Health, Law, And Ethnicity: The Disability Administrative Law Judge And Health Disparities For Disadvantaged Populations

Social determinants play into who gets to die prematurely while others get to have healthy productive lives—these are loosely called health disparities. Health disparities are typically understood socially, economically, and politically, but rarely analyzed within the legal system.

The Social Security Administration (SSA)—the federal program for providing Americans with disabilities benefits and resources—recorded that in 2018 approximately ten thousand people died while waiting for the backlogged SSA and SSA administrative law judges (ALJs) to decide on whether they deserved aid for their disability. In regard to this death toll, the president of the association of ALJs expressed frustration about inadequate Congressional funding: “I know that people will die waiting . . . We have decided it’s better for people to die than to adequately fund this program.”[2]

SSA is the primary accessor of the propriety of denials by ALJs that otherwise go unchecked. But studies show that ALJs incorrectly decide SSA benefit applications at high rates without a properly functioning check on ALJ bias. Because an incorrect decision can lead to the worsening health, and even death, of the claimant, it is critical that we understand the factors that impact ALJ decision-making.

The legal system is a contributing factor to health disparities. In this Note, I explain this by analyzing a legal aid database of plaintiff- side SSA disability cases via a mixed methods study utilizing statistics and content analysis.[3] I find that Black claimants are statistically more likely to be rejected for disability aid over their peers—this mirrors what the SSA found in their prior reports. However, SSA reports fail to explore in depth why these results occur. Here, I find that the higher rejection rates for Black claimants is due to two primary factors. First, a minority of judges disproportionately deny Black claimants at significantly higher rates than their ALJ peers. Second, some ALJs rely on medical documentation that does not reflect surrounding factors like implicit bias and cultural context. Low-cost solutions can combat these inconsistent outcomes. For example, the SSA has previously instituted implicit bias training, which although rarely provided, found great success in rooting out ALJ bias. An even more cost-effective method is releasing redacted informative data for independent researchers to analyze. The alternative is the continued perpetuation of a public health crisis caused by courts and judges, not by hospitals and doctors, and the continued deleterious health effects on disabled claimants.

Table of Contents Show

Introduction

John Jordan,[4] a Black high school dropout, was homeless when he became the program manager of a non-profit organization that helped other homeless people obtain housing. He later supported at-risk youth by managing a non-profit that provided mentorship and community service. After serving his community for many years and turning his own life around, John was diagnosed with HIV. Soon after, John was divorced and then incarcerated for driving without a license. In prison, guards mocked John because of his HIV status. Guards publicly announced his illness to the prison and quarantined him in solitary confinement without a flushing toilet. John was subsequently diagnosed with posttraumatic stress disorder, depression, and generalized anxiety disorder. Upon release, John was ashamed of his HIV, and developed psychological “feelings of worthlessness and guilt” along with thoughts of “giving up” and “not want[ing] to be here anymore.”[5] He was “unable to engage in the type of community work that he was passionate about prior to his diagnosis.”[6]

John applied for disability aid through the federal Social Security Disability Insurance Program (SSDI). John was physically deteriorating with a T-cell count so low that he was constantly at risk of death. After waiting four years for the opportunity to prove his disability before an ALJ, the judge ultimately denied his claim. It is unknown what happened to John afterwards.

John’s denial as a Black claimant is reflective of a government study by SSA, which found that Black claimants are statistically more likely to be denied over other groups.[7] It is impossible to investigate whether John’s case was properly decided because disability cases are not publishable due to concerns over patient privacy and health care data regulations. But SSA cases are being denied improperly.[8] Thousands of denied claimants appeal their cases to Article III courts where, about 50 percent of the time, the judge will find that the ALJ grossly misapplied the law.[9] When the ALJ misapplied the law by exhibiting bias, safeguards for claimants are almost always inoperative,[10] which is especially concerning because administrative courts lack the procedural safeguards that are normally available in state and Article III courts.[11] For instance, Article III courts provide claimants an appeal process to ALJ bias that must meet higher constitutional due process requirements, whereas the SSDI system’s appeal process is a standard so high that my research found one case where the standard was met.[12] Thus, ALJs function unchecked for unconscious behaviors over the SSDI system.

The lack of a check is concerning when a potentially erroneous denial could lead to the worsening health, and even death, of a claimant due to inability to afford food, shelter, and medicine. Improper denials are especially significant for Black claimants who face higher health disparity rates than any other group, including higher rates of mortality, higher rates of numerous serious illnesses, and shorter life spans.[13] Given ALJ’s immense responsibility over eleven million Americans and its budget of over $150 billion dollars in the 2019 fiscal year, it may be surprising that there are almost no independent studies of disability ALJs.[14]

In this study, by using a mixed methods approach that includes both a statistical and content analysis, I find that the legal system is perpetuating health disparities by improperly rejecting Black claimants for disability aid at higher rates than their peers. These results help to illuminate a need for: (1) Congressional and SSA intervention in the disability hearing process; (2) more institutional study; (3) funding for independent research; (4) implicit bias trainings for ALJs; and (5) re-examination of implicit biases in disability hearings. This study also supports practitioners, especially in disability law, by providing information that helps them to better understand their clients and the underlying cultural circumstances that may impact their adjudicatory outcomes.

This study proceeds in four Parts. In Part I, I discuss the history and progress of health disparities and health disparities research. In Part II, I provide a broad overview of SSA and the ALJ hearing process, introduce background information about ALJs, and explain the complexity of the system for claimants both to gain disability aid and to allege bias by an ALJ. I highlight the lack of ALJ studies and explain why the system for claimants to allege bias is inoperative as a functioning check. In Part III, I detail the study’s methodology and results, which includes a statistical finding that ALJs deny Black claimants at significantly higher rates. In the content analysis section, I find that three ALJs (out of twelve ALJs total) were responsible for 77 percent of denials of Black claimants—those same three ALJs only granted 7.6 percent of Black claimants. I find that the result may be due to a lack of cultural understanding, physicians underreporting the pain of Black patients, and Black patients being more reluctant to speak about their pain because of factors such as trauma. Thus, these three ALJs may receive medical records that do not accurately reflect the disability of Black claimants, contributing to their improper denial. In Part IV, I discuss how the SSA, ALJs, and practitioners can work to remedy the current public health crisis of the hearing process.

I. Health Disparities: History, Study, and Progress

Health disparities are the health differences that disadvantaged populations face versus non-disadvantaged populations.[15] The American population facing disability is large, with 12.8 percent of Americans living with a disability.[16] Of that 12.8 percent, those who come from disadvantaged backgrounds, including certain ethnic groups and those from lower incomes, will suffer higher rates of disablement and mortality.[17] Conversely, wealth and higher education levels are associated with longer lifespans and better health.[18] Disadvantaged populations can also include people from different genders and sexual orientations.[19] Disadvantaged populations typically face health disparities as a result of social, economic, or environmental disadvantages, which can include poverty, poor education, and discrimination.[20] Overall, health disparities are complex, and they result from a powerful and complex relationship between factors such as biology, socioeconomic status, racism, and legislative policies.[21]

Health disparities impact Black Americans at higher rates than all other tracked groups and are linked to historic inequality in power structures. For example, health disparities are historically and contemporarily ingrained in the civil rights discourse. In the 1900s, W.E.B. DuBois and Dr. Martin Luther King Jr. were prominent voices in the health disparities discourse. Martin Luther King noted that that “of all the forms of inequality, injustice in health care is the most shocking and inhumane.”[22] W.E.B. DuBois noted that the higher levels of poor health for Black Americans were connected to social inequalities such as the “‘vastly different conditions’ under which blacks and whites lived.”[23] Further, DuBois advocated that high mortality rates in minority groups would improve with better health and education.[24] Today, Section 1557 of the Affordable Care Act tries to address this disparity by building on federal civil rights law pursued by the civil rights leaders. It prohibits health care discrimination, expands access to care, and eliminates barriers to healthcare access.[25]

Health disparities were first given prominent academic attention in a 1986 study that found Black Americans faced “excess deaths” when compared to White Americans.[26] Since then, discrimination due to bias has arisen as a public health threat. Discrimination harms both the target and the actor: it leads to health disparities due to changes in behavioral and psychophysiological responses to stress, changing the individuals interactions with others, and increasing the risk of stress-related disorders.[27] Psychosocial stressors due to bias can lead to unequal education, economic opportunity, and material resources, which is in turn a pathway to other health-related illness such as high blood pressure, hypertension, or diabetes.[28]

Researchers have yet to fully explain the factors that create and maintain health disparities but they suggest that implicit bias is a substantial source of health disparity.[29] For instance, physicians are more likely to provide higher quality care to White patients as opposed to their Black counterparts.[30] But conducting health disparities research is challenging because databases are difficult to obtain and typically do not include socioeconomic data.[31]

Through the years, substantial progress has been made to help alleviate health disparities through trainings to help professionals recognize their implicit biases. In medical institutions, physicians that undergo training and acknowledge their implicit biases are more likely to contribute positively to alleviating health disparities.[32] In the legal setting, similar trainings have helped reduce implicit bias.[33] Given the success that implicit bias trainings have had, their further implementation would help greatly in alleviating health disparities by physicians and by ALJs who facilitate health-related legal causes.

In the next Section, I introduce how the legal system is perpetuating health disparities through the SSDI system. Specifically, I explain the numerous barriers that low-income, disabled claimants face when dealing with the complex administrative system, and the inadequate check for ALJ bias in the process.

II. SSDI: Barriers, Bias, Dysfunction, And Obscurity

SSDI is part of the Social Security Act, which academics posit as the most important social justice law because of its ability to uplift marginalized communities.[34] SSDI has been “remarkably successful” in lifting low-income disabled people out of poverty by providing access to food, shelter, clothing, healthcare, and protection for children.[35] Conversely, those denied are likely to remain in poverty. Keeping groups who are in need out of the program harms the general public interest that is best served by the provision of essential services.[36]

ALJs are “presumed to be unbiased” by reviewing courts[37] despite unavoidable human unconscious biases that permeate our institutions.[38] Yet, claimants who are Black, women, and non-English speaking disproportionately suffer in the SSDI process, encountering barriers to aid and implicit biases.[39] This bias in ALJ judges is almost completely unchecked.[40] Moreover, the constitutional Article III courts’ due process requirement is absent in administrative judicial hearings, leading to an “absence . . . of procedural safeguards normally available in judicial proceedings.”[41] The result is an increased danger of ALJ bias going unchecked without recourse for claimants.

Part II proceeds in three parts. In Part II.A, I describe the claimants’ barriers to accessing SSDI, including socioeconomic barriers such as navigating a lengthy and expensive application process without the provision of an attorney. In Part II.B, I describe implicit bias among ALJs. In Part II.C, I discuss the inoperative process for claimants alleging ALJ bias through the system provided by SSA and through the federal courts.

A. Barriers and Application Complexity For Low-Income Claimants

SSDI claimants experience specific socioeconomic barriers throughout the SSDI application process. For instance, some claimants from low-income backgrounds are unable to attend hearings due to a lack of childcare or transportation.[42] Other claimants lack the ability to read or write in English, which is especially problematic because disability notices need not be written in any language other than English.[43] The result for non-English-speaking or illiterate claimants may be denial of SSDI because of misunderstanding the date of a hearing, the instructions for a hearing, or the adjudication of a hearing.[44] Furthermore, claimants must seek their own attorney because the SSA is not required to provide legal support.[45] Without an attorney, low-income individuals are left alone to face a complex system that Congress lightly describes as intricate,[46] and others critique as “unintelligible.”[47]

Assuming SSDI claimants are able to overcome initial barriers, they will ultimately be denied a majority of the time once they reach the ALJ stage.[48] Consider the complexity of the process. Claimants begin applying for disability aid with an initial application with an SSA hearing officer[49] who will make a determination on the merits of the claim.[50] The denial rate at this initial stage is 77.2 percent whereupon the claimant may request reconsideration with another hearing officer where nearly all claimants will be denied again at a rate of 92.9 percent.[51] If denied there, a claimant may appear before the ALJ.[52] Appearance before the ALJ occurs about a year later.[53] 62.6 percent of claimants at the hearing will be rejected once more.[54] By the time they appear before an ALJ, 90 percent of all claimants will have been denied.[55] Throughout this multiyear long process, the health of claimants may worsen, potentially leading to death. After this process, the claimant may then appeal their case to an ALJ.

The ALJ decision-making process is complex and expensive, creating another barrier for low-income claimants. This is especially true for those that cannot afford an attorney. There is no constitutional requirement for the state to provide an attorney in these administrative hearings, and claimants must cover their own cost of evidentiary discovery.[56] The ALJ begins by reviewing all of the evidence of the disability, including medical records and physician testimony, with the support of an SSA vocational and medical expert.[57] The ALJ then determines if the claimant is truly disabled,[58] such that the claimant is so physically or mentally disabled that the claimant is unable to do any work.[59] The ALJ decides this by looking at the totality of the circumstances in a five-step analysis. The claimant has the burden of proving in the first four steps that they are currently unemployed, that their disability is severe, that their disability will last for a year or more, and that their disability prevents them from doing any past work in which they engaged.[60] Claimants may have difficulty here because they can only prove what is in their medical record, and if the treatment is not extensive or well-articulated on the record, they are very likely to lose.[61] A fifth step shifts the burden back onto the ALJ, who must establish that the claimant is unable to do any work.[62] However, the ALJ does not have to defer to the claimant’s medical testimony, and may decide independently. This is problematic because ALJs are not medically trained.[63]

Through the five-step process, low-income claimants who cannot afford an attorney must represent themselves as pro se claimants in a difficult legal process. Claimants must cover their own costs, which includes proof of disability derived from various sources. This can entail having to pay for their own discovery and provide hundreds of pages of costly evidentiary materials including medical records, physician testimony, doctor’s letters, mental health letters, witness letters, caregiving notices, workplace notices, etc. For the low-income litigant fighting for their own life-sustaining aid without any legal training, this can be an emotional, scary, and intimidating process that includes having to go through a complex process that requires motions, objections, and general legal knowledge.[64] These claimants may be suffering from poverty, advanced age, low educational levels, inability to see or hear, mental disability, and general disability, which can inhibit their capacity to articulate and present a coherent argument in court.[65]

Even after a claimant wins their case, SSA may still terminate their benefits, to which the claimants again carry the burden of covering the cost of discovery and providing evidence that they are still disabled.[66] Low-income disabled claimants may be unable to meet the burden of the continued fight for benefits due to barriers they face in their life, such as the illness of which they are trying to gain relief from, lack of access to transportation, or inability to afford an attorney.[67] To complicate the already difficult process of gaining disability aid, certain claimants must also overcome implicit bias, which I discuss in Part II.B.

B. Implicit Bias Among ALJs

In addition to facing structural burdens throughout the application process, many SSDI claimants face the additional hurdle of bias. Judges have implicit biases which can result in disparate outcomes for Black and other ethnic groups. Psychological measures of implicit biases find that groups hold unconscious racial biases against Black Americans.[68] For instance, a 2014 study found that the lives of Black Americans are more likely to be devalued and deemed worthless by juries in death penalty states.[69] In the criminal justice context, Black Americans are implicitly associated with retribution, while White Americans are implicitly associated with mercy.[70] Judges hold these same implicit biases and these biases may influence their judgment.

The SSA indicates that those most often improperly denied are claimants who are Black, women, and non-English speaking. In 1992, Congress requested an investigation of the apparent disparity in Black claimants’ denials. The investigation found that Black claimants obtained favorable outcomes only 55 percent of the time as compared to White claimants who obtained a favorable outcome 66 percent of the time.[71] These findings were unexplained and deemed by the SSA to be unknown,[72] suggesting that other factors, such as racial or gender bias, may be at play.[73] As far as my research has found, no other large-scale studies exist on ALJ hearing outcomes based on ethnicity or bias.

Black claimants at the intersection of gender and national origin are even more likely to face bias. A study found that women may experience ALJ bias in the form of harmful interpersonal interaction with the ALJ, such as by the ALJ imposing gender stereotypes onto female claimants, or an ALJ not believing female claimants based on a female stereotype.[74] Bias also plays a role in the way non-English speaking groups are perceived.[75] For example, an ALJ exhibited “cavalier and insensitive treatment” specifically towards a non-English-speaking Hispanic claimant, including “sarcastic comments, constant interruptions and callous disregard . . . leav[ing] th[e] court dismayed.”[76]

In response to the study, the SSA began training ALJs in cultural awareness and diversity, a program which found success.[77] Some ALJs verbally resisted the training due to their pride and self-perception as objective adjudicators.[78] However, ALJs undergoing the training—including those that opposed the training—recognized at least some of their biases and preferences.[79] Despite the program’s success in rooting out possible biases, the trainings occur infrequently, likely due to lack of funding.[80]

Overall, there is very little other research on SSA disability adjudications, so we must turn to older studies for support that the overall population is being incorrectly adjudicated. A 1976 study by the General Accounting Office (GAO) found that two SSA officers presented with the same case would decide differently from each other 24 percent of the time, meaning only one of them would have been correct.[81] A subsequent 1978 study by the GAO found that decisions were being subjectively decided and that uniform outcomes may never be possible.[82] An academic study of ten states found similar discrepancies; the disposition of the same case would be agreed upon by all ten states only 22 percent of the time.[83] Professionals also consistently disagree with the SSA on correctness of case outcomes. An independent clinical assessment disagreed with the SSA’s decisions 33 percent of the time.[84] Despite the inconsistent decision-making outcomes, administrative attempts to improve disability decision-making carries a wider concern by Congress, Article III judges, and scholars as “impermissible political interference.”[85]

The next Section explains that despite implicit bias in our systems and some question as to the rate of correct adjudications, claimants are unable to seek relief from improper adjudications when they face implicit bias.

C. Inoperative Checks for ALJ Implicit Bias

Claimants who face implicit bias in the ALJ adjudicatory process will find that the mechanisms to seek relief are inadequate. Claimants can bring their allegations of ALJ impropriety to the SSA or a federal court. In Part II.C.1, I explain why the SSA process is inadequate. In Part II.C.2, I explain why the federal court system is inadequate. The result is that claimants are, in essence, unable to seek relief from implicit bias by an ALJ.

1. Inoperative SSA Process

ALJs primarily misapply the legal standard because they inappropriately deem the disabled claimant to lack credibility.[86] The SSA does not release data to help explain why ALJs consider some claimants to lack credibility with any specificity. One possibility is that ALJs may perceive claimants from backgrounds and genders outside of their own to be less credible due to unconsciously biased views about them. Another possibility is that ALJs have too many cases and lack support to give each case adequate scrutiny. ALJs adjudicate a docket that is about ten times larger than the combined federal docket,[87] amassing 330,000 cases in 2019 alone. So, it is understandable that an ALJ may not have the bandwidth to give each case sufficient scrutiny.[88]

About nine thousand denied claimants appeal their case to a federal court where the federal judge will find that the ALJs grossly misapplied the legal standard about half the time.[89] But there is little academic study seeking to understand the reasoning behind improper adjudications. The SSA itself stated that “[t]he literature evaluating factors that affect allowance rates in Social Security’s disability programs is extremely sparse.”[90] A congressional report aimed at demystifying the SSA hearing process acknowledged a lack of available data. The report stated that “[a]lthough there had been attempts in the past to shed light on aspects of the disability programs, those attempts were often hampered by the lack of available data to help those outside of SSA understand how the disability programs operate.”[91] However, the report failed to provide any detailed demographic information.[92] This lack of detail is not an anomaly. ALJs receive so little systematic study that they have been “aptly named the ‘hidden judiciary’” and the “invisible judiciary.”[93] The lack of ALJ study is alarming considering that ALJs are tasked to use their ample discretion to assess the credibility and deceitfulness of a claimant,[94] which inherently involves some human bias.[95]

Claimants who experience implicit bias will find it nearly impossible to succeed on a claim due to the strangeness of the process and the great difficulty in finding evidence of covert bias. The process for claimants to allege bias is peculiar. A claimant requesting investigation of ALJ bias must present the allegation to the same ALJ being accused. Then the claimant must await the ALJ’s decision as to whether the ALJ believes they were biased or not.[96] But evidence of bias is nearly impossible to find in a singular case because bias may only occur once or twice on the record. [97] To compound the difficulty, those few instances of bias on the record may appear with subtlety, covertness, and unbeknownst to the ALJ. Furthermore, even if a claimant successfully finds a few instances of implicit bias on the record, such a showing is too vague for a judge to admit to their own bias. Nor can a claimant introduce evidence outside the record. This means that general bias, meaning bias recurring across several cases, is impossible to prove.[98] The result is that the SSA, which states that all people from different backgrounds should be treated with neutrality,[99] fails to provide a functional safeguard to protect vulnerable claimants that experience covert bias.[100]

2. Inoperative Federal Court Process

Once claimants exhaust recourse through the Social Security Administration’s process, they may bring their allegations of ALJ bias to the federal courts. The federal court’s system functions similarly to SSA’s system, providing an inoperative check on bias. Claimants have a “substantial burden”[101] to prove that an ALJ’s behavior was so extreme that it displayed a “clear inability to render fair judgment.”[102] This burden is not met when an ALJ displays only isolated instances of extreme behavior. But rather, the record must show as a whole that there was fundamental unfairness.[103] In addition, claimants may not introduce evidence outside the established ALJ record, except in very limited circumstances.[104]

The federal court standard for ALJ bias was met only once, as far as my research has found, in Grant v. Commissioner, Social Security Administration. Here the court found the ALJ to be biased due to a lengthy history of explicitly racist and derogatory claims.[105] The ALJ held numerous deeply racist beliefs. On the record, the judge stated that Hispanics often pretended not to understand English, that certain ethnic groups often faked mental illness, and that all African Americans should return to Africa.[106] The judge also used racial slurs when refering to claimants.[107] The ALJ often gave no credibility to Blacks, Hispanics, and poor Whites because of the ALJ’s belief that they were drug addicts and alcoholics bent on not working and preferring to live on public money.[108]

The federal court found bias here, whereas the SSA’s internal investigations found none. The SSA conducted an internal investigation of the ALJ, analyzed 212 of the judge’s cases, and found no evidence of bias against claimants.[109] They gave no explanation for their finding. This outcome is curious considering that the investigation also concluded that sixty-nine out of eighty-two denials by the ALJ exhibited some unlawful behavior.[110] Given the court’s finding of bias, it may be that the SSA’s internal investigations are also following a substandard process.

In other cases, the district courts consistently avoided or denied a finding of bias. In Pronti v. Barnhart, a judge refused to decide whether an ALJ was biased against claimants, despite an SSA investigation finding general ALJ unfairness.[111] There, the court requested the SSA investigate possible ALJ bias and unfairness. The SSA conducted a complex fifteen-month investigation, but failed to comply with the court’s orders to assess bias. Instead, the investigators searched only for unfairness, which they conclusively found.[112] As a result, the federal judge refused to adjudicate on a finding of bias. The judge cited, among other reasons, that a decision would serve “no useful purpose” and that the claimants were not harmed by the lack of an assessment of bias.[113] Many other cases similarly punted on the question of bias, instead finding that an ALJ was generally unfair.[114] Still other cases simply found neither bias nor unfairness.[115] Claimants, absent one found exception above in Grant, always lost on the question of biases that existed covertly and subtly.

Even statistically significant evidence of ALJ bias was not enough to prove bias in Article III courts. In Perkins v. Astrue, a law firm introduced statistically significant results that an ALJ held a bias against obese claimants with mental impairments.[116] The court, guided by Doan v. Astrue, rejected the results. The court reasoned that without evidence of bias in the present case, statistically significant data alone was insufficient.[117] Similarly in Doan, a statistically significant variance in an ALJ’s adjudication rates when compared to the national average was insufficient to establish bias without other evidence. The court determined that bias was not the only basis for the variance. But rather, the court determined that the unique mix of the ALJ’s cases could explain the higher denial rate.[118] In both cases, the courts reasoned that statistically significant data, without evidence of bias in the case at issue, was insufficient.

In the next two Parts, I find that inadequate checks on ALJ bias contribute to the perpetuation of health disparities within our legal system. The study empirically identifies that improper disability aid denials cause public health disparities. The study finds that a minority of ALJs deny Black claimants at significantly higher levels. The data reveals that ALJs unconsciously misinterpret medical records and that Black claimants resist admitting to their subjective health related pains.

III. Dataset, Methodology, And Results

This study conducts the only two-tier statistical and substantive study of SSDI ALJ decisions. The data for this project comes from four months of work at a non-profit legal aid center in a major metropolitan area. During this time, I worked at the center as a student attorney. As a student attorney, I interacted with disabled clients, assisted disabled clients with their legal cases, wrote a brief for an ALJ hearing, and attended the hearing for that case. Through these experiences, I gained a personal understanding of the documented and undocumented narratives of the disabled claimants from this study. This includes an emic and etic perspective (from within the social group and from the outside of the social group) on the physical pains, subjective struggles, and life histories of the disabled claimants.

The study explores whether claimants in the dataset suffer outcome differences in their cases based on their ethnicity. To do so, I analyze whether there is a statistically significant difference in the outcome of claimants based solely on their ethnicity. Statistical significance only provides a partial picture. Thus, I also interpret the statistical results with a content analysis of every case in the dataset to find factors that might explain discrepancies in case outcomes. The analysis reveals that Black claimants are statistically more likely to lose their case. In addition, the case readings indicate that the medical records of Black claimants might be misinterpreted or misunderstood due to cultural factors.

A. Descriptive Statistical Method

The Pearson Chi Square test was used to discover if differences in outcomes from the data, via the null hypothesis, were by chance or statistically likely to be replicated. The data itself was presented in a raw format, which required extensive cleaning. The data was presented in an excel spreadsheet, with the total number of n=55,000 cases. However, this number included every single case for every single matter by the Health Office. Upon further review, the number of cases with the variable relevant to ALJ hearings was n=537. A further review of these cases was required because the variable relevant for ALJ hearings is used concurrently with other cases unrelated to ALJ hearings, such as initial level hearings. After extensive review of the physical files of each client, n=136 cases were used for statistical analysis.

The data was again considered for its variable categories for study, which included the following categorical variables, most of which were irrelevant for our study: office, last_name, first_name, problem, sp problem, zip code, city, gender, ethnicity, hispanic, and close_code. The office variable included every office at the legal aid, which provides numerous services, such as immigration. This study is only concerned with disability cases, so only data from the Health and Welfare office was considered. The variables for the Health and Welfare Office were HV (HIV) and HM (MLP), which reflect the program clients were in. Both variable groups were predominantly from Medical-Legal Partnership referrals, which are referrals from different clinical settings to the legal office. The variables “problem” and “sp problem” were intended to identify specific problems in the cases. However, the variables were too inconsistently recorded to be of any statistical use, and were thus disregarded. “Zip code” and “city” were also disregarded, although future study may look to analyze results by location. “Gender” was included as a variable. I replaced the values of male with a “0” and female with a “1.” “Ethnicity” was included.

Ethnicity played a significant factor in outcomes. The data presented a number of ethnicities: American Indian and Alaskan Native; Asian; Black or African American; Native Hawaiian/Other Pacific Islander; White; American Indian or Alaskan Native and White; Black or African American and White, American Indian or Alaskan Native and Black or African American; Balance/Other; Unknown; Asian and White; and Multi-Ethnic. Note that Hispanic was not included in this group and instead was separately annotated. From these groups, I cleaned the data into usable groups, which became “White” (data value 1), “White-Hispanic” (data value 2), “Asian” (data value 3), “Hispanic” (data value 4), “Black or African Americans” (data value 5), “Multi-Ethnic” (data value 6), “Multi-Ethnic Hispanic” (data value 7), and “Balance/Other.” The Hispanic variable was of particular concern, which was difficult to collect due to its separate category. I suspect that the size of this category is larger than collected data suggests. I decided to run a regression coefficient to see if the group value changed when further collapsed. I proceeded to collapse those who identified as “White-Hispanic” and “Multi-Ethnic Hispanic” into the “Hispanic” group. No change was found.

The final values for ethnicity were as follows: (1) “White,” (2) “Asian,” (3) “Hispanic,” (4) “Black or African American,” (5) “Multi-Ethnic,” (6) “Balance/Other.” Any group with even partial Black or African American affinity was categorized as “Black.” I annotated all Black and mixed Black into a single field. I annotated Asian and mixed Asian into a single field. White was given its own category. Native Americans, Native Hawaiians, and Pacific Islanders were combined into the Other/Balance category. Any mixed Hispanic ethnicity was collapsed into the Hispanic category. However, as an exception, those who identified as both Hispanic and African American were collapsed into the African American category.

Close codes included numerous values, but for purposes of this study, I only considered four values: “Administrative Agency Decision–Favorable,” “Administrative Agency Decision–Unfavorable,” “Court Decision–Favorable,” and “Court Decision–Unfavorable.” Data points with missing categories were disregarded. The legal aid department I derived the data from only functions in administrative hearings, so any report under “court hearing” refers to administrative court hearings. All favorable decisions were collapsed as a “1,” and all unfavorable decisions were collapsed into a “2.” Ultimately, these data points were too inconsistently reported, so I manually checked every single data point to ensure that there was a case file associated with that data point and that it was for an SSA disability client file.

B. Content Analysis Code Index

As the courts have long stated, even if statistical significance is found, such a finding is not enough on its own to indicate, at least by law, any bias.[119] As such, I conducted a content analysis within the cases to search for variables, which contextually explain the outcomes. The legal aid office from which this dataset was obtained has a cloud with computer files for every client. Every client file was redacted before extensive search, to protect the identity of the client. The content analysis specifically sought to find out why certain clients lost their case. The searched files included the briefs submitted to the judge, denial or approval letters from the judge when available, letters to doctors, letters from doctors, letters to client, letters to the judge, medical records, psychological records, police reports, and arrest histories. I searched for five content analysis codes: (1) severity of disability, (2) whether specific judges were more prone to adjudicating a certain way towards specific groups, (3) indications on the record of bias by the ALJ, (4) reluctance by the claimant to speak about their illness, and (5) independent factors. These independence factors addressed whether the claimant was incarcerated, had taken drugs, died, withdrew due to fear, or never returned.

C. Results: Black Claimants Are Denied At Statistically Significant Rates Explained By Three Factors

With statistical significance, I found that Black claimants lost their cases at higher rates than all the other claimants combined. The results indicated a chi-square statistic of 4.3956 with a p-value of .036033. The result was significant at p < .05.

The content analysis revealed three interesting results, especially as to Code 3 (indications on the record of bias by the ALJ). Firstly, Black claimants were denied at statistically significant rates, handed out disproportionately by a minority of ALJs. Out of the twelve ALJs sampled in this study, 77 percent of the denials to Black claimants came from three ALJs. Secondly, those same three ALJs accounted for only 7.6 percent of the awards to Black claimants—all of whom were among the most ill of the entire data pool. Disability in these cases was most obvious because these Black claimants were consistently among the sickest of the entire group.

To take one example, Sam was a Black claimant awarded by one of these three ALJs. But, Sam also had a T cell count of twenty-six (200 and below is considered AIDS), which was the lowest found in the study. Another example is Sarah, who had such severe paranoid schizophrenia that she consistently thought the devil was speaking to her. Another schizophrenic claimant consistently stated that he saw or heard unspeakable things, such as demons that asked him to hurt himself or to burn things. Absent severe illness, Black claimants were significantly more likely to be denied by these three ALJs. These results are consistent with the findings in academic literature studying the bias Black Americans face in clinical and judicial settings, such as by being undertreated for pain relative to other groups.[120]

The majority of denied Black claimants had medical records that misunderstood and underestimated their pain and illnesses under Code 1 (severity of disability) and Code 4 (reluctance by the claimant to speak about their illness). The study revealed that this is likely due to physician bias and claimants’ subjective discomfort with expressing their physical and psychological pain. The literature on non-legal settings finds similar results.[121]

Physicians were underreporting the pain of their Black patients Code 1 (severity of disability). The Black patients in this dataset were among the sickest of the entire data pool of this study. They had higher rates of HIV, lower T cell counts, higher rates of depression, and more severe psychological trauma. Yet, the physicians’ documented narratives failed to describe the higher pain experienced by these Black claimants at appropriate levels, as opposed to other groups. For example, the physicians consistently used similar language for patients with the most serious forms of trauma and for those with the least serious forms of trauma. The medical record language could influence an ALJ to conflate the severity of illness for two dissimilar patients, despite the medical differences in the seriousness of their respective illnesses. Content analysis Code 1 found that physicians of Black patients consistently utilized narrative language in their medical charts that was similar to that used for other less-ill patient groups—despite the fact that Black patients had higher T cell counts or records of more prevalent health issues. This finding coincides with studies that show physicians tend to misunderstand, underestimate, stereotype, and undertreat the disability of Black patients.[122]

Black claimants in the study were more likely to not talk about their hardship as a coping mechanism to avoid reliving the traumatic emotional pain under Code 4 (reluctance by the claimant to speak about their illness). For example, Mary expressed that she “disliked discussing her depression unless absolutely required to because ignoring her depression has become a coping mechanism to deal with the painful effects of her impairments.” Mary also “declined to report the extent of her condition . . . [and] complete depression questionnaires to medical providers for fear of being further medicated.” The ALJ presiding over Mary’s case perceived the lack of medically documented pain as grounds for dismissing her application. From the sample, proportionally, Black patients were more likely to provide language to their physician indicating reluctance to admit to their pain.

The record had little to offer for Code 2 (whether specific judges were more prone to adjudicating a certain way towards specific groups) and Code 3 (indications on the record of bias by the ALJ). This is likely because direct hearing transcripts were not available. Only the attorneys’ hearing notes or anecdotal evidence from the claimant could serve as evidence of biased ALJ language.

Overall, results indicate that Black claimants are being denied at higher rates than their peers, likely due to medical records that are inadequate. There are a number of interesting conclusions to draw here, but the overall results indicate that improvements can be made to offer better health outcomes.

IV. Fixing the Public Health Crisis of The Hearing Process

Disability hearings are a covert public health threat to Black claimants due to their higher denial rates and incomplete medical records. Black claimants who apply for SSDI are already heavily impacted by other societal health disparities.[123] Such disparities include higher death rates,[124] higher mortality rates from disease,[125] and shorter average life spans.[126] Furthermore, Black claimants face increased health disparities due to discrimination. Discrimination deteriorates the health of the target via psychosocial stressors that increase illness rates, disadvantage educational and economic outcomes, and arouse susceptibility to risky behaviors.[127] There is also support that suggests health disparities can be passed down generationally. A study of Holocaust survivors who suffered racism and their children found that their offspring may suffer behavioral and psychobiological harms.[128] Already struggling with these health disparities, Black claimants enter the disability application process facing even more obstacles—higher denial rates at the hands of a minority of ALJs and medical records that are incomplete and inaccurate.

Based on this dataset, I make four recommendations that address this public health crisis to ALJs, the SSA, legislators, and practitioners. They are as follows: ALJs might consider the cultural biases and norms of claimants, ALJs should be provided with implicit bias trainings, policy-makers can make low-cost changes to improve outcomes, and physicians and attorneys on the front-lines of caring for claimants should understand that medical records might be unreflective of the claimants’ real pain.

First, ALJs should consider cultural norms when evaluating a record’s documentation of pain and suffering. Health disparities begin with the way physicians interpret their Black patients’ health. Physicians may engage in covert healthcare discrimination by unconsciously providing inferior care to Black patients[129] or by underreporting the pain of Black patients.[130] As noted above, this study found a similar effect—physicians are underreporting the pain of their Black patients.

In turn, ALJs are presented with medical records that do not reflect the true severity of the claimants’ pain and suffering, leading to the denial of otherwise disabled claimants. This effect is compounded by Black claimants’ subjective discomfort with expressing their physical and psychological pain. Consequently, their reluctance to speak about their pain factored into ALJ denials. However, the outcome for these claimants may have been different if the ALJs gave weight to these cultural biases and norms. Thus, when a record may at first appear to lack documentation, the ALJ should consider the totality of the circumstances. For instance, an ALJ may consider a physician’s report stating that a claimant felt discomfort expressing their pain or trauma as a factor weighing in favor of the applicant. Or an ALJ can consider that physicians implicitly underreport the pain of their Black applicants, and thereby give more weight to other factors outside the medical report. This may reduce the number of claimants who otherwise qualify as disabled persons, but for these cultural biases and norms, from incorrect denial.

Second, considering that a minority of ALJs disproportionately deny Black claimants unless severely ill, the SSA should account for these discrepancies by providing ALJs with trainings, such as implicit bias training. Implicit bias trainings have been proven to improve the rooting out of implicit biases. Implicit bias training would bring about critical self-awareness as to possible unconscious factors that are influencing adjudications, which is the most critical step toward minimizing health disparities.[131] Because this study finds that bias is impacting SSA adjudications, it is strongly suggested that these programs be expanded to improve outcomes for claimants.

Third, because the current system for alleging ALJ wrongdoing is essentially inoperative, the SSA and legislators must focus on creating low-cost and effective strategies for improving adjudications. Increased funding for SSDI would help alleviate the massive deluge of cases before ALJs, thereby allowing ALJs more time to scrutinize their cases for factors of culturally competent care. However, there are also readily available low-cost solutions. For instance, to alleviate the concerns of fiscal conservatives, ALJ implicit bias trainings are low-cost and have been proven to minimize ALJ bias. Additionally, simply making data publicly available for independent study would bring transparency to the ALJ adjudication process, thereby lifting the veil of the hidden judiciary. Adequate study can help bring to light the exact disparities that exist, thereby facilitating cost-effective informed decisions. Increased funding to universities and institutions for adequate study of the topic may be a higher cost, but still feasible, solution. These studies should include analysis of how race, gender, socioeconomic status, and sexual orientation impact the health of disabled claimants.

Armed with such knowledge, the SSA or legislatures can implement directed strategies to safeguard claimants from improper adjudicatory outcomes. Policies should be intentional in their focus on ending systematic health disparities. Otherwise, policy decisions are prone to unconsciously contribute to systemic and institutional policies that negatively impact entire communities.[132] As the system is currently setup, absent explicit discrimination, subtle implicit biases will continue to be overlooked, increasing health disparities among vulnerable groups.

Finally, physicians working with Black patients should encourage their patients to openly discuss their pain with physicians and ensure that it is on the medical record. Legal practitioners working with Black claimants should also provide similar encouragement. Unfortunately, legal practitioners usually welcome claimants once their medical records are fully developed or nearly fully developed. Thus, it is the physician who must ensure that they are adequately maintaining a medical record that accurately reflects their patient’s disability. Preventative medicine tactics, such as catching disabilities early on and fully developing them on the record, would be most useful. Ultimately, the synergy between doctors and lawyers has great potential for efficacy in assisting vulnerable disabled communities.

Conclusion

Despite countless health disparity studies, researchers have yet to fully explain the factors that create and maintain health disparities.[133] This study identifies our legal system as one such contributor. This study focuses on the barren landscape of disability ALJ adjudications and finds that Black Americans are statistically more likely to be denied due to underreported pain levels. These findings are in line with the body of health disparities research and criminal justice research, which both find that the lives of Black Americans are being undervalued.

As the legal system stands today, little is known about how implicit biases are impacting the adjudication of disabled claimants. Furthermore, those claimants have no reliable methods to address unconscious biases that impact their physical and mental health. Legal aid centers help alleviate the burden on claimants, in part because they are the primary provider of legal assistance to claimants in disability cases. However, they are severely underfunded and unable to care for all of the claimants in need.[134] But as this study shows, Black claimants will nonetheless be denied civil justice because ALJ bias is still left without a reliable check. If mechanisms to protect Black disabled applicants are not implemented, especially improvements to their physician reported medical records, the civil justice movement will be harmed. In the meantime, disability claimants will continue to face a system that denies certain claimants over others, thereby facilitating higher illness and death rates for those most disadvantaged.

DOI: https://doi.org/10.15779/Z389W0903K.

Copyright © 2020 Shawn Trabanino. In dedication to my mother and father, who sacrificed endlessly to give their children a better life, for providing a loving home. To my sister, the first in my family to go to college, for paving the road for my brother and me to also attend University of California, Los Angeles (UCLA) as she did. To my brother, for furthering that example of educational excellence by obtaining his M.D. from UCLA and PhD from CalTech. If not for both of you going before me, I would not be where I am today, thank you. To my wife, for being my best friend and support. To Kara Gordon for her edits. To God, for all his glory.

2. Terrence McCoy, 597 Days. And Still Waiting., Wash. Post (Nov. 20, 2017), https://www.washingtonpost.com/sf/local/2017/11/20/10000-people-died-waiting-for-a-disability-decision-in-the-past-year-will-he-be-next [https://perma.cc/BD9J-LBEE]. ↑

3. As far as my research has found, this study is the only mixed methods study of SSA adjudications. ↑

4. John Jordan is a real client from this study’s legal aid database of cases. Names used in this Note are all pseudonyms for real clients of a major northern California non-profit legal aid provider. Pseudonyms are used to protect the client’s ethical rights to privacy. The descriptions of the clients’ health were all recorded within their respective case file documents. ↑

5. The quote is taken directly from the medical records of the patient. ↑

6. Id. ↑

7. U.S. Gen. Accounting Office, GAO/HRD-92-56, Social Security: Racial Difference in Disability Decisions Warrants Further Investigation 40 (1992). At reconsideration, 14 percent of Whites were allowed benefits while 11 percent of Blacks were allowed benefits. Id. at 39. Further discussion can be found in Part II.B. ↑

8. See infra Part II.D. ↑

9. See Universal Camera Corp. v. NLRB, 340 U.S. 474, 491 (1951) (holding that remands occur when the standard has been grossly misapplied); Court Remands as a Percentage of New Court Cases Filed, Soc. Sec. Admin.: Hearings And Appeals, https://www.ssa.gov/appeals/DataSets/AC05_Court_Remands_NCC_Filed.html [https://perma.cc/HZ9D-T3WS]. ↑

10. See infra Part I.C. ↑

11. See Hummel v. Heckler, 736 F.2d 91, 93 (3d Cir. 1984) (finding that the absence of procedural safeguards in the administrative law process requires a stricter application of impartiality). ↑

12. My extensive research revealed only one instance when a federal judge found an ALJ to be bias. This will be discussed in depth in Part I.C. See Grant v. Comm’r, Soc. Sec. Admin., 111 F. Supp. 2d 556 (M.D. Pa. 2000). ↑

13. See infra Part II. ↑

14. See Soc. Sec. Admin., Fiscal Year 2019 Budget Overview (2018), https://www.ssa.gov/budget/FY19Files/2019BO.pdf [https://perma.cc/3ZXU-SSNR]. ↑

15. Paula A. Braveman et al., Health Disparities and Health Equity: The Issue Is Justice, 101 Am. J. Pub. Health 149 (2011). ↑

16. Lewis Kraus et al., Inst. on Disability: Univ. of N.H., 2017 Disability Statistics Annual Report 2 (2018). ↑

17. See Paul L. Kimmel et al., Racial Disparities in Poverty Account for Mortality Differences in US Medicare Beneficiaries, 2 SSM Population Health 123, 128 (2016); see also Tracie Harrison, Health Disparities Among Latinas Aging with Disabilities, 32 Fam. Community health S36 (2009); Ahmad R. Hosseinpoor et al., Socioeconomic Inequality in Disability Among Adults: A Multicountry Study Using the World Health Survey, 103 Am. J. Pub. Health 1278 (2013). ↑

18. See Youlian Liao et al., Socioeconomic Status and Morbidity in the Last Years of Life, 89 Am. J. Pub. Health 569 (1999). ↑

19. See Disparities, Off. of Disease Prevention & Health Promotion (2019), https://www.healthypeople.gov/2020/about/foundation-health-measures/Disparities [http://perma.cc/GK3J-WK3Z]. ↑

20. See U.S. Dep’t of Health & Human Servs, The Secretary’s Advisory Committee on National Health Promotion and Disease Prevention Objectives for 2020 - Phase I Report: Recommendations for the Framework and Format of Healthy People 2020 19–36 (2008), http://www.healthypeople.gov/sites/default/files/PhaseI_0.pdf [https://perma.cc/WA99-D788]. ↑

21. Disparities, supra note 18. ↑

22. Amanda Moore, Tracking Down Martin Luther King, Jr.’s Words on Health Care, Huffington Post (Mar. 20, 2013), https://www.huffpost.com/entry/martin-luther-king-health-care_b_2506393 [https://perma.cc/PA4E-24PE]. ↑

23. David R. Williams & Michelle Sternthal, Understanding Racial-Ethnic Disparities in Health: Sociological Contributions, 51(S) J. Health & Soc. Behav. S15, S16 (2010) . ↑

24. The Health and Physique of the Negro American, (W. E. Burghardt Dubois ed., 1906). ↑

25. Section 1557 of the Patient Protection and Affordable Care Act, Dep’t of Health & Human Servs.: Civil Rights, https://www.hhs.gov/civil-rights/for-individuals/section-1557/index.html [https://perma.cc/BAT2-XX5J]. ↑

26. See William W. Dressler et al., Race and Ethnicity in Public Health Research: Models to Explain Health Disparities, 34 Ann. Rev. Anthropology 231, 233 (2005). ↑

27. See Elizabeth Brondolo et al., Race, Racism and Health: Disparities, Mechanisms, and Interventions, 32 J. Behav. Med. 1, 3–4 (2009). ↑

28. See id. at 2–4. ↑

29. See id. at 2. ↑

30. See Alexander R. Green et al., Implicit Bias Among Physicians and its Prediction of Thrombolysis Decisions for Black and White Patients, 22 J. Gen. Internal Med. 1231, 1236 (2007). ↑

31. Nancy Krieger, Race/Ethnicity, Gender, and Monitoring Socioeconomic Gradients in Health: A Comparison of Area-Based Socioeconomic Measures—The Public Health Disparities Geocoding Project, 93 Am. J. Pub. Health 1655, 1655 (2003). ↑

32. See Chapman et al., Physicians and Implicit Bias: How Doctors May Unwittingly Perpetuate Health Care Disparities, 28 J. Gen. Internal Med. 1504, 1508 (2013). ↑

33. Daniel L. Skoler, Fighting Racial Bias - How One Federal Agency Confronted the Problem, Hum. Rts., Winter 1997, at 18, 19, 21. ↑

34. Lisa Brodoff, Lifting Burdens: Proof, Social Justice, and Public Assistance Administrative Hearings, 32 N.Y.U. Rev. L. & Soc. Change 131, 132 (2008). ↑

35. Id. at 141. ↑

36. Lopez v. Heckler, 713 F.2d 1432, 1437 (9th Cir. 1983) (“Our society as a whole suffers when we neglect the poor, the hungry, the disabled, or when we deprive them of their rights or privileges. Society’s interest lies on the side of affording fair procedures to all persons, even though the expenditure of governmental funds is required. It would be tragic, not only from the standpoint of the individuals involved but also from the standpoint of society, were poor, elderly, disabled people to be wrongfully deprived of essential benefits for any period of time.”) ↑

37. Rollins v. Massanari, 261 F.3d 853, 857 (9th Cir. 2001) (quoting Verduzco v. Apfel, 188 F.3d 1087, 1089 (9th Cir.1999) (citation and internal quotation marks omitted)); see also Withrow v. Larkin, 421 U.S. 35, 47 (1975) (finding “a presumption of honesty and integrity in those serving as adjudicators”). ↑

38. Judges’ implicit biases result in disparate outcomes for minority groups. The Implicit Association Test, a well-known measure of implicit biases, has found that groups hold unconscious racial bias against Black Americans. Jeffrey J. Rachlinski et al., Does Unconscious Racial Bias Affect Trial Judges, 84 Notre Dame L. Rev. 1195, 1198–1201 (2009). Judges hold these same biases, and these biases may influence their judgment. Id. at 1197; see also Jonathan Feingold & Karen Lorang, Defusing Implicit Bias, 59 UCLA L. Rev. Discourse 210, 227 (2012) (finding shooting implicit bias in the way Black faces are perceived as a threat over peer group faces); Justin D. Levinson & Robert J. Smith, Systemic Implicit Bias, 126 Yale L.J. F. 406, 415 (2017). ↑

39. See infra Part I.A. ↑

40. See infra Part I.C. ↑

41. Hummel v. Heckler, 736 F.2d 91, 93 (3d Cir. 1984). ↑

42. Brodoff, supra note 33. ↑

43. Id. ↑

44. Id. at 151. ↑

45. There is no right to court-appointed counsel in administrative hearings. See Brandyburg v. Sullivan, 959 F.2d 555, 562 (5th Cir. 1992) (“The Supreme Court has never recognized a constitutional right to counsel at an SSA hearing.”). ↑

46. Jon C. Dubin, Poverty, Pain, and Precedent: The Fifth Circuit’s Social Security Jurisprudence, 25 St. Mary’s L.J. 81, 140 (1993) (“The Social Security Act has been described as Congress’s most intricate law . . . .”). ↑

47. Friedman v. Berger, 547 F.2d 724, 727 n.7 (2d Cir. 1976). ↑

48. Soc. Sec. Admin., Annual Statistical Report on the Social Security Disability Insurance Program, 2016 149 (2017), https://www.ssa.gov/policy/docs/statcomps/di_asr/2016/di_asr16.pdf [https://perma.cc/7RHP-QLYM]. ↑

49. See 20 C.F.R. § 404.603 (2019). ↑

50. See id. § 404.902–06. There is no data on the time-frame between adjudication levels, but legal aid attorneys provide anecdotally that initial applications take an average of six months to be processed. ↑

51. . See Soc. Sec. Admin., Annual Statistical Report on the Social Security Disability Insurance Program, 2017 152, 161 (2018), https://www.ssa.gov/policy/docs/statcomps/di_asr/2017/di_asr17.pdf [https://perma.cc/SGC7-2DAE]. ↑

52. Upon denial by an ALJ, the claimant may appeal again to SSA’s Appeals Council. See 20 C.F.R. § 404.966–67 (2019). Thereafter, the claimant may again appeal to federal court. 42 U.S.C. § 405(g) (2018). Appeal to the federal court is near non-existence, with only .03 percent of cases going to federal court. Brodoff, supra note 33, at 145. ↑

53. Average Wait Time Until Hearing Held Report, Soc. Sec. Admin. (June 2020), https://www.ssa.gov/appeals/DataSets/01_NetStat_Report.html [https://perma.cc/X26F-XL6V]. (considering data from the court in Oakland, CA, which is the same general geographic area where data from this study was collected). ↑

54. Soc. Sec. Admin., supra note 47, at 163. ↑

55. This number is calculated by adding up the aforementioned initial, reconsideration, and hearing level denial rates. ↑

56. See Brandyburg v. Sullivan, 959 F.2d 555, 562 (5th Cir. 1992). ↑

57. See 20 C.F.R. §§ 404.929–43, 404.944–61, 404.1501–99, 416.901–98 (2019). Disability is defined statutorily as the “inability to do any substantial gainful activity by reason of any medically determinable physical or mental impairment which can expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months.” Id. at. §§ 404.1505. There is no data on the time frame for these hearings to the actual notification of the ruling; legal aid attorneys provide anecdotally that appeals with an ALJ hearing take on average eighteen months. ↑

58. See 42 U.S.C. § 1382c(a)(3)(A) (2018). ↑

59. Id. § 1382c(a)(3)(B). ↑

60. 20 C.F.R. § 416.920(a)(4)(i)–(iv) (2019). ↑

61. See S. Sandy Sanbar, Winning Social Security Disability Cases, 31 J. Legal Med. 91, 117 (2010). ↑

62. 20 C.F.R. § 416.920(a)(4)(v) (2019). ↑

63. See Amrita Maharaj, Note, The Lack of Deference to Medical Opinions in Adjudicating Social Security Disability Claims, 63 Buff. L. Rev. 207, 210 (2015). Ultimately, the ALJ may have SSA’s own medical expert perform an examination on the claimant, which can undermine the testimony of the claimant’s own physician who treated the patient firsthand. ↑

64. Brodoff, supra note 33, at 148–150. ↑

65. Id. ↑

66. See Patti v. Schweiker, 669 F.2d 582, 587 (9th Cir. 1982) (superseded by statute, 42 U.S.C. § 1382 (2018), as recognized in Warren v. Bowen, 804 F.2d 1120 (1986); contra Gist v. Sec’y of Health & Human Servs., 736 F.2d 352 (6th Cir. 1984) (finding that there is no presumption of ongoing disability). ↑

67. Brodoff, supra note 33, at 143. ↑

68. See Rachlinski, supra note 37. ↑

69. . Justin D. Levinson et al., Devaluing Death: An Empirical Study of Implicit Racial Bias on Jury-Eligible Citizens in Six Death Penalty States, 89 N.Y.U. L. Rev. 513, 541 (2014). ↑

70. Rachlinski, supra note 37, at 1197. ↑

71. U.S. Gen. Accounting Office, supra note 6. At reconsideration, 14 percent of Whites were allowed benefits while 11 percent of Blacks were allowed benefits. Id. at 39. ↑

72. See Skoler, supra note 30, at 18–19. ↑

73. U.S. Gen. Accounting Office, supra note 6, at 47. ↑

74. Dorothy W. Nelson, Introduction to the Effects of Gender in the Federal Courts: The Final Report of the Ninth Circuit Gender Bias Task Force, 67 S. Cal. L. Rev. 731, 732–34 (1994). ↑

75. Dubin, supra note 45, at 93 (describing the way Hispanic homeless claimants face additional hurdles because of the lack of interpreters and documents in Spanish). ↑

76. Kendrick v. Sullivan, 784 F. Supp. 94, 103 (S.D.N.Y. 1992) (citing a Magistrate’s Report adopted in Pedroza v. Bowen, No. 85 Civ. 3650, 1989 WL 296875 (S.D.N.Y. 1989)); see also Lora v. Bowen, No. 85 Civ. 7063 (SWK), 1987 WL 16151, at *6 (S.D.N.Y. 1987) (“[T]he ALJ’s failure to probe . . . the claims of an individual who cannot speak English leads this Court to direct . . . the case on remand to another ALJ.”). ↑

77. Skoler, supra note 32. ↑

78. Id. at 19. ↑

79. Id. at 18, 21. ↑

80. Id. at 21. ↑

81. . U.S. Gen. Accounting Office, HRD-76-105, Department of Health, Education, and Welfare: The Social Security Administration Should Provide More Management and Leadership in Determining Who Is Eligible for Disability Benefits 35 (1976). ↑

82. U.S. Gen. Accounting Office, HRD-78-146, Department of Health, Education, and Welfare: A Plan for Improving the Disability Determination Process by Bringing it Under Complete Federal Management Should Be Developed 20 (1978). ↑

83. See Sal Gallicchio & Barry Bye, Dep’t of Health, Educ. & Welfare, Consistency of Initial Disability Decisions Among and Within States 1 (1981). ↑

84. See Saad Z. Nagi, Disability and Rehabilitation: Legal, Clinical and Self-Concepts and Measurements (1969). ↑

85. Elaine Golin, Solving the Problem of Gender and Racial Bias in Administrative Adjudication, 95 Colum. L. Rev. 1532, 1533 (1995). ↑

86. See Top 10 Remand Reasons Cited by the Court on Remands to SSA, Soc. Sec. Admin.: Hearings And Appeals, https://www.ssa.gov/appeals/DataSets/AC08_Top_10_CR.html#content [https://perma.cc/X7PV-CMV3]. ↑

87. 2 Richard J. Pierce, Jr., Administrative Law Treatise 693 (4th ed. 2002). ↑

88. Appeals Council Requests for Review FY 2019, Soc. Sec. Admin.: Hearings And Appeals, https://www.ssa.gov/appeals/DataSets/07_AC_Requests_For_Review.html [https://perma.cc/E23Z-QNVL]. ↑

89. Universal Camera v. NLRB, 340 U.S. 474, 491 (1951); Court Remands as a Percentage of New Court Cases Filed, Soc. Sec. Admin.: Hearings And Appeals, https://www.ssa.gov/appeals/DataSets/AC05_Court_Remands_NCC_Filed.html [https://perma.cc/HZ9D-T3WS]. ↑

90. Javier Meseguer, Outcome Variation in the Social Security Disability Insurance Program: The Role of Primary Diagnosis, 73 Soc. Security Bull. 39, 43 (2013). ↑

91. Soc. Sec. Admin., Aspects of Disability Decision Making: Data and Materials 1 (2012), https://www.ssab.gov/Portals/0/OUR_WORK/REPORTS/Chartbook_Aspects%20of%20Disability%20Decision%20Making_2012.pdf [https://perma.cc/5SUR-QYLN]. ↑

92. See generally id. ↑

93. Golin, supra note 84, at 1532; Paul R. Verkuil, Reflections Upon the Federal Administrative Judiciary, 39 UCLA L. Rev. 1341, 1342 (1992) (“[T]he ‘invisible judiciary[]’ . . . ha[s] not yet been subjected to systematic study.”). ↑

94. See, e.g., 20 C.F.R. §§ 404.1527, 416.927 (2019) (discussing how opinion evidence is evaluated); id. §§ 404.1529, 416.929 (informing how symptoms are evaluated); Deborah A. Stone, The Disabled State 133 (1984); Linda G. Mills, A Calculus for Bias: How Malingering Females and Dependent Housewives Fare in the Social Security Disability System, 16 Harv. Women’s L.J. 211, 232 (1993); Skoler, supra note 32, at 18, 21. ↑

95. Rachlinski, supra note 37, at 1197. ↑

96. See 20 C.F.R. §§ 404.940, 416.1440 (2019). ↑

97. Jason D. Vendel, General Bias and Administrative Law Judges: Is There A Remedy for Social Security Disability Claimants?, 90 Cornell L. Rev. 769, 787 (2005) (“[T]here may be only one or two indications of bias per case . . . .”). ↑

98. Afterwards, an Appeals Council Review (ACR) may view the allegation of bias, but no new evidence is allowed. The next mechanism to allege bias are interim procedures, but they are also inadequate because they don’t specify how evidence should be presented, only provide a short sixty-day window to claim bias, and don’t appear to overturn the decision even if the ALJ was biased. Finally, the claimant may take his allegation of bias to federal court. Again, there is a tremendous hurdle to prove bias, with uncertainty as to whether the court can permit discovery, and a strong possibility that even with discovery, wrongdoing is difficult to assess. See id. at 789–90, 794. ↑

99. Soc. Sec. Admin., Fact Sheet: Social Security is Important to African Americans (2019), https://www.ssa.gov/news/press/factsheets/africanamer-alt.pdf [https://perma.cc/7CBF-6ECX]. Social Security is neutral with respect to race or ethnicity – individuals with identical earnings histories are treated the same in terms of benefits. ↑

100. See supra Part I. ↑

101. See Williams v. U.S. Dep’t of Labor, 879 F.2d 327, 331 (8th Cir. 1989) (citing Ouachita Nat’l Bank v. Tosco Corp., 686 F.2d 1291, 1300–01 (8th Cir. 1982)); see also Leazenby v. Colvin, 654 F. App’x 301, 302 (9th Cir. 2016) (“[Claimant] did not carry her high burden of showing the Administrative Law Judge (‘ALJ’) was biased against her.”). ↑

102. See Rollins v. Massanari, 261 F.3d 853, 858 (9th Cir. 2001) (quoting Liteky v. United States, 510 U.S. 540, 555–56 (1994)). Note that anything less than extreme behavior does not prove bias. As the Court notes, “expressions of impatience, dissatisfaction, annoyance, and even anger, that are within the bounds of what imperfect men and women . . . sometimes display do not establish bias.” Id. ↑

103. See Bayliss v. Barnhart, 427 F.3d 1211, 1215–1216 (9th Cir. 2005). ↑

104. Haseltine v. Astrue, 668 F. Supp. 2d 1232, 1233–34 (N.D. Cal. 2009) (citing Hummel v. Heckler, 736 F.2d 91 (3d Cir. 1984)). ↑

105. Grant v. Comm’r, Soc. Sec. Admin., 111 F. Supp. 2d 556 (M.D. Pa. 2000). ↑

106. . Id. at 559–61. ↑

107. Id. ↑

108. Id. at 559–60 (“[C]ertain characteristics put a claimant at risk of being classified . . . by ALJ Rowell. Those characteristics were if the claimant were black, Hispanic, a poor white, a union member, obese, allegedly mentally impaired, a workmen’s compensation claimant, a controlled substance addict, a Department of Welfare employee, or an accident victim.”). ↑

109. Id. at 558. ↑

110. Id. at 559. ↑

111. 441 F. Supp. 2d 466, 473–74 (W.D.N.Y. 2006). ↑

112. Id. at 474, 469–70. ↑

113. Id. at 473, 475. ↑

114. See, e.g., Ventura v. Shalala, 55 F.3d 900, 903–05 (3d Cir. 1995) (remanding claim because the current ALJ’s line of questioning to claimant and his representative was intimidating, coercive, irrelevant to the claim of disability, offensive, and unprofessional, and ALJ did not allow a full and fair development of the record when he ignored subjective complaints of pain despite the presence of objective evidence and interfered in the representative’s introduction of evidence); Cooper v. Barnhart, 345 F. Supp. 2d 1309, 1311 (S.D. Ala. 2004) (remanding because current ALJ used hostile and antagonistic language towards claimant such as “little skinny twerp”, “[s]kinny little white guy”, “ex-con”, and “junkie”); Sutherland v. Barnhart, 322 F. Supp. 2d 282, 293 (E.D.N.Y. 2004) (finding that an ALJ was found to be unfair, but not biased, when he sarcastically dismissed aspects of the case record and expert witness’s opinions, while devoting considerable writing space in the decision to sarcastically discuss the credibility of claimant); Rosa v. Bowen, 677 F. Supp. 782, 783–85 (D.N.J. 1988) (remanding on the basis of an ALJ exhibiting behavior that was offensive, wrongful, contemptuous, impatient, and irritated). ↑

115. See, e.g., Perkins v. Astrue, 648 F.3d 892, 903 (8th Cir. 2011) (finding no bias in a verbal exchange in which the claimant stated her tendency to watch the Lifetime channel, to which the ALJ responded, “[t]hat’s the girl channel . . . I could have touched a feminist’s nerve there with the girl thing.”); Rollins v. Massanari, 261 F.3d 853, 858 (9th Cir. 2001) (finding that despite an ALJ’s sarcasm and impatience, no bias was found); Verduzco v. Apfel, 188 F.3d 1087, 1089 (9th Cir. 1999) (finding no bias in an ALJ that stated it was hard to believe that the claimant spoke very little English, despite being in the United States for thirty years.). ↑

116. See Perkins, 648 F.3d at 902. ↑

117. See id. at 902–03. ↑

118. See Doan v. Astrue, No. 04CV2039 DMS (RBB), 2010 WL 1031591, at 15 (S.D. Cal. Mar. 19, 2010) (finding statistical evidence of an ALJ’s high denial rates was not enough to establish bias without substantive evidence); see also Johnson v. Comm’r of Soc. Sec., No. 08–4901 (WJM), 2009 WL 4666933, at 4 (D.N.J. Dec. 3, 2009) (stating that statistical evidence is not, by itself, sufficient to find that an ALJ is biased); Smith v. Astrue, No. H–07–2229, 2008 WL 4200694, at *5–6 (S.D. Tex. Sept. 9, 2008) (finding that a 7.19 percent grant rate to claimants was not enough to prove bias, but was nonetheless a cause of concern). ↑

119. See, e.g., McCleskey v. Kemp, 481 U.S. 279 (1987) (finding that statistically higher rates of the death penalty for African Americans were not sufficient to prove bias). ↑

120. See, e.g., Kelly Hoffman et al., Racial Bias in Pain Assessment and Treatment Recommendations, and False Beliefs About Biological Differences Between Blacks and Whites, 113 Proc. Nat’l Acad. Sci. 4296, 4296 (2016) (“Black Americans are systematically undertreated for pain relative to white Americans . . . These findings suggest that individuals with at least some medical training hold and may use false beliefs about biological differences between blacks and white to inform medical judgments . . . .”). ↑

121. See, e.g., Brian D. Smedley et al., Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care (2003). ↑

122. See Hoffman et al., supra note 119. ↑

123. Braveman et al., supra note 14, at S152. ↑

124. Id. ↑

125. Dressler et al., supra note 25. ↑

126. Liao et al., supra note 17. ↑

127. See Brondolo et al., supra note 26, at 2. ↑

128. . Rachel Yehuda et al., Holocaust Exposure Induced Intergenerational Effects on FKBP5 Methylation, 80 Biological Psychiatry 372, 372 (2016). However, some conflict has arisen in recent years as to the immediacy of the trauma. ↑

129. See Green et al., supra note 29, at 1231. ↑

130. See Hoffman et al., supra note 119. ↑

131. See Chapman et al., supra note 31, at 1504. ↑

132. . See Levinson & Smith, supra note 37, at 413–14. ↑

133. Id. ↑

134. In 2003, LSC-funded programs closed over 126,000 cases concerning income and benefits. LSC is tasked with funding legal aids around the country. Legal Services Corporation, 2003-2004 Annual Report 8 (2004), https://www.lsc.gov/sites/default/files/LSC/pdfs/LSC_2003-2004_Annual_Report.pdf [https://perma.cc/3645-H7M4]. ↑