Kari Hong, The Absurdity of Crime-based Deportation, 50 University of California Davis Law Review 2067 - 2147 (June, 2017) (272 Footnotes Omitted) (FULL ARTICLE)
The belief that immigrants are crossing the border, in the stealth of night, with nefarious desires to bring violence, crime, and drugs to the United States has long been part of the public imagination. In 1994, this imagery was used in a notorious commercial in support of California's Proposition 187, an anti-immigration voter initiative (which passed by nearly 60% of the vote but was struck down by federal courts). Donald Trump ratcheted up the rhetoric during the 2016 presidential campaign, calling Mexican nationals "rapists" and "killers" and promising to stop this newly-defined national problem of immigrants who cross the border to kill Americans. Indeed, in his first interview after the election, President-Elect Trump promised to round up and deport up to three million non-citizens who had committed crimes. In one of his first significant acts as president, he enacted executive orders that expand the types of crimes that can serve as the basis for deportation.
The studies and statistics overwhelmingly establish the falsehood of this rhetoric. The facts are that non-citizens commit fewer crimes and reoffend less often than citizens. And, the facts also show that there are not three million non-citizens with criminal convictions that President Trump could target for deportation. But facts do not stop the myth. Even those who support immigration reform often will be quick to point out that they want to help deserving immigrants but will deport the undeserving ones, those with criminal convictions, and especially those who committed violent crimes.
My Article counters that there will be--and should be--an end to crime-based deportation. It is already happening quickly and quietly in federal courts. Moreover, I suggest that as a normative matter, using criminal convictions as proxies to determine who can stay and who must leave is a doomed project: it is arbitrary, disproportionate, and unnecessary.
In Part I, I begin with a discussion of two specific federal laws that have led to a remarkable reform in federal courts. The 1984 Armed Career Criminal Act ("ACCA") and the Illegal Immigrant Responsibility and Immigration Reform Act of 1996 ("IIRIRA") have schemes whereby collateral consequences--substantially increased federal sentences (ACCA) and potential and actual deportation (IIRIRA)--will attach if the person has a prior conviction for a different crime. These statutes are from the Tough on Crime era and add more than simply "direct consequences" to an offense, which the American Bar Association defines as a component of a sentence that the legislature authorized and the criminal court imposed at the time of punishment.
In contrast to direct consequences, ACCA and IIRIRA are examples of statutes that impose collateral consequences. A collateral consequence is a "penalty, disability, or disadvantage" that attaches (either by mandate or by discretion) to a criminal conviction--not by the sentencing court, but by legislatures, agencies, and officials in mostly civil contexts. Under ACCA, if someone commits a federal firearm offense, instead of being punished for that particular crime, he or she will be subjected to at least an additional fifteen-year mandatory minimum prison sentence if he or she committed prior offenses that Congress has defined to be a crime of violence or involving controlled substances. The prior offense attaches no matter which jurisdiction it occurred in and no matter how remote in time. In an era when recidivism was believed to be a social ill, ACCA was the cure.
IIRIRA typically is not thought of as imposing collateral consequences because it is an entire scheme that regulates entries, admissions, and violations of immigration law. But its provisions related to criminal convictions very much are. Historically, immigration law graded crimes based on how the criminal courts treated them. Crimes that state courts considered minor, not meriting lengthy sentences, or expunged under state law did not have immigration consequences. IIRIRA changed that scheme. IIRIRA replaced the measured, individualized determinations with a new legal framework that subjected non-citizens to deportation by lumping types of crimes into crude categories, attaching consequences based on potential--and not actual--sentence length, no longer recognizing the lines between felony and misdemeanor and vacated versus continuing offenses, and no longer offering second chances to those whose equities outweighed their mistakes. As a result, a college student who urinates in public, who is convicted of a misdemeanor public indecency crime, and whose conviction is expunged under state law will be treated identically to a child molester who willfully preys on children in public.
The greatest value of a criminal statute is that it does cover both minor and serious conduct and lets the prosecutor charge and sentencing court impose the appropriate sentence. A prosecutor and criminal court judge will absolutely seek and impose a harsh sentence for the child molester who bears markers of being a public danger but will give the proverbial slap on the wrist for the drunken college student whose actions are borne of stupidity. But, under immigration law, both the child molester and the college student are treated the same. Under IIRIRA, they committed the identical "sex offense," which merits the denial or stripping away of legal status.
The fundamental problem with collateral consequences, and IIRIRA's reliance on them, is that they impose an additional penalty for a crime in a separate context that is not willing or able to distinguish the degrees of harm swept up by criminal law. Direct consequences are related in a meaningful way to the underlying crime or to the offender's future rehabilitation. A person convicted of a DUI may be ordered to abstain from alcohol as a condition of probation but a person convicted of another crime not involving substance abuse often will not. But collateral consequences, inherent to their nature, have no gradations or nuance, rendering them arbitrary and often disproportionate to the original crime. The current immigration law is failing in sorting out the dangerous from the non-dangerous individuals. In a rush to get the "bad guys" and "bad hombres" out, the law sweeps in too many people, which leads to absurdity. For example, a man who had been in the United States with a green card for forty years was deported over stealing a $2 can of beer. The Ninth Circuit reversed the deportation on a technical matter, but the case illustrates that figuring out who is dangerous and who is not based on a criminal record is not an efficient or effective method of immigration enforcement.
Part II looks at the practical reality of how the Supreme Court is quietly and effectively ending crime-based deportation. In 2013, in Descamps v. United States, Michael Descamps was convicted of the federal crime of being a felon in possession of a firearm. Without ACCA, a federal judge would impose a sentence between zero and ten years, depending on the existing mitigating or aggravating circumstances of the possession offense. But that was not the sentence Mr. Descamps was facing. Mr. Descamps had a prior California burglary offense, which under ACCA, triggered a fifteen-year mandatory minimum sentence as a crime of violence. The prosecutor alleged that the California offense matched ACCA's generic definition of a burglary (unlawful entry of a residence with intent to commit a crime) because California criminalized an unlawful entry in defining the crime under state law. The Supreme Court disagreed. In a technical decision, the Court set forth a three-step framework to analyze whether a California's burglary statute will have ACCA or IIRIRA consequences--which it clarified as being the categorical approach, divisibility step, and modified-categorical approach. As to the California statute at issue, the Court held it was overbroad and indivisible because California defined the element of entry as involving both lawful and unlawful entries and did not direct juries to agree on the actual method that occurred in that crime. Mr. Descamps was not subjected to ACCA's sentencing enhancement, and any other person convicted of that crime would not be subjected to ACCA or IIRIRA's collateral consequences.
This area of the law had been mired with confusion, conflicting decisions, and ever-shifting doctrines. The Descamps majority decision, authored by Justice Kagan, announced that its three-step methodology was an obvious and natural extension of twenty-five years of precedent. Of note, in the divisibility step, the majority introduced the role of state law in interpreting statutes. In the above example, California jury instructions and case law would illuminate how a state defined the elements of the burglary offense. In the sole justice writing in dissent, Justice Alito, made a pointed criticism that the purpose of the Court was to "simplify the work of ACCA courts." Justice Alito argued that Descamps' new methodology was failing that end because parsing state law to figure out elements will be a difficult, technical rabbit hole.
After Descamps, a circuit split arose over whether the text of a statute or state law authority would be best used in gleaning whether the elements of a prior offense were or were not divisible. The majority of courts followed Justice Alito's warning to avoid looking to state law, and only two-- the Ninth Circuit and the Fourth Circuit--instead referred to state law--jury instructions, court decisions, and other statutes--to determine the meaning of the predicate crime. In 2016, Mathis v. United States resolved the technical dispute in favor of relying on state law to determine divisibility.
Part III focuses on two significant empirical aspects of the operation of Descamps and now Mathis. In weighing in on the Justice Kagan and Justice Alito debate, there is evidence to strongly suggest that Justice Kagan was likely correct in asserting that Descamps' approach is more consistent and workable compared not just to pre-Descamps case law--but also to the textual approach recommended by Justice Alito's dissent. In the pre-2013 landscape, courts often employed Justice Alito's recommended methodology, gleaning a defendant's conduct from court records. The result was ever-shifting doctrines, methodologies, and confusion. Compared to the pre-2013 cases, Descamps' return to an elements-only approach introduced clarity that had been lacking.
But more to the heart of Descamps, Part III compares the Eighth Circuit's textual analysis (Justice Alito's dissenting approach) with the Fourth Circuit's state-law approach (Descamps' majority approach). Looking at the published cases issued in the time period between Descamps and Mathis, there is a strong argument that the state-law approach is surprisingly workable. Among the fourteen Eighth Circuit decisions, one panel disagreed over whether the state statute could be interpreted by text or by a case that provided a different interpretation in context. Among the eleven published Fourth Circuit decisions, not a single judge disagreed with whether state law resolved the divisibility question. To the extent that consensus across political ideologies is evidence of consistency, this survey of cases suggests that Descamps in fact is providing more clarity and workability than its critics forewarned.
But a more notable discovery from the different approaches is that the state-law approach--the one endorsed by Mathis--is scaling back collateral consequences of convictions in a much more pronounced way. When Descamps was properly applied in the circuits, there were noticeably fewer prior convictions that aligned with the generic offenses set forth in ACCA and IIRIRA. Stated another way, under Descamps and Mathis, there will be fewer statutes to which ACCA and IIRIRA consequences will attach. Federal judges have long voiced concern over the arbitrariness in ACCA and IIRIRA cases, colloquially described as one bad guy wrongfully getting off and another long-term resident being deported over a petty offense. I argue that the arbitrariness attributed to Descamps is in fact the fault of ACCA and IIRIRA and inherent to the very nature of collateral consequences. Whereas direct consequences of criminal law--conviction and sentence--metes out a tailored punishment to any given offender, ACCA and IIRIRA in particular are resulting in the absurdity of sweeping in minor and non-serious offenses. The greatest value of Descamps and Mathis is their ability to restore proportionality to ACCA and IIRIRA by eliminating their collateral consequences in their entirety.
In Part IV, I contend that, as a normative matter, the actual result of reducing crime-based deportations is the correct one to reach. After Mathis and Descamps, in the ACCA context, prosecutors, judges, and policy makers are embracing the reduction and even elimination of collateral consequences. That is to be expected given that ACCA's additional mandatory punishment is superfluous. Whatever the crime the defendant did when committing a federal firearm offense, the prosecutor and sentencing judge can assess the appropriate sentence. Avoiding a mandatory and lengthy sentencing enhancement for a past crime whose sentence was already served is neither unreasonable nor undesirable. In the Tough on Crime era, the additional penalty was believed to end recidivism. Mass incarceration has exposed this premise to be misguided and some contend that lengthy sentences are a leading contributor to recidivism.
This Article suggests that that same result should be embraced in full in the immigration context. IIRIRA's use of criminal convictions to sort out desirable from undesirable immigrants has failed. The use of a conviction does not distinguish who is dangerous from who is not. This existing system results in a man with a green card, who has a citizen wife and children and has been in the United States since his childhood, to be deported over stealing a $2 can of beer. Limiting the crimes to only violent ones also fails under this approach. Under IIRIRA's definition, many individuals are included as violent offenders even though their offenses would not be viewed as such. For instance, until the Supreme Court corrected the issue, numerous individuals were deported because the federal courts wrongfully found their DUI convictions to be violent crimes. Likewise, a teenager who spat at a police officer during an arrest has been found to commit a crime of violence, even though few would deem that conduct serious, dangerous, or worthy of lost immigration status.
It is time then to reconsider and repeal IIRIRA and its emphasis on deporting individuals based on convictions alone. I am far from alone in calling for such reform. A number of immigration scholars have been arguing for this result out of principles of proportionality, fairness, and human rights norms. Likewise, criminal law scholars are engaged in an invigorating project to question the reason for--and unintended costs of--collateral consequences. This article contributes to the existing conversations by embracing the practical realities that are happening in the federal courts and voicing a normative defense of that result.
In the current political climate, defending non-citizens who commit crimes as valuable members of our country is not popular. But as a matter of history, it is IIRIRA's radical reconfiguration for a great number of convictions to result in deportation that is the aberration. For 100 years, criminal aliens were only seven percent of all deportations. In 1996, IIRIRA expanded the types of crimes that could kick someone out of the country from the most serious felonies to minor crimes, including drug possession, misdemeanors, and even those that are expunged and pardoned. Now, six out of every ten deportations are for crimes, at a tangible cost in billions of dollars in heightened enforcement.
But not everyone who breaks the criminal law is a hardened recidivist or even dangerous. Some have served in our military, have children and spouses who are citizens, have lengthy employment histories, have been paying taxes, and are good neighbors. Some commit crimes that in fact involve non-serious conduct. Criminal judges who knew the facts of a case used to be able to stop deporting someone whose crime arose from stupidity rather than depravity. Moreover, these crime-based deportation grounds apply retroactively, meaning individuals who pose no risk of any criminal activity are newly subjected to deportation based on a crime they committed--and reformed from--decades ago. Immigration judges too used to have power to weigh the good and the bad and give a second chance to individuals who exhibited remorse, rehabilitation, and contributions to the community. IIRIRA simply ended giving anyone a second chance. As illustrated in the cases described above, kicking these people out, and disrupting their families, over drug possession and minor offenses is harsh and absurd.
This article was written before the election of Donald Trump. It has taken on new meaning in light of the Trump administration's heightened commitment to increasing crime-based deportations. In light of this shift in immigration enforcement priorities, it is critical to ask why are crimes used as a means to deport people? As argued here, the main problem is that the current use of categories to identify which crimes will have immigration consequences is incredibly overinclusive. A deportable offense includes misdemeanors; a violent crime includes spitting at a police officer during an arrest. The Trump administration is taking the absurdity found in existing law and making it ridiculous. The new policies do not fix the overinclusive problem. To the contrary, someone is now a priority for deportation if they have been arrested or engaged in conduct that could be a crime.
Instead of sorting out immigrants based on how they entered or penalizing them for minor crimes, immigration law needs to sort out contributing from non-contributing immigrants. Criminal convictions are neither effective nor efficient in drawing these lines. After Mathis, the federal courts will be significantly reducing collateral consequences on a case-by-case, statute-by-statute basis. The federal courts will reach this result out of the technical dissection of individual criminal statutes. This Article argues that this practical result is the better way to avoid the absurd and arbitrary results that otherwise attach when criminal convictions are considered relevant to the questions of who can stay and who can remain in this country. Those who pay taxes, have citizen children, serve in the military, work in jobs citizens will not take, or help those around them, need a path to legalization. And those who cause more harm than good should be deported. Criminal convictions can no longer be the only factor in this consideration. A full repeal of IIRIRA and a defense of non-citizens who commit crimes is in order.
Assistant Professor, Boston College Law School.