Cornell University Law School Social Science and Law

Sessions v. Morales-Santana

Compiled by Emily Le, Leonardo Mangat, and Natalia San Juan

Overview: Luis Ramón Morales-Santana alleged that 8 U.S.C. §§ 1401 and 1409(a) and (c), the statutes that impose a gender-based difference in the physical presence requirement concerning the transfer of derivative citizenship to a child born abroad, violated the equal protection guarantee of the Fifth Amendment.

Part 1: The Legal Question and Facts

Major Legal Question:
Does the gender-based difference in the physical presence requirement for the transmission of derivative citizenship to a foreign-born child for unwed citizen mothers and unwed citizen fathers violate the Fifth Amendment's equal protection guarantee?

Facts:
    Luis Ramón Morales-Santana was born in the Dominican Republic in 1962 to an unwed U.S.-citizen father and a Dominican mother. Morales-Santana v. Lynch, 804 F.3d 521, 524 (2nd Cir. 2015). He was denied derivative citizenship because, under the Immigration and Nationality Act in effect at the time of Morales-Santana’s birth, his father had to be physically present in the United States for ten years, and at least five of those years must have been after the age of fourteen, in order to meet the requirements to transfer derivative citizenship to a child born abroad. Though his father was born in Puerto Rico, and lived there for eighteen years, he left Puerto Rico for the Dominican Republic twenty days prior to his nineteenth birthday.
    Morales-Santana was “legitimated” by his father when his parents married in 1970 and was admitted to the U.S. as a lawful permanent resident five years later. He was placed in removal proceedings in 2000 after being convicted of several felonies.
   The applicable rule for transmitting citizenship to a child born abroad at the time of Morales-Santana’s birth required the physical presence of U.S.-citizen parents in the United States for ten years prior to the child’s birth. Though § 1409(a) applies this rule to unwed citizen fathers, § 1409(c) contains an exception for unwed citizen mothers, who may transfer their citizenship to a foreign-born child after living continuously in the United States for only one year prior to the child’s birth. Sessions v. Morales-Santana, No. 15-1191, 582 U.S.___, slip op. at 1 (2017).
    Morales-Santana argued that the denial of derivative citizenship violated the equal protection guarantee of the Fifth Amendment because the gender difference meant that although his father did not satisfy the requirements for unwed citizen fathers, he met the one-year requirement for unwed citizen mothers. Id. at 6.
    In an 8-0 decision for Morales-Santana, the Court found that the gender-based differences in the physical presence requirements violated the Fifth Amendment's equal protection guarantee because it disadvantaged unwed citizen fathers on the basis of gender. The Court determined that the gender based differences in the Immigration and Nationality Act did not survive the heightened scrutiny used to evaluate “all gender-based classifications” because the gender differential is based on the unjustified assumption that “the unwed mother is the natural and sole guardian of a non-marital child.” Id. at 8, 10.

Part 2: The Social Science Evidence

    A. What Did Amici Offer?
     Interestingly, respondent Morales-Santana did not rely on heavy social science evidence within his brief. In fact, respondent made passing remarks to “empirical evidence” within a footnote in his brief. See Brief for Respondent, Luis Ramon Morales-Santana, Sessions v. Morales-Santana, 582 U.S.___ (2017), at 24 n.8. Yet a group of leading population and family scholars and researchers filed a brief as amici curiae in support of respondent.
    All told, the amici offered twenty pieces of social science evidence. Brief of Population and Family Scholars as Amici Curiae in Support of Respondent, Sessions v. Morales-Santana, 582 U.S.___ (2017), at ii–iv. Generally, the social science evidence offered within amici’s brief could be classified according to the following categories:

• (1) Literature reviews, which summarizes the current evidence without reporting new work, like Professor Judith A. Seltzer’s article, Families Formed Outside of Marriage.

• (2) Population data, like surveys from federal agencies like the United States Census Bureau or the United States Department of Health & Human Services; and

• (3) Empirical, statistical social science analysis.

    Within that third category falls the principal study that amici relied upon in their brief: the Fragile Families and Child Wellbeing Study, by Sara McLanahan—one of the amici—and other researchers. This study informed much of the amici’s brief, including providing the dataset for other social science studies cited within the brief, so we have chosen it as an exemplar of the studies introduced and explore and analyze it in detail—
• Purpose of the Study: A descriptive study whose purpose was to discover the consequences of nonmarital childbearing for parents and children.
• Design: The study was a longitudinal survey—that is, a survey that repeats observations of the same variable over a length of time—that began in 1998. The researchers conducted initial interviews with the parents, and subsequently conducted follow-up interviews when the children were one, three, five, and nine years old.
• Sample: The parents of approximately 5,000 children born across twenty cities were chosen. In selecting this sample, the social science researchers performed the following procedure: (1) first, the social science researchers stratified seventy-seven cities with populations of more than 200,000 people; (2) next, sixteen cities were randomly selected and four cities were chosen for particular research interests; (3) finally, groups of hospitals were identified and beds within each hospital were randomly selected to obtain a sample of new mothers.
• Control: The parents selected for this study were a mix of married and unmarried couples; in fact, approximately three-quarters of the couples were unmarried at the time of birth. The married one-quarter served as a control group.

See Sara McLanahan et al., Fragile Families and Child Wellbeing Study (RWJF Program Results Rep., Jan 28, 2014), at 6–9 (2014).
     The initial interviews conducted involved questions designed to measure the expectations regarding the fathers’ rights and responsibilities, and the subsequent interviews inquired into the child’s health; the use of health care and child-care services; the parents’ relationship; experiences with welfare or child support agencies; and, for the ninth year, a DNA component. The purpose of this last measurement was so researchers could analyze any environmental forces that affected genetic development. Id.

     B. What Did the Fragile Families Study Reveal?
     The Fragile Families data revealed that “paternity establishment rates” (that is, the rate at which a father voluntarily signs a form that acknowledges that he is in fact the father) for nonmarital families were approximately 70%. See Brief of Population and Family Scholars as Amici Curiae in Support of Respondent, Sessions v. Morales-Santana, 582 U.S.___ (2017), at 10. In fact, approximately 81% of acknowledgements happened in the hospital. Id. Moreover, 50% of unmarried parents lived together at the time of the child’s birth, and over 80% of unmarried fathers are “very involved” during pregnancy and immediately after birth. See Fragile Families and Child Wellbeing Study, Fact Sheet 1. Indeed, across three different types of relationships, cohabiting, visiting, and non-romantic, the mother wanted the father involved 99%, 99%, and 74% of the time, respectively. Id.

    C. What Did Other Social Science Evidence Reveal?
     The Fragile Families study lined up with other surveys that revealed similar results that amici discussed in their brief. For example, the U.S. Census Bureau found that, in 2013, 648,000 custodial fathers were owed child support, and that custodial fathers were due $4.17 billion in child support from their children’s mothers. See Brief of Population and Family Scholars as Amici Curiae in Support of Respondent, Sessions v. Morales-Santana, 582 U.S.___ (2017), at 11. Moreover, the U.S. Census Bureau also found that, according to 2014 census data, 40,000 male couples are raising children together, and two-thirds of those couples are married. Id. at 11–12.

     D. How Did the Supreme Court Use This Data?
     The Fragile Families Study—indeed, all of the social science evidence that amici discussed in their brief—provided a comprehensive, widely applicable set of empirical findings with which to answer amici’s key question: Are nonmarital fathers regularly in parental roles at the time of their child’s birth? In light of this evidence, the answer is a resounding yes. How, then, did the Supreme Court handle this social science evidence? While discussing the way that overbroad, gender-based generalizations—though constitutionally impermissible—may in fact still describe “the way many people still order their lives,” the Supreme Court tucked one sentence into their thirteenth footnote:

In fact, unwed fathers assume responsibility for their children in numbers already large and notably increasing. See Brief for Population and Family Scholars as Amici Curiae 3, 5–13 (documenting that nonmarital fathers “are [often] in a parental role at the time of their child’s birth,” and “most . . . formally acknowledge their paternity either at the hospital or in the birthing center just after the child is born”)

Sessions v. Morales-Santana, No. 15-1191, 582 U.S. ___, slip op. at 14 n.13 (2017).

     Perhaps unsurprisingly, the Supreme Court disposed of the social science evidence in a footnote, without determining whether the evidence would pass Daubert or evaluating the soundness of the methodology employed. In light of the Court’s choice, we decided to take up the task of evaluating the Fragile Families Study ourselves.

Part 3. Our Analysis: How Does the Fragile Families Study Fare?

    A. Does It Pass Daubert?
     Assuming that this study was introduced at the trial level, it would be subjected to the Daubert test that governs admissibility of social science evidence. The touchstone for admissibility of social science evidence is whether the methodology used was sound. A common, though non-exclusive, list of factors considered in this inquiry are: (1) whether the theory or technique in question can be and has been tested; (2) whether the evidence has been subjected to peer review and publication; (3) the evidence’s known or potential error rate; (4) the existence and maintenance of standards controlling its operation; and (5) whether it has attracted widespread acceptance within a relevant scientific community. See Daubert v. Merrell Dow. Pharms., Inc., 509 U.S. 579, 589–93 (1993).
     Here, it is very likely that the Fragile Families study passes Daubert. The techniques in question—longitudinal surveys with stratified random samples—have been widely used throughout a litany of social science fields. Their efficacy and soundness is evident through their long and continued use. In fact, their continued and widespread use evinces a widespread acceptance with social scientists, the relevant scientific community. Furthermore, the Fragile Families researchers employed techniques like random sampling which had the effect of reducing potential error rates. Finally, both the Fragile Families study and related studies that make use of its dataset have been published in reputable, peer-reviewed publications. E.g., Ronald Mincy et al., In-Hospital Paternity Establishment and Father Involvement in Fragile Families, 67 J. Marriage & Fam. 611 (2005) (using Fragile Families’ dataset to evaluate paternity establishment rates).

    B. Is It a Sound Study?
     Additionally, we thought that analyzing the Fragile Families study in light of the concepts we have learned this semester would be instructive. To that end, we inquired into the study’s design, the threats—both internal and external—to the study’s validity, and the reasonable generalizations that might be made from this study.
     The design of this study, as mentioned above, is partly a time-series design because it studies responses at multiple times throughout the life of the study. But the presence of a control group (the married couples) served as a comparison group to compare nonmarital couples against. In that sense, the Fragile Families study is best thought of as a combination design: a quasi-experimental (because the cities were not fully randomly chosen), time-series, comparison group design. The time-series aspect of this design is particularly beneficial, because it allows the researchers to identify and control for some time-based changes that might affect the characteristics of the study.
     Next, we evaluated potential threats to internal validity; that is, threats that bear on the soundness of the methodology of the study itself and potentially inhibit any sound inferences. Here, the internal threats of testing and mortality were the most salient. First, the testing threat—where familiarity with a test might enhance performance—may be an issue here, because the parents might change their responses to conform with what they think the “correct” answers are; in other words, if the interview questions have asked about the child’s health or the parent’s relationship with one another, then the parents, as they encounter similar questioning during the child’s upbringing, gain greater familiarity with what might be “expected” of a good parent, and might conform their answers accordingly. Next, the internal threat of mortality—that is, when effects throughout a study might be due to the different kinds of persons who dropped out of a particular treatment group during the course of an experiment—is likely present here, because, as the Fragile Families study notes, participation in the study dropped over time. Response rates for married and unmarried eligible mothers fell from 90% in the first follow-up interview to 76% in the last; response rates for fathers fell from 74% to 59% for the same period. See Sara McLanahan et al., Fragile Families and Child Wellbeing Study (RWJF Program Results Rep., Jan 28, 2014), at 9 (2014).
But these drops in response rates are not so staggering as to render the study fatally flawed.
     We also evaluated the study’s external validity, which allowed us to consider whether the findings of the study could be generalized from the study-specific circumstances to a broader context. An inquiry into external validity turns on whether its findings may be generalized across settings, across persons, and over time. Because the study only focused on cities with populations of 200,000 or more people, we would hesitate to generalize these findings to smaller cities, which might not share similar characteristics with these larger cities, like access to child services, hospitals, or specialized medical clinics. Moreover, although the families were randomly chosen, there may be socioeconomic or demographic differences between these families and other families because of the kinds of families that may, for example, afford to live in larger cities. On the other hand, because this study began relatively recently (in 1998), we feel fairly confident in generalizing these results over time. This study is more likely to capture modern trends in parenting than, say, a study conducted earlier in the 20th century. Yet, on balance, it is reasonable to generalize this study’s findings to a broader context because of the soundness of its design, thoroughness of its data collection, and overall comprehensiveness.

Part 4: Critiques of the Fragile Families and Child Well-Being Study

     The Fragile Families Study (FFS) is highly regarded in the social science and legal communities, though some scholars have pointed out methodological weaknesses that detract from the study’s effectiveness. Cabrera et al. note the difficulty in locating “men who became fathers under extremely harmful conditions, or who refuse to acknowledge their children or do not know they fathered a child.” Cabrera et al., The DADS Initiative: Measuring Father Involvement in Large-Scale Surveys, in Conceptualizing and Measuring Father Involvement 417, 441 (Randal D. Day & Michael E. Lamb eds., 1st ed. 2004). Fathers under these categories are likely underrepresented in the FFS sample, constituting less than 10% of new fathers in the study. The authors also note that the FFS only collected data from biological fathers and interviewed the majority of fathers in the hospital when they were visiting the mother and new baby. See id. at 442–45. These methods exclude data on social fathers and skew the sample toward fathers who visit mothers in the hospital after birth. (Despite its weaknesses, hospital sampling yielded the highest response rates and was the most successful sampling method in the surveys studied.)
     By limiting its scope to city hospitals, the FFS sampling method also likely resulted in underrepresentation by socially- and economically-advantaged mothers. Robert L. Wagmiller, Jr., How Representative Are the Fragile Families Study Families?: A Comparison of the Early Child Longitudinal Study-Birth Cohort and Fragile Families Samples 5 (University at Buffalo, SUNY, Working Paper No. 2010-01-FF, 2010). When comparing data from the FFS to that of the Early Childhood Longitudinal Study, Birth Cohort of 2001 (ECLS-B), Wagmiller found that parents in the FFS sample reported lower mean household incomes and fewer years of completed education. They were also more likely to be African-American and less likely to be non-Hispanic white. Id. at 16. Between the studies, fathers in the FFS were more likely to be in “very good health” and be employed but earn less. Id. These differences stemmed from the studies’ sampling and data collection strategies: to minimize non-response bias and encourage response from non-resident fathers, the FFS conducted interviews specifically during the “magic moment” shortly after birth when fathers were most likely to be present and agree to participate in the study. As noted before, these interviews were limited to city hospitals. The ECLS-B sampled births from large cities, suburbs, and rural areas, and collected data by interviewing mothers and fathers and assessing children about nine months after the child’s birth. Id. at 5. (This method likely resulted in underrepresentation by non-resident fathers—past studies have shown great difficulty in collecting data about unwed fathers after the child’s birth.) Despite these differences, Wagmiller notes that between the studies, overall distributions of children and families were similar. Id. at 17.