An Inquiry into Higher Education, Applied Skills, and Early-Career Hiring

The argument: AI may be changing the value of higher education for early-career workers. As skill requirements shift more rapidly and employers place greater weight on applied capability, the differentiating value of another degree may increasingly depend on the experiences and opportunities attached to it – not simply the credential itself.

Do Degrees Differentiate?

For Gen Z and millennial individuals in North America and many other developed countries, the tough conditions of the current labor market, reflected by a persisting discrepancy between job demand and supply of entry-level workers, create sentiments of anxiety and frustration. While conversations are dominated by anecdotal evidence, multiple data sources are confirming the reality of a decline in entry level jobs in the US and North America across sectors. For example, the U.S. market has seen a rapid decline in the portion of salaried job openings with no experience, going from 1 in every 15 postings to 1 in 50 postings in under 4 years. This trend is aligned with an overall decrease in salaried postings (for employers across sectors, and for postings across experience levels), and with senior-level candidates being the true winners in the current market – in the US, the portion of roles requiring experience has doubled.

In Prior Periods of Job Scarcity, Higher Education Offered a Way to Differentiate

In prior periods of job scarcity, including recessions, early career stage individuals increasingly sought out higher education to improve their employment prospects in response to economic uncertainty. Intuitively, they likely viewed forms of post-secondary education as a vehicle for competitive differentiation through knowledge specialization or ‘signalling’ (i.e., in absence of more work experience, may use further education as a proxy for ambition, intelligence and skills). However, the labor market is shifting substantially as AI develops – regardless of individual organization’s AI maturity and adoption levels, there are broader trends indicating that AI is changing the (incremental) value of higher (post-secondary) education for early career workers.

AI May Be Contributing to Skill Instability

The development of AI technologies, in various domains (e.g., Generative AI, agentic AI, predictive analytics) may be contributing to skill instability across sectors, especially observable in knowledge intensive industries. In other words, the half-life of a skill is decreasing – World Economic Forum’s Future of Jobs 2025 survey stated that across employers and sectors surveyed, 39% of workers’ core skills are expected to change by 2030. Naturally, the degree of skill disruption a sector faces is likely related to that sector’s exposure to AI – with employer respondents from knowledge-intensive industries like financial services and electronics confident that AI exposure will be significant (97% and 95% of respondents, respectively). Importantly, WEF estimated that over one third of early career workers are in medium/high exposure occupations. This skill disruption is partially explained by the overlap between AI’s capacity for substitution of the skills that roles require (driving automation) and potential complementarity between human and AI collaboration where human input is still valuable (driving augmentation). GenAI model usage data suggests that both cases are occurring today: Anthropic’s analysis of Claude usage in 2025 showed that at the task level, “57% of usage “suggests augmentation of human capabilities, while 43% suggests automation.”

One way we can conceptualize the extent of this skill disruption at a macroeconomic level is thinking through the portion of tasks that are human-only today and projection of this in future – for example, the financial services industry is expected to see a 18 percentage point decline in human only tasks by 2030, and the electronic industry is expected to see a 15 p.p. decline in this same period. If the share of human-only tasks is declining substantially, the role and unique value of the human workers is likely shifting dramatically as well.

What Does Skill Disruption Mean for the Value of Higher Education?

In a market of rapid skill disruption, the value of post-secondary education is somewhat dependent on institutions’ ability to adapt curricula and program offerings in alignment with changing market demands. Existing human capital research suggests that historically, formal education institutions have struggled to keep up with the rate of change for labor market requirements. Therefore, employers seem to be thinking differently about whether education should remain as the best proxy for individual capability and knowledge.

In previous labor market conditions, there was a clear premium in compensation based on formal education level, across sectors. Today, this trend remains for jobs that are not AI related, employers are willing to pay more for candidates with higher levels of education – and the premium they’re willing to pay increases substantially as the candidate climbs the ladder (i.e., in the UK, the wage premium for PhD holders is 3x greater than for premium for masters degrees). Interestingly, this gradient does not really exist for AI roles – in M. Bone et. Al’s 2025 research it was observed that the wage premium associated with incremental higher education was minimal for AI roles in the UK job market. In essence, this means that if two UK candidates were equal in experience, an employer may not be willing to pay substantially more for the PhD candidate compared to the non-degree candidate. While wage premiums and compensation do not directly represent desirability and candidate competitiveness, similar hiring patterns North America may suggest that the incremental value for obtaining ‘additional’ higher education may be lower for AI roles than in other occupations.

AI-Related Skills Are Spreading Across Occupations

Of course, AI roles only represent a portion of jobs in the economy, but this trend is significant for two core reasons: (a) the skills that AI roles demand are proliferating into roles across other sectors and job types (e..g, OECD’s 2025 blog post, ‘The state of AI jobs in Canada: What 12 million job postings reveal about hiring trends’ states what we know intuitively – AI related work is spreading across multiple sectors, despite current concentration in a handful of sectors including Professional services, financial services, manufacturing, and publishing). This is also reflected in OECD analysis of AI-related job postings by occupation in Canada, where the largest share of roles consistently falls into the “Other” category rather than traditional AI-intensive occupations such as software engineering or data science. This suggests that demand for AI-related skills is increasingly extending beyond the technical roles most commonly associated with AI.

Could Skills-Based Hiring Practices Spread Beyond AI Roles?

As demand for AI-related skills continues to spread across a broader set of occupations, trends in skills-based hiring practices observed for AI role hiring may become evident across other professions. If employers prioritize applied capabilities, the broader perception of the incremental value of formal higher education for early and mid-career roles across the market may shift as well. Employers are seemingly looking to credibly prove that the individual has a dynamic set of skills as evidenced by application and experience – and are increasingly measuring skills through contribution/evidence generated in MOOCs/micro-certifications, portfolios, on-the-job projects, apprenticeships.

The Value of Higher Education May Increasingly Depend on Access to Applied Experience

As such, the value of higher, incremental education may increasingly depend on the extent to which it grants the individual access to unique opportunities for technical, applied experience; e.g., employer partnerships, etc. The marketplace and frameworks of credentialing (such as microcredentials/certifications) are evolving, and this may prompt early/mid career individuals (especially those in high exposure industries) to think through avenues that allow them to gain the technical experience at lower opportunity costs (i.e., time and money spent to obtain the degree). What seems less clear is whether employers have a consistent way of evaluating different forms of applied experience in a skills-based hiring model – for example, how a Data Science capstone project compares with three months of freelance project work as evidence of capability.

Sources

• Michael T. Nietzel, “What the Changes in Entry-Level Jobs Mean for New College Graduates,” Forbes, August 2, 2026.
• Peter Q. Blair and Bobby W. Chung, “How Much of a Barrier to Entry Is Occupational Licensing?” The Quarterly Journal of Economics 136, no. 4 (2021): 2495–2564.
• World Economic Forum, The Future of Jobs Report 2025 (Geneva: World Economic Forum, 2025).
• World Economic Forum, Artificial Intelligence and the Future of Entry-Level Work: A Framework for Safeguarding and Reinventing Early Career Pathways (Geneva: World Economic Forum, 2026).
• Carlo Pizzinelli, Augustus J. Panton, Marina Mendes Tavares, Mauro Cazzaniga, and Longji Li, “Labor Market Exposure to AI: Cross-country Differences and Distributional Implications,” IMF Working Paper No. 2023/216 (2023).
• Kunal Handa, Alex Tamkin, Miles McCain, Saffron Huang, Esin Durmus, et al., “Which Economic Tasks Are Performed with AI? Evidence from Millions of Claude Conversations,” Anthropic (2025).
• Wolfgang Lutz, Claudia Reiter, Caner Özdemir, Dilek Yildiz, Raquel Guimaraes, and Anne Goujon, “Skills-Adjusted Human Capital Shows Rising Global Gap,” Proceedings of the National Academy of Sciences 118, no. 7 (2021): e2015826118.
• Matthew Bone, Eugenia Gonzalez Ehlinger, and Fabian Stephany, “Skills or Degree? The Rise of Skill-Based Hiring for AI and Green Jobs,” Technological Forecasting and Social Change 214 (2025).
• Diego Eslava, Fabio Manca, and Caroline Paunov, “The State of AI Jobs in Canada: What 12 Million Job Postings Reveal About Hiring Trends,” OECD, March 18, 2025.