Capital social, ¿negativo?

Esto es, al parecer, lo que averiguan Bentolila y otros (2004):

We have shown, with both US and European Union data, that there is indeed a wage discount of around 5% to 7% for jobs found through contacts. This evidence is reinforced once we control for whether the contact person was working at the same firm (which should purge the positive effect of referrals). The evidence for both the US and Europe also supports the presence of a trade-o. between quicker jobfinding and lower wages. We have also shown that the wage discount on contact jobs persists over time, is increasing in the worker’s endowment of contacts (as captured, for instance, by the number of older siblings), and is largely orthogonal to measures of individual cognitive ability. Overall, the evidence supports the claim that social contacts tend to distort workers’ occupational choices, inducing mismatch. Additionally, for European regions we have also found evidence of negative externalities since the regional importance of contacts for jobfinding depresses average regional wages beyond what would result from the simple aggregation of the wage discounts estimated with individual-level data.

Todo ello puede ser verdad, pero no se me ocurre qué otra alternativa tienen quienes suelen buscar trabajo a través de amigos o familiares. Suelen tener menos cualificaciones formales y menor nivel de estudios, luego es natural que suplan con "capital social" sus carencias de "capital humano". ¿Acaso sería mejor para ellos que alguien, al llegar al puesto de trabajo conseguido a través de contactos, les dijera: espera, busca otro trabajo, que quizás éste no te conviene? ¿O quizá imaginamos que un INEM omnisciente debiera tomar las decisiones de búsqueda por ellos?