Abstract
Licensure by examination is a major component of social work regulation. This gatekeeping process is intended to increase professional credibility, promote ethical integrity, and provide an avenue for accountability (Council on Social Work Education [CSWE], 2018; Donaldson et al., 2014; Garcia, 1990). By regulating the practice of social work, those seeking social work services, particularly marginalized and oppressed individuals and communities, are believed to be better protected from misconduct or exploitation. However, there are ongoing concerns of bias and disparity within the regulatory system, specifically the licensing exams. Identifying if licensing exams mitigate harm and how social work regulatory systems perpetuate power imbalance is essential.
The use of high-stakes examinations is a common prerequisite to obtaining a license to practice in many professions, including teaching, medicine, nursing, and law. Social workers followed suit and many now view licensing regulations as an essential component of the profession (GlenMaye & Bolin, 2015; Grise-Owens et al., 2016), or as stated by Thyer (2011), a “driving influence” (p. 300), on professional practice. Passing an exam is equated with having the minimum competency required to practice and is a compulsory precondition of licensed social work practice in nearly all United States (U.S.) jurisdictions and most of Canada.
Several authors have noted the lack of research on how social work regulation is impacting professional practice and social work education (Bibus & Boutte-Queen, 2011; Donaldson et al., 2014; Grise-Owens et al., 2016; Miller et al., 2015). Research on the role of licensure in maintaining standards of practice and mitigating harm against patrons is needed, specifically an investigation into how social work licensure exams perpetuate or reduce disparities. It is also unclear from the limited research where the disparity may manifest itself: within total pass rates between racial groups, or the number of times separate racial groups must take the exam to pass. Both represent a form of bias. For example, social workers who must take the exam multiple times are now burdened with the additional costs associated with the process, representing an unfair economic disadvantage. Recent demographic data released by the Association of Social Work Boards (ASWB) on those who take the social work licensure exam appears to show that these inequities are occurring at multiple levels (ASWB, 2022b)
Although the processes used to develop licensure exams have been found to be valid and reliable (Marson et al., 2010), it has also been shown that the licensure exam may negatively impact the recruitment and retention of social workers of diverse minority backgrounds who graduate with social work degrees (Senreich & Dale, 2021). A recent report by the Council on Social Work Education (CSWE), based on responses from 3,400 new graduates from Master of Social Work (MSW) programs, found that nearly 90% of MSW graduates in 2017–2019 were women. “More than 22% of new social workers were Black/African American, and 14% were Hispanic/Latino and 4.8% Asian/Pacific Islander and American Indian/Alaska Native were 1.3%” (Salsberg et al., 2020, p. 14).
Pass rate data for the multiple social work licensure exams administered in the U.S. and several Canadian provinces is collected by the ASWB, and prior to August of 2022 had to be requested by individual state social work board members or schools of social work. Historically there were no concerted efforts to make this data available to the public, but in the last decade, there have been increasing calls for ASWB to release pass rate data, especially by race and ethnicity (Board of Social Work, 2021; Dettlaff, 2021; Minnesota Board of Social Work, 2021; NASW-IL Staff, 2021; S. Maguire, personal communication, January 24, 2022; Texas State Board of Social Work Examiners, 2021). In response to these calls for greater exam outcome transparency, the ASWB approved an initiative in November of 2021 to release demographic exam performance data (ASWB, 2021). In August of 2022, a comprehensive breakdown of exam takers across the U.S. and five Canadian provinces and their pass rates by gender, age, race, and first or subsequent time taking the exam was released. In addition to this national data, social work education program-specific pass rates connected to demographic data were also released.
Efforts to defy Abraham Flexner's claims that social work is not a profession (Flexner, 2001), have fixated on establishing regulation structures to protect the public but have failed to adequately examine the ways in which the existing regulatory frameworks are perpetuating racist policies. Social work licensing exams are intended to determine whether social workers have the minimum knowledge required to practice safely and competently (Thyer, 2011). Achieving equity is not just about addressing individual behavior but must also involve dissecting and analyzing the systems for the presence of racist policies, or “any measure that produces or sustains racial inequity between two groups” (Kendi, 2019, p. 18).
This article presents a brief history of the regulation models in social work and salient historic and contemporary concerns regarding the lack of transparency with regard to pass rates on licensure exams. Next, it addresses how the recent national racial reckoning increased pressure on ASWB to release pass rate data by specific demographic characteristics. The article concludes with insights on the potential consequences now that pass rate data show significant racial disparities do exist. The authors make moderate recommendations in terms of next steps and potential solutions to address the disparities going forward.
Brief History of Regulation Models in Social Work
The creation of robust regulatory standards within social work emerged through a desire to promote greater public protection and higher professional standards (Garcia, 1990). Efforts to regulate the social work profession began nearly 100 years ago, but the evolution of the social work regulatory landscape has been somewhat slow and cumbersome. It wasn't until the 1990s that the U.S., the District of Columbia, and the U.S. territories of Guam, Puerto Rico, the U.S. Virgin Islands, and the Northern Mariana Islands had adopted some form of social work regulation. Each jurisdiction remains responsible for defining the role of social worker and outlining the requirements needed to practice and recently there have been discussions underway to examine state-level definitions of social work practice (Hill et al., 2017). In most cases, licensure requires completion of an accredited social work program, a criminal background check, and passage of an exam. Licensing has become an instrumental aspect of social work regulation and is primarily awarded through passing a qualifying licensure exam (Hoffman, 2002).
The American Association of State Social Work Boards (AASSWB) was established in 1979 and later became known as the ASWB. ASWB was established with the goal of creating examinations that could measure minimum competency and to support state regulatory boards in promoting greater public protection (ASWB, 2022a). Since the 1980s, ASWB has created and administered the licensing exams for the Associate, Bachelor, Master, Advanced Generalist, and Clinical categories. The organization touts a rigorous process for the construction of the exams, which includes the use of psychometricians partnering with subject matter experts, regular practice analysis, and routine audits to prove validity and reliability. ASWB (2020) states that after undergoing statistical analysis and review, any questions which reflect potential bias are removed, ensuring all its “examinations are fair measures of competency regardless of the race or gender of the test-taker” (p. 3).
Lack of Data Transparency
Whether or not a standardized exam is the most fair and valid way to assess professional competence is something social work has grappled with since as early as 1945 (Padgett, 1946). Prior to social work licensure becoming widespread across the U.S., Gandy and Raymond (1979), in an analysis of the impacts state examinations would have on the social work profession, predicted one consequence would be the restriction of access to the profession for Black, Indigenous or People of Color (BIPOC). Licensure and exams for licensure have always been opposed by the National Association of Black Social Workers (NABSW) because of these same concerns (Garcia, 1990).
Because California was one of the first states to pass licensure for social workers (Thyer & Biggerstaff, 1989), one of the earliest studies regarding variables impacting passage of social work licensure exams came from Borenzweig (1977) who studied California's clinical social work exam. This exam was developed and administered by the State of California and not by ASWB. At this time, social work licensure exams were developed and administered by individual states as a part of their regulatory frameworks (Bibus & Boutte-Queen, 2011), whereas today every state uses the exams developed by ASWB. In Borenzweig's study (1977), the only statistically significant variable found to predict successful pass rates on California's clinical social work exam was if the individual was participating or had participated in individual therapy. While Borenzweig indicated that 60% of his sample were white women, the number of people of color and how ethnicity was defined were not presented. In the late 1990s and again in 2008, California would go on to reject ASWB's clinical social work exam citing concerns with representativeness and validity of the exam (Alexander & Johnston, 2008; Woodcock, 2016). They did eventually start requiring the ASWB exam in 2016, which according to Woodcock (2016), had more to do with budget restraints than any changes to the fairness of the exam for California's BIPOC test-takers.
Ten years after Borenzweig (1977), Johnson and Huff (1987) looked at the examination process for entry-level (those with a BSW) social workers in Idaho. The exam they reviewed was developed by Professional Educational Services, not ASWB, which began administering exams in 1983 (Hoffman, 2002). In addition to finding the exam was not a valid predictor of social work knowledge, Johnson and Huff (1987) also found “whites scored significantly higher than people of color” (p. 160); although as described by Thyer (1987) the same year, only 2% of the proportion of variance explained (PPE) can be attributed to the independent variable of race.
At the time ASWB was rolling out its national licensing exams, Garcia (1990) offered a prescient warning, “While the disproportionate exclusion of minorities may not be an explicit goal of legal regulation, it might be a latent effect of that process” (p. 493). Garcia (1990) recommended continual monitoring of the regulating process and removal of cultural biases as much as possible. In one of the first studies investigating one of the ASWB exams, Randall and Thyer (1994) called into question the validity of ASWB's Advanced Practice Examination (nonclinical exam taken by those with an MSW). Similar to Johnson and Huff (1987), findings from Randall and Thyer's (1994) study showed a lack of both content and discriminant validity. Echoing the warnings from Garcia (1990), the finding that the exam lacked discriminant validity means demographic factors such as racial or ethnic identity could be influencing performance (Thyer & Vodde, 1994).
One may argue these early studies simply represent the initial stages of testing within the profession, with increasing sophistication and attention to the issues raised (Thyer & Biggerstaff, 1989). However, 15 years later, Albright and Thyer (2010) repeated the 1994 study, this time with the ASWB Clinical Exam. They obtained even more statistically significant results showing a lack of discriminant and content validity than had been found in the 1994 study. Discussion of the findings included a much more explicit condemnation on the lack of transparency regarding the psychometric properties of the exam, especially regarding race. No empirical studies done since that time have disputed their findings.
While ASWB was eventually successful in developing an exam adopted in every U.S. state, information sharing on how test takers were doing was not forthcoming. Researchers resorted to collecting data through convenience samples (Albright & Thyer, 2010; Randall & Thyer, 1994; Senreich & Dale, 2021) or examining study guides and practice exams (Albright & Thyer, 2010; Castex et al., 2019; Randall & Thyer, 1994) to try and obtain some sense of the various exam impacts. Thyer (2011) went so far as to submit a Freedom of Information Act (FOIA) request to the Florida Department of Professional Regulation just to obtain the pass rates on the Licensed Clinical Social Work (LCSW) exam of students from MSW programs across Florida. What he found were significant discrepancies in pass rates of social workers depending on the university attended; and called on ASWB to provide relevant exam passage rate data by demographics. Thyer (2011) even recommended that the CSWE request the data from ASWB and publish it or make it a part of the accreditation process as is required of law schools and the bar exam scores.
A similar call for relevant data was made by the National Association of Social Workers-New York City (NASW-NYC) (2010) chapter in an issue paper examining social work licensure in New York. It was reported that NASW-NYC received numerous reports of differentials in the pass rates of white and BIPOC social workers on the Licensed Master of Social Work (LMSW) exam (nonclinical exam for those with an MSW). Without access to comparative data from ASWB at the time, the anecdotal stories seemed to be undermining confidence in the fairness of the exam (NASW-NYC, 2010). This issue was so concerning to social workers in New York that the Latino Social Work Task Force hosted a roundtable discussion in New York City on cultural competence and the social work exam in 2014. The panel included major stakeholders, such as representatives from the ASWB, NASW-NYC, the New York State Department of Education, and deans at major New York schools of social work.
When asked directly about releasing the demographic data on passage rates of its exams, ASWB representatives indicated this data was used to ensure individual exam questions “operated properly” (Castex et al., 2019, p. 224), but the data had not been aggregated collectively and apparently there was no desire on the part of ASWB to do so at that time (Castex et al., 2019). And while ASWB claimed their exam was free of racial biases (Marson et al., 2010) in a review of a Master level practice exam in the 2010 ASWB Study Guide, Castex et al. (2019) found several examples of, “ethnic, cultural and racial bias in the content of several items…and…use of unnecessarily complex language lacking cultural sensitivity” (p. 218). The recent disparities seen in the pass rate data released by ASWB seem to support the claims made by Castex et al. (2019).
To date, two significant studies have examined racial disparities between licensed social workers in an effort to better understand how the ASWB licensure exam may impact these differences. Senreich and Dale (2021) examined the licensure status of 5,000 MSW graduates from City of New York programs. From their sample, they were able to determine white graduates had significantly higher rates of licensure than those who identified themselves as BIPOC. What they were unable to determine was whether other variables could be impacting the licensure rates across racial groups.
Kim’s (2022) study attempted to examine these additional variables identified by Senreich and Dale (2021). Using a national sample, she examined racial differences in licensure rates, based on structural barriers, such as disparate educational access to social work programs, varying state licensure regulations, sectors of social work practice, and occupational requirements for a social work license. What Kim (2022) found was that the significant variation in licensure rates between white and Hispanic as well as white and other races (e.g., Asian–Pacific Islander) could be explained by these structural barriers. Differences in licensure rates between African American and white social workers at the Master level were mitigated only by type of practice and type of employer. However, the racial disparity between Black and white social workers at the Bachelor level of licensure was not moderated and persisted even after accounting for a variety of variables.
The recently released ASWB exam pass rate data indicate significant disparities between racial groups in all exam categories, with particularly stark differences between white test takers and Black test takers. The first-time pass rate data for the clinical exam: Black (46.2%), Native American/Indigenous (59.1%), Hispanic/Latino (65.8%), and white (85%), exemplifies the disparity between white test takers and test takers of other racial groups (ASWB, 2022b). Pass rates for the bachelor exam indicate that white test takers are more than twice as likely to pass on their first attempt than black test takers (ASWB, 2022b). These unsettling outcomes bring into question the credibility and validity of the licensure process.
At its inception, ASWB was formed to share information across states working toward regulation of social work practice (Hoffman, 2002). A desire for consistency and uniformity in regulation led the organization to develop a licensure examination that could be used across jurisdictions (Hoffman, 2002). ASWB also established the American Foundation for Research and Consumer Education in Social Work Regulation in 1985 (Thyer & Biggerstaff, 1989). According to Thyer and Biggerstaff (1989), the foundation was meant to provide funding for research on social work regulation. Between 1985 and 2017, the Foundation provided support to approximately twelve research projects (ASWB, 2019), however access to information about these projects can only be obtained through an email request. This established lack of data transparency contributes to the creation of unnecessary barriers to addressing inequity within the profession.
Disparities in Other Professional Licensing Exams
Social work is by no means alone when it comes to grappling with potential disparities in standardized professional exams. In psychology, Sharpless (2021) has found significant differences in the first-time pass rate on the Examination for Professional Practice in Psychology between white and Asian American exam takers versus Black and Hispanic examinees. Similar results have been found among exams for teachers (Harris, 2015a, 2015b), doctors (Williams et al., 2020), surgeons (Yeo et al., 2020), police (United States Department of Justice, n.d.), firefighters (Newman, 2007), and attorneys (Wightman & Ramsey, 1998), to name just a few. In the summer of 2021, the American Bar Association released Bar exam passage data by race, ethnicity, and gender. Distinct disparities between racial groups, along with gender differences, emerged from the data (American Bar Association, 2021). It remains to be seen what the ABA ultimately does to address the disparities.
Concern over leaked test questions, the difficulty describing statistical procedures used to a lay public (Marson et al., 2010; Taube, 1997), or even legal challenges due to adverse impact (Harris, 2015a, 2015b) can be significant barriers to data transparency (DeCarlo, 2022). However, the tide is changing when it comes to standardized exams. Many colleges, including all University of California and several Ivy League universities, including Harvard, have chosen not to require SAT/ACT scores as a part of their admissions process. While many of these decisions have been an acknowledgment of the lost educational time due to Covid (Jaschik, 2022), the simultaneous recognition of the inherent racial bias within standardized exams has also been influential (Watanabe, 2019).
Post-George Floyd: Pressure Mounts
Despite decades of repeated calls (Albright & Thyer, 2010; Collins et al., 2002; Randall & Thyer, 1994), innuendos (Woodcock, 2016), and requests for release of information (Board of Social Work Examiners, 2021; Florida Association of Deans and Directors, 2019; Hymans, 2020; Minnesota Board of Social Work, 2021; NASW-IL Staff, 2021; National Association of Social Workers-MI, 2021; S. Maguire, personal communication, January 24, 2022; Social Worker Professional Standards Committee, 2020; Texas State Board of Social Work Examiners, 2021) concerning passage rates on all types of social work exams by specific demographics, ASWB remained resolute in its refusal to release such information. Then came a moment of national reckoning on race and race relations. On May 25, 2020, as a result of a live-streamed video, the world witnessed the murder of George Floyd, an African American man, by Derek Chauvin, a white police officer. Something shifted in the U.S. and across the world. A different type of consciousness was born for many white people. Calls for racial justice, anti-racist practices, and a re-examination of all entrenched policies, procedures, and ways of knowing made their way into the discourse of the dominant culture. The need for critical analysis was true for the profession of social work as well.
In August 2020, all major national social work organizations, including NASW, CSWE, the National Association of Deans and Directors (NADD), NABSW, the Society for Social Work and Research (SSWR), American Academy of Social Work and Social Welfare (AASW) and the Group for the Advancement of Doctoral Education (GADE), and ASWB came together for a virtual “town hall” to discuss what social work organizations were doing to advance anti-racism. As could be expected, the CEO of ASWB, Dwight Hymans, was again asked about the release of pass rates on social work licensure exams. At one point in the conversation, he indicated ASWB did not collect this type of information, but then later in the town hall, said the ASWB Board of Directors had voted repeatedly in the past not to release such aggregated information for any of the social work exams. In an attempt to clarify these contradictory remarks, one of the present authors reached out to Mr. Hymans directly. In his response, Mr. Hymans indicated ASWB does ask exam takers for their racial/ethnic identity, but providing it is voluntary. He clarified this information is only used to analyze individual test questions, specifically ones that are being pretested and do not count toward the passing score. He went on to write, The data that was referenced during the townhall meeting was a request made by the National Association of Deans and Directors (NADD) to ASWB for pass rates broken down by demographics. We do not collect the pass rates of various demographic groups but instead publish a single pass rate for all test takers. NADD made a formal request to our board of directors for pass rates broken down by demographics. Our board considered the request but declined it (D. Hyman, Personal Communication, August 16, 2020).
Mr. Hymans included with his email the letter sent to NADD (Hymans, 2020, December 21) and the ASWB Board of Directors Policy Statement on Disclosure of Examination Performance Data (ASWB, n.d.)
Subsequently, over the next year, ASWB continued to meet with stakeholders demanding release of the data. For example, in November of 2020, the Macro Practice Licensure Workgroup of the Special Commission to Advance Macro Practice (Licensure Workgroup) held their annual meeting, this time virtually, due to the pandemic, with leadership at ASWB. Release of the exam passage rates was a key issue on the agenda. Members of the Licensure Workgroup urged ASWB to operate in the spirit of transparency and cooperation, promising support, and collaboration from the Licensure Workgroup. A formal letter was subsequently sent from the Special Commission reiterating the request and an offer of solution-building no matter what the results. The response from Hyman indicated several other organizations and groups had made similar requests, He asked for patience as the Board considered all of the requests and planned to meet at the end of April 2021 to, “explore options that fit ASWB's mission and move toward viable solutions to the issues raised.” (D. Hymans, Personal Communication, February 16, 2021).
During the ensuing months, members of the Workgroup and many other coalitions, higher education institutions, and NASW chapters continued to advocate for the release of social work exam data through their own boards of social work (Board of Social Work Examiners, 2021; Dettlaff, 2021; Minnesota Board of Social Work, 2021; NASW-IL, 2021; S. Maguire, personal communication, January 24, 2022; Texas State Board of Social Work Examiners, 2021).
In May 2021, ASWB leadership was asked to address the issue before the Minnesota (MN) Board of Social Work. It was clear from their presentation that the late April 2021 board meeting had not resulted in a change in their policy. After a presentation of the procedures used to ensure fairness and reduce bias in all their social work exams, members of the public were allowed to give testimony. Several individuals spoke of their own experiences with passing the exam, along with members of social work organizations such as NASW-MN, The MN Association of Black Social Workers, and the Hmong Social Work Coalition. All testimony urged ASWB to release the demographic data in an effort to address any racial discrimination which may exist. In response to the testimony, Mr. Hyman stated ASWB was not a social work organization and only beholden to its membership and Board of Directors, who had chosen not to release such data (Minnesota Board of Social Work, 2021).
Several members of the MN Board of Social Workers also expressed their desire for the data, and voted to put together a task force to collect, analyze and then release the demographic data of all MN examinees for the past 5 years, as well as a system for better data collection moving forward. The Board also voted to formally request such data from ASWB at a national level. To the authors’ knowledge, the MN Board of Social Work was only the second Board to willingly agree to release this information. The Michigan Board of Social Work also released the data they had on examinee pass rates to their Schools of Social Work (National Association of Social Workers-MI, 2020). The released data led the Michigan chapter to begin discussions with social work programs about the potential presence of racial bias in the licensure exams.
On November 9, 2021, the ASWB Board voted unanimously to, “gather, analyze, and release examination performance data” (Association of Social Work Boards, 2021). An update on their process was promised within the first quarter of 2022. On April 22, 2022, ASWB announced that they were working with the Human Resources Research Organization (HumRRO) to release pass rate data “broken down by race, ethnicity, age, primary language, and gender” by late summer of 2022 (ASWB, April 22, 2022, para 12). Just prior to the publication of this article, ASWB released pass rate data for the period of 2011 to 2021 revealing an alarming discrepancy of outcomes among several demographic groups.
Addressing Consequences of Racial Disparity in Social Work Licensure Exams
Among the greatest challenges the nation is facing now, and making steady but slow progress on, is the redressing of various forms of institutional racism since its founding. The social work profession also faces a moment of reckoning regarding systemic racism, specifically within its regulatory system. Whereas nearly all graduate social work programs require a minimum cumulative grade point average of 3.0, based upon observation and application of practice behavior, the licensure exam process relies upon assessment of knowledge or cognitive processes only (Apgar, 2021). It appears this discrepancy may result in outcomes that disproportionately deny social work graduates of BIPOC origin the opportunity to earn licensure. Perhaps the more difficult question to answer is how many BIPOC social workers are denied social work positions due to failing licensure examinations. What emotional, behavioral, and physical health consequences have BIPOC social workers experienced after having invested academically and financially toward earning social work degrees, only to be denied licenses to practice in the profession? What is the total impact of the many lost opportunities for clients, constituents, and communities to be served by BIPOC social workers who could not practice because they did not achieve licensure? What consequences can be envisioned for the perpetuation of racial disparities in the licensure apparatus? These are the questions the profession must prepare itself to answer in order to embody its values of integrity and social justice.
The efforts to analyze and disseminate the recently revealed disparity in social work licensure passage rates are ongoing. To address the implications will require a prompt and comprehensive response. Any discussion of remedies must be guided by a formal admission of wrongdoing and a spirit of reconciliation and healing. There will need to be both structural and policy solutions explored to redress the legitimate grievances of BIPOC social workers impacted by the racial disparities within the licensure apparatus. A review of how some states have started to address these issues provides guidance for moving forward. However, in this historic moment of racial reckoning, the role and relevance of licensure examinations will need to be reexamined and may need to be radically reimagined.
Footnotes
Author Note
Dr. Mary Nienow is a special editor of this issue of the Research on Social Work Practice. This journal uses an anonymous review process in which authors do not review their own work and reviewers do not know authors’ identities. Dr. Nienow had no influence on the selection of the reviewers nor were their identities disclosed.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article
