The global financial crisis has exposed the absurd effects of short-termism when it comes to bonuses on long-term sustainability and efficiency of enterprises. However, the idea that bonuses can be effective in creating a compensation wedge over relatively standardised salary scales to reward performance and/or human capital (on the supply side of labour) and to provide competitive advantage to firms in attracting human capital (on the demand side of labour) is not necessarily out of touch with reality in many other sectors and occupations. Still, some worrying lessons that we should learn about the distortions introduced by bonuses from the crisis do apply to other sectors as well.
An interesting paper, albeit purely theoretical, titled "Bonus Culture: Competitive Pay, Screening, and Multitasking" by Roland Bénabou and Jean Tirole (NBER Working Paper No. 18936, April 2013) looked at "the impact of labor market competition and skill-biased technical change on the structure of compensation."
The authors found that "Competition for the most talented workers leads to an escalating reliance on performance pay and other high-powered incentives, thereby shifting effort away from less easily contractible tasks such as long-term investments, risk management and within-firm cooperation. Under perfect competition, the resulting efficiency loss can be larger than that imposed by a single firm or principal, who distorts incentives downward in order to extract rents. More generally, as declining market frictions lead employers to compete more aggressively, the monopsonistic under-incentivization of low-skill agents first decreases, then gives way to a growing over-incentivization of high-skill ones. Aggregate welfare is thus hill-shaped with respect to the competitiveness of the labor market, while inequality tends to rise monotonically. Bonus caps and income taxes can help restore balance in agents' incentives and behavior, but may generate their own set of distortions."
Furthermore, "The extent to which [such a correction via bonus caps and income taxes] is achievable depends on how well the government or regulator is able to distinguish the incentive versus fixed parts of compensation packages, as well as on the distortions that may arise as firms try to blur that line or resort to even less efficient screening devices."
One issue with the study is that the model does not allow for heterogeneity between agents and between various sectors of economy. The authors acknowledge this much by stating that "…task unobservability may be less of a concern for some (e.g., private-equity partnerships) and more for others (large banks), but if they compete for talent the high-powered incentives efficiently offered in the former may spread to the latter, and do damage there. Heterogeneity also raises the question of the self-selection of agents into professions and their matching with firms or sectors, e.g., between finance and science or engineering."
Other shortcoming, also mentioned by the authors in their 'what can be done next' discussion is that in some "settings in which high-skill workers become more valuable as firms compete harder for customers, for instance because the latter become more sensitive to quality."