[*PG485]IN DEFENSE OF OUR LAW OF
SOVEREIGN IMMUNITY
Abstract: Professor Hill maintains that the Constitution was grounded on an understanding that the states would not be suable without their consent, either in the federal or state courts; the Eleventh Amendment, within its purview, is declaratory of this understanding. The Supreme Court has consistently treated sovereign immunity as of constitutional dimension. As such, the immunity has been deemed exempt from congressional modification under the Commerce Clause. However, without overt challenge to the immunity’s constitutional status, it has been held subject to congressional modification under Section 5 of the Fourteenth Amendment. The Supreme Court’s decision in this regard does not withstand critical analysis. Sovereign immunity is not the malign doctrine it is commonly thought to be. In general, it has not served as a bar to effective relief for lawless conduct by government officers. For the most part, it has operated to defeat claims arising from consensual relations with the government—and here the immunity has been almost completely eliminated by the federal and state legislatures within their respective areas of competence.
Introduction 486
I.The Sovereign Immunity of the States 493
A.The Constitution’s Formative Period 493
1.The Case for the Original Understanding 493
2.The Case Against the Original Understanding: The Dissents 498
a.Atascadero 498
b.Seminole 503
c.Alden 504
d.Kimel 506
3.Academic Commentators 508
B.The Eleventh Amendment 513
C.Hans 517
D.Cases Following Hans, culminating in Monaco 520
E.Alden Revisited 521
F.Sovereign Immunity Conceived as Common Law 523
[*PG486] 1.Sovereign Immunity as Internal Law 523
2.Federal Judicial Power to Override State Sovereign Immunity 526
G.Congressional Power to Override State Sovereign Immunity 526
1.Under Article I 526
2.Under Section 5 of the Fourteenth Amendment 529
H.Legislative Consent to Suit 537
1.Congressional Consent 537
2.State Legislative Consent 538
I.Local Subdivisions 540
II.The Sovereign Immunity of the United States 540
III.The Wrongdoing Officer 541
A.Ex parte Young 541
1.Trespassory Conduct 541
2.Non-trespassory Conduct 542
3.Bivens 545
4.Larson: A Preview 546
B.Breach of Duty 546
C.The Government as the Real Party in Interest 551
1.In General 551
2.The Effect-on-the-Government Test 552
3.The Prospective-Retrospective Test 553
4.Offensive and Defensive Uses of Sovereign Immunity 557
IV.The Pattern in the Decisions 558
A.Government Contracts 558
B.Recovery of Property 562
C.Larson: Old Wine in a New Bottle 565
D.Other Parts of the Pattern 569
E.Pennhurst: A Nadir 571
V.Getting Money from the Government 573
A.Mandamus and Mandatory Injunction 573
B.Reich: Suing the State in Tax Cases 573
C.First English: Compensation for Takings 574
VI.The Proper Forum 579
A.Federal Jurisdiction Founded on a State’s Consent to Suit in Its Own Courts 579
B.Pennhurst: Pendent Jurisdiction 580
C.Nevada v. Hall: Suit in a Sister State 581
Conclusion 584
This Article is not written in defense of the doctrine of sovereign immunity as such, but rather in defense of our law of sovereign im[*PG487]munity, which is under continual attack. Academic opinion has been overwhelmingly hostile.1 In addition, for some fifteen years a substantial number of the justices, usually not less than four, has stood poised to eliminate the doctrine root and branch.2 The Supreme Court has [*PG488]created a potentially large hole in the structure of state sovereign immunity by its holding that Congress, under Section 5 of the Fourteenth Amendment, can provide for suits against the states without their consent.3 It will be argued that the Court erred in construing Section 5 as conferring such power on Congress. To be sure, the Court is not likely to change its course in this regard. Assuming the authority to exist, it will be argued that, in invalidating virtually every congressional attempt to exercise this authority, the Court has unduly narrowed the scope both of the Fourteenth Amendment and of Section 5. Thus, congressional diminution of state sovereign immunity exists largely as a potential.
Currently, sovereign immunity stands as an absolute bar to suit against the federal government, and against state governments as well, subject only to such power as Congress may have to override the immunity of the states under Section 5. There may of course be consent to suit on both levels, but only if consent is given by the legislatures of the respective jurisdictions.
In brief, the opponents of sovereign immunity argue: (1) that there is error in allowing the doctrine to defeat claims founded on the Constitution, especially when it is considered that the Constitution makes no provision for sovereign immunity in the first place; (2) that sovereign immunity came to us from England, where it was founded on the notion that the king can do no wrong, and as such it has no place in a regime of written constitutions that set limits on the powers of government; (3) that in any event sovereign immunity is at most an aspect of the common law and is subject to modification or elimination by judges and legislators, as in the case of common law generally; and (4) that while the Supreme Court does in fact, through the fiction of Ex parte Young,4 override sovereign immunity to vindicate constitutional rights, it has, without meaningful explanation, left large gaps where constitutional rights go unprotected, and its opinions on the point are in utter confusion. It will be argued that all but the last of these objections are without merit. As to the last objection, the opinions of the Court are indeed confusing, but if we look, not to what the Court has said, but to what it has done, it will be seen that a defensible pattern emerges from the confusion.
[*PG489] So far as concerns the provenance of sovereign immunity,5 our conception of the doctrine is seriously skewed if we conceive of it as deriving from English law.6 We derived it independently, in the same way as did England—and Italy and Japan. The immunity is an inherent attribute of sovereignty, without regard to the form of government prevailing within the borders of the particular sovereign. There is probably not a country in the world that permits itself to be sued except on terms satisfactory to it. Conversely, if we except countries where there is little if any law to speak of, there is probably not a single one that disallows suit against itself in all circumstances. The question is which governmental organs have authority to consent to such suit. In the United States, on both the federal7 and state8 levels, it has been assumed from the start that exclusive competence in this regard is vested in the legislative branch.
It will be argued that, when adopted, the Constitution was understood as embodying an understanding that the federal and state governments were free to invoke the doctrine of sovereign immunity for themselves, even if this meant that rights given by the federal Constitution would go unenforced.9 It will be further argued that the Eleventh Amendment, in the cases to which it applies, is merely an em[*PG490]bodiment of the original understanding underlying the Constitution, adopted only because the decision in Chisholm v. Georgia10 was thought to have ignored the original understanding.11 Judicial and academic critics dispute both points. They see sovereign immunity as a common-law doctrine. They maintain that if the federal courts are obliged to honor the sovereign immunity of the states, this is only because of the ouster of their jurisdiction in such cases by the Eleventh Amendment. They maintain further that the champions of state sovereign immunity were not concerned with protecting the states from claims founded on federal law, but only from claims founded on state law. They construe the Eleventh Amendment as embodying a similar limitation, contrary to its express language.12
Part of the confusion arises from the frequent assertion that sovereign immunity is common law. The confusion is compounded by failure to distinguish the two distinct dimensions in which the problems arise. One of these is the internal law of a particular jurisdiction, such as Georgia or the United States. It will be argued that even in this dimension, sovereign immunity is not common-law doctrine. The other dimension is a vertical one. In this dimension, the question is one of federal power, legislative or judicial, to set aside the immunity of the states. This is part of the larger question of federal power to override state law. Discussing this power as an aspect of common-law doctrine is absurd.13
Although academic critics and minority Justices have ignored the distinctions between these two dimensions, the Court as a whole has not. The Court has denied congressional power to set aside state sovereign immunity under Article I and sustained such power under Section 5 of the Fourteenth Amendment. It will be argued that the Article I Supreme Court decisions are correct, and, as earlier observed, that the Section 5 decisions are dubious.
While the federal and state governments have been immune from suits not consented to, their officers in general have not. The doctrine of Ex parte Young has been employed to permit suits against government officers acting contrary to law. Such relief is founded on the theory that the suit is not against the government but against the officer personally. The problem is that, in what seem to be the over[*PG491]whelming majority of such cases, the officer is essentially a nominal party, with the government, though not named, the real party in interest. Thus, the doctrine of Ex parte Young is an obvious fiction.
Acceptance of the views of the Supreme Court minority and of virtually all scholars would in effect transform the fiction into law. But that is not the only way of dealing with what is thought to be a blatant fiction. It can be argued that sovereign immunity should be recognized as constituting the basic rule and that what is wrong is the undermining of this rule by an obvious sham. Not long ago, a majority of the Supreme Court justices inclined to just this view, expressing doubt that there is a “principled basis” for the “fiction of the [Ex parte] Young opinion,” and stating that it should be kept as “a very narrow exception” to the sovereign immunity doctrine.14 The Court added: “For present purposes, however, we do no more than question the continued vitality of the . . . doctrine in the Eleventh Amendment context.”15 This challenge to the Ex parte Young doctrine has not borne fruit so far.
The Ex parte Young suit against the officer has presented the Court with a serious dilemma. There must be limits to the redress afforded against the officer, or else nothing would be left of the government’s immunity. The Court’s attempts to devise a formula for dealing with this dilemma have been unsuccessful. For a long time the Court maintained that a judgment against the officer would be disallowed if it resulted in interference with the government’s administration of its laws. Consistent application of such a test would have barred most Ex parte Young suits, and this did not happen. This test was succeeded by the one now in force—relief may be granted if it operates prospectively, but not if it operates retroactively. The test is simple but unworkable, and in fact the Court does not follow it.
It is submitted that the problem is solved if we look, not to the rationalizations attempted by the Court, but rather to its actual holdings. These form a pattern. It appears that in the suit against the officer, a plea of sovereign immunity is disallowed when the immunity would operate offensively, but not when it would operate defensively. If the claimant is seeking only to be left alone and charges that past or [*PG492]prospective conduct of government officers is unlawfully intrusive, judicial inquiry into the validity of such conduct would be precluded if a plea of sovereign immunity would be sustained. This would constitute what is here called offensive use of the immunity, and such use is disallowed. On the other hand, when the claimant is seeking some affirmative advantage from the government, like payment of its debt, a plea of sovereign immunity is sustained, in what is here called defensive use of the immunity.16
Apart from recognition of this pattern, our understanding of the law pertaining to officer liability will be enhanced by recognition that not every suit against an officer that also affects the government calls for exercise of the Ex parte Young fiction. If the pertinent statute is not under attack, and the claim is based only on the officer’s failure to perform a nondiscretionary duty, the officer is routinely held to performance of this duty. Such a suit is not deemed to be one against the government. When this is not recognized, as in the notorious Supreme Court decision in Pennhurst State School & Hospital v. Halderman,17 mischief can result.
There will be some discussion of topics bearing on obtaining money from the government itself, apart from legislation expressly consenting to such suit. One of these involves judicial use of the writ of mandamus for access to funds from the general treasury, with the unexplained assertion that sovereign immunity is no bar to such relief. An explanation will be ventured. Of special interest are recent Supreme Court decisions in a tax refund case and in an inverse condemnation case, which can be read to herald the demise of sovereign immunity in a broad number of contexts. The question is whether such a reading is justified.18
One of the concluding points to be discussed will be the question of how far federal jurisdiction may be exercised on the basis of a state’s consent to be sued in its own courts—a troublesome area where the Court has been accused of inconsistency in answering this question differently on the appellate and trial levels.19 In another concluding point, it will be argued that the Court erred when it ruled generally in Nevada v. Hall20 that a state need not recognize the sover[*PG493]eign immunity of a sister state, although the writer believes the result was proper under the circumstances of that case.21
When ratification of the proposed Constitution was being considered in New York, Alexander Hamilton made the following much-quoted statement:
It has been suggested that an assignment of the public securities of one State to the citizens of another would enable them to prosecute that State in the federal courts for the amount of those securities; a suggestion which the following considerations prove to be without foundation.
It is inherent in the nature of sovereignty not to be amenable to the suit of an individual without its consent. This is the general sense and the general practice of mankind; and the exemption, as one of the attributes of sovereignty, is now enjoyed by the government of every State in the Union. Unless, therefore, there is a surrender of this immunity in the plan of the convention, it will remain with the State, and the danger intimated must be merely ideal. . . . [T]here is no color to pretend that the State governments would, by the adoption of that plan, be divested of the privilege of paying their own debt in their own way, free from every constraint but that which flows from the obligations of good faith. . . . [T]o ascribe to the federal courts, by mere implication, and in destruction of a pre-existing right of the State governments, a power which would involve such a consequence, would be altogether forced and unwarrantable.22
Hamilton’s contemporaries did not express disagreement concerning the universality of sovereign immunity, apart from possible questions arising from adoption of the Constitution. James Wilson, who had been a delegate to the Philadelphia Convention, was strongly [*PG494]of the view that the immunity had no place in a polity which lacked a king, and in which a written constitution prescribed the limits of government—from which it followed that the immunity would be denied to the federal as well as the state governments. These views were expounded in his opinion in Chisholm v. Georgia, which was decided in 1793 shortly after adoption of the Constitution.23 Chief Justice John Jay, in his own opinion in Chisholm, expressed somewhat similar views, save as to the suability of the United States.24 No evidence exists, however, that these views were shared to any significant extent. In general, the sovereign immunity of the United States was not questioned. Further, it was assumed that the states were free to bar suits against themselves in their own courts.25 The debate among the leading statesmen of the time centered almost exclusively on whether the states, without their consent, were suable in the federal courts, in light of the provision in Article III extending the federal judicial power to controversies “between a State and Citizens of another State . . . and between a State . . . and foreign States, Citizens or Subjects.”26
Some argued that by reason of this grant of jurisdiction a state was suable despite a plea of sovereign immunity. Among those so contending were George Mason,27 Edmund Randolph28 and Patrick Henry.29 Alexander Hamilton, as has been seen, argued differently, as [*PG495]did James Madison30 and John Marshall.31 In accordance with their view, the states could be plaintiffs in litigation under the Article III clauses indicated above, but could rely on sovereign immunity if sued.32 If sovereign immunity could not be interposed, the states faced large claims by reason of their heavy indebtedness and by reason of their violations of the terms of the Peace Treaty of 1783 with Great Britain through confiscations and escheats.33
In ratifying the Constitution, several of the states proposed amendments to preserve the sovereign immunity in whole or in part. It has been argued that this shows an understanding that, absent amendment, the states were suable.34 But two of the amendments, proposed by Rhode Island and New York respectively, cannot be so described. Rhode Island proposed an amendment that would have eliminated any suit by any person against a state in federal court. But the drafters was apparently assumed that the states would not lose [*PG496]their sovereign immunity under the Constitution, inasmuch as the purpose of the amendment was said to be “to remove all doubts or controversies respecting the [issue].”35 In New York, the proposed amendment itself announced an understanding that the federal judicial power did not “authorize any suit by any person against a state,”36 and the ratifying convention declared that the Constitution was being ratified “[u]nder these impressions.”37 Thus, it seems that the amendments proposed by Rhode Island and New York had only a clarifying purpose.
On the other hand, Virginia and North Carolina advanced amendments that were far-reaching.38 It may be doubted, however, that any of the proposed amendments reflected serious apprehension that state sovereignty would be lost by adoption of the Constitution. The fact is that virtually no attempt was made to protect the immunity of the states when the First Congress was considering what became the first ten Amendments,39 and also the Judiciary Act of 178940—in sharp contrast to the alacrity with which the Eleventh Amendment was adopted after the decision in Chisholm v. Georgia41 showed that the states were indeed vulnerable to suit.
The general understanding regarding sovereign immunity was manifested by the reaction to Chisholm. In this case a citizen of South Carolina invoked the original jurisdiction of the Supreme Court in an assumpsit action against Georgia, for non-payment under a contract to furnish supplies to the state during the Revolutionary War. The [*PG497]Court, by a vote of four to one, held that under Article III the state could be sued without its consent by a citizen of another state.42 The reaction was speedy and angry. Georgia’s House of Representatives adopted a bill making it a capital offense to attempt to levy a judgment in the case.43 The Massachusetts and Virginia legislatures called for a constitutional convention to reverse the decision; and such a call was soon under consideration by the legislatures of eight additional states, where it had “strong support.”44 But the Eleventh Amendment quickly went through Congress, and the requisite number of state ratifications was achieved within two years of the Chisholm decision.45
In sum, at the time of adoption of the Constitution, it was generally assumed that the states were protected by sovereign immunity if sued in their own courts. Hamilton and those of like views insisted that the states would be similarly protected if sued in the federal courts, and that the state-noncitizen clauses of Article III did not contemplate a contrary result. Others said they were unconvinced. But subsequent to ratification there was virtually a total lack of effort to secure such protection for the states, despite ample opportunity to do so—which supports the conclusion that no significant doubt existed that sovereign immunity was expected to survive the Constitution. So too does the speedy adoption of the Eleventh Amendment following Chisholm, and the silence on the sovereign immunity issue when provision was briefly made for federal question jurisdiction in the Judiciary Act of 1787. Therefore, the writer is persuaded that the case for an original understanding on state sovereign immunity is a strong one. As will be shown, the Court has given effect to this understanding [*PG498]from the start, save in one decision which was soon overruled. The essentials of the argument set forth above were advanced recently in Alden v. Maine.46 Clear holdings along the same line were Hans v. Louisiana47 and Principality of Monaco v. Mississippi.48
Justice Brennan’s dissenting opinion in Atascadero State Hospital v. Scanlon, in 1985, was the first full-scale judicial assault on the doctrine of sovereign immunity.49 It remains a principal basis for the continuing attack on sovereign immunity by a minority of the justices, usually not less than four in number.50 Further, in its basic approach it reflects the dominant view of academic writers.51
Justice Brennan emphasized the inconclusive character of the discussion of sovereign immunity at the ratification debates.52 He maintained that the amendments suggested by some of the ratifying states to preserve state immunity in whole or in part were indicative of a “felt need” on this point,53 yet he failed to explain, or even discuss, the fact that subsequently these states, through their legislative representatives, made virtually no effort to act in accordance with this “felt need,” despite ample opportunity to do so.
But in major part, he purported to show that the proponents of sovereign immunity were not concerned with loss of the immunity when claims were founded on federal law but only when founded on state law. He said that “virtually” all the discussion during the ratification debates was centered on suits against the states on their debts.54 Such suits, he maintained, would be based on state law and [*PG499]would be brought in federal court under one of the diversity clauses.55 On the other hand, he said, “[t]he debates do not directly address the question of suits against States in . . . federal-question cases, where federal law and not state law would govern.”56 From this he inferred that the debates disclosed a willingness to surrender state sovereign immunity in regard to federal claims.57
This conclusion is wildly implausible. Consider Charles Warren’s statement of some of the problems facing the newly independent colonies immediately prior to adoption of the Constitution:
In the crucial condition of the finances of most of the States at that time, only disaster was to be expected if suits could be successfully maintained by holders of State issues of paper and other credits, or by Loyalist refugees to recover property confiscated or sequestered by the States; and that this was no theoretical danger was shown by the immediate institution of such suits against the States in South Carolina, Georgia, Virginia and Massachusetts.58
This state of affairs suggested that the states would be beset by claims founded on the Contracts Clause, and on Peace Treaty violations, unless they had the protection of sovereign immunity. It should take weighty arguments to persuade us that the states were content to give up such protection. The arguments advanced by Justice Brennan are singularly unpersuasive.
Thus, Justice Brennan was wrong in insisting that when jurisdiction rested on diversity of citizenship the claimant would necessarily [*PG500]be “asserting a cause of action based on state law.”59 More importantly, even if correct in this regard, he was wrong in his assumption that such a case would be “governed,” presumably in its entirety, by state law.60 Take, for example, a suit founded on violation of the Contracts Clause. Justice Brennan remarked that “it was certainly not clear at the time . . . that the Contracts Clause provided a plaintiff with a private right of action for damages.”61 This was indeed true.62 But if suit could not be brought “on” the Contracts Clause,63 it did not follow that the Clause was a dead letter. It could always have been invoked (as could the Peace Treaty) to challenge the constitutional adequacy of a state defense.
The system of common-law pleading then universally in force provided ample opportunity for vindication of federal rights. The plaintiff had to choose the particular form of action appropriate under the circumstances, and the permissible contents of the several pleadings under that form of action were rigidly prescribed. If, say, a plaintiff wanted to recover land seized by a state in violation of the Peace Treaty and now in the possession of a third person under deed from the state, an appropriate form would be the one for ejectment. The declaration, as the initial pleading was called, would allege only [*PG501]that the plaintiff was entitled to possession and that the defendant was wrongfully in possession; possible federal issues determinative of the outcome could not be anticipated.64 In the answer, under the rules applicable to that plea, the defendant would claim entitlement under a deed from the state. In the replication, which was the plaintiff’s responsive plea, the claim would be made that the deed was void because the state’s seizure of the land was in violation of the Peace Treaty.65 In form, the suit was instituted as one for vindication of a state-created right. In substance, the suit was “on” the Peace Treaty.66
This, essentially, is what happened in Martin v. Hunter’s Lessee, where an action in ejectment ended with a holding that Virginia’s escheat of British-owned property was in violation of the Peace Treaty.67 The suit was instituted in a Virginia state court. In Sturgis v. Crowninshield, an action in assumpsit brought in a federal circuit court, resulted in a decision vindicating the plaintiff’s claim that the State had impaired a contractual right under the Contract Clause.68 There was a similar holding in Fletcher v. Peck.69 There the action was in covenant [*PG502]and was also instituted in a federal circuit court. So too, the Dartmouth College Case, enforcing a claim under the Contract Clause, was commenced in a New Hampshire state court as an action in trover.70 These were all suits against private persons. The system of common-law pleading prevailed in actions at law in the federal courts irrespective of the source of jurisdiction. Even after the advent of general federal question jurisdiction in 1875, with the passage of the Judiciary Act of 1875, that jurisdiction could not properly be invoked if a declaration in an action at common law would not have disclosed the federal character of the claim.71 The states themselves were protected against such suits only by their sovereign immunity.
Furthermore, underlying Justice Brennan’s analysis was a fundamental error concerning the nature of jurisdiction. The source of a court’s jurisdiction is immaterial to the legal issues that come before it. Having jurisdiction, a court administers justice in accordance with the law applicable to the particular controversy, whether it be federal law, state law or “the laws of the most distant part of the globe.”72 This, Hamilton said, is “the nature of judiciary power . . . the general genius of the system.”73 The point is obvious. A court’s jurisdiction may be solidly founded, but it would be the grossest injustice if, in the exercise of that jurisdiction, judgment were rendered on a basis other than the applicable law. There can be no doubt that the Founding Fathers, or the lawyers among them, were aware that in any litigation, whether in federal or state court, federal law would be considered if presented to the court. Whether preservation of sovereign immunity was contemplated by the Framers was a separate question; and, indeed, the history recounted above affords solid basis for the understanding that the sovereign immunity of the states would be preserved.74
Justice Brennan was aware that claims might be made for vindication of rights claimed under federal law, but he thought this could be done only be invoking the federal question jurisdiction. Since the First Congress did not provide for such jurisdiction, he thought it evi[*PG503]dent that “Congress had decided that [such] cases, even those arising under the Treaty of Paris, should be heard in the first instance only in state courts.”75 He did not recognize that such cases could “in the first instance” be heard in general courts on the basis of diversity of citizenship.
In sum, Justice Brennan’s basic position in his Atascadero dissent was that the stated fears of the Framers concerning the loss of state sovereign immunity related only to cases arising under the diversity clauses, which were the only clauses they talked about, and which he thought were governed in their entirety by state law. On the other hand, he argued that the absence of similar stated apprehensions regarding the federal question clause shows that the Framers acquiesced in the overriding of state immunity in regard to federal claims. He did not consider the possibility that issues of controlling federal law might arise in cases brought under the diversity jurisdiction.
Why was there relative silence on the issue of federal question jurisdiction during the ratification debates? Justice Brennan’s answer was convoluted and unpersuasive. It is submitted that there is a more plausible explanation—namely, that Article III’s provision for federal question jurisdiction was not perceived to be a threat to state sovereign immunity, in contrast to the diversity clauses, which expressly spoke of the state as a party. Madison and Marshall argued that, despite the generality of the diversity clauses, the state could be a party only as plaintiff.76 Their opponents argued that the language of Article III did not warrant such a limitation.77 That was the only issue debated. It is impossible to find in the ratification debates or elsewhere any suggestion, direct or indirect, that the states would lose their sovereign immunity by virtue of the constitutional provision for federal question jurisdiction.
In Seminole Tribe v. Florida, the Court, in reliance upon what it took to be the original understanding, invalidated a federal statute overriding state sovereign immunity.78 Justice Souter’s dissenting opinion—joined by Justices Ginsburg and Breyer—rested primarily on Justice Brennan’s dissenting views in Atascadero. But Justice Souter [*PG504]made some additional points. Declaring that sovereign immunity is a common-law rule, “derived from the laws and practices of our English ancestors,”79 he emphasized that the reception of the English common law by the states did not have a counterpart on the federal level.80 He noted that the Framers had “an aversion to a general federal reception of the common law.”81 Justice Souter concluded that, given these circumstances, “the Court today cannot reasonably argue that something like the old immunity doctrine somehow slipped in as a tacit but unenforceable background principle.”82
But a strong policy against a general reception of the common law did not preclude use of aspects of the common law compatible with the basic federal scheme. Justice Souter noted that the Constitution itself in effect incorporated aspects of the common law, including its provision for habeas corpus and its distinction between law and equity.83 To state that the rule of sovereign immunity is not comparable and that it was “slipped in” is to ignore the fact that the issue was extensively debated in the state ratification conventions. This of course assumes that sovereign immunity is a common-law doctrine. If, as is argued below, it is not, Justice Souter’s argument becomes irrelevant.84
In Alden v. Maine, in 1999, the Court held that Congress could not, under its Article I powers, subject a state to suit in a state court without state consent.85 Justice Souter’s dissenting opinion incorporated by reference the arguments advanced in the dissenting opinions in Atascadero and Seminole Tribe, and then presented at great length an argument never advanced in any prior opinion, nor, so far as the writer is aware, in any academic commentary. Justice Souter argued that the holding in Alden, and the Court’s prior holdings to the same effect, are supportable only if sovereign immunity is a natural law doc[*PG505]trine.86 Justice Souter saw only two possibilities: either sovereign immunity is a common-law principle, in which case it is defeasible by legislative action,87 or it is an aspect of “natural law, a universally applicable proposition discoverable by reason.”88 Conceived as natural law, sovereign immunity is “unalterable,”89 “indefeasible,”90 “untouchable and untouched by the Constitution.”91 Justice Souter declared that the majority in Alden relied on sovereign immunity as a natural law doctrine in rejecting congressional power to override state sovereign immunity.92
The dissent understood Justice Holmes’s opinion in Kawananakoa v. Polyblank93 as “embodying . . . [the] natural law theory of sovereign immunity.”94 The relevant passage in Holmes’s opinion was as follows: “A sovereign is exempt from suit, not because of any formal conception or obsolete theory, but on the logical and practical ground that there can be no legal right as against the authority that makes the law on which the right depends.”95 This passage was not a statement that the immunity is immutable.96 If the legislature, which is [*PG506]the “authority that makes the law,” abolishes the immunity, a suit thereafter instituted is not the assertion of a right against this “authority.” What is more to our purpose is that, as emphasized by the dissent, under the above formulation, “sovereign immunity may be invoked only by the sovereign that is the source of the right upon which suit is brought.”97 The dissent did not understand that it was undermining its own position. Since the right in issue in Alden was federally created, only the federal government—and not the states—could make a claim of sovereign immunity under the natural law formulation of Justice Holmes. Hence, in upholding a state claim of sovereign immunity, the Court was not applying natural law doctrine. On the other hand, it was the dissenting opinion that comported with natural law doctrine, in its insistence that the state could not interpose sovereign immunity in regard to a federal claim.98
In the course of their discussion, the dissenters said that ultimately their position rested on their conception of sovereign immunity as a common-law rule, defeasible by statute.99 But here, we find the minority laboring under another misconception. Whether or not the sovereign immunity of the federal government is common-law doctrine, it is defeasible by Congress. Similarly, the sovereign immunity of, say, the state of Illinois, whether or not resting on common law, is defeasible by the Illinois legislature. However it does not follow that Congress can override the sovereign immunity of Illinois. The question of federal power in this regard is subsumed in a larger question—federal power to override state law. The answer must be found in the Constitution.100
In Kimel v. Florida Board of Regents, in 2000, the Supreme Court invalidated the Age Discrimination in Employment Act of 1967 [*PG507](“ADEA”) insofar as it abrogated state Eleventh Amendment immunity from suit by private individuals.101 The dissenting opinion of Justice Stevens advanced yet another argument against the constitutional status of sovereign immunity as it pertains to the states.102 His view was that the subject was entirely within the province of Congress. Drawing on a well-known article by Herbert Wechsler,103 Justice Stevens declared that “the normal operation of the legislative process itself would adequately defend state interests from undue infringement.”104
Wechsler had concluded that the Supreme Court should move slowly in invalidating congressional legislation affecting the states.105 Wechsler wrote broadly, without mention of sovereign immunity. In any event, his insistence that the states could be counted upon to protect themselves overlooked the reality that state interests often clash, and often on a regional basis. These clashes have occurred throughout our history as a nation. One need only recount the pre-Civil War disputes over tariffs; federal spending on internal improvements; the Bank of the United States; and the struggles between North and South over the status of African-Americans that culminated in the Civil War, and continued after that war. There have been continual regional clashes over water rights. Moreover, it has long been evident that state interests vary widely in such matters as labor and regulatory legislation generally. The list is endless. Indeed, the recent federal statutes adopted under Section 5 of the Fourteenth Amendment, which provided for suit against states in a variety of situations, were presumably the work of pressure groups not equally potent in all states.106 Accordingly, Justice Stevens was unpersuasive in his Kimel dis[*PG508]sent when he said that “once Congress has made its policy choice, the sovereignty concerns of the several States are satisfied.”107
Academic critics are for the most part in agreement regarding the basic point made by Justice Brennan in his Atascadero dissent—the Framers contemplated that claims founded on federal law would not be barred by sovereign immunity.108 They commonly argue that delegates to the state ratification conventions who feared loss of the immunity were actually in the majority of those who spoke on the subject. Often, however, these critics do not go beyond the debates, ignoring the subsequent events that tended to show that such fears were not entertained seriously.109 Others do not discuss the question of an original understanding at all, in reliance on the work of earlier critics, or on the apparent assumption that parts of the problem can be analyzed adequately without regard to what this writer believes to be the overarching principle of an original understanding.110 The commentators who have examined the problem in depth are relatively few in number, and they are predominantly hostile to the conception of an original understanding protective of state sovereign immunity.111
[*PG510][*PG509] Some of the critics who reject the notion of an original understanding112 rely on general language employed by Chief Justice Marshall in Cohens v. Virginia.113 In this case, the state of Virginia imposed [*PG511]a criminal sentence on two of its residents for violation of a state statute forbidding the sale of lottery tickets, in the face of a defense that federal law permitted such sale.114 When the defendants sought review in the Supreme Court, the State contended that the judgments of its courts were not federally reviewable on the ground of sovereign immunity.115 The Supreme Court said flatly: “We think a case arising under the constitution or laws of the United States, is cognizable in the Courts of the Union, whoever may be the parties to that case.”116 But the opinion as a whole belied the generality of the quoted language, a point commonly ignored by critics.117 In the first place, the opinion makes plain that this remark concerned the Constitution as it “originally stood,” prior to adoption of the Eleventh Amendment.118 What is more important, the Court limited the scope of the quoted language by suggesting that judicial disregard of state sovereign immunity was not contemplated.119
[*PG512] Some commentators, who assume that there is legislative power to abolish sovereign immunity, claim the same power for the judiciary, in reliance on the maxim that the judicial power is coextensive with the legislative.120 The problem, as with maxims generally, is that the law refuses to conform with what the maxim prescribes. Under the American constitutional system, the legislature of each component—the nation and the states—has authority to waive sovereign immunity for that component. It does not follow that Congress (let alone the federal judiciary) has power to override the immunity of the states. The writer has found no indication that this distinction is recognized by the critics. Apart from that, judicial power is not necessarily coterminous with legislative power even within a particular jurisdiction. For example, in the case of the federal government, Congress has power, under the Commerce Clause, to establish a regulatory regime for railroads and other utilities, or a law of labor relations—in both instances with a scope unknown to the common law. There is no basis in the Supreme Court’s jurisprudence for the view that the federal courts are free to make like rules, on the theory that they are acting in areas within the potential reach of Congress.121
It has been argued in effect that even if sovereign immunity had been contemplated by the Founders, the subsequent expansion of federal power and the corresponding contraction of state competence—most notably in consequence of the Civil War Amendments—warrant elimination of the doctrine.122 In an earlier article, the writer advanced the following thesis: (1) that the Constitution should always be construed in accordance with the intent of the Framers, or at least their probable intent, as best that can be ascertained; (2) that when the provisions of the Constitution permit of only one interpretation, like those establishing the two houses of Congress, the intent is sufficiently clear, and such provisions should be modifiable only through the amendment route; (3) that when the language of a provision is open-ended in that it is fairly susceptible to more than a single interpretation, the provision should be construed as the Framers most probably would have wanted it construed at the time of decision, assuming foreknowledge on their part of conditions existing at that [*PG513]time; and (4) that the greater the public consensus on a particular issue at such time, the greater the likelihood that the Framers would have wanted the provision to be construed in accordance with that consensus.123
Where the structural provisions of the Constitution are concerned, modification in light of later developments is singularly inappropriate, for the structural provisions are not open-ended. A statement in the Constitution that the states are not suable, as in the case of the Eleventh Amendment, does not lend itself to more than one construction. So too, whether deemed structural or not, is the general understanding on state sovereign immunity, which the Court later referred to as a controlling “postulate”124—as part of “the understood background against which the Constitution was adopted.”125 At least so much seems clear when the construction is one that in effect eliminates sovereign immunity altogether. It is arguable, however, that the immunity is accorded its intended scope when implemented in accordance with a current consensus on its meaning. On this basis, governmental liabilities arising out of their commercial activities could be deemed unprotected by sovereign immunity.126
The Eleventh Amendment provides as follows: “The Judicial power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the [*PG514]United States by Citizens of another State, or by Citizens or Subjects of any Foreign State.”127 By its terms, the Eleventh Amendment applies to “any” suit brought against a state by a noncitizen, regardless of the source of jurisdiction. Further, read literally, it would embrace a suit founded on federal law as well as state law. It is submitted that the literal reading should prevail, and others have also taken this view.128 The overwhelming majority of academic commentators, however, claim that the Eleventh Amendment should not be deemed to apply in cases founded on federal question jurisdiction.129 Rather, they construe the Amendment as if it read: “Any suit in which jurisdiction is founded on diversity of citizenship.”
For this construction, the commentators observed that the Framers of the Eleventh Amendment took no action on an earlier version of the Amendment under which the states would have been protected from suit by “any person or persons, citizens or foreigners.”130 Virtually without exception, they maintain that a suit against a state by one of its own citizens could have been brought only under the federal question jurisdiction.131 In the circumstances they stress what seems to [*PG515]them to be the pointed omission of any reference to federal question jurisdiction in the final form of the Eleventh Amendment, which they say simply tracks the diversity language of Article III.132 Finally, they contend that if the drafters wanted to protect the states from federal claims, the Eleventh Amendment was poorly drafted to that end because it applied only to suits by noncitizens. Since the drafters could easily have protected the states from suits by their own citizens as well, but failed to do so, the argument goes, it is reasonable to read the Amendment as not protecting against federal claims at all.133
Even if the critics are correct in contending that the Eleventh Amendment should be read otherwise than in accordance with its express language—that it should be taken to apply only when jurisdiction was founded on diversity of citizenship—the Eleventh Amendment has an effect opposite to the one for which they contend. Critics assume that the source of jurisdiction is determinative of the issues, federal or state, that the court is called upon to determine. Justice Brennan made this assumption in his dissenting opinion in Atascadero, and this was no accident, for he credited his views to academic commentators.134
Whatever the source of jurisdiction, as has been seen, the system of common-law pleading universally in effect at the time, allowed ample scope for the injection of issues of controlling federal law into the case.135 As it happened, the only relevant head of jurisdiction at the time was the diversity jurisdiction.136 If, as the critics contend, the only jurisdiction to which the Eleventh Amendment applies is the diversity jurisdiction, the federal courts were stripped of power to hear any claims that might be presented by those invoking that jurisdiction, including claims founded on federal law.
[*PG516] Moreover, it is evident that the Amendment’s drafters understood this problem. Thus, when the Eleventh Amendment was under consideration by Congress, Senator Albert Gallatin of Pennsylvania proposed that the Amendment should not apply “in cases arising under treaties made under authority of the United States.”137 The proposal was not adopted.138 The reason should be apparent. As has been observed, the “states . . . most ardent in advocating the amendment—Massachusetts, Virginia, and Georgia—all faced pending claims in the Supreme Court that posed issues turning upon interpretation of the Constitution or federal treaties.”139 It is plain that the Amendment as drafted was expected to protect the states from treaty claims.
The absence in the Eleventh Amendment of any protection for the states against suits by their own citizens should not be taken to indicate acquiescence in such suits. It had been the general understanding that, absent consent, the states were immune from suit by citizens and noncitizens alike.140 Chisholm disputed this understanding, and the Eleventh Amendment should be read as no more than a repudiation of Chisholm.141 As has already been remarked, during the Constitution’s formative period, those who expressed apprehension of possible overriding of state sovereign immunity by virtue of Article III made the point that the Article’s diversity clauses expressly spoke of states as parties.142 Since they did not express similar apprehension of the federal question jurisdiction, and indeed scarcely mentioned it at all, this bespeaks confidence on their part that the states would not be suable under that jurisdiction, and sufficiently explains why there was no mention of the jurisdiction in the text of the Eleventh Amendment. Finally, even if it is assumed that the Eleventh Amendment is not applicable to federal claims, it is submitted that the states are protected against such claims by virtue of the original understand[*PG517]ing that their sovereign immunity survived adoption of the Constitution.
In Hans v. Louisiana, the State of Louisiana defaulted on state-issued bonds, and one of its own citizens sued the state in a federal circuit court on coupons attached to the bonds, claiming that the State’s default violated the Contracts Clause of the Constitution.143 The claimant maintained that, under Article III of the Constitution and the implementing jurisdictional legislation, a state was suable by its own citizens on cases arising under the Constitution, laws and treaties of the United States. The Court acknowledged that the Eleventh Amendment was inapplicable by its terms because no diversity of citizenship existed,144 but sustained the defense of sovereign immunity.145 The Court observed that in earlier cases, where the claimant was a non-citizen and the Eleventh Amendment was therefore applicable, the holding had been that a non-consenting state is not suable on such a claim.146 The Court added that it would be “anomalous”147 to allow such a claim when brought against a non-consenting state by one of its own citizens.
The academic verdict on Hans has been overwhelmingly negative.148 It has been contended: (1) that, despite the Court’s disclaimer, the Court did indeed rely on the Eleventh Amendment;149 and (2) that the Amendment did not apply when a federal claim was asserted against a state.150 Regarding the first point, critics have relied heavily on the Court’s statement that it would be “anomalous” 151 to allow a claim against a state by one of its own citizens, in light of the Amendment’s bar of suit by a noncitizen. But the Court also wrote more broadly, declaring in effect that the broad consensus at the time of adoption of the Constitution was that a state could not be sued by any [*PG518]person on any cause of action. Thus, the Court said: “[T]he cognizance of suits and actions unknown to the law, and forbidden by the law, was not contemplated by the Constitution when establishing the judicial power of the United States. . . . The suability of a State without its consent was a thing unknown to the law.”152
The Court also said that the “shock of surprise”153 occasioned by the Chisholm holding, leading as it did to prompt adoption of the Eleventh Amendment, represented a repudiation of the notion of federal judicial power “to entertain suits by individuals against the states . . . [a power that] had been expressly disclaimed, and even resented, by the great defenders of the constitution while it was on its trial before the American people.”154 Adoption of the Amendment, said the Court, showed that “the highest authority of the country was in accord rather with the minority than with the majority of the court in . . . Chisholm.”155
With particular reference to a federal-question claim the Court declared:
Can we suppose that, when the eleventh amendment was adopted, it was understood to be left open for citizens of a state to sue their own state in the federal courts, while the idea of suits by citizens of other states, or of foreign states, [*PG519]was indignantly repelled? Suppose that congress, when proposing the eleventh amendment, had appended to it a proviso that nothing therein contained should prevent a state from being sued by its own citizens in cases arising under the constitution or laws of the United States, can we imagine that it would have been adopted by the states? The supposition that it would is almost an absurdity on its face.156
As for the argument of critics that the Eleventh Amendment does not cover the case of a claim against a state based on federal law, this argument is based on both their narrow reading of the Amendment and on their rejection of the idea that there was an original understanding that the sovereign immunity of the states would be preserved.157 The writer has already attempted to show that such criticism is without merit.158 Hans has commonly been regarded as a major and surprising turning point in American law—as the first case to suggest, let alone hold, that state sovereign immunity could be sustained as matter of constitutional right, independent of the Eleventh Amendment.159 However, Hans was foreshadowed in Cohens v. Virginia160 and in Osborne v. Bank of the United States.161
The Supreme Court has followed Hans many times in suits by persons against their own states on constitutional claims.162 The Court has also extended the Hans holding to other types of cases that involved federal claims and which were outside the express terms of the Eleventh Amendment. In Smith v. Reeves, the Court sustained a plea of sovereign immunity where a state was sued by a federal corporation, which was not a citizen by the terms of the Eleventh Amendment.163 Also, in Ex parte New York, the Court held that the Eleventh Amendment was inapplicable because the suit originated as an admiralty proceeding in a federal district court.164 Although the Eleventh Amendment spoke only of suits in “law or equity,” not admiralty,165 the state’s plea of sovereign immunity was sustained “because of the fundamental rule of which the amendment is but an exemplification.”166
In Principality of Monaco v. Mississippi,167 in 1934, the Supreme Court gave the issue of sovereign immunity further explication. The Principality of Monaco had invoked the original jurisdiction of the Supreme Court in a suit against the state of Mississippi on defaulted state bonds.168 The Eleventh Amendment does not apply to suits by a foreign state. Monaco argued that the Supreme Court’s jurisdiction under Article III was clear and that the Article made no reference to the need for consent by a defending state in the case of a suit by a foreign state.169 The Court, however, observed that the Article similarly conferred jurisdiction in cases to which the United States is a [*PG521]party, and that the consent of the United States to suit against it had always been required.170 The crux of the opinion was as follows:
Manifestly, we cannot rest with a mere literal application of the words of section 2 of article 3, or assume that the letter of the Eleventh Amendment exhausts the restrictions upon suit against non-consenting States. Behind the words of the constitutional provisions are postulates which limit and control. There is the essential postulate that the controversies, as contemplated, shall be found to be of a justiciable character. There is also the postulate that States of the Union, still possessing attributes of sovereignty, shall be immune from suits, without their consent, save where there has been “a surrender of this immunity in the plan of the convention.” . . . The question is whether the plan of the Constitution involves the surrender of immunity when the suit is brought against a State, without her consent, by a foreign State.171
The Court went on to show why, under the “plan of the convention,” a state should not enjoy sovereign immunity when sued by the United States,172 or when sued by a sister state in the Supreme Court.173 On the other hand, the Court gave reasons why the constitutional scheme should not be deemed to subject a non-consenting state to suit by a foreign state.174
Although Hans and the cases just discussed made clear that state sovereign immunity exists independently of the Eleventh Amendment, in recent decades the Court, in what seems to have been careless usage, often spoke of the Eleventh Amendment as the source of the immunity in cases where the Amendment clearly did not apply.175 This point was clarified by the Court in Alden:
[*PG522][W]e have . . . sometimes referred to the States’ immunity as “Eleventh Amendment Immunity.” The phrase is convenient shorthand but something of a misnomer, for the sovereign immunity of the States neither derives from nor is limited by the terms of the Eleventh Amendment. . . . Rather . . . the States’ immunity is a fundamental aspect of the sovereignty which the States enjoyed before the ratification of the Constitution, and which they retain today (either literally or by virtue of their admission into the Union upon an equal footing with the other States) except as altered by the plan of the Constitution or certain constitutional Amendments.176
Sovereign immunity is widely understood by judges177 and academic critics178 to be a branch of the common law, and subject to ju[*PG523]dicial or legislative modification, like any other branch of the common law. Professor Martha Field has made the clearest exposition of this viewpoint.179 Her reasoning is as follows: (1) Sovereign immunity was in existence as common-law doctrine immediately prior to adoption of the Constitution, and the Constitution left it untouched; (2) there was no consensus on the subject at the one time when the historical record shows it to have been considered—namely, during the debates on ratification; (3) the only constitutional provisions having any possible bearing on the subject are the jurisdictional clauses of Article III; and (4) it has been accepted that Article III by itself affords no basis for overruling sovereign immunity. For these reasons, she argued, the common-law doctrine of sovereign immunity has survived adoption of the Constitution.180
Concerning this argument, it may be helpful to distinguish two questions: (1) whether the sovereign may be sued without its consent as an aspect of the internal law of the particular jurisdiction, federal or state; and (2) whether the federal government has authority to override the sovereign immunity of the states. These questions are governed by different considerations (a point not noted by the critics). As will now be argued, the concept of sovereign immunity as common-law doctrine has little if any pertinence to the first, and no pertinence at all to the second.
The sovereign immunity law traditionally applied has none of the significant incidents of common-law doctrine. First, it is doubtful that even the English law of sovereign immunity at the time of the American Revolution can be meaningfully described as an aspect of the English common law. The term, common law, is stretched beyond its generally recognized meaning if it is said to embrace judicial acquiescence in Royal prerogatives. It would be much like saying that the acquiescence of the federal judiciary in the exclusive power of the President and the Senate in the making of treaties is an aspect of the federal common law.
[*PG524] Second, if sovereign immunity is nevertheless regarded as common law, it is a unique kind of common law. The proposition that the sovereign may not be sued without its consent means that the consent must be that of the sovereign. In the American constitutional systems it is the legislative rather than the judicial branch that speaks for the sovereign, and it has never seriously been suggested otherwise. Certainly, it has never been thought that the judiciary can consent for the sovereign in the particular case. Yet if sovereign immunity is merely common-law doctrine, the judiciary can in effect consent in all cases, simply by abolishing the doctrine.181
Third, the notion of sovereign immunity as common law is ill-suited to a constitutional system that limits the power of government and confers rights upon people. In such a system, either the assertion of the immunity should prevail or it should not. The idea that courts should have power to decide this fundamental question one way or the other would be surprising to an intelligent person not versed in the law. For such judicial power to exist, it should be inferable from the constitution of the particular jurisdiction and success in finding such a constitution is most unlikely.
Fourth, traditionally, courts have decided the basic immunity issue in only one way—the government, as distinct from its officers, was never suable. This tends to show that the law they were applying was not common law at all, even if they called it that. Fifth, in response, presumably, to a changed public consensus on the propriety of suit against the government, American legislatures for the most part have consented to such suit in all but narrow circumstances.182 If sovereign immunity were no more than common law, one would have expected courts to respond to the same public concerns. But, except as noted below,183 they did not. Thus, on the federal level, where legislative consent to suit has proceeded in stages for over a century, the judiciary has never asserted authority to restrict the immunity along similar lines.
[*PG525] Currently courts, especially state courts, have continued to speak of sovereign immunity as an aspect of common-law doctrine.184 For the reasons given above, the law traditionally applied has none of the significant incidents of common-law doctrine. On the other hand, it is evidence of a limitation on judicial authority that can have only a constitutional source.185
In the mid-twentieth century, there was a break, on the state level, with what had previously been universal judicial practice in this regard. Convinced that sovereign immunity was an aspect of the common law, a number of courts abolished it in whole or in part. Some of the cases involved states or their instrumentalities.186 Others involved political subdivisions, whose claims to sovereign immunity seemingly did derive from the common law,187 but the judicial language was usually broad enough to encompass states as well.188 It is worth noting that in many cases the immunity was abolished only as to tort.189 There is irony in this, for the most common ground for a tort claim is negligence, which, at least on the federal level, furnishes no basis for a constitutional claim.190 Those states that have gone no further have left intact the immunity regarding constitutional violations, which is the problem of special concern to the academic critics.
In the case of sovereign immunity as an aspect of the internal law of a particular jurisdiction, the governing law rests ultimately on the [*PG526]constitution of that jurisdiction. The question of federal judicial power to override state sovereign immunity is governed exclusively by the federal Constitution. On this point, the English experience is irrelevant, because pertinent only to the problem of English internal law. The extent of the federal judicial power, if any, to override state sovereign immunity is subsumed in the more basic issue of a federal power—legislative or judicial—to set aside state law. To the extent that provisions of the Constitution are self-executing, their implementation by the federal judiciary has never been though to constitute an application of common law. When the Constitution is not self-executing and its implementation depends on legislative action, the action may be taken only pursuant to one or more of the enumerated powers of Congress. It would be anomalous in the extreme if the federal judiciary, in the administration of what it deems to be federal common law, should have a power to override state law comparable to that of Congress but exercisable in the absence of the type of constitutional limitations that circumscribe congressional action.191
The question now to be considered is whether the original understanding is pertinent to the issue of federal legislative power. The Court has analyzed the congressional power to abolish state sovereign immunity under Article I and Section 5 of the Fourteenth Amendment. We now turn to an analysis of such powers.
In Parden v. Terminal Railway,192 the statute involved, adopted under the Commerce Clause, provided that suit could be brought against interstate railroads by employees claiming injury by reason of violation of the statute. The Court held that the state had impliedly consented to be sued in that it had undertaken operation of the railroad subsequent to enactment of the statute. The question of suit [*PG527]based on a theory of implied consent is an important one, and will be discussed in a separate section below.193
Subsequently, in Pennsylvania v. Union Gas Co., also involving a statute based on the Commerce Clause, the Court ruled in effect that there was no need to find consent on a case-by-case basis. The Court reasoned that congressional power under that Clause “would be incomplete without authority to render States liable in damages,”194 and, further, that the States “gave their consent all at once, in ratifying the Constitution containing the Commerce Clause rather than on a case-by-case basis.”195 This facile approach to the problem has been abandoned, presumably in light of the repeated declarations by the Court that consent to suit must be shown by unequivocal language.196
Union Gas was subsequently overruled in 1996 by Seminole Tribe.197 This case invalidated a federal statute providing for suits against states that had been adopted under the Indian Commerce Clause. The holding was that Congress lacked such power under Article I. In Alden v. Maine,198 where the Court again invalidated an Article I statute providing for suit against a state, the Court spoke more generally, emphasizing the incompatibility of such a statute with the original understanding on the place of state sovereign immunity in the constitutional scheme. It is submitted that the Court had ample basis for this disposition of the cases. While the ratification debates were focused on the federal judicial power, the original understanding embodied a conception of the role of the states in our federal system that, in the view of this writer, limits the power of Congress (putting aside for the moment Section 5 of the Fourteenth Amendment) in the same degree that it limits that of the federal judiciary.
Some additional remarks may be ventured here concerning the Commerce Clause. First, congressional power under that Clause has expanded enormously, in step with the enormous changes in the economy. Starting about fifty years ago, successive legislative measures under the Clause were judicially validated almost as a matter of course, under the rational basis mode of review.199 The recent deci[*PG528]sions in United States v. Lopez,200 and United States v. Morrison,201 marked an abrupt halt, but the full implications of these decisions remain to be seen. Challenges to legislation as exceeding the bounds of the Commerce Clause have been treated by the Court very much like challenges based on denial of economic due process. In effect, the rational basis test has been applied in both contexts, and that is understandable. But that test is ill-suited to resolving the question of congressional competence under the Commerce Clause when a competing constitutional principle is invoked to challenge such competence. The minority Justices do not expressly rely on the rational basis test to support breach of state sovereign immunity under the Commerce Clause, but it seems to this writer that the arguments they employ are essentially rational basis arguments.202
Second, it is anomalous to require more extensive enforcement of legislatively-created rights than of rights created by the Constitution itself. Given the constitutional status of state sovereign immunity, it would be odd if the Constitution sanctioned judicial enforcement against non-consenting states of a sub-constitutional right created by Congress, such as the right to overtime pay involved in Alden.203 Suppose, however, that Congress tries to assure enforcement of a constitu[*PG529]tional right. Consider a possible statute in which Congress purports to find that non-payment of state bonds has harmful consequences for the national economy, and vests the federal courts with jurisdiction to make states pay. It is submitted that the Commerce Clause is an unlikely source of such authority, in view of the constitutional basis for state sovereign immunity.
More fundamentally, it is submitted that, whatever the scope of congressional power under Article I, it cannot be employed to set aside other provisions of the Constitution. In the section immediately below, it is contended that Congress lacks such power even under Section 5 of the Fourteenth Amendment.204 If that contention is correct, the lack of such power under Article I follows a fortiori.
The first case to find congressional authority to abolish state sovereign immunity under Section 5 of the Fourteenth Amendment was Fitzpatrick v. Bitzer.205 The suit had been instituted in a federal district court, and the state had relied on the Eleventh Amendment.206 The [*PG530]Court held that, under Section 5, Congress could override state sovereign immunity despite the Eleventh Amendment.207 Actually the Eleventh Amendment was not applicable in Fitzpatrick, where the state was not being sued by a citizen of another state.208 The Court, however, said: “[W]e think that the Eleventh Amendment, and the principle of state sovereignty which it embodies, see Hans v. Louisiana . . . are necessarily limited by the enforcement provisions of section 5 of the Fourteenth Amendment.”209 The Court’s reasoning in Fitzpatrick did not support its holding. The Court relied, as it said, on a “line of cases [that] sanctioned intrusions by Congress, acting under the Civil War Amendments, into the judicial, executive, and legislative spheres of autonomy previously reserved to the States.”210 In the same vein, the Court added:
In [section 5] Congress is expressly granted authority to enforce “by appropriate legislation” the substantive provisions of the Fourteenth Amendment, which themselves embody significant limitations on state authority. When Congress acts pursuant to § 5, not only is it exercising legislative authority that is plenary within the terms of the constitutional grant, it is exercising that authority under one section of a constitutional Amendment whose other sections by their own terms embody limitations on state authority.211
A federal statute that provides for unconsented suit against a state obviously embodies a limitation on state authority. The state, however, argued that the Eleventh Amendment was a bar to adoption of the statute. The issue before the Court, which it failed to recognize, was whether Congress can supersede a constitutional provision (and “the principle that it embodies”) that imposes a limitation on federal authority. After all, the Eleventh Amendment and its underlying principle forbid federal courts to exercise jurisdiction in specified circumstances.212 The line of cases that the Court thought controlling is rep[*PG531]resented by Ex parte Virginia,213 where the Court upheld a federal statute prohibiting exclusion from state juries on the basis of race, and Katzenbach v. Morgan,214 upholding a federal statute forbidding use of state literacy tests for voter qualification. These cases involved limitations on state authority.
In Seminole, the Court, with reference to Fitzpatrick, observed that “the Fourteenth Amendment was adopted well after adoption of the Eleventh.”215 In the same vein, a distinguished federal judge spoke of “the familiar premise that the Fourteenth Amendment trumps the Eleventh Amendment because the Fourteenth Amendment was adopted later in time.”216 As a later amendment, the Fourteenth undoubtedly supersedes the Eleventh insofar as inconsistent with it. A finding of inconsistency, however, should not be made lightly. If Congress, under section 5, is free to set aside the Eleventh Amendment, it is also free—in an indeterminate degree—to set aside other provisions of the Constitution, or at least those previously adopted. Any assumption that this follows because the Fourteenth was adopted later in time is unpersuasive, to put it mildly.
After the Fitzpatrick decision, the Court, in City of Boerne v. Flores, made clear, on the basis of numerous precedents, that congressional power “to enforce” Section 5 of the Fourteenth Amendment extends to measures of remedy but not of substance.217 Further, the Court concluded that that legislation which “alters the meaning of [a constitutional clause] cannot be said to be enforcing the . . . Clause.”218 If Congress may not alter the meaning of a constitutional clause with a view to effectuating the rights afforded by that clause,219 it decidedly lacks power to achieve this objective by superseding other provisions of the Constitution. In Fitzpatrick, the Court was oblivious to limita[*PG532]tions. The power of Congress under Section 5 was said to be “plenary,”220 and the overriding of the Eleventh Amendment seemed to follow from that.221
Two years after Fitzpatrick, the Court followed that decision in upholding an award of costs against a state.222 But statutes authorizing suits against states have been struck down in five subsequent cases.223 These invalidations have followed in part from inattention to the implications of the Due Process Clause and in part from misapplication of the “congruence and proportionality” test of Boerne.224
In Boerne, the Court observed that Congress may adopt remedial measures under Section 5 “even if in the process it prohibits conduct [*PG533]which is not itself unconstitutional and intrudes into ‘legislative spheres of autonomy previously reserved to the States.’”225 On the other hand, if a Section 5 statute is to be sustained as remedial, the Court concluded that “[t]here must be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end. Lacking such a connection, legislation may become substantive in operation in effect.”226 As an example of valid remedial legislation that has intruded on state constitutional prerogative, the Court noted the problem of state literacy tests.227 While the Constitution vests states with control over voter qualifications, the Court has upheld Section 5 legislation striking down such tests.228 Thus, the Court upheld a statute setting aside New York’s requirement of literacy in English because its effect was to deny the right to vote to “large segments” of the State’s Puerto Rican population.229 Had the number of persons adversely affected by the literacy requirement been insubstantial, the congressional intrusion on New York’s constitutional prerogative would presumably have failed Boerne’s proportionality test.
Now consider the Boerne test as it was applied in Kimel v. Florida Board of Regents.230 This case involved the Age Discrimination in Employment Act of 1967 (“ADEA”), which prohibits an employer, including a state, from failing to or refusing to hire or to discharge any individual because of such individual’s age.231 The constitutionality of this statute under Article I had been already been decided.232 The question in Kimel was whether Congress, acting under Section 5, could validly subject the states, as employers, to the liability created by the ADEA.233 The Court noted that age is not a suspect classification, so that discrimination on the basis of age does not violate the Equal Protection Clause of the Fourteenth Amendment so long as the classification is reasonably related to a state’s legitimate interest (i.e., [*PG534]if it meets the rational basis test).234 Further, the Court found that the ADEA “prohibits substantially more state employment decisions and practices than would likely be held to be unconstitutional under the applicable equal protection, rational basis standard.”235 That being so, the Court concluded that the statutory remedy was excessive under the proportionality rule of Boerne.236
Unaccountably, the question whether the state had violated the Due Process Clause in rejecting a claim under the statute was not an issue in the case. The ADEA had created a cause of action against violators. This cause of action was property, protected against state impairment under the Due Process Clause.237 The Court’s determination of lack of proportionality rested on its conclusion that only a small number of the rights created by the statute were protected by the Fourteenth Amendment. In fact, however, all the rights created by the statute were protected by the Amendment’s Due Process Clause.
A case in all fours with Kimel is Board of Trustees of University of Alabama v. Garrett, where the Court invalidated the Americans with Disabilities Act of 1990 (“ADA”) insofar as the Act authorized suit against the states.238 And a comparable case is Florida Prepaid Postsecondary Education Expense Board v. College Savings Bank, which involved a suit against Florida for patent infringement.239
[*PG535] The foregoing decisions are questionable if Fitzpatrick, which declared that Congress has “plenary” authority under Section 5, represents the governing law. It does not necessarily follow that they are wrong in the result.
The Court has decided that Congress cannot abrogate state sovereign immunity under Article I, but that it can do so under Section 5. There is a resulting anomaly in that property rights created under the Commerce Clause seemingly become enforceable against the [*PG536]States under Section 5. The Supreme Court has never come to grips with this anomaly. The Fifth Circuit Court of Appeals, however, has.
In Chavez v. Arte Publico Press, the plaintiff, invoking several Section 5 statutes,240 sued a state university for copyright infringement and violation of the Lanham Act.241 The plaintiff’s claim that her rights were protected by the Due Process Clause was clearly presented. The court confronted the question whether Section 5 legislation may be relied on to enforce rights created Article I.242 Judge John Minor Wisdom, in dissent, answered this question in the affirmative: “Congress can combine its authority under Article I and section 5 of the Fourteenth Amendment to achieve a result that would not be possible in the absence of that combination.”243 The majority of the judges disagreed. They argued that rights created Article I can easily be dressed up in due process clothes if that is all it takes to render them enforceable under Section 5.244 They said that even the statute involved in Seminole Tribe itself was amenable to “this style of constitutionalization.”245 To allow Article I rights to be enforced against states under Section 5, concluded the court, “would require us to ignore the result in Seminole [Tribe].”246
This view of the matter would severely limit the scope of Fitzpatrick.247 But that decision, as previously argued, was a dubious [*PG537]one.248 If the Fitzpatrick result is constitutionally sustainable, it is not on any ground advanced by the Court.
The Court has shown increasing reluctance to read a federal statute as a consent to sue the United States. Thus, it recently stated that such consent must be “unequivocally expressed in statutory text . . . and will not be implied.”249 It is submitted that the Court’s resistance to a finding of consent (arguably excessive in any event) is inappropriate in the case of federal statutes. While sovereign immunity is an aspect of the constitutional scheme, so too is the plenary power of Congress to waive it for the United States. Construing the exercise of this power with excessive strictness is to thwart effectuation of the congressional will without advancing any constitutional purpose.250
[*PG538] On another footing is the Court’s reluctance to find a congressional purpose to overrule state sovereign immunity. The Court, as it has said, adopted a “stringent test” in determining whether Congress intended to abrogate state immunity, because “that abrogation of sovereign immunity upsets ‘the fundamental constitutional balance between the Federal Government and the States,’ . . . placing a considerable strain on ‘[t]he principles of federalism that inform Eleventh Amendment doctrine.’”251 This same consideration does not apply in the case of congressional waiver of federal sovereign immunity.
While constitutional considerations counsel greater reluctance to find state legislative consent, the Supreme Court is unduly rigid in its insistence that a state consents to suit only if its legislature has made that purpose clear “‘by the most express language or by such overwhelming implications from the text as (will) leave no room for any other reasonable construction.’”252 If, though such a statement is lacking, the available state materials suggest that the state courts would probably construe the statute as a waiver, it would be perverse for the federal courts to conclude otherwise. In Edelman v. Jordan, involving a joint federal-state welfare program, the Court held that the State’s assertion of sovereign immunity under the Eleventh Amendment barred a suit against it in a federal court by a recipient for back benefits withheld in violation of federal law.253 Applying its clear-statement rule, the Court rejected an argument that the State had impliedly consented to suit in federal court by participation in the program through which the federal government provided assistance for the operation by the State of a public aid system.254 Furthermore, [*PG539]the Court did so without inquiry into what the state’s law on the subject might be.
In partial response to this problem, Congress in 1986 provided by statute that “[a] State shall not be immune under the Eleventh Amendment . . . from suit for a violation of . . . any . . . Federal statute prohibiting discrimination by recipients of Federal financial assistance.”255 The Supreme Court, in dictum, and speaking evidently of acceptance of a conditional federal grant, has called this “an unambiguous waiver of the . . . immunity.”256 The lower federal courts have imposed liability on states accordingly.257
It should be noted that congressional power to impose conditions on grants of assistance is not unlimited.258 One need not accept the result in New York v. United States, to acknowledge the justice of the Court’s observation that without limits “the spending power could render academic the Constitution’s other grants and limits of federal authority.”259 So far as is here pertinent, conditions must be “reasonably related to the federal interest” in the grant.260 The Supreme Court has also recognized that although a state can avoid onerous conditions by declining the federal grant, the circumstances may be such as [*PG540]to place undue pressure on the state in making that choice. In South Dakota v. Dole, the Court said:
Our decisions have recognized that in some circumstances the financial inducement offered by Congress might be so coercive as to pass the point at which “pressure turns into compulsion.” . . . Here, however, . . . all South Dakota would lose if she adheres to her chosen course as to a suitable minimum drinking age is 5% of the funds otherwise obtainable under specified highway grant programs.261
From our beginnings as a nation, it has been understood that cities, counties, and other subdivisions or agencies have no entitlement to a claim of sovereign immunity because they do not have the attributes of sovereignty.262 This at least is the rule when the suit is based on asserted violations of federal law. In many states, local subdivisions are not suable for violations of local law, usually but not always on the theory that sovereign immunity applies as a matter of state law.263
The first holding that sustained the sovereign immunity of the United States came in United States v. McLemore,264 and this position is one from which the Court has never swerved.265
[*PG541] The immunity of the federal government seems not to have drawn attention in the ratification debates nor generally during the Constitution’s formative period. The constitutional and statutory provisions endowing the federal judiciary with jurisdiction in cases to which the United States was a party did not lead to discussion regarding the suability of the United States. Nor was there litigation raising that issue, though it is possible that such litigation was headed off by early adoption of the Eleventh Amendment, which reflected a strong bias in favor of sovereign immunity generally.
The critics of sovereign immunity have focused almost exclusively on the doctrine in its application to the states. Some of their basic arguments have equal pertinence on the federal level, namely (1) that the doctrine was not meant to protect from claims founded on federal law, and (2) that sovereign immunity is at most an aspect of the common law. In the foregoing pages the writer has attempted to refute these arguments.266
Traditionally, the officer acting without valid authority was personally liable for acts that, if committed by a private person, would have been deemed trespassory at common law. When damages were assessed against the officer, it was for the commission of a common-law tort. When an injunction was granted, it was to restrain the commission of a common-law tort. This was the law in England, and it was the law on both the federal and state levels before the decision in Ex parte Young.267
It has been contended that Ex parte Young represented a departure from this basic pattern.268 In that case, the Court concluded that the Attorney General of Minnesota could be enjoined from instituting a proceeding in the state courts to enforce the terms of a regulatory [*PG542]statute alleged to be unconstitutional.269 Argument has been made that the Attorney General’s readiness “to prosecute for conduct in violation of state law was probably not tortious under traditional common law concepts.”270 But if the state law was unconstitutional, enforcement of that law would have been trespassory. Institution of the enforcement proceeding was an integral part of a process that, uninterrupted, could have culminated in confiscatory rates or statutory penalties of imprisonment and fine.271 Officers are always subject to personal liability for such conduct. But the Court dealt with a larger concern in Ex parte Young. As the Court put it, whether the conduct was trespassory turned on whether the statute was “void because unconstitutional.”272 Authority under state law is of no avail, the Court said, when the “officer in proceeding [pursuant to such authority] comes into conflict with the superior authority of the Constitution.”273 If valid authority is lacking, the officer is liable for the trespass as any private person would be for the same conduct.274 But is liability in tort an essential feature of the Ex parte Young doctrine as it has evolved? Seemingly not, as will be shown below.
Participation of state officers in maintenance of a segregated school system is not readily characterized as trespassory. By way of contrast, consider the government officers who operate a prison under inhumane conditions. Ordinarily, involuntary confinement is justified under the valid judgment of the sentencing court. But the justification fails when prison conditions do not meet minimal constitutional standards, and at that point continued confinement can arguably be viewed as trespassory. Educational segregation does not lend itself to such characterization. Yet the Court has relied on Ex parte Young in desegregation suits against state officers.275
[*PG543] Other cases of this type are not readily found in the Supreme Court Reports, but comparable situations exist, though seemingly uncommon. Thus, impairment of voting rights through unconstitutionally constituted legislative districts does not necessarily call for trespassory conduct on the part of election officials. Yet these officials—not the federal and state governments—are the named defendants in such cases.276 In such cases (and in the desegregation cases) claimants do not ask to be let alone—as when trespassory conduct takes place or is threatened—but what they seek is somewhat similar—not to be singled out for denial of constitutional rights.
Ex parte Young has also been invoked to secure statutory benefits unlawfully denied by government officers, as in the case of the future benefits involved in Edelman v. Jordan.277 In this instance, too, the official conduct is not trespassory.278 Whether or not the official con[*PG545][*PG544]duct at issue in any of these cases is otherwise tortious under the common law need not detain us. For after Bivens v. Six Unknown Named Agents of Federal Bureau of Narcotics,279 it is clear that the cause of action derives directly from the Constitution or governing statute, to the extent that the Supreme Court is willing to imply a cause of action from the particular provision.
Prior to Bivens,280 persons suing federal or state officers in the federal courts to restrain threatened violations of federal law were granted injunctive relief as a matter of course. In some of these cases there was no diversity of citizenship, from which it seemed to follow that federal law created the basic cause of action. If the federal or state officers were sued for damages arising from their conduct, it was generally assumed that the cause of action if any arose under state tort law and could not be litigated in federal district court in the absence of diversity of citizenship.281 The writer has attempted elsewhere to account for this anomaly.282
In Bivens, the Supreme Court confronted the issue for the first time. The claimant had sued federal narcotics agents for violations under the Fourth Amendment and had sought damages for this trespassory conduct. The suit had been instituted in a federal district court, and, there being no diversity of citizenship, it was understood that federal jurisdiction could be sustained only upon a showing that the claim arose under the Constitution, laws or treaties of the United States.283 The Court held that the right to damages could be implied directly from the Fourth Amendment itself.284
Earlier holdings in cases involving trespassory conduct can be rationalized in the same manner, insofar as they have rested on violations of the Constitution or other federal law. And this is true a fortiori in the non-trespassory cases, in which the existence of a common-law tort of any kind is usually dubious. Therefore, Bivens has been effectively incorporated into the Ex parte Young doctrine.
In Larson v. Domestic and Foreign Commerce Corp.,285 the Court modified the Ex parte Young doctrine by rejecting tort and substituting unauthorized conduct as the basis of the officer’s liability. Although this looks like a revolutionary modification of the doctrine, it is submitted that Larson has worked no substantial change. Larson is best understood after detailed analysis of a number of the cases decided under Ex parte Young.286
In the case of a statute not under attack for invalidity, courts routinely hold officers to performance of ministerial duties by mandamus,287 mandatory injunction,288 or negative injunction,289 as circumstances warrant. Such cases are legion. Sovereign immunity is not a bar to such relief because the suit is not deemed to be one against the government.290 As the Court said in a suit of this kind involving a federal statute, “[t]he suit rests upon the charge of abuse of power, and its merits must be determined accordingly; it is not a suit against the [*PG547]United States.”291 The officers being sued in such cases are named in their official capacities.292
Does this mean that in such a case on the federal level the Court will (1) order payment from the public treasury, or (2) order conveyance of title to government land, or (3) hold the government to performance of its contracts? The answer to the first two questions is clearly yes. Regarding the third question there is considerable confusion.
The Supreme Court has repeatedly sustained mandamus to require payment of funds from the public treasury,293 sometimes declaring that a suit to require an officer to perform a ministerial duty is not one against the government.294 When the pertinent statute imposes the duty of making the payment sought, the mandamus can fairly be regarded as in aid of the government rather than against it.295 The outcome is the same as if the officer had executed the duty in the first place.296
[*PG548] This explains why the Court has sustained the use of mandamus, not only to require payment of money from the treasury, but also to require transfer of title to government lands.297 Professor (as he then was) Scalia, in a study of litigation involving public lands, concluded that the cases constituted a special category, without an articulated general principle, in which the Court routinely addressed on the merits claims to public lands—except in a number of cases, that he thought aberrational, in which the Court sustained a plea of sovereign immunity.298 It is submitted that there is a unifying principle underlying the cases, namely, the principle that the judiciary holds officers to performance of their statutory duties. Few of Scalia’s sovereign immunity cases contravene this principle. Where, in result, some do, this stems from a summary declaration that overlooks the principle.299
[*PG549] Some precedent exists for holding officers to their statutory duties even though the effect is to compel the government to perform its obligation under a contract.300 In a later case there was a dictum to the effect that such relief would be denied if the contract was executory.301 In this dictum, the Court did not account for earlier cases where the contracts were in fact executory, and the officers were subjected to liability for violating their statutory duties in respect to the contracts.302 It is not immediately apparent why mandamus should be [*PG550]unavailable in such cases. Yet in Missouri v. Jenkins, the Court, in dictum, denounced employment of “the writ of mandamus as a ruse to avoid the Eleventh Amendment’s bar against exercising federal jurisdiction over the State.”303 For this proposition, the Court cited Louisiana v. Jumel,304 but in fact Jumel sanctioned use of mandamus (though not in the circumstances before it).305 The subject is mired in confusion.306
Of course, far more common than the statutory contracts heretofore discussed are contracts whose terms are set by officers having general statutory authority to do so. A statute creating this general authority is not likely to yield a construction that such contracts, made [*PG551]by the officer on behalf of the government, are judicially enforceable. In any event, relief is invariably denied in such cases.307
If the foregoing analysis is correct, some earlier decisions bear reconsideration. Thus, the denial of relief in Pennhurst State School and Hospital v. Halderman,308 with pernicious consequences for the administration of justice,309 rested on the Court’s assumption that a suit to compel state officers to comply with a concededly valid statute of their state was a suit against the state.310 Another such case is Hawaii v. Gordon.311
If the object of the suit is injunctive relief, the officer typically has no personal interest in the outcome and the government, though not named, is the only real party in interest.312 Even when the relief sought against the officer is a judgment in damages, frequently the only issue is the constitutionality of the statute under which the officer acted. Further, the officer is often indemnified by the government.313 Indeed, a decree that requires the officer to pay large amounts of money may be essentially pro forma in its application to the officer, the expectation of the court being that the money will be paid by the government. A conspicuous example is Milliken v. Bradley,314 where state officers315 were ordered to end unconstitutional school segregation and to pay $5,800,000316 “to wipe out continuing conditions of inequality produced by the inherently unequal dual school system [*PG552]long maintained by Detroit.”317 Indeed, the Court said in Milliken that “state officials [may be ordered] to conform their conduct to requirements of federal law, notwithstanding a direct and substantial impact on the state treasury.”318
Such impact is not diminished by calling it, as the Court sometimes does, “ancillary.”319 Particularly striking are the numerous cases where, on the basis of an assertion of personal liability on the part of the officer, the claimant recovers property in which ownership is asserted by the government.320 Indeed, in suits against state officers the Court has sometimes given a remedy tantamount to specific enforcement of a contract.321 How then is it to be determined whether the suit against the officer is or is not barred by sovereign immunity? The Court’s several attempts to answer this question have been notoriously unhelpful.
The first test traditionally employed to determine whether an Ex parte Young suit was against the sovereign focused on the effect a judgment would have on governmental operations. A typical formulation of this test was as follows:
[A] suit is against the sovereign if the judgment sought would . . . interfere with the public administration . . . or if the effect of the judgment would be to restrain the government from acting, or to compel it to act.322
Under this formulation, virtually every Ex parte Young suit would be one against the government, including, conspicuously, Young itself.323
In recent years another test has emerged—the prospective-retrospective test. It is now invoked whenever the question arises whether the suit against the officer should be deemed one against the state.324 This test originated in Edelman v. Jordan.325 In that case Illinois officers administering a joint federal-state welfare program had followed state regulations that were incompatible with the corresponding federal regulations, with resulting underpayment to the beneficiaries.326 Invoking Ex parte Young, the claimants sued the officers to compel future compliance with the federal requirements, and to recover past benefits wrongfully withheld.327 Both demands were granted by the lower courts. Review was sought in the Supreme Court only with respect to the decree ordering payment of the arrears. In this regard, the decree, though directed only against the officers, was held to operate essentially against the state.328 Thus, the Court said:
[T]he relief awarded in Ex parte Young was prospective only; the Attorney General of Minnesota was enjoined to conform his future conduct of that office to the requirement of the Fourteenth Amendment . . . . The funds to satisfy the award in this case must inevitably come from the general revenues of the State of Illinois, and thus the award resembles far more closely [a] monetary award against the State itself . . . than it does the prospective injunctive relief awarded in Ex parte Young. . . . [This award] is in practical effect indistinguishable in many aspects from an award of damages against the State.329
The Court recognized that compliance by the officers with the “prospective” features of the decree would also take state money, but said [*PG554]that “[s]uch an ancillary effect on the state treasury is a permissible and often an inevitable consequence of the principle announced in Ex parte Young.”330 Apparently, the Court thought that the distinguishing feature of the arrears ordered to be paid was that they were “compensatory.”331 The result in Edelman was defensible.332
Furthermore, discussion of the issue as turning on the difference between prospective and retrospective relief was understandable in the circumstances of the case. But introduction of this mode of analysis was unnecessary, and has proved troublesome. This problem is illustrated by the several opinions in Idaho v. Coeur d’Alene Tribe.333 In this case, the Tribe sued Idaho officers for interfering with their asserted rights in regard to certain submerged lands.334 Five of the Justices voted to deny relief, describing the requested remedy as “the functional equivalent of a quiet title action.”335 The Tribe apparently conceded that this was a fair characterization of its claim.336 In her concurring opinion, Justice O’Connor stated as follows:
[T]he Tribe seeks to eliminate altogether the State’s regulatory power over the submerged lands at issue—to establish not only that the State has no right to possess the property, but also that the property is not within Idaho’s sovereign jurisdiction at all.337
In effect, Idaho was asserting its sovereign immunity to bar inquiry into whether the property it claimed to own was not state property at all, but rather, federally-owned property in which the state had no regulatory competence. There was no need to discuss the retroactive-prospective distinction, yet it was discussed in the three opinions. Yet, relief was denied even though it was to operate prospectively. None of [*PG555]the majority Justices contended otherwise. In short, the case was one in which the retroactive-prospective distinction was ignored, presumably because it would have produced the wrong answer.
After Edelman, the Court declared that the essence of prospective relief is that it is addressed to a “continuing violation.”338 Consider in this connection the decision in Milliken v. Bradley.339 That case involved the Detroit school system, which had long been segregated de jure as a result of state and local action. This was no longer the situation, and the issue in the Supreme Court concerned only that aspect of the district court’s decree that ordered institution of remedial programs, these being thought necessary by reason of past practices.340 As the Supreme Court said, the conditions to be remedied resulted from the “unequal dual school system”341 formerly maintained by Detroit. However, the Court stated that the decree was “wholly prospective”342 because the decree was designed to eliminate “vestiges of state-imposed segregation” and thus “to wipe out continuing conditions of inequality.”343
Milliken should not and indeed cannot be confined to desegregation cases. Relief should be available generally not only for a continuing violation, but also for continuing consequences of a past violation. It would be an absurdity in contravention of settled law to refrain from granting such relief. Thus, in Osborne v. Bank of the United States, the Court held that the Bank of the United States was not barred from recovering specie unlawfully seized in the past by the state of Ohio.344 That has been the universal rule governing past seizures.345 The Court has never deviated from the principle that the continuing consequences of a past violation may be the subject of a recovery.346
[*PG557][*PG556] But it should be apparent that absent a continuing violation, or the continuing consequences of a past violation, there is no entitlement to relief of any kind. Thus the prospectivity fork of the test is no more than the statement of a truism. As for the retrospectivity fork, this loses much of its utility when it is considered that, as in Milliken, recovery may be had for past misconduct. The question that remains is when relief should be denied. What seems to emerge from the decisions is that the forbidden remedy is one that is “compensatory.”347 But this formulation is too loose. For one thing, it is broad enough to encompass possessory relief against officers for unlawful seizures, which is granted as a matter of course.348 Further, bearing in mind that the suit is not one against the government but against the officer, there should be hesitation in abolishing outright the officer’s historic personal liability.349 If the foregoing analysis is valid, the prospective–retrospective test is unhelpful, to say the least. It is noteworthy that all the cases discussed in this subsection could have been decided as they were without mention of the prospective-retrospective test, which, indeed, was deemed irrelevant by all the Justices in Coeur d’Alene Tribe.350
It is submitted that the problem of a suitable test is solved if we look, not to the rationalizations attempted by the Court, but rather to their actual holdings. Indeed, the Court’s holdings in this area form a pattern. In a suit against an officer, a plea of sovereign immunity is [*PG558]disallowed when the immunity would operate offensively, but not when it would operate defensively. If the claimant is seeking only to be left alone and charges that past or prospective conduct of government officers is unlawfully intrusive, judicial inquiry into the validity of such conduct would be precluded if a plea of sovereign immunity is sustained.351 This would constitute what is here called offensive use of the immunity, and such use is disallowed. On the other hand, when the claimant is seeking some affirmative advantage from the government, like payment of outstanding indebtedness, a plea of sovereign immunity is sustained, in what is here called defensive use of the immunity.352 The two lines of cases are not distinct if we look to the language the Court uses, but they are distinct in the result, with relatively few exceptions. We now turn to an analysis of the cases.
Poindexter v. Greenhow involved an issue of state bonds backed by a guarantee that the interest coupons would be accepted in payment of taxes.353 This guarantee was repudiated by subsequent legislation. Accordingly, a tax officer had rejected a proffer of interest coupons, and had seized a desk belonging to the claimant, for nonpayment of taxes.354 In an action in detinue against the officer, the Court sustained recovery of the desk. The Court held that since repudiation of the State’s obligation violated the Constitution, the tax collector was a wrongdoer who had been “stripped of his official character,” and li[*PG559]able as such for seizing the claimant’s property.355 The obvious effect was to hold the State to its contract; for if tax coupons were proffered in payment of taxes, the taxes were not collectible by other means. It was understandable, therefore, that four dissenters contended that the suit was “virtually” one against the State for “specific performance.”356 As a matter of fact, the Poindexter majority made it explicit that it was the contract that was being enforced. Thus, one of the arguments advanced by the State was that the case presented no more than a question of remedy, since the State contended that under state law a taxpayer claiming a tax to be invalid could pay under protest and then sue to recover the amount paid.357 The Supreme Court did not challenge the reasonableness of such a remedy.358 But, said the Court, the State under its “contract” had “bound herself that it shall be otherwise,”359 and the contractual obligation was enforced.
The Court also emphasized that the taxpayer, though nominally a plaintiff, was essentially a “defendant, passively resting on his rights,”360 and said in the same vein: “[The taxpayer’s] object is merely to resist an attempted wrong and to restore the status in quo as it was when the right to be vindicated was invaded. In this respect, it is upon the same footing with the preventive remedy of injunction in equity.”361 This position was elaborated soon afterwards in McGahey v. Virginia,362 where the Court declared that one tendering tax-receivable coupons in the circumstances of Poindexter is:
entitled to be free from molestation in person or goods . . . and may vindicate such right in all lawful modes of redress—by suit to recover his property, by suit against the officer to recover damages for taking it, by injunction to prevent such taking . . . or by a defense to a suit brought against him.363
[*PG560]In McGahey, the Court sustained a defense under this principle. In other cases, the Court has approved injunctive relief against state officers seeking to collect taxes in violation of the contract.364
As the foregoing cases show, sovereign immunity, as such, does not render contracts with the state unenforceable, and does not exempt such contracts from the operation of the Contracts Clause; the doctrine does no more than give the state an immunity from suit without its consent.365 The Court’s use of the term “defensive” to describe the taxpayer’s posture in Poindexter was another way of saying that the redress sought was essentially to be left alone.366 The holding in Poindexter and like cases shows that sovereign immunity may not be offensively interposed to defeat such redress.
There is another class of cases in which government contracts are in effect enforced even in the absence of unconstitutional statutes. In these cases, the claimant has a vested property interest, the property having been obtained from the government or from a private source. A government officer, having statutory authority, enters into a contract relating to such property. Thereafter, the officer (or another officer) seizes the property, or threatens to do so, contending that the contract has been breached and that it provides for such seizure in event of breach. When the claimant sues the officer for an injunction against the seizure, or for return of the property if it has been taken, the defense of sovereign immunity is denied. The court makes an independent determination whether there has been a breach or whether the contract authorizes a seizure. In this general category, a crucial issue is whether a seizure of property is lawful.367
There are several contract cases where sovereign immunity has been sustained. In Hagood v. Southern,368 the State had repudiated a statutory obligation to accept certain scrip in payment of taxes. The [*PG561]Court held that state officers could not be judicially compelled to receive the scrip, on the ground that this would be tantamount to coerced “performance of the alleged contract by the state.”369 But the only damage claimed by the plaintiffs, most of whom had not proffered the scrip in payment of taxes, was that their holdings of scrip had been rendered worthless by the State’s repudiation.
In In re Ayers,370 there was a similar statutory obligation involving, not use of scrip, but use of interest coupons attached to state bonds. A subsequent statute had directed state officers to recover taxes from persons who had used tax-receivable coupons in payment.371 A suit to enjoin the officers from enforcing this statute failed as essentially an attempt “to compel the specific performance of the contract.”372 Under the Poindexter line of cases, persons who had actually paid their taxes with the coupons would presumably have been entitled to injunctive relief.373 But the claimants in In re Ayers were British owners of coupons who had purchased them with a view to resale to Virginia property owners; the claimants did not contend that they themselves had paid taxes with such coupons.374 Both in Hagood and In re Ayers, the claimants were seeking only to protect their investments; their position was comparable to that of owners of state bonds suing for payment.
Wells v. Roper375 was a case in which the plaintiff had a contract with the Post Office to supply vehicles for use in the District of Columbia. The plaintiff asserted that, after adoption of a statute enabling the Post Office to purchase its own vehicles, the Post Office had repudiated the contract. The plaintiff sought an injunction against certain named postal officers to restrain them from proceeding in violation of the contract. The Court held that the injunction was properly denied since its “effect . . . would have been to oblige the United States to accept continued performance of plaintiff’s contract.”376 To be sure, the plaintiff had invested funds in reliance on the [*PG562]government’s performance; but, again, the purchaser of a government bond is in the same position.377
Thus, in the first group of cases the Court in effect enforced state contracts, and in the second group the Court denied enforcement. Critics have castigated the Court for such disparate treatment, for which, they have said, there was no principled basis.378 But what distinguishes the cases is that in the first group sovereign immunity, if allowed, would have operated offensively, while in the second group its operation was defensive only. In the first group claim was made of seizures or threatened seizures, and sovereign immunity would have barred judicial consideration of whether the officers had acted with legal justification. In the second group the plea of sovereign immunity spoiled an investment entered into voluntarily in anticipation of the state’s performance of its contracts.379
A similar pattern is revealed in suits against officers for recovery of property said by the officers to belong to the government. Thus, in United States v. Lee,380 the Court upheld an ejectment judgment against Army officers in charge of property that was used in part as a military cemetery and in part as a fort.381 The United States had purchased the property at a tax sale that was defective by reason of the government’s own misconduct.382 In the ejectment action, the government [*PG563]conceded that the plaintiff had legal title as heir of the original owner but resisted the ejectment claim as an unconsented suit against the United States.383 Disallowing this defense, the Court declared in effect that occupation of the land by the officers was tortious.384 The Court likened the case to one where “the citizen has been deprived of his property by force, his estate seized and converted to the use of the government without lawful authority, without process of law, and without compensation.”385 The Court declared that the United States, not being a party, was not bound by the judgment,386 and the claimant’s relief was possessory only. But it is unrealistic to infer from this that the claimant was getting only half a loaf. In view of the uses to which the land was being put, the government was subject to an overriding necessity of quickly purchasing the property or acquiring it by condemnation.387
In Davis v. Gray, the Court sustained a decree prohibiting the Governor of Texas from alienating land to which the State had title.388 The claim was founded upon a contract the State had made with a railroad company whereby certain lands were to be transferred to the company as it made progress with its construction program. The State countered that the requirements of the contractual timetable had not been met. But construction had been frustrated by the onset of the Civil War and the Court held that the State itself had made performance within the time limits impossible.389 In a later case, the Court spoke of Davis as “rest[ing] on the same principle it would if patents had been actually issued to the company, and the State, through its officers, was attempting to place a cloud on the title by granting subsequent patents to others.”390 Davis also illustrates the point that a court will enforce a contract against the state when necessary to afford redress against trespassory conduct.
Similarly, in Pennoyer v. McConnaughy,391 the Court sustained an injunction against state officers to restrain them from selling lands in which the claimant asserted ownership under a contract with the [*PG564]State. A statute that in effect terminated the rights of the claimant to the land was held to violate the Contracts Clause.392 Cases other than those previously mentioned in which the Court, rejecting pleas of sovereign immunity, upheld recovery of specific property, have involved not only land,393 but also stock certificates,394 barges,395 a shipyard,396 specie,397 and (as earlier seen) a desk.398
On the other hand, an interesting and suggestive case where the defensive use of sovereign immunity was upheld is Malone v. Bowdoin.399 In Malone, an ejectment action against a federal officer was successfully resisted. The plaintiff asserted title to land under an 1857 will of the then-owner of the land. Specifically, the plaintiff claimed that under the will a life estate in the land was left to one Martha A. Sanders, with the remainder to her children, and that in 1873 she had “devised” the land “in fee” to another.400 The United States acquired title in 1936.401 The Court observed that there had been no allegation that government officers had acted in violation of the Constitution or of any federal statute.402 Relief was denied by reason of the absence of any showing that the officers had acted without authority.403 The Court emphasized that there was no claim of a taking.404 The claimant had been wronged, it seemed, not by the government, but by Mrs. Sanders who, having only a life estate, had disposed of the land “in fee” to strangers. To the claimant’s assertion of superior title, the government invoked sovereign immunity defensively.405
[*PG565] In sum, the property cases resemble the contract cases. Sovereign immunity is allowed when employed defensively but not when employed offensively. To the extent that the dispute over property turns on the contract from which the right derives, the law pertaining to contracts is followed, and the contract is enforced when appropriate. The courts will not award the claimant title, but will award possessory or monetary relief against the officer if there has been a taking.
Larson v. Domestic & Foreign Commerce Corp. is notable for advancing a new formulation of the law governing specific relief against government officers.406 The dissenters in the case decried this formulation and viewed the result in the case as a regressive step, expanding the scope of sovereign immunity.407 Commentators have also taken this view.408 It is submitted, however, that Larson has made no meaningful change in the law.
The case involved a contract for the purchase of coal from the War Assets Administration, a federal agency.409 The Administrator, contending that the terms of the contract had not been fulfilled, was arranging to sell the coal to others.410 The buyer, on the other hand, claimed title to the coal, and maintained that the Administrator’s action constituted a conversion.411 The relief sought was an injunction [*PG566]restraining sale or delivery of the coal to other persons.412 The Administration’s defense of sovereign immunity was sustained.413
The Court’s new formulation of the law was this: When specific relief is sought against government officers, the crucial issue is not whether they acted tortiously but whether they acted in the exercise of valid authority. When official action is validly authorized, such “action is the sovereign’s,”414 and if a claimant’s demand is that such action “be prevented or compelled,” then “the demand . . . must fail as a demand against the sovereign.”415 Officers, however, remain liable for their torts whether the commission of these torts is authorized or not.416 The holding was that the Administrator’s action fell within the range of his authority and that the suit was therefore one against the United States.417
In dissent, Justice Frankfurter observed that in previous cases upholding specific relief, the predicate for such relief was tortious conduct on the part of the officers. He argued that the claimant, having charged a conversion, should therefore not be barred by a plea of sovereign immunity.418 The Court, however, declared in effect that every previous case involving “specific relief” in connection with property “held or injured” by governmental officers had involved a “taking.”419 It seems to the writer that they did indeed involve takings, consummated or threatened.420 It is notable that Justice Frankfurter, [*PG567]despite the seeming thoroughness of his dissenting opinion, made no attempt to show otherwise.
The nature of the holding in Larson is illuminated if we consider the Court’s treatment of United States ex rel. Goldberg v. Daniels,421 and Goltra v. Weeks.422 In Daniels, the plaintiff had contended that he had submitted the highest bid for a surplus naval vessel; that ownership of the vessel had vested in him when the bids were opened; and that the Secretary of the Navy had refused to deliver the vessel.423 The relief sought was mandamus against the Secretary to compel delivery.424 The Court held that such relief should be denied, on the ground that even if title had passed and the Secretary was acting tortiously, the suit was essentially against the United States.425 In this case, it should be noted, the plaintiff was not asking to be left alone. The contract was executory on the government’s part, and what the plaintiff was demanding was specific performance. The government was using sovereign immunity as a shield. The Larson Court saw Daniels as on all fours with the case before it.
But consider now a case in which the government sells and delivers a chattel, and government officers subsequently repossess it or threaten to do so, claiming a right of repossession by the terms of the contract. If the buyer resists or attempts to prevent repossession, the buyer is acting defensively. The government may not justify its trespassory conduct under the contract and at the same time interpose sovereign immunity to bar judicial inquiry into the adequacy of the justification. That would be attempted use of sovereign immunity not as a shield but as a sword; and that, under the precedents, the government may not do.
The facts just posited correspond to those alleged in Goltra upon which Justice Frankfurter relied in his dissent in Larson. The govern[*PG568]ment had delivered possession of a number of barges to the plaintiff, under a lease whose validity was not in question.426 Thereafter, complaining of non-compliance with the lease terms, government officers had seized some of the barges and were threatening to seize others.427 The plaintiff contended that there had been compliance and sought return of the barges seized and an injunction against seizure of the others. The Court held that, inasmuch as the officers were charged with trespassory conduct, sovereign immunity was not a bar to the action.428
Justice Frankfurter argued that Goltra was being overruled in Larson.429 But there was no overruling. The Larson majority made clear that its only quarrel was with “the theory of the Goltra opinion,”430 not with the result in that case. The Court added:
Whether the actual decision in the Goltra Case, on the basis of the facts there presented, was correct or not is not relevant to the disposition of the present case, and we express no opinion on that question. Goltra, unlike Goldberg, does not present a parallel to the facts in the case at bar. The action complained of there was a seizure with a strong hand which was claimed to be unconstitutional, as an arbitrary taking of property without due process of law. . . . There is no such claim in the present case.431
Thus, pace Justice Frankfurter, Larson did not repudiate Goltra;432 nor did it “overrule” United States v. Lee “and the cases which have applied it.”433 There is simply no basis for reading Larson as changing the law in this regard.
On the other hand, the Court was unhelpful in stating that the exclusive basis for specific relief against government officers is lack of valid authority. In this regard, it is instructive to compare Larson with [*PG569]Goltra. In Larson, there was no challenge to the constitutionality of the statute under which the Administrator acted, and there was no challenge to his authority. As the Court observed, the Administrator had power to make contracts, to determine whether their terms were met, and to act accordingly.434 But in these regards, the situation in Goltra was identical. There was no challenge to the underlying statute or to the authority of the officers to lease the barges in question and protect the government’s interest in accordance with the terms of the lease. Absent a holding that the governing statute was unconstitutional, the decision has turned on the interpretation of a contract or statute without challenge to the authority of the officers to act as they did if their interpretations were proper.435
The appeal of the rationalization founded on lack of valid authority is that it bolsters the view that the suit is only against the officer and not in any respect one against the government. Lack of valid authority can plausibly be found when the officer acted under an unconstitutional statute, or when the claim of authority under a valid statute is based on what the court finds to be an unfounded reading of that statute. But it cannot be said, without a good deal of strain, that the officer’s lack of authority stems from the unfounded reading of a contract.436 If the officer is subject to liability on that account, it is not for action without authority, but for action without legal justification. The same rationalization—lack of legal justification—could also be employed in all cases where the courts now talk of lack of authority. But there is no need to employ a single rationalization in the entire area.
As has been seen, in cases where officers are enjoined from unlawfully withholding statutory benefits from the intended beneficiaries, their conduct is not trespassory in the common law [*PG570]sense.437 However, their conduct can be analogized to a taking insofar as it prevents the benefits from reaching the intended beneficiaries. Cases that do not lend themselves to rationalization along this line are those in which the officers are ordered to desist from denying the claimant the equal protection of the laws, as in the desegregation decisions.438 Here, the claimant is asking to be left alone in the sense of not being singled out invidiously for denial of rights enjoyed by the general public. In both classes of cases, recognition of a claim of sovereign immunity would mean denial of opportunity to show the illegal character of the official action. In short, it would constitute the use of sovereign immunity offensively, and this is not permitted.439
In Pennhurst State School & Hospital v. Halderman,440 the confusion regarding the Ex parte Young doctrine reached a nadir. The Court expressed doubt as to whether the doctrine, in its most important applications, was compatible with the principle of sovereign immunity, which it called a “constitutional limitation” on the authority of the federal judiciary.441 The suit was instituted in a federal district court against Pennhurst—a state institution for the mentally retarded—and certain of its officers,442 based on allegations of mistreatment of the patients.443 While relief was sought on federal and state grounds, it was granted in the lower courts solely on the state ground.444 Reversing, the Supreme Court declared that the “fiction” of Ex parte Young “rests on the need to promote the vindication of federal rights.”445 It follows, said the Court, that “Young . . . [is] inapplicable in a suit against state officials on the basis of state law.”446 This statement was true, but the Court erred regarding its implications.
First, it may be noted that the Court doubted the validity of the Ex parte Young doctrine even in its application to rights claimed under federal law.447 Calling the doctrine “a narrow and questionable exception” to sovereign immunity,448 the Court had no trouble with it insofar as it was used to impose on the officer personal liability in damages. But the Court was apparently dubious about any relief against the officer that as a practical matter placed a financial burden on the [*PG572]government.449 Observing that injunctive relief against the officer might impose such a burden,450 the Court said:
In this light, it may well be wondered what principled basis there is to the [Young] doctrine as it was set forth in Larson. . . . For present purposes, however, we do no more than question the continued vitality of the . . . doctrine in the Eleventh Amendment context.451
The Court relied in part on the Larson decision, in the belief that Larson represented an earlier restriction on the scope of the Ex parte Young doctrine. Thus the Court said that the numerous cases involving takings, upon which Justice Frankfurter had relied in his dissent in Larson, were now “moribund.”452 As has been seen, however, the Larson majority had disapproved of those cases only insofar as they rested on a theory of tort.
As to where this leaves us, consideration must be given to some of the Pennhurst Court’s additional remarks. The Court said: “[A]n injunction based on federal law stands on very different footing, particularly in light of the Civil War Amendments. . . . [I]n such cases this Court is vested with the constitutional duty to vindicate ‘the supreme authority of the United States.’”453 Since the Court had earlier said that a decree against a state officer was in effect one against the state if it could not be satisfied without funding by the state, it may be that the Court’s conception of a proper injunction was one that cast no financial burden on the government, as was seemingly the case in Ex parte Young itself. But the Court also suggested that if, upon remand, consideration was given to the Fourteenth Amendment claims of the Pennhurst plaintiffs, account should be taken of the then recent decision in Youngberg v. Romeo.454 This was a case in which the Court declared a state to be under a duty to meet certain standards in the institutional care of mentally retarded persons.455 But this is precisely the kind of case in which additional state funding may be necessary.
[*PG573] The most plausible view of Pennhurst is that the Court did not contemplate an actual change in the law applicable to federal claims. It opposed what it considered to be any extension of Ex parte Young and believed that federal enforcement of claims founded on state law would be such an extension. Its views of Ex parte Young, however, if taken up by the Court, would severely limit the application of that doctrine to claims founded on federal law.456
This subject is discussed in the earlier section on breach of duty.457
Reich v. Collins is notable for a dictum to the effect that a state may not interpose sovereign immunity to bar a suit in its courts for recovery of illegally-exacted taxes.458 Some preliminary comments may help place Reich in perspective. When sovereign immunity is not in issue, it is established doctrine that state courts, when called upon to implement federal rights, must provide a remedy that is adequate by federal standards.459 The typical case has involved state taxation. When state taxes are attacked on federal grounds, the Supreme Court has held that the states must provide either an adequate pre-deprivation remedy or an adequate post-deprivation remedy.460 The states typically prefer the latter, which amounts to consent to suit.
[*PG574] In Reich, the state had provided a postdeprivation remedy in the state courts for illegally-exacted taxes, and the Court concluded that plaintiff’s reliance on this remedy was justified.461 However, relief had been denied in the state courts on the ground that the plaintiff should have resorted to a pre-deprivation remedy. The Supreme Court held this to be error, because the plaintiff, under then state law, was not on adequate notice that the pre-deprivation remedy was the exclusive one.462 That was all the Court needed to say.
But the Court remarked, summarily, that there was an obligation to grant “recovery of taxes [illegally] exacted . . . the sovereign immunity States traditionally enjoy in their own courts notwithstanding.”463 The remark was doubly gratuitous, in that (1) the issue of sovereign immunity had not been raised and (2) the state had waived its immunity in allowing itself to be sued. The Supreme Court’s only holding was that a limitation on the remedy that prevented recovery in the particular case was violative of due process; its requirement that any remedies provided must be constitutionally adequate had of course been amply set forth in earlier cases.464 The Court cautioned that a state waiver allowing suit in a state court could not be the basis for overriding a defense of sovereign immunity in a federal district court; this too accorded with settled doctrine.465 But in Reich the Court was reviewing the judgment of a state court, in a suit consented to by the state. There was no need to discuss sovereign immunity at all.
However, Reich’s dictum regarding sovereign immunity may now be controlling law by virtue of the decision now to be discussed. In First English, the Court ruled in effect that the constitutional requirement of compensation for takings cannot be avoided by the defense of sovereign immunity.466 Of course, the immunity is not a problem in formal condemnation proceedings, which are instituted by the gov[*PG575]ernment, not against it. First English is pertinent in cases of inverse condemnation, where the owner of property, alleging intrusion amounting to a taking, sues the government for compensation, typically under a consent-to-suit statute.467 But the decision has broader ramifications as well.
First English arose as an inverse condemnation proceeding in a California state court, as permitted by statute.468 The Supreme Court understood California law to deny compensation for temporary takings in a particular class of such cases.469 That, apparently, was the basis for denial of redress below. The validity of such denial was the only issue on review.
Sovereign immunity was injected into the case by the United States, which in an amicus brief took the position that the Takings Clause, standing alone, was no ground for invalidating the California law.470 The argument was that the Clause operates only as a limitation on government power: a taking without compensation could be prevented, or, if consummated, could be invalidated, but the compensation feature of the Clause did not operate of its own force.471 The argument was based in part on the language of the Clause. But it was also based in part on the contention that, as a practical matter, the Clause could have no other meaning, inasmuch as sovereign immunity barred an unconsented suit for compensation.472 A number of decisions were cited that supported this contention.473
This view of the Takings Clause was rejected in First English. The Court did not challenge the cases cited in the amicus brief. Instead, it relied on cases that contained language indicating that the compensation requirement was an integral aspect of the Takings Clause.474 Sovereign immunity, however, was not mentioned, and was not remotely in issue, in any of those cases. Most of them involved suits against the [*PG576]United States under statutes consenting to suit.475 Others involved formal condemnation proceedings instituted by the United States.476 Another involved construction of a federal statute.477 The cases relied on by the Court were simply unresponsive to the argument made in the amicus brief. But the Court insisted that these cases “refute the argument of the United States that ‘the Constitution does not, of its own force, furnish a basis for a court to award money damages against the government.’”478
The approach that the Court took to sovereign immunity was an oblique one. To repeat, the immunity was not directly in issue because the suit was consented to and because the suit was against local subdivisions not enjoying the immunity in any event. The United States interjected the immunity issue into the case in support of its argument that the Takings Clause was not self-executing. The Court’s response was that the Clause itself is a command for compensation. But that should not have been deemed conclusive. After all, the Court has repeatedly held that the assertion of a constitutional right is not itself sufficient basis for overriding sovereign immunity.479 Still, it should always be appropriate to inquire whether the Constitution itself, in one of its particulars, requires that the federal and state governments be amenable to suit whether they consent or not. The analysis in Principality of Monaco v. Mississippi480 is illustrative. There the Court said that the question should be considered in light of “the plan of the Convention”; and observed that, under the “plan,” a state may be sued without its consent when another state or the United States makes a claim against it under the Supreme Court’s original jurisdiction, and also when sued by the United States.481 The tenor of the opinion in First English suggests that the Court, faced with the precise [*PG577]question, might well have deemed the Takings Clause to be a provision contemplating suit irrespective of state consent. Again, however, the case was one in which the state had in fact consented.
First English is of particular interest for what it may portend. As the Court said in Logan v. Zimmerman Brush Co., a leading case under the Takings Clause, “[t]he hallmark of property is an individual entitlement grounded in . . . law, which cannot be removed except ‘for cause.’”482 Seemingly, any purposeful impairment of the value of property can amount to a taking; the property interest need not be one in land.483
With the immunity of the government to unconsented suit hitherto taken as axiomatic, the only remedy available has been the Ex parte Young suit against the officer. For a variety of reasons that remedy may be inadequate. Property repossessed may have been damaged, or there may have been substantial pecuniary harm from deprivation of its use. In theory, the officer remains liable for such losses, especially for conduct that can be classified as tortious; but the officer may be protected by a privilege, or may lack sufficient resources to satisfy a judgment. The implication of First English is that the government itself must make good such losses, and indeed that it may be sued in the first instance.
It is of course doubtful that the Court had in mind these ramifications of its position in First English. Against such a sweeping change in the law, it can be argued that First English, like Reich before it, involved suits against consenting states, and that these two cases should be taken as controlling only in that context. But, having said time and again, in regard to challenged taxes, that the state must afford an adequate pre-deprivation remedy or an adequate post-deprivation remedy, is the Court likely to say that this obligation applies only if the state consents to be sued? And having held in First English that the Constitution requires compensation to be made for a taking, is the Court likely to say that compensation can be avoided by a plea of sovereign immunity?
[*PG578] An affirmative answer to both questions is possible. The constitutional basis for the doctrine of sovereign immunity has been sustained for two centuries. The anomalies that would be presented by confining First English and Reich to their own facts would be more theoretical than real. After all, consent-to-suit statutes are now relatively universal. Thus, so far as the writer is aware, no state shuts the doors of its courts to a suit for inverse condemnation. The Court could take the position that Congress and the state legislatures, in the exercise of their exclusive constitutional roles in regard to waivers of sovereign immunity, have reflected popular sentiment on holding governments financially responsible for their wrongs; that it is not to be supposed that this process has been completed; and that there is no pressing reason for the courts to usurp the role of the legislatures in this regard.
Yet another possibility remains. The Court could take the position that the near-universality of consent-to-suit statutes is warrant for abrogating the sovereign immunity of the nation and the states. This would present problems already discussed.484
It is submitted that the problem of enforcing a money judgment against the government, if it comes to that, should not be one of practical concern. In general, the Court does not allow uncertainties relative to enforceability of judgments to interfere with its disposition of controversies—as shown when it ordered President Richard Nixon to surrender the Watergate tapes,485and when it in effect ordered the House of Representatives to reinstate Representative Adam Clayton Powell.486 It may be added that collecting money in a suit directly against the government should not encounter greater obstacles than achieving the identical result indirectly in a suit against the officer.
Indeed, the Court has routinely authorized judgments against states as such, including judgments for the payment of money, when its original jurisdiction is invoked in controversies between states.487 Further, in the Ex parte Young line of cases, the Court has recognized, as has been shown, that a judgment casting a heavy financial burden on an officer will in many cases almost certainly be borne by the gov[*PG579]ernment.488 In all these cases, the expectation, justified in the result, is that, at least eventually, the political branches of the government will implement the decisions of the judicial branch.
In Smith v. Reeves, the Court held that a state’s consent to be sued in its own courts cannot be deemed a consent to be sued in a federal trial court.489 The Court added that this was “subject always to the condition . . . that the final judgment of the highest court of the State in any action brought against it with its consent may be reviewed . . . as prescribed by the Act of Congress, if it denies to the plaintiff any [federal] right.”490 Professor Jackson interprets the quoted passage from Smith v. Reeves as making the state’s consent to suit in its own courts suffice as consent to suit in the Supreme Court.491 She therefore sees a discrepancy in the effect accorded to state consent on the appellate and trial levels of the federal courts.492
It is submitted that there is a sound constitutional basis for the discrepancy, if that is what it is. Apart from the Eleventh Amendment, it should be clear that the constitutional scheme contemplates that the Supreme Court, as ultimate arbiter of federal law, can review state decisions of federal questions. Chief Justice Marshall made this point strongly in Cohens v. Virginia.493 The question is whether the Amendment makes a difference.
The Amendment covers the Supreme Court, since it is a limitation on the “judicial power of the United States.”494 In Cohens v. Virginia the Court held that the Amendment does not apply when the [*PG580]state is plaintiff.495 The apparent implication was that the Amendment would bar review in the case where an individual sues the state. However, as the Court declared more recently, “it is inherent in the constitutional plan that when a state takes cognizance of a case, the state assents to appellate review of federal issues raised by the case.”496 It could hardly be otherwise, for absent such review, the states would have the last word on question of federal law. As for the Eleventh Amendment, assuming it to be other than declaratory of the original understanding, it is waivable,497 as of course is the immunity inherent in the states under the original understanding. Considering the Amendment in light of the constitutional plan, it can strongly be argued that there has been a waiver, for purposes of Supreme Court review, when the state “takes cognizance” of a case presenting a federal question.
On the other hand, such an argument is unavailable to support federal district court jurisdiction on the basis of a state’s consent to suit in its own courts. Imputing consent to the Supreme Court’s jurisdiction on that basis rests in the final analysis on the constitutional role of that Court as having the final say on federal law. The federal district courts have no such role.
In Pennhurst, the Court rejected a claim of pendent jurisdiction.498 Such rejection was based on the bewildering assumption that the “principles established in our Eleventh Amendment decisions” would not apply in cases adjudicated on the basis of pendent jurisdiction.499 The Court remarked that the Eleventh Amendment barred even federally-based suits aimed at securing “damages against the state [*PG581]treasury,”500 or “brought directly against a state.”501 It added that the “Amendment should not be construed to apply with less force”502 to suits founded on state law, through invocation of pendent jurisdiction. Otherwise, said the Court, “a federal court could award damages against a state on the basis of a pendent claim.”503 No one argued for such a view of pendent jurisdiction, and there was no basis for it. Since the claim was founded on state law, and in the absence of any questions of federal substantive law, state law, whatever it might be, would have governed in all respects.
More fundamentally, the Court was also in error in assuming that the suit was against the state. To repeat, it was one against state officers to compel them to perform their duties under a concededly valid statute. For reasons developed earlier in this Article, such a suit is not one against the state but rather one in aid of the state.504 Such a case is governed entirely by state law; the Eleventh Amendment and the Ex parte Young doctrine are irrelevant.505
Giving stare decisis effect to this holding would be most unfortunate. Litigants claiming, as they commonly do, that the conduct of state officers violates both federal and state law, are confronted with two unpalatable choices: (1) to bring the federal claims in a federal court and the state claims in a state court, with the resulting inconvenience, expense, and uncertainty;506 or (2) to bring both claims in a state court, with only the remotest prospect of federal consideration of the federal claim, considering the unlikelihood of Supreme Court review. It is difficult to perceive any public reliance on the pendent jurisdiction ruling that would mitigate in favor of stare decisis.
The case of Nevada v. Hall involved an accident in California resulting from the negligent operation of a Nevada-owned vehicle on official Nevada business.507 An injured person sued the State of Nevada in a California state court, with service based on California’s long-arm statute.508 Nevada’s plea of sovereign immunity was rejected.
The Supreme Court held that California was under no constitutional obligation to respect Nevada’s claim of immunity.509 The Court explained that, on the international level, the question whether one sovereign is bound to respect the immunity of another is governed entirely by considerations of comity.510 The Court explained that when the Constitution was being drafted and ratified, the states, “heavily indebted”511 as they were, “presumably” assumed that they had “adequate protection” against suit in the courts of sister states by virtue of “prevailing notions of comity,” but that they neglected to write this into the Constitution, which accordingly gave them no protection on this point.512 A dissenting opinion argued that there were arguments for implying such protection from the Constitution that were more persuasive than the Court’s “literalism.”513
It is submitted that the Court was beguiled by “notions of comity,” and inattentive to the implications of our constitutional arrangements. On the international level, if nation A has suffered injury in consequence of action by nation B that constitutes a violation of international law, nation A’s opportunity for redress is sharply limited. An international tribunal could not exercise jurisdiction over the case absent consent by nation B, unless nation B has in effect given a general consent under the terms of a treaty. Absent nation B’s consent on either basis, nation A might resort to retaliation, or even to war if it deemed itself sufficiently aggrieved.
[*PG583] Avoidance of such confrontations between the states was of course a major reason for creation of the Supreme Court.514 The Supreme Court is a tribunal to which one state can summon another without the latter’s consent. Absent any indication in the Constitution of the law to be applied in these controversies between states, the Supreme Court applies rules that it deems appropriate to the occasion, drawing heavily on rules of customary international law.515 Clearly, Nevada, if wronged, would have had a sound basis for redress against California in an original proceeding in the Supreme Court.
To return to the situation in Hall, suppose that Nevada had argued in the California courts for an outcome the same as that which could be achieved if Nevada sued California in the Supreme Court. Our jurisprudence would be wasteful and formalistic if Nevada could have this claim recognized only by such a suit in the Supreme Court. In fact, there is precedent indicating that the California courts were bound to apply the law that the Supreme Court would apply if adjudicating the controversy itself; that law is binding on all courts as federal common law.516 It is not suggested that one state should be allowed to [*PG584]sue another state in a federal district court, but only that, in a suit to which a state is a party or in which it has a substantial interest, account should be taken of the law that the Supreme Court would apply in a comparable case between the two interested states.
However, it does not follow that the wrong result was reached in Nevada v. Hall. Since Nevada had, in effect, entered California and affected that state’s interests adversely, Nevada should not be able to avoid such consequences by a plea of sovereign immunity. To a substantial extent the original jurisdiction of the Supreme Court is exercised in suits by one state against another for relief from harmful consequences caused by the second state within the borders of the first, as in cases of pollution of water and air, or impairment of water rights.517 The Supreme Court touched upon this point in an ambiguous footnote in Nevada v. Hall.518 The problem is that the thrust of the opinion would allow one state to override the sovereign immunity of another even when the latter’s conduct has not produced harmful consequences within the borders of the first.
It has been argued in this Article that there was an understanding at the time of adoption of the Constitution that the nation and the states were not suable without their consent, except in special particulars like a suit by a state against a sister state. Modern critics and a minority of the Justices have challenged the existence of such an understanding. They have maintained that the historical record shows that the proponents of sovereign immunity were concerned with protecting the states only from claims founded on state law, as distinct from federal law. This is implausible, and incompatible with the historical record. The same fallacy has engendered confusion about the meaning of the Eleventh Amendment. Read in light of the original understanding the Amendment loses its mystery. The original understanding also illuminates the problem of congressional power to set aside the sovereign immunity of the states.
[*PG585] The Supreme Court’s seeming zigzagging in decisions under the Ex parte Young doctrine assumes a rational and defensible pattern when it is recognized that the Court has permitted the use of sovereign immunity for defensive but not offensive purposes. In these decisions, the Court has gone far in recognition of constitutional and statutory claims. The result can be seen as a defensible accommodation of the requirements of sovereign immunity to the requirements of the rule of law.
Consent-to-suit statutes aside, the relief that can be afforded through the suit against the officer is imperfect. Property may be restored, but damages assessed against an officer may not be collectible. Hardship may be incurred by reason of inability to compel a transfer of title by the government. It has been suggested, however, that, as the law is tending, governmental interference with a property right may in all cases be tantamount to a taking, with relief allowable on that basis. Contracts with the government are not enforceable when sovereign immunity is asserted defensively, although this rule may not be applicable where the government’s obligation is spelled out in a statute which also specifies clearly the duties of the officer in regard to the obligation. If an inability to hold the government to its executory contracts, as in the case of defaulted bonds, presents a rule-of-law problem, it is one a good deal less serious than government intrusion on persons who ask only to be left alone, or to be free from governmental denial of rights enjoyed by others. The government’s ability to escape liability in tort, at least when the tort consists of negligence, as is usually the case, arguably presents even less of a rule-of-law problem.
These are the principal costs of sovereign immunity, absent consent-to-suit statutes. But these statutes are now so widespread, on the federal and state levels, that these costs have in greater part been eliminated. Paradoxically, the prevalence of such statutes offers a possible basis for a principled judicial modification or elimination of sovereign immunity, at least as applied to governmental commercial activity, if there is merit in a thesis advanced by this writer eleven years ago.519