Summary
holding that we consider the delay until the defendant's “final revocation hearing” without indicating whether the defendant's sentencing occurred at that hearing or at a subsequent court date
Summary of this case from United States v. Morales-IsabarrasOpinion
No. 00-50303.
Argued and Submitted January 11, 2001.
Filed June 11, 2001.
Todd W. Burns, Federal Defenders of San Diego, Inc., San Diego, California, for the defendant-appellant.
Christopher P. Tenorio, United States Attorney Office, San Diego, California, for the plaintiff-appellee.
Mark Rehe, United States Attorney Office, San Diego, California, for the plaintiff-appellee.
Appeal from the United States District Court for the Southern District of California; Barry T. Moskowitz, District Judge, Presiding. D.C. No. CR-97-01115-BTM
Before: TROTT, THOMAS, and BERZON, Circuit Judges.
Appellant Thomas James Garrett, Jr., appeals the revocation of his supervised release by the United States District Court for the Southern District of California ("the district court"), and the district court's imposition of an additional nine months incarceration, followed by two more years of supervised release. Garrett challenges the district court's jurisdiction to revoke his supervised release on the ground that there was no reasonable necessity for the approximately ten-month delay between the expiration of his supervised release term and the district court's revocation of his supervised release. We have jurisdiction pursuant to 28 U.S.C. § 1291, and AFFIRM the district court.
DISCUSSION
A. Background
On May 2, 1997, Garrett pled guilty to bringing an undocumented alien into the United States in violation of 8 U.S.C. § 1324(a)(2)(B)(iii). On June 23, 1997, Garrett was sentenced to time served and two years of supervised release, which was scheduled to end on June 22, 1999. One condition of Garrett's supervised release was that he notify his probation officer of any arrest.
On June 15, 1999, Garrett was arrested and placed into California state custody for burglary, forgery, and making a false financial statement. Garrett failed to notify his probation officer of this arrest. Subsequently, on June 21, 1999 — one day before the expiration of Garrett's supervised release term — the district court issued a supervised release violation warrant ("violation warrant") for Garrett's arrest and signed an order to show cause why Garrett's supervised release should not be revoked.
On June 28, 1999, Garrett pled guilty to grand theft and burglary in state court and was sentenced to sixteen months in state custody. Immediately upon Garrett's release from state custody in late March, 2000, the outstanding federal warrant for Garrett's arrest was executed, and he was taken into federal custody.
Garrett appeared before Federal Magistrate Judge Porter on April 7, 2000, and denied the alleged violations of his supervised release. On April 17, 2000, Garrett appeared before Federal District Judge Moskowitz for his revocation hearing and argued that the district court lacked jurisdiction to revoke his term of supervised release because it had expired nearly nine months earlier on June 22, 1999.
Before ruling on Garrett's jurisdictional challenge, the district court ordered briefing from both parties and held two additional hearings. The first hearing was held on May 8, 2000, and the second on May 11, 2000. At the latter hearing, the district court rejected Garrett's claim, concluding that it did have jurisdiction to revoke Garrett's supervised release. The court held that the government's decision to postpone the adjudication of Garrett's supervised release violations until after Garrett was released from state custody comported with 18 U.S.C. § 3583(i), which extends the power of a federal court to revoke a defendant's term of supervised release after its expiration for "any period reasonably necessary for the adjudication of matters arising before its expiration." 18 U.S.C. § 3583(i). The district court sentenced Garrett to nine months custody, to be followed by two additional years of supervised release. This appeal followed.
B. The District Court Did Not Err In Revoking Garrett's Supervised Release
1. Standard of Review
"Jurisdiction is a question of law subject to de novo review." United States v. Neville, 985 F.2d 992, 994 (9th Cir. 1993).
2. Garrett's Term of Supervised Release Expired On June 22, 1999.
The threshold issue in this dispute is whether Garrett's term of supervised release expired as scheduled on June 22, 1999, or was tolled, as argued by the government, because of his earlier June 15, 1999 arrest and consequent incarceration on state charges.
A defendant's term of supervised release is tolled "during any period in which the person is imprisoned in connection with a conviction for a Federal, State, or local crime. . . ." 18 U.S.C. § 3624(e). However, we have held that "pretrial detention does not constitute an `imprisonment' within the meaning of § 3624(e) and thus does not operate to toll a term of supervised release." United States v. Morales-Alejo, 193 F.3d 1102, 1106 (9th Cir. 1999). Here, Garrett did not plead guilty to the state charges until June 28, 1999, and was not sentenced on those charges until September 8, 1999. Therefore, Garrett's confinement from June 15, 1999 — the time of his arrest — to June 22, 1999 — the final day of his scheduled term of supervised release — constituted pretrial detention that was insufficient to toll the running of his supervised release. Accordingly, Garrett's term of supervised release expired on June 22, 1999.
3. The government's decision to postpone the execution of Garrett's violation warrant until after he was released from state custody did not violate 18 U.S.C. § 3583(i).
Title 18 U.S.C. § 3583(i) empowers a district court, under certain circumstances, to revoke a term of supervised release after that term has expired. Specifically, 18 U.S.C. § 3583(i) provides:
(i) Delayed revocation. — The power of the court to revoke a term of supervised release for violation of a condition of supervised release, and to order the defendant to serve a term of imprisonment and, subject to the limitations in subsection (h), a further term of supervised release, extends beyond the expiration of the term of supervised release for any period reasonably necessary for the adjudication of matters arising before its expiration if, before its expiration, a warrant or summons has been issued on the basis of an allegation of such a violation.
18 U.S.C. § 3583(i) (emphasis added). Thus, a district court can revoke a term of supervised release after that term has expired if: (1) a violation warrant or summons was issued before the term expired; and (2) the delay between the end of the term of supervised release and the district court's revocation order is "reasonably necessary for the adjudication of matters arising before [the term's] expiration." Id. Both elements were satisfied here.
The first requirement was satisfied by the district court's issuance of a violation warrant for Garrett on June 21, 1999, which predated the expiration of Garrett's term of supervised release by one day. It is the second requirement — that the delay be "reasonably necessary" — that is in dispute here. Garrett contends that the government was not justified in awaiting his release from state custody to execute the warrant, and that he was prejudiced by the delay because he was deprived of any chance to serve his federal and state sentences concurrently. We disagree.
Ultimately, this case boils down to whether awaiting a defendant's release from state custody to execute a violation warrant is "reasonably necessary for the adjudication" of matters arising prior to the expiration of the defendant's supervised release term. As explained below, the difficulty in resolving this issue stems from the fact that while the delay complained of by Garrett has repeatedly been deemed "reasonable," see, e.g., United States v. Bartholdi, 453 F.2d 1225, 1226 (9th Cir. 1972), it is not technically "necessary" given the federal government's power to execute a writ of habeas corpus ad prosequendum; thus leaving the question of whether such a delay is "reasonably necessary."
In support of its argument that the delay in this case was permissible under § 3583(i), the government relies on the Supreme Court's holding in Moody v. Daggett, 429 U.S. 78, 97 S.Ct. 274, 50 L.Ed.2d 236 (1976). The Supreme Court in Moody addressed the constitutionality of the government's ability to postpone the execution of a parole violation warrant until a defendant is released from state custody. The defendant in Moody shot and killed two individuals while on parole for a previous federal offense. Moody, 429 U.S. at 80, 97 S.Ct. 274. During the defendant's incarceration for the homicides, the United States Board of Parole issued, but did not execute, a violation warrant. Id. The warrant was lodged with prison officials as a detainer. Id. The defendant requested that the Board execute the warrant so that any imprisonment he received for his parole violation could run concurrently with his homicide sentences. Id. at 80-81, 97 S.Ct. 274. The Board declined the defendant's request and refused to execute the warrant until he finished his homicide sentences. Id. The defendant then sought "dismissal of the violation warrant on the ground that he had been denied a prompt hearing at which the pending parole revocation issues could be aired." Id.
The Court in Moody rejected the defendant's claim, ruling that the ten-year delay between the issuance and execution of the parole violation warrant was constitutionally acceptable because a parolee is not constitutionally entitled to a revocation hearing immediately upon the issuance of such a warrant. Id. at 86, 97 S.Ct. 274. A revocation hearing, the Court reasoned, need only be tendered promptly after the violation warrant is executed, because a parolee does not suffer a loss of liberty as a parole violator until he is taken into custody under the violation warrant. Id. at 87, 97 S.Ct. 274; see also United States v. Sanchez, 225 F.3d 172, 175 (2d Cir. 2000) ("[D]elay between a defendant's violation of supervised release and the execution of the violation warrant does not, in and of itself, violate a defendant's due process rights."); United States v. Tippens, 39 F.3d 88, 90 (5th Cir. 1994) (finding a thirty month delay between the issuance of a violation warrant and its execution to be justified where the defendant had been in state custody during the thirty months).
Moreover, the Supreme Court held that the defendant had not been prejudiced by the delay because the Parole Commission had "the power to grant, retroactively, the equivalent of concurrent sentences and to provide for unconditional or conditional release upon completion of the subsequent sentence." Moody, 429 U.S. at 86, 97 S.Ct. 274; see also United States v. Throneburg, 87 F.3d 851, 853 (6th Cir. 1996) ("[The defendant] cites no authority for the proposition that a delay that may affect one's ability to serve sentences concurrently either implicates due process or violates the provisions of 18 U.S.C.A. § 3583(i).").
The Court in Moody unambiguously held that the federal government is not constitutionally required to writ a defendant out of state custody and into federal custody for purposes of executing a violation warrant. Moody, 429 U.S. at 87, 97 S.Ct. 274. Furthermore, the opinion clarifies that a defendant cannot claim prejudice from such a delay on the ground that he is unable to serve his multiple sentences concurrently. Id.
Given the Supreme Court's position that the rights of probationers and parolees in revocation hearings are virtually identical, Gagnon v. Scarpelli, 411 U.S. 778, 783, 93 S.Ct. 1756, 36 L.Ed.2d 656 (1973), the holding in Moody is instructive here. The applicability of Moody to this case, however, is more attenuated than suggested by the government and the district court. The absence of a constitutional right to a prompt parole revocation hearing prior to the execution of a violation warrant does not foreclose the argument that a defendant may possess such a right under the express statutory language of § 3583(i). Thus, Moody is instructive, but not dispositive.
The government also emphasizes our holding in Bartholdi, 453 F.2d at 1226, in which we lent considerable attention to the issue of whether postponing the execution of a probation violation warrant until the defendant is released from state custody is reasonable. In Bartholdi, the district court issued a violation warrant for a federal probationer who was in state custody. Bartholdi, 453 F.2d at 1226. Although the warrant was issued within the probationary period, it was not executed until the defendant was released from state custody nearly a year later, by which time the original probationary period had expired. Id. The defendant in Bartholdi argued that "[t]he warrant was not executed and hearing held within a reasonable time." Id. We rejected this argument, holding that the execution of the warrant could await the expiration of pending state sentences without amounting to an unreasonable delay. Id.; see also Barr v. Parker, 453 F.2d 865, 867 (9th Cir. 1971) ("A warrant may be executed after the term has expired and may await the outcome of pending criminal charges and sentences entered thereon without amounting to an unreasonable delay.").
Garrett challenges the government's reliance on Bartholdi by emphasizing the differences between a "reasonable" delay and a "reasonably necessary" delay. In fleshing out this distinction, Garrett relies heavily on United States v. Dworkin, 70 F.Supp.2d 214 (E.D.N.Y. 1999). The court in Dworkin addressed whether it had jurisdiction under 18 U.S.C. § 3565(c) to revoke the defendant's probation after it had expired. Section 3565(c) contains language identical to that embodied in § 3583(i), permitting only a "reasonably necessary" delay for the adjudication of matters arising prior to the expiration of probation. Id. at 216.
Unlike the case before us, Dworkin was never arrested or incarcerated for the state crime he allegedly committed while on parole. The government in Dworkin delayed the execution of Dworkin's violation warrant in anticipation of the issuance of an indictment against both Dworkin and a key witness against him for the same conduct the government believed had violated the terms of Dworkin's parole. Id. at 215-16. The government asserted that its actions were permissible under § 3563(c) because similar delays had been found by other courts to be "reasonable." Id. at 216. The court rejected the government's argument, holding that none of the cases relied upon by the government was decided under a statute "providing for a time period `reasonably necessary' to adjudicate matters that arose before the expiration of a probationary term." Id. Focusing on the fact that the government took no action against Dworkin until more than three years after it learned of his alleged violation and two years after his probationary term expired, during which time Dworkin was never incarcerated, the court held:
[T]he government offers no "necessity" for the delay in bringing about an "adjudication" of the alleged violation. This is not a case in which, for instance, the government learned of an alleged violation shortly before the expiration of the probationary term. Nor was the government unable to execute the warrant because the probationer was imprisoned in another state or otherwise absent from the jurisdiction.
Id. at 216-17.
This language does not support Garrett's reliance on Dworkin. While the distinction drawn by the Dworkin court between a "reasonable" delay and a "reasonably necessary" delay is applicable to the case at bar, the Dworkin court did not suggest that a delay attributable to a defendant's incarceration on state charges could never be "reasonably necessary." To the contrary, the quoted language suggests that a defendant's incarceration on state charges could present a situation in which a delay would be "reasonably necessary."
However, Garrett's argument concerning the differences between a "reasonable" delay and a "reasonably necessary" delay does not fall on deaf ears. We appreciate the crucial distinction between the two terms and do not purport to view them as one in the same. While Moody and Bartholdi clearly illustrate that the government's delay in executing Garrett's violation warrant was constitutionally permissible and "reasonable," neither case squarely resolves the issue before us — that is, whether the delay was "reasonably necessary" under § 3583(i). So how do we reconcile the "reasonably necessary" language in § 3583(i) with the holdings in Moody and Bartholdi? The answer, we believe, is that they do not intersect. We view the precedent established by Moody and Bartholdi not as our guiding light, but merely as the canvas upon which we paint our interpretive picture.
Our determination of whether the government violated § 3583(i) by waiting until Garrett was released from state custody to revoke his term of supervised release requires us first to ascertain the type of adjudication contemplated by § 3583(i). We conclude that the language of § 3583(i) — "any period reasonably necessary for the adjudication of matters arising before its expiration" — refers to the federal adjudication of the defendant's supervised release violations. Thus, § 3583(i) extends the jurisdiction of the federal court only to the period of time reasonably necessary to adjudicate pending supervised release revocation issues.
The next, and significantly more complex, issue is whether the "reasonably necessary" period of time referenced by § 3583(i) encompasses delays attributable to a defendant's incarceration on state charges. Specifically, we must determine whether the sands of this statutory hourglass should begin to fall immediately upon the expiration of a defendant's term of supervised release, or in the alternative, whether those sands should remain suspended during the defendant's incarceration on state charges. We embrace the latter approach in concluding that once a defendant's supervised release term expires, a district court may, in accordance with § 3583(i), postpone the federal adjudication of matters arising before the expiration of that term until after the defendant is released from state custody. This holding reflects our belief that the purpose of § 3583(i) was to assure reasonable speed of federal adjudication after the defendant is in federal custody; it was not to overrule Moody and Bartholdi.
A contrary interpretation would be tantamount to holding that the federal government is statutorily required to writ a defendant out of state custody and bring him before the federal district court for his revocation hearing. The obvious problem with such a conclusion is that § 3583(i) imposes no such duty on the federal government, nor has any court imposed such a duty. Furthermore, we are mindful of the fact that requiring the federal government to writ a defendant out of state custody for a supervised release revocation hearing could prove extremely burdensome. While the task of transferring Garrett may not have been excessively burdensome due to his detention in a state facility near the federal courthouse, one can foresee numerous situations in which the onus would not be so slight.
Therefore, the government's postponement of Garrett's revocation hearing until his release from state custody did not violate § 3583(i).
4. The delay between Garrett's release from state custody and his final revocation hearing was "reasonably necessary" for purposes of 18 U.S.C. § 3583(i).
The delay between Garrett's release from state custody and his final revocation hearing clearly fell within the purview of § 3583(i). Garrett was held in state custody until late March, 2000. Immediately after being released, the government executed the violation warrant and Garrett was taken into custody. One week later, on April 7, 2000, Garrett appeared before a federal magistrate judge and denied the alleged violations of his supervised release. A revocation hearing was then scheduled in district court for April 17, 2000. At this hearing, Garrett voiced his concern about the district court's jurisdiction to revoke his supervised release. The district court then granted both parties approximately two weeks to fully brief the issue, and held two additional hearings on May 8, 2000, and May 11, 2000. This sequence of events conclusively demonstrates that the delay between March 31, 2000 and May 11, 2000, was "reasonably necessary" for the adjudication of matters arising prior to the expiration of Garrett's supervised release, and thus did not violate the tenets of § 3583(i).
CONCLUSION
For the foregoing reasons, the revocation of Garrett's supervised release term is AFFIRMED.
The majority recognizes that it is not "technically `necessary'" for the federal government to await a defendant's release from state custody before federal adjudication of the revocation of his supervised release can take place. Majority Opinion at 447. Rather, the government can issue a writ of habeas corpus ad prosequendum for the period of time necessary to conduct a hearing, determine whether violations occurred and, if so, impose the federal sentence for the supervised release violations charged. The majority holds, nonetheless, that under 18 U.S.C. § 3583(i), the government can wait until the end of a defendant's state pretrial detention and imprisonment on a state charge to begin the revocation proceedings. I read the statute, instead, as quite explicitly stating otherwise: The court's authority to revoke a term of supervised release extends for as long as it reasonably takes to hold a fair and adequate hearing on revocation, not longer.
If the statutory question were simply whether the delay in this case was "unreasonable," then it might be sensible to conclude that Congress intended in enacting § 3583(i) to codify the standard applied in United States v. Bartholdi, 453 F.2d 1225, 1226 (9th Cir. 1972) and similar cases, and the majority's reliance on that case would be understandable. But § 3583 did not simply enact the general "unreasonable delay" standard articulated in Bartholdi. Rather, the statute is considerably more specific.
See, e.g., Barr v. Parker, 453 F.2d 865 (9th Cir. 1971); Shelton v. United States Board of Parole, 388 F.2d 567, 571 (D.C. Cir. 1967) (cited in both Barr and Bartholdi).
Under the statute, the term of supervised release is not itself extended or tolled by a close-to-end-of-term violation. Rather, it is only the "power of the court to revoke" the supervised release that is extended, and only for so long as " reasonably necessary for the adjudication of matters arising before its expiration." § 3583(i) (emphasis added). So the statute does not simply add the "necessary" requirement, but also provides a benchmark for measuring what the delay must be "reasonably necessary" to accomplish — namely, "the adjudication of" any supervised released violations that occurred before the expiration of the supervised release term.
The majority, moreover, imports into § 3583(i) something that is simply not there — an additional power automatically to further delay revocation proceedings until the defendant is in federal custody. The absence of any statutory basis for such post-term authority is particularly significant, in my view, because the statutory provisions governing supervised release do adjust in another respect for periods of incarceration on charges other than those giving rise to the term of supervised release in question. See 18 U.S.C. § 3624(e) (suspending the running of a term of supervised release while the defendant is imprisoned); cf. United States v. Morales-Alejo, 193 F.3d 1102 (9th Cir. 1999) (holding that § 3624(e) does not apply until after conviction). Had Congress intended, as the majority supposes, similarly to extend, automatically, the post-expiration extended revocation period contained in § 3583(i), it had only either to modify § 3624(e) accordingly, or to include language similar to § 3624(e) in § 3583(i). Congress did neither when it amended § 3583 in 1994.
See § 110505 of the Violent Crime Control and Law Enforcement Act of 1994, Pub.L. No. 103-322, 108 Stat. 1796.
The majority nonetheless reads a provision parallel to § 3624(e) into § 3583(i)'s limited grant of post-expiration jurisdiction, justifying this statutory revision by stating its firm "belief that the purpose of § 3583(i) was to assure reasonable speed of federal adjudication after the defendant is in federal custody; it was not to overrule Moody and Bartholdi." (Maj. Op. at 450.) We do not ordinarily disregard the words of a statute on the strength of our own belief that Congress could not have meant what it said. Pinter v. Dahl, 486 U.S. 622, 653, 108 S.Ct. 2063, 100 L.Ed.2d 658 (1988); see also United States v. Wells, 519 U.S. 482, 497, 117 S.Ct. 921, 137 L.Ed.2d 107 (1997).
Moreover, as the majority initially recognizes and then later forgets, Moody was a case concerned only with whether the constitution requires that revocation proceedings occur promptly even when a prisoner is incarcerated, not with whether it would be preferable as a policy matter to provide for prompt hearings, when feasible, even for incarcerated defendants. ( See Maj. Op. at 448) ("The absence of a constitutional right to a prompt revocation hearing prior to the execution of a violation warrant does not foreclose the argument that a defendant may possess such a right under the express statutory language of § 3583(i)."). The dissent in Moody presented strong policy arguments favoring a contrary approach to revocation hearings: The uncertainty associated with pending charges can adversely affect "efforts to involve the offender in correctional programs" while imprisoned, 429 U.S. at 94 n. 9, 97 S.Ct. 274 (Stevens, J., dissenting) ( quoting Bureau of Prisons, Policy Statement 7500.14A (Jan. 7, 1970)), while a long-delayed hearing is likely to make it difficult for the defendant to marshal evidence relevant to the revocation proceedings as well as to shift the focus of those proceedings to the defendant's behavior while incarcerated. Moody, 429 U.S. at 95, 97 S.Ct. 274 (Stevens, J., dissenting).
It bears noting that although in this case the revocation hearing was delayed only ten months because of the state incarceration, under the majority's holding the revocation could take place many years after the events giving rise to the revocation proceedings.
Moody held that these considerations do not create a constitutional due process right to a prompt revocation hearing if the defendant is incarcerated. But Moody did not limit Congress' authority to credit the policy considerations spelled out by the Moody dissent, and to enact legislation in accord therewith. By adopting as the touchstone for the permissible period of delay not all stretches of incarceration but only any period "reasonably necessary for the adjudication" of an alleged supervised release violation, Congress, it seems to me, chose — once a term of supervised release had expired — to prefer prompt adjudication, where possible, over the indefinite delay involved in awaiting the end of any and all terms of imprisonment.
I recognize that where § 3624(e) does apply, there can be a delay in the instigation of federal revocation proceedings until after the end of state incarceration, however long that might be, because the term of supervised release will not end until then and there is no other limitation in the statute on the timing of revocation hearings. The statute as I read it, in other words, permits delay for incarceration due to a conviction as long as that incarceration begins before the termination of the term of supervised release but not if it begins thereafter. As such, the statutory scheme is certainly not a tidy one. And why Congress chose to make that distinction I can only speculate. Perhaps the factors stressed by the Moody dissent were thought entitled to greater weight where a defendant had gotten through his entire term of supervised release without a second conviction and resulting incarceration than where he had not. Perhaps the defendant's right to a prompt hearing was thought stronger once he was no longer subject to court supervision on account of his earlier conviction. Whatever the case, the distinction is one that could not be clearer in the statute — which is quite explicit regarding the fact that the delayed revocation period does not extend the term of supervised release but only the power of the court to revoke an expired period of supervised release — and is therefore a distinction that we are obliged to respect.
Several courts of appeal have also held, applying Moody, that there is no due process violation where there is a lengthy delay because of the § 3624(e) tolling rule, unless there is an showing of actual prejudice. See United States v. Sanchez, 225 F.3d 172, 175-77 (2d Cir. 2000), and cases cited therein.
Whether the practical difficulties involved in arranging for the presence in court of a defendant incarcerated by a state can sometimes meet the statutory "reasonably necessary for . . . adjudication" standard is a question that need not be addressed in this case. Obviously, the defendant needs to be present for any revocation hearings, and obtaining the defendant's presence may at times be quite burdensome. So it may at times be "reasonably necessary for the adjudication" of revocation to wait some period of time before obtaining custody over the defendant.
That is the case, though, not only when the defendant is imprisoned, but also when he or she is for other reasons out of the jurisdiction, or, while in the jurisdiction, has not yet been arrested for violation of supervised release. The statute as written draws no distinction between incarcerated defendants and others. I doubt that, under the terms of the extended jurisdiction conferred by § 3583(i), the government could wait indefinitely to arrest a defendant on a revocation warrant if the defendant was in his known dwelling place and therefore easy to find. Yet the majority's supposition that "the purpose of § 3583(i) was to assure reasonable speed of federal adjudication after the defendant is in federal custody" (Maj. Op. at 450) appears to suggest otherwise.
In any event, whatever burden might be involved in other circumstances in obtaining temporary custody over the defendant for purposes of holding the revocation and sentencing hearings, there was none here. During the period when the revocation proceedings could have taken place, Garrett was in pretrial state custody a few blocks from the federal courthouse. Obtaining the defendant's presence would have been simply a matter of arranging with state authorities for his temporary attendance in federal court. Indeed, the government in this case has essentially conceded that the delay was in no respects necessary or the result of an undue burden, but argues only that no burden at all, or even inconvenience, need be shown if the defendant is incarcerated. I would reject that proposition, and therefore I respectfully dissent.