Pay enough, don’t pay too much or don’t pay at all? An empirical study of the non-monotonic impact of incentives on job satisfaction

Pouliakas, K1 Centre for European Labour Market Research (CELMR), University of Aberdeen Business School, Scotland Keywords: Incentives, intensity, job satisfaction, non-monotonic JEL- Code: C23, J28, J33.

Abstract
This paper attempts to test the non-monotonic effect of monetary incentives on job satisfaction. Specifically, 8 waves (1998-2005) of the British Household Panel Survey (BHPS) are used to investigate the ceteris paribus association between the intensity of bonus/profit-sharing payments and the utility derived from work. After controlling for individual heterogeneity biases, it is shown that relatively ‘small’ bonuses exert a significant negative effect on worker satisfaction. In contrast, job utility is found to rise only in response to ‘large’ bonus payments, primarily in skilled, non-unionized private sector jobs. The empirical evidence of the paper is therefore consistent with a ‘V-effect’ of incentives, suggesting that employers wishing to motivate their staff should indeed “pay enough or don’t pay at all”.

1

Research Fellow, Address: CELMR, University of Aberdeen Business School, Edward Wright Building, Dunbar Street, Old Aberdeen AB24 3QY, UK; Tel: ++44 01224 272172; e-mail: k.pouliakas@abdn.ac.uk.

1. Introduction
The principal-agent model, with its convincing illustration of the trade-off that arises between risk and incentive provision when attempting to align the conflicting interests of two contracting parties, remains central for our understanding of the compensation strategies employed by firms (Mirlees, 1976; Holmstrom, 1979). According to the standard model of agency theory, the introduction of financial incentives as part of an agent’s remuneration package will increase his/her productivity