Decade-Long Analysis of Skills in Mechanical Engineering Job Advertisements 2010-2022
Plain Language Summary
This study examines over one million online job postings for mechanical engineering positions requiring a graduate degree, spanning 2010 to 2022, to see how the skills employers ask for have shifted over time. We find that engineers are increasingly expected to pair their technical training with business-oriented duties like budgeting and supplier management, alongside a steadily growing reliance on workplace software. Understanding these shifts matters because graduate programs, which can adapt curricula faster than undergraduate ones, need an evidence base for deciding what to teach next.
Contribution
We contribute a decade-long, discipline-specific analysis of graduate-level mechanical engineering labor demand by combining a Bayesian-optimized, marker-based classifier for isolating graduate ME postings with a multi-model (null, linear, logarithmic, exponential, quadratic) trend-selection approach applied to 76 employer-articulated skills, moving beyond the short-term, cross-disciplinary snapshots common in prior job-posting research.
Research Questions
- How has the prevalence of employer-articulated technical and professional graduate mechanical engineering skills changed from 2010 to 2022, and what do these changes reveal about employer demand for skills?
Methods
We analyzed 1,142,245 graduate-level mechanical engineering job postings from the Burning Glass Technologies (now Lightcast) database, filtered from an initial pool of over 372 million postings using a manually-validated, Bayesian-optimized confidence-scoring classifier (F1 = 0.83) to isolate graduate-level ME roles. Skills were extracted from BGT's CanonSkillCluster field, retaining the 76 skills with at least 2% average prevalence across the full period; for each, we fit five candidate trajectory models (null, linear, logarithmic, exponential, quadratic) and selected the best fit using corrected AIC, while controlling for job posting length and verifying regression assumptions via Durbin-Watson, Shapiro-Wilk, and Breusch-Pagan tests.
Key Findings
Of the 76 tracked skills, "Drafting and Engineering Design" was the most prevalent, appearing in roughly half of postings each year, while 16 skills qualified as high-demand (≥10% average prevalence). The skills with the largest percentage gains clustered around business and organizational functions—"Regulation and Law Compliance" (+184%), "Product Management" (+138%), and "Contract Management" (+98%) among them—alongside software and data-oriented skills like "Mathematical Software" (+137%) and "Data Analysis" (+103%). We interpret these patterns through two complementary profiles, the "business-aware technical liaison" and the "proficient software executor," though noting that the dataset ends just after ChatGPT's November 2022 launch, so these trends predate the current generative AI landscape rather than reflecting it.
Implications
Since program directors cannot reasonably address every rising skill given finite curricular space, we suggest graduate programs prioritize skills with sustained, not-yet-plateaued growth (e.g., simulation, data analysis, statistics) through coursework, capstones, or electives, while relying on cross-campus partnerships, dual degrees, or certificates to cover business-adjacent skills like contract management that sit outside typical ME departmental boundaries. Deeply workplace-contextual and already-saturating skills, meanwhile, may be better left to co-ops, internships, and on-the-job learning rather than formal instruction, and we flag post-2022 AI-related competencies as an explicit direction for future research.