In its early years, P-1 operated mainly as a vehicle for individual referrals, primarily coming from UNHCR. But the system as designed also allowed UNHCR to refer groups that meet the standards of this priority. At various times, the Office did refer a few groups for U.S. resettlement, such as specific groups of Benadir and Barawans from Somalia in 1996 and 1997.⁵ By the turn of the century, however that UNHCR group referral practice had declined, fading from the memory of many persons involved in the admissions system (who by then associated group access solely with the P-2 category), to the point that the 2003 UNHCR referral of a P-1 group of vulnerable Liberians in Cote d'Ivoire was widely regarded as a novel development. But that referral amounted instead to a revival of an earlier practice, not a novelty.

2. The P-2 category: groups of special humanitarian concern

P-2 is for groups of special humanitarian concern to the United States who are designated by the US government for resettlement processing. Specific categories from the former Soviet Union, Cuba, and Vietnam, processed inside their countries of origin, have consistently been designated as P-2 groups, and other categories have been added at various times. These have included the so-called Lost Boys of Sudan, certain Burmese ethnic leaders and political activists, and mixed-marriage families from Bosnia (along with a wider list of victimized groups from Bosnia). For FY 2004 the congressional consultation document listed these additional P-2 groups: Somali Bantu in Kenya, Baku Armenians in Russia, and Iranian religious minorities, primarily in Austria.⁶ It also stated that other groups were under consideration, and some have been added during FY 2004, including Meskhetian Turks in Krasnodar, Russia, certain Hmong refugees in Thailand, and a specific population of roughly 2,000 Vietnamese in the Philippines.

The history of this priority is instructive. As of the 1994 change in the priority system, P-2 was not expected to be the major or exclusive vehicle for group admissions, because continuing P-1 group referrals from UNHCR were anticipated. P-2 instead was adopted primarily to allow for the admission of groups that are covered by US law and policy but for which UNHCR involvement or referral could not be expected. For example, UNHCR's authority normally extends only to persons who are outside their country of nationality.⁷ For historical reasons, however, growing out of U.S. experience with the Cuban Freedom Flights and the Vietnamese Orderly Departure Program, the Refugee Act of 1980 authorized U.S. processing, as refugees, of persons not yet internationally displaced but who face persecution in their country of residence.⁸ UNHCR would lack the legal authority to screen such persons in their countries of nationality. Therefore, such "in-country refugees" from the former Soviet Union, Vietnam, and Cuba have been dealt with through the P-2 category. They comprised a high percentage of U.S. refugee admissions during the 1980s and 1990s.

Later, as political or financial limitations on UNHCR's potential involvement in other refugee situations became apparent, PRM began to use P-2 more widely. Certain Bosnian groups were designated in 1997, for example, under the P-2 framework, based in part on the initial rationale that these were persons whom UNHCR would have referred, if it had sufficient resources for full engagement there. The Bosnian group designations were quite specific, such as persons in ethnically mixed marriages or former detainees who had been held on account of ethnicity or political or religious opinion.⁹ The designations were specific because this part of the USRP used an "open access" model. That is, Bosnians who believed that they fit one of the P-2 group specifications could present themselves directly to the processing agency. Obviously they still had to pass the initial screening and INS adjudication of their claim to refugee status before being approved for resettlement. But access to the process did not depend on a prior official notification or appearance on a specific registration list; individuals could take the initiative to start the process. Today's P-2 designations for Baku Armenians, Meskhetian Turks, and Iranian religious minorities in Vienna also follow the open access model. Open access can work because these specific groups are rather narrowly or specifically defined or because other factors (such as limited permission to travel to Austria) impose constraints on the potential universe of applicants.

As PRM began to make wider use of the P-2 category, it then developed another model as well, sometimes called a "circumscribed group" model. Such a group designation commonly requires not only a finding of common characteristics that indicate a risk of persecution, but also a registration or other way of obtaining a definitive list of persons, in advance of group designation. This practice started with a specific group of Burundian and Rwandan families of mixed marriages between Hutus and Tutsis, located in the Mkugwa camp in Tanzania, and fully registered by UNHCR.¹⁰ Later, PRM added P-2 designations, based on a consolidated registration or verified list, for Sudanese Somalis in Dadaab camp in Kenya, the so-called Lost Boys of Sudan, Ogoni refugees in Benin, and Hmong Lao in Thailand. The circumscribed group model is seen as necessary in these circumstances in order to avoid overwhelming the processing agencies, to defeat attempts at fraud, and to minimize any magnet effect. PRM officers indicated that they had felt it important to have a solid way to "draw the circle" i.e., to assure themselves of ways to identify the intended beneficiaries definitively B before announcing any designation. This model amounts essentially to the "finite group" approach discussed in Chapter I B which is likely, as that Chapter discussed, to be of increasing importance in the conditions under which refugee resettlement must function in the twenty-first century.

3. Categories P-3 through P-5, plus Visas 93: family-based access

P-3 through P-5 are based on family relationships to persons in the United States, but these priorities are made available only to specific nationalities. P-3 includes spouses, minor children, and parents, whereas P-4 and P-5 apply to more distant relatives.¹¹ The nationality must be on the list at the time that the anchor relative in the United States files the Affidavit of Relationship (AOR) that starts the process going. For most of the time that these priorities have been in use, anyone lawfully resident in the United States (including citizens, lawful permanent residents, persons having refugee or asylum status, and certain parolees) could file the necessary documents to initiate resettlement for a relative, if that nationality was covered by an operative priority. For FY 2004, however, this privilege has been limited to persons initially admitted to the United States as refugees or asylees. This latest restriction has been explained on the basis that the U.S. Government will have sufficiently complete family information only for those categories of anchor relatives B information needed to provide a good basis for checking the validity of the claimed family relationship.¹²

In addition to the P-3 category, a provision added by the Refugee Act of 1980 keeps families together by providing categorically that the spouse and minor unmarried children of a refugee are entitled to derivative refugee status themselves, even if they do not have an individual risk of persecution.¹³ (Such derivative relatives may be of any nationality.) This provision works straightforwardly when families move together from a refugee camp to their new home in the United States. It has raised more complications, of both process and proof, when a spouse or child uses this provision to "follow to join" the principal refugee, as the Refugee Act specifically permits. As Section E and Chapter V explain in greater detail, the follow-to-join provision for spouses and children is implemented through a process known as "Visas 93," often handled by consular officers through procedures that differ markedly from the normal refugee admissions machinery. Because the Visas 93 procedure is unfamiliar to many consular officers, delays or misunderstandings often arise. P-3 eligibility, for the nationalities listed in a given year as eligible for that priority, overlaps a fair amount with the Visas 93 process. Both provisions cover spouses and unmarried minor children; P-3 differs in that it also includes parents of the principal refugee. Although P-3 requires each relative to satisfy the refugee definition independently, while the Visas 93 process does not, P-3 is often favored by refugees and the volags that assist them because the procedures are seen as more reliable and familiar.

P-4 and P-5 have not been made available since the end of registration for the Bosnia P-4 program in November 1999. P-3 was made available for only four nationalities in FY 2003, and nine in FY 2004, although it has been open to as many as 18 in the past. Reductions in the list of eligible nationalities in later years derived in significant part from concerns about fraud (discussed in Section E below).¹⁴ Actual P-3 admissions for any given year, however, typically include a wider array of nationalities, because eligibility is governed by the nationality list that applied at the time the AOR was filed. An AOR provides only access to the process, not a guarantee of admission. DHS still must find that each person to be admitted under P-3 individually satisfies the refugee definition, or is the spouse or child of an accompanying relative found to be a refugee. Many persons who fit P-3 through P-5 might also be eligible, now or after the anchor relative naturalizes, for immigration in one of the family preference categories of the regular immigrant admission system. Backlogs in that system usually make the refugee route advantageous, however, if the applicant can demonstrate a well-founded fear of persecution.¹⁵

B. Broad restructuring of the priority system?

Dissatisfaction with the priority system usually focuses on the relatively limited use of P-2 group designations (which is addressed in Chapter II) and on the limited list of nationalities eligible for P-3. Many people argue for a universal P-3, available to all nationalities. Other critics suggest that too much reliance in the overall system is placed on UNHCR's role in referring refugees (for P-1 admissions) B a gatekeeper role for which, they argue, UNHCR has not been adequately staffed or supported. They want to see a greater push for embassy referrals and a wider referral role for NGOs. Some consider that the P-1 category is too diffuse, including a wide range of different types of potential refugees that would better be separated for admissions planning purposes. If it were divided so as to create a wider array of categories, they assert, the priority system might better allow for giving precedence to the truly urgent case. And some critiques call for a complete revamping of the priority structure. This Section considers proposals for broad restructuring. Later sections then move on to review specific suggestions for each of the primary functional categories: individual, group, and family processing.

1. The Frelick proposal

Bill Frelick, who has extensive experience in refugee issues and is currently Director of the Refugee Program for Amnesty International USA, proposed a thoughtful restructuring of the priority system in testimony before the Senate Subcommittee on Immigration in 2002.¹⁶ Frelick begins by noting the unevenness of the list of factors that govern the existing P-1 category. That list covers a wide variety of situations, some far more compelling than others. He proposes regrouping them with closer attention to their relative urgency, as well as introducing new priority categories based on some factors not directly addressed in the current system.

Frelick would pull together the most urgent refugee cases in countries of first asylum as a new P-1 category, still to be based on UNHCR or embassy referral. These would be the first three subcategories in the list set forth in Section A1 of this Chapter, plus persons in urgent need of medical attention (the eighth factor listed). P-2 would be a new category, although one that is similar to categories used in the pre-1994 system: persons whose fear of persecution is based on actual or imputed association with the U.S. government or U.S. nongovernmental entities. P-3 and P-4 would also be break-outs from the current P-1 factor list. P-3 would be for women at risk in the first-asylum setting and P-4 would cover physically or mentally disabled refugees and survivors of torture or violence. Frelick's P-5 would be what is essentially our current P-2 category, State-Department designated groups of special humanitarian concern, selected based on specific group circumstances. P-6 would become what is essentially today's P-3, close family members, but open to all nationalities. And P-7 would be for long-stayer refugees whose situation in the country of first asylum does not amount to a satisfactory durable solution. He lists specific factors to consider in making that determination.

Frelick has subdivided the list and paid close attention to the ordering because he intends these categories to be true priorities: it is more important to use admission spaces for his P-1 refugees than for P-2s, and so on, and admissions must take that hierarchy into account. Significantly, his P-7, long-stayers, is meant to be a category that comes into play in order to assure use of PD admission slots that would otherwise go unused B and there have of course been such unused slots, sometimes in large quantities, throughout the history of the program. If a given year appears unlikely to make use of the full allotment for persons in the first six priorities, he suggests, then the U.S. should process long-stayers so as to make full use of the PD resource. He is aware of the difficulties of initiating resettlement in the midst of a long-term refugee situation, and he offers specific steps to minimize resentment by those not chosen and to reduce any pull factor.

2. Evaluation and recommendation for more modest revision

Frelick's proposal is a worthy effort to retool the priority system to respond better to real humanitarian need, and the system he proposes would certainly be workable. I recommend making more modest changes in the overall priority system, however. In my view, the major difficulties in the current system do not result from the priority system as such. Numbers have gone unused, for example, not because of the lack of a P-7 category that can serve as a reserve pool, but because of difficulties with the decision-making structure and a failure to treat the PD number as a target. These problems are addressed directly by the recommendations in Chapter II. Moreover, even if we agreed to fill unused slots with long-stayers, we would still face difficult decisions about just when to trigger that resettlement (Frelick does not contemplate that we wait until the final quarter of the fiscal year to see how large a shortfall is looming) and above all, which long-stayer situations to focus on. Making those decisions will inevitably require group-based determinations by those in charge of the USRP, not unlike what is now required for the current P-2 process. Although it is definitely wise to look more systematically at protracted refugee situations as the source for significant admissions in this new resettlement era, this objective is best accomplished by incorporating it into an established group designation process, as outlined in the previous Chapter.

Frelick's ranking of the relative urgency of current P-1 factors (as reflected in the different break-out categories) makes sense, but it seems unlikely that there will often be situations where we must choose the higher ranking categories to the exclusion of the lower. The mechanism will still be individual referrals, if not from UNHCR, then from a U.S. embassy or, under expanded arrangements, from NGOs or other partners. Increasing the overall system's capacity for individual referrals could proceed without explicitly subdividing the different subcategories of P-1. Also, giving priority to admission of persons who are persecuted precisely because of their ties to the United States is a worthy objective, but that factor can be accommodated within the individual referral system. It might be wise to add that consideration explicitly to the P-1 list B presumably a factor that is more likely to be acted upon by U.S. embassies than by UNHCR. If identifiable groups are targeted for persecution because of such U.S. ties, that fact would also constitute a strong argument in favor of P-2 designation, which of course can be so linked, on a case-by-case basis, without any change in the overall priority system.

Frelick's proposal does helpfully point out a major disconnect between one of the P-1 factors and virtually all the rest. The factor that is based on lack of another durable solution combined with an unsatisfactory situation in the country of first asylum (the final factor in the list set forth in Section A1) presents less of the immediately compelling short-term risk that characterizes most of the other entries on the list. Moreover, such a characteristic is less likely to apply to isolated individuals, but instead will probably be shared by a larger populace. That is, it is more likely to lend itself to a P-1 group referral. UNHCR is working to develop a new set of procedures that will facilitate systematic use of group referrals on a larger scale in the future (see Chapter VI), and a great many of these are likely to derive from the comparable no-other-durable-solution category used by UNHCR in its own resettlement referral system.

I would prefer to use this functional insight and implement only a modest revision in the priority system. At at time when the refugee admissions system is going through several other significant changes, there are advantages to keeping the priority system as stable and straightforward as possible. We essentially have operated with three broad functional sources of refugee access: individual referrals, group decisions, and family-based access. These correspond roughly, but not precisely, to the current P-1, P-2, and P-3 categories. We should bring those correspondences more completely into line, by treating all group-based admissions under the P-2 category and employing for all a systematic group decision mechanism, one version of which is outlined in Chapter II. When UNHCR refers a group, as it may well do with increasing frequency over the coming years, that referral should then be given prompt consideration for inclusion as a priority group within the USRP. Any such UNHCR group referral should ordinarily come to the Refugee Admissions Committee with considerable momentum for approval.¹⁷ But there should be a deliberate US decision process for deciding whether we want to participate in that resettlement effort (as opposed to, say, leaving a certain situation primarily for other resettlement countries). That decision should be worked into the same procedures that will come to be used on a more frequent basis for the traditional P-2 group decisions. Hence we would cease using the terminology of "P-1 groups," instead coming to recognize that P-2 groups could originate from different sources B the familiar U.S.-initiated group designation or a UNHCR referral.

Beyond this, I agree with Frelick that we now have little need for the current P-4 and P-5 categories as such. Such factors B more distant family relationships to persons resident in the United States B have been sparingly used and do not deserve their own priority designations. Nonetheless, these relationships may in limited circumstances remain of importance. For example, the United States has sometimes felt called upon to respond to a large refugee flow (such as from ex-Yugoslavia or Afghanistan), but under circumstances that make it clear that resettlement can address only a small fraction of the overall need. In these settings, we may occasionally want to employ family ties to US residents (more distant than those relationships covered by P-3) as an access factor, probably just one among many B that can help select a manageable resettlement pool out of a potentially enormous and overwhelming refugee population. But we do not need separate priority categories for these purposes. Such qualifications can be worked into the description of the precise group given priority access under the P-2 process.

Recommendation III-1: The priority system should be revised modestly, so that P-1 becomes the priority for individual referrals from UNHCR, U.S. embassies, or, in some circumstances, other referring entities; P-2 covers all decisions for designated groups; and P-3 remains the category for close family members of U.S. residents. P-4 and P-5, now covering more distant family relationships, should be removed as priorities, although those same family connections might sometimes be characteristics used in specific P-2 designations. UNHCR group referrals should be worked into the group designation process under P-2, and should cease being thought of as "P-1 groups." Such UNHCR referrals should generally enjoy additional momentum for approval as an access category. PRM should consider refining the list of factors for P-1 referrals accordingly, as well as adding a factor for persons facing persecution that is based on their real or imputed ties to the U.S. Government or U.S. entities. PRM should also regularly post on its website a list of P-2 groups currently being given access to the U.S. system.

C. Individual referrals

1. UNHCR referrals

P-1 admissions derive primarily from UNHCR referrals. This system has worked reasonably well, although a few reforms would be worthwhile. The main complaints I heard about individual referrals were that there are too few of them. This theme has been particularly urgent in the past two years of low overall admission totals. Some have blamed the 1994 redesign of the priority system for the problem, believing that it unwisely gave too much of a gatekeeping role to UNHCR without assuring that the latter was funded and staffed at a level that could keep admissions at historic US levels, as the Soviet and Indochinese programs phased down. Nonetheless, it should be noted that the 1994 changes did not necessarily require that most admissions come through UNHCR. P-2 has always played a significant role in sustaining a large-scale program, and PRM's use of the P-2 category has adjusted over time in light of limitations affecting UNHCR, as described in section A2.

PRM has provided the UNHCR protection division with several million dollars worth of additional resources for referrals in recent years, eventually insisting upon performance criteria based on actual referrals produced as a result of the enhanced funding.¹⁸ UNHCR has added staff for these purposes and has increased its output of individual referrals. It has also made increasing use of a cooperative deployment scheme with the International Catholic Migration Commission (ICMC) that was first established in 1998. Through these arrangements, ICMC maintains a roster of persons with expertise in various phases of resettlement work, including refugee identification, claim verification, case processing, and the best interests determinations (BIDs) needed before unaccompanied minors are assigned for resettlement. UNHCR can call upon them as needed for assignments of several months' duration. As of December 31, 2003, 114 persons were on the roster, 43 of them deployed to the field.¹⁹

PRM should keep up the pressure for the system to continue growing its capacity, including use of cooperative deployments from ICMC or other NGOs. The individual referral process at UNHCR can be quite cumbersome, often involving an individual refugee status determination (RSD) based on lengthy individual interviews, followed by the completion of a multi-page resettlement registration form (RRF). Some UNHCR field offices have experimented with a streamlined process, particularly using abbreviated forms for resettlement referrals to the United States. Although there are tradeoffs to be made, UNHCR should consider further streamlining, at least on a selective basis.²⁰ A one-size-fits-all model of individual referrals may not be appropriate. Some countries in Europe accept UNHCR referrals based on the dossier only, without doing their own separate pre-admission interviewing. For those countries, the elaborate UNHCR investigation and paperwork may make sense. But, for interview-based systems like that of the United States, which inevitably include their own separate and thorough inquiry into refugee status and other qualifying characteristics, UNHCR could simplify its process.

PRM should also continue to support ongoing UNHCR reforms that are meant to enhance quality control and consistency in the standards for individual referrals. I was told of situations where some offices are more likely to refer troublesome or pushy refugees from a first-asylum situation, simply in order to be rid of the hassle, without much attention to the merits of the individual's case for resettlement. The UNHCR individual referral system must reliably focus on individually worthy cases, according to the criteria set forth in the UNHCR resettlement manual. UNHCR is aware of this issue, and has been developing new measures, such as greater use of regional resettlement hub offices, that are improving quality-control monitoring and overall performance.²¹

2. U.S. embassy referrals

Although U.S. embassies have not produced a high volume of referrals, this mechanism should certainly remain available as a source for P-1 admissions. The Department of State should include in the initial training of foreign service officers more systematic instruction on refugee and humanitarian programs generally and on the specific opportunity and procedures for referrals of this type. It should also improve and simplify the annual State Department instruction cable regarding embassy referrals.²² The current procedures are daunting, very likely to deter an embassy officer who might initially learn of a local refugee in danger, for whom U.S. resettlement seems otherwise appropriate. Creative thinking needs to go into mechanisms that could make it easier for embassy officers to watch for, and then act speedily on, the cases of qualifying refugees who come to their attention. With some simplification of existing processes, embassy officers would be in a better position to act on urgent cases informally brought to their attention by NGOs. Nonetheless, one should not expect more than modest increases in referrals from embassies, which are already stretched thin by their other responsibilities.

3. NGO referrals

Some people have proposed to expand possibilities for NGO referrals, looking to this mechanism in part as a potential source for a significantly increased volume of admissions. The International Rescue Committee (IRC) developed an NGO referral system several years ago that referred a modest number of individual Afghan refugee cases in Pakistan (mostly women at risk), and the Hebrew Immigrant Aid Society (HIAS) has such a program in East Africa. PRM also initiated a pilot program for enhanced NGO referrals with a two-day training session in Nairobi in early 2003, targeted at the staff of assistance-focused NGOs who are regularly present in refugee camps or settlements. These initiatives are worthwhile, and PRM should systematically evaluate the experience under each of them, in order to refine the system and carefully deploy it elsewhere. In particular, persons interviewed suggested that a less cumbersome procedure than the one set up as part of the Nairobi training would be more likely to produce better results. If the paperwork required is too extensive, busy NGO personnel in the camps will be deterred from triggering the process.

Although a refined and more widely deployed system for NGO referrals would be worthwhile, it is important to be realistic about what the program can expect to accomplish through this mechanism. In fact, it was not until after many months of interviews for this project that I came to appreciate that persons speaking about NGO referrals may have two quite different kinds of programs in mind.

One model, which was used at certain stages in the Indochina program, I was told, gave voluntary agency representatives a rather extensive and proactive role in going out to find refugees, but according to carefully defined criteria set out by the State Department. Such a model can generate significant numbers of refugee admissions, but it would appear to work well only in conditions where there are reasons not to worry greatly about a magnet effect or where the criteria employed can be readily applied based on accessible documents or other objective factors. Such an NGO role would probably have to be accomplished by special NGO staff (probably affiliated with an Overseas Processing Entity) designated specifically for admissions purposes; it is not at all realistic to add this function as a kind of sidelight to be carried out by NGO staff who work primarily on assistance. For the reasons set forth in Chapter I, however, today's refugee situations are unlikely to lend themselves to this kind of approach, because in most of them concerns about magnet effects, fraud, and exaggerated resettlement expectations loom large B for host governments and for many other players, often including NGOs and UNHCR. This first model, despite its important historical antecedents under somewhat different circumstances, should not be seen as a promising method for improving today's refugee admissions program. But it could remain on reserve in the program toolbox, to be used if circumstances generate a large-scale resettlement initiative for which such an approach is appropriate.

The second model is more in line with the IRC Afghan program, or the recent pilot project in Nairobi. Such programs are meant to draw, discreetly and carefully, on the contacts NGOs already have in refugee settlements in order to find particularly vulnerable cases who could then quietly be moved to safety in the United States. The premise is that the NGO's other work, not expressly resettlement-oriented, places its staff in a better position than UNHCR or US personnel to learn of such vulnerable persons. In order for this model to operate discreetly, it cannot involve any significant deployment of new personnel, and certainly not any who are expressly designated as resettlement staff. Many people interviewed for the project, including several NGO representatives, emphasized B vigorously B that NGO assistance staff in camps must not be put in a position where they become known as a kind of ticket to admission to the United States. In many first-asylum settings, hunger for resettlement possibilities is so great that knowledge of such a role might leave the NGO staff besieged with supplicants, interfering greatly with their primary operational work, and sometimes even generating security or crowd management problems. (The desire to avoid this risk, perhaps more than cumbersome paperwork, may account for the apparently low usage of the Nairobi pilot referral system.) For these reasons, referral programs of this type will at best involve modest numbers of refugees, although they can help the refugee program meet a highly important objective: identifying persons quietly facing acute dangers who might otherwise be overlooked.

Under today's conditions, the individual referral role for NGO staff is best done under the second model. It will be able to focus only on those in the most urgent need to escape from dangers in the camp or settlement, and the NGO function must remain discreet, almost covert. NGOs do have a role in helping to facilitate access decisions on a larger scale, but that should come through their involvement in the group designation process, as discussed in Chapter II and in the following Section of this Chapter.

Recommendation III-2: PRM should press UNHCR to continue expanding its individual referral capacity, possibly streamlining the referral process for U.S.-destined cases, and enhancing its mechanisms for quality control and consistency. Procedures for embassy referrals of P-1 cases should be simplified, and State Department training should better equip embassy personnel for this role. PRM should systematically evaluate the past experiences with NGO individual referral schemes, so that the process can be refined and made available elsewhere. Such referrals will ordinarily be done quietly by NGO personnel present in the camp or settlement for other reasons, and can be expected to produce only modest numbers, because they will focus on urgent, compelling cases.

D. Group access

Chapters I and II addressed many key questions surrounding the process of deciding on the groups to be granted priority access to the U.S. Refugee Program, recommending an attitudinal and institutional framework for improvements. An additional word about NGO roles, initially discussed in Chapter I, Section B1, is appropriate. NGOs perform a considerable service in recommending group initiatives to the Department of State, and they have provided highly valuable information sources for these purposes for several years. Publications like the World Refugee Survey, produced by the U.S. Committee for Refugees, as well as the human rights reports of some other organizations, provide well-organized accounts that can be of great assistance in identifying potential groups for access to the US program. The most focused NGO resource of this type, however, consists of the annual Recommendations volumes produced by the Refugee Council USA, the umbrella organization for the resettlement volags.²³ Each year they provide details, region by region, of needy refugee populations, and make specific recommendations for enhanced individual access or P-2 designation, as appropriate.

No official sanction, such as legislative authorization, is needed for this process. RCUSA and others should continue to provide this type of detailed non-official input, which should be of assistance in preparing and refining the list of potential groups for consideration by the recommended Refugee Admissions Committee. More visible demonstrations of a Departmental intention to examine these suggestions seriously, coupled with an effective and sustained increase in the use of the group-designation process B which is decidedly needed anyway (and is occurring in FY 2004) B should decrease the momentum for enshrining this initial NGO process in legislation. Further, RCUSA will maximize the impact of its recommendations to the extent that they openly deal, in close detail for each specific refugee situation, with the trade-offs, barriers, and obstacles that government players must confront, as sketched in Chapter I, Section A. (It could be expected that such considerations would sometimes result in the removal or postponement of specific resettlement proposals.)

Once a potential group has been chosen as a candidate group, NGOs can play a highly useful but focused and selective role as part of the investigatory process contemplated before a decision on final designation. PRM's current efforts to establish targeted response teams seem to afford a useful model B teams assigned to undertake a field visit to look in detail at particular populations who might be appropriate for a resettlement initiative. The teams are meant to be composed of governmental experts as well as NGO representatives who have particular knowledge of the targeted country or population, coupled with familiarity with the resettlement program, and who would be available for service, sometimes on fairly short notice.

Those teams should also include a representative of DHS. Until now, DHS (like its predecessor INS) has tended to want to leave the group designation process to the Department of State. It then takes up its role only after designation has occurred. I was told that this reticence traced in part to the diffuse process for group designation. DHS did not want to be fingered as the source of denial of a group designation, when all it might have done was to suggest certain problems or obstacles B objections which it often did not see as fatal to a proposed designation. But if the process becomes more clearly defined, as suggested in Chapter II, a DHS role should fit far more comfortably. The Department of State would take responsibility for developing the list of potential groups to bring to the Admissions Committee. Only after a formal decision to name a candidate group would DHS be asked to play a role in the future investigation and development of possible plans for resettlement.

Such a DHS role at the stage of investigating a candidate group is needed to assure ample consideration of operational issues and means of addressing them, before a final designation is made. Under past practice, I heard of several instances where rather significant problems affecting DHS's role had not been spotted until the circuit ride team arrived to begin interviews B resulting in damaging inefficiencies. Sometimes these had to do with operational issues, like the quality and identity of translators, or the adequacy of arrangements for site security. Other times the overlooked issues were legal questions. For example, a common issue for Colombians has been whether payment of ransom to retrieve a relative from paramilitary kidnappers should be considered "material support" to terrorists, which would render the person inadmissible to the United States. And with regard to Liberians, an issue arose over whether those who had married Ivorians were "firmly resettled" in Cote d'Ivoire, and therefore ineligible for the refugee program. Not all such operational and legal questions can be spotted in advance, and some capacity to respond quickly in the midst of a DHS circuit ride will always be required. But any gain in advance preparation can make for better use of resettlement resources.

Moreover, DHS involvement at this stage can help provide a better estimate of the approval rates likely once interviewing begins B which can be useful both to assure that all players share realistic expectations about the program and to allow better planning for the later stages of the process. Approval rates have proven a sore point in interagency dealings over the past few years, but my interviews suggest that this usually resulted from an inadequate understanding of the roles and perspectives among the various players in the process. (See Chapter V, Section A4.) Including DHS on the investigatory team considering a candidate group will go a long way toward providing better understanding of this factor, and also toward educating PRM and NGO staff about the realities and constraints of DHS's role. This process can also bring advantages in the other direction. DHS will thereby learn in more detail about the perspectives of PRM and the NGOs regarding the group, and the DHS member may be better able to convey the real level of need presented by the refugee situation back to all the potentially involved bureaus in DHS headquarters B thereby expanding support for efforts to facilitate the process at every stage where DHS in involved. DHS participation in the field investigatory and preparation process should help assure a full buy-in and commitment by DHS when the decision is made to move ahead with admissions.

Recommendation III-3: The Department of State should give close attention to NGO suggestions, including the annual RCUSA Recommendations report, when developing the potential group list. Concomitantly, NGOs should work to address more concretely the trade-offs, barriers, and obstacles that would affect any resettlement initiative B and that sometimes counsel against undertaking it. No legislative sanction is needed or desirable with regard to that sort of input. Once a group has been chosen as a candidate group, further investigation of group needs and characteristics should take place. It will often prove advantageous to proceed through a field mission by a targeted response team, which should ordinarily include NGO representation and one or more participants from the Department of Homeland Security.

E. Family-based priorities, including issues of processing bottlenecks and of fraud

As noted in Section A, family reunification has been a constant and enduring value in modern U.S. immigration law and policy, and U.S. law provides that the spouse and minor unmarried children of a refugee, if accompanying or following to join that person, may be admitted without having to meet the refugee definition independently.²⁴ This applies to refugees of all nationalities, and does not require further specific authorization through the priority system. Thus the nuclear family may be kept together even if only one parent has become the target of persecution. This provision becomes somewhat more problematic in "following to join" cases, wherein the principal refugee files only after admission to bring the spouse and minor children. The reasons derive largely from problems with the process, known as Visas 93, used to handle "following to join" cases.

If that process worked smoothly, one would probably see far fewer applicants trying to use the P-3 process, because a major part of the eligible P-3 pool overlaps with the class of persons who could use Visas 93. In fact, because P-3 applicants must independently satisfy the refugee definition individually, at first glance there would appear to be significant incentives to use Visas 93. But that has not been the experience. For any nationality to which P-3 applies, the P-3 process has been favored over Visas 93. In any event, a well-designed P-3 process remains a real need, because some of those eligible could not use Visas 93 B primarily parents of the anchor relative. In addition, there is no derivative status to a Visas 93 admission. To appreciate the significance of this limitation, consider, for example, an unmarried teenage daughter of a previously admitted refugee who gives birth (perhaps as a result of a sexual assault) before she can travel as a Visas 93 beneficiary. Because she is a derivative relative, rather than someone qualifying for refugee status in her own right, the law does not permit her to bring her child as part of this process.²⁵ The P-3 category provides a possible means to keep such a mother and child together, although she would then have to show independently that she meets the refugee definition. (And when P-3 is not available, PRM and DHS normally find another way to assure that the family stays together.)

The P-3 program has encountered a significant amount of fraud. Although new measures in place for the past couple of years provide improved means to detect and deter fraud, the prior experience soured many government players on the whole category. The following Sections explore these issues, leading toward recommended improvements for both the Visas 93 and P-3 processes, so that both can be used more readily. It also evaluates a frequently voiced proposal to adopt a universal P-3 category B that is, to open the P-3 category to persons of any nationality, and not limit it to a specified list of countries of origin.

1. The Visas-93 process for immediate family following to join

A person already admitted to the United States as a refugee under INA '207 who wishes to bring in his or her spouse and minor unmarried children begins the process by filing a Form I-730 with the Nebraska Service Center of DHS, accompanied by proof of the family relationship. Once the I-730 is conditionally approved, it is sent via the National Visa Center to the overseas post where the family members are expected to complete the rest of the process. DHS officers often tend to refer to these as I-730 cases, because the Form I-730 provides their main initial contact with the such applications. But the State Department refers to them as Visas 93 cases, a sometimes confusing nomenclature.²⁶

In a few locations, Visas 93 work is handled by those involved in the general refugee processing procedures, including preparation of the case by an Overseas Processing Entity (OPE) and often adjudication by a DHS officer. As will be discussed in Chapter V, such case preparation is the bread-and-butter work of OPEs, and they are expert in the overall refugee admissions process. Such expertise helps the Visas 93 process work efficiently in those locations. But in posts where such processing is not employed, Visas 93 processing often does not proceed smoothly. These types of cases are a relatively exotic form of application, and many consular officers, as well as the local-hire staff who often play a key role in consular processing, are not familiar with the specialized steps needed to complete them. The cases are dauntingly labor-intensive, meaning that the consular officer may have to postpone work on several of the more familiar types of visa applications in order to process one Visas 93 case. As a result, Visas 93 applications may simply be laid aside, languishing for many months on a desk in the consulate. Moreover, consular officers may have difficulty arranging for the needed interview with the family members, because the family members living in a refugee camp may not be able, at least not without special clearances, to travel to the consular post. Delays and uncertainty over the Visas 93 process have often led volags to counsel their clients trying to bring in spouse and minor children to use the P-3 process, if it is available for their nationality.

PRM has taken steps to address these problems with Visas 93, including assigning an officer in headquarters to spend a major portion of her time troubleshooting, helping to get delayed cases unstuck, and walking consular officers through the process. The staff of the Refugee Processing Center in Arlington (a contract unit that manages the basic data processing system for the refugee program), also spends a fair amount of time assisting consular officers or local-hire staff to work through these cases. PRM has also worked for more coverage of Visas 93 in the consular training program. But these cases remain a problem, and more should be done. At the very least, Consular Affairs should incorporate Visas 93 training more thoroughly into the normal consular training course, and the Foreign Affairs Manual section dealing with these cases should be updated.²⁷ It would also be good to route approved I-730s through the Refugee Processing Center (or at least assure inclusion of full tracking data in the WRAPS database system), so that PRM could better monitor such cases and act promptly when processing bogs down in particular locations or for particular cases.

One source of confusion could be reduced if DHS would agree to change its I-730 form, so as to signal better to the consular officer who will eventually process the case the distinctive requirements in Visas 93 cases. This particular confusion arises because the Form I-730 is used for two related, but in the end importantly different, types of cases B to bring in both the family members of persons who gained status in the U.S. as a result of an asylum claim under INA ' 208 and the family members of persons admitted as refugees under INA ' 207. The State Department labels the former

'Visas 92" cases and the latter "Visas 93." Visas 93 cases count fully as refugee admissions in the annual totals and therefore occupy one of the admission spaces made available by the Presidential Determination. Visas 92 family members (spouses and children of asylees) do not. Most importantly, Visas 93 cases benefit from the special treatment and subsidies that the U.S. system provides for refugees, while Visas 92 cases do not.

Visas 92 cases therefore look a lot more like the normal kind of case a consular officer might handle. The individual applicants are responsible for completing the medical examination at their own expense, making their own travel arrangements, and paying for their air tickets. But in Visas 93 cases, the U.S. government pays for the medical examination and obtains the ticket through the International Organization for Migration (IOM). The government advances the cost of the ticket, but in the form of a loan that must be repaid over time after resettlement in the United States. Hence in Visas 93 cases, the consular officer must take additional steps quite different from all other consular cases B arranging for the medical exam through or in cooperation with IOM, obtaining the ticket, getting the individual to sign the loan note, and also initiating the process that will result in the required sponsorship assurance from one of the resettlement volags in the United States. (Chapter V explains these distinctive processes in greater detail.)

Using wholly different DHS forms for the two types of cases would therefore be beneficial. Alternatively, DHS might at least issue a revision to Form I-730 that would signal to the consular officer more prominently just which of the two types of cases is involved. Furthermore, for Visas 93 cases, a larger "official use only" box could provide spaces for the consular officer to check off completion of the distinctive steps that are to be completed at government initiative (and usually expense): medical exam arrangement, sponsor assurance, IOM contact for the ticket, and the loan note. Although these would be unusual changes to make on a DHS form, they are worthwhile in this setting, precisely because the Visas 93 procedures have proven to be so daunting for consular posts that see them infrequently. The high importance of reuniting a resettled refugee with his or her spouse and minor children would justify such relatively unusual steps.

A more ambitious change to the process, suggested to me during the interviews for this project, might provide a more broadly applicable and reliable fix. The basic problem is that Visas 93 cases are shoe-horned into the consular caseload, whereas most of the process follows instead the distinctive procedural contours that have evolved for refugees B quite different from the normal consular routine. It would be better to use the refugee-specific procedures, to the greatest extent possible, for the Visas 93 cases. This would mean processing by an OPE and adjudication (ordinarily) by a DHS officer (with the concomitant changes in budgeting). Of course, Visas 93 cases could crop up anywhere in the world, and are not necessarily confined to the major refugee processing sites. Thus some additional circuit riding, for both OPE staff and DHS, would be required B often to service a rather small caseload. Some allowance would probably still have to be made for consular processing where a caseload is so small as to make OPE and DHS travel there impractical.

Recommendation III-4: The Visas 93 process needs to work reliably and efficiently, because it serves a vital function: reuniting a resettled refugee with his or her spouse and minor children. At the least, Visas 93 training should be a regular part of consular training, and the State Department should adopt additional procedures (possibly including a greater monitoring role for WRAPS) to assure that such cases do not languish. DHS should also either revise the Form I-730 or develop two separate forms for Visas 92 and Visas 93 cases, so that the form will better guide the consular officer through all the distinctive steps needed for Visas 93 cases, owing to their inclusion in the special benefits of the overseas refugee program. DHS and PRM should also consider arrangements that could take most Visas 93 work from consular officers and have it handled through standard OPE procedures, with ultimate adjudication by DHS.

2. P-3 processing

The P-3 category is available for spouses, minor unmarried children, and parents of persons resident in the United States. As noted, it is an access category, not a derivative benefit. This means that such persons can get access to a DHS interview, but must independently satisfy the refugee definition.²⁸ Historically the process began with the filing with a local volag office of an affidavit of relationship (AOR) by the already admitted anchor relative. The volag office then typically forwarded the affidavit on to the overseas processing location near the site where the family members were located. For many years different volags used their own diverse affidavit forms, and efforts on the part of the volags to verify the family information set forth in the affidavit were uneven. As the volume of P-3 filings grew, government officers involved in the process began to suspect a significant amount of fraud. (The extent of fraud varied by refugee situation, and people interviewed for this project speculated that the variances trace to some extent to cultural differences and perhaps more to whether or not good civil records systems existed, recording births and marriages for the refugee population at issue.) Access to US resettlement is a highly prized commodity. It is likely that some of those committing fraud used the P-3 process to help more distant relatives or acquaintances who may have been quite needy in their own right B even if not technically eligible for P-3. But the scale of the fraud appeared to go beyond that, and some of the misrepresentation was much more cynical, involving the buying and selling of access. The requirement that all P-3s meet the refugee definition might help screen out non-needy impostors, persons not at risk of persecution, but that process too has hardly been foolproof. Some INS officers pushed for reforms for many years, suggesting especially that new AORs be checked against family tree information gathered by the processing team around the time of the principal alien's interview and now contained in his or her INS file, known as the A-file.

Those suggestions did not get far²⁹ until the September 11, 2001 attacks brought heightened attention to all the vulnerabilities of the refugee admission system. At that time, INS set up a Refugee Access Verification Unit (RAVU) to manage a verification process that would apply to all future P-3 admissions. This screening is separate from, and in addition to, added screening for national security purposes. It is not directly targeted at the kinds of terrorist dangers that September 11 revealed. Instead, September 11 merely provided the impetus to tighten up many parts of the admissions system, and it invited close attention to long-pending suggestions for building better means to respond to fraud in the P-3 process.

RAVU verification involves obtaining the A-file of the anchor relative and checking the currently claimed family relationship against the family information submitted at the time of the initial application. If discrepancies appear, denial or revocation of approval is likely. But the process generally allows the anchor relative an opportunity to submit additional information, often including DNA testing at his or her expense, to explain the variances and offer more solid proof of the relationship.

All P-3 cases that had not yet traveled to the United States as of September 11, 2001, were subjected to the new verification process. This meant that a significant number of persons who thought that they had been approved for admission saw their cases reopened and their approvals suspended. Notifications of discrepancies were sometimes delayed, as were notifications of final revocation of admission approval. Therefore many refugees who thought they had been fully approved for resettlement in the United States found themselves in limbo for months, and a great many for years. These limbo cases have received a good deal of media attention, often confusing the issue and blaming the delays and disappointments solely on security screening rather than anti-fraud review. The inordinately lengthy delays affecting these persons once approved for resettlement have attained special prominence in the strong criticism the program has incurred since September 11. Final resolution of these cases and speedy notification of the results must be given a high priority.³⁰

Despite these rocky patches in the initial implementation of RAVU, the basic process is sound and much-needed. Fraud has been a genuine problem that needs to be addressed systematically. NGOs agree fully with this position, although many suggest that the prevalence of fraud often reflects the depth of need in the refugee camp. They urge that tightening up on fraud be matched with other efforts to expand access to the refugee program. That idea is fully consistent with the general approach recommended in this report, but one must be realistic about the impact of expanding other forms of access. Even in the best of scenarios, there will still be enormous demand or desire for US resettlement that will go unmet. Temptations to commit fraud will still be substantial. Hence the program must definitely continue to make use of available tools to deter and detect falsehood. In the absence of good civil records, which is the case with many refugee situations, the A-file information (i.e., the family tree information developed when the principal alien applied to the program) may provide the best possible documentary handle on the issue. Of course, there can be innocent explanations for discrepancies, particularly for applicants from societies that are less oriented toward linking important events to clock and calendar. The system needs to make allowances for these problems, and provide genuine opportunities to explain family tree variances.

Other anti-fraud steps, some initiated before the September 11 attacks, have also been implemented in connection with the P-3 category. They include the development of a standard AOR form and instructions that must be used by all volags. The instructions now include a prominent warning about the penalties for submitting false information as part of the process. AORs may no longer be sent directly to the overseas processing location from a local volag office, but instead must be sent to the national volag, which is called upon to review the information for quality control and to report anomalies to DHS. The RAVU process is now done up-front, before a P-3 case is cleared for interview by a DHS circuit-ride team. Thus, in marginal or uncertain cases, DHS interviewers can be equipped with the RAVU information to use in exploring family relationships during the circuit-ride interview.

The NGOs have been supportive of most of these steps, reserving, of course, the right to criticize particular aspects of implementation. The most successful of the post-September-11 PRM-DHS-NGO working groups was the one dealing with fraud issues. It came up with 19 specific action suggestions, transmitted in the spring of 2003, many of which are in the process of implementation.³¹ DHS also has its own "Operation Fair Refuge," with an action plan containing additional steps in the process of exploration or implementation. Continued vigilance and continued evolution of anti-fraud measures will be necessary, because entrepreneurs in the camps will continue to seek ways to defeat current measures. Situations have already been discovered where persons have planted dummy family members when reporting family trees during their own refugee interviews, so as to lay the groundwork for a later fraudulent family admission that would not be detected using current RAVU techniques. I was told that in a few instances, RAVU-cleared families were found to be composed entirely of unrelated individuals. As a general matter, however, DHS's anti-fraud framework is now sufficiently institutionalized and sufficiently aware of the dynamic nature of anti-fraud efforts that it should be able to find reasonable counters to new fraudulent techniques. DHS is also considering making the AOR, with revisions, into a full-fledged government form. Such a step would be worthwhile, because it would probably strengthen the ability to apply criminal punishment to those who commit fraud in such a filing (although one should not expect a dramatic impact).

The ultimate safeguard in checking most claimed family relationships would of course be DNA testing.³² It is currently in use in a limited fashion, largely in connection with anchor relatives' efforts to overcome a RAVU denial based on discrepancies in family information.³³ But it might be possible to use it more widely, as a far more effective deterrent and obstacle to fraud. One well-thought out and detailed proposal that I received during the interviews would call for DNA testing for all refugee cases (save solo applicants) as a routine part of the medical examination required before travel. Once the mandatory use of such a procedure became known, the proponents explained, it would put certain kinds of "refugee brokers" and fraud rings out of business, and would end the nascent practice of claiming false relatives during the initial family tree interview so as to defeat RAVU screening. The proponents of this proposal also pointed out that it would carry genuine protection benefits for large numbers of innocent refugees, because it would largely shield them from intimidation by criminal syndicates pressing them to add impostors to their applications.

The major drawback to this proposal, of course, is cost. DNA testing currently costs about $300 per case, although the price has declined steadily over recent years.³⁴ Proponents therefore suggested alternatives to across-the-board DNA testing. For example, DNA testing could be employed on a random sample basis for a stated percentage of cases. Particularly if the sampled percentage is reasonably high, this prospect might increase the risk of detection enough to deter some organized fraud B although to a lesser extent than the main proposal. (Some were skeptical that it would deter the worst refugee brokers, who could still claim that most of their clients would get through.) Alternatively, if measures could be developed for adequate preservation, DNA samples could be taken routinely as part of every medical examination but tested only upon a later determination that such a step is necessary or advisable. (Taking a sample is a simple and inexpensive procedure, whereas the testing requires more expensive laboratory work.) Publicizing this step and what it means as a possible benchmark for judging later family filings could also help deter family fraud. Or across-the-board testing could be triggered at a particular site or for a particular population when other objective indicators of fraud, such as the RAVU rejection rate, reach a pre-set benchmark level.

These proposals for wider use of DNA testing carry considerable merit, and should be pilot-tested in selected locations. There will probably come a point when DNA testing costs decline sufficiently that it will make sense to include such testing in the medical exam for all family cases. Some NGOs have expressed skepticism about wider use of DNA procedures. But if such a change could greatly reduce the concern about fraud in refugee family cases, it might well make possible far wider use of the P-3 category. Nonetheless, there remain some important ethical and logistical issues to address before launching wide-scale DNA testing. Some test results could have a traumatic or tragic impact on the persons involved, such as a revelation that a child raised since infancy as part of the family unit is actually not related. Or a DNA test that reveals that a child was actually the product of a previously unknown adulterous relationship could trigger spouse or child abuse. An IOM paper helpfully considers the various possible roles of DNA testing in family reunification cases, including discussion of logistical and ethical constraints and cautions.³⁵

Recommendation III-5: Review of family cases by DHS's Refugee Access Verification Unit (RAVU), which includes checking asserted family relationships against earlier family information in the anchor relative's A-file, has been a highly worthwhile development. RAVU also provides an institutional location for ongoing innovations to respond to what will inevitably be new schemes and patterns of fraud. DHS should continue to work cooperatively with PRM and the NGOs in developing such refinements, and PRM should make the affidavit of relationship, with revisions, into an official government form. DHS and PRM should also pilot-test wider use of DNA testing, and should closely consider the costs and benefits of routine DNA testing as part of the medical examination all refugees must pass, particularly as testing costs decline. Ongoing improvements in the safeguards against fraud should permit an expansion in the availability of P-3 admissions.

3. A universal P-3 category?

Many in the NGO community have long supported a universal P-3 category B that is, making P-3 access available to family members of US residents who prove that they are refugees, no matter what their nationality. These calls became more urgent after September 11, 2001, as usage of refugee admission numbers declined steeply. Why not, the proponents have asked, use those admission slots more aggressively for family members, at least until P-1 and P-2 admissions rebound? Instead, these proponents noted, the trend has run in the other direction. Eighteen nationalities qualified for P-3 in FY 1999, but the list declined to four in FY 2003. (For 2004, however, the list has rebounded to nine nationalities.)

These are worthy questions, but a bit of perspective is helpful. Although it would seem shocking to allow only a limited group of refugees to reunite with their families, and to parcel out those numbers based on nationality, in fact the situation is not nearly so stark. We do in fact have a highly important universal family reunification provision for the spouses and unmarried minor children of refugees B the Visas 93 program. This is not limited by nationality and is available to the closest family members of all refugees. The call for a universal P-3, therefore, is a call for universal access by refugee parents of U.S. residents, who are not eligible for Visas 93. (And perhaps the call also reflects a strategy that might allow a higher number of spouses and minor children to avoid the processing problems that have often beset the Visas 93 program.) Reunion of adult U.S. residents with refugee parents can be important, but it is less compelling than the forms of family reunification covered by Visas 93. Some government officials I interviewed also thought that the admission of parents was more vulnerable to fraud than some other categories B not so much because of fraud regarding the parent (for whom DNA testing could potentially identify impostors), but because parents often have remarried by the time of the interview and have a whole new step-family to bring in as derivative relatives B a step family that lacks a blood relation, and perhaps even an acquaintance, with the anchor relative who filed the AOR. Opportunistic marriages could result, perhaps at the instance of refugee brokers in refugee settlements.

PRM tries to balance multiple objectives in setting the P-3 nationality list. The admissions office of course understands the value of family reunification, but on the other side lie concerns about fraud and also B less widely recognized B the desire to assure that qualified P-3 applicants can receive timely processing in a program whose reach is manageable. DHS circuit rides, with the necessary preparation by an Overseas Processing Entity, are costly to schedule and perhaps not worthwhile if only a handful of cases await processing in a particular location. One way to identify nationalities likely to yield significant numbers of bona fide P-3 applicants is by consulting the UNHCR document on resettlement need for the coming year. For the last several years, therefore, PRM has determined the composition of the P-3 nationality list based largely on those nationalities with the greatest need for resettlement overall, as identified in this UNHCR document.³⁶ No set projection of numbers from UNHCR provides a minimum cut-off, but in general nationalities would not be considered for inclusion absent identified resettlement need numbering in the hundreds. PRM in the past has also experimented with different formulae. The FY 1999 list of 18 nationalities derived from an effort to incorporate a complicated multi-factor matrix in which UNHCR projections played a more limited role.³⁷ Unfortunately, that expansion triggered such a large number of new filings, including a high percentage of claims that proved fraudulent but could only be weeded out by a time-consuming process, that PRM sought ways to create more restricted lists in following years.³⁸

Shortening the nationality list may have been an understandable step to take before other and more targeted anti-fraud measures were in place. But it is much better to target fraud directly, rather than simply disqualifying large categories that may contain hundreds of genuine applicants. The system has now equipped itself far better for that task, largely but not exclusively through the RAVU process B not only its current review procedures but also its role as an institutional home for ongoing anti-fraud innovations. These improvements should be seen as setting the stage for expanding the P-3 nationality list, at least until such time as the other priorities are positioned to fill more of the admissions slots made available in the annual Presidential Determination.

In fact, it may be worthwhile, at least on a test basis for a year or two, to implement a universal P-3 category, if only to provide solid data about just what such a system might accomplish and about what logistical difficulties it would spawn. Several NGO personnel with whom I spoke were convinced that such a step would generate thousands of new, valid refugee admissions. PRM personnel and some others were far more skeptical. They believe that the current formulas, although they could be tweaked to add a few more nationalities to the P-3 list, already result in P-3 processing in the main locations where significant numbers of family members could be expected to apply. New admissions, they thought, might only number a few hundred, and at a high logistical cost. A universal P-3 could spark unrealistic expectations that would go unmet because processing teams could not realistically be deployed to low-density locations.

Several NGO representatives with whom I spoke, sensitive to concerns about fraud, expressed support for a kind of compromise on these issues, one that would still allow a cautious form of a universal P-3. They referred to a processing change introduced in FY 2004, of which many NGOs have otherwise been critical. This is the restriction on the categories of persons who may apply for family members through P-3: for FY 2004, only persons initially admitted to the United States as refugees or asylees can file an AOR. (Before that, virtually any lawful US resident could file.) The rationale is that better family tree information exists in these anchor relatives' files, which can serve as a useful check against false filings. Although these NGO representatives would have preferred a universal P-3 category using the old filing rules, they would find the new restrictions on those who can file far more acceptable if refugees of all nationalities became eligible for admission.

When I discussed this possibility later with some PRM officers, they expressed a wary interest, but they still worried about stale cases and about processing capacities if there is such expansion. As we talked through these problems and possible remedies for them, I concluded that a viable pilot test of a universal P-3 category might best be constructed on the basis of the NGO representatives' tentatively voiced compromise, but with two additional constraints. First, to avoid stale cases, there should be a time limit between the refugee or asylee anchor relative's admission and the time when the AOR is filed. For an initial pilot trial, this period might appropriately be longer than what one would expect if a universal P-3 category became a regular part of refugee admissions.³⁹ After all, in the first year, filing would be possible for several persons whose family members (or at least whose parents) would not have had any earlier chance to qualify, because their nationality was not on an earlier list. Hence the lateness of the filing would be based on prior ineligibility, not lack of interest. Thus the pilot test might allow filing within, say, three or four years of the principal relative's admission. Second, publicity for the expanded program should make it clear that some applicants in low-volume locations may find that their cases cannot be scheduled for interview at all, or that they might have to wait several years until enough cases accumulate at that location to make the dispatch of processing teams worthwhile.

That proposal, a prudently constrained version of a universal P-3, merits serious consideration. Trying a universal P-3 on these terms for two or three years would enable a full assessment of what the impact might be, both on processing and on numbers. It should afford as well an opportunity to measure the risk of fraud through false step-families or other devices. If the risks are unacceptably high or the logistical problems significant B or if P-1 and P-2 numbers independently expand to the point of using a far higher percentage of the annual refugee admission spaces B then the P-3 category could later be scaled back.

Recommendation III-6: Now that anti-fraud capacity has significantly improved, PRM should continue expanding the list of nationalities for whom P-3 access is available. It should also give serious consideration to implementing a carefully designed universal P-3 category for a few years on a trial basis, at least until the other priorities come closer to using all available admission spaces. This universal program, however, should use the AOR filing restrictions implemented in FY 2004 (permitting filings only by those persons admitted to the United States as refugees or asylees). Additionally, it should allow filings only within a stated number of years after the principal relative's admission, and should warn explicitly that cases in low-volume locations may not be processed, or may have to wait many months or years for processing. Such a pilot test should be carefully monitored for problems with either processing or fraud.

4. Functional family relationships

Interviews for this project produced many comments that the refugee program should make better provision for functional family relationships in the chaotic aftermath of refugee flight. For example, in a conflict like that in Liberia, orphaned children may be picked up and cared for by distant relatives or by neighbors. This relationship may be established in the country of origin before the flight or only in the refugee settlement. Sometimes it leads to very close ties, functionally the same as family affinities, between the children and their caretakers. In many settings the caretakers refer to themselves as foster parents, but rarely are such relationships formalized through legal procedures. In this context, it can be traumatic for all concerned if only a part of the functional family unit is approved for resettlement in the United States. Some therefore suggested that the concept of family to be used in the refugee program should be broader. Foster children should be treated like blood-children so as to keep the functional family unit together during resettlement.

The impulse behind this suggestion is unassailable. Resettlement should not add family separation trauma for those who have already suffered such separation once before, through the death of their actual parents or children. The system currently takes account of such a situation through the case composition rules that go some distance toward accommodating this need. They provide:

For humanitarian reasons, other family members [besides the spouse or unmarried minor children] may be presented to DHS as part of the same case if they resided together prior to flight, continue to reside in the same household, and are part of the same economic unit as the PA [principal applicant]. Such add-on family members may be assigned the same priority as the PA but must establish their own individual claims to refugee status.⁴⁰

These guidelines cover many of the functional family situations of concern here, especially since I was told that "other family members" for these purposes can include in practice so-called foster children who lack blood ties, as long as the other qualifications are met.

But because the rules invariably require residence together both before and after the flight, they will miss the situation of those children taken in to another family only in the refugee settlement. Some have urged that co-residence or dependency, such that the members all form part of the same economic unit, should be sufficient even if it occurs only in the refugee camp.⁴¹ When I inquired about making such a change, DHS officials explained that such a rule would be too open to fraud. Requiring that the persons in question have shared life together in both settings provides opportunities for questioning about both locations that is more likely expose outright impostors. If impostors only had to present consistent stories regarding camp life, it would be easier to carry off the fraud. These officers candidly acknowledged that the existing rule will unfortunately bar from being considered as part of the same case some persons who have formed genuinely close family-type affections through perhaps many years of life together in the refugee camp. They pointed out, however, that the others might be able to gain access to the program anyway in their own right, particularly when group designations are the source of access. If that occurs, case allocation procedures covering ultimate sponsorships in the United States can often be used to make sure that the functional family members wind up in the same destination community to restore the family unit, and may even be used to assure that they travel together.

This concern is real and worthwhile, but I have not found a solution that is superior to the current arrangements without opening up unacceptable chances for fraud. PRM and DHS should nonetheless keep this issue active and continue searching for better methods that might enable keeping functional family units of this type together.

F. Urgent cases

1. General considerations

The popular image of refugee rescue often includes spur-of-the-moment efforts to pluck people from danger and move them swiftly to the state of refuge. The reality of most resettlement is quite different B understandably, in view of the complexity of the system and the many moving parts that must be put in motion and kept in synchrony. Nonetheless, the system can sometimes gear up for extremely urgent action for large groups, as happened with the Kosovo crisis of 1999.⁴² Such a mobilization of course requires a very high degree of political commitment, running all the way to the very top levels of the U.S. government, and so will perforce be infrequent.

But small-scale urgent cases crop up more often, involving an individual, a couple, or an extended family placed in immediate and life-threatening danger. The volume still is quite low, but the need can be acute. Part of the impulse behind the creation of the current P-1 category was to allow better for timely action to deal with these kinds of urgent individual cases. Unfortunately, as the program has developed, and particularly with the additional screening added after September 11, 2001, it is very hard to get the U.S. Refugee Program to act with the speed needed for these isolated individual cases. UNHCR officials reported numerous occasions where cases of this type were referred to the United States but where the process took so long that UNHCR turned instead to other resettlement countries that are better equipped to act speedily, such as Netherlands or the Nordic countries, or occasionally Canada.

I discussed this issue during several interviews. One opinion voiced during the conversations was that perhaps this is not a wholly bad situation B but instead marks out a healthy division of labor. The United States is quite good at running a high-volume program, a kind of aircraft carrier among the fleet of resettlement nations. But this means it cannot turn quickly to deal with smaller scale urgent issues. Nations with smaller programs, especially ones that can accept cases based only on a review of a UNHCR-prepared dossier, are the PT-boats of the resettlement world, and they should be deployed for these purposes. Others acknowledged the difficulties the United States faces, but insisted that we should restore the capacity to move very quickly in a limited number of urgent cases B and that we should especially have this capacity for rescuing persons who have been placed in immediate danger precisely because of actions they took in support of U.S. objectives or policies. They pointed out that even after September 11, we do sometimes move quickly for such purposes, as in the widely reported case of Mohammad al Rehaief, the lawyer who was protected by the United States because of his role in assisting U.S. forces to find the wounded U.S. soldier Jessica Lynch in the early days of the Iraq conflict.⁴³

Today such cases are usually handled through the mechanism of parole, which is discretionary permission to establish physical presence in the United States, but without a formal admission.⁴⁴ Parole permits such presence without regard to the usual screening and inadmissibility requirements. And in principle, because parole does not count as an admission, the individual can be removed more easily if disqualifying acts or traits later come to light that were missed in the foreshortened review that preceded his or her evacuation. In practice, however, it is very hard B nearly impossible B to remove someone paroled in these circumstances (when there are indications of threatened severe harm in the home country), even if derogatory information is discovered later. For this reason, pre-parole review or screening, even if compressed or unorthodox, has remained quite important. Parole also has several drawbacks for the individual. Most importantly, it gives the person no clear immigration status in this country, and in most cases no direct avenue toward qualifying for lawful permanent residence. (Parolees in these circumstances may well qualify for asylum upon application filed after their arrival here, but success is not guaranteed.)

In reality, people paroled in these urgent circumstances are being brought to the United States as resettled refugees. It would make abundant good sense to treat them, as much as possible, within the framework of the refugee program, with its established immigration status, provisions for both private and public assistance, and direct avenue toward permanent resident status. What I learned during interviews about the steps needed to achieve governmental approval of urgent paroles led me to believe that it would be no more difficult to put together a similar set of waivers and urgent decisions in order that such persons could qualify as refugees under INA ' 207. The program at various times has set up expedited procedures for particularly urgent cases. These should be revisited, updated, and revised for use in the post-September 11 environment. They will still be invoked only in a small number of cases, probably below 100 per year.

Recommendation III-7: PRM and DHS should work together to restore the capacity to act in a matter of days or weeks to approve and resettle as ' 207 refugees persons who are in grave and immediate danger and whose cases are referred by UNHCR or a U.S. embassy. This procedure for urgent action cases should replace the use of parole to the greatest extent possible. Such cases will be exceptional and the volume of such cases can be expected to be quite low, thus making such special arrangements feasible.

2. Proposals for a universal in-country designation

As it happens, some urgent cases have to do with persons still within their countries of origin. Such persons cannot meet the Convention refugee definition, which requires that a person be outside the country of origin before qualifying as a refugee, and so they cannot be referred by UNHCR. But their need may be no less acute. If such cases arise in one of the three countries normally listed in the Presidential Determination as approved for in-country processing, then inclusion in the refugee program is possible. Some have recently suggested a new approach to in-country designation in order to make provision for this kind of urgent case on a wider scale. They propose that the President designate all countries for in-country processing, although specifying in some fashion more limited criteria to make it clear that this will be invoked only in exceptional circumstances. Others have argued that such a universal in-country designation is not legally authorized.

In my view, such a designation could be done in a fashion that is consistent with the statute. INA ' 101(a)(42)(B)⁴⁵ authorizes in-country refugee determinations "in such special circumstances as the President after appropriate consultation [with Congress] may specify." To date, the "special circumstances" specifications have been done by means of limiting this access to a short list of countries, although the PD typically does further state that such admissions shall be available only for those "otherwise qualified." Other guidance then spells out more precise criteria that govern access to the in-country program. But nothing in the statute requires geographic limitations as the way of honoring the "special circumstances" requirement. Nonetheless, the "special circumstances" requirement must be honored in some fashion. A Presidential specification that covers all countries would have to limit the reach of in-country processing in some other significant fashion. If all that is contemplated is a handful of in-country urgent cases, then it should be quite possible to spell out limiting criteria in the Presidential Determination, in a manner that would be legally sufficient.⁴⁶ Whether such a universal specification would constitute wise policy is a closer question. Fuller exploration of the policy disadvantages is necessary before taking that step.

Recommendation III-8: The President has the legal authority to designate all countries for in-country processing, provided that other precise limitations confine its effect, thus honoring the "special circumstances" requirement of the statute. Such a designation would hold advantages for a handful of urgent cases each year, involving the rescue of individuals from immediately dangerous circumstances in their country of nationality. But a full exploration of possible drawbacks should be undertaken before deciding on such a step.

Footnotes:

_______________

¹ INA ' 207(c)(1), (3), 8 U.S.C. ' 1157 (c)(1), (3) (2000).

² See United States Department of State, Proposed Refugee Admissions for Fiscal Year 1994: Report to the Congress 18-19 (Sept. 1993).

³ United States Department of State, Proposed Refugee Admissions for Fiscal Year 1995: Report to the Congress 19-21 (Sept. 1994). What we now know as P-2 (groups of special concern to the United States) was initially set forth, in the 1994 announcement, as a subset of P-1. In order to provide greater clarity on the differences between the system for UNHCR or embassy referrals and the group identification process, the priority description was modestly reorganized in 1995 (effective in FY 1996), by splitting off current P-2 from P-1 and renumbering the remaining categories accordingly. Department of State, Department of Justice, and Department of Health and Human Services, Report to the Congress on Proposed Refugee Admissions for Fiscal Year 1996, at 18-19 (July 1995). At that time the priority system took essentially its current shape, although of course the nationalities or precise groups offered access under each numbered priority (save P-1, which is not limited by nationality) have changed over time.

⁴ See United Nations High Commissioner for Refugees, Resettlement Handbook: Division of International Protection, Chapter 4 (rev. ed. 2002).

⁵ Department of State, Department of Justice, and Department of Health and Human Services, Proposed Refugee Admissions for Fiscal Year 1999: Report to the Congress 8 (June 1998).

⁶ Department of State, Department of Homeland Security, Department of Health and Human Services, Proposed Refugee Admissions for Fiscal Year 2004: Report to the Congress 7 (Sept. 2003).

⁷ See Statute of the Office of the United Nations High Commissioner for Refugees, UNGA Res. 428 (V) (Annex), para. 6 (Dec. 14, 1950).

⁸INA ' 101(a)(42)(B), 8 U.S.C. ' 1101(a)(42)(B) (2000). See David A. Martin, The Refugee Act of 1980: Its Past and Future, 1982 Mich.Y.B. Int'l L. Stud. 91, 101-04.

⁹ Department of State, Department of Justice, and Department of Health and Human Services, Proposed Refugee Admissions for Fiscal Year 1999: Report to the Congress 15 (June 1998).

¹⁰ Id. at 21.

¹¹ P-4 covers married sons and daughters, siblings, grandparents, and grandchildren. P-5 covers uncles, aunts, nieces, nephews, and first cousins. Department of State, Department of Justice, and Department of Health and Human Services, Report to the Congress on Proposed Refugee Admissions for Fiscal Year 1996, at 20 (July 1995).

¹² I was also told that a growing proportion of P-3 cases, before this restriction was imposed, involved filings by beneficiaries of diversity visas B that is, persons who win one of the 50,000 spaces provided annually as a result of a visa lottery. INA ' 203(c), 8 U.S.C. ' 1153(c) (2000). Because such persons could have brought their spouses and minor children with them at the time of admission under the lottery (see INA ' 203(d)), this increase apparently raised suspicions about this use of P-3, although the 2004 limitation on those who can apply for P-3 family members was never expressly justified to me on this ground.

¹³ INA ' 207(c)(2), 8 U.S.C. ' 1157(c)(2) (2000).

¹⁴ See Department of State, Department of Justice, and Department of Health and Human Services, Proposed Refugee Admissions for Fiscal Year 2001: Report to the Congress 6 (June 2000).

¹⁵ If the person can qualify as an immediate relative of a US citizen (spouse, minor unmarried child, or parent) B a category to which no quota limits and hence no backlogs apply B then he or she must ordinarily use that route rather than gain admission as a refugee. 8 C.F.R. ' 207.1(d) (2004); 9 Foreign Affairs Manual (FAM), Appendix O, ' 204. Under the regulation, this restriction may be waived "in the public interest," however, and in recent years, when it was clear that many refugee admission numbers would go unused, INS and DHS have allowed a blanket waiver.

¹⁶ Empty Seats in a Lifeboat: Are There Problems with the U.S. Refugee Program?, Hearing Before the Subcomm. on Immigration, S. Comm. on the Judiciary, 107th Cong. 40 (2002) (statement of Bill Frelick, Director of Policy, U.S. Committee for Refugees). The proposal also appeared in Bill Frelick, Rethinking U.S. Refugee Admissions: Quantity and Quality, 2002 World Refugee Survey 28.

¹⁷ That is, most of the time the Committee should look favorably on promptly accepting the UNHCR referral and designating the group as a P-2 group for purposes of access to the US admissions process. Actual admissions, of course, would still come only after individual DHS interview and approval.

¹⁸ See Joanne van Selm, Tamara Woroby, Erin Patrick, & Monica Matts, Feasibility of Resettlement in the European Union 113 (Migration Policy Institute 2003). The U.S. demand for a higher level of resettlement referrals also helped strengthen the momentum for UNHCR to complete the development of group-referral procedures, described in greater detail in Chapter VI.

¹⁹ UNHCR-ICMC Resettlement Deployment Scheme: Update for Tripartite Resettlement Meeting (March 17, 2004), available on the UNHCR website, . A similar cooperative arrangement for temporary deployment of skilled protection officers, known as the Protection Surge Capacity Project, has been worked out with the International Rescue Committee. Launched in July 2001, it is largely funded by PRM. IRC's description may be found on its website, at www.theirc.org/index.cfm/wwwID/555.

²⁰ The more elaborate procedures were developed and have been perpetuated in part as a way of helping to deter or detect integrity problems in the UNHCR resettlement system. See Chapter VI. Streamlined procedures, to be sure, would make it easier to disguise fraud or other manipulation B but it should be possible to find an alternative to current UNHCR arrangements that would be less cumbersome, while still retaining adequate checks and balances for integrity purposes.

²¹ Van Selm, et al., supra note 18, at 11-12.

²² U.S. Dept. of State, How a Post Can Refer Cases to the Refugee Admissions Program, State 326248, Nov. 24, 2003 (unclassified cable).

²³ See, e.g., U.S. Refugee Admissions Program for Fiscal Year 2004: Recommendations of The Refugee Council USA (May 2003).

²⁴ INA ' 207(c)(2), 8 U.S.C. ' 1157(c)(2) (2000). Persons admitted in this way are often called derivative relatives, because they derive their eligibility for the program from the characteristics of the principal alien, who is often called the anchor relative. The specified family relationship must have existed at the time that the anchor relative was admitted to the United States. A further administrative limit requires that the application for follow-to-join cases must be filed within two years of the principal’s admission as a refugee. In contrast, most other follow-to-join provisions in the immigration laws allow such benefits without time limit. The two-year limitation was explained to me as a function of the special benefits and public assistance that apply to refugees. After two years the principal alien should be better established B and should then be able to use normal immigration provisions to seek the entry of close family members. See also Procedures for Filing a Derivative Petition (Form I-730) for a Spouse and Unmarried Children of a Refugee/Asylee, 63 Fed. Reg. 3792, 3793 (1998) (explanation accompanying final rule adopting this limitation). The two-year limit can be waived for humanitarian reasons, such as a situation where the family members were missing or unreachable for a lengthy period. 8 C.F.R. ' 207.7(d) (2004).

²⁵ It is hard to conceive of a legitimate policy reason for barring the entry of mother and child in these circumstances, but this outcome stems from the particular wording of INA ' 207(c)(2). See 63 Fed. Reg. 3792, 3794 (1998) (rejecting a suggestion, lodged in response to a proposed regulation, that derivatives of derivatives be allowed admission, on the basis that such a change is foreclosed by the wording of the statute). A technical amendment should be adopted to remedy this situation. Replacing "paragraph (1)" the first time it appears in INA ' 207(c)(2) with "this subsection" would appear to cure the problem.

²⁶ The precise reference of "Visas 93" is to a type of cable used as part of the processing, to communicate between the consular post and main State. Technically, persons admitted as refugees do not receive U.S. visas, INA ' 211(c), 8 U.S.C. ' 1181(c) (2000), but instead receive a different form of travel authorization from U.S. officials that permits them to board the aircraft and apply for admission at a U.S. port of entry. Also, the Form I-730 is used for two types of cases that look quite similar for the DHS processing center's purposes, but quite different to a consular officer, for reasons described in the text below. It covers both the family members of persons who gained status in the United States as a result of an asylum claim under INA ' 208, 8 U.S.C. ' 1158 B called "Visas 92" cases by the State Department B and the family members of persons admitted as refugees under INA ' 207, 8 U.S.C. ' 1157 B "Visas 93" cases.

²⁷ 9 FAM, Appendix O, Part 1700.

²⁸ Small exceptions to this generalization exist for derivative relatives of a qualifying P-3 who accompany that P-3 for admission purposes. For example, if a P-3 wife meets the definition, her minor children in the interview room with her need not be found to have a well-founded fear of persecution. Similarly, if a P-3 father qualifies under the definition, DHS need not make a separate refugee status determination with regard to his accompanying wife. INA ' 207(c)(2), 8 U.S.C. ' 1157(c)(2) (2000).

²⁹ I was told that some version of this type of checking against A-file information actually existed in the early years of the Indochina program, and was done by the Joint Voluntary Agency staff that was responsible for initial refugee processing. But any such systematic checking process had disappeared by the 1990s.

³⁰ See, e.g., Mary Beth Sheridan, Terrorism's Other Victims: Refugees Cleared to Join Family in U.S. Stuck in Limbo After Attacks, Wash. Post, Dec. 2, 2001, at C1; Rachel Swarns, U.S. Security Backlog Strands Many Refugees in Camps Abroad, N.Y. Times, June 4, 2003, at A10; Anastasia Hendrix, Post-9-11 Delays Hurt U.S.-bound Refugees: Security Checks Leave Immigrants in Dangerous Limbo, San Fran. Chronicle, Nov. 30, 2003, at A1. Although most of these backlogs have now been overcome at the screening stage, not all notifications of negative results had been delivered to the applicants at the time of my major interviewing in fall 2003. Sometimes this further delay resulted from UNHCR or host country concerns about the notification process, particularly when RAVU screening resulted in a high percentage of revocations for once-approved cases. Some notifications were therefore delayed so that a comprehensive notification plan could be developed in order to deal with expected public-order problems in the camps. Such caution is understandable, but some notification plans seem to have languished for far too long. It is important that all final notifications in these cases be completed.

³¹ Memorandum from Joseph D. Cuddihy to Kelly Ryan, et al., Recommendations Paper from Fraud Working Group (Aug. 11, 2003) (attaching report submitted by co-convenors of the working group, Joe Martin of the Refugee Office of the DHS Bureau of Citizenship and Immigration Services and Anastasia Brown of the U.S. Conference of Catholic Bishops).

³² Obviously such testing applies directly only to claimed blood relationships and could not verify an asserted marriage. Where there are offspring in the family, however, their DNA results could provide relevant evidence (with adequate allowance, of course, for adoptions and stepchild relationships).

³³ See PRM to Allow DNA Testing for Refugee Family Reunification, Refugee Reports, May 2003, at 9.

³⁴ This is not the true measure of the net per capita cost, however, because implementation of such a system should be offset by savings deriving from likely reductions in other anti-fraud measures, particularly in resources now devoted to catching and prosecuting fraud after admission.

³⁵ Jackie Taitz, Exploring the Use of DNA Testing for Family Reunification (IOM publication, Dec. 2001).

³⁶ See, e.g., UNHCR Projected Global Resettlement Needs 2004 (June 2003).

³⁷ See State Department Expands Family-based Refugee Processing for Africans, Refugee Reports, Sept. 1998, at 7-9.

³⁸ See Department of State, Department of Justice, and Department of Health and Human Services, Proposed Refugee Admissions for Fiscal Year 2001: Report to the Congress 6 (June 2000).

³⁹ In an ongoing program, the two-year limitation on filing I-730s in "following to join" cases (Visas 93) would seem to provide a logical benchmark.

⁴⁰ U.S. Dept of State, FY 2002 Refugee Admissions Processing Guidelines, State 43737, March 6, 2002, at para. 24 (unclassified cable).

⁴¹ See, e.g., Commission on Immigration Reform, U.S. Refugee Policy: Taking Leadership 46 (1997).

⁴² See Kosovo: Outpouring of Misery, Refugee Reports March/April 1999, at 1; Here Come the Kosovars, id., May 1999, at 1; Citing Kosovo as "Example of U.S. Leadership," State Department Proposes Increasing U.S. Refugee Admissions, id., July/August 1999, at 1.

⁴³ See, e.g., Jerry Seper, Iraqi Lawyer Who Saved Pfc. Lynch Granted U.S. Asylum, Wash. Times, April 30, 2003, at A15.

⁴⁴ INA ' 212(d)(5), 8 U.S.C. ' 1182(d)(5) (2000).

⁴⁵ 8 U.S.C. ' 1101(a)(42)(B) (2000).

⁴⁶ For example, the PD could designate persons within their countries of nationality or habitual residence who face grave threats of immediate harm, as determined by the chief of the US mission to that country.