The Competency Framework Illusion

Why Your Development System Might Be Making Performance Worse

Here's a question that should trouble every L&D leader: What if the competency framework you spent months building is actually hindering the performance it's meant to improve?

That's not hyperbole. It's what the research tells us, though you'd never know it from walking around most L&D departments. We've built an entire professional edifice around detailed behavioural descriptors, elaborate rating scales, and comprehensive assessment matrices, all whilst ignoring the mounting evidence that these systems don't predict performance any better than flipping a coin.

Campion et al. (2011) delivered what should have been a death blow to competency modelling when they found that most frameworks demonstrate rigour on only one of ten critical variables compared to traditional job analysis. More damning still, they concluded that "most competency models are implemented in organisations without validation procedures". Think about that for a moment. We're making career-defining decisions about people based on systems that haven't been tested to see if they actually work.

This isn't about being anti-framework or anti-systematic. It's about being pro-performance. And if we're serious about performance, really serious, we need to stop pretending that complex behavioural taxonomies are the answer when the evidence suggests they're often part of the problem.

The Research Reckoning

The academic demolition of competency frameworks isn't new; it's been building for over two decades, though our profession seems remarkably skilled at not noticing inconvenient truths.

Barrett and Depinet (1991) systematically challenged David McClelland's foundational 1973 assertions about competency testing. Their review found that empirical evidence failed to demonstrate that competency testing showed superior predictive validity compared to cognitive ability testing. In other words, the intellectual foundation upon which the entire competency industry was built doesn't hold up under scrutiny.

The validation studies that do exist paint an even bleaker picture. Lievens et al. (2016) found "very weak correlations across stations" for identical competencies in healthcare assessment centres, with "higher, positive correlations between the scores for different competencies within the same station than between the same competencies across different stations". What this means in plain English is that assessors were rating the person, not the competency. The entire premise, that different competencies measure different things, collapsed.

When trained assessors can't agree on what they're seeing, and the same competency produces different ratings depending on context, we're not measuring anything meaningful. We're engaging in "measurement theatre", elaborate performances that feel scientific but lack the predictive power that would make them useful.

But here's what's really interesting: most L&D teams don't even attempt validation. We build frameworks because they feel professional, comprehensive, and reassuringly complex. The illusion of scientific rigour becomes more important than actual scientific rigour

.

The Psychology of Professional Self-Deception

Understanding why competency frameworks persist despite evidence of their ineffectiveness requires a brief detour into human psychology. We're not dealing with malicious incompetence here; we're witnessing the predictable operation of cognitive biases that make these systems feel effective regardless of their actual impact.

Kahneman and Klein (2009) identified what they called the "illusion of validity"; our tendency to maintain confidence in predictions when we can construct coherent narratives, even when those predictions have zero correlation with outcomes. Competency frameworks, with their detailed behavioural descriptors and logical structures, create precisely this kind of seductive consistency. The more elaborate the framework, the stronger the illusion becomes.

This combines devastatingly with what Heath and Heath (2007) describe as "complexity bias", our professional preference for elaborate solutions over simple ones. In L&D contexts, this manifests as the belief that detailed frameworks signal intelligence and thoroughness. Simple interventions get dismissed as "too obvious" or unsophisticated, despite often being more effective.

Then there's the "competency trap", organisational inertia around ineffective practices. Companies become skilled at using frameworks that don't work, making change seem risky. HR systems get built around competency models. Professional certification programmes reinforce competency-based approaches. The sunk cost fallacy prevents abandonment of substantial framework investments.

Perhaps most powerfully, competency frameworks satisfy fundamental human needs for control and predictability in inherently uncertain environments. They provide the illusion of understanding complex human performance through manageable categories. They create shared language for discussing subjective performance judgements. These psychological benefits persist regardless of whether the frameworks actually improve performance.

What Actually Drives Performance

Whilst L&D has been obsessing over behavioural competencies, performance science has been quietly revolutionising our understanding of what makes people effective at work. The findings should fundamentally reshape how we approach development, if we're brave enough to listen.

Thomas Gilbert's groundbreaking work identified the critical insight that competency frameworks miss entirely: approximately 75-80% of performance factors are environmental rather than individual (Gilbert, 1978). His Behavioural Engineering Model distinguished between accomplishments (valuable results) and behaviour (what people do), arguing that management's lack of information and support, not individual deficiencies, represents the greatest barrier to performance.

Gilbert's framework prioritises six factors through a simple matrix.

• Environmental factors come first: data and information (clear performance expectations, standards, feedback),

• Instruments and resources (tools, materials, processes),

• Incentives and consequences (meaningful rewards and accountability)

Individual factors follow:

• Knowledge (required skills and information),

• Capacity (physical and mental capabilities),

• Motives (personal drives and motivation)

The implications are profound. Instead of analysing behavioural competencies, we should be asking: Do people know what good performance looks like? Do they have the tools and resources to achieve it? Are there meaningful consequences for success and failure? Only then should we consider individual knowledge, capacity, and motivation.

Recent research has further undermined traditional assumptions about performance predictors. Sackett et al. (2021) found that structured interviews may actually be stronger predictors of job performance than cognitive ability, challenging fifty years of accepted wisdom. Job knowledge tests and work sample tests consistently outperform complex competency assessments.

The pattern is clear: specific, context-relevant measures predict performance better than abstract, generalised competencies. A structured interview focused on actual job requirements beats elaborate behavioural frameworks. A work sample demonstrating real skills trumps competency ratings. Job knowledge directly related to performance outcomes surpasses generic leadership competencies.

This isn't cynicism. It's clarity. We're not here to look busy with sophisticated frameworks. We're here to make work work better.

The Performance-First Alternative

Gilbert's Behavioral Engineering Model provides a practical alternative that's both scientifically sound and remarkably straightforward. Instead of dissecting behaviour into elaborate competencies, it focuses on the environmental and individual factors that actually create worthy performance.

The beauty of BEM lies in its diagnostic power and clear prioritisation. When performance problems arise, Gilbert's framework provides a systematic approach to identifying root causes. Environmental assessment comes first: information systems, resource availability, and consequence structures. Individual factors follow only after environmental barriers are addressed.

Consider a typical competency framework response to poor sales performance: identify leadership competencies, assess behavioural gaps, design development programmes, implement training modules, and measure competency acquisition. It's elaborate, expensive, and usually ineffective because it assumes the problem lies within individuals rather than systems.

The BEM alternative starts with environmental factors. Do salespeople have clear performance standards and regular feedback on results? Do they have effective tools, current market intelligence, and streamlined processes? Are there meaningful rewards for success and consequences for failure? These questions often reveal system deficiencies that no amount of individual development can overcome.

When environmental factors are sound, BEM turns to individual elements, but with laser focus on accomplishments rather than behaviours. Instead of measuring generic leadership competencies, we assess specific job knowledge. Instead of rating communication skills, we evaluate actual results achieved through communication. Instead of developing elaborate behavioural taxonomies, we focus on the knowledge, capacity, and motivation required for specific accomplishments.

Leading organisations are demonstrating how this works in practice. Adobe abandoned their traditional competency-based annual reviews, which consumed 80,000 manager hours annually whilst producing minimal performance improvement (Buckingham & Goodall, 2019). Their new "Check-in" system features ongoing conversations focused on goals, growth, and impact rather than competency ratings, resulting in 30% reduction in voluntary turnover.

Netflix's approach eliminates performance management entirely in favour of continuous coaching and their "keeper test": would you fight to keep this employee if they wanted to leave? Their context-not-control philosophy focuses on providing comprehensive business context rather than specific directives (McCord, 2018). The result: 4% voluntary turnover compared to 14% industry average.

Google's OKR system demonstrates how goal-setting can replace competency frameworks. Instead of rating behavioural competencies, employees set quarterly objectives with measurable key results. Success is measured at 60-70% achievement, encouraging stretch goals whilst maintaining psychological safety (Doerr, 2018).

These examples share common elements: focus on outcomes rather than activities, continuous feedback over periodic assessment, and prioritising environmental factors over individual development. When performance issues arise, the first question isn't "what competencies are missing?" but "what barriers exist in the system?"

Making It Practical

Implementation begins with environmental assessment using Gilbert's framework. Information systems: do people have clear performance expectations, regular feedback, and transparent standards? Resources: are tools, processes, and materials adequate for successful performance? Consequences: are there meaningful rewards for success and accountability for failure?

This assessment typically reveals system deficiencies that explain performance problems better than individual competency gaps. Addressing these environmental factors often produces immediate improvements whilst reducing the need for extensive development programmes.

Individual development, when needed, focuses on specific knowledge, capacity, and motivation gaps directly related to accomplishment requirements. Instead of generic leadership development, programmes address specific knowledge gaps that prevent successful performance. Instead of broad communication skills training, interventions target specific capacity limitations that inhibit results.

The measurement system transforms accordingly. Key performance indicators focus on business impact, goal achievement, and results delivery rather than competency scores. Development ROI is calculated through performance improvement attribution rather than training completion rates.

If you're serious about impact, this is where you start, not with another competency framework, but with an honest assessment of whether people have what they need to succeed. Most of the time, they don't. And no amount of behavioural development will fix environmental deficiencies.

The Uncomfortable Truth

Here's what the research tells us that most L&D teams don't want to hear: competency frameworks persist not because they work, but because they satisfy our psychological need for control and professional credibility. They make complex human performance feel manageable through detailed categories and systematic approaches.

The evidence against competency frameworks is overwhelming. The alternatives are proven. The choice is ours. We can continue building elaborate systems that feel professional whilst delivering minimal performance improvement, or we can focus on the environmental and specific individual factors that actually drive results.

This isn't about abandoning systematic approaches; it's about using systematic approaches that work. Gilbert's model is systematic. Environmental assessment is systematic. Focus on accomplishments rather than behaviours is systematic. But it's systematic in service of performance, not systematic in service of looking systematic.

The competency framework illusion persists because it serves our professional identity more than our learners' performance. Acknowledging this reality doesn't make us failures; it makes us honest. But professional maturity requires moving beyond what feels comfortable toward what actually works.

Your learners deserve systems designed for their success rather than your professional comfort. The evidence is clear, the alternatives are proven, and the choice is yours.

What would you do differently if you designed for performance instead of competency? And here's the bigger question: if competency frameworks can't predict performance, what does that tell us about everything else we assume to be true about learning and development?

---

References

Barrett, G. V., & Depinet, R. L. (1991). A reconsideration of testing for competence rather than for intelligence. American Psychologist, 46(10), 1012-1024.

Buckingham, M., & Goodall, A. (2019). Nine lies about work: A freethinking leader's guide to the real world. Harvard Business Review Press.

Campion, M. A., Fink, A. A., Ruggeberg, B. J., Carr, L., Phillips, G. M., & Odman, R. B. (2011). Doing competencies well: Best practices in competency modeling. Personnel Psychology, 64(1), 225-262.

Doerr, J. (2018). Measure what matters: How Google, Bono, and the Gates Foundation rock the world with OKRs. Portfolio.

Gilbert, T. F. (1978). Human competence: Engineering worthy performance. McGraw-Hill.

Heath, C., & Heath, D. (2007). Made to stick: Why some ideas survive and others die. Random House.

Kahneman, D., & Klein, G. (2009). Conditions for intuitive expertise: A failure to disagree. American Psychologist, 64(6), 515-526.

Lievens, F., Dilchert, S., & Ones, D. S. (2016). The importance of exercise and dimension factors in assessment centers: Simultaneous examinations of construct-related and criterion-related validity. Human Performance, 22(5), 375-390.

McCord, P. (2018). Powerful: Building a culture of freedom and responsibility. Silicon Guild.

Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2021). Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. Journal of Applied Psychology, 107(11), 2040-2068.