University of Virginia School of Law
April 3, 2006
Judicial and Administrative Review of Immigration Decisions
Testimony before the Senate Committee on the Judiciary
April 3, 2006
David A. Martin
Warner-Booker Distinguished Professor of International Law
University of Virginia
Mr. Chairman and Members of the Committee: I appreciate the invitation to appear before you today to address important issues of judicial and administrative review that have arisen in connection with the immigration reform proposals now pending before Congress. I have taught and written about immigration and constitutional law for 25 years. I also spent two and a half years as General Counsel of the Immigration and Naturalization Service in the mid-1990s, an experience that afforded me close inside acquaintance with how review affects the operations of the agencies charged with administering and enforcing the immigration laws.
It is quite true that the current system for administrative and judicial review of immigration decisions is under stress. The greatly increased volume of immigration appeals in the federal courts has imposed difficulties on the courts and their staff, the Department of Justice, and the private bar. The Committee is right to be concerned about these issues. But it would be profoundly unwise, in my opinion, to consolidate all judicial appeals in the Federal Circuit, as was proposed in section 701 of the Chairman's initial bill before the Committee. The nation - and indeed the agencies involved in immigration matters - benefit significantly from the involvement of the generalist regional courts of appeals in the consideration of immigration issues. And those courts have been finding ways to adapt to the new caseload. Their efforts should be allowed time to mature. Also, I believe that the single-judge screening mechanism provided by section 707 of the Chairman's mark would risk denying court consideration in cases where careful review should be provided. It might also prove counterproductive, ultimately creating more work for the courts involved. The remedy for the current problems should focus instead on restoring sound functioning by the Board of Immigration Appeals. This requires both additional resources and the return, in essence, to a system of administrative appellate review that pertained before a set of changes was imposed in the 2002 streamlining regulations, with unintended effects that have brought many negative consequences.
I. The Proposed Transfer of Jurisdiction to the Federal Circuit Would be Unsound
The arguments offered by some for transferring jurisdiction over immigration appeals to the Federal Circuit are thin, and cannot begin to overcome the significant transition costs as detailed by Judge Michel, Chief Judge of the Federal Circuit, in his testimony - to say nothing of other practical disadvantages. Nor do they outweigh the importance of involvement by judges from generalist courts in a caseload that frequently involves questions of individual liberty or the rights afforded by our statutes and treaties for refugees to gain protection from persecution. Generalist courts help counteract the inevitable tendency of specialist agencies to become overly preoccupied with their own missions, narrowly conceived. American administrative law greatly benefits from this balance and counterpoise, and immigration law should not be isolated from these salutary effects. Immigration law can be highly technical, to be sure, but I agree wholly with Judge Newman that most of the cases now reaching the courts of appeals do not present close technical issues. They usually involve factual disputes or controversies over credibility determinations.
The two main arguments for such consolidation are apparently the risk of forum shopping and the need for uniformity and consistency in the administration of the immigration laws. Forum shopping was arguably a problem (though one that presented itself only in a tiny minority of cases) before 1996. Amendments that year, however, eliminated the possibility that an alien could move to another circuit during or after his administrative proceedings in order to take advantage of more advantageous case law. The 1996 legislation added section 242(b)(2) to the Immigration and Nationality Act (INA) to provide that judicial review must be had in the circuit with jurisdiction over the place where the immigration judge issued the initial ruling. The Department of Homeland Security initially determines that venue by filing the charging documents with an immigration court. The regulations afford a limited opportunity to change venue, but change requires an order of an immigration judge, issued for good cause shown.
Consistency and uniformity are of course important values in administrative law, but one can easily exaggerate the extent to which the current immigration system departs from consistency. Most issues of legal interpretation are initially determined by the administering agency, and the vast majority of those then remain undisturbed, assuring a core consistency in operations. When Congress passed the 1996 Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA), it introduced a host of complex new provisions into the INA. I was then at INS, and we developed a careful and systematic process, often involving other components of the Department of Justice or other agencies, to bring interpretive questions to the surface and to consult about the best ways to resolve them. We then built those understandings into the implementing regulations and other guidance. I know that in the overwhelming majority of cases, those interpretations simply became part of the shared understanding, by agencies, the bar and the public, of the new provisions. Most such interpretations were not questioned in court and have provided uniformity in the implementation of the law. A mere handful, compared with the range of issues initially presented, have been challenged in litigation, but principles of administrative deference (including the Supreme Court's Chevron doctrine) usually have resulted in judicial acceptance of the authoritative agency interpretation. It is only in a small number of instances that the circuits have split on such questions.
Some such disagreement is wholly understandable, and temporary persistence of differences among the circuits on this small number of issues does not significantly impede the agencies involved. The immigration laws raise important questions that touch on broad considerations of national policy, foreign relations, sound administration of a complex administrative scheme, individual liberties, and our shared national heritage as a country of immigration and refuge. Circuit splits historically serve the purpose of helping to signal when there are ambiguities in the law, significant constitutional issues, or difficulties in reconciling the many policy objectives our immigration laws serve. Ultimate resolution by the Supreme Court of these kinds of difficult questions benefits from the efforts of seasoned judges from different circuits to analyze the issues afresh, and from acquaintance with many different fact patterns, as presented in the various circuit decisions, that may foster the Court's understanding of the full stakes. If all appeals went only to the Federal Circuit, a prematurely uniform resolution of truly difficult questions of statutory interpretation or constitutional application might impede this percolation process, as it is often called. The system usually benefits from circuit splits in the relatively small number of instances where they occur - because deliberation by several courts helps to think through the best way ultimately to resolve the issue.
I say this even though I remember occasional frustration, during my time at INS, with judicial decisions I thought deeply wrong - a frustration often shared by my DOJ colleagues. But consolidation in a single court affords no guarantee against occasional unsound decisions, and I am pleased to see that the DOJ testimony offered here today does not say much in support of the proposal to move all cases to the Federal Circuit. Moreover, the fact of different rules in different circuits on a few questions does not significantly hamper the agencies involved. It has been quite common to issue legal and operational guidance that is circuit-specific - until such time as court decision or statutory change resolves the differences.
It is said by some that the Supreme Court is not equipped adequately to resolve circuit splits on immigration matters. That claim is belied, in my opinion, by the fairly active immigration docket the Supreme Court has maintained - resolving circuit splits in recent years, for example, over questions of habeas corpus review, the retroactivity of certain changes to relief from removal, the application of the "aggravated felony" definition to certain crimes, and the detention provisions of the statute.
More importantly, Supreme Court review is not required in order to restore uniformity on most such differences in the immigration field. Congress has often responded to circuit splits on important statutory questions by amending the underlying provision. The REAL ID Act of 2005, for example, contained important new provisions on credibility determinations and corroboration requirements in asylum cases - precisely one of the circuit splits that is highlighted by some who favor consolidation of immigration cases in a specialized court. Because those new rules apply only with respect to asylum claims filed after May 11, 2005, we have little experience so far with their impact on judicial review. But I expect to see far less in the way of variance among the circuits once the amendments take full effect. Furthermore, other provisions in pending immigration reform legislation, clarifying existing provisions in light of controversies that have arisen in ongoing litigation, are likely to resolve differences on other key issues. For example, section 705 of the Chairman's mark would amend the reinstatement of removal provision, INA § 241(a)(5), in a way that would end the circuit split on this issue, which is highlighted in the testimony of Judge Bea.
It would be far better to serve the objectives of uniform and consistent administration through strengthening the administrative agencies that make the initial judgments. I speak below of ways to strengthen the BIA and restore its historic role as a primary venue for consistent and uniform rulings on issues of law and fact, one that will consistently inspire the full measure of judicial deference. But uniformity has also been hampered by unintended effects of the split of immigration functions among three separate bureaus of the Department of Homeland Security (DHS) - without, as yet, a truly effective departmental mechanism for timely resolution of differences among those bureaus on difficult questions of law and policy. The Committee could strike a far more effective and important blow for consistency and sound administration by addressing that issue - for example, by consolidating ICE and CBP within DHS, or by placing both those bureaus, plus USCIS, under a single Under Secretary for Immigration Affairs (or Immigration and Customs Affairs).
II. The Proposed Certificate of Reviewability Risks Jeopardizing Important Values and May Not Save Significant Judicial Resources
Section 707 of the Chairman's mark also proposes that appeals could be heard in an article III court only following a grant by a single judge of a certificate of reviewability (COR). The Department of Justice strongly supports this procedure, and analogizes it to the procedure for issuance of a certificate of appealability by a single judge under 28 U.S.C. § 2253. That analogy is deeply flawed, and the proposed procedure should not be adopted. The individuals involved in standard removal cases deserve at least one opportunity for full consideration by Article III judges, although of course it is appropriate to resolve insubstantial appeals through summary measures or other court management practices like those already in place in the courts of appeals. The existing steps taken by the regional circuits are beginning to master the recent jump in immigration filings - as described, for example, in the testimony of Judge Walker and Judge Newman, and in other communications the committee has received. We should at least gain more experience with the full impact of those court management measures before adopting so sharp a departure (as the COR represents) from our historic commitment to court access in cases where individual stakes may be quite high and constitutional claims may be implicated. Moreover, as a practical matter, experience with similar screening measures in other settings suggests that adopting this prior review system might even slow down the resolution of immigration cases.
Section 2253, the allegedly analogous procedure, governs appeals to the courts of appeals of decisions by federal district court judges on habeas corpus petitions filed by persons challenging criminal convictions entered by state or federal courts. Those convictions have already been ruled upon by judges - not solely by administrative adjudicators - and have been subject to a full spectrum of direct judicial review, including possible access to the U.S. Supreme Court on a petition for certiorari. Section 2253 then governs collateral review that becomes applicable, if at all, after multiple layers of direct judicial review. Even there, the petitioner has access to full consideration of the habeas petition by a federal district court judge. The gatekeeper provision (the certificate of appealability) comes into play only after that judge's consideration, and in part it is meant to reflect sensitivity to the important separate status of the states under our federal system of government.
The immigration setting is strikingly different. It presents no federalism issues requiring deference to the courts of the states. Immigration control is clearly a federal function. Moreover, Section 707 covers direct review - the only opportunity for an individual to present his case to an Article III judge, in a field where limitations on judicial access at least raise significant constitutional doubts (as the Supreme Court ruled in INS v. St. Cyr). I am sure that the single judges performing the screening would carry it out conscientiously. But there is always a risk of erroneous dismissals of meritorious claims, a risk that can be diminished if there is panel consideration or more complete briefing. This is particularly a risk at present, because, under administrative changes adopted in 2002, the BIA often provides only single word affirmances or brief rulings that do not greatly help the reviewing judge in understanding the record, the issues, or the relevant case law.
Much will turn on the governing standard for granting the COR, as it actually comes to work in daily operation. If that standard is too stringent, the procedure will effectively dispose of potentially important claims based on limited information and without briefing by the government, and may miss significant statutory or constitutional issues. Or if, as is likely, a great many of the appeals present factual issues, the single judge may have to delve deeply into a voluminous record to decide on the certificate - and likely will end up passing such factual questions on to a full panel of three judges. In the latter case, the court will in essence have to plunge into the merits twice. Judge Michel's testimony points out that the Federal Circuit considered a similar procedure for screening Merit Systems Protection Board appeals, but ultimately rejected it because the court concluded that the new procedure might actually require the court to consider most cases twice.
In sum, the screening procedure will achieve judicial economies only if it actually screens out a fairly high percentage of the cases. The proposal is accompanied by no studies, to my knowledge, suggesting that that would be the case. My understanding of the caseload suggests quite the opposite - and some of the judges testifying today, having a far closer acquaintance with the current caseload, bear out this impression. Nonetheless, if stringently administered, the procedure seems likely to prevent meritorious cases involving high stakes and possible constitutional issues from receiving the kind of review they deserve from Article III courts, and that our nation has historically provided. But if the standard is softened to avoid that highly negative result (a later amendment proposed by the Chairman, as I understand it, would move strongly in that salutary direction), then the process is likely to screen out relatively few cases - certainly few that are not currently handled by speedier procedures already implemented by the circuit courts. It is highly significant that the Executive Committee of the Judicial Conference of the United States met last week and specifically decided to oppose the screening procedure set forth in section 707, in part because of the inadequacy of the administrative record currently being transferred from the BIA. I understand that the March 31 letter from Leonidas Ralph Mecham, secretary of the Judicial Conference, is being entered into the record. That letter explains: "Streamlining both the administrative and appellate review of immigration cases raises concerns about whether the process would provide a meaningful review . . . ."
III. Concentrate on Resource Additions and Reforms for the BIA and Immigration Courts
As that letter indirectly suggests, the current stresses on the system for judicial review of immigration decisions are best addressed by undoing the counterproductive elements in the administrative streamlining adopted in 2002 and pursuing reforms that will restore sound functioning of the adjudication and appeals system at the administrative level. It is bound to be more cost-effective and efficient - and more likely to promote uniformity and consistency - to assure full and fair appellate review at the administrative level, rather than having to make up for systemic deficiencies through judicial review. To the extent that litigants feel that their claims were fully and fairly heard, judicial appeals will recede, and courts will also be more likely to defer to the administrative result. Title VII of the Chairman's mark contains many promising provisions to this end, and a later amendment drafted by the Chairman but, as I understand it, not formally considered by the Committee, would introduce further improvements.
The major jump in immigration appeals to the courts of appeals derives from at least three sources.
First, a significant increase in caseload was predictable once Congress stepped up enforcement activity through significant additions to INS resources in the 1990s. Activity essentially doubled, and so did the number of immigration judges. That expansion itself was likely to bring to the courts an eventual doubling of the appellate caseload (a need not adequately factored into budget allocations for the judiciary in the wake of those enforcement enhancements), but the courts were shielded from the full impact for several years by backlogs at the BIA. When the BIA cleared much of the backlog in a burst of activity in 2002-04, a large spurt of appeals suddenly arrived at the courts. Once that backlog-clearance spurt is played out, we should see some easing of the strains, but we can still expect total numbers of appeals in a steady state at least double that of the 1990s, simply because of increased enforcement activity.
The increase in judicial appeals over the last year or so has not been a doubling, however; it has been roughly a five- or six-fold increase. This is because appeal rates have climbed from the historic level of approximately 10 percent of BIA decisions taken to federal court, up to 25-30 percent. Some of that higher rate derives from a second factor: Congress made major and complex changes in the immigration laws in 1996 in IIRIRA, raising a host of interpretive questions. One can expect an increase in court challenges in the wake of any massive legislative change, until the new statutory provisions receive authoritative interpretation (and rulings on their constitutionality) - but that is an effect that eventually wears off. We are probably still in the midst of this settling process for the 1996 changes. (Statutory stability would contribute significantly toward eventually reducing the number of appeals.) This factor may account for some of the increase in appeal rate, but probably only a portion of it.
Most of the jump in court appeals derives, I believe, from profound dissatisfaction with changes in the way the BIA handles appeals, imposed in 2002. This change was done in the name of streamlining, and it involved the following:
? limiting BIA authority to review factual determinations by immigration judges;
? imposing tight time limits on consideration by the BIA
? providing that most cases would be resolved by single members of the BIA
? allowing for affirmance without opinion in a wider array of cases.
In addition, the 2002 regulations decreed a cut in the size of the Board from an authorized 23 members down to 11, to be imposed six months after the effective date of the regulations - without waiting to see whether the changes really mastered caseload pressures or adversely affected the quality of decisions.
Experience has shown that these regulations, designed, as I understand it, by assistants to the Attorney General with relatively little input from the BIA itself, have had most unfortunate impacts. Why it was thought that a cut in the Board size was a good idea, in the face of a massive caseload, is mystifying. That bit of staff paring - ostensibly made possible by authorizing more summary affirmances - proved to be a wholly false economy. For, in my view, these changes quickly helped trigger the higher appeal rate that has consumed so much time and energy of appellate lawyers drawn from far and wide in DOJ, with the full range of negative consequences detailed in the statement of Jonathan Cohn of the Department of Justice. Restoring the perception and reality of full and fair consideration of appeals at the BIA should eventually reduce the number of appeals to the federal courts, although the impact will not be immediate.
I certainly do not claim that the streamlining regulation was the sole reason for the appeals. There are frivolous or clearly ill-founded appeals. Other witnesses also mention the desire by aliens to delay their removal by filing an appeal, That risk is one reason why the law was changed in 1996 to eliminate automatic stays of removal throughout the pendency of judicial review. But that factor - the desire to delay removal in the hopes of developing additional equities that might allow the person ultimately to stay - operated in exactly the same fashion before 2002, in the days when appeal rates ran at approximately 10 percent. And even before 2002, the BIA had more focused ways of dealing with clearly meritless appeals in summary fashion.
The BIA itself, through a patient process initiated and shepherded by its former chairman, Paul Schmidt, had adopted an earlier streamlining regulation in 1999 that was making significant inroads on the administrative appeals backlog by 2002, without triggering significant negative effects on judicial review dockets. We should, in general, now go back to that system. It enabled the Board to deal expeditiously with weak cases, but it had a much more limited list of instances in which cases could be resolved by single members of the Board or without a full BIA opinion. It also maintained BIA authority to engage in full review of factual determinations, a power taken away in the 2002 streamlining. The Board did not use that power indiscriminately. Under a set of precedent decisions issued in the late 1990s, the Board made it clear that it would ordinarily defer to the factual findings of the immigration judge in the vast majority of cases, but it retained authority to delve deeper if there appeared to be significant problems with the immigration judge's assessment of the factual record. Occasional exercise of that factual review authority by the BIA can be quite useful, because immigration judges will continue to handle high caseloads and will continue ordinarily to dictate their decisions at the close of the hearings. When IJ mistakes occur, it is far better for the BIA either to correct the factual conclusions, offer a sound rationale that really justifies the factual conclusions reached (if the defect is simply in summarizing the record or drawing conclusions from it), or else remand the matter because of factual mistakes that may not rise entirely to the level of "clearly erroneous," before the matter ever reaches the federal courts. (It is precisely such IJ mistakes regarding the development or analysis of factual records that now often seem to trigger harsh criticism of the administrative system by federal judges.) An independent review of the 1999 streamlining by an outside management consulting firm found that those reforms had been largely successful, were generally well received, and were having an appreciable effect in eliminating the BIA's case backlog.
I would urge that the BIA's procedures revert in most respects to the 1999 regulations. Much of section 712 in the Chairman's mark would move in that helpful direction. For example, it would limit the occasions in which single-member decisions, affirmance without opinion, or other summary disposition would be authorized. Importantly, it would also expand the size of the Board, returning (under the Chairman's later proposed amendment) to an authorized level of 23. The Committee should consider a still larger increase in Board size. Section 712 also contains other provisions, however, that raise further concerns and should not be adopted without a good deal more study of their likely impact. It is definitely important to assure the decisional independence of Board members and immigration judges in resolving individual cases, and I applaud the bill for its efforts toward that end. But the exact measures for such insulation must be carefully designed to allow for the fact that precedent decisions also may touch on delicate issues of foreign policy or national security. The changes in section 712 may go too far in disabling the Attorney General from stepping in, through a carefully structured (and infrequently used) certification or referral process, to resolve by adjudication certain overarching questions that entail both policy and law. Exact measures for appointment and removal of members also need also to be carefully scrutinized. And as indicated, I think a wider range of BIA authority to reconsider factual determinations would be advisable - though it should be sparingly employed.
In significant part the issue is resources. The most promising investment Congress could make in solving the problems that rightly motivated the Chairman's proposals to alter judicial review would be in additional resources for the immigration courts and the Board of Immigration Appeals, plus related government functions. Such augmentation is authorized in section 702 of the proposed Title VII and is supported by several of the judges testifying today - and by others who have looked closely at the unfortunate byproducts of the 2002 streamlining. This means both an increase in the size of the Board, as mentioned above, and in the staff of attorneys who assist the members in preparing the cases for decision. Permitting ample consideration and the preparation of a full opinion in the vast majority of cases would, first, reduce the number of appeals that result from litigant frustration over perceptions of insufficient Board attention to the individual's arguments. Second, it would also better assist the courts in disposing efficiently of appeals that do result.
Although the Department of Justice was right to take steps to catch up with the BIA backlog, the administrative system took a wrong turn in 2002, with unintended consequences that have imposed significant costs on courts, the Department of Justice, and the respondents in removal cases. The Committee is rightly concerned about the problems that have resulted, but the answer, at least for now, should not be sought in consolidating appeals in a single circuit or imposing a problematic screening barrier to judicial review. I urge the Committee instead to let the regional courts of appeals address these issues through their current court management initiatives. The Congress should concentrate its attention in this realm on restoring a fully reliable system for administrative appellate review, through greater resources and other reforms, rather than introducing at this time the sharp and problematic changes implicit in sections 701 and 707 of the Chairman's mark.