Citation: Hecker, I., & Loprest, P. (2019). Foundational Digital Skills for Career Progress. Urban Institute, Income and Benefits Policy Center. August 2019.
URL: https://www.urban.org/sites/default/files/publication/100843/foundational_digital_skills_for_career_progress_2.pdf
Pulled by: Claude on 2026-04-29
PDF: urban-institute-2019-foundational-digital-skills.pdf in this folder.
What this is
A short Urban Institute brief synthesizing research on the demand for foundational (non-specialized) digital skills, the share of US workers who lack them, and the strategies and challenges of teaching them. Funded as part of Urban's collaboration with JPMorgan Chase's $250M New Skills at Work initiative. Includes interviews with five providers (urban library, community college, city workforce board, adult basic education program, senior community service employment program).
Definition of "foundational digital skills" (load-bearing for our pitch)
The brief explicitly distinguishes:
- Specialized digital skills — programmer, software engineer, IT support.
- Foundational digital skills — non-specialized digital skills used in many roles, ranging from basic (turn on a computer, use a mouse, access the internet) to digital literacy (combining base knowledge and problem-solving to approach new platforms and uses).
"A higher level of foundational digital skills involves using these tools to carry out specific digital tasks... A higher level... is being able to take the knowledge of how to accomplish specific digital tasks and applying it to new circumstances, contexts, or platforms... We refer to this as being 'digitally literate,' being able to combine base knowledge and problem solving to approach new platforms and uses." (p. 2)
This framing — digital literacy = transfer to new platforms — directly aligns with our pedagogy thesis (schemas, not procedures).
Key findings useful for the pitch
- 78% of middle-skill jobs (those typically requiring a bachelor's degree and paying a living wage) require baseline digital skills (spreadsheets, word processing).
- Brookings 2002→2016: share of jobs with high digital content quadrupled (4.8% → 23.0%); low-digital share fell 55.7% → 29.5%.
- PIAAC 2011–12 US data (the source the pitch is using):
- 16% of 16–65 year-olds — ~32M Americans — have no digital skills (failed core screener / no computer use).
- 51% scored Level 1 or below.
- 33% scored Level 2 or above.
- Age stratification (from Mamedova et al. 2018, NCES): "no digital skills" rate by age — 25–34: 11%; 55–65: 28%.
- By education: less than high school: 41% no digital skills; associate's or higher: 5%.
- By origin: born outside US: 36%; born in US: 13%.
The transfer-failure finding (the load-bearing quote for our pedagogy doc)
"Another digital skill provider noted that fluid use of a smartphone does not always translate to broader digital skills. Some young people who were experts with smartphones were not able to easily transfer their knowledge into a work setting where they needed to use computers and office software and tasks." (p. 12)
"A key point in the literature is that moving from knowledge of specific skills and tasks to broader digital literacy involves confidence, familiarity, and interest. Confidence likely affects a person's ability to translate learned skills from task to task or occupation to occupation. These concepts are broadly applicable to people learning anything new (perhaps especially to adult learners)." (p. 3)
These two quotes are the empirical anchor for the mental-models-not-just-UI-patterns design constraint in our spec.
Provider-reported strategies (useful for technical-approach.md and spec)
From interviews with five providers — these are field-tested heuristics for adult digital-literacy training:
- Teach skills in context. "Learning digital skills worked better for participants when taught in the context where they would be used... learning how to use a tool or do a specific digital task is easier if done as part of a real work task." (p. 12) — Supports our task-based, real-world-scenario pedagogy.
- In-person teaching matters. Self-guided online modules (Google Digital Training, Northstar) work for some, but providers emphasized in-person support, especially for low-skill learners. "One provider spoke of 'humanizing' the experience by having staff available." (p. 13) — Implication: an AI co-pilot may need to play this "humanizing" role explicitly, not just be a hint dispenser.
- Match training to needs. Providers offered tiered programs because skill levels varied widely. — Supports adaptive sequencing.
- Access to tools is part of the intervention. One program loaned laptops; another loaned Wi-Fi hotspots. — Outside our scope but relevant for distribution.
Provider-identified challenges
- Choosing trainings/assessments: providers had limited info on efficacy and population fit. Our project could fill this gap.
- Funding: persistent issue.
- Non-English speakers: dual needs (language + digital). One provider taught digital skills in native language first; another deferred digital training until English proficiency.
- Moving from initial familiarity to fluency: "Multiple respondents suggested it is not clear how to train people to move from this initial level to more fluency." (p. 14) — This is exactly the gap our project targets.
Research gaps the brief flags (= opportunity for our project)
"Providers seeking to use existing digital skill assessments and trainings need more information on their effectiveness and whether employers value associated credentials. Lastly, providers need to know more about what kinds of training are most effective for teaching foundational digital skills, including basic and discrete tasks and the ability to transfer this knowledge across jobs and platforms, solve problems, and adapt to new technological settings." (p. 15, emphasis added)
This is the literature explicitly asking for what we propose to build. Worth quoting directly in the pitch.
How we'll use this in the docs
- Pitch — Problem section: confirms the PIAAC age-stratified figures already in the doc.
- Pedagogy doc — transfer / mental-models section: the smartphone-to-office quote is the sharpest design constraint; cite directly.
- Pedagogy doc — falsification section: the "moving from initial familiarity to fluency" gap defines what our metrics must show.
- Pitch — Why fund this: cite the closing research-gap quote as evidence that the field itself has identified this as the missing intervention.