Entry Level. Experience Required.

The sentence arrives before the explanation

The practical answer is that “entry-level” is usually an employer’s label, not a shared legal or statistical category. A vacancy may be intended for a new entrant while still asking for evidence gained through study, internships, volunteering, projects or adjacent work. That does not make every requirement reasonable. Nor can the available evidence tell us how many entry-level postings demand experience across all countries and occupations.

The number we do not have

A useful warning comes before any percentage. Youth unemployment describes people without work who are available for and seeking it. A NEET rate describes young people who are neither employed nor in education or training. A vacancy study describes what employers write in advertisements. A callback audit measures whether an application produced an employer response. None is a count of entry-level jobs, recent graduates, rejected applicants or completed junior hires.

One indicative figure is worth discussing precisely because it is limited. The UC Davis paper cites an analysis of almost four million LinkedIn postings since late 2017 in which 35% of postings classified as entry-level asked for years of prior relevant experience. That is evidence about one platform’s analysis and classification, not an authoritative cross-country prevalence rate. It does not tell us whether the experience wording was a genuine minimum, a preference or inherited language in a job description.

The scale of the distinction matters. Eurostat reported that 11.0% of EU residents aged 15–29 were NEET in 2025, including 12.8% of those aged 20–24 and 14.7% of those aged 25–29. In the United States, the Bureau of Labor Statistics reported a 10.0% unemployment rate for people aged 16–24 in its 2025 annual estimates. BLS notes that these estimates use an 11-month average because October data were not collected during the federal government shutdown, so they are not strictly comparable with ordinary annual averages for other years. Neither figure measures experience-related rejection.

Why employers ask for the past

There is a mundane cost behind the paradox. A University of California, Davis working paper models training as an investment of experienced employees’ time. It describes a feedback loop: firms may avoid inexperienced entrants because training is costly, while entrants need a first job to acquire the experience those firms request. The model explains a possible mechanism; it does not prove that every employer behaves this way or that every requirement is justified.

The underlying Danish employer study gives the training problem a clearer boundary. In a survey linked to administrative data, one-third of Danish employers considered the time to train new recruits a hiring obstacle. The finding concerns employer perceptions in Denmark, not entry-level recruitment specifically and not employers everywhere. It does, however, show why a company may prefer someone who appears ready to contribute sooner.

A randomized resume audit in metropolitan Manila offers a different kind of evidence. Researchers submitted 7,172 resumes to 1,793 entry-level vacancies between October 2015 and March 2016. Among 18- and 19-year-olds applying to jobs that did not explicitly require experience or postsecondary education, one or two years of relevant experience increased callback rates by 2.4 percentage points, or about 11%. The result concerns callbacks, not offers, pay or long-term employment, and it does not show that experience is always more valuable than education.

Experience is not one thing

A paid internship, a regulated apprenticeship, a university project, volunteering and a personal project can all produce evidence. They do not produce the same evidence, carry the same protections or demand the same resources. Treating them as interchangeable is convenient for a job description and unfair to the person reading it.

The ILO’s 2021 review finds that evidence for a direct employment benefit from unpaid internships remains contested and difficult to measure. It also discusses unequal access: students with greater financial resources may be better positioned to take unpaid placements and capture their networking and CV benefits. Someone working to pay rent cannot treat three months without income as a neutral “opportunity”.

An apprenticeship has a more defined architecture in the European framework. It combines company-based training with school-based education and leads to a nationally recognised qualification. The EU framework also points to a contractual relationship, pay, social protection, trained supervisors and health and safety, while recognising that national systems determine the legal details. A placement using the word “apprenticeship” does not automatically prove that it meets every criterion.

The European Commission describes traineeships as limited periods of work practice with a significant learning and training component. Its current page also records 2024 proposals to strengthen quality, fair pay, social protection and enforcement; those proposals are not the same thing as enacted law across Member States. The practical questions remain useful anywhere: who teaches, what is being learned, what is agreed in writing, who is protected and whether the arrangement is genuinely developmental rather than ordinary work without ordinary pay.

What should count before the first job?

For employers, years of tenure are a blunt proxy. A more defensible process is to define the capability the role actually needs, then assess it through proportionate, structured evidence: a short scenario, a limited work sample, a practical conversation or a review of relevant coursework. That is editorial analysis, not a universally validated hiring formula. The studies reviewed do not establish that portfolios reliably bypass experience screens.

For applicants, credible evidence does not need to pretend to be employment. A course project can show the problem, the person’s contribution and the result. Volunteering can show reliability in a real setting. A personal project can show judgement if its limits are clear. As a practical safeguard, any work sample should be bounded, use no confidential information and avoid producing unpaid commercial work that the employer could simply use. A short demonstration is one thing; a working week disguised as an interview is another.

There is also a useful distinction between asking whether experience is required and asking what problem the requirement solves. If the role needs regulated permission, safety training or a technical credential, the employer should say so. If it needs familiarity with a tool that can be taught in two weeks, that is a different conversation. Precision helps applicants decide whether to apply and helps employers stop calling a training plan a candidate requirement.

The newer explanation should stay in its lane

AI may be changing one part of this story, but it did not invent the first-job paradox. A 2026 IZA study using near-universe U.S. online vacancy data found a 14–15% relative decline in junior software-developer vacancies compared with senior vacancies after ChatGPT’s public release. It also found that rising experience requirements were driven mainly by employers asking for more experience within the same job titles. This is evidence about software-development vacancies in the United States, not proof of economy-wide junior job elimination or a universal AI explanation.

The central dilemma remains older and more ordinary: employers want evidence that someone can contribute, while many routes to that evidence are unpaid, informal or available mainly to people with spare money and time. The answer is not to declare degrees useless, portfolios magical or employers dishonest by default. It is to make the bargain visible. What does the role teach? What must the applicant already know? What is paid, protected and assessed? And if the job is truly entry-level, why is its first demand a past the applicant could only have obtained elsewhere?