Summary
vacating district court's award of back pay because the district court acted outside of its jurisdiction under § 1346 where the claim exceeded $10,000
Summary of this case from Warren v. U.S.Opinion
No. 85-1514.
Submitted January 15, 1986.
Decided June 27, 1986.
Constance Wynn, Washington, D.C., for appellants.
James L. Sloan, Little Rock, Ark., for appellee.
Appeal from the United States District Court for the Eastern District of Arkansas.
Before ARNOLD, JOHN R. GIBSON and WOLLMAN, Circuit Judges.
The primary issue in this case is whether a district court, presented with a claim for less than $10,000 accrued back pay against the United States, has jurisdiction pursuant to the Tucker Act, 28 U.S.C. § 1346(a)(2), 1491 (1982), to ultimately award back pay in an amount greater than that statutory maximum. The District Court for the Eastern District of Arkansas made such an award in reinstating Wallace B. Shaw to his position as U.S. property and fiscal officer for Arkansas pending his proper removal by the U.S. Army in accordance with due process and the Army's own procedural regulations, 604 F. Supp. 880. The Army argues that the back pay award of more than $10,000 was exclusively within the jurisdiction of the Claims Court. In addition the Army argues that the district court erred in granting even equitable relief because Shaw's removal from his position was an exercise of military discretion not subject to judicial review and because due process and Army regulations were not implicated since Shaw was by law removable and was in fact removed at the will of the Arkansas Governor.
The Honorable Garnett Thomas Eisele, Chief Judge, United States District Court for the Eastern District of Arkansas.
Before reaching any of the Army's contentions, however, we must consider an overriding question regarding our appellate jurisdiction. See McGowne v. Challenge-Cook Bros., 672 F.2d 652, 658 (8th Cir. 1982). Pursuant to the Federal Courts Improvement Act of 1982, see 28 U.S.C. § 1295(a)(2) (1982), the Court of Appeals for the Federal Circuit now has exclusive jurisdiction over most final decisions of district courts based "in whole or in part" on section 1346(a)(2). Thus, if we agree with Shaw that the district court had jurisdiction of his back pay claim under section 1346(a)(2), we must transfer this appeal to the Federal Circuit. E.g., Van Drasek v. Lehman, 762 F.2d 1065 (D.C. Cir. 1985); Kidde, Inc. v. E.F. Bavis Associates, 735 F.2d 1085 (8th Cir. 1984) (per curiam) (patent case under section 1295(a)(1)); see 28 U.S.C. § 1631 (1982). If we agree with the Army that the district court acted beyond its power in awarding back pay, that court's decision could not have been based "in whole or in part" on section 1346(a)(2), and we have jurisdiction to review questions properly before the district court regarding significant prospective equitable relief. Doe v. United States Department of Justice, 753 F.2d 1092, 1101-02 (D.C. Cir. 1985).
Determination whether this appeal properly lies here or in the Federal Circuit thus requires that we go beyond the plaintiff's or district court's characterization of the source of district court jurisdiction and analyze the underlying bases for the district court's assertion of or refusal to exercise its power. E.g., Van Drasek, supra; see also Wronke v. Marsh, 767 F.2d 354, 355 (7th Cir. 1985) (per curiam) (appellate jurisdiction to be based on the "real rather than ostensible source of the district court's jurisdiction"); cf. Maier v. Orr, 754 F.2d 973, 982 (Fed. Cir. 1985) (Federal Circuit had the power to determine that district court jurisdiction actually rested on section 1346(a)(2), even though the complaint premised jurisdiction on mandamus ( 28 U.S.C. § 1361 (1982)) and on its face would have been appealable only to the regional circuit). Contra Oliveira v. United States, 734 F.2d 760, 762 n. 6 (11th Cir. 1984) (per curiam). This practice seems in keeping with the emphasis in the legislative history of the Federal Courts Improvement Act that there be no "presumption" in favor of jurisdiction and that jurisdiction be affirmatively shown. S.Rep. No. 275, 97th Cong., 2d Sess. 18, reprinted in 1982 U.S. Code Cong. Ad. News 11, 28. We vacate the district court's award of back pay as beyond its power but affirm the order of reinstatement.
The Federal Circuit, relying on additional portions of the legislative history regarding the need to limit forum shopping and achieve decisional uniformity in specified areas of the law, once stated that section 1295 vests it with the exclusive power to determine all issues of appellate subject matter jurisdiction arising thereunder. C.R. Bard, Inc. v. Schwartz, 716 F.2d 874, 877-78 (Fed. Cir. 1983). Uniformity achieved in that manner, however, would require that a regional circuit refer to the Federal Circuit all issues of section 1295 jurisdiction even when addressing an appeal on its face within regional jurisdiction or when raising jurisdiction on its own motion. We decline to lightly read into the language of section 1295 a requirement which would thus subordinate a regional circuit to the Federal Circuit with regard to such traditional and inherent functions as determining its own jurisdiction and supervising the exercise of jurisdiction by the district courts below it.
I.
The underlying jurisdictional question in this case arises from Shaw's diligence in attempting to prevent his removal in June 1982 as property and fiscal officer. Having received advance notice of the impending employment action, he filed suit four days prior to its effective date seeking an injunction. Preliminary relief was denied on the ground that reinstatement would be an adequate remedy were Shaw eventually to prevail, and Shaw on the day his removal was to be effective filed an amended complaint requesting reinstatement and back pay. After receiving briefs on the Army's initial motion to dismiss, the district court on March 30, 1984, ruled that Shaw's removal was not shielded from review as an exercise of military discretion and that Shaw had been entitled to the pretermination process established by the relevant military regulations. Shaw v. Gwatney, 584 F. Supp. 1357 (E.D.Ark. 1984) ( Shaw I). The Army subsequently filed a second motion to dismiss, alleging that the district court lacked jurisdiction to award relief because Shaw's back pay claim had by that time grown to approximately $50,000. The court, however, found its jurisdiction proper under section 1346(a)(2), observing that Shaw had not yet been entitled to back pay when he filed his initial complaint and invoking the traditional rule that "jurisdiction is determined by reference to the time of the institution of the action." Shaw v. Gwatney, 604 F. Supp. 880, 884 (E.D.Ark. 1985) ( Shaw II) (citing 1 J. Moore, J. Lucas, H. Fink, D. Weckstein J. Wicker, Moore's Federal Practice ¶ 0.91[3] (2d ed. 1986)); see 14A C. Wright, A. Miller E. Cooper, Federal Practice and Procedure § 3702, at 28 (2d ed. 1985).
While we do not dispute the rule relied on by the district court or its practical advantages (or even necessity), see Shaw II, 604 F. Supp. at 884, it is not properly applied here. The district court in looking to Shaw's initial complaint ignored the essential change wrought in the nature of the action when the monetary claim against the United States was added. Because of the doctrine of sovereign immunity, such a claim requires an additional jurisdictional basis beyond that which supports a claim for injunctive relief. See generally Van Drasek v. Lehman, 762 F.2d 1065, 1068-69 (D.C. Cir. 1985); 14 C. Wright, A. Miller E. Cooper, Federal Practice and Procedure §§ 3654-3655, 3657 (2d ed. 1985). Thus, the district court's invocation of the time of filing rule would be analogous to holding that amendment of a diversity complaint to add a nondiverse party would not destroy federal jurisdiction. To the contrary, it has clearly been held that when the amendment changes the action in such an essential fashion, jurisdiction is determined as of the date of the amendment. Carlton v. Baww, Inc., 751 F.2d 781, 785 (5th Cir. 1985); Lewis v. Odell, 503 F.2d 445, 446 (2d Cir. 1974). A rule determining jurisdiction as of the time of filing despite a significant subsequent amendment would actually encourage the very manipulation of jurisdiction the rule was designed to minimize because a plaintiff, by initially asserting only part of his action, would then be able to gain ancillary or pendent jurisdiction over claims as to which independent jurisdiction otherwise would have been required. Sovereign immunity cannot be avoided in this manner.
Shaw further argues, however, that even if jurisdiction is measured from the time of his amendment, his claim was properly before the district court pursuant to section 1346(a)(2) because the amount of back pay then accrued was less than the $10,000 statutory maximum. This contention misapplies the time of filing jurisdictional rule by assuming that the amount of the claim asserted is equal to the amount of damages accrued. To the contrary, the amount of the claim seems to depend more on the amount the plaintiff is entitled to recover through the single suit. For example, in installment contract cases, if, under state law, judgment may be entered only for the amount of the installments due at the commencement of the suit, only those installments may be considered in determining whether diversity jurisdiction, given the $10,000 minimum, is proper. If, however, state law allows recovery of future installments also in a single action, those installments too may be considered in determining the amount in controversy. Aetna Casualty Surety Co. v. Flowers, 330 U.S. 464, 467-68, 67 S.Ct. 798, 800, 91 L.Ed. 1024 (1947); see 14A C. Wright, A. Miller E. Cooper, supra, § 3710, at 171-72. Similarly, the Supreme Court in Weinberger v. Wiesenfeld, 420 U.S. 636, 642 n. 10, 95 S.Ct. 1225, 1230 n. 10, 43 L.Ed.2d 514 (1975), stated in a footnote that future social security benefits to which a plaintiff would be entitled upon a favorable judgment were to be considered in determining whether the $10,000 minimum then applicable to federal question jurisdiction had been met. The Fourth Circuit has held diversity jurisdiction proper when a buyer's expenses in caring for a lame horse pending rescission of the purchase would accrue to more than $10,000 before the court could rule on a motion to dismiss, even though damages were less than $10,000 when the complaint was filed. Broglie v. MacKay-Smith, 541 F.2d 453 (4th Cir. 1976).
The parties cite only one case in which the back pay accrued at the time of filing was less than $10,000, Giordano v. Roudebush, 617 F.2d 511 (8th Cir. 1980), and the plaintiff in that instance did not challenge the district court's decision to vacate an ultimate award of $70,000 and transfer the back pay question to the court of claims.
In light of these principles we believe that the amount of a claim against the United States for back pay in the contemplation of section 1346(a)(2) is not the amount of back pay accrued at the time of filing but the total amount of back pay the plaintiff stands ultimately to recover in the suit. Cf. Graham v. Henegar, 640 F.2d 732 (5th Cir. 1981) (a plaintiff's "civil action or claim * * * not exceeding $10,000" under section 1346(a)(2) in certain circumstances must include attorney fees to which the plaintiff might be entitled). Such a rule would be consistent both with the pronouncements of the Supreme Court that "limitations and conditions upon which the Government consents to be sued must be strictly observed and exceptions thereto are not to be implied," Soriano v. United States, 352 U.S. 270, 276, 77 S.Ct. 269, 273, 1 L.Ed.2d 306 (1957); see Palmer v. General Services Administration, 787 F.2d 300, 301 (8th Cir. 1986); 14 C. Wright, A. Miller E. Cooper, Federal Practice and Procedure § 3654, at 194-99 (1985), and with the legislative intent in granting concurrent district court jurisdiction over claims of less than $10,000 that small claimants be relieved of the expense of traveling to Washington, D.C., to litigate while large claims remained centralized at the seat of the government so that department heads would be better able to protect the government's interests. Sutcliffe Storage Warehouse Co. v. United States, 162 F.2d 849, 852 (1st Cir. 1947); Oliver v. United States, 149 F.2d 727, 728-29 (9th Cir. 1945).
This position is consistent with the result in Goble v. Marsh, 684 F.2d 12 (D.C. Cir. 1982), and we reject suggestions that the holding of that case does not extend to the situation here. Goble involved an attempt by plaintiffs to remain within district court jurisdiction under the Tucker Act by waiving back pay already accrued in excess of $10,000 but continuing to claim back pay that would again accrue in excess of that amount. Shaw argues that the Goble court specifically disclaimed deciding whether jurisdiction would exist when no prefiling back pay had to be waived and points to the court's statement in footnote 4 that "[w]e need not reach the question whether the District Court may under any circumstances award a monetary judgment in excess of $10,000; our holding is limited to the waiver situation presented in these cases." Id. at 15 n. 4. The court in footnote 5, however, rejects the very "amount accrued at time of filing" argument relied on by Shaw, emphasizing that "a waiver of prefiling claims over $10,000 does not reduce the 'claim' below $10,000, for the reasons discussed in [the] text." Id. at 16 n. 5. The relevant textual passage holds that
[b]y its very nature a back pay claim continues to accrue between filing of the complaint and entry of final judgment. Therefore, if a plaintiff waives only pre-filing claims in excess of $10,000, it is certain that if he prevails additional accruals in the intervening period will result in an aggregate claim of more than $10,000. In this situation the partial waiver does not reduce the back pay claim to a "civil action or claim against the United States, not exceeding $10,000 in amount."
Id. at 15-16 (citation omitted). Furthermore, the court in Goble in defining the issue characterized the district court opinion as having held the plaintiffs' attempted waivers inadequate "because they did not waive any sums exceeding $10,000 that might accrue between the date of filing the complaint and the date of judgment," id. at 13; and the court identified the difference between the parties' positions on appeal as lying "in the treatment of back pay claims that accrue between filing and judgment." Id. at 15. We believe that footnote 4 is a caution by the Goble court only that its opinion was not to be read to have addressed instances where the ultimate award might exceed $10,000 for reasons other than the foreseeable — and desired — accrual of the damages inherent in the nature of the relief sought.
Furthermore, this method of determining whether a plaintiff's monetary claim exceeds $10,000 is consistent with — and as certain and easy of application as — the practice in other cases involving jurisdictional amounts, because the plaintiff's good faith estimate at the time of filing regarding his ultimate entitlement will be controlling. Hahn v. United States, 757 F.2d 581, 587 (3d Cir. 1985); see St. Paul Mercury Indemnity Co. v. Red Cab. Co., 303 U.S. 283, 288, 58 S.Ct. 586, 590, 82 L.Ed. 845 (1938). The strictness with which conditions accompanying the waiver of sovereign immunity are to be construed, however, requires a plaintiff in essence to guarantee the good faith of his estimate by waiving back pay in excess of $10,000. Cf. Zumerling v. Devine, 769 F.2d 745, 748 (Fed. Cir. 1985) (district court jurisdiction proper over claim arguably involving more than $10,000 when plaintiff waives recovery in excess of that amount); Goble v. Marsh, 684 F.2d 12, 15 (D.C. Cir. 1982) (same); 14 C. Wright, A. Miller E. Cooper, supra, § 3657, at 286-87 n. 32 (same, citing cases). This method of determining the amount claimed raises no specter of a district court "losing jurisdiction" as back pay accrues because, with back pay fairly estimated, the court never will have had jurisdiction. There is thus no danger of the government delaying a trial while back pay accrues so that it may then seek to transfer a case to the Claims Court. Also, the waiver requirement should prevent most preliminary litigation over the true amount of back pay to which the plaintiff might be entitled. If this leaves a plaintiff litigating in two forums, it is not because we measure jurisdiction in terms of ultimate rather than accrued back pay but because of the plaintiff's choice to pursue the nonmonetary claims in district court rather than filing the entire suit in Claims Court and because of Congress' choice to centralize adjudication of monetary claims against the United States with no clearly expressed intent to disturb the traditional powers of district courts over nonmonetary claims of constitutional dimension. See Hahn, 757 F.2d at 590.
Applying this rule to the present case, we believe that Shaw's monetary claim at the time of filing was for an amount in excess of $10,000. Shaw's amended complaint included an allegation that the matter in controversy exceeded $10,000. He argues that that allegation was superfluous for federal question jurisdiction and included the value also of the injunctive relief sought; however, he still bore the burden of affirmatively showing jurisdiction in his pleadings, 14A C. Wright, A. Miller E. Cooper, supra, § 3702, at 16, which means that he had to allege that his claim was for less than $10,000. The pay and allowances for the position from which Shaw was removed totaled about $4,600 a month, making it likely that were he to prevail, he would be entitled to more than the $10,000 statutory maximum, see Doe v. United States Department of Justice, 753 F.2d 1092, 1101 (D.C. Cir. 1985), and he did not waive any excess. Furthermore, Shaw's jurisdictional argument to this court is inconsistent with any claim that he ultimately seeks less than $10,000 in back pay.
Since we conclude that the portion of Shaw's claim seeking monetary damages was within the exclusive jurisdiction of the Claims Court, the district court's jurisdiction was not based even in part upon section 1346(a)(2), and we have the power to dispose of this appeal. We vacate the back pay award and remand that portion of Shaw's action to the district court for a waiver of excess damages or transfer to the Claims Court. See Hahn v. United States, 757 F.2d 581, 587-88 (3d Cir. 1985).
II.
The Army, while expressing disagreement that the district court, absent jurisdiction over the monetary claim, should retain jurisdiction over and decide Shaw's claim for reinstatement, concedes that under Giordano v. Roudebush, 617 F.2d 511 (8th Cir. 1980), the law of this circuit allows such bifurcation. The Army instead argues that the removal of Shaw was not subject to judicial review pursuant to the rule of Mindes v. Seaman, 453 F.2d 197 (5th Cir. 1971), requiring deference to military decisions regarding internal military affairs. Under Mindes, adopted by this circuit in Nieszner v. Mark, 684 F.2d 562 (8th Cir. 1982), cert. denied, 460 U.S. 1022, 103 S.Ct. 1273, 75 L.Ed.2d 494 (1983), a court may consider a claim against the military only if the claim alleges a violation of the Constitution or of a statute or military regulation, the claimant has exhausted any available intraservice remedies, and the balance of four factors reflecting the policy reasons for deferring to the military is in the claimant's favor. Mindes, 453 F.2d at 201; Nieszner, 684 F.2d at 563. The Army challenges only the district court's conclusion upon balancing these four factors — the nature and strength of Shaw's claim, the injury to Shaw if review is refused, the type and degree of interference with the military if review is granted, and the extent to which military expertise or discretion is involved. See Mindes, 453 F.2d at 201-02; Nieszner, 684 F.2d at 563-64.
Consideration of the nature and strength of the claim, we have emphasized, does not involve an evaluation of the merits of a plaintiff's position or of the likely outcome of the suit; instead, this factor weighs against a claimant only if the claim is "tenuous," for example, in that it relies on a legal proposition that has previously been rejected. Nieszner, 684 F.2d at 564-65. Shaw's claim is that his removal was defective because he was not accorded the process required by the Army's own regulations; and his interpretation of those regulations, as demonstrated by the district court's discussion, Shaw I, 584 F. Supp. at 1363-64, is far from being patently implausible or tenuous. There is no issue regarding the wisdom of the Army's choice to remove Shaw or even regarding the appropriateness of the military regulations. If the regulations were, as Shaw argues, applicable, the Army was bound to follow them, with no military discretion or expertise involved. Determination of individual rights under military regulations is within the role envisioned for courts in Mindes. Rucker v. Secretary of the Army, 702 F.2d 966, 971 (11th Cir. 1983); see also Hodges v. Callaway, 499 F.2d 417, 419 n. 2 (5th Cir. 1974) (noting that it is "well-settled" that federal courts may review claims that the military failed to comply with its own regulations); Denton v. Secretary of the Air Force, 483 F.2d 21, 24-25 (9th Cir. 1973) (reviewing plaintiff's allegation that his discharge was not accompanied by the fair hearing guaranteed by regulation), cert. denied, 414 U.S. 1146, 94 S.Ct. 900, 39 L.Ed.2d 102 (1974). Shaw's claim, in fact, is in nature identical to the claim the Eleventh Circuit in Rucker found "peculiarly appropriate for judicial review." 702 F.2d at 971 (plaintiff charged with fraudulent entry into service was separated from the military without the medical evaluation, assistance of counsel, or opportunity to submit statements on his own behalf allegedly provided for by regulation).
The Army's argument that the Mindes factors point toward nonreviewability is flawed by its assumption of its own position on the merits. For example, the Army asserts that Shaw's claim is tenuous because Shaw was removed by the Governor of Arkansas rather than by the Army, with the result that the military regulations did not apply and no process was due. Even accepting as fact that Shaw was removed by the Governor, the district court could still properly review Shaw's argument that the Governor thus acted beyond his authority. See Harmon v. Brucker, 355 U.S. 579, 582, 78 S.Ct. 433, 435, 2 L.Ed.2d 503 (1958) (per curiam). Judicial review cannot be rejected as interfering with the Governor's ability to replace quickly a property and fiscal officer in whom he has lost confidence or as interfering with the Governor's discretion in selecting such officer until it has been determined that the Governor does indeed have such authority. As discussed above, Shaw asks us to review not the Army's or the Governor's good sense, intelligence, or fairness in removing him but only the Army's or the Governor's power to so act and the procedures thereby required. The district court properly declined to defer to the military on those issues.
III.
Finally, turning to the merits, the Army seems to concede that if Shaw's removal is deemed to have been effectuated by the Army, he did not receive the process which, under the regulations, he was due. The Army instead argues that the Governor exercised an independent power of removal that he possessed concomitant to a statutory power of appointment. See 32 U.S.C. § 708(a) (1982). The district court rejected this argument, as a matter of law, on the ground that the Governor did not in fact possess a true power of appointment or that, if he did, the unique circumstances surrounding the property and fiscal officer position required that the traditional rule of concomitant removal power not be applied. Shaw I, 584 F. Supp. at 1365-69.
The parties at oral argument, though not in their briefs, engaged in some discussion of whether Shaw was deprived of a liberty interest. Since we agree with the district court that Shaw is entitled to relief based on the Army's violation of its own regulations, we need not consider the liberty issue not reached below.
The statutory language governing selection of property and fiscal officers somewhat ambiguously provides that a state's governor "shall appoint, designate or detail, subject to the approval of the Secretary of the Army and the Secretary of the Air Force, a qualified commissioned officer of the National Guard * * * to be the property and fiscal officer of that jurisdiction." 32 U.S.C. § 708(a). The Army argues that, despite the approval clause, this language confers full appointment authority on governors because Congress repeatedly used the term "appointment" in the legislative history and rejected a version of the legislation that provided for appointments to be made by the military secretaries with the approval of the state governors. Finally, the Army asserts that, because it is the agency charged with enforcing section 708(a), we should accord substantial deference to its interpretation of the statutory language.
Starting with this last argument first, we do not believe the present case presents an appropriate occasion for deferral. The line of precedent relied on by the Army is distinguishable because section 708(a) does not embody substantive provisions for the protection of the public welfare that the Army can "administer" in the sense that the National Labor Relations Board oversees collective bargaining between unions and employers, see Ford Motor Co. v. National Labor Relations Board, 441 U.S. 488, 495, 99 S.Ct. 1842, 1848, 60 L.Ed.2d 420 (1979), or the Food and Drug Administration oversees the development of new drugs. United States v. Rutherford, 442 U.S. 544, 553-54, 99 S.Ct. 2470, 2475-76, 61 L.Ed.2d 68 (1979). It is not evident — and in fact seems unlikely — that Congress "assigned to the [Army] the primary task of construing" section 708(a). See Ford Motor Co., 441 U.S. at 495, 99 S.Ct. at 1848. Resolution of the question of the proper relationship of state and federal governments in the exercise of authority over the National Guard does not involve factors peculiarly within military expertise; the issue, rather, is an important one of federalism which is in fact addressed in the Constitution. U.S. Const., art. I. § 8, cls. 15-16.
Were we giving more deference to the Army's interpretation of section 708(a), we would be forced to observe that, while the Army has now rewritten its regulations to reflect its position in this case, at the time Shaw was removed both its regulations and its correspondence with the Arkansas Governor seemingly contemplated only that the Governor had the power to nominate a property and fiscal officer. E.g., National Guard Regulation 130-6, ¶ 2-3 (1981).
The Army's arguments based on legislative history are similarly unconvincing. The House Report on which the Army relies deals not with the selection of property and fiscal officers but with an amendment to section 708(a) giving such officers active duty status in the U.S. Army. H.R. Rep. No. 1879, 83d Cong., 2d Sess., reprinted in 1954 U.S. Code Cong. Ad. News 2504. The references to "appointment," because of their casual context and infrequency (the whole report is only three pages long), simply will not bear the weight the Army attempts to place on them. "Appointment" is never used without the qualification of the Army's power of approval, and the House Report in other references in fact speaks of governors "initiating" appointments. Id., reprinted at 2505. The choice of whether the Governor was to "appoint" and the Army was to "approve," or vice versa, is characterized in the report as a choice "of technique rather than substance"; by either method the individuals selected, in keeping with the legislative concern, would be acceptable to both their states and to the Army. Id. This court has previously held that state governors "merely designate or detail the officers of the National Guard for the consideration of the Secretary of War as proposed property and disbursing officers of the United States," Woodford v. United States, 77 F.2d 861, 864 (8th Cir. 1935), and neither the differing context nor the subsequent minor changes in statutory language are sufficient to cause us to vary our interpretation of this appointment provision.
Even if this role of governors in selecting property and fiscal officers were considered in some sense a power of appointment, we could not find a concomitant power of removal. The above-mentioned House Report emphasizes that after property and fiscal officers are ordered to active duty, they are under the "direct control of the President." H.R. Rep. No. 1879, supra, reprinted at 2506. Also, as noted above, under section 708 property and fiscal officers are entitled to active duty status in the U.S. Army, and Shaw was ordered to duty by and assigned to the National Guard Bureau of the Departments of the Army and Air Force, Washington, D.C. Thus, it must be considered that a National Guard member appointed as his state's property and fiscal officer has been called into the service of the United States. The exercise by a governor of a power to remove an officer so detailed would interfere with the exclusive authority of the federal government over the militia when called into federal service. See Houston v. Moore, 18 U.S. (5 Wheat.) 1, 17, 5 L.Ed. 1 (1820); U.S. Const., art. 1, § 8, cl. 16.
The Army's suggestion at oral argument that Shaw was a state officer hardly seems consistent with those portions of its own regulations which provide, for example, that a property and fiscal officer may not hold a state position or be assigned any state National Guard duties and that a property and fiscal officer is qualified to serve as a federal, but not a state, contracting officer. National Guard Regulation No. 130-6, ¶¶ 1-4(b)(2), 3-1(f)(1) (1981). Furthermore, if a property and fiscal officer were a state officer, the federal government's assertion of the authority to approve such appointments, 32 U.S.C. § 708(a), and to prescribe extensive qualifications for the position, National Guard Regulation No. 130-6, ¶ 2.2 (1981), would be constitutionally suspect as infringing state power over the appointment of militia officers. U.S. Const., art. 1, § 8, cl. 16.
We conclude that only the Army, and not the Governor of Arkansas, had the power to remove Shaw as property and fiscal officer and thus that Shaw was entitled to the procedural protections of the Army regulations. We affirm the district court's order of reinstatement and vacate the award of back pay for disposition in accordance with this opinion.