Employers who hold negative stereotypes about older workers hire and train fewer of them—but the strength of that bias depends heavily on the circumstances in which decisions are made. That is the central finding of a new study analyzing representative data from 1,310 German establishments, published open access in the European Journal of Ageing. Using the 2024 wave of the BIBB Training Panel, a survey of human resource managers, business owners, and managing directors, researchers Philipp Heyer, Wiebke Schmitz, and Kathrin Weis show that age-based stereotypes measurably shape personnel decisions, and that workplace digitalization can blunt one of their most damaging consequences.
The study arrives at a moment when Germany, like much of Europe, faces a demographic squeeze. The country has one of the oldest populations and lowest fertility rates in Europe, and policy has shifted decisively away from the early-retirement deals that once let firms shed older workers. The employment rate of 55- to 64-year-olds rose from 64 percent in 2013 to 75 percent in 2023, according to the Federal Statistical Office. Yet stereotypes depicting older employees as less productive or adaptable persist in many organizations, often without an evidence base, and they can quietly distort the cost–benefit calculations behind hiring and training choices.
To quantify those distortions, the researchers drew on seven survey items measuring how respondents on behalf of their organizations rate the performance and adaptability of employees aged 50 and above, adapted from a validated instrument known as the Work-Related Age-Based Stereotypes scale. Items phrased favorably were reverse-coded, so higher values consistently indicate more negative stereotypes. Factor analysis supported combining the items into a single scale with acceptable internal consistency. The two outcomes of interest were shares: the proportion of older employees among all hires a company made in 2023, and the proportion of older employees who received employer-sponsored further training among all older employees in the firm.
Because both outcomes are proportions bounded between zero and one, the team used fractional-logit regression models, reporting average marginal effects. The headline result is stark. A one-point increase on the stereotype scale—which ranged from −2.25 to 2.94 in the sample—was associated with a 2.3 percentage point decrease in the share of older workers hired, relative to an average hiring share of 20 percent, and a 2.1 percentage point decrease in the share of older workers receiving training, relative to an average of 32 percent.
The more intriguing findings emerged when the researchers added interaction terms testing whether structural conditions moderate the stereotype effect. For hiring, firms anticipating staffing difficulties for most or all positions over the next three years showed a conditional effect of stereotypes of only 0.7 percentage points, compared with a 5.7 percentage point decrease among firms expecting no problems—and in the shortage-expecting group the stereotype effect was no longer statistically significant. However, the interaction itself was only weakly significant at the 10 percent level, and the authors are careful to describe this as limited empirical support. Labor shortages may soften hiring bias, but the evidence is not conclusive.
For training, the moderating role of technology was far clearer. In firms using a below-average number of technologies relative to their industry, a one-point increase in negative stereotypes reduced older employees’ training participation by 3.9 percentage points. In firms with above-average technology usage, the corresponding decline was just 0.7 percentage points and not statistically significant. The researchers measured digitalization by counting how many of 13 technology types each establishment used, benchmarked against industry-specific averages to avoid misclassifying firms from technology-intensive sectors as laggards. Notably, neither technology level nor perceived labor shortage had any direct effect on hiring or training shares; their influence operated indirectly, by reshaping how stereotypes fed into decisions.
The theoretical frame is Resource Dependency Theory, which holds that organizations facing environmental uncertainty—resource scarcity or rapid technological change—adjust their strategies to secure what they need. Qualified personnel is a resource, and when external replacements become hard to find or external human capital obsolesces quickly, investing in the existing workforce regardless of age becomes rational. Consistent with that logic, digitalization appears to offset age bias in training by making continuous skill development essential for everyone. Yet the same logic did not extend to hiring: technology level did not amplify or reduce the stereotype effect on hiring, contradicting the authors’ own hypothesis that digital firms would favor younger, supposedly more adaptable candidates.
Robustness checks added an important layer of nuance. The effects of age-based stereotypes on personnel decisions were driven largely by small firms with up to 99 employees. The authors suggest a plausible mechanism: small firms often lack formal HR departments and standardized procedures, leaving individual managers’ attitudes with greater scope to shape outcomes directly, and their smaller, more flexible training budgets make a single manager’s preferences more consequential. Alternative specifications of the technology measure, and a disaggregation into ICT, non-ICT, and AI usage, largely confirmed the main results, though the moderating pattern for training held for ICT and non-ICT technologies but not for AI usage.
The authors are explicit about the study’s limits. The data are cross-sectional, capturing a snapshot rather than the dynamic process of technological change, so causal claims about digitalization eroding bias over time remain beyond reach. The stereotype scale excluded the original instrument’s warmth dimension, training was measured only as received or not, without content or duration, and potential reverse causation—skilled older employees shaping their own evaluations through positive contact—could only be partially addressed through controls for qualification levels and the share of older staff. Firm-size differences may reflect not size itself but variation in HR professionalization and decision formalization, which the data cannot disentangle.
Even with those caveats, the message for policymakers is pointed. Age-based stereotypes and their consequences do not operate in isolation; they are shaped by structural constraints and opportunities. In an economy defined by labor shortages and digital transformation, the authors argue, policy should tackle stereotypes directly while also creating structural incentives for hiring and upskilling older workers—repositioning them not as a burden on welfare systems but as a strategic asset for firms competing for scarce skills.
Subject of Research: How age-based stereotypes among German employers influence hiring and training of older workers, and how labor shortages and workplace digitalization moderate these effects.
Article Title: Age-based stereotypes in hiring and training: how digitalization and labor shortages reshape employer decisions
Article References: Heyer, P., Schmitz, W., & Weis, K. (2026). Age-based stereotypes in hiring and training: how digitalization and labor shortages reshape employer decisions. European Journal of Ageing, 23(1), Article 33. https://doi.org/10.1007/s10433-026-00936-7
Image Credits: AI Generated
DOI: 10.1007/s10433-026-00936-7
Keywords: age-based stereotypes, older workers, hiring discrimination, further training, digitalization, labor shortages, Germany, BIBB Training Panel, human resource decisions, demographic change, fractional-logit regression, firm size
News Source: Ophelia Keating. (October 30, 2026). Digitalization Erodes Age Bias in Training, German Employer Data Show. Scienmag.
