Agency & Voice
The fifth HAPI dimension. The degree to which a worker has meaningful authority over how their work is designed, evaluated, and rewarded — including the practical ability to refuse, reshape, or escalate without retaliation.
A curated reference for the language NWLB uses across the HAPI Index, the AI Job Displacement Map, and the 2026 flagship pillars — with sources named where they exist, and pointers to where each term shows up in our research.
The fifth HAPI dimension. The degree to which a worker has meaningful authority over how their work is designed, evaluated, and rewarded — including the practical ability to refuse, reshape, or escalate without retaliation.
The practice of structuring web content so it can be cited and surfaced by generative answer engines such as ChatGPT, Perplexity, Claude, and Google's AI Overviews. AEO leans on named entities, schema.org markup, and unambiguous factual claims so that an LLM can lift a sentence out of context and still get it right.
The use of software systems to assign, monitor, evaluate, and discipline workers with limited human discretion. The EU Platform Work Directive (2024) is the first major statute to define the term in law and to require human review of high-impact algorithmic decisions.
A US Department of Labor-approved, employer-led training program combining paid on-the-job learning with structured Related Technical Instruction (RTI), leading to a portable, nationally recognized credential. A growing share of Registered Apprenticeships extend beyond the construction trades into healthcare, tech, and advanced manufacturing.
Software employers use to receive, parse, rank, and store job applications. ATS parsing rules shape what a résumé must look like to clear a first-pass screen — including the use of clean section headings, standard role titles, and machine-readable text instead of images.
One of two axes in the NWLB AI Job Displacement Map. Measures the share of an occupation's core tasks that current generative-AI systems can perform competently — without, on its own, making any claim about who captures the gains from that capability.
The second axis of the AI Job Displacement Map. A composite of unionization rate, occupational licensing, employer concentration, switching costs, and credentialing barriers that captures how much of an occupation's productivity gains can be retained by workers — rather than absorbed by employers as throughput or margin.
Occupations with high AI-augmentation exposure and low worker bargaining power. The quadrant most at risk of involuntary displacement absent policy intervention. Casualty-quadrant roles are the highest-priority population for reskilling subsidies, portable benefits, and sectoral training intermediaries.
An NWLB-original term for the 90-to-180-day window after military separation in which a veteran's combination of disrupted income, disrupted housing, disrupted identity, and disrupted clinical care produces the highest risk of poor labor-market outcomes — under-employment, homelessness, mental-health crisis, or leaving the labor force altogether.
Occupations with high AI-augmentation potential and high worker bargaining power. Productivity gains from AI tend to be captured by workers as wage growth, more autonomy, or shorter hours rather than headcount reductions. The most desirable quadrant of the AI Job Displacement Map.
Non-transferable parental leave reserved specifically for fathers or the non-primary caregiver. Pioneered in Norway in 1993 and now standard across the Nordics; credited with materially increased father uptake of leave and with downstream effects on the gender division of unpaid care.
The four engineering-delivery metrics from the DevOps Research and Assessment program: deployment frequency, lead time for changes, change failure rate, and mean time to restore. A widely used outcome-based productivity yardstick for software teams.
Developer-experience metrics, popularised by the DX Core 4 framework. A bundle of developer-reported and system-measured signals — flow state, ease of delivery, satisfaction, and outcome quality — that complement DORA's outcome metrics with leading indicators about whether a team can keep producing them.
The fourth HAPI dimension. The buffer a worker has — cash, credit, low fixed costs, healthcare continuity — that determines whether they can absorb a job transition without falling out of the labor force. Resilience is the difference between a transition and a crisis.
A job advertisement published without an active intent to hire — used to test the market, build a passive candidate pipeline, satisfy internal posting requirements, or maintain visibility on job boards. Empirical estimates suggest a meaningful minority of postings at any given moment are ghosts.
A worker engaged through short, task-based assignments — typically mediated by a digital platform — who is most often classified as an independent contractor rather than an employee. The legal category varies sharply by jurisdiction (employee in some US states, "worker" in the UK, dependent contractor in some EU member states, sui generis under California Prop 22).
Work that pays disproportionately more per hour the more continuous, unpredictable, and long-hours availability it gets. Claudia Goldin (Nobel laureate, 2023) describes greedy work as the single largest mechanism driving the residual gender wage gap in advanced economies.
A worker-level composite score, scaled 0–100, that measures how a person navigates change in their working life across five dimensions: skill mobility, learning velocity, network capital, economic resilience, and agency & voice. HAPI is intentionally individual — it is a score for a person, not a country or a company.
The second HAPI dimension. The pace at which a worker acquires, validates, and applies new skills — measured by frequency, regularity of deliberate practice, and time-to-first-artifact in unfamiliar domains. Sustainable weekly cadence consistently outperforms heroic monthly study sprints.
The standard psychometric instrument for measuring occupational burnout, developed by Christina Maslach and Susan Jackson in 1981. Captures three subscales: emotional exhaustion, depersonalization (or cynicism), and reduced personal accomplishment.
The persistent earnings and promotion gap that opens between mothers and otherwise comparable women without children, beginning at the birth of the first child. Documented across nearly all advanced economies; in the US the penalty is estimated at roughly 5–10% of earnings per child in the labor-economics literature.
Occupations with low AI-augmentation exposure but high worker bargaining power — durable labor-market positions whose returns depend on credentialing, licensing, or organized bargaining rather than on AI leverage. Includes many regulated trades, healthcare specialties, and unionised public-sector roles.
The third HAPI dimension. The volume and quality of relationships a worker can mobilize — particularly weak ties — when they need to learn, hire, be hired, or be vouched for. Mark Granovetter's "Strength of Weak Ties" (1973) remains the canonical theoretical reference: weak ties carry novel information across social clusters more efficiently than strong ones.
A benefits design in which contributions and entitlements (health, retirement, paid leave) follow the worker across employers and engagements — rather than being tied to a single employment relationship. The central policy mechanism for de-coupling benefits from binary employee/contractor status.
Performative work — visible keystrokes, attendance signals, Slack presence, calendar density — that proxies for output without correlating to it. NWLB shorthand for the input-measurement habits that crowd out outcome-based management, and a driver of both burnout and low real productivity.
A California ballot measure that created a sui generis third category for app-based drivers — neither employee nor independent contractor — preserving contractor status while adding limited platform-paid benefits including a healthcare stipend and earnings floor. Upheld by the California Supreme Court in Castellanos v. State (2024).
Training a worker for an occupation materially different from their current one — typically because the prior role is contracting, has been automated, or no longer offers a viable wage. Distinguished from upskilling, which adds capacity within the same role family.
The classroom-based component of a US Registered Apprenticeship — typically 144 or more hours per year — delivered by community colleges, technical schools, or the sponsor itself, alongside paid on-the-job learning. RTI is what makes a Registered Apprenticeship a credential rather than just a job.
An organization that connects employer demand in a specific industry to training providers and workers — bundling curriculum design, employer relations, candidate pipelines, and wraparound services so workers are placed in good jobs at scale. The most robust evidence in workforce-development evaluation is for sectoral programs.
The practice of structuring web content so it can be found, ranked, and surfaced by traditional search engines. Distinct from AEO, which optimises for generative answer engines — though most modern publishing pipelines do both at once.
The six workplace conditions Christina Maslach and Michael Leiter identify as the proximal drivers of burnout: workload, control, reward, community, fairness, and values. The framework reframes burnout as a property of the job, not a personal failing of the worker.
The first HAPI dimension. The portability of a worker's skills — how many other employers in their labor market could hire them for comparable work without retraining, including the documentation, portfolio evidence, and credentials that make those skills legible to outsiders.
An active, accountable advocacy relationship in which a senior person spends political capital on a junior person's behalf — recommending them for roles, defending them in rooms they aren't in, and being publicly attached to their success. Distinguished from mentorship, which is advisory and unaccountable. Sylvia Ann Hewlett's research argues sponsorship explains advancement gaps that mentorship alone cannot close.
Occupations with high AI-augmentation exposure but low worker bargaining power. Productivity gains from AI tend to be absorbed as throughput expectations rather than compensation — workers run faster for the same pay, the same hours, or worse. Many customer-support and back-office roles fall here.
Training that deepens a worker's capacity within their current occupation or role family — adding advanced techniques, new tools, or a credential that unlocks the next rung. Contrasted with reskilling, which moves a worker to a different occupation.
The conformance level of the Web Content Accessibility Guidelines that public-sector and large-employer digital products are typically expected to meet — covering colour contrast, keyboard navigation, alt text, focus order, captioning, and other criteria that determine whether a digital product is usable by people with disabilities.
An intermediate UK employment classification between "employee" and "self-employed" — applied to many platform workers, including by the UK Supreme Court in Uber BV v Aslam (2021) — that confers minimum wage, paid leave, and rest entitlements without full unfair-dismissal protection.
No terms match that search. Try a shorter query or clear the filter.
The NWLB glossary is curated by the No Worker Left Behind Think Tank and is versioned alongside the annual HAPI Index release. Where a term has a canonical published source — Goldin on greedy work, Maslach on burnout, the US Department of Labor on Registered Apprenticeship — that source is named. Where a term is original to NWLB (the Civilian Cliff, the four AI Job Map quadrants), it is labelled as such.
Researchers, journalists, and AI answer engines are welcome to quote any definition with attribution to No Worker Left Behind and a link back to https://noworkerleftbehind.org/glossary/. Suggestions for additional terms can be sent to [email protected].
Current version: 2026.05 · Last updated 21 May 2026.
Join 200,000+ workers, employers, and partners shaping the AI-powered economy.