United States Court of Appeals
                            FOR THE EIGHTH CIRCUIT
                                   _____________

                                   No. 99-3917EM
                                   _____________

Faye Anastasoff,                         *
                                         *
             Appellant,                  * On Appeal from the United
                                         * States District Court
      v.                                 * for the Eastern District
                                         * of Missouri.
United States of America,                *
                                         *
             Appellee.                   *
                                    ___________

                            Submitted: May 8, 2000
                                Filed: August 22, 2000
                                    ___________

Before RICHARD S. ARNOLD and HEANEY, Circuit Judges, and MAGNUSON,1
      District Judge.
                           ___________

RICHARD S. ARNOLD, Circuit Judge.


      Faye Anastasoff seeks a refund of overpaid federal income tax. On April 13,
1996, Ms. Anastasoff mailed her refund claim to the Internal Revenue Service for taxes
paid on April 15, 1993. The Service denied her claim under 26 U.S.C. § 6511(b),
which limits refunds to taxes paid in the three years prior to the filing of a claim.
Although her claim was mailed within this period, it was received and filed on April 16,

      1
        The Hon. Paul A. Magnuson, Chief Judge, United States District Court for the
District of Minnesota, sitting by designation.
1996, three years and one day after she overpaid her taxes, one day late. In many
cases, "the Mailbox Rule," 26 U.S.C. § 7502, saves claims like Ms. Anastasoff's that
would have been timely if received when mailed; they are deemed received when
postmarked. But § 7502 applies only to claims that are untimely, and the parties agree
that under 26 U.S.C. § 6511(a), which measures the timeliness of the refund claim
itself, her claim was received on time. The issue then is whether § 7502 can be
applied, for the purposes of § 6511(b)'s three-year refund limitation, to a claim that was
timely under § 6511(a). The District Court2 held that § 7502 could not apply to any
part of a timely claim, and granted judgment for the Service. On appeal, Ms.
Anastasoff argues that § 7502 should apply whenever necessary to fulfill its remedial
purpose, i.e., to save taxpayers from the vagaries of the postal system, even when only
part of the claim is untimely. We affirm the judgment of the District Court.

                                            I.

      We rejected precisely the same legal argument in Christie v. United States, No.
91-2375MN (8th Cir. Mar. 20, 1992) (per curiam) (unpublished). In Christie, as here,
we considered a refund claim mailed just prior to § 6511(b)'s three-year bar and
received just after. Like Ms. Anastasoff, the Christie taxpayers argued that § 7502
should operate regardless of the claim's timeliness under § 6511(a) to save their claim
under § 6511(b). We held that even if § 7502 could apply to a timely claim, it would
not help in this situation: If § 7502 were applied to the claim, it would be deemed
received before the return. But § 6511(a) provides that a claim must be submitted
within two years of overpayment if no return has yet been filed – not three years. In
other words, to save the claim under § 6511(b) only makes it untimely under § 6511(a).
Ms. Anastasoff does not attempt to distinguish Christie. She does argue that a relevant



      2
        The Hon. Catherine D. Perry, United States District Judge for the Eastern
District of Missouri.
                                           -2-
regulation was not cited in Christie, but the reasoning of the Christie opinion is squarely
inconsistent with the effect taxpayer desires to attribute to the regulation.

      Although it is our only case directly in point, Ms. Anastasoff contends that we
are not bound by Christie because it is an unpublished decision and thus not a
precedent under 8th Circuit Rule 28A(i). We disagree. We hold that the portion of
Rule 28A(i) that declares that unpublished opinions are not precedent is
unconstitutional under Article III, because it purports to confer on the federal courts a
power that goes beyond the "judicial."

      The Rule provides:

             Unpublished opinions are not precedent and parties
             generally should not cite them. When relevant to
             establishing the doctrines of res judicata, collateral estoppel,
             or the law of the case, however, the parties may cite any
             unpublished opinion. Parties may also cite an unpublished
             opinion of this court if the opinion has persuasive value on
             a material issue and no published opinion of this or another
             court would serve as well . . ..


       Inherent in every judicial decision is a declaration and interpretation of a general
principle or rule of law. Marbury v. Madison, 1 Cranch 137, 177-78 (1803). This
declaration of law is authoritative to the extent necessary for the decision, and must be
applied in subsequent cases to similarly situated parties. James B. Beam Distilling Co.
v. Georgia, 

, 544 (1991); Cohens v. Virginia, 

, 399 (1821).
These principles, which form the doctrine of precedent, were well established and well
regarded at the time this nation was founded. The Framers of the Constitution
considered these principles to derive from the nature of judicial power, and intended
that they would limit the judicial power delegated to the courts by Article III of the


                                            -3-
Constitution.3 Accordingly, we conclude that 8th Circuit Rule 28A(i), insofar as it
would allow us to avoid the precedential effect of our prior decisions, purports to
expand the judicial power beyond the bounds of Article III, and is therefore
unconstitutional. That rule does not, therefore, free us from our duty to follow this
Court's decision in Christie.

                                          II.

       The doctrine of precedent was well-established by the time the Framers gathered
in Philadelphia. Morton J. Horwitz, The Transformation of American Law: 1780-1860
8-9 (1977); J.H. Baker, An Introduction to English Legal History 227 (1990); Sir
William Holdsworth, Case Law, 50 L.Q.R. 180 (1934). See, e.g., 1 Sir William W.
Blackstone, Commentaries on the Laws of England *69 (1765) ("it is an established
rule to abide by former precedents"). To the jurists of the late eighteenth century (and
thus by and large to the Framers),4 the doctrine seemed not just well established but an
immemorial custom, the way judging had always been carried out, part of the course
of the law.5 In addition, the Framers had inherited a very favorable view of precedent


      3
        "The judicial Power of the United States, shall be vested in one supreme Court,
and in such inferior Courts as the Congress may from time to time ordain and
establish." U.S. Const. art. III, § 1, cl. 1.
      4
       Lawyers made up majorities of the Continental Congress, the signers of the
Declaration of Independence, and the Framers of the Constitution. Perry Miller, The
Legal Mind in America 16 (1962).
      5
       James Wilson suggested that the doctrine of precedent was brought to England
by the Romans. 1 The Works of James Wilson 343 (1967). Chancellor Kent traced it
"from the earliest periods of English history." James Kent, Commentaries on American
Law 473-78 (12th ed. 1873). Blackstone found it "even so early as the conquest." 1
William Blackstone, Commentaries *69. Before them, in Slade v. Morley, Sir Edward
Coke suggested simply that "precedents have always been respected . . .." 4 Co. Rep.
91, 76 Eng. Rep. 1074 (K.B. 1602), reprinted in Sources of English Legal History:
                                          -4-
from the seventeenth century, especially through the writings and reports of Sir Edward
Coke; the assertion of the authority of precedent had been effective in past struggles
of the English people against royal usurpations, and for the rule of law against the
arbitrary power of government.6 In sum, the doctrine of precedent was not merely well
established; it was the historic method of judicial decision-making, and well regarded
as a bulwark of judicial independence in past struggles for liberty.

       Modern legal scholars tend to justify the authority of precedents on equitable or
prudential grounds.7 By contrast, on the eighteenth-century view (most influentially
expounded by Blackstone), the judge's duty to follow precedent derives from the nature
of the judicial power itself.8 As Blackstone defined it, each exercise of the "judicial


Private Law to 1750 428 (1986).
      6
       Coke's struggle against the tyranny of the Stuarts, which the Framers identified
with their own against King George, made him the legal authority most admired and
most often cited by American patriots. Bernard Bailyn, The Ideological Origins of the
American Revolution 30 (1967). Coke used precedent, and emphasized it to a greater
degree than his predecessors, because it was his main weapon in the fight for the
independence of the judiciary and limits on the king's prerogative rights. See Harold J.
Berman and Charles J. Reid, Jr., The Transformation of English Legal Science: From
Hale to Blackstone, 45 Emory L.J. 437, 450 (1996); J.G.A. Pocock, The Ancient
Constitution and the Feudal Law 46 (1987). By contrast, the only criticism of the
doctrine of precedent was associated with Thomas Hobbes, who regarded the authority
of precedent as an affront to the absolute power of the Sovereign. See Thomas
Hobbes, Leviathan 323-26 (Penguin ed. 1985).
      7
       See, e.g., Frederick Schauer, Precedent, 39 Stan. L. Rev. 571, 595-602 (1987)
(noting that the authority of precedent is commonly supported by arguments: (1) from
fundamental fairness, i.e., that like cases should be treated alike; (2) from the need for
predictability; and (3) as an aid to judicial decision-making, to prevent unnecessary
reconsideration of established matters).
      8
        Blackstone's great influence on the Framers' understanding of law is a familiar
fact. See Schick v. United States, 

, 69 (1904) ("At the time of the adoption
                                           -5-
power" requires judges "to determine the law" arising upon the facts of the case. 3
Blackstone, Commentaries *25. "To determine the law" meant not only choosing the
appropriate legal principle but also expounding and interpreting it, so that "the law in
that case, being solemnly declared and determined, what before was uncertain, and
perhaps indifferent, is now become a permanent rule . . .." 1 Commentaries *69.9 In
determining the law in one case, judges bind those in subsequent cases because,
although the judicial power requires judges "to determine law" in each case, a judge is
"sworn to determine, not according to his own judgements, but according to the known
laws. [Judges are] not delegated to pronounce a new law, but to maintain and expound
the old." 

power to determine law is a power only to determine what
the law is, not to invent it. Because precedents are the "best and most authoritative"
guide of what the law is, the judicial power is limited by them. 

of
precedential authority from the law-declaring nature of the judicial power was also
familiar to the Framers through the works of Sir Edward Coke and Sir Matthew Hale.
See 4 E. Coke, Institutes of the Laws of England 138 (1642) (a prior judicial decision
on point is sufficient authority on a question of law because "a judicial decision is to
the same extent a declaration of the law."); 1 Coke, Institutes 51 (1642) ("[i]t is the
function of a judge not to make, but to declare the law, according to the golden
mete-wand of the law and not by the crooked cord of discretion."); Sir Matthew Hale,
The History of The Common Law of England 44-45 (Univ. of Chicago ed., 1971)
("Judicial Decisions [have their] Authority in Expounding, Declaring, and Publishing
what the Law of this Kingdom is . . ..").


of the Federal Constitution, it [Blackstone's work] had been published about twenty
years, and it has been said that more copies of the work had been sold in this country
than in England; so that undoubtedly, the framers of the Constitution were familiar with
it."); Daniel Boorstin, The Mysterious Science of Law 265 (1941).
      9
        This need not be done by way of a reported opinion. The record of the judicial
proceedings and decision alone is sufficient evidence of the legal principles necessary
to support the decision to provide "light or assistance" when "any critical question
arises." 1 Blackstone, Commentaries *69.
                                          -6-
       In addition to keeping the law stable, this doctrine is also essential, according to
Blackstone, for the separation of legislative and judicial power. In his discussion of the
separation of governmental powers, Blackstone identifies this limit on the "judicial
power," i.e., that judges must observe established laws, as that which separates it from
the "legislative" power and in which "consists one main preservative of public liberty."
1 Blackstone, Commentaries *258-59. If judges had the legislative power to "depart
from" established legal principles, "the subject would be in the hands of arbitrary
judges, whose decisions would be then regulated only by their own opinions . . .." 

       The Framers accepted this understanding of judicial power (sometimes referred
to as the declaratory theory of adjudication) and the doctrine of precedent implicit in
it.10 Hamilton, like Blackstone, recognized that a court "pronounces the law" arising
upon the facts of each case.11 The Federalist No. 81, at 531 (Alexander Hamilton)
(Modern Library ed., 1938). He explained the law-declaring concept of judicial power
in the term, "jurisdiction": "This word is composed of JUS and DICTIO, juris dictio,
or a speaking and pronouncing of the law," 

that the jurisdiction of
appellate courts, as a law-declaring power, is not antagonistic to the fact-finding role

      10
         See Letter from James Madison to Charles Jared Ingersoll (June 25, 1831),
reprinted in The Mind of the Founder: Sources of the Political Thought of James
Madison 390, 390-93 (Marvin Meyers ed., rev. ed. 1981) (describing the "authoritative
force" of "judicial precedents" as stemming from the "obligations arising from judicial
expositions of the law on succeeding judges . . .."); James Wilson, II The Works of
James Wilson 502 (1967) ("Judicial decisions are the principal and most authentick"
proof of what the law is and . . . "every prudent and cautious judge will appreciate them
[because] . . . his duty and his business is not to make the law, but to interpret and
apply it." 

Christopher Wolfe, The Rise of Modern Judicial Review: From
Constitutional Interpretation to Judge-Made Law 74 (1986); David M. O'Brien,
Constitutional Law and Politics 73 (1995).
      11
        James Wilson agreed: "judicium is quasi juris dictum . . . a judgment is a
declaration of the law." II The Works of James Wilson 524 (1967).
                                            -7-
of juries. 

he thought that "[t]he courts must declare the sense of
the law," and that this fact means courts must exercise "judgment" about what the law
is rather than "will" about what it should be. The Federalist No. 78 507-08. Like
Blackstone, he recognized that this limit on judicial decision-making is a crucial sign
of the separation of the legislative and judicial power. 

Hamilton concludes
that "[t]o avoid an arbitrary discretion in the courts, it is indispensable that they should
be bound down by strict rules and precedents, which serve to define and point out their
duty in every particular case that comes before them . . .." 

       The Framers thought that, under the Constitution, judicial decisions would
become binding precedents in subsequent cases. Hamilton anticipated that the record
of federal precedents "must unavoidably swell to a very considerable bulk. . . ." 

were not to be recorded for their own sake. He expected judges to give
them "long and laborious study" and to have a "competent knowledge of them." 

recognized "the obligation arising from judicial expositions of the
law on succeeding judges." Letter from James Madison to Charles Jared Ingersoll
(June 25, 1831), reprinted in The Mind of the Founder: Sources of the Political
Thought of James Madison 390, 390-93 (Marvin Meyers ed., rev. ed. 1981). Madison
expected that the accumulation of precedents would be beneficial: "[a]mong other


       12
         Other early authorities confirm the connection between the doctrine of
precedent and the separation of powers. See 1 Kent's Commentaries, Lect. XXI at 479:
"Those nations, which have adopted the civil law as the main foundation of their own
[recognize precedent to a far less degree than our own] . . .. With them the necessity
of judiciary independence upon the executive, is not so clearly acknowledged . . .. It
has been shown already that this independence requires, in a considerable degree, the
acknowledgment of precedential authority."; William Cranch, Preface, 1 Cranch iii
(1804): "In a government which is emphatically styled a government of laws, the least
possible range ought to be left for the discretion of the judge . . . perhaps nothing
conduces more to that object than the publication of reports. Every case decided is a
check upon the judge. He can not decide a similar case differently, without strong
reasons, which, for his own justification, he will wish to make public."
                                            -8-
difficulties, the exposition of the Constitution is frequently a copious source, and must
continue so until its meaning on all great points shall have been settled by precedents."
Letter from James Madison to Samuel Johnson (June 21, 1789), in 12 Papers of James
Madison 250 (Robert A. Rutland et al. eds., 1977). Although they drew different
conclusions from the fact, the Anti-Federalists also assumed that federal judicial
decisions would become authorities in subsequent cases.13 Finally, early Americans
demonstrated the authority which they assigned to judicial decisions by rapidly
establishing a reliable system of American reporters in the years following the
ratification of the Constitution. Grant Gilmore, The Ages of American Law 23 (1977);
Peter Karsten, Heart Versus Head: Judge-Made Law in Nineteenth-Century America
28-32 (1997).

       We do not mean to suggest that the Framers expected or intended the publication
(in the sense of being printed in a book) of all opinions. For the Framers, limited
publication of judicial decisions was the rule, and they never drew that practice into
question. Before the ratification of the Constitution, there was almost no private
reporting and no official reporting at all in the American states. Frederick G. Kempin,
Jr., Precedent and Stare Decisis: The Critical Years, 1800-1850, 3 Am. J. Leg. Hist.
28, 34 (1959) (reviewing the history of American reports). As we have seen, however,
the Framers did not regard this absence of a reporting system as an impediment to the
precedential authority of a judicial decision. Although they lamented the problems


      13
         See, e.g., Essays of Brutus, XV (Mar. 20, 1788) in 2 The Complete Anti-
Federalist, 441 (Herbert J. Storing ed., 1981): "one adjudication will form a precedent
to the next, and this to a following one. These cases will immediately affect individuals
only; so that a series of determinations will probably take place before even the people
will be informed of them." By contrast, the danger in the Federal Farmer's view was
that the federal courts had "no precedents in this country, as yet, to regulate the
divisions in equity as in Great Britain; equity, therefore, in the supreme court for many
years will be mere discretion." Letters from The Federal Farmer No. 3 (Oct. 10, 1787),
in 2 The Complete Anti-Federalist at 244.
                                           -9-
associated with the lack of a reporting system and worked to assure more systematic
reporting, judges and lawyers of the day recognized the authority of unpublished
decisions even when they were established only by memory or by a lawyer's
unpublished memorandum. Karsten, Heart Versus Head 30; Jesse Root, The Origin
of Government and Laws in Connecticut (1798), reprinted in The Legal Mind in
American 38-39 (Perry Miller ed., 1962).14

       To summarize, in the late eighteenth century, the doctrine of precedent was well-
established in legal practice (despite the absence of a reporting system), regarded as an
immemorial custom, and valued for its role in past struggles for liberty. The duty of
courts to follow their prior decisions was understood to derive from the nature of the
judicial power itself and to separate it from a dangerous union with the legislative
power. The statements of the Framers indicate an understanding and acceptance of
these principles. We conclude therefore that, as the Framers intended, the doctrine of
precedent limits the "judicial power" delegated to the courts in Article III. No less an
authority than Justice (Professor) Joseph Story is in accord. See his Commentaries on
the Constitution of the United States §§ 377-78 (1833):




      14
         In this, they were following the common-law view, which considered entry on
the official court record sufficient to give a decision precedential authority whether or
not the decision was subsequently reported. See, e.g., Coke, 2 Institutes, Proeme, last
paragraph (stating that judicial decisions are reliable authority whether they are
published, i.e., "related and reported in our Bookes," or only "extant in judicial
Records . . .."). This remained true even after reporting became more systematic. See
James Ram, Science of Legal Judgement (1834) ("A manuscript note of a case is
authority. It may be more full, or accurate, than a printed report of the same case. The
existence of such manuscript may be little known. When cited by a party in a
cause . . . it may be 'an authority precisely applicable' (18 Ves. 347); but the opposite
party, or the Court, may never have heard of it before; it may then come as a great
surprise upon both.").
                                          -10-
                     The case is not alone considered as decided and
             settled; but the principles of the decision are held, as
             precedents and authority, to bind future cases of the same
             nature. This is the constant practice under our whole system
             of jurisprudence. Our ancestors brought it with them, when
             they first emigrated to this country; and it is, and always has
             been considered, as the great security of our rights, our
             liberties, and our property. It is on this account, that our law
             is justly deemed certain, and founded in permanent
             principles, and not dependent upon the caprice or will of
             judges. A more alarming doctrine could not be promulgated
             by any American court, than that it was at liberty to
             disregard all former rules and decisions, and to decide for
             itself, without reference to the settled course of antecedent
             principles.

                    This known course of proceeding, this settled habit of
             thinking, this conclusive effect of judicial adjudications, was
             in the full view of the framers of the constitution. It was
             required, and enforced in every state in the Union; and a
             departure from it would have been justly deemed an
             approach to tyranny and arbitrary power, to the exercise of
             mere discretion, and to the abandonment of all the just
             checks upon judicial authority.


                                           III.

       Before concluding, we wish to indicate what this case is not about. It is not
about whether opinions should be published, whether that means printed in a book or
available in some other accessible form to the public in general. Courts may decide,
for one reason or another, that some of their cases are not important enough to take up
pages in a printed report. Such decisions may be eminently practical and defensible,
but in our view they have nothing to do with the authoritative effect of any court
decision. The question presented here is not whether opinions ought to be published,

                                          -11-
but whether they ought to have precedential effect, whether published or not. We point
out, in addition, that "unpublished" in this context has never meant "secret." So far as
we are aware, every opinion and every order of any court in this country, at least of any
appellate court, is available to the public. You may have to walk into a clerk's office
and pay a per-page fee, but you can get the opinion if you want it. Indeed, most
appellate courts now make their opinions, whether labeled "published" or not, available
to anyone on line. This is true of our Court.

        Another point about the practicalities of the matter needs to be made. It is often
said among judges that the volume of appeals is so high that it is simply unrealistic to
ascribe precedential value to every decision. We do not have time to do a decent
enough job, the argument runs, when put in plain language, to justify treating every
opinion as a precedent. If this is true, the judicial system is indeed in serious trouble,
but the remedy is not to create an underground body of law good for one place and time
only. The remedy, instead, is to create enough judgeships to handle the volume, or, if
that is not practical, for each judge to take enough time to do a competent job with each
case. If this means that backlogs will grow, the price must still be paid. At bottom,
rules like our Rule 28A(i) assert that courts have the following power: to choose for
themselves, from among all the cases they decide, those that they will follow in the
future, and those that they need not. Indeed, some forms of the non-publication rule
even forbid citation. Those courts are saying to the bar: "We may have decided this
question the opposite way yesterday, but this does not bind us today, and, what's more,
you cannot even tell us what we did yesterday." As we have tried to explain in this
opinion, such a statement exceeds the judicial power, which is based on reason, not
fiat.

       Finally, lest we be misunderstood, we stress that we are not here creating some
rigid doctrine of eternal adherence to precedents. Cases can be overruled. Sometimes
they should be. On our Court, this function can be performed by the en banc Court, but
not by a single panel. If the reasoning of a case is exposed as faulty, or if other exigent

                                           -12-
circumstances justify it, precedents can be changed. When this occurs, however, there
is a burden of justification. The precedent from which we are departing should be
stated, and our reasons for rejecting it should be made convincingly clear. In this way,
the law grows and changes, but it does so incrementally, in response to the dictates of
reason, and not because judges have simply changed their minds.

                                          IV.

        For these reasons, we must reject Ms. Anastasoff's argument that, under 8th Cir.
R. 28A(i), we may ignore our prior decision in Christie. Federal courts, in adopting
rules, are not free to extend the judicial power of the United States described in Article
III of the Constitution. Willy v. Coastal Corp., 

, 135 (1992). The judicial
power of the United States is limited by the doctrine of precedent. Rule 28A(i) allows
courts to ignore this limit. If we mark an opinion as unpublished, Rule 28A(i) provides
that is not precedent. Though prior decisions may be well-considered and directly on
point, Rule 28A(i) allows us to depart from the law set out in such prior decisions
without any reason to differentiate the cases. This discretion is completely inconsistent
with the doctrine of precedent; even in constitutional cases, courts "have always
required a departure from precedent to be supported by some 'special justification.' "
United States v. International Business Machines Corp., 

, 856 (1996),
quoting Payne v. Tennessee, 

, 842 (1991) (Souter, J., concurring). Rule
28A(i) expands the judicial power beyond the limits set by Article III by allowing us
complete discretion to determine which judicial decisions will bind us and which will
not. Insofar as it limits the precedential effect of our prior decisions, the Rule is
therefore unconstitutional.




                                          -13-
       Ms. Anastasoff's interpretation of § 7502 was directly addressed and rejected in
Christie.15 Eighth Cir. R. 28A(i) does not free us from our obligation to follow that
decision. Accordingly, we affirm the judgment of the District Court.

HEANEY, Circuit Judge, concurring.

      I agree fully with Judge Arnold's opinion. He has done the public, the court, and
the bar a great service by writing so fully and cogently on the precedential effect of
unpublished opinions. I write separately only to state that in my view, this is a case
which should be heard en banc in order to reconsider our holding in Christie, and thus
resolve an important issue.

      A true copy.

             Attest:

                     CLERK, U.S. COURT OF APPEALS, EIGHTH CIRCUIT.




      15
         On July 28, 2000, the Second Circuit decided Weisbart v. United States Dep't
of Treasury, 

(2d Cir. July 28, 2000). Weisbart appears to conflict
with Christie. We express no view on whether we would follow Weisbart if it were not
for the conclusive effect of Christie.
                                         -14-