TL;DR. A credential is a signal: a cheap way for a stranger to guess what you can do. In 2026 the signals are not equal, and they are not what the headlines suggest. Skills-based hiring is real in employer surveys and thin in actual hiring. Firms that removed degree requirements raised the share of non-graduates hired into those roles by about 3.5 percentage points, which across the economy came to roughly 97,000 extra workers a year out of 77 million hires, “not even 1 in 700” (Burning Glass Institute and Harvard Business School, 2024). Degree requirements in US job postings rose again, from 16.6 percent in November 2023 to 19.3 percent in November 2025 (Indeed Hiring Lab). When over 1,000 employers were asked how they will assess skills from 2025 to 2030, 81 percent named work experience, 48 percent pre-employment tests, 43 percent a university degree and only 14 percent short courses and online certificates (World Economic Forum, 2025). Online certificates still matter, but for a specific group: in a randomised trial with more than 800,000 learners, making certificates visible on LinkedIn raised new employment by 6 percent, with gains of 12 percent for the third of learners with the weakest predicted employability and negligible effects for the strongest third (Athey and Palikot, 2025). The CEO move is to treat credentials as a portfolio of signals, each bought for a specific buyer. The student move is to collect the one signal no course can give you: proof of real work that someone else can check.
This article is part of the skill strategy series. If you have not yet decided which skill to signal, start with the skill audit and the 4-factor decision matrix for what to learn next. This piece answers the next question: once you have the skill, what will convince someone else?
Why do credentials exist at all?
Hiring is a decision made under uncertainty. A recruiter cannot watch you work, so they rely on cheap proxies. The Burning Glass Institute put it bluntly in its 2022 study of 51 million job postings: degree requirements “seemed to function instead as a proxy used to simplify the hiring process, a shorthand for certain skills and a screen to reduce the number of candidates whom recruiters needed to evaluate”.
That framing explains most of the evidence below. A signal is valuable when it reduces the employer’s uncertainty more cheaply than the alternatives. A study of skill certificates on a large online freelancing platform found exactly that mechanism: the earnings effect “is not driven by increased freelancer productivity but by decreased employer uncertainty” (Kassi and Lehdonvirta, 2019). Credentials do not make you better at the job. They make you easier to believe.
The supply of signals has exploded. Credential Engine counted 1,850,034 unique credentials in the United States in 2025, including 264,099 degrees, 486,352 certificates, 6,892 certifications, 14,331 occupational licences and 1,022,028 badges. When there are more than a million badges, any single badge says less.
What do degrees still buy in 2026?
Earnings and unemployment still rise with education. The US Bureau of Labor Statistics’ 2025 figures for people aged 25 and over show median weekly earnings and unemployment rates by attainment. The 2025 figures are 11-month averages because October was not collected.
| Highest education (age 25+, 2025) | Median weekly earnings | Unemployment rate | Earnings vs high school diploma |
|---|---|---|---|
| Less than a high school diploma | $770 | 6.1% | -20.3% |
| High school diploma | $966 | 4.3% | baseline |
| Some college, no degree | $1,062 | 3.8% | +9.9% |
| Associate’s degree | $1,135 | 3.0% | +17.5% |
| Bachelor’s degree | $1,578 | 2.8% | +63.4% |
| Master’s degree | $1,876 | 2.6% | +94.2% |
| Professional degree | $2,294 | 1.9% | +137.5% |
| Doctoral degree | $2,307 | 1.8% | +138.8% |
Source: BLS, Education pays, 2025. The last column is a CEOtudent calculation. These are averages across people who differ in many ways besides education; they are not the causal return to a degree.
The degree screen is not disappearing. Indeed Hiring Lab reports that 19.3 percent of US postings required a bachelor’s degree or higher in November 2025, “up from 16.6% in November 2023”. After holding the mix of job titles constant, the share “has been trending upwards since March 2024”. At the same time, 51 percent of postings ask for no formal education at all. The market is split: degree-gated roles stay gated, and many other roles never were.
Skills-based hiring is mostly an announcement. The 2024 Burning Glass and Harvard Business School study of 11,300 roles at large firms found an almost fourfold rise from 2014 to 2023 in roles where employers dropped degree requirements. But about 45 percent of those firms made the change “in name only”, with no meaningful difference in actual hiring. The real change was driven by 37 percent of firms. Where it happened, it worked: at firms that followed through, non-degreed workers had a retention rate 10 percentage points higher than graduates, and those hired into formerly degree-gated roles saw a 25 percent salary increase on average.
A degree is not a guarantee either. The Burning Glass Institute and Strada’s 2024 “Talent Disrupted” report found that “52 percent of graduates are underemployed a year after graduation” and 45 percent a decade later, meaning they work in jobs that do not require a degree. Underemployment is sticky: graduates who start in a college-level job mostly stay there, and those who start underemployed mostly stay underemployed. The Federal Reserve Bank of New York’s second-quarter 2026 update put recent-graduate unemployment at about 5.6 percent and underemployment at 42 percent, using a narrower definition.
Do certifications and licences pay?
The BLS separates two things people often confuse. “Certifications are issued by a non-governmental certification body”, while “a license is awarded by a government agency and conveys a legal authority to work in an occupation”. In 2025, 24.0 percent of employed people held one or the other, but only 2.5 percent held a certification without a licence.
The earnings pattern is the most useful original comparison in the official data. Below, each row compares full-time workers at the same education level.
| Education level (2025) | Without certification or licence | With certification or licence | Premium | Certification, no licence | Premium |
|---|---|---|---|---|---|
| High school, no college | $943 | $1,129 | +19.7% | $1,195 | +26.7% |
| Associate degree | $1,089 | $1,229 | +12.9% | $1,295 | +18.9% |
| Bachelor’s degree only | $1,588 | $1,558 | -1.9% | $1,876 | +18.1% |
| Advanced degree | $1,924 | $1,910 | -0.7% | $2,261 | +17.5% |
Original analysis by CEOtudent editorial framework from BLS Current Population Survey Table 54, 2025 (median weekly earnings, full-time wage and salary workers). Premiums are CEOtudent calculations. These are cross-sectional comparisons, not causal effects: people who earn certifications differ in occupation and motivation.
Two patterns stand out. For workers without a degree, any certification or licence goes with noticeably higher pay. For graduates, a licence on its own goes with no premium at all, while a non-government certification goes with roughly 17 to 18 percent more. A plausible explanation is occupation mix: licences cluster in fields such as healthcare, where 71.3 percent of practitioners hold a licence, while stand-alone certifications are more common in fields such as computer and mathematical occupations, where 6.9 percent hold a certification without a licence against 2.5 percent of all workers. That is still a useful signal for planning. A certification that a profession recognises tends to sit in better-paid work.
Experimental evidence points the same way. In a resume audit of 10,492 fictitious applications, Deming and colleagues (2016) found no significant difference in callbacks by type of institution “for health jobs (such as practical nursing and pharmacy technician) that require both a certificate and a valid occupational license”. When a licence verifies competence, employers stop caring where you studied.
Do online certificates and bootcamps work?
The evidence here is newer and more nuanced than either the marketing or the dismissal.
Where you studied can count against you. In the same audit study, business degrees from large online for-profit institutions were “about 22 percent (2 percentage points) less likely to receive a callback” than similar degrees from non-selective public schools. Associate degrees from public or for-profit institutions produced no more callbacks than identical work experience with no degree.
Visible certificates help people with thin profiles. Athey and Palikot (2025) ran a randomised experiment with more than 800,000 online certificate earners in developing countries who lacked college degrees, nudging them to share their credentials on LinkedIn. New employment rose by 6 percent. Gains were 12 percent for the bottom third by predicted employability and negligible for the top third. In a separate resume-scoring exercise, credentials improved perceived candidate quality “by 8.5 points on average (on a scale from 1 to 100), with effects of 15-20 points for weak resumes but near-zero for strong resumes”. Most of the certificates in the study were short: individual courses made up 82 percent.
Platform certificates work the same way. On a large online freelancing platform, an additional skill certificate led to “a 9.7% increase” in project value, but “experienced workers who have already accumulated a significant work history on the platform do not benefit” (Kassi and Lehdonvirta, 2019).
Retention, not raises. A European Commission study using a differences-in-differences design found that MOOC participation “can improve workers’ employment retaining but not their wages” (Castano-Munoz and Rodrigues, 2021).
Bootcamps can redirect, at a short-term cost. The one randomised trial we found, a women-only coding bootcamp in Buenos Aires and Bogota studied by the World Bank (Aramburu, Goicoechea and Mobarak, 2021), increased “the probability of working at a technology job by 9.2 percentage points” but decreased the probability of being employed at all by 9.8 percentage points in the short term, during a follow-up that overlapped COVID-19. Commercial surveys report better numbers: a Gallup study conducted in partnership with the bootcamp operator 2U found median income growth of 17 percent among 3,824 graduates, with no comparison group.
Company-reported outcomes are not evidence of effect. Google reports one million Career Certificate graduates and says “more than 70% of certificate graduates report a positive career outcome”; Coursera reports that 91 percent of surveyed learners “achieved a positive career outcome”. Both are self-reported surveys of people who chose to respond, with no comparison group. They describe satisfaction, not the value of the signal.
What about work experience and portfolios?
Experience is the signal employers name first. In the World Economic Forum’s Future of Jobs survey, evaluation of work experience (81 percent) leads every other assessment method employers plan to use through 2030. The National Association of Colleges and Employers’ Job Outlook 2026 found that “nearly all respondents cited U.S.-based internships as valuable”. Talent Disrupted found that “the odds of underemployment for graduates who had at least one internship are 48.5 percent lower than those who had no internships”.
The first piece of verified work is disproportionately valuable. In a field experiment on the freelancing platform oDesk, Pallais (2014) hired 952 randomly selected inexperienced workers and gave them either detailed or coarse public evaluations. Compared with a control group that was not hired, being hired and receiving even a coarse public evaluation “almost tripled the fraction of these workers with any employment from 12% to 30% and the average worker’s earnings from $10 to $27”. The signal was not a credential. It was a short, verified record of real work.
Portfolios are the least studied signal. We could not find a credible published field experiment on portfolios, GitHub profiles or work samples as hiring signals. One audit study of candidates skilled through alternative routes was registered in 2023, and its design “will be made public when the trial is complete”. The closest measured relative is the pre-employment test, which 48 percent of WEF employers plan to use. The honest conclusion: portfolios are plausibly strong when an employer can check them quickly, but the evidence does not yet exist to rank them.
Employers are moving away from proxies like grades. NACE reports that 73 percent of employers screened candidates by GPA in 2019 and just 42 percent do so in its 2026 survey, while 70 percent now report using skills-based hiring, mostly at the interview (87 percent) and screening (65 percent) stages. This is a small, non-random survey of 183 employers, but the direction matches the WEF data.
The Signal Strength Scorecard
The table below summarises the evidence for each signal. The ratings in the last column are an editorial judgement, not a measured score.
| Signal | Best evidence found | What the evidence shows | Who gains most | Evidence quality | Editorial signal rating |
|---|---|---|---|---|---|
| Bachelor’s degree | BLS earnings; Indeed postings; Burning Glass hiring data | Still a dominant screen; 19.3% of postings require it; skills-based hiring changed under 1 in 700 hires | Anyone targeting degree-gated roles | Strong descriptive, weak causal | Strong as a gate, weak as proof of a specific skill |
| Graduate degree | BLS earnings | $1,876 to $2,307 a week vs $1,578 for a bachelor’s | Fields with graduate gates | Descriptive only | Strong where required, costly elsewhere |
| Occupational licence | Deming and colleagues 2016; BLS Table 53 | Equalises callbacks across schools in licensed health jobs | Regulated occupations | Experimental (callbacks) | Very strong where legally required, irrelevant elsewhere |
| Professional certification (no licence) | BLS Table 54 | 17.5% to 26.7% higher pay at every education level | Technical and professional roles | Descriptive only | Moderate to strong if the profession recognises it |
| Online course certificate | Athey and Palikot 2025 RCT; Kassi and Lehdonvirta; Castano-Munoz | +6% new employment; 15-20 resume points for weak profiles, near-zero for strong | Early-career, no degree, thin resume | Experimental | Moderate for thin profiles, near-zero for strong ones |
| Bootcamp | World Bank RCT 2021 | +9.2 points tech job, -9.8 points any job, short term | Career switchers who can absorb a gap | One small RCT | Mixed, depends on the program |
| Work experience and internships | Pallais 2014; Talent Disrupted; WEF 2025 | First verified job nearly tripled employment; internship cuts underemployment odds by 48.5% | Everyone, most of all newcomers | Experimental and large-scale observational | Strongest signal in the evidence |
| Portfolio and work samples | No completed field experiment found | Unmeasured; closest relative is the skills test (48% of employers) | People whose work is visible and checkable | Gap | Plausibly strong, unproven |
Signal Strength Scorecard: CEOtudent editorial framework built from the sources listed below.
How should you build your signal stack?
The evidence suggests a simple rule: the weaker your current record, the more a cheap, verifiable signal is worth, and the stronger your record, the more only real work counts. The plan below is a CEOtudent editorial framework, not a tested intervention.
If you have no degree and little experience. Your first goal is any verified work. A short online certificate is worth showing; the trial evidence says it helps exactly this group. Pair it with one small, real project for a real client or employer and get it reviewed in public. In the oDesk experiment, the first verified job did more than any credential.
If you are a graduate early in your career. The degree has already done its screening work. The risk is underemployment, and the strongest protection in the data is an internship or equivalent work experience in the target field. Add one professional certification that your target employers name in their postings; at bachelor’s level, certification without a licence goes with about 18 percent higher pay in the BLS data.
If you are experienced and switching fields. Certificates add almost nothing to strong profiles, so do not expect them to carry the switch. Lead with a portfolio of two or three work samples in the new field that a hiring manager can check in minutes, and use a recognised certification only where the new field requires it. Our guide to reskilling after 40 covers the learning side.
If you are in or entering a licensed profession. The licence is the signal. Everything else is secondary.
The 6-question signal test
Before you pay for any credential, answer these:
- Do employers in my target role name this credential in their job postings?
- Is it issued by a body the profession recognises, or by the seller of the course?
- Can an employer verify it in under a minute?
- Does my current record already signal this skill? If yes, the credential adds little.
- Would the same time spent on one real, reviewable piece of work signal more?
- Will this credential still mean something if AI tools make the underlying task easier?
Test: CEOtudent editorial framework.
How is AI changing the value of signals?
Two trends make proof of real work more valuable. First, entry-level pathways are under pressure. The Stanford AI Index 2026 reports that “employment among software developers aged 22-25 has plummeted nearly 20% since 2024, even as their older colleagues’ headcount grows”, and the Burning Glass Institute found unemployment for 20-24 year olds with at least a bachelor’s degree “increased from 5.2% in 2018-19 to 6.2% in the two years through June 2025”. When the first job is harder to get, the signals that substitute for it matter more.
Second, AI makes polished applications cheap, and screening software decides what gets seen. In a 2022 SHRM survey, 45 percent of HR professionals said their organisation used automated resume prescreening, but only 32 percent of those said the system recognised alternative credentials. A certificate an applicant tracking system cannot parse is a signal no one receives. For the screening side of that problem, see resume writing when AI reads first. For which skills employers are paying for right now, see the most in-demand skills of 2026.
The common thread: as text becomes cheap, evidence of work that someone has checked becomes the scarce signal.
Frequently asked questions
Is a degree still worth it in 2026?
In the US data, bachelor’s holders earn a median 63 percent more per week than high school graduates and have lower unemployment, and degree requirements in postings rose between 2023 and 2025. Those are averages, not guarantees: about half of graduates start in jobs that do not need a degree. The degree is worth most when paired with relevant work experience.
Are Google or Coursera certificates worth putting on a resume?
Yes, if your resume is thin. The best trial evidence shows a 6 percent rise in new employment from making certificates visible, concentrated among learners with weaker job prospects. For those with the strongest prospects the effect was close to zero.
Do employers really practise skills-based hiring?
They say so more often than they do it. NACE found 70 percent of surveyed employers report using skills-based hiring, but the largest hiring-data study found the policy changed fewer than 1 in 700 hires. The practice is real at a minority of firms.
Are portfolios better than certificates?
Probably, for experienced people and in fields where work is visible, but no completed field experiment has measured it. What the evidence does show is that verified real work, such as a first job with public feedback or an internship, outperforms every credential studied.
Should I do a bootcamp?
The only randomised trial found it moved graduates into tech jobs but lowered overall employment in the short term. Treat it as a career switch with a transition cost, and check whether the program publishes complete, audited outcomes.
Sources
- U.S. Bureau of Labor Statistics. Employment Projections, Table 5.1: Unemployment rates and earnings by educational attainment, 2025. Last modified 1 September 2026.
- U.S. Bureau of Labor Statistics. Current Population Survey annual averages 2025, Tables 52, 53 and 54: Certification and licensing status of the employed.
- Federal Reserve Bank of New York. The Labor Market for Recent College Graduates, 2026 Q2 update.
- Sigelman M, Fuller J, Martin A. Skills-Based Hiring: The Long Road from Pronouncements to Practice. Burning Glass Institute and Harvard Business School Project on Managing the Future of Work, February 2024.
- Burning Glass Institute and Harvard Business School. The Emerging Degree Reset. February 2022.
- Burning Glass Institute and Strada Institute for the Future of Work. Talent Disrupted: Underemployment, College Graduates, and the Way Forward. February 2024.
- Levanon G, Sigelman M, and colleagues. No Country for Young Grads. Burning Glass Institute, July 2025.
- Deming DJ, Yuchtman N, Abulafi A, Goldin C, Katz LF. The Value of Postsecondary Credentials in the Labor Market: An Experimental Study. American Economic Review. 2016;106(3):778-806.
- Pallais A. Inefficient Hiring in Entry-Level Labor Markets. American Economic Review. 2014;104(11):3565-3599.
- Kassi O, Lehdonvirta V. Do Digital Skill Certificates Help New Workers Enter the Market? Evidence from an Online Labour Platform. CESifo Working Paper 7810, 2019.
- Athey S, Palikot E. The Value of Non-Traditional Credentials in the Labor Market. arXiv 2405.00247, 2025.
- Castano-Munoz J, Rodrigues M. Open to MOOCs? Evidence of their impact on labour market outcomes. Computers and Education. 2021;173:104289.
- Aramburu J, Goicoechea A, Mobarak AM. Coding Bootcamps for Female Digital Employment: Evidence from an RCT in Argentina and Colombia. World Bank Policy Research Working Paper 9721, 2021.
- World Economic Forum. The Future of Jobs Report 2025. January 2025.
- National Association of Colleges and Employers. Job Outlook 2026 and related releases, November 2025 to April 2026.
- Indeed Hiring Lab. Where Do College Degrees Still Matter in a Skills-First Job Market? 28 January 2026.
- Credential Engine. Counting U.S. Secondary and Postsecondary Credentials 2025. Executive summary.
- SHRM. Survey on alternative credentials in hiring, April 2022 (press release; funded by a Walmart grant).
- Gallup and 2U. Boot Camp Graduates Study, 2022. Commercial survey.
- Google. Google Career Certificate graduates reach 1 million, 3 December 2024; Coursera. 2025 Learner Outcomes Report. Company-reported.
- Stanford HAI. The 2026 AI Index Report. Summary pages.
- Oseguera M. Do Firms Like STARs? AEA RCT Registry AEARCTR-0012586, 2023.
The earnings premium table, the Signal Strength Scorecard, the signal stack plan and the 6-question test are CEOtudent analyses built on the sources above. All other figures are reported as printed in those sources. Several widely repeated figures were not used because they could not be found in primary text, including a “75 percent” Google certificate outcome, older bootcamp placement rates from 2017 press coverage, and a commonly cited share of recruiters who check GitHub.
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
