📄 Abstract
Employers in the United States and comparable developed economies lose an order of magnitude more output to impaired and absent working time than they spend on the wellbeing programmes intended to recover it, yet the return to those programmes remains one of the least securely established quantities in applied human resource economics. This article argues that the difficulty is not primarily a difficulty of intervention design but of measurement regime. The prevailing regime is retrospective: it prices output loss after the loss has occurred, through claims run-out, absence registers and annual survey instruments whose reporting lag typically runs from nine to twenty-four months. A systematic review of 74 studies, of which 31 supplied extractable quantitative parameters, establishes three economic consequences of that regime. First, attribution failure: the reported return to wellbeing expenditure falls monotonically as identification strengthens, from 3.27:1 in observational meta-analysis to 1.50:1 in quasi-experimental employer studies and to no statistically detectable effect in the two large randomised trials, whose confidence intervals exclude the great majority of prior estimates. Second, latency: retrospective measurement can price a loss but cannot avert it, so the recoverable share of the loss decays with detection lag. Third, denominator error: presenteeism, the larger component of impaired output, is the component measured least precisely, so programme design optimises against the smaller and better-measured claims margin. Predictive wellbeing analytics is then assessed as a forward-looking alternative. The review finds that its economic value derives not from prediction as such but from targeting, which is the single feature common to the interventions that have survived rigorous evaluation. That value is bounded by the base rate of the risk detected, by the efficacy of the interventions the score triggers, and by a privacy and trust cost that is economic rather than merely ethical. The article concludes that neither regime is sufficient alone, and sets out a policy architecture in which prediction performs the targeting and embedded prospective randomisation performs the evaluation, supported by organisational rather than individual risk features, a restored federal incentive standard, and disclosure of measurement provenance alongside claimed returns. A comparative extension to India shows which parts of the argument are general and which are artefacts of American benefit financing: attribution failure, measurement latency and the base-rate bound on targeting all transfer, while the private return to investment is lower where sickness cost is partly socialised through the Employees' State Insurance scheme and the great majority of workers lie outside any measured perimeter. Because the claims apparatus that drew American programme design toward the wrong margin is largely absent, India's organised sector is positioned to adopt predictive targeting with embedded evaluation without first replicating the retrospective regime.
🏷️ Keywords
📚 How to Cite:
Dr. N Subbukrishna Sastry, Dr. Manjula Mallya M , EMPLOYEE WELLBEING INTERVENTION AND THE ECONOMIC COST OF PRODUCTIVITY LOSS: FROM RETROSPECTIVE COST MEASUREMENT TO PREDICTIVE WELLBEING ANALYTICS , Volume 13 , Issue 9, September 2026, EPRA International Journal of Socio-Economic and Environmental Outlook(SEEO) , Pages: 1 - 18 ,