v. United States, 498 U.S. 103, 121 (1990) (Scalia, J., dissenting).
83 Presidio Ind. Sch. Dist. v. Scott, 309 S.W.3d 927, 929–30 (Tex. 2010) (“Before parsing the language of § 21.307(a), a brief survey of the surrounding statutory landscape provides a helpful context for that section’s use of the term ‘party’ . . . .”).
84 See, e.g., Molinet v. Kimbrell, 2011 WL 182230, at *6–7 (Tex. Jan. 21, 2011) (looking to other semantic cues but declining to excavate legislative history when divining relevant statutory context); DeQueen, 325 S.W.3d at 635–37 (same); Taylor v. Firemen’s & Policemen’s Civil Serv. Comm’n of Lubbock, 616 S.W.2d 187, 189–90 (Tex. 1981) (same).
85 See John F. Manning, What Divides Textualists From Purposivists?, 106 Colum. L. Rev. 70 (2006).
86 __ S.W.3d at __.
87 Tex. Gov’t Code § 311.023 (emphasis added).
88 __ S.W.3d at __ (citing Smith v. City of Jackson, 544 U.S. 228 (2005); Gen’l Bldg. Contractors Ass’n, Inc. v. Pennsylvania, 458 U.S. 375 (1982); Griggs v. Duke Power Co., 401 U.S. 424 (1971)).
89 The Court principally cites Smith, 544 U.S. at 238, to show that “courts have looked to a statute’s legislative history when determining whether the statute gives rise to a disparate impact theory of liability.” __ S.W.3d at __. Some courts may have done so, but not this one. In its certified question to us, the Ninth Circuit asks whether Texas law prohibits a credit-score factor from working a racially disparate impact, and I would answer the question applying ordinary Texas rules of statutory interpretation. In any event, as the Court today acknowledges, the legislative-history piece of Smith garnered only plurality, and not majority, support. Pre-Smith High Court cases consulted legislative history to determine the existence of disparate-impact liability under various antidiscrimination laws, but that approach lacked majority support in the Court’s most recent disparate-impact decision.
90 __ S.W.3d at __.
91 __ S.W.3d at __.
92 __ S.W.3d at __.
93 Holberg, 440 S.W.2d at 42.
94 Entergy, 282 S.W.3d at 443 (quoting Heller, 554 U.S. at 590).
95 Dutcher, 647 S.W.2d at 950.
96 Scalia, A Matter of Interpretation 37.
97 __ S.W.3d at __.
98 Schwegmann Bros., 341 U.S. at 396 (Jackson, J., concurring).
99 So says the author of a dissent longer than the majority. “Physician, heal thyself.” Luke 4:23.
100 Conroy v. Aniskoff, 507 U.S. 511, 519 (1993) (Scalia, J., concurring in the judgment).
101 Schwegmann Bros., 341 U.S. at 397 (Jackson, J., concurring).
102 __ S.W.3d at __.