Case supporting the view. When a trained employee leaves, the firm loses the embodied skill it invested in; the receiving firm captures the benefit. If the trained employee's tenure after training is short, the firm has paid the training cost without recouping it in extra productivity. For firms in sectors with high mobility — technology, consulting, finance — this leakage is real and persistent, and some firms respond rationally by minimising training investment and hiring already-trained staff from competitors.
Case against the view. The 'wasted' framing ignores several things. First, trained employees are typically more productive while they are still at the firm — the firm captures value during the employment, which may well exceed the training cost even for short tenures. Second, firms with strong training reputations recruit better: candidates value development opportunities and accept lower pay for them, so the training investment pays back in recruitment savings. Third, firms that train well typically have higher retention — ironically, the firms that most 'protect' themselves against the training-and-leaving risk by not training are often the firms with the highest turnover, which is a bigger drain than any training cost. Fourth, the alternative — a workforce without ongoing training — produces lower quality, lower productivity, and worsening competitive position over time. 'Not training' is not a cost-free alternative.
Case for a middle view. The right policy is not to abandon training but to design it so the firm captures most of the benefit even if an employee eventually leaves. Make training specific to the firm's processes and products (less portable to competitors), bond training investments with retention periods (common in management training), focus on building team-wide capability rather than individual skills (which cannot leave even if individuals do), and treat training as a recruitment and engagement tool (which benefits whether or not the specific trainee stays).
Reasoned judgement. The view is substantially wrong. Training is not wasted when employees leave because the firm captures productivity, retention and recruitment benefits during the employment and from the firm's reputation as a developer of talent. The alternative — don't train, hire already-trained from elsewhere — works at the margin for individual firms but fails if too many firms adopt it (there is no one to hire from), and does not avoid the long-run cost of a less capable workforce. The strongest statement is: training produces returns during employment, raises retention, improves recruitment and builds capability that outlasts individual employees. It is 'wasted' only if the firm fails to design or evaluate it well — a problem of training quality, not of training as such.