How and Why Do Technology Apprenticeships Matter: Mixed-Methods Evidence from the Launchcode Program and Equifax

Background. Despite an increase in employer-aligned certificate and apprenticeship programs, there is limited research examining the impact of these programs on economic outcomes. Moreover, for much of the research that has explored the impact of these programs, it is unclear whether the outcomes are a product of the courses alone or the apprenticeships and other work-related experiences that many of these programs provide. While our previous research has leveraged survey data to understand the economic outcomes associated with these programs, administrative records on employment and earnings allow for more robust inferences to be made. We conduct a mixed-methods study on LaunchCode, leveraging administrative records from Equifax and in-depth interviews with program instructors and participants. LaunchCode is a technology certificate and apprenticeship program that it is free, includes a paid apprenticeship, and serves a larger share of women and Black individuals than traditional technology programs.

Data and Methods. We first merge application data containing entrance exam scores with administrative employment and earnings data from Equifax for (a) individuals that were not accepted into the program, (b) individuals that were accepted but did not complete the course, (c) individuals that completed the course but not the apprenticeship, and (d) individuals that completed both the course and the apprenticeship (n=2,141). First, by using entrance exam scores as an instrumental variable, we conduct an intent-to-treat analysis (ITT) to understand the impact of program acceptance on earnings and STEM employment. Second, by using machine-learning generated multinomial propensity score weights, we conduct a treatment-on-treated (TOT) analysis to understand heterogeneous treatment effects across different levels of program participation. Finally, we conduct in-depth interviews with program instructors and participants (N=24) to better understand how LauchCode program components might explain these findings.

Preliminary Findings. We find that the majority of the impacts on earnings are driven by the apprenticeship component, which confirms one of the main findings from our survey analysis. However, the impact on income is considerably lagged for apprenticeship completers; increased earnings trajectories are most pronounced three to four years after completion, which has implications for how we understand economic impacts in programs that include apprenticeships. Moreover, we find that the majority of these impacts are driven by lower-income persons and Women, which has implications for social mobility and gender equity in STEM. Finally, we find that key supports in the program, such as support in developing resumes and preparing for interviews helped build students’ confidence, especially for those experiencing a career change.

Significance. At the local level, our findings can be used to incentivize businesses to offer apprenticeships and partner with local education institutions to create new training-to-employment pipelines. At the state level, our findings may cause stakeholders to consider supporting more holistic workforce development programs that consist of both education and apprenticeship components. Finally, at the federal level, our findings lend support for the use of federal funds for alternative education programs and apprenticeships, like those found in LaunchCode.