The most consequential feature in American hiring may be a tiny checkbox. Bachelor’s degree required. Click it, and an applicant tracking system can make millions of capable people disappear before a manager sees a résumé. Opportunity@Work, a Washington nonprofit, has spent a decade arguing that this is not a neutral shortcut. It is a design choice - and frequently a costly one.
The organization calls those filtered workers STARs, short for Skilled Through Alternative Routes. They learned through military service, community college, certificate programs, apprenticeships, partial college, self-study, or the oldest classroom in capitalism: doing the job. More than 70 million U.S. workers fit the definition, roughly half the workforce. “People without degrees” tells you what is absent. STARs tells you what is present. That semantic flip may be Opportunity@Work’s neatest product.
Act IThe first theory met an employer-shaped wall
Opportunity@Work began in 2015 with TechHire, an Obama White House initiative conceived by economist and former McKinsey partner Byron Auguste. The pitch was practical: employers needed technology talent; accelerated training programs could prepare people overlooked by conventional recruiting; cities could connect the two. The effort grew to more than 70 communities. Opportunity@Work was incubated at New America, became an independent nonprofit in 2016, and took over the network.
The concept was not foolish. Thousands trained, and early participants connected to jobs earning above local median wages. But the first independent evaluation of selected federal TechHire programs found the plumbing unfinished. Three programs struggled to recruit enough participants. Career awareness, job development, and placement services were not fully integrated. Most tellingly, employer engagement fell short, especially around work-based learning and jobs for graduates. Early follow-up found no detectable labor-market effect.
The original pipeline - and the loose joint
That is the answer to “what failed first?” Not the workers. The handoff. Training supply was easier to assemble than durable employer demand. A person could complete a program and still collide with the same degree screen. The lesson changed the scale of the problem: worker preparation mattered, but labor-market infrastructure decided whose preparation counted.
“If employers don’t have a STARs talent strategy, they only have half a talent strategy.”
Act IIFrom job pipeline to market repair shop
Opportunity@Work’s 2020 research supplied the new map. Reach for the STARs analyzed workers’ occupations and skills, arguing that millions had the capability to move into jobs paying at least 50 percent more. The nonprofit acquired Skillist, a skills-first hiring startup, that August for an undisclosed price. It increasingly stopped behaving like another training provider and started acting like a repair crew for the rules, data, software, and stories that govern hiring.
Today its customers are not individual job seekers in the usual commercial sense. They are employers, state and local governments, regional economic-development groups, HR platforms, philanthropies, and workforce organizations. STARs are the beneficiaries and, through an advisory council and published stories, contributors to the work. The product bundle is unusual: labor-market research; STARSight regional data; employer guidance; public-sector implementation; technology partnerships; coalition management; and selective grantmaking.
STARSight lets public and regional leaders examine talent pools, occupations, skills, and mobility routes through a STAR lens. Partnerships with Indeed, Jobcase, and Lightcast push that logic into systems with much larger reach. With the Ad Council, Opportunity@Work launched Tear the Paper Ceiling in 2022 alongside nearly 50 organizations. In its first year, the campaign reported 3.3 billion impressions and nearly 1,300 media mentions. A good phrase traveled farther than a workforce-policy memo ever could.
The moneyWhat it costs, and who pays
This is a nonprofit business model, so the invoice usually goes to philanthropy rather than to a worker seeking a better job. Opportunity@Work reported roughly $30.4 million in fiscal 2024 revenue on its Form 990 and about $20.7 million in expenses. That year, MacKenzie Scott’s Yield Giving announced a $20 million unrestricted grant. The organization also turned funder: its $2 million Powering STAR Breakthroughs program backed eight training and talent organizations for 18 months.
Unrestricted money matters because systems change is a poor fit for tidy project boxes. Research must become a usable tool; a campaign must become changed manager behavior; policy must survive implementation. Still, grant-funded scale has a condition attached even when the grant does not: employers and platforms must adopt the work. Opportunity@Work cannot order a company to change an algorithm or make a supervisor trust a new assessment.
The honest scorecard: the 2026 report says employers in its network opened almost 600,000 jobs to STARs and that 90,000 STARs had experienced upward mobility since 2022. Those are meaningful network measures, not proof that one nonprofit caused every outcome across a 70-million-person population.
The differenceStandards stay. Lazy proxies go.
Opportunity@Work sits between advocacy groups and workforce vendors. Year Up United, Merit America, and local providers train or place people. Lightcast sells labor-market intelligence. LinkedIn, Workday, and Indeed operate hiring infrastructure. Jobs for the Future and the Burning Glass Institute publish research and organize employers. Opportunity@Work borrows a little from each category, then concentrates on one market failure: skills exist, but the signals and pathways do not reliably transmit them.
Its argument is not that degrees are worthless, or that every role should accept every applicant. It is that a degree should be required when the work genuinely needs one, not because an inherited job description had the line in 2007. Skills-first hiring replaces a rough proxy with explicit evidence: relevant tasks, assessments, prior performance, portfolios, certifications, and structured interviews. The standard can become clearer and tougher while the pool becomes wider.
The catchWhen the playbook does not work
Skills-first hiring fails when it is merely cosmetic. Removing “B.A. required” does little if recruiters still search by college, algorithms learn from degree-heavy historical hires, managers equate pedigree with polish, or promotion ladders quietly restore the same screen. It also cannot replace legally required credentials, deep occupational training, or the wraparound support some workers need to complete demanding programs. The early TechHire evidence was blunt on that last point: participants entering with lower baseline skills often needed tutoring or preparatory bridges.
It works best where employers face persistent talent demand, can specify job-relevant skills, have credible ways to assess them, and are willing to change several steps at once. It weakens in a slack labor market, where employers can afford arbitrary filters, or in organizations without a senior owner, operational data, and manager incentives. A coalition can offer the map. Someone inside the company still has to move the furniture.
Inside the shopA mission-driven startup with startup seams
Opportunity@Work describes its culture through five unusually specific values: Better Together, Intentionally Inclusive, Clarity to Inform Action, Always Learning, and Champion Empowerment. Its careers page promises flexible scheduling and work-life balance while asking staff to embrace ambiguity, healthy conflict, and responsibility when outcomes miss the plan. That is a fair description of systems-change work, where the customer may be an employer, the partner a government, and the person who must benefit is a worker neither one can neatly command.
The outside view is less polished. Anonymous workplace reviews praise the mission, benefits, colleagues, and flexibility, while some former employees describe shifting priorities, weak leadership, and painful layoffs in 2024. Anonymous reviews are anecdotes, not an audit, but the tension is worth noticing. An organization asking employers to value STARs is judged by whether its own people experience the dignity and mobility in its vocabulary. Culture, like skills-first hiring, becomes real in operations rather than slogans.
Act IIIAI arrives with a pencil and an eraser
The next fight is not only over degree filters. It is over whether AI learns yesterday’s exclusions or exposes better evidence of skill. Opportunity@Work’s 2026 research with Brookings found more than 15 million STARs in jobs highly exposed to AI, including nearly 11 million in “gateway” occupations that historically lead to higher wages. Customer service, administrative, and clerical roles are both career footholds and prime automation targets.
In July, the partners launched a six-month AI Readiness Lab with business-led groups in Birmingham, central New York, Indiana, Cleveland, Minneapolis-Saint Paul, and San Diego. Each region gets localized analysis, peer learning, and a playbook. The premise is wonderfully unmagical: national AI forecasts are interesting, but people get hired in local economies. Nearly three-quarters of American workers live and work in the same county.
Opportunity@Work’s goal is now audaciously measurable: by 2035, enable upward mobility for one million STARs, open ten million good-paying jobs, and increase earnings by $100 billion. Those numbers are targets, not trophies. The organization’s useful contribution may be making the labor market’s defaults visible enough to argue with. A degree filter feels inevitable until somebody calls it a filter.
The company’s transferable insight is simple: do not train around a broken gate. Fix the gate, too.