Abstract
Consistent with psychology’s call to action for an inclusive and intentional focus on social class, we conducted a content analysis examining class variables relative to race and gender variables in articles over a 15-year period in The Counseling Psychologist and the Journal of Counseling Psychology. Articles were classified by degree of inclusion of these variables, resulting in three categories: mentioned, integrated, and primary. Despite the recent trend toward class inclusion, only 560 of 1,440 studies (39%) included all three variables at any level. Articles where all variables were “Primary” comprised only 1.9% (n = 28) of the articles we reviewed. Using a qualitative content analysis and intersectionality lens, only 15 of the 28 studies thoroughly integrated class, race, and gender variables. Results highlight a recent, significant increase in attention to class issues, as well as the continued need to place class analysis on par with other important diversity variables. Implications are discussed.
Keywords
Social class is a central organizing construct in understanding human behavior, social organization, power, privilege, equality, and lack thereof. Class is foundational in shaping human identity and is embedded at every level of all social structures. Counseling psychologists have long approached diversity and social justice with a fundamental concern for identifying, illuminating, and righting injustices; however, the discipline of psychology as a whole has fallen short with respect to issues of social class, classism, and poverty (Blustein, McWhirter, & Perry, 2005; Lott & Bullock, 2007; Smith, 2008, 2010). Even within the multicultural movement, social class has historically been understudied as a demographic or contextual identity variable in psychology compared with other diversity variables such as race and gender (American Psychological Association [APA], 2000, 2006; Arredondo, Rosen, Rice, Perez, & Tovar-Gamero, 2005; Liu et al., 2004; Smith, 2008, 2010). Despite decades of research showing that chronic poverty causes pervasive psychological and physical harm, Smith (2010) suggested that psychologists’ relative neglect in exploring social class and economic justice in practice, theory, and research is noteworthy, unacceptable, and tacitly endorses and reinforces classism.
Steps in rectifying the marginalization of class have been slow to emerge. APA’s Resolution on Poverty and Socioeconomic Status in 2000 (see http://www.apa.org/about/policy/poverty-resolution.aspx) and the Report of the Task Force on Socioeconomic Status (SES) in 2006 (see https://www.apa.org/pi/ses/resources/publications/task-force-2006.pdf) provided a call to action for psychologists to address issues regarding economic injustice. In 2007, at the APA convention in New Orleans on the heels of Hurricane Katrina, APA Council approved the creation of the organization’s permanent Office on Socioeconomic Status and Committee on Socioeconomic Status (OCES & CSES, respectively), whose mission, in part, is to “promote appropriate attention to SES in psychological research and practice” (APA, 2014, p. 1). We argue (and demonstrate) that social class remains understudied in the counseling psychology literature, although paradoxically, the importance of class issues remains starkly evident.
The Significance of Class
As noted more than 30 years ago, The correlation between social class and prevalence of symptoms of psychological disorder is one of the most thoroughly documented relationships in epidemiological research. So consistent are these findings, with so few exceptions, that we will assume this relationship as given. (Allen & Britt, 1983, p. 149)
Thus, economic power is central to our well-being. As poverty rates climb, the gulf between the rich and the poor in the United States continues to increase, leading to unprecedented income equality. For example, in 2012, those in the top 20% of U.S. earners took home more than half of all the income in the United States, whereas the poor received 3% (Chin & Culotta, 2014). The largest share of U.S. income is now held by the top 1% of the population, more than at any other time since 1928 (Center on Budget and Policy Priorities, 2009), a fact now entrenched in the popular mind through the Occupy Wall Street movement.
In 2013, nearly 16% of Americans (48.8 million people) were struggling to exist with income below the poverty level (U.S. Census Bureau, 2014). Twenty-two percent of all children in the U.S. (more than 16 million children) live in families with incomes at or below poverty level, and the majority of those children come from non-White families (National Center for Children in Poverty, 2014); in addition, 1 in 30 children from the U.S. (2.5 million children) are homeless (Homelesschildrenamerica, 2014). Extreme poverty (living on less than $2 a day) in the United States has doubled since the mid-1990s, and now impacts more than 6 million children (Children’s Defense Fund, 2014).
The poor are generally blamed for not pulling themselves up by their bootstraps, yet residential segregation and isolation by class and race adversely affect education, employment opportunities, safety, and health, resulting in limited chances for poor individuals to mitigate their circumstances (Downing, 2014; Smith, 2010; Stoll & Raphael, 2002; Wilson, 1996). This segregation and isolation leads to what Wilson described as horizontally organized communities, where, ironically, being poor is expensive and there are no role models to guide or encourage a way out of poverty. As poverty rates continue to soar, poor people are being forced into further invisibility by new laws that prohibit providing food for the homeless or allowing homeless people to sit on the sidewalks (Landen, 2014; Villas, 2014).
Poverty is not just about having limited resources, it is about chronic poverty-related stress. These pressures include income fluctuation and insufficiency, lack of safe and affordable housing, child care concerns, food insecurity, and few alternatives to assist when illness or emergency situations arise (Kurtzleben, 2014). The working poor often do the dirty and strenuous jobs no one else wants to do, yet these jobs routinely lack compensation plans that the middle class and the rich take for granted, such as livable wages, health care coverage, paid time off, or sick pay (Downing, 2014; Smith, 2010). Deeply embedded in dominant cultural values that support meritocracy (Knowles & Lowery, 2012; Wiley, Deaux, & Hagelskamp, 2012), poverty results in shame and exclusion. Smith (2013) stated that often, out of the awareness of those with class privilege, exclusion from common opportunities and experiences in society are markers in defining poverty.
Differences in social class allow the wealthy to enrich their children’s lives, to open educational and employment doors, and to provide material safety nets for their children when they fail, a phenomenon that economists call “opportunity hoarding” (Reeves & Howard, 2013, p. 3). In contrast, children of low-income parents face obstacles even when excelling in college. They receive a large student loan debt along with their diploma and often return to live in impoverished neighborhoods that keep them disconnected from mentors and opportunities to succeed (O’Brien, 2014). It would be remiss not to acknowledge that the rich have their own set of psychological issues (Luthar, 2003), providing further rationale for the need to include social class in our scholarship; however, the plight of the poor is an ethical problem of epic proportion for everyone, including psychologists, and could reasonably be described as a crisis (Cozzarelli, Wilkinson, & Tagler, 2002; Padilla & Sherraden, 2005).
The crisis of poverty cannot be understood in isolation. Race and gender factor strongly into poverty statistics, with 27% of Blacks and 26% of Latino/as being poor compared with 10% of non-Latino Whites (Downing, 2014). Women are disproportionately poor and in poverty more often than men, and women of color are twice as likely to be in poverty compared with non-Latina White women. Single mothers represent 31% of those who live in poverty; that number is higher for single mothers of color (Downing, 2014). In addition, although rates of sexual activity are comparable for all women, a poor woman is about 5 times as likely as an affluent woman to have an unintended child, limiting reproductive justice and furthering the cycle of poverty for women and children (Reeves & Venator, 2015). Indigent African American males have disproportionate incarceration rates compared with all other groups; nearly all death row inmates are poor, and most are racial minorities (Amicus, n.d.; Amnesty USA, 2012). The unmistakable evidence of bias in multiple aspects of law enforcement in poor, non-White communities has recently been documented and made public by the U.S. Department of Justice (DOJ; 2015). In making sense of and addressing these clear inequities, understanding the complex interplay between class, race, and gender (CRG) is critical. Thus, an intersectional perspective can assist in this regard.
Intersectionality
When intersections of critical diversity factors determining resources and life chances fail to be examined together, psychologists’ learning and ability to advocate are thwarted. Although beyond the scope of this article, there are countless examples of how CRG intersect to create unimaginable hardship and unfair treatment. Examples include distant historical events to more recent tragedies such as the trafficking of poor ethnic minority women (Mir, 2013), the school-to-prison pipeline for poor ethnic minority children (Henderson, 2014), and the mishandling of Eric Duncan’s Ebola diagnosis (Akhenaten, 2014).
Too often, psychologists who study gender and/or race limit their overall focus, therefore missing the impact of class on gender and racial differences in power and privilege (Silverstein, 2006; Smith, 2008; Smooth, 2006). The omission of social class from psychological scholarship in general, and from intersectional research that explores other diversity variables, is of serious concern (Jackson & Williams, 2006; Lott & Bullock, 2007). The authors of this study are not suggesting that race and gender are studied too much, but rather that social class is not studied enough. Specifically, an accurate and context-rich consideration of human behavior is not realized when social class is absent or omitted from intersectional studies.
Certain subdisciplines within psychology have paved the way in recognizing that the conceptualizations of culture and multiculturalism must include social class. Tebes (2010) suggested that community psychology has been at the forefront of this integration for decades. Cole (2009) and Smith (2008, 2010) paid tribute to social, feminist, and vocational psychologists for considering multiple forms of social location in their research approaches. Such approaches can be broadly understood under the framework of intersectionality.
Intersectionality Theory and Methodology
Feminist sociological legal theorist Kimberle Crenshaw introduced the term intersectionality in 1989, but the theory can be traced back to the 19th century. Intersectionality theory seeks to examine how socially and culturally constructed categories of discrimination interact on multiple levels, often simultaneously, contributing to systematic inequalities and social injustice (Collins, 2000; Crenshaw, 1989; Novotney, 2010). Intersectionality is also a systemic approach to conceptualizing and analyzing how inequalities function together in people’s lives at numerous levels (e.g., individual, family, institutional, society). This integrative approach may include a complex and historical consideration, examining the role and effect of marginalization and exclusion, often yielding transformative social justice power. Structural patterns found in intersectionality perspectives highlight institutional power, privilege, oppression, and marginalization in areas such as education, health care, and the workforce, providing a lens through which to consider changing the system. Without an intersectionality approach, many scholars address CRG as singular, unconnected issues. In reality, the synergy and complexity of these variables shape the experiences of all people. Intersectionality is both a theory and a methodology for considering relationships among multiple dimensions of social structures and identities (Cho, Crenshaw, & McCall, 2013). Such relational thinking moves beyond simple group comparisons to examine connections, interactive processes, and patterns.
Intersectionality approaches are being heralded by some disciplines, such as criminology and marriage and family studies, as a cutting edge way to advance theory, research, and practice (Burgess-Proctor, 2006; Few-Demo, 2014). The field of psychology is beginning to embrace intersectionality (Bowleg, 2008; Cole, 2008, 2009; Shields, 2008; Warner, 2008), allowing psychologists to shift focus by examining current sociopolitical structures in our culture that have historically and systematically oppressed certain groups, rather than viewing poverty as an individual shortcoming or voluntary choice (Diemer, 2009; Lott & Bullock, 2007; Smith, 2010). Cole (2009) suggested that the discipline of psychology has most likely been slow in integrating intersectionality due to a lack of framework and guidelines. Cole argued for a new mindset in understanding the multiple factors that constitute oppression and privilege, rather than a need for new methodologies or unruly sample sizes. In contrast, Warner (2008) stated that incorporating an intersectionality perspective necessitates an expanded methodological scope and adds complexity to potential analyses. It seems reasonable that both perspectives could coexist and be useful. Intersectionality approaches abound and are as apt to vie as to coalesce, and as a result, intersectionality methodology has been criticized due to its vagueness as well as its diversity and complexity. Despite the tensions and disagreements, according to Smooth (2006), it is “a mess worth making” (p. 400).
McCall (2005) described three methodological approaches that attempt to manage inherent complexities in taking an intersectional approach. The first approach, anticategorical complexity, rejects categorization as arbitrary, deconstructing divisions that have been artificially created. This approach posits that finite divisions do not cover the complexities that exist in intersecting identities and do little to understand the intersections of CRG or the inequalities or oppression at the root of these intersections. The second approach, intracategorical complexity, can best be described as a midpoint on a continuum between anticategorical and intercategorical categories. This approach examines intricacies of identity by using case-study methods, for example. These two approaches are amenable to qualitative work. The third approach, intercategorical complexity, accepts categorization and, while recognizing its limitations, uses these classifications as an anchor point to compare differences between groups as they unfold. This third approach allows quantitative research to be more readily implemented.
In sum, social class has been inadequately addressed in psychological scholarship, both alone and relative to race and gender. Intersectionality theory and methods are a way to ensure a more concerted focus on class along with its undeniable interaction with other critical features of identity. Research using intersectionality helps place psychology at the forefront of how we think about multiculturalism and social justice.
The current investigation rests conceptually on an intersectionality approach and is closely related to McCall’s (2005) intercategorical complexity methodology. Categories in the present study were defined, maintained, counted, and compared quantitatively, as well as analyzed through a qualitative content analysis. Content analyses have an extensive track record in helping the discipline of psychology self-assess and then move forward; a selective review of relevant content analyses is presented.
Relevant Content Analyses
Content analysis, with a focus on diversity, has a long history in the psychology literature in general, and in counseling psychology in particular. The first appears to be Ponterotto’s (1988) landmark analysis of race/ethnicity. In the last 10 years alone, there have been several content analyses focusing on multiculturalism broadly defined (Arredondo et al., 2005; Lee, Rosen, & Burns, 2013), race/ethnicity (Delgado-Romero, Galván, Maschino, & Rowland, 2005; Edwards & Pedrotti, 2008; Neville & Carter, 2005; Noltemeyer, Proctor, & Dempsey, 2013), sexual diversity (Huang et al., 2010; Singh & Shelton, 2011), religion (Leach & Sato, 2013), international populations (Pieterse, Fang, & Evans, 2011), men and masculinity (Wong, Steinfeldt, Speight, & Hickman, 2010), and disability (Foley-Nicpon & Lee, 2012). A number of content analyses have incorporated social class into their reviews; however, it was not the central focus and was often combined with another single variable, such as race or gender.
Of particular note, Lee et al.’s (2013) analysis of 50 years of multicultural content in the Journal of Counseling Psychology (JCP) revealed that only 17 articles had been published between 1954 and 2009 on social class, representing 1.41% of all multicultural articles and only 0.46% of all articles. Social class was the least frequent of all diversity variables analyzed in this extensive review, last after gender, race/ethnicity, age, sexual orientation, religion, and disability. These authors included a category called “intersections” that included any two (or more) of the social locations; this category contained 213 articles, representing 17.72% of all multicultural articles and 5.73% of all articles. However, it cannot be determined from this review how many of those might have included social class and, given the dearth of articles on social class alone, it is unlikely to have been a substantial number. Such results provide important validation for the present investigation’s focus on social class as the most ignored of the CRG triad.
In 2004, Liu et al. conducted the first content analysis of social class in the counseling psychology literature, in an investigation spanning 20 years. The authors identified considerable problems with studying social class, as well as a paucity of information about class compared with other significant diversity issues such as race/ethnicity. For their study, they reviewed three counseling journals, including JCP, Journal of Counseling and Development (JCD), and Journal of Multicultural Counseling and Development (JMCD), and examined articles between the years 1981 and 2000. Of the 3,915 articles they reviewed, 710 (18%) articles included social class as a variable. Among theoretical articles, only 10 (1.4%) included social class as a primary aspect of the study. Liu et al. noted that it seemed that many researchers discussed social class in general theory papers and that the tendency was to mention social class in parenthetical terms. The researchers found 448 words that were used to describe social class and classism. Liu et al. noted that terms related to social class were used inconsistently and class was assessed using archaic instruments (i.e., Duncan, 1961; Hollingshead, 1958, 1975), making it difficult for psychologists to research social class issues effectively. The authors called for researchers to continue to address these problems so “that future research will use social class in intentional and meaningful ways” (Liu et al., 2004, p. 16).
In summary, what appears to be relatively absent from the research database in counseling psychology are specific studies examining the critical intersection between class and the more often studied variables of race and gender (APA, 2000, 2006; Arredondo et al., 2005; Liu et al., 2004; Smith, 2008, 2010). The current study was conducted with this goal in mind; it is the first content analysis of the counseling psychology literature to simultaneously address all three variables. As the most neglected of the three areas, special attention was focused on class issues throughout. Also, it has now been a decade since Liu et al.’s (2004) review and call to attend to class issues. Thus, the present review sheds light on the degree to which their call has been answered. The three overarching research questions for this investigation were as follows:
Method
Overview
Employing an archival mixed method design, this study examined the intersection of the variables of CRG in counseling psychology literature spanning the years 1996 to 2011. TCP and JCP were selected as they are the leading U.S. counseling psychology journals, capturing the zeitgeist, values, beliefs, and perceptions of professional identity of counseling psychology as a discipline. The 15-year span was chosen as a practical compromise between resources and comprehensive coverage. In addition, given the timing of the multicultural movement as the “fourth force” in psychology (Pederson, 1991) coming into its own in the early and mid-1990s, the interval chosen parallels the development of this important advance. Finally, the analysis was done in two waves: A 10-year analysis conducted as the first author’s dissertation, and then a 5-year update following the initial work. Data collection began with journal issues dated on or after July 1, 1996 and ended with the last issue of each journal prior to July 2011, resulting in a total of 15 years of articles.
Content analysis methodologies have been developed for both quantitative and qualitative work. Quantitative coding of the 1,440 articles in the 15 years of TCP and JCP was conducted first. In keeping with recommended guidelines for such work (e.g., Nandy & Sarvela, 1997; Neuendorf, 2011), researchers attended to issues regarding both preparation and methodological processes.
Quantitative Procedure and Analysis
Preparation for quantitative content analysis involves immersion in the field of study to understand what has already been examined and how variables have been measured, as well as to determine what theoretical orientation might be used to guide the work and define the scope of the investigation (Neuendorf, 2011). These preparation steps are evidenced earlier in this article. Methodological decisions, such as unitizing, sampling, and determining a consistent way to measure and record results are presented below.
Initial steps
Following training, all journals were divided among the research team members. The initial step involved sorting articles in each journal into one of two categories: empirical or theoretical. Articles were coded into only one category. Theoretical articles were nonempirical and/or conceptual. The decision to include theoretical articles was modeled on Liu et al.’s (2004) content analysis for consistency and because Liu et al. found that theoretical articles included class more frequently than did empirical articles. Empirical articles could be either quantitative or qualitative but, in all cases, involved the collection and presentation of new data. Each article was read and reviewed in its entirety, and results were coded on a tracking form.
The next step in the analysis was to examine, categorize, and quantify information from each article (n = 1,440) from the two journals based on whether or not an article included mention of any of the three variables (CRG). Every article was coded for the presence or absence of all three variables. Decisions regarding inclusion of CRG variables were based on affiliated synonyms or key words. A list of key words was used as a starting point. During research team meetings, affiliated synonyms or key words that were debatable were discussed as a group until consensus was achieved. Research team members also had access to a preexisting key word list from Liu et al.’s (2004) study on social class to aid in this process, although parallel lists for gender and race were not available. Key word lists for race and gender were developed during the course of the study and are available from the primary author.
Articles using CRG together
For empirical research studies, each article was evaluated and coded on a tracking form to determine whether the variable under examination was used in just one location (i.e., introduction) or whether it was used in multiple locations (i.e., abstract, introduction, and discussion). Borrowing from the Liu et al. (2004) content analysis categories, 19 different combinations of categories were formed to reveal the location of CRG in each article. Both for consistency in comparing with Liu et al.’s work and because class continues to be routinely ignored (e.g., Lee et al., 2013), we felt it was important to document any attention given to this variable.
Not all articles used all three variables equally. For example, in a given article, race and gender may be mentioned throughout the entire article, but social class may only be mentioned in a particular location such as the Method section (i.e., demographics of the participants). To account for this, each variable’s location was assessed separately according to the 19 possible categorical locations. Thus, for empirical articles, the tracking form was used three times for each article: once to determine the location of the class variable, subsequently to determine the location of the race variable, and finally to determine the location of the gender variable. When the location of the three variables had been assessed, the researcher entered the results of the location (e.g., 1-19) on a document summary form. A similar procedure was followed for theoretical articles, using appropriately modified criteria.
For both types of articles, a three-tiered categorization system was used once all document summary forms were completed. Articles could be classified as mentioned, integrated, or primary. The system is detailed in Table 1.
Coding Categories.
Interrater agreement
In the first wave of data collection (the first 10–year content analysis), the last two issues of each volume of TCP and JCP for 1996 were used to establish initial interrater agreement. This resulted in 49 training articles. There were six members of the research team involved in coding. Ongoing interrater agreement was rechecked using one randomly selected article from TCP and one from JCP for every even year (1998, 2000, 2002, 2004, 2006).
In the second wave of data collection (the 5-year update), the second to last issue of TCP for 1996 (Vol. 34, No. 5) and the last issue of JCP for 1996 (Vol. 43, No. 4) were used to establish initial interrater agreement. This resulted in 22 training articles. There were four members of the research team involved in coding. Ongoing interrater agreement was rechecked using one randomly selected article from TCP for every even year (2008, 2010), and one from JCP for every odd year (2007, 2009, 2011).
Percent agreement was calculated for each training period and for each period of ongoing agreement checks. The percentages for coder agreement were as follows: original 10-year data, training (91.0%); original 10-year data, ongoing (85.1%); 5-year follow-up data, training (89.3%); and 5-year follow-up data, ongoing (83.4%).
Qualitative Procedure and Analysis
Qualitative content analysis methodologies are used to bring conceptual and theoretical coherence to textual data via either inductive or deductive approaches (Elo & Kyngäs, 2008; Finfgeld-Connett, 2014; Hsieh & Shannon, 2005). A deductive approach was used in this investigation, as coding categories of CRG and their intersections were already determined as a focus for analysis. Within this framework, the researchers further examined how intersectionality was manifest and tracked types of designs, samples, analyses, and theories used to undergird studies, as well as the implications for practice, research, and social justice. Deductive coding systems in qualitative work remain open to alteration to accommodate emergent patterns (Finfgeld-Connett, 2014). Thus, research team members were instructed to make note of additional thoughts and observations regarding intersectionality during analysis.
Articles that were determined to be at a primary level on all three variables of CRG in the quantitative content analysis were selected for the qualitative content analysis. This resulted in 28 articles. Each article that met criteria was examined in its entirety. Team members took notes as they conducted their separate analyses.
The primary author independently read and analyzed each article without referring to the data that the research team members had compiled. After the primary author had completed her own summaries, the team members’ summaries were reviewed for consensus, discussed, and integrated. All team members’ notes were compiled, and a matrix was created, tracking key elements in these articles. After the matrix was completed, as a final step, the second author reviewed the matrix and read each of the articles independently and noted any incongruities or questions. These were resolved via discussion and consensus with the first author. This iterative and reflective process is recommended to offset individual biases or oversight by any single analyst (Finfgeld-Connett, 2014).
Results
Research Question 1
The first research question asked in what number of articles and with what frequency the variables of CRG were studied together in TCP and JCP over the 15-year period from 1996 to 2011. A total of 1,440 studies were reviewed in the analysis. Of these, approximately 40% contained some mention of all three variables of CRG. See Table 2.
Use of Class, Race, and Gender Variables in Articles in TCP and JCP, 1996 to 2011.
Note. Inclusion = the variable was used alone or in any combination with the other two variables. TCP = The Counseling Psychologist; JCP = Journal of Counseling Psychology.
Across both journals, very small percentages (1-2%) of articles included class alone or in combination with gender or race. In contrast, it is notable that one third of articles included the intersection of race and gender, representing the most common intersection of two social locations. These results reinforce the contention that class is understudied in comparison with race and gender, and when two social locations are examined conjointly, class is least likely to be included. Interestingly, if the 481 articles that reported on both race and gender had included class, the overall inclusion rate would have jumped from 39% to 72%.
A positive trend was evident in the last 5 years of the analysis, however, with the number of articles including class steadily growing besides race and gender. See Figure 1 and Table 3.
Trends comparing qualifying articles across journals.
Number of Qualifying Articles and Number of Total Articles by Year.
Note. TCP = The Counseling Psychologist; JCP = Journal of Counseling Psychology; CRG = class, race, and gender.
Percentage of total articles that included the use of CRG.
We conducted a Mann-Whitney U test to examine the difference between the average of the proportions of articles including class–race–gender for the 10 years, 1996 to 2006 (n = 11), and comparing this with the average of the proportions for the 5-year update, 2007 to 2011 (n = 5). A nonparametric test was preferred over a t test due to the small sample size involved. The test yielded a significant result, as the distribution of scores did not overlap at all (U = 0, Critical U = 9, p < .05). This demonstrates significantly greater attention to class issues and intersectionality of the three variables under consideration in the most recent 5-year period of the 15-year analysis.
Research Question 2
Research Question 2 stated, “Of those articles that included all three variables of CRG together, at what level were these variables used?” Class was used most often in the Mentioned category in both theoretical and empirical articles. Gender was used most often in the Primary category in empirical articles, and race was used most often in the Primary category in theoretical articles. In the Integrated category, there were minor differences between the three variables. See Figure 2.
Percentage of articles incorporating class, race, and gender categorized by level.
Keeping in mind that the only category representing the actual integrated use of a variable in analysis is the primary level, it is again notable that class was much less frequently used (13%) at the primary level compared with race (42%) or gender (38%). These numbers reveal a consistent story: In comparison with both race and gender, class, if included at all, is most often used in the Methods section only, and is typically solely reported as a demographic variable, failing to receive any additional attention. Researchers have been more consistent in the application of race and gender variables, as compared with class variables, throughout their studies.
Research Question 3
Research Question 3 asked, “From an intersectional lens, what important qualitative themes might emerge from an analysis of those articles that did study CRG at the primary level?” Only 28 of 1,440 articles (1.9%) used all three variables (CRG) at a primary level. Of the 28 articles, 10 were found in TCP (37%) and 18 came from JCP (63%). These 28 articles were reviewed for qualitative emergent themes, specifically to search for intersectionality approaches and how intersectionality was operationalized, as well as the integration of social justice perspectives, which are key to intersectional work (Bowleg, 2008; Cole, 2009). Results are summarized below. The 28 studies used in this analysis are not cited within this article, but are available as supplemental materials at tcp.sagepub.com/supplemental.
Use of intersectionality
This analysis revealed that 15 of the 28 articles (54%) that used CRG at the Primary level truly examined these variables together as nonadditive, interlocking, and simultaneous intersections of inequality/oppression, privilege, and power. In other words, intersectionality was clearly evident in these exemplary works. Vocational researchers continue to be leaders in this area, with 8 of 15 (53%) of these outstanding articles having a vocational focus. TCP published 4 of the 15 articles (three theoretical and one empirical), whereas JCP published 11 of the 15 articles (all empirical).
A number of themes related to intersectionality are notable in these 15 exemplary articles. Interestingly, no authors referenced intersectionality theory, and only one article used the word intersection (though the word interaction was common). Using McCall’s (2005) system for understanding methodological approaches to intersectionality, there were no examples of any researcher using an anticategorical approach. Quantitative articles could arguably be said to have embraced an intercategorical approach via the statistical analysis of interaction effects and intercorrelations of discrete CRG variables. These statistical analyses were quite sophisticated, most typically using structural equation modeling. Qualitative studies fell in McCall’s intracategorical approach by recreating boundaried conceptualizations of CRG, yet providing verbatim examples of how individuals construct their own integrated sense of dynamic and blurred social locations. Most researchers acknowledged the difficulties in, and artificiality of, separating aspects of identity. The most often-cited theoretical perspective in the 15 exemplary articles was social cognitive career theory (Lent, Brown, & Hackett, 1994), and many articles made reference to various models of identity development related to gender, acculturation, and/or race and ethnicity. The three theoretical articles that embodied intersectional approaches all went beyond mere summaries of literature in a particular domain to synthesize information and bring together data on CRG that had not previously been integrated.
In the remaining 13 articles (46%), CRG were used in a manner that met qualifications for “primary” inclusion, but did not synthesize the three variables in an intersectional way. This was most often due to a stronger focus on two social locations with a clearly lesser emphasis on a third. In two-thirds of cases, that lesser emphasis was on class. Two other reasons for missing the intersectional mark were (a) the use of separate analyses for each variable and largely separate discussion of those variables, or (b) reverting to generic statements about the importance of these variables in Discussion sections rather than specifically addressing the dynamic and synergetic interplay between CRG.
Implications for social justice/advocacy
All the articles were found to have a broader purpose of advocacy for a particular group or issue targeted in the article. Many suggestions were education based, focusing on topics such as how to enhance economic independence, teach effective family coping, develop adaptive health behaviors, or to better understand the influence social class backgrounds can have on people’s career aspirations and choices. When articles focused on a broader social issue, suggestions for policy and social action were routinely included.
Throughout the 28 articles reviewed, applied actions that counseling psychologists can take, including revisiting existing vocational theories and developing new theories addressing the realities of people who are not male, middle class, and/or White, were provided. Counseling psychologists were encouraged to expand their role repertoire, consistent with a social justice agenda, to include roles such as consultants, liaisons, program developers, and evaluators.
Other findings
The 28 primary articles all included implications for research, such as considering within-group differences (be that related to class, race, or gender), expanding underlying theoretical perspectives (e.g., vocational development) to include social class, reexamining the comparative roles of individual vs. contextual factors, focusing on health and mental-health prevention research, and addressing ongoing concerns regarding assessment of social class. Many articles also included implications for psychotherapy or training.
Discussion
The results of this investigation show that 39% of articles published in TCP and JCP in the 15-year period under review contain some mention of CRG together, and the rate of articles published including the class–race–gender triad has been growing, particularly in the last 5 years.
However, class issues continue to be relatively ignored and are infrequently integrated as a meaningful variable throughout a study. In comparison with race and gender, class continues to be used most often only as a demographic variable or briefly mentioned in Discussion sections as an item for future research consideration. When demographic variables are considered in only pairwise conjunction, race and gender are addressed jointly at a rate approximately 11 times higher than class in conjunction with either gender or race. At a purely numerical level, it is therefore unsurprising that studies incorporating all three variables remain in the minority. In addition, a much smaller percentage (1.9%) used these three variables in an integrated way throughout an article, and even fewer (1%) achieved a fully intersectional approach.
Implications for Theory
As noted earlier, intersectionality theory is gaining in popularity among psychologists (Bowleg, 2008; Cole, 2008, 2009; Shields, 2008) as the discipline grapples to define it, solidify its theoretical constructs, and learn how to capitalize on its full advantages in research and practice. However, the explicit use of intersectionality theory (i.e., use of the term and/or citations by major researchers in intersectionality theory) was not evident in the articles reviewed. Yet the results of the qualitative content analysis in the present investigation highlighted the usefulness of an intersectional approach. Readers who wish to see exemplary illustrations of intersectional theory operationalized may benefit from examining the studies from the supplemental materials list. The present study itself is also framed by intersectionality; the simultaneous examination of CRG lead to a clear portrayal of how class is neglected in comparison with race and gender and what is missed when doing so.
The results of the primary articles also confirmed some key aspects of the social class worldview model (SCWM; Liu et al., 2004), a theory that includes attention to the subjective and individual perceptions of social class groupings. Studies analyzed in the current study often reflected the deep internalization of social class experiences and their lasting influence on people’s worldview, identity, class mobility and career aspirations, agency, and life outcomes. One of the most fundamental shifts psychologists can make is understanding that experiences of classism, racism, and sexism are the result of institutional inequities and broader cultural context, rather than individual, internal qualities (Lott & Bullock, 2007; Smith, 2010). Intersectional research is just one way in which this shift in mind-set can manifest, and serves as an important and transformational one to the profession. As noted by Cho et al. (2013), intersectional scholars “. . . embody a motivation to go beyond mere comprehension of intersectional dynamics to transform them” (p. 786).
It is clear from the present investigation that although authors who published in counseling psychology journals are beginning to use a range of both quantitative and qualitative methodologies that support an intersectional perspective, few have claimed intersectionality theory as a starting point for their work. This is somewhat paradoxical, given the deep roots in advocacy and social justice that counseling psychology embraces (Mallinckrodt, Miles, & Levy, 2014). The present authors suggest that increased focus on intersectionality theory could provide a fruitful conceptual framework as scholarship advances. Across disciplines, intersectionality theorists have started to articulate specific ways in which to do this, such as combining theories that examine identities with those that examine power differences (e.g., Mattsson, 2014) or Weldon’s (2006) nuanced discussion of how to think about CRG as being inextricable/simultaneous as well as having independent additive and multiplicative effects.
Implications for Research
The results of the current study provide continued evidence supporting the lack of class inclusion relative to race and gender, highlight the inconsistent ways in which social class has been measured, and bring to light that archaic instruments are still being used (i.e., Duncan, 1961; Hollingshead, 1958, 1975). These results support the conclusions of Liu et al. (2004) that there is an urgent need to develop effective and sophisticated psychological measurements to assess class; similar measures are lacking in relation to the intersection of CRG. APA’s (2006) Report of the APA Task Force on SES provided some initial guidance on these issues for interested researchers. Recently, Diemer, Mistry, Wadsworth, Lopez, and Reimers (2012) provided a more in-depth guide to conceptualizing and measuring social class, detailing and evaluating measures of subjective social status, educational attainment, income, occupational prestige, and markers of poverty or wealth. Such a guide allows researchers to more carefully match assessments of class with their research questions.
In theoretical articles analyzed in the current study, which were mainly found in TCP, race was used most often in the primary category, with gender and then class less frequently coded in this category. In the empirical articles reviewed across both journals, class tended to be reported in the Methods section but was not analyzed. Ideally, researchers would not only consistently include CRG in both their theoretical and empirical studies, but would also use these variables purposefully and meaningfully (Cole, 2009).
On an optimistic note, the articles that used CRG at primary levels in the present study demonstrated significant congruency with leading antipoverty researchers’ recommendations (APA, 2006; Lott & Bullock, 2007; Silverstein, 2006; Smith, 2010), which include research that is related to broad sociopolitical realities and change. This congruence implies that certain segments of the research community understand the complexity of intersectionality and the need to study it. Although studying intersectionality may require innovative methodology or sophisticated analyses (Warner, 2008), the potential value of this approach was clearly demonstrated by the meaningful conclusions of the primary articles in this review.
Examples of both qualitative and quantitative methods are available for scholars who wish to pursue intersectional research. In addition to the exemplars found in the present study, for instance, Robson, Anisef, Brown, and Parekh (2014) used multilevel modeling and multinomial logistic regression to examine both individual- and school-level variables in their study of high school students with special needs. Sarno, Mohr, Jackson, and Fassinger (2015) used an innovative combination of hierarchical linear regression with measures that were intersectional themselves to capture tensions in the multiple identities of lesbian, gay, and bisexual (LGB) persons of color. These sophisticated quantitative techniques complement the historical focus on the use of qualitative methods such as constructivist phenomenology (e.g., Greyerbiehl & Mitchell, 2014) to uncover lived experiences of those with intersecting identities in their own words.
The results of the present study also have implications for counseling psychology’s top journals as they relate to editor and editorial board selection, journal missions, journal reviewers, and researchers. Although editors are allowed to implement their own vision of how to direct a journal, these data could usefully inform publication directions from a multicultural and social justice perspective. It is encouraging to see the significant increase in publication of research including the class–race–gender triad over the 5-year period from 2007 to 2011; this momentum should be maintained. This expanded examination of class in the literature reflects not only editorial decisions but also, at a fundamental level, increased production and valuing of research on class, as well as of intersectional research that includes social class. Our hope is that this shift in published research indicates the continual embracing of social class as an important diversity variable that will continue to be examined by scholars in the field.
Quite to this last point, the primary author, a former CSES member, notes the current work of the APA CSES. They have recently begun a #stopskippingclass campaign, a letter-writing effort to the APA Council of Editors, Publications and Communications Board, which has a grassroots mechanism targeting authors published in APA journals in the last year. This campaign is in direct response to APA policy from the 2000 Resolution on Poverty and SES, as well as the charge to the CSES. APA is encouraged to implement the recommendations of the resolution, in particular, that APA “will recommend that where possible and appropriate socioeconomic status be identified for published reports of social sciences research.” At a minimum, key aspects of SES should be assessed and reported for the participants, and when possible, included in key analyses. The resolution also encourages researchers to report the intersection of SES with gender, race and/or ethnicity, age, sexual orientation, and disability status. Readers interested in learning more can follow the link http://www.apa.org/pi/ses/ for ways to get involved.
Limitations and Future Research
The content analysis and methodological review conducted in this study examined only the two leading U.S. counseling psychology journals, limiting the overall generalizability of this study; coverage of the intersection of CRG may be more (or less) evidenced elsewhere. Interrater agreement, although certainly satisfactory, could have been improved and some categorization errors may have gone uncorrected. Replicating this current study using journals, local or international, with a multicultural or gender focus to see to what extent intersectional approaches are being used, would provide a meaningful future contribution.
When determining articles for inclusion, gender was primarily used as a binary (male or female) variable, even though terms such as transgender were included in the initial search. In a related vein, although race and ethnicity are not equivalent, both words were used interchangeably for the purposes of this study. Other critical intersections are worth noting, but were outside the scope of this article including sexual orientation, disability, religion, formal education, citizenship status, size, nationality, and first-language status, to name a few. The present authors’ focus on CRG limits our knowledge of the degree to which researchers are including class as it intersects with other diversity variables.
Indeed, intersectionality theory arose and remains as one way for Black women to counter the exclusion and marginalization they experienced from both Black men and White, middle-class feminists (Barretti, 2015); hence, intersectionality has been consistently associated with oppression (Collins, 2000). As intersectionality theory emerged, it has been critiqued at many levels for both its contributions and its limitations in scope and content (Cho et al., 2013). A logical extension is to consider expanding intersectional approaches to include the spectrum of diversity variables, making it comprehensively multicultural and inclusive. For example, even though inclusive intersectional research may construct contradictory social patterns of both oppression and privilege, there is much to be learned about how individuals negotiate such tensions. Similarly, important information about the experiences of privilege can be learned from the study of White men and those with wealth. The field would benefit from exploring these psychological issues, sharing both common and disparate perspectives (Levine-Rasky, 2011; Luthar, 2003; Thomas, Dovidio, & West, 2014).
Conclusion
The sustained voices of psychologists working at the forefront of class issues appear to be making an impact. The present study documented higher levels of inclusion of the intersection of CRG in counseling psychology’s top two journals, particularly in the last 5 years. However, the results of this investigation indicated that class, when included, was given less thorough attention than race or gender, and that intersectionality theory is yet to be embraced as a conceptual framework for counseling psychology research.
Looking back to the 1970s, race and gender were ignored as important variables of interest to counseling psychologists. Buboltz, Miller, and Williams (1999) stated that in 1973, studies focusing on women as participants were rare (7%) and Ponterotto (1988) concluded in his content analysis spanning the years 1976 to 1986 that race and ethnicity were infrequent variables studied in psychological research (6%). Hopefully, many years from now, others will look back with near disbelief and say, “Remember when only 2% of the articles in our journals included an intersectional consideration of CRG?”
Footnotes
Acknowledgements
We wish to thank our student research team members Melissa Adams, Dominique Ash, Kimberly Boyd Bowman, Tasha Dorsey, Veronica Eisenhardt, Angela Hungrige, Kim Jones, Lisa Renee Jordan, Rachel McChesney, and Heather Rodges for their contributions. We are also indebted to Kathryn Keller Lamar for her thoughtful and helpful review.
Authors’ Note
Parts of these data were presented in August 2007 at the 114th Annual American Psychological Association Convention, San Francisco, California, in a symposium, “Class, Race, and Gender—Intersections, Inequities, and Implications for Change.” Also, parts of these data were presented in January 2013 at the National Multicultural Summit, Houston, Texas, in a symposium, “Class Intersections: Integrating Theory, Research, and Advocacy.”
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
