Fixed salary or incentive contract? The effect of stickiness of compensation contracts

ABSTRACT

This study adopts behavioural contract theory through a mathematical model and clarifies the situation in which a fixed–salary contract is preferable to an incentives–based one for the principal. Theoretically, the expected utility for the principal is higher under an incentives–based contract but, in reality, there are companies that employ people via fixed–salary contracts.

This study fined that, in the multitasking case, a case exists in which the principal should offer a fixed–salary contract, and due to behavioural elements, compensation contracts will become inefficient; however, inefficiency may be reduced by using appropriate performance measures.

KEYWORDS:

• Stickiness of compensation contract
• behavioural contract theory
• fixed salary contract
• performance evaluation

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. For example, theoretically Lambert (2001) surveyed, and empirically Ichinowski, Shaw, and Prennushi (1997), Kauhanen and Piekkola (2006) and Lazear (2000) showed.

2. LEN stands for ‘linear contract’, ‘exponential utility’ and ‘normally distributed performance measure’. Refer to Lambert (2001) for further discussions.

3. Another pioneer study in behavioural contract theory is Casadesus–Masanell (2004), who incorporated three intrinsic motivations, norms, ethical standards and altruism in an agency model and identified conditions under which these motivations mechanisms arise and show how they promote trust.

4. For example, to improve profitability, firms prioritize sales districts, seasons or customers. Therefore, directors often instruct managers to follow a priority policy. From a cost management perspective, manufacturing department heads might offer a working procedure to middle managers to reduce spoilage in their departments. Moreover, to save the contingent loss, directors instruct managers to pursue a risk management policy. Complying with these instructions does not relate to achieving output (performance) target but exerting the input (effort) for achieving output; even if the agent complies with these instructions, the agent necessarily achieves output target because of measurement error or environmental uncertainty.

5. Even while feeling stress, it might be rational for an agent to engage in behaviour that deviates from expectations. For example, Kaplow and Shavell (1994) demonstrated that it is rational for people to sometimes confess their own ‘illegal’ behaviour even if they might be punished for it. In terms of accounting control, agents select actions depending on the benefits he/she might derive from performance evaluations and the costs entailed in performing the task and deviating from the targets.

6. According to Milgrom and Roberts (1992) and Zimmerman (2016), influence activities are self-interested and intended to impact the decision-making of others through lobbying activities and intra-organisation politics, whereas influence costs are the expenses incurred by undertaking influence activities, causing decision-making to be inefficient and, thus, preventing influence activities.

7. This is calculated by the author based on the chart in Reference 2 (Institute of Labor Administration 2016, 30).

8. Of the 35.8% companies adopting the performance-linked bonus system, 87.2% adopted it for all employees, whereas 12.8% adopted it only for managers or employees in senior positions (Institute of Labor Administration 2010, 81). The subjects comprised 4,990 companies from 3,585 listed companies on nationwide securities markets and 1,405 non-listed companies. The data were compiled from the 283 companies that responded to the survey from among the survey subjects (Institute of Labor Administration 2010, 10). Here, a performance-linked bonus system is, for example, a system to determine the funds for bonuses according to the results of the organization, of either a department or the entire company, based on a predetermined formula, such as ‘The funds for the bonuses for the performance-linked part of the company’s compensation shall be 7% of the operating profit’. It excludes the allocation of bonuses that were simply from the appraisals (assessments) of individual personnel (Institute of Labor Administration 2010, 80).

9. This assumption is consistent with Feltham and Xie (1994).

10. This model setting can be seen in various papers, such as in Akerlof and Kranton (2005), Bruggen and Moers (2007), Fischer and Huddart (2008), Heinle, Hofmann, and Kunz (2012) and Stevens and Thevaranjan (2010), which used a model assuming that costs are incurred when the agent takes actions that deviate from the identity and norms.

Additional information

Funding

This work was supported by the JSPS KAKENHI [18K12904].