We present a model in which the owner of the firm enjoys a private benefit from developing a personal relationship with the executives. This may lead the owner to retain a senior executive in office even though a more productive replacement is available. The model shows that the private returns of the employment relationship distort executive selection, reducing the executives’ average ability and the firm productivity. We estimate the structural parameters of the model using a panel of Italian firms with information on the nature of the controlling shareholder, matched with individual records of their executives. These estimates are used to quantify the relevance of private returns and the related productivity gap across firms characterized by four different types of ownership: government, family, conglomerate and foreign. We find that private returns are large in family and government controlled firms, while smaller with other ownership types. The resulting distortion in executive selection can account for TFP differentials between control types of about 10%. The structural estimates are fully consistent with a set of model-based OLS regressions, even though the sample moments used by the two approaches are different.