Methods

The purpose of this analysis is to better understand the landscape of short-term training programs, who pursues short-term training, and the labor market outcomes of those who complete these programs. Three main questions guided this descriptive analysis:

1. Which entities deliver short-term training programs, and for what fields do short-term training programs aim to prepare students?
2. What are the demographic characteristics, in terms of gender, race, and prior education, of individuals who enroll in short-term training?
3. What are the completion, employment, and earnings outcomes of short-term training program graduates?

In summer 2020, New America requested and received data on Washington’s training programs from the Washington Workforce Training and Education Coordinating Board (WTECB), the state’s governor-appointed workforce board. The WTECB is responsible for implementing standardized accountability measures for the state’s workforce development system. As part of its accountability function, the workforce board maintains training program participation and performance data. Additionally, it links individual participant records reported by each training provider with unemployment insurance (UI) wage records and other administrative databases, such as the State Wage Interchange System (SWIS),¹ to compute statistics on the employment and earnings of training program completers. Through the Washington Career Bridge website, the workforce board publishes detailed enrollment and outcome information for education and training programs that charge tuition in the state.² The information analyzed for this report was compiled from two sets of extracted data from the Career Bridge site.

The information contained in this report is based on an analysis of 10,976 training programs offered in the State of Washington since 2013. To understand the scope, demographic composition, and outcomes associated with training programs for which current federal legislative proposals could make Pell Grants available, this analysis is primarily focused on short-term training programs that are fewer than 600 clock hours and 15 weeks in length.³ These short-term training programs account for 2,560, or 23.3 percent, of Washington’s training programs included in the data set. Where relevant, we highlight how the participant characteristics and outcomes of short-term training programs compare to longer training programs—those greater than 600 clock hours and between 15 and 52 weeks, one to two years, and two or more years in length.

The compiled data set included comprehensive information about each of the state’s training programs, including a unique program identification number (ID), the type of training provider, the program length, associated clock hours, and six-digit Classification of Instructional Programs (CIP) code. The data set also included aggregate-level data on the demographics (gender, race, and prior education level) of program participants and the outcomes of graduates for each training program, namely the completion rate, the employment rate one year after program completion, and the median hourly and annual earnings one year after program completion. Appendix A provides more detailed information about the key variables used in this analysis.

The analysis contained in this report is descriptive in nature, with much of the analysis consisting of sample mean comparisons. Using CIP codes, which classify instructional programs offered by postsecondary institutions by the occupational field for which program participants prepare, we were able to examine the 20 most common short-term training programs of study, which account for nearly 55 percent of short-term training programs in the data set (see Appendix B for list of common short-term programs of study and corresponding occupations). We also explore which providers tend to offer certain programs of study as well as the demographic profile of participants and the outcomes of graduates based on program of study.

Using aggregate data on the demographics of participants in each training program, we calculated the overall percentage of male and female participants; the percentage of participants who identify as white, Black, Hispanic/Latinx, Asian, Native American, and multiracial; and the percentage of participants whose highest level of education before enrolling in training was some college or less (includes high school diploma and GED holders and those without a high school diploma or GED), a certificate or associate degree, and a bachelor's degree. Using aggregate-level data on the outcomes of graduates for each training program, we were able to calculate the average completion rate across all short-term training programs as well as the average employment rate and median annual earnings one year after completion for each training program.⁴

Limitations

Because participant demographic data was only available in the aggregate, it was not possible to do cross-tabulations between any two demographic groups (e.g., race/ethnicity and gender). Furthermore, because the data set lacked demographic information about program graduates specifically, it was not possible to assess how the demographic profile of graduates differs from that of those who initially enrolled in the program.

For some programs, the data set lacked complete participant demographic and completer outcome information. The WTECB, which requires that training providers collect and submit participant data on an annual basis, suppresses participant demographic data in Career Bridge if a training program does not enroll at least 10 participants. The WTECB also suppresses the employment and earnings information for training programs that have fewer than 25 completers. Some of the missing outcome data can be attributed to lags in the collection of employment and earnings information for recently completed programs, as the employment and earnings of completers is measured one year after completion. Furthermore, in Washington, much like in other states, UI records exclude self-employed individuals, federal workers, military personnel, and foreign-employed workers, and therefore these groups are underrepresented in the employment and earnings data. The exclusion of these workers from employment and earnings records is more likely to skew the outcome data for entrepreneurship/entrepreneurial studies programs, which have a sizable number of self-employed graduates; or marine science/merchant marine officer programs, many graduates of which go on to contract, federal, or foreign employment.

Given WTECB publication cutoffs and reporting lags, the data set did not include participant and/or completer outcome data for all programs within a common short-term training program of study category. We excluded a common short-term program of study from the in-depth analysis if the data set lacked participant or completer outcome data for at least 10 programs. In all, the data set included sufficient program data for 14 of the 20 common short-term programs of study (the availability of program data on marine science/merchant marine officer; carpentry/carpenter; data entry/microcomputer applications; computer programming/programmer; plumbing technology/plumber; and electrical, electronic, and communications engineering technology/technician programs of study varied across participant or completer outcome indicators and thus are sometimes excluded from certain in-depth analyses). Appendix C includes a list of common short-term programs of study for which sufficient program demographic and outcome data exist. The overall analysis of short-term training programs includes all programs of study, regardless of the number of programs for which participant or outcome data were available.

In the absence of student-level Standard Occupational Classification (SOC) data indicating the specific occupations in which individuals are employed, it was not possible to determine the likelihood of graduates finding employment related to their program of study. Graduates employed in occupations misaligned with their program of study could skew—up or down—the average median earnings associated with a program of study.