Criminal Justice & the Rule of Law

Biden’s Border Problem, and How to Fix It

Peter Margulies
Monday, April 19, 2021, 10:48 AM

The United States is seeing a troubling rise in unaccompanied children crossing the U.S.-Mexico border. Resolving the problems at the border will require action from the Biden administration, Congress and the courts.

The U.S.-Mexico border wall in Nogales, Arizona. (Ignatian Solidarity Network,; CC BY-NC 2.0,

Published by The Lawfare Institute
in Cooperation With

President Biden has a border problem. The U.S. is seeing a deeply troubling increase both in total unaccompanied children (UACs) and in very young UACs—children as young as six or seven years old. The current problems at the border have both short- and long-term causes. In the short term, the Biden administration contributed to the problem with two related steps. First, it chose to continue the Title 42 coronavirus border restrictions that severely limit family entry. Second, it chose to exempt UACs from Title 42. In the longer term, addressing the exponential increase in both UACs and immigrant families over the past 10-12 years will require action by Congress and the courts. This post addresses both short- and long-term issues, centering on UACs and families.

The story of sharply increased Central American immigration to the United States from 2007 on includes push, pull and operational factors. Push factors are forces that drive Central Americans out of their home countries. Those forces include poverty, climate change and violence. For example, climate change has reduced coffee production, which is a prime source of employment in Central America. High rates of gang activity and violence in Central America also spur immigration, although many accounts overstate the role of violence: Death rates in Guatemala—the principal home country of Central American immigrants—have dropped 50 percent in recent years.

If escape from push factors like poverty, climate change and violence is immigrants’ goal, the operational factor of smuggling networks is the means. Smugglers ease the logistical and safety challenges of travel to the U.S. border. Like other businesses, smuggling networks charge a price for their work and try to cultivate new business opportunities.

Smuggling networks also operationalize pull factors. Pull factors are forces in the United States that attract immigrants from Central America, such as economic opportunity and family reunification. U.S. economic opportunity contrasts favorably with diminishing opportunities in Central America. Immigrants from Central America may also want to reunite with family members already in the United States. (See this analysis by Juliette Kayyem, former assistant secretary of homeland security in the Obama administration).

Once smuggling networks have gotten Central American immigrants to the border, a second operational factor allows immigrants to remain here: U.S. law. Just as smugglers are a means for immigrants to arrive at the border, U.S. law is a means for immigrants to realize economic opportunity and family reunification in the United States. As we shall see, features of U.S. substantive immigration law, procedure and enforcement—such as the burgeoning asylum backlog, the inclusion of families in the 1997 Flores settlement agreement (FSA), and provisions of the Trafficking Victims Protection Reauthorization Act (TVPRA)—allow many immigrants to obtain speedy release from detention and sometimes qualify for legal permanent resident status. To be sure, each of these legal features also provides needed relief for persons who can stake out a meritorious claim and in some cases—such as asylum—allows the United States to fulfill duties under international law. But those successes are only one side of the ledger. U.S. law also erects formidable practical obstacles to the removal of those immigrants who ultimately lose on the merits.

U.S. law complements the operational role of smuggling networks. For hard-pressed Central Americans, paying smugglers has always been a gamble. While Central Americans living in poverty may not initially know the features of U.S. law such as asylum backlogs that delay adjudication, smugglers fill that information gap. Additional information comes from Central Americans’ relatives in the United States. Fulfilling their side of the wager, Central Americans invest their savings, hoping that with smugglers’ help they can attain a more safe and prosperous future in the United States for themselves or their children. Favorable trends in U.S. law increase the odds of success. It is difficult to precisely calculate the role of U.S. law when weighed against push factors such as poverty, climate change and violence. But statistics and common sense tell us that the role of U.S. law is significant.

Opportunistic entities such as smuggling networks always use the law for their own purposes. As California Supreme Court Justice and Stanford Visiting Professor Mariano-Florentino Cuellar has explained, U.S. corporations use the law in the same way. U.S. employers benefit from weak enforcement of the employer sanctions provisions of the Immigration and Nationality Act, which on paper bar employment of undocumented immigrants. Echoing Justice Cuellar, Adam Cox and Cristina Rodriguez have observed that the absence of remedies in U.S. law for labor law violations against undocumented workers fuels exploitation of this vulnerable group. Smugglers are merely playing the same game as U.S. firms.

This analysis of push, pull and operational factors sets up recommendations for reform. To restore control at the border, the Biden administration should halt the Title 42 program for all entrants. Immigration advocates may well applaud this change. However, in the longer term, Congress, immigration officials and the courts will have to take other actions that immigration advocates may criticize. Congress will need to limit Special Immigrant Juvenile Status (SIJS) under the TVPRA of 2008. Immigration officials will need to ask the courts to modify the Flores settlement to clarify that the agreement’s rapid release requirements apply only to UACs, not to accompanied children in immigrant families. The government should be able to detain immigrant families for as long as 45 days, to complete the expedited removal process. In addition, the government should implement “last-in, first-out” (LIFO) asylum processing to reduce the massive asylum backlog.

Currently, the Biden administration’s layering of Title 42 and the UAC exemption creates an operational factor encouraging entry of UACs, including unprecedented numbers of very young children.

Under the Title 42 program—named after the volume of the U.S. statutory code that deals with health legislation—the Centers for Disease Control and Prevention can limit entry into the United States to address public health emergencies. As part of the Trump administration’s response to the coronavirus pandemic, the CDC sharply limited entry at the U.S.-Mexico border. While immigration advocates challenged the Title 42 program and many public health experts criticized the program as misguided and inadequately tailored to public health threats, the program continues in effect. Biden has maintained the program, although administration sources have conceded that the program’s rationale—if it was ever plausible—will continue to erode as coronavirus vaccinations become more widely available.

The Biden administration has exempted one group of foreign nationals from Title 42: UACs. In late January, the U.S. Court of Appeals for the D.C. Circuit stayed a decision by Judge Emmet Sullivan of the U.S. District Court for the District of Columbia in P.J.E.S. v. Wolf enjoining the Title 42 program. Sullivan had found that the summary removal of UACs at risk of torture, persecution and abuse would be an “irreparable injury” that outweighed any public health benefits of the Title 42 program. Notwithstanding the stay of Sullivan’s order, the Biden administration determined that subjecting UACs to summary removal would be inappropriate and inhumane. As a result, the Department of Homeland Security is handling UACs under the pre-Title 42 framework: the TVPRA. That gap between the treatment of families under Title 42 and the more generous treatment of UACs under the TVPRA has become a pull factor for increased UAC entries. Explaining why requires a deeper dive into the TVPRA’s provisions.

Because Congress has recognized that the government has a responsibility to protect UACs, who have no parent to care for them, the TVPRA requires speedy release of UACs from detention. Under the TVPRA, the Department of Homeland Security must within 72 hours of a UAC’s arrest at the border transfer custody to the Office of Refugee Resettlement (ORR) of the Department of Health and Human Services. Due to provisions of the TVPRA and related government undertakings based on the Flores settlement (more later), ORR then has 20 days to place the child safely in a community in the U.S. interior.

The requirement of prompt community placement makes the TVPRA a powerful operational factor. To illustrate this, consider asylum, which requires a well-founded fear of persecution based on race, religion, nationality, political opinion, or membership in a particular social group. A UAC will often first make an “affirmative” application to the Asylum Office in Homeland Security’s office of Citizenship and Immigration Services (USCIS). If that application does not succeed, the child will end up in a removal proceeding in the Department of Justice’s Immigration Court, for a full asylum hearing before an immigration judge. Immigration courts now have a backlog of more than 1.2 million cases.

Because of these prolonged delays, an asylum case that starts now will not be resolved for three years or more. Asylum-seekers can file for work permits 150 days after making her asylum claim, and long-established rules give Homeland Security 30 days to decide on the application, which is almost always granted (the Trump administration issued a final rule eliminating the 30-day requirement; a U.S. district court enjoined the rule, but only for members of two organizations). After 180 days, an applicant receives a work permit. For UACs who are 16 years old or older—historically the vast majority—a work permit is a valuable immigration benefit.

The ability to stay and work in the United States legally for years while an asylum claim is pending creates perverse incentives. Many asylum-seekers have meritorious asylum cases. But the chance to remain in the United States and work legally for a lengthy period also elicits unfounded claims. Consider allegations of gang violence, which figure heavily in Central American asylum applications. As the U.S. Court of Appeals for the Second Circuit observed in Ucelo-Gomez v. Mukasey, the Immigration and Nationality Act sharply limits gang-related asylum claims.

Under the statute, asylum requires more than harm from generic crime and greed. While it is certainly dangerous to live in a neighborhood with a high crime rate caused by gang activity, that danger in itself is not a basis for asylum. If it were, tens of millions of people around the world could claim asylum in the United States. But both the U.N. Refugee Convention and U.S. law require evidence of more specific types of persecution. A substantial majority of Central American asylum-seekers—including UACs—cannot make that more specific showing and thus fail to meet the legal requirements for asylum status.

At that future date when a UAC who has relied on an asylum claim receives a final adverse decision, the government lacks an effective remedy. It is true, as Ingrid Eagly and Steven Shafer have shown, that a substantial majority of asylum applicants attend their asylum hearings. However, few unsuccessful asylum applicants who have been released from detention cooperate in removal efforts after they have lost their cases. A 2003 report by the Justice Department’s inspector general criticized as “ineffective” post-merits removal efforts by the Homeland Security office of Immigration and Customs Enforcement (ICE), in large part because of this marked lack of cooperation.

Over time, the efficacy of removal has declined even more sharply. According to a fiscal 2019 report by ICE, a cumulative total of 595,430 foreign nationals who at some previous point had received final orders of removal had failed to cooperate with removal. While that reluctance is understandable, given the lives that immigrants have built here, it still undermines immigration enforcement. Thanks to the smuggling networks’ efforts and information from relatives already in the United States, prospective entrants to the United States know that assertion of an asylum claim at the border and subsequent release to the U.S. interior under the TVPRA effectively ensure that a UAC asylum applicant can remain in the United States indefinitely. That result is hardly optimal for parents. Given Title 42’s applicability to adults, it leaves parents outside the border, looking in. But for parents concerned about a better life for their children, helping children enter is a worthwhile “second-best” option. Driven by this dismal calculus, many Central American parents make the wrenching choice of sending children as young as six or seven years old to the U.S. border.

In sum, it is likely that the Biden administration’s maintenance of Title 42 and exemption of UACs would produce exactly the result that has occurred: a substantial increase at the border of UACs, including those far younger than any seen previously. While the current border situation has a range of causes, including underlying poverty and violence in Central America, the Biden administration’s choices have exacerbated the problem.

Expedited Removal

The readiness and remedies gaps that characterize the interaction of Title 42 and the TVPRA also adversely affect expedited removal, a process that Congress established in 1996 to close the readiness and remedies gaps of that period. Congress hoped to speed up asylum decisions for foreign nationals arrested at the border. Under expedited removal, an asylum officer conducts an interview after the foreign national’s arrest. The timing of the interview can range from days after arrest to four to six weeks after the foreign national’s apprehension. If the asylum officer finds that the foreign national does not have “credible fear” of persecution and an immigration judge agrees in a truncated hearing, the United States can remove that person. Judicial review is virtually nonexistent, a result the Supreme Court recently upheld.

In practice, however, expedited removal has not achieved Congress’s goal of speeding up removals. Here again, the TVPRA has played a role. Under 8 U.S.C. 1232(a)(5)(D)(i), UACs from Central America bypass expedited removal. Instead of expedited removal, UACs from the Northern Triangle of Guatemala, Honduras and El Salvador get a full hearing with an immigration judge. In most recent years, immigration courts have decided cases in the order in which those cases arose, an approach called “first-in, first-out” (FIFO). But the adjective “first” in “first-out” is a relative term, given the extraordinary backlogs in Immigration Court. As noted above, a case that starts now will not be resolved for three years or more.

The exemption of UACs from expedited removal expands the readiness and remedies gaps in immigration enforcement. Those gaps in U.S. law are a pull factor for entry of UACs. According to research by migration experts Catalina Amuedo-Dorantes and Thitima Puttitanun, the TVPRA is a “key determinant” driving UAC entries into the United States, which have increased 1,000 percent since the TVPRA’s enactment, according to a report by the Congressional Research Service. Because of the need for U.S. officials to treat UACs humanely in their parents’ absence, letting UACs bypass expedited removal is the right call. But policymakers rightly concerned about humane treatment of UACs should recognize that UACs’ exemption from expedited removal is a double-edged sword. Because of the interface of anxious Central American parents and smuggling networks, the UAC exemption from expedited removal generates more UACs. That is not a humane result, whatever policymakers’ intentions.

The Flores Settlement

The courts’ unduly expansive interpretation of the Flores settlement agreement has further broadened the readiness and remedies gaps. The original FSA covered only UACs. But a recent decision by the U.S. Court of Appeals for the Ninth Circuit upholding injunctions against a final rule issued by the Department of Homeland Security during the Trump administration has incorrectly read the FSA as also covering immigrant families containing minor children (accompanied minors). Because of this and other flawed rulings, expedited removal is effectively inapplicable to families containing both adults and accompanied minors. That de facto exemption from expedited removal has become a formidable pull factor.

As I noted in an earlier post, the 1997 FSA resolved a lawsuit brought on behalf of UACs. The Supreme Court in Reno v. Flores stated that the lawsuit concerned immigration detention of “juveniles who are not accompanied by their parents” (emphasis added). The Flores court was deferential, upholding a federal rule that provided for government custody of UACs who could not be released to a parent, legal guardian or other close adult relative in the United States. The court’s deference will surely inform any future Supreme Court decision on the scope of the FSA.

The 1997 FSA established ground rules for UACs’ custody, detention and release. Such settlements are called consent decrees, in which the parties negotiate terms that a court approves in a judicial order. As the Supreme Court explained in the consent decree case United States v. ITT Continental Baking Co., the overall structure of the agreement is important, as are the “surrounding circumstances.”

Let us apply that contextual approach to the FSA. While the FSA defined the class protected by the settlement as “[a]ll minors who are detained in the legal custody” of immigration officials, the structure and context of the FSA show that UACs were the only persons included in this category. The duties of the government under the FSA all referred to UACs. For example, paragraph 14 included detailed provisions authorizing release of minors “to” parents, legal guardians or other close adult relatives, while also permitting release to licensed programs and to other adults and entities at the discretion of immigration officials. Another provision of the FSA barred transportation of minors with unrelated adults. These provisions addressed the care and safety of UACs; they would have been irrelevant to accompanied minors who remained with their parents and relied on parents for their care. Moreover, between 1997 and 2011, the overwhelming majority of unauthorized immigrants at the southern border were either single adults or UACs. Immigrant families were a tiny fraction of the total. Under the circumstances, an agreement regarding accompanied minors would have been largely unnecessary. In sum, from the start, the FSA was all about UACs and not about families.

Unfortunately, the Ninth Circuit’s recent decision on the FSA failed to follow the contextual approach. The court did not adequately address the agreement’s overall structure and language, which refer almost exclusively to UACs. In addition, the Ninth Circuit failed to acknowledge the context of the agreement, which arose in litigation exclusively about UACs, as the Supreme Court explained in Reno v. Flores. Rather than interpret the FSA, the Ninth Circuit in effect wrote a new agreement that was more to its liking.

Even assuming the Ninth Circuit was correct in its reading of the original FSA, the court incorrectly rejected the government’s effort to modify the FSA in light of changed circumstances. As the Supreme Court observed in Horne v. Flores, a judicial tribunal approving a consent decree such as the FSA sits as a court of equity, meaning that judges can modify their orders in light of changed circumstances. The Supreme Court has signaled that it will generally defer to reasonable government arguments that current conditions have rendered an earlier settlement obsolete.

Equitable modification would be appropriate regarding the FSA. The vast increase in accompanied children has made the agreement obsolete regarding this group, if it ever applied in the first place. Perhaps the FSA made sense for accompanied children and their families when their numbers were infinitesimal. But the exponential increase in the volume of family entrants made the original agreement outmoded.

The sharp increase in the number of accompanied families at the border suggests that the FSA has become an operational factor facilitating increased family immigration. A framework that is purportedly designed to protect accompanied children at the border should provide appropriate care for the children it covers. But it should not in itself increase the number of those children. That unintended consequence undermines humanitarian goals. Because the FSA now has this operational effect, it requires equitable modification.

Special Immigrant Juvenile Status as a Pull Factor

The FSA interacts with the TVPRA to form another operational factor: the availability of SIJS. Many UACs qualify for SIJS, which provides a basis for gaining lawful permanent resident status, followed by U.S. citizenship. UACs without well-founded asylum claims still often qualify for SIJS. For UACs, SIJS’s main challenge is the time to apply and then wait for adjudication to unfold. Here, the readiness gap takes another turn. Thanks to the FSA’s quick-release mandate, UACs’ exemption from expedited removal, and the massive asylum backlog, UACs usually have the time required.

To qualify for SIJS, UACs apply to USCIS. That application requires findings from a state family court that a UAC has been abused, neglected or abandoned by “1 or both parents”; reunification with one or both parents is not “viable”; and the return of the child to her home country would not be in the child’s “best interest.” Upon USCIS approval of an application, a SIJS recipient will still have to wait for a substantial period—often years—for a visa to become available (or “current,” in immigration parlance). During that period, most administrations have given SIJS applicants deferred action—the same benefit available for Deferred Action for Childhood Arrivals (DACA) recipients.

I started doing SIJS cases in the 1990s and can testify to the program’s benefits and flaws. On the positive side of the ledger, SIJS has changed lives for the better in thousands of cases, while the United States has often gained hard-working members of the community. However, SIJS also has adverse effects. While Congress aimed to provide support for UACs who lacked parental care, the unprecedented entry of UACs as young as six or seven years old shows that SIJS doesn’t merely aid UACs. Along with that aid, the pull factor of SIJS also increases UACs’ ranks.

Consider again the statutory requirement of a family court finding that a UAC has been abused, neglected or abandoned by “1 or both parents.” In some cases, that finding rests on demonstrable acts or omissions by a parent or parents in the UAC’s home country—typically one of the Northern Triangle countries of Central America. In contrast, in other cases, at least one parent abroad is entirely fit. However, many state courts, such as those in New York, interpret the statutory standard as merely requiring proof that one parent is unfit. In courts that construe the “1 or both” standard broadly, the presence of a fit parent in the UAC’s country of origin is legally irrelevant. Some state courts, such as the Nebraska Supreme Court, have pushed back with a more carefully tailored interpretation, finding that the statutory reference to “1” parent’s lack of fitness refers only to cases where just one parent is still “in the picture,” with the other parent unaccounted for or deceased.

The Nebraska Supreme Court rightly noted that the statutory language does not support the broader reading of “1 or both.” Courts usually interpret a statute to ensure that every key term in the statute means something, on the theory that the legislature drafts statutes carefully. The word “both” in the TVPRA would be unnecessary if Congress meant to base SIJS merely on a finding that one of two available parents was unfit. As the Nebraska court noted, it is reasonable to infer that Congress included the one-parent option as guidance when only one parent was a possible candidate for reunification, since the other parent was deceased or had deserted the family either before or after the child’s birth. But these courts are in a distinct minority.

As with the asylum backlog and the Flores agreement, one should take care to acknowledge that SIJS is not an operational factor for all Central American immigration. Some families and UACs would come to the United States solely because of push factors such as violence and persecution. But the analysis above shows that U.S. law is a factor for many Central American immigrants, reinforced by the smuggling networks on which those immigrants rely.


To address the current situation, Biden should make the following changes:

First, Biden should withdraw the Title 42 restrictions. Those restrictions have outlived their usefulness. Instead of stopping most families from entering the United States, the administration should require that all families receive vaccinations prior to entry and should provide those vaccinations to all those seeking asylum at the border. While that might seem like a large undertaking, the ability to operate on that scale is straightforward, given the huge strides made in vaccination rates nationwide since January. Withdrawing the Title 42 restrictions will end the de facto family separation occurring near the border, as families frustrated by the restrictions send young children to the border as a “Plan B.” While dealing with families seeking asylum presents challenges, those challenges are more manageable than addressing the needs of UACs as young as six or seven years old.

To get better control over the challenges of family immigration from Central America, the Biden administration should seek to modify the FSA to exclude families from its scope. The inclusion of this group is a misreading of the FSA that created perverse incentives, generating more accompanied children and families at the border. Those perverse incentives constitute changed circumstances that require equitable modification of the FSA. The Supreme Court’s handling of changed circumstances in past cases such as Horne v. Flores suggests that courts would ultimately approve this change.

However, the dangers of family detention require the imposition of safeguards. Particularly for children, even short stays in detention can be harmful. Congress should authorize access to habeas corpus in federal court for any family detained longer than 30 days. In the habeas proceeding, a family should be able to assert an asylum claim and seek relief for harsh or unhealthy conditions of confinement. In addition, Congress should require the release of any person who has received a positive credible fear finding, as well as that person’s immediate relatives. Access to court will create an incentive for the government to improve conditions of detention and quickly decide asylum claims.

On SIJS, Congress should amend the TVPRA to clarify its ambiguous language on reunification with parents. Instead of the current language that requires a finding that reunification with “1 or both” parents is impracticable, Congress should require a negative finding on the prospects for reunification with each parent. To ensure clarity, Congress should provide that if reunification with one of the child’s parents is practicable, that would disqualify the applicant from relief. This change would mean that applicants with a fit parent in their home countries would not meet SIJS requirements and, hence, would not be eligible for legal permanent resident status. That change would significantly reduce the SIJS pull factor.

On asylum, as former Department of Homeland Security senior lawyer and University of Virginia law professor David Martin has long advocated, the U.S. should move from the inefficient FIFO system to a LIFO system. Under a LIFO regime, new cases would be adjudicated first. If those cases were denied, the losing asylum claimant would be swiftly removed. A pivot to LIFO would narrow the readiness and remedies gaps that make asylum a pull factor. A bipartisan Customs and Border Patrol subcommittee recommended this step in 2019. The Biden administration appears to be considering a similar proposal by former top immigration official Doris Meissner. This change would be a welcome step in the right direction. It would take time and resources to implement. But that commitment would be worthwhile.

In the realm of adjudication, the Biden administration has also announced that it will restart the Central American Minors program, which provides processing in home countries for children of Central American immigrants lawfully present in the United States. The parents here typically have deferred action or Temporary Protected Status, so they would otherwise not be able to legally petition for children who are still located abroad. The CAM is a promising form of in-country processing that will relieve pressures on the U.S. asylum system. In addition, as Martin recommended in the Vox piece cited above, the United States should develop more in-country and regional alternatives to U.S. asylum. (Former Obama White House Domestic Policy Council director Cecilia Muñoz agrees.) The Trump administration gestured toward that strategy with asylum cooperative agreements (ACAs) with Iron Triangle countries that the Biden administration has wisely rescinded. The Trump ACAs lacked a serious commitment to developing the fragile rule-of-law institutions necessary for fair and accurate asylum adjudication. However, careful preparation and collaboration can develop more functional institutions over time.

Speaking of commitment, the Biden administration is right to also stress long-term economic development and political reform in Mexico and Central America. Former Obama Secretary of Homeland Security Jeh Johnson has long urged that step. The development and reform effort is necessary to reduce push factors like poverty, climate change and violence. To operationalize that focus, the Biden administration needs to partner with indigenous nongovernmental organizations, given the corruption that plagues governance in the region.

Unfortunately, a recent move by the Biden administration deepens its reliance on the region’s flawed governments. On April 12, the Biden administration announced agreements with Mexico, Honduras and Guatemala. Under the terms of those agreements, these countries will dispatch troops to their respective borders to curtail the movement of prospective immigrants heading for the United States. Given the scope of the U.S. border problem, it is understandable that the Biden administration would want to enlist the cooperation of other countries in the region. However, the use of other countries’ military forces is too blunt an instrument for this occasion. Those forces lack the expertise or infrastructure to process asylum claims. As a result, the new Biden agreements will endanger asylum-seekers with claims that might ultimately prevail on the merits. Moreover, these agreements will draw the United States closer to the region’s corrupt governments, which have often exacerbated the push factors of poverty and violence that help prompt immigration to the United States. That isn’t a sustainable policy; it’s a surrender to expedience. The Biden administration can and should do better.

In conclusion, the current border situation is a symptom of both short- and long-term policy choices. Policymakers need to get a handle on the rising number of UACs, particularly very young children. That effort will require ending the Title 42 program and reforming asylum adjudication, the Flores settlement agreement and Special Immigrant Juvenile Status. A show of political will by the Biden administration will be necessary, along with actions by Congress and the courts. Despite the complexity of the problem, a whole-of-government approach can achieve visible results.

Peter Margulies is a professor at Roger Williams University School of Law, where he teaches Immigration Law, National Security Law and Professional Responsibility. He is the author of Law’s Detour: Justice Displaced in the Bush Administration (New York: NYU Press, 2010).

Subscribe to Lawfare