When labor economists say “work experience,” they usually just mean (Age-Education-5).  Industrial psychologists are far more subtle.  Quiñones, Ford, and Teachout, The
Relationship Between Work Experience and Job Performance: A Conceptual and
Meta-Analytic Review.
”  (Personnel Psychology, 1995) begins by distinguishing three measures:

Time-based measures are perhaps the most familiar to researchers. These include typical measures such as job and organizational tenure (e.g., months or years in the job). Amount measures refer to numerical counts such as the number of times a task was performed or the number of different jobs held in an organization. Finally, measures that categorize experience qualitatively (e.g. management, accounting, etc.) are referred to as type measures.

Measure matters:

The meta-analyses results also revealed some variation in the relationship between work experience and job performance as a function of measurement mode. The strongest relationship occurred between amount of experience and performance. Time and type measures showed the weakest relationships. Finally, variability in the relationship between experience and performance as a function of level of specificity was found. Task-level experience had the strongest relationship with performance whereas organizational level showed the weakest.

Also of note:

The relationship was stronger when hard performance measures such as work samples were used as compared to soft performance measures such as supervisory ratings.

Plausible explanation:

[I]ndividuals may rate an employee’s performance relative to his or her level of job experience. A poorly performing newcomer may be rated the same as an average performing veteran. Thus, subjective ratings may attenuate actual performance level differences which are captured by more objective “hard” performance criteria.

This article is yet another reason why economists should stop laughing at psychologists and start listening to them.  They have so much to teach us.