Louisiana s Conflicts Codification: Some Empirical Observations Regarding Decisional Predictability

• Author: Patrick J. Borchers
• Position: Dean and Professor of Law, Creighton University School of Law. B.S., University of Notre Dame
• Pages: 1061-1070

Page 1061

Dean and Professor of Law, Creighton University School of Law. B.S., University of Notre Dame, 1983; J.D., University of California, Davis, 1986.

I Introduction

Of all of Professor Symeonides's accomplishments during his tenure at the Louisiana State University Law Center, perhaps the one with the greatest public impact was serving as the Reporter for the Louisiana State Law Institute's preparation of a draft bill to comprehensively codify the state's conflicts law.¹ That bill, unanimously passed by both houses of the Louisiana Legislature, became law for cases filed as of January 1, 1992.² The Louisiana conflicts codification is not your father's conflicts statute. Sleek and modern, it represents a serious effort to consolidate the gains of the conflicts revolution.

It also provides a unique opportunity in the United States to study the advantages and disadvantages of comprehensive statutory solutions to conflicts problems. While conflicts codifications are fairly common abroad-particularly in countries with a civil law heritage³-Louisiana (doubtlessly motivated in part by its civil law heritage⁴) is the first American jurisdiction to undertake such an endeavor. Louisiana makes an interesting laboratory for the further reason that for the nineteen years preceding the codification, it mostly followed a modern common law conflicts approach derived from Professor Currie's writings and the Second Conflicts Restatement.⁵ Thus, if one were to look at Louisiana choice-of-law decisions immediately preceding and following the codification, one might find clues as to whether comprehensive conflicts statutes could benefit other states.

In this article, I endeavor to begin that task. I selected the project of measuring the affirmance rate in Louisiana by appellate courts of trial court conflicts decisions before and after the codification. As discussed more extensively below, I chose to look at affirmance rates because they are a reasonable proxy for decisional predictability. A higher affirmance rate (and correspondingly a lower reversal rate) probably indicates greater decisional predictability. Correspondingly, a lower affirmance rate (and correspondingly a higher reversal rate) probably indicates less Page 1062 predictability, and brings with it greater incentives for parties to appeal adverse decisions, and with that higher litigation costs (and transaction costs generally), as well as greater costs imposed on the judicial system and thus, the public at large.

As reported in detail below, for the pre-codification cases that I sampled, the affirmance rate was 52.9%, which is statistically indistinguishable from a coin flip. For post-codification decisions, however, the affirmance rate improved to 76.2%. Any statistical analysis has inherent limitations. While the sample size is large enough that the before-and-after rates do not fall within a standard deviation of each other, more data is needed before one can say with virtual certainty that the Louisiana codification has produced an improved affirmance rate. Moreover, I do not mean to suggest that merely analyzing case results in terms of affirmances and reversals can ever provide the complete picture. Nonetheless, the results are hopeful and suggest that comprehensive conflicts codifications can produce significant benefits.

II Louisiana Conflicts Law: Before And After

Beginning with the Louisiana Supreme Court's 1973 decision in Jagers v. Royal Indemnity Co.,⁶ Louisiana became a pretty typical "modern" conflicts jurisdiction. Jagers-a case involving the application of intrafamilial immunity- seemed to adopt interest analysis,⁷ at least in the case of loss-allocating tort rules in which the contesting parties are all domiciled in the same state. These sorts of cases, conventionally called false conflicts,⁸ are the most appealing for departing from the conventional state-of-the-injury rule in tort cases, and account for many instances in which states have gone modern.⁹ Jagers, however, cited the Second Conflicts Restatement as well. As things eventually developed in Louisiana, courts took to applying interest analysis for the purpose of identifying so-called "false conflicts," but then applying the Second Conflicts Restatement in the event that the case was not a false conflict.¹⁰

One of the charges that is often brought against such methodologies is that they are radically uncertain in their operation.¹¹ Interest analysis itself is notoriously vulnerable to the criticism that what counts as an "interest" is neither self-evident nor well understood by courts,¹² and the Second Restatement (particularly for torts Page 1063 conflicts) is generally only relied upon for its open-ended, "grab-bag" provisions,¹³which are popular with courts but woefully indeterminate.¹⁴ There are suggestions, old and new, that such free-form methodologies may be heaping higher-than- necessary costs on parties and taxing judicial resources.¹⁵ If, then, such methodologies suffer this vice, there is little reason to think that pre-codification Louisiana would have been atypical.¹⁶

All this changed in Louisiana for cases filed beginning in 1992. Of course, one might ask whether there is any reason to think that the new codification is more determinate than the old system. Louisiana's codification is not, to be sure, laced with hard-and-fast rules. For instance, Article 3544, a general torts provision, states:

Issues pertaining to loss distribution and financial protection are governed, as between a person injured by an offense or quasi-offense and the person who caused the injury, by the law designated in the following order:

(1) If, at the time of the injury, the injured person and the person who caused the injury were domiciled in the same state, by the law of that state. Persons domiciled in states whose law on the particular issue is substantially identical shall be treated as if domiciled in the same state.

(2) If, at the time of the injury, the injured person and the person who caused the injury were domiciled in different states: (a) when both the injury and the conduct that caused it occurred in one of those states, by the law of that state; and (b) when the injury and the conduct that caused it occurred in different states, by the law of the state in which the injury occurred, provided that (i) the injured person was domiciled in that state, (ii) the person who caused the injury should have foreseen its occurrence in that state, and (iii) the law of that state provided for a higher standard Page 1064 of financial protection for the injured person than did the law of the state in which the injurious conduct occurred.¹⁷

This statute should, at least in some cases, reach results similar to those expected to be reached under the modern approaches. It is not tied to the state of the injury or any other single connecting factor. It also contains terms that are open to debate in a significant number of cases. For instance, determining whether tort rules involve "loss distribution" is possible in most instances, but is not an enterprise that is always free from doubt.¹⁸ The same could be said for determining when the laws of the connected states are "substantially identical" or when the tortfeasor "should have foreseen" the injury being caused.

But, from the standpoint of predictability, there are many reasons to be optimistic that such a statute will be an improvement over the pre-existing approach. For one, it gives litigants and courts an indisputably authoritative text in which to turn. When contrasting a statute with the judicially invented approaches-which rely upon a confusing amalgamation of law review articles, cases from other states, quotations from Restatements and the like¹⁹-one ought not underestimate the benefits of having, at the very least, a commonly understood starting point for analyzing multistate problems. Another reason to be sanguine that such a statute will improve predictability is that even with some open-ended terminology, the universe of competing considerations is restricted. If, say, a case under Article 3544 were to turn upon whether the locus of the injury was "foreseeable," at least all concerned could address themselves to that question. By contrast, with the judicially invented approaches, one only need survey the cases briefly to discover that there is intense debate as to what counts as an "interest" in any particular case.²⁰

Of course, speculation on these matters accomplishes little. The question to which I now turn is whether one can measure any impact on predictability made by the codification.

III Methodology

Attempting to measure anything as elusive as decisional predictability is a difficult task. I settled on attempting to measure the affirmance rates by appellate courts of trial court choice-of-law decisions. I decided on looking at affirmance (and by implication reversal) rates, because they are a reasonable proxy for predictability. If the trial court and the appellate court agree on the result, even though the losing party thought it worth the trouble to prosecute an appeal, this is Page 1065 a systemic success. It shows that both of the neutral observers agreed on the result. A reversal, however, is a systemic failure. If a trial court-presumably neutral in its outlook and doing its level best to reach the "correct," non-reversible result-fails in the task, this suggests that resolution of the issue was not apparent. If so, the resolution of similar issues will not be apparent to contesting parties, and...