Abstract
Apprenticeship is a time-tested learning model and workforce development strategy. In the United States, registered apprenticeship (RA)—a structured form of apprenticeship that is regulated and overseen by government agencies—has been expanding from skilled trades into new industries. To inform this expansion and contribute to the understanding of how the model serves apprentices and employers, the author undertook a systematic literature review and a qualitative synthesis of empirical research, interpreting findings on RA over a 30-year period. The review identified 36 scholarly studies and an additional 98 articles from the practitioner literature. The synthesis derived three themes: expansion, benefits, and outcomes for minoritized groups, revealing gaps in research on the curricular and instructional design elements of RA that promote retention, completion, career advancement, and learning for apprentices.
Apprenticeship has long been used to facilitate the passing of skills and knowledge from experienced workers to the next generation. The principles of this time-honored model have been incorporated into influential theories of learning (Brown et al., 1989; Lave & Wenger, 1991) and have informed thriving systems of transition from school to work in countries such as Canada, Austria, Germany, Switzerland, and the United Kingdom (Dummert, 2020; Fortwengel et al., 2021).
In the United States today, a common and well-funded apprenticeship model is registered apprenticeship (RA). Federalized in 1937 by the Fitzgerald Act (also known as the National Apprenticeship Act) and further regulated by the Taft–Hartley Act of 1947, RA is both a model of adult learning and an approach to workforce development. The essential elements of RA are a paid job with a participating employer; on-the-job training (OJT), supervised and mentored by an experienced other; related technical instruction (RTI), which incorporates classroom-based instruction of job-related concepts and skills; structured wage progression; and career advancement along a predetermined pathway, leading to an industry-recognized credential. Outside of these essential elements, models for RA programs (RAPs) can vary based on whether unions are involved (joint or nonjoint) and whether one or several employers sponsor the training (nongroup or group). RAPs take 1 to 6 years to complete, depending on the occupation and industry. Sponsors and apprentices finance training together: sponsors contribute resources, instruction, and funds, while apprentices may accept lower wages for the duration of the RAP. Most RAPs have historically been in the building trades (Bilginsoy, 2005; Sharp & Dvorkin, 2018). This fact is due in part to the Davis–Bacon Act of 1950, which requires the payment of prevailing wages on federally funded construction projects and stimulated the growth of RAPs in this industry. It is also important to note that around half of all apprenticeship programs are unregistered, and it is likely that there are more unregistered apprentices than registered (Jacoby & Lerman, 2019). While the current study excludes these programs, they are a key area of current and future research.
RA persists within a rapidly changing context. The current U.S. labor market can be characterized as increasingly based on knowledge and service work, with a growing polarization between high- and low-quality jobs and between higher- and lower-educated workers (Buera & Kaboski, 2012). Employment opportunities in manufacturing and agriculture have significantly declined, while lower-quality, more precarious service-sector jobs (i.e., jobs with low levels of compensation and fringe benefits, minimal job security, and few opportunities for advancement) have proliferated (Kalleberg, 2011). Noncollege-educated workers are much more likely to hold these lower-paying jobs (Kearney et al., 2015), as are non-White and immigrant workers (Kalleberg & von Wachter, 2017). As these lower-quality, frontline service occupations have grown, so has economic inequality (Kalleberg, 2011). Since RA is a notably low-cost career advancement option for apprentices, this frontline worker population is poised to benefit from the growth and expansion of RA in the United States, particularly in the service industry (Kuehn, 2019).
While the promise of RA for these workers seems evident, a scholarly basis for this enthusiasm is difficult to find. Much of the research on apprenticeship has been conducted in European and Australian contexts, which differ greatly from the United States RA model (Bosch & Charest, 2008). As RA expands into new industries, notably the service industry, sponsors are changing the model in ways that raise important questions about what promotes learning, retention, and completion in RAPs, particularly among frontline service and health care workers. Without an empirically based understanding of what contributes to RA's success and failure in the U.S. context, there is a risk of diluting the model and threatening its efficacy. Therefore, this study answers the research question, “What are the trends and impacts of registered apprenticeship on apprentices and employers based in the United States in the last 30 years?”
Methodology
To answer the research question, I conducted a systematic review of all empirical literature on RA in the United States in a 30-year period (January 1991–January 2021) and undertook a qualitative research synthesis of the findings of those studies. A systematic literature review provides a comprehensive overview of existing literature, summarizes findings, informs policy and practice, and identifies gaps in research, with the goal of informing future research (Okoli, 2015). I chose a critical interpretive synthesis approach (Dixon-Woods et al., 2006; Heyvaert et al., 2016), which aims for integration of qualitative, quantitative, and mixed-methods research findings, viewed through the lens of theory. Critical interpretative synthesis incorporates the constant comparative method of data analysis and an inductive process for generating themes. This approach allows for consideration of the impacts of RA in various contexts and for different groups of learners but requires a critical stance toward the data. In the case of this study, I ask questions about the primary authors’ emphases on economic outcomes of apprenticeship and their de-emphasis of the role that program design choices play in learning and success outcomes for apprentices (Heyvaert et al., 2016). This inductive methodological approach allows for theory generation (Dixon-Woods et al., 2006).
Search Strategy and Data Screening
I began by defining the parameters used in the completion of a series of searches. My initial review identified many conceptualizations of apprenticeship (e.g., unregistered youth apprenticeship, pre-apprenticeship, internships, and mentoring arrangements). These models, while usually sharing several of the components, do not offer all the benefits of structured and regulated RA, such as guaranteed employment, wage progression, the pairing of classroom-based training and OJT, and industry-recognized certification. To learn about the complex interplay of RA's structural and programmatic elements, curriculum and instruction, and a particular learner population—that is, U.S. apprentices employed in an apprenticed occupation—I limited my investigation to RA.
I prepared an initial protocol with a data extraction form and developed a search strategy that culminated in the use of apprentice* in the subject and United States OR America OR USA OR U.S. in any part of the text. To narrow the focus, I excluded papers that included the term cognitive apprentice*. I limited the search to articles published between January 1991 and February 2021, a 30-year period that has seen much societal and economic change in the United States. The databases I searched were Education Research Complete, APA PsycInfo, CINAHL Complete, MEDLINE, and Business Source Complete. I followed this library search with two Google Scholar searches of the terms apprenticeship and registered apprenticeship AND United States from 1991 to 2021 and used the first 300 citations from each search (Haddaway et al., 2015). After the removal of duplicates, I had an initial set of 1,206 articles. My search yielded practice literature, such as technical reports, white papers, and dissertations, which I included to comprehensively map the domain and avoid publication bias (Kitchenham, 2004). It is important to note that much of the important research on U.S. apprenticeship is funded by federal contracts, so technical reports and papers are a major source of information on this topic.
Next, I reviewed titles and abstracts using my initial inclusion and exclusion criteria, yielding 182 sources, 62 of them scholarly and empirical. My final eligibility criteria were (a) empirical research, (b) RA, (c) data from the United States, (d) published between January 1991 and February 2021, and (e) published in or translated to the English language. Articles were excluded if they instead reported (a) forms of apprenticeship other than RA, (b) apprenticeship history, or (c) conceptual, advocacy, or journalistic articles. I read each of the articles, applying these eligibility criteria. Reading and the scholarly review process led to the inclusion of several additional eligible articles, garnered from reference lists, which had not been identified in initial searches. This screening resulted in my sample of empirical publications for analysis: 36 peer-reviewed articles and 98 nonpeer-reviewed articles, technical reports, white papers, and theses.
Data Analysis
Using a thematic analysis approach (Braun & Clarke, 2006), I familiarized myself with the data by reading and re-reading all the papers and recording initial analysis notes in my review diary. I generated initial codes from the peer-reviewed articles and inductively derived themes and subthemes from the data. I then reviewed these themes by checking them against the coded extracts of the entire data set, including the empirical papers from the practice literature, to refine the analysis.
Lincoln and Guba (1985) encouraged the enhancement of qualitative studies’ trustworthiness by establishing their confirmability, credibility, and transferability. I contribute to confirmability in this study by listing all included studies in the online supplementary Appendix, so other researchers may confirm my findings. To enhance credibility, another educational researcher selected and examined three studies in the scholarly research sample and compared them to my data extraction form. The researcher agreed with my categorization and analysis of the studies. A research synthesis approach contributes to the transferability of findings through the analysis of many studies that span a 30-year period and through the rich description generated through thematic analysis. My diary of review choices further enhanced trustworthiness by establishing an audit trail (Carcary, 2009).
Results and Discussion
I identified three themes: expansion (n = 14 scholarly and 49 practitioner sources), benefits (n = 18 scholarly and 49 practitioner sources), and outcomes for minoritized groups, including women and people of color (n = 13 scholarly and 32 practitioner sources). The analysis also uncovered attending questions and gaps in the literature. Below, I summarize and discuss each theme, then share questions, gaps, and implications related to learning. Apart from illuminating recent statistics from the National Skills Coalition (2020) and selected recent reports from the U.S. Department of Labor, all papers cited in the discussion of themes are those identified in my set.
Theme 1: Expansion
The first theme relates to the status of RA in the United States. Three subthemes comprise this theme: (a) increasing popularity, (b) changes in mindset, and (c) expansion to service occupations. This theme raises questions about the promotion of retention, completion, and higher wages in new RAPs and questions about learning.
Increasing Popularity
RAPs serve a very small number of workers in the United States, relative to the approximately 19.4 million students enrolled in higher education in 2020 (National Center for Education Statistics, 2021) and to the ubiquity of apprenticeship in many European countries (Decker, 2019; Gospel, 1994; Lazaryan et al., 2014). There are currently over 600,000 active registered apprentices in the United States, and numbers of new apprentices have increased 70% since 2011—more than 13,500 new RAPs were created between 2015 and 2020 (U.S. Department of Labor, 2021). It is important to note that over 100,000 of these are military apprentices and therefore do not contribute to the number of civilian apprentices available to the labor force (U.S. Department of Labor, 2021). The federal government and public policy organizations increasingly tout RA as an answer to secondary and postsecondary education systems that do not produce enough workers to fill available jobs. RA is held up as an affordable training and career advancement opportunity for adults with less than a college degree (Amoyaw & Brown, 2018). Indeed, most Americans have favorable views of apprenticeship and support increases in government funding for these programs (Snell et al., 2017).
Changes in Mindset
Yet, RA is often presented in U.S. society as a second-rate alternative for students who are not perceived as college-ready or cannot afford it (Workman, 2019). Strong emphasis on a college degree, coupled with the decline of labor unions and U.S. employers’ wariness of organized labor, has created a situation in which few young people are aware of apprenticeship as an option and are dissuaded from pursuing it if they are aware (Gospel, 1994; Rossmeier, 2015). In response to these issues, several papers encouraged a shift in the way students, parents, and our broader society view apprenticeship, citing successful apprenticeship systems in Europe as evidence for their arguments (Browning & Sofer, 2017; Cai, 2018; Craig & Bewick, 2018). Decker (2019) found that technical college students wanted their secondary and postsecondary institutions to provide more information about apprenticeship, particularly since it is so much less costly than college, and to contribute to a reduction in stigma around apprenticeship.
Expansion to Service Occupations
Most U.S. apprentices are concentrated in the building and skilled trades (Sharp & Dvorkin, 2018; U.S. Department of Labor, 2021). As several of the papers reported on one or more of these trades, most of what we know about RA in the United States is from the skilled trades context (i.e., RA in industrial and construction trades such as welding, ironworking, plumbing, and carpentry). Today, the RA model is also expanding into many new industries, such as tech and STEM fields (Gaudet et al., 2010; Kuehn & Jones, 2018; Kuehn et al., 2019; Sharp & Dvorkin, 2018), the insurance industry (Cheney, 2017), transit maintenance occupations (Glover et al., 2007), and early childhood education (Copeman Petig et al., 2019; Uttley & Horm, 2008; Workman, 2019). Hecker and Kuehn (2019) encouraged the expansion of RA within the U.S. prison system. RA has also made substantial inroads into the public sector, where RAPs were once uncommon (Riccucci, 1991; U.S. Department of Labor, 2021). The United Services Military Apprenticeship Program (USMAP), comprising around 180 occupations within the U.S Navy, Marine Corps, and Coast Guard (Hanson & Lerman, 2016; Lerman et al., 2015), saw its considerable number of active apprentices more than double between 2008 and 2020 (U.S. Department of Labor, 2021).
However, the area of RA expansion most salient in the literature review was the service sector (Browning & Sofer, 2017; Dimeny, 2019; Jopson et al., 2019; Kuehn, 2019; Mauldin, 2011). Defined broadly, service occupations include retail sales; frontline health care practitioners like nurses, nurses’ aides, and personal care assistants; cleaning and maintenance; and food preparation. The diverse workers in these positions are often referred to as frontline workers because they perform hands-on, essential tasks on the front lines of service-sector industries, often dealing directly with the public, representing their employer to others, delivering services or doing hands-on manual work, and juggling the needs of multiple stakeholders (Nichols et al., 2016). Approximately 32% of the U.S. workforce, or around 48 million people, are frontline workers (Bergson-Shilcock, 2017). In a study on apprenticeship diversity and expansion that falls outside this review's inclusion criteria, Sattar et al. (2020) found that 27 states used federal grants to target health care or biotechnology for apprenticeship expansion.
What Contributes to Retention, Completion, and Higher Wages?
Many studies encouraged increasing the number of RAPs in the United States to address major economic and social problems, such as youth unemployment, the middle-skills gap, and income inequality (Becht, 2019; Browning & Sofer, 2017; Koller, 2018). However, relatively few studies presented scholarly research on the design features of apprenticeship that contribute to retention, completion, and higher wages, much less mobility along a career path.
Studies show that, despite the increasing popularity of RAPs in the United States, less than half of participants complete their programs and retention rates are low (de Alva & Schneider, 2018; Hanson & Lerman, 2016; Lerman et al., 2015; Reed et al., 2012; U.S. Department of Labor, 2021). Among low-wage service workers, relatively short, joint programs had more completers (Kuehn, 2019). Berik and Bilginsoy (2000) noted that limited data and anecdotal evidence suggest that quitting apprenticeship is largely voluntary on the part of the apprentices, as reasons given included failure to pay union dues, show up for work, or attend classes. Rarer were instances of involuntary quits due to failed drug tests or fights. In the USMAP program, factors contributing to quits included changes in posts of duty and poor initial communication about the nature of the RAP, possibly leading to high levels of attrition early in the program and low perception of the value of the training among apprentices (Lerman et al., 2015). In a dissertation study, Swegle (2017) identified financial issues such as layoffs, lack of reliable transportation and one's own tools, and poor money management as primary reasons for quits. Swegle also identified poor attitude and work ethic as primary causes of attrition. Conversely, Browning and Sofer (2017) attested to successful practices that reduced barriers to apprenticeship for opportunity youth, such as waiving the requirement for a car and wait periods to re-take exams and appointing mentors who can help younger mentees deal with work-related frustration and provide other individualized supports. When it comes to higher wages, Kuehn (2019) concluded that longer RTI and training with the apprenticeship sponsor (as opposed to with a community college) correlate with higher exit wages among service workers. Outside of these disparate findings, which suggested some design implications, the literature is inconclusive on what elements contribute to increased retention, completion, and higher exit wages. This gap in the literature raises the concern that apprenticeship is expanding without a research foundation that points to the design elements that contribute to successful apprenticeship. This fact is likely due to the lack of federal and state data collection on these particular features. It is also likely because educational researchers have not focused much attention on apprenticeship as it is practiced in the United States. Because apprenticeship researchers are most often economists and workforce development evaluators, most of the research on apprenticeship has focused on the impacts and benefits of apprenticeship on employment, earnings, and other economic and employer-related outcomes. Addressing this gap in the literature will require educational researchers to take more interest in this model of workforce learning.
As RA Expands, What Features of New RAPs Promote Learning?
Much of adult learning theory tells us that people learn by doing (Brown et al., 1989; Lave & Wenger, 1991). In apprenticeship, the combination of active OJT for which apprentices are paid, reinforced by in-class theoretical instruction that prepares the worker for lifelong learning in a dynamic industry, can result in “education that transcends the textbook and extends to physical and even cultural learning” (Rossmeier, 2015, p. 1). However, due to the lack of state and federal data collection on these topics, the reviewed studies are largely silent on which design and programmatic elements contribute to learning in RA and how they do so. This is partially due to the unique position of RA in the United States and underlines why researchers cannot always draw on studies conducted about overseas contexts to inform RA design at home. Though RA is regulated by federal and state government agencies, it is an employer-led and largely decentralized system. Individual U.S. apprenticeship programs can have their own content and quality standards, which can make programs difficult to compare to each other. Unlike European systems, which incorporate a strong social partnership between industry and education at the national level, RAPs are often wholly separate from the secondary and postsecondary educational systems. Moreover, proprietary RA curricula and approaches are often quite guarded, so educational researchers do not have ready access to them to conduct studies that would shed light on the benefits or drawbacks of instructional practices and curricular approaches. These realities hinder expansion and create the risk that RA expands in the United States without enough knowledge about how to optimally design these experiences for learning, particularly among diverse, frontline service workers.
Theme 2: Benefits to Society, Economy, Apprentices, and Employers
The second theme summarizes the benefits of the RA model, organized by three subthemes: (a) benefits to society and economy, (b) benefits to apprentices, and (c) benefits to employers. This theme raises questions about how learning is promoted in RAPs.
Benefits to Society and Economy
Over half of all jobs in the United States are described as “middle skill”—that is, requiring more than a high school degree but less than a bachelor's degree—but only 43% of workers can access training for these occupations (National Skills Coalition, 2020). The term middle skill can be misleading, as it can be interpreted as disparaging workers’ knowledge and abilities. That is not how it is used here; instead, the term is meant to refer only to the amount of education typically required to obtain and persist within these occupations in the United States. I found frequent references to this middle-skills gap and the promise of apprenticeship for upskilling the service workforce, replacing retiring Baby Boomers, and addressing unemployment, particularly among those with lower levels of education (Beer, 2019; Bosch & Charest, 2008; Crewe, 2020; Decker, 2019; Hanson & Lerman, 2016; Lerman et al., 2020; Shaw et al., 2019). Several authors pointed out that youth unemployment in the United States is high relative to Europe and that apprenticeship could help disadvantaged young people find more success in a job market where employers seek experienced workers (Ayres, 2014; Boi, 1993; Messing-Mathie, 2015).
The literature identifies social benefits as well. In a cost–benefit analysis of RA in 10 states, Reed et al. (2012) not only found that RA worker productivity was higher but also found that RA participants used social safety net government programs (such as welfare, food stamps, and unemployment insurance) less, particularly when tracked over a participant's career. Evaluation studies of early educator and child development RAPs found better classrooms and positive outcomes for young children (Uttley & Horm, 2008) and perceptions of higher quality of child care and engagement with parents (Copeman Petig et al., 2019). Becht's (2019) findings suggested apprenticeship expansion can reduce income inequality.
Benefits to Apprentices
Nearly all apprentice completers (91%) maintain or find employment after completion (Helper et al., 2016). RA participants earn substantially more than nonparticipants (Berik & Bilginsoy, 2006; Hanks et al., 2018; Kuehn, 2019; Manzo et al., 2019; Reed et al., 2012). Hollenbeck and Huang (2016) found that RA had a greater impact on future earnings than any other workforce training program in Washington state; though outside the study's time parameters, more recent return-on-investment research in Washington has reinforced these findings, showing that apprentices earn over $30,000 more after exiting their RAP than they would have if they had never participated (Dula, 2021). The low-wage frontline worker population (Kuehn, 2019) is particularly well positioned to benefit from the growth and expansion of RA, as it is much less costly than a traditional college education, provides an industry-recognized credential at completion (Lerman, 2016a), and leads to permanent employment (U.S. Department of Labor, 2021) and more opportunities for advancement (Lerman et al., 2014) after completion.
Benefits to Employers
Employers who sponsor apprenticeships invest significant resources into worker and instructor pay, materials, and overhead, but research indicates they reap a substantial return on their investment. Benefits to employers include higher levels of performance among apprentices, including increased productivity, flexibility, and enhanced problem-solving ability (Helper et al., 2016); reduced turnover and improved recruitment (Payne, 2016); and a ready-made pipeline of skilled, enculturated, engaged, and loyal employees with less need for supervision and a reduction in overtime costs (Helper et al., 2016; McCarthy et al., 2017). RA has also been shown to improve workers’ soft skills, quality of care, and patient outcomes and increase client satisfaction when implemented in the service industry (Helper et al., 2016; Lerman et al., 2014; Love & McCarthy, 2018).
What Should Be the Role of Community Colleges in RA?
The articulation between the secondary and postsecondary education systems and the RA system is weak in the United States in comparison to the European apprenticeship powerhouses, such as Germany and Switzerland. While community college partnerships are increasingly relied upon to provide the RTI component of RAPs (Beer, 2019) and the practice literature encourages this reliance (McCarthy et al., 2017), research on the benefits of this practice has been contradictory. Partnerships with community colleges can have many potential benefits, including a reduction of sponsors’ costs, since expenses related to materials, curriculum development, recruitment, and instruction could be shared with the public education system (Helper et al., 2016). Such a partnership can also benefit apprentices by offering them college credit that is aligned with degree programs (Workman, 2019). Cantor (1994, 1995) analyzed factors that contributed to successful partnerships with community colleges and U.S. Navy apprenticeships, concluding that collaboration was successful when it provided access to funding the parties could not obtain on their own and when it was structured to allow for the mediation of conflicts. Fuller (2016) found that completion of RA within the community college system in California led to the probability of employment in one's chosen industry after completion. Authors of several studies in the practice literature encouraged connecting RA to the higher education system to expand the model, improve outcomes after completion, and reduce students’ financial burden (Beer, 2019; McCarthy et al., 2017; Welton & Owens, 2017). Community colleges may also be able to contribute to the diversification of apprenticeship through their existing community-based recruitment mechanisms, their experience serving nontraditional students with barriers (Messing-Mathie, 2015), and strong partnerships with local employers, ultimately contributing to the expansion of apprenticeship, particularly in the service industry (Beer, 2019; Doty & Odom, 1997). Model programs in states such as South Carolina (Lerman, 2016b) spotlight partnerships with community colleges, which provide instructional technology and bring students to apprenticeship with local employers at a fast clip.
However, others warn that partnerships with community colleges can have drawbacks. Kuehn's (2019) quantitative study found that partnerships with community colleges resulted in reduced RAP completion rates and were associated with lower exit wages for some service occupations. Other challenges to partnering with community colleges include a lack of flexibility on degree program credit requirements, slow and unsystematic change mechanisms within colleges, and the heavy lift required for college commitment to RA, including equipping classrooms with appropriate technology, building strong relationships with local employers, and hiring qualified instructional staff with industry expertise (de Alva & Schneider, 2018; Gunn & De Silva, 2008; Nehls, 2019). Limited and disconnected federal and state funding for postsecondary education and apprenticeships, ongoing cuts to federal workforce training programs, and difficulties overcoming the traditional separation in the United States between education and business all potentially hinder the success of community college involvement in apprenticeship (Beer, 2019). In addition, there is concern that some of the benefits that community colleges offer to RAPs, including their ability to develop curricula, support instructors, and offer college credit to apprentices, lead to higher costs that will be passed on to the apprentice, undercutting one of the key benefits of apprenticeship to many disadvantaged learners—that is, the low cost (Prebil, 2019).
On What Evidence Do We Base Our Assumptions About the Learning Benefits of RA?
Many of the studies in the practice literature are evaluative and point to positive outcomes for different groups but stop short of the design and educational research that would help the field understand what combination of elements can lead to these desired outcomes. For example, Copeman Petig et al. (2019) found, in an evaluation of early educator apprenticeship programs, that apprentices report learning valuable knowledge and skills in their apprenticeship, greatly improving their practice with children and families and the quality of care. But the study does not describe the apprentices’ learning activities that led to these benefits.
Many of the reviewed studies do provide limited insights on what promotes or creates conditions for learning, but the studies are so distinct from one another in context, learning population, and purpose that it is difficult to draw conclusions for RA design. For example, in their analysis of benefits and costs of RA for businesses, Gaudet et al. (2010) evaluated a geospatial technology apprenticeship program, concluding that apprentices passed skills tests and perceived course work as relevant, suggesting some learning occurred, but the authors did not venture guesses as to why. In a small, quantitative study of electrical industry apprentices and instructors, one researcher found evidence that respondents reported a higher occupational work ethic than other study contexts in which their instrument had been used, but the author did not explain why or how (Hatcher, 1995). Porter et al. (1997) analyzed predictors of success in a medical service apprenticeship course in the U.S. Air Force and concluded that certain scores on select parts of the screening test were more predictive of success but used these findings only to recommend minimum test scores for admission into the RAP. Gordon (2003) surveyed apprenticeship instructors and reported the principal finding that beginning instructors were less likely to feel that they were effective teachers than those with more experience.
On the other hand, several studies suggested promising directions for future educational and design research. Several of these concluded that the experienced mentor, a key feature in any RAP, played an important role in apprentices’ learning and development, supporting the development of attitudes toward unions (Fullagar et al., 1992), in the outcomes for children served by child-development specialist apprentices (Uttley & Horm, 2008), and for opportunity youth in health care apprenticeships (Browning & Sofer, 2017). In addition, Helper et al. (2016) found that apprenticeships led to greater creativity and problem-solving ability and enhanced flexibility in completing tasks for apprentices, surmising that this is owed partially to the dual RTI and OJT components of the model, which create a strong foundation for understanding of the principles of the work. The role of RTI in learning, behavior modification, and reduction of injuries was highlighted in a mixed-methods study by Kaskutas et al. (2010) that detailed the development of safety training to reduce falls from heights in the residential construction industry and reported favorable evaluations from apprentices. In that program, designers introduced enhanced safety practices in RTI before workers went to the job site and reported that they infused their instructional approach with attention to adult learning principles, problem-solving activities, authentic practice, and role playing. Unfortunately, the authors gave very few examples of these approaches. In a report on the outcomes of a survey and interviews with employer sponsors, Colborn (2015) pointed out that curriculum was often designed by staff that work the job and thereby identify the skills to be taught, while Craig and Bewick (2018) criticized the siloed design approach that results in very different curricula in similar programs, with no standardization. A small study in the practice literature by Johnson and Lambeth (2017) recommended shortening the time delay between the demonstration and practice of a skill in a stagecraft RAP, advocating for teaching technical skills in small steps.
A few studies identified barriers to learning. Uwakweh (2006) surveyed construction apprentices and found that their motivation to proceed as apprentices was low and that they did not perceive value in their program. Uwakweh concluded that managers did not do enough to support and incentivize apprentices’ learning and retention. These findings were echoed by Hanson and Lerman (2016). Byrd (1999) discussed how women carpentry apprentices can feel marginalized in RTI classrooms by the competitive learning culture that values previous experience and the jargon, cursing, and off-color jokes used by the men.
At their core, RAPs are learning experiences. Their purpose is to support the learning of job-related skills and knowledge, guiding new workers through a structured experience that brings them closer to full participation in their chosen occupation and their community of practice. The reviewed studies highlighted the benefits of RAPs for society and the economy, employers, and apprentices themselves. However, because educational researchers have not focused on apprenticeship, we do not yet have many studies that investigate the specific design elements these learning experiences should have to ensure these benefits.
Theme 3: Outcomes for Minoritized Groups
Many more men than women are registered apprentices, and Whites comprise the largest racial group in RAPs (U.S. Department of Labor, 2021). In this theme, I explore the research on outcomes for groups that are minoritized in RA through two subthemes: (a) different outcomes in joint and nonjoint programs and (b) outcomes for women and other marginalized groups. I use the term minoritized to refer to groups of people that have been historically disenfranchised or pushed to the margins of mainstream U.S. society with economic consequences. In my use of the term, I do not refer to sizes of demographic groups in the population. In this paper, I refer principally to women, workers of color, and those with foundational academic skills gaps who are or may be at an economic disadvantage. This theme raises questions about how the learning needs of frontline service workers, disproportionately members of minoritized groups, are addressed in RAPs.
Different Outcomes in Joint and Nonjoint Programs
RAPs can differ by types of sponsorship. Most U.S. registered apprentices are in RAPs sponsored jointly by labor unions and employers (U.S. Department of Labor, 2021). These are joint programs. Other RAPs, nonjoint, are not sponsored by joint labor–management organizations. There are many more nonjoint programs, but they train a smaller share of apprentices overall (U.S. Department of Labor, 2021). One of the most salient findings from my analysis was that joint programs have generally better outcomes, including the highest exit wages (Kuehn, 2019), higher enrollments, longer programs and higher completion rates, and lower attrition rates than nonjoint programs (Argyres & Moir, 2008; Bilginsoy, 2005, 2007, 2018; Bradley & Herzenberg, 2002; Public Sector Consultants Inc., 2017). Joint programs have a well-established history of emphasizing quality and safety and successfully preparing apprentices for stable careers (Nehls, 2019).
Outcomes for Women and Other Marginalized Groups
These better outcomes extend even to nontraditional apprentices. Joint programs today employ a larger share of women and minoritized apprentices than nonjoint programs. Also, outcomes are better for minoritized groups in joint than nonjoint programs (Argyres & Moir, 2008; Berik & Bilginsoy, 2000, 2002; Olinsky & Ayres, 2013). Building trades unions’ reputation as historically discriminatory against women and other minoritized groups is well earned. However, today, after considerable regulatory intercession, joint construction RAPs have significantly better completion rates for these groups than nonjoint construction RAPs have overall for all groups combined, including White men (Glover & Bilginsoy, 2005).
This record for joint programs is an important finding because, historically, women are rare in RAPs compared to White men, particularly in the building trades (Berik & Bilginsoy, 2000; Byrd, 1999; Eisenberg, 2018; Riccucci, 1991). In the 1970s, antidiscrimination policies attempted to promote more participation by women in these programs. Women's representation in the construction industry has barely changed since then (Eisenberg, 2018), and studies show a pattern of decline (Berik & Bilginsoy, 2006). Barriers for women abound. Reports of severe discrimination, isolation, harassment, and violence against women in trades apprenticeship programs have been common for decades, and women of color report racism (Bower, 2007; Byrd, 1999; Eisenberg, 2018). Despite barriers and occupational segregation (i.e., the overrepresentation of demographic groups in particular careers), many women remain determined to participate due to RA's potential for quality training, career advancement, and higher wages (Berik & Bilginsoy, 2006; Bower, 2007), and it is recommended that RAPs promote access for women through outreach and information, support the development of foundational skills to help them succeed in the apprenticeship, assist with child care and peer support, and better address the harassment they experience on worksites that are dominated by men (Byrd, 1999; Reed et al., 2012). Berik and Bilginsoy (2000) placed the responsibility for eliminating barriers in joint programs with the unions, noting it “requires vigilance to change the present discriminatory practices and sexist attitudes” and that “their resources and mechanisms ought to be strengthened and put into use to promote equity” (p. 61).
Outside of the trades, newer apprenticeships in early education, frontline health care, and service occupations are much more successful in recruiting and retaining women (Workman, 2019), perhaps because many of these care jobs are historically performed by women. In an evaluation study of joint early educator apprenticeships in California, Copeman Petig et al. (2019) found that 98% of apprentices were women and most were of color. In a study of inclusion in apprenticeship, Kuehn (2017) found that the share of women in South Carolina RAPs was high, particularly those in nonconstruction occupations, compared to the national average due to systematic efforts by policy makers to promote diversity.
While women received the most attention in the RA literature, the experiences of other minoritized groups are also the subject of research. Building trades unions have had an ugly history of exclusionary behavior toward Black workers and other workers of color (Penn, 1998). Starting in 1969, the proportion of Black and Hispanic workers in trades apprenticeship programs has increased, owing partly to civil rights legislation, but overall progress has been modest. Today, Hispanic workers have particularly high representation in the building trades apprenticeships, while nearly a quarter of apprentices in service occupations are Black (Kuehn, 2019). These groups’ overall trade union participation rate stands at less than 15% (Helper et al., 2016).
Reflecting long-standing systemic inequities in rates of incarceration of Black men relative to Whites, apprentices in prison RAPs are more likely to be Black (Hecker & Kuehn, 2019), and while completion rates are slightly higher, wages are much lower, and some apprentices in prison are not paid at all (McGrew & Hanks, 2017). Authors have advocated for expanding RA in prisons, accompanied by higher wages, because RAP participation can facilitate justice-involved individuals’ access to jobs after release and significantly reduce recidivism (Saylor & Gaes, 1997).
Mollica's (2020) analysis of interviews with English as a Second Language teachers, program staff, and thought leaders led them to conclude that few English language learners participate in RA. Mollica also encouraged the use of pre-apprenticeship programs and inclusive training models to contribute to a more diverse and qualified workforce. Indeed, in a study that falls outside this review's inclusion criteria, Copson et al. (2021) observed that federal grantees used pre-apprenticeship programs to promote access to apprenticeship for marginalized groups. Browning and Sofer (2017) explored positive outcomes for opportunity youth in urban RAPs; praised the ability of these programs to connect this population to marketable skills, certification, and employment; and recommended attention to young people's social–emotional development in design.
How Are New RAPs Designed for Frontline Service Workers?
Frontline workers face several barriers to participation in education and training. Understanding these barriers and what research in adjacent fields tells us about how to address them are two steps in facilitating the expansion of apprenticeship to the service and health care industries. In an analysis of the Organisation for Economic Cooperation and Development (OECD) Survey of Adult Skills, Bergson-Shilcock (2017) found that 62% of service-sector workers have limited literacy, 74% have limited numeracy, and 73% have limited digital problem-solving skills. This is even though their frontline positions require the use of these skills; hence, workers who lack these skills tend to have less mobility and lower wages and are more likely to end up unemployed (Ross & Holmes, 2017). These workers are also not offered learning and development opportunities by employers with the same frequency that these are offered to more highly educated, higher-wage employees (Devins & Gold, 2014; Hedayati Mehdiabadi & Li, 2016; Institute for Corporate Productivity et al., 2016). In addition, frontline workers are less likely to take advantage of training opportunities when they are offered (Fouarge et al., 2013; Kyndt et al., 2011). Illeris (2006) found that negative past educational experiences can lead to a lack of self-confidence in these workers, which discourages further learning, while societal stigma associated with low basic skills can further discourage adults from admitting need and seeking support (Windisch, 2016). In addition, because low-paid, low-skill frontline work is perceived as low quality, repetitive, and offering few opportunities for advancement, these workers themselves often do not see a need to engage in training. They often have less free time, access to transportation, and money that further education and training would require, which further disincentivizes participation (Lloyd & Mayhew, 2010). When they do participate in adult education programs and workplace training, dropout rates are high (Windisch, 2016).
These barriers strongly suggest the need for evidence-based approaches to designs for learning for these workers, for whom RA may be a particularly beneficial approach. Kuehn (2019) found that apprenticeship leads to advancement and higher exit wages for service workers, is affordable in comparison to college (Amoyaw & Brown, 2018), and can promote the development of interpersonal, communication, and other “soft” skills that are especially useful in customer service–oriented industries (Lerman, 2016a). Since dropout rates are high among the frontline worker population, understanding how these workers build confidence and positive learning and occupational identities can help us design RAPs that emphasize retention, encouragement, and support, leading to higher rates of completion. In short, while apprenticeship is a useful model for many jobs and skill levels, RA can be a particularly practical and accessible path to middle-skills occupations—and the middle class—for millions of low-wage service workers. U.S. literature on RA in the last 30 years provides few insights on how to design curricula and programs to meet the learning and support needs of this large and in-demand group of workers.
Conclusions and Implications
In summary, through analysis of 134 empirical publications on the topic of RA in the United States over a 30-year period, I identified three themes: (a) RA continues to expand, including to frontline workers; (b) RA benefits employers, employees, and society despite costs by providing increased wages and stability; and (c) marginalized groups disproportionally benefit from but have historically lacked access to RA, though they are now increasingly accessing apprenticeship opportunities. The analysis also revealed that, due to lack of federal and state data collection on these elements and the orientation of apprenticeship researchers to economic outcomes, the reviewed studies offered little insight about how specific RAP design components contribute to apprentice retention, program completion, attainment of higher exit wages, and learning itself. The studies were notably ambivalent about the benefits of community college partnerships and inconclusive regarding what designed elements support frontline workers in particular, an issue given the increasing expansion of RA for frontline workers.
The results of this analysis have both practical and scholarly implications. Apprenticeship sponsors may benefit from the insights on the ongoing expansion of the model into new industries and find reasons to invest resources into its development, justified in their decision-making by research on the benefits for employers, apprentices, the economy, and meaningful social outcomes. Educators, instructional designers, and economic and program development professionals may take lessons from U.S.-based research on the barriers experienced by women, marginalized groups, and those with foundational skills gaps and use this enhanced understanding to create RAPs that are more accessible and equitable. Given the substantially better outcomes for all apprentices, particularly women and minoritized groups, in joint programs and the ongoing diminution of the U.S. labor movement, results of this analysis may also encourage practitioners to take a long look for inspiration and guidance at the educational and training practices that U.S. unions have honed over more than a century.
This analysis has revealed a deep gap in our knowledge of the design elements that contribute to learning, promote the development of agency and professional identity, and support success for all apprentices, including (but not limited to) those in frontline service and health care occupations. This systematic literature review showed that some researchers view learning and success in the apprenticeship context through a variety of different lenses, such as human capital theory and transformational learning theory (Howze, 2015); program theory (Hunt, 2020); and expectancy theory (Uwakweh, 2006), among others. However, most papers presented no perspective on how learning or identity development occurs in RA. Theory tells us that apprenticeship promotes learning. The model is social and experiential, providing ample opportunities for situated practice in authentic, on-the-job contexts, guided by experienced mentors (Brown et al., 1989; Lave & Wenger, 1991). In addition, RA seems particularly well suited to workers with foundational skills gaps who may prefer hands-on, active learning to a traditional classroom. Unfortunately, the designs of the RTI and OJT components of RA are rarely described in detail in existing research, leaving scholars with an impoverished understanding of how they work together with other programmatic elements to promote learning and success for apprentices.
Moreover, very few studies approached the development of agency, confidence, and professional identity among apprentices, though decades of educational research have indicated a fundamental relationship between workplace learning and the development of these characteristics (Cruess & Cruess, 2018; Freire, 1970; Lave & Wenger, 1991; Tett, 2016; Toom et al., 2021). Scholarly work has been conducted on the development of professional identity in apprentices in non-U.S. contexts (Chan, 2013; Duemmler & Caprani, 2017). For instance, Copeman Petig et al. (2019) found that nearly all surveyed apprentices reported more confidence in their roles as early educators because of their apprenticeship program and felt more likely to continue as an educator. Their responses indicated a growth in their sense of being a valued professional in a career field. The gap in our understanding of the development of apprentices’ agency and identity points to a need for scholarly research that could shed much light on effective designs for learning in apprenticeship.
The gaps in our understanding of how learning and identity development happen in RA impact our ability to design programs that promote these processes. As the model expands into new occupations and is used with apprentice populations who have different needs and barriers, RAP sponsors and designers necessarily adjust the model to fit these new contexts. With such sparse knowledge about what promotes learning in RA, we risk sacrificing those elements that have contributed to the model's persistence and success.
Theories of learning are useful because they give us a way to make sense of phenomena we observe. A theoretical lens helps us understand and explain what happens in apprenticeship; make decisions about program design, curriculum, and instruction; and predict outcomes. If we have no perspective on learning and if our approach does not arise from a point of view, we have very little basis on which to make decisions as we design an apprenticeship program, which is at its core a learning experience. Each sponsor or instructional designer thus begins essentially from scratch, building programs that reflect idiosyncratic understandings of how learning occurs and is supported, instead of being guided by well-developed theories based on analysis of research and practice in adult learning and the learning sciences.
I propose that RA practitioners become familiar with two social theories of learning that have explanatory power for the elements and outcomes of apprenticeship that were most salient in this literature review and analysis. The first is Lave and Wenger's (1991) theory of situated learning and communities of practice. This theory presented a framework for how people learn and reflects a view of learning that is not merely the cognition of the individual but instead is a fundamentally interactive and participatory process that is situated; in other words, it is meaningful to the extent that is relevant to and reflective of the authentic activities of daily life for those learners. In this theory, learners participate in communities of practice, which are groups of people that share a common set of understandings, interests, or problems and work together toward common goals. Through a process called legitimate peripheral participation, new community participants gradually move from the periphery of the community to its center, toward full participation in the community. The theory of situated learning and communities of practice naturally reflects some of the most intriguing aspects of apprenticeship as a learning model, including its highly situated nature, learning in an authentic occupational context and necessarily among knowledgeable members of a community that guide an apprentice to full participation over time. Among the design decisions this theory could guide are choices of learning activities in both RTI and OJT.
A second social theory of learning that can inform apprenticeship design is cognitive apprenticeship (Brown et al., 1989). In this theory, knowledge is fundamentally situated; it cannot be abstracted and separated from the situations in which it is used. Learners construct meaning together through social interaction and interaction with the material world through authentic tools and tasks. Learning is a form of enculturation, and cognitive skill acquisition is guided by coaches or master teachers (Collins et al., 1989). The theory of cognitive apprenticeship can provide rich insights into the process of mentorship in RA, among many other design decisions.
These social theories of learning contribute a perspective on the development of identity as well. For Lave and Wenger (1991), learning itself entailed the development of identities, which they defined as “long-term, living relations between persons and their place and participation in communities of practice” (p. 53). The movement to full participation in the community is the most salient characteristic of learning. Through learning, the individual changes fundamentally: they view themself differently and are viewed differently by others. Identity changes slowly; participation in the community and the journey from novice to expert is an ongoing and transformative process. This theoretical viewpoint has implications for the types of experiences we prioritize in our apprenticeship designs, and the time we allow for them, if we want to contribute to the evolution of the individual as a full community participant.
Limitations and Future Research Directions
The findings of any research synthesis are limited by the studies selected for inclusion and the quality of those studies. I did not include research on un-RAs, which may have led me to miss key innovations present in this diverse group of apprenticeship programs. I also did not include a systematic analysis of study quality as part of my inclusion criteria; as such, I may have included some studies of low quality. In combining results from several studies that used different theoretical approaches and methodologies, I ran the risk of sacrificing some rigor and transferability of findings. In addition, it is possible that my somewhat emergent review protocol led to missing some studies that would have enhanced or contradicted my findings.
More empirical research must be carried out on the impacts of apprenticeship overall, as well as on how RAP design impacts retention, completion, exit wages, and learning. In particular, increased scholarly attention must be paid to the learning needs of frontline apprentices and their ability to access high-demand, family-sustaining, middle-skills occupations. Many conceptual and nonscholarly papers in the practice literature make claims for the promise of RA, without rigorous educational research that provides a foundation for such claims. Targeted research into the role of community colleges in RTI is necessary to inform the rapid expansion of service industry RAPs that are utilizing these systems. More research on un-RAs would also add depth and new perspectives to elements that promote and detract from apprentice success across industries. Cross-walking adult learning research from other scholarly domains can assist workforce learning and economic development researchers in identifying factors that support apprentices to enroll, persist in, and complete their RAP and enjoy increased mobility along a career pathway in the years after completion. In addition, it should be noted that much of the practice literature centers on the needs of employers; we need more research that focuses on the design of RTI and OJT and how both can promote apprenticeship success over the course of a career and a lifetime of learning.
Supplemental Material
sj-docx-1-edq-10.1177_08912424231196792 - Supplemental material for What We Know About Registered Apprenticeship: A Systematic Review and Synthesis of 30 Years of Empirical Research
Supplemental material, sj-docx-1-edq-10.1177_08912424231196792 for What We Know About Registered Apprenticeship: A Systematic Review and Synthesis of 30 Years of Empirical Research by Amber Gallup in Economic Development Quarterly
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
Author Biography
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.
