In this article, I argue that a state court
interpreting the law of a
sister state
is subject to the same obligations that Erie puts upon a federal court
sitting in diversity: The state court must decide as it
predicts the sister state’s
supreme court would. Surprisingly, the Supreme Court has not
required
state courts to use the predictive method with respect to
unsettled
sister state law. And state courts have taken advantage of this
freedom, by
employing a presumption that the unsettled law of a sister state
is the
same as
their own.
This article fills a big hole in
the literature. Although the issue of a state court’s
constitutional
obligations when
interpreting the law of a sister state is clearly fundamental,
it has never received detailed academic discussion before. It also
has
practical consequences, because the presumption of
similarity to forum law can play an important role in class action
certification.
This article introduces a fundamental puzzle
about Justice Brandeis's argument in Erie
that has never been explored in detail before: Under Erie
a federal court interpreting state law must defer to the
decisions of the supreme court of that state. But what if the
state
supreme
court
does not want deference?
What
if it wants federal courts to come to
their own
conclusions about the state’s law? Brandeis never
explains why federal courts should not adopt the approach of Swift v. Tyson toward a
state's
law, if that is what the
state supreme court wants them to do.
A similar puzzle arises in connection with the predictive method.
Many
states do not care whether federal courts use the
predictive method concerning their unsettled law. If that is so,
how
can the Supreme Court claim that the predictive method is demanded
by Erie?
One
solution to this puzzle is to bite the bullet and conclude that
federal
courts should decide
questions about the binding effect of state court decisions by
reference to
state law. The result would be that federal courts would choose
between
Swift
and Erie on the
basis
of the state’s own views. A similar state-by-state
approach would apply to whether a federal court is obligated to
use the
predictive method.
I offer a different solution. Erie,
I argue, does not merely put obligations on federal courts. It
also
obligates a state supreme court not to free federal courts of the
duty
to defer to its decisions about the state's law, unless
it is willing to free the courts of its own state of the same
duty. A
state supreme court’s views about the binding effect of its
decisions
must be
jurisdictionally neutral.
They
cannot discriminate on the
basis of the court system where this effect is felt.
For the same reason, a state supreme court cannot free federal
courts
of the duty to interpret the state's unsettled law according to
the
predictive method, except through inconceivable changes in the way
that
unsettled law is treated in the state's own courts.
I end with a brief discussion of a similar puzzle in connection
with
state choice-of-law rules. For those familar with conflicts, the
issue
concerns renvoi, that
is,
whether the forum should look to a state's choice-of-law rules
when
determining whether the state's law applies.
Although it is less likely to be of concern to practitioners, I
like
this article even better than the Michigan one. Despite the
academic
scrutiny that Erie
and the predictive method have
received over the years, I was able to say something new, and
pretty
radical, about both. I'm also proud of the historical research in
the
article, in which I uncover state courts' Swiftian views about the
common law
during the nineteenth and early twentieth centuries. Law’s
Dark Matter, 54 William & Mary Law Review
845-84 (2013)
Dark matter is matter that can be observed only
through its
gravitational effect on other bodies. The extrajurisdictional
effect of
a state's laws -- that is, whether they are binding on sister
state and
federal courts -- is like dark matter, because, absent
certification, the state's courts will never have a reason to make
an
authoritative statement about the issue. Questions of
extrajurisdictional effect will be addressed only by
sister state and federal courts, which are
not in a position to provide authoritative answers. In this symposium piece, I
argue that Swift v. Tyson
and Erie Railroad v. Tompkins
are best understood as guesses about dark matter, that is, the
extrajurisdictional effect of state court decisions concerning the
general common law. So understood, Swift is not the problematic decision thatJustice Brandeismade it out to be in Erie.
I also argue that the problem of dark matter remains to this day
and produces positions similar to Swift and Erie.
The example I concentrate on is the extrajurisdictional effect of
a
state's choice-of-law rules. All current choice-of-law approaches
are Swiftian:
the forum makes claims about the territorial scope of a state's
law
without considering the state's choice-of-law rules. An Erie
approach to choice-of-law rules has been advocated by Lea
Brilmayer,
Larry Kramer, and Kim Roosevelt. They argue that the forum must
respect
a state's choice-of-law rules when determining the territorial
scope of
the state's law. I argue that their Erie alternative is more problematic than the current Swiftian
approach to choice of law and that the issue can be resolved only
through certification to state supreme courts.
My Michigan article was about how the constitutional argument in Erie
applies horizontally: Just as federal courts sitting in diversity
must
respect state lawmaking power by deciding as the state supreme
court
would, so state courts entertaining sister state causes of action
must
respect sister state lawmaking power by deciding as the sister
state
supreme court would. My Minnesota article was also about the
constitutional argument in Erie
- whether the Erie
obligations that a federal court sitting in diversity has to
respect
state supreme court decisions can be controlled by the state
supreme
court itself.
This article is about a very different issue, unrelated to the
constitutional argument in Erie.
According to Hanna v. Plumer, 380 U.S. 460, 471 (1965), a federal
court
sitting in diversity must use forum state law if it is necessary
to
avoid “forum shopping” and the “inequitable administration of the
laws.” It has a reason to use forum state law even if the forum
state
supreme court does not care
whether its law is used - thus rendering Erie
irrelevant. I argue that the twin aims apply in diversity cases
not
because they protect state interests, but because they serve the
federal purposes standing behind the diversity statute. They are
about
separation of powers, not federalism. Through the twin aims, state
law
is incorporated into federal procedural common law in order to
serve
federal jurisdictional goals.
My justification of the twin aims
solves a number of outstanding puzzles, including why Justice
Warren
mentions the inequitable administration of the laws as a second
aim
separate from forum shopping; why horizontal differences in
procedure
between state court systems do not violate the twin aims (but
vertical
differences can); and why federal courts, but not Congress, are
bound
by the twin aims.
But the most important part of this article
is the discussion of the twin aims outside diversity.
The
twin aims apply in non-diversity cases, I argue, only if the
purposes
of the federal statute giving the court jurisdiction recommend
them. I
therefore examine four jurisdictional scenarios in order to
assess the role of the twin aims in each: a federal court
entertaining
a federal cause of action, a state court entertaining a federal
cause
of action (sometimes called reverse-Erie),
and a federal court entertaining a state law action
under supplemental jurisdiction and bankruptcy. In the course
of
my argument, I offer a resolution to the
current circuit split about whether a federal court sitting in
bankruptcy should use federal or forum state choice-of-law rules.
I
also argue that the Supreme Court has wrongly assumed that the
twin
aims apply in a reverse-Erie
context. As a result, it has improperly limited state courts’
powers
when entertaining federal civil rights actions – most recently in
Haywood v. Drown, 129 S. Ct. 2108 (2009).
Previously published at 107
Northwestern University Law Review Colloquy 165-79 (2012).
This is a response to Trey Childress's
article, When
Erie
Goes International, 105 Northwestern University Law Review
1531 (2011). In it I express skepticism
about whether the Erie
doctrine is significantly different when a federal court
entertains a non-federal action with international elements.
A fundamental issue, which I have not discussed in my writing on Erie
before, is the extent to which a federal court must respect state
procedural law that a state's officials think should be applied by
a
federal court. An example is when a state binds up a procedural
rule
(such as a statute of limitations) into a state cause of action
that is
then brought in federal court. Notice that the twin aims of Erie
do not concern this question, for the twin aims are about
uniformity
between federal and forum state courts whatever the forum state's
officials think about the matter.
This article addresses one
piece of the puzzle - namely the extent to which a state may
legitimately extend its procedural law to federal courts even in
the
absence of competing federal law. Many
legal scholars and federal judges—including Justices Ginsburg and
Scalia—have implicitly assumed that a state can extend its
procedural
law solely to federal courts within its borders. For example, a
state
might extend its statute of limitations for tort to, and only to,
federal courts within the state that are entertaining tort actions
(including actions under the tort law of sister states). To date,
however, no
one has identified this assumption, much less defended it. Drawing
upon
an example discussed by Chief Justice Marshall in Wayman v.
Southard,
23 U.S. (10 Wheat.) 1 (1825). I argue that such vertical power
does not
exist. Not only do states lack a legitimate interest in extending
their
law vertically, a state’s assertion of vertical power would
improperly
discriminate against federal courts. If state law applies beyond
the
state court system, it must do so on the basis of a criterion that
can
be satisfied by sister state as well as federal courts. This
requirement, which I call the principle of coordinancy, has
important
consequences, not merely for Erie
cases,
but for any situation in which the federal government seeks to
identify
legitimate state interests in the activities of federal courts.
The symposium piece offers what I believe is the
first complete flowchart for Erie
problems. Although it differs from past efforts in many respects,
perhaps the most important difference is that it accommodates all
the
jurisdictional contexts in which Erie problems can arise in
federal
court, not just diversity jurisdiction. My hope is that this
flowchart
will help demystify Erie,
by showing that Erie problems are, by and large, standard
choice-of-law problems, much like those faced by state courts.
An Erie problem can be understood as
concerning the extent to which the court of one sovereign (the
interpreter), when determining the content of the law of another
sovereign (the target), should defer to the decisions of the
target’s courts (and especially its highest court of appeals). This essay looks at Erie problems
outside of the normal context of federal courts
interpreting state common law, including how a court of a member
of the British Commonwealth (such as Australia) should interpret
the common law of another member (such as Canada). I like this
piece a lot.