Accounting Conservatism and Incentives: Intertemporal Considerations

63 Pages Posted: 16 Aug 2016

See all articles by Jonathan C. Glover

Jonathan C. Glover

Columbia Business School

Haijin Lin

University of Houston

Multiple version iconThere are 2 versions of this paper

Date Written: August 13, 2016

Abstract

We study the intertemporal properties of conditional accounting conservatism, focusing on its role in determining the timing of payouts to managers. In our first model, conservative accounting introduces downward bias; any understatement of performance is reversed in a second period. If a new agent is hired in the second period, conservative (aggressive) bias rewards (penalizes) the second-period agent for part of the first-period agent’s performance and is costly because of the performance measurement contamination. If the same agent is retained in both periods, a conservative bias is not costly because any first-period bonus underpayment is self-corrected in the second period; an aggressive bias is costly because first-period bonus overpayments cannot be fully “clawed back” in the second period as the agent is subject to limited liability. With performance-contingent firing, a conservative bias may be costlier than an aggressive bias because the agent is fired in the circumstance in which performance may be misstated under a conservative bias but is correctly stated under an aggressive bias. If the same agent is retained in both periods and the principal learns about the agent’s productivity over time, a conservative bias strictly dominates not only an aggressive bias but also unbiased accounting. A conservative bias allows the principal to effectively retroactively fine-tune the first-period compensation using the principal’s learning by applying the second-period bonus rate to good performance that was underreported in the first period and reversed in the second period.

In the second model, conservative accounting is a choice to recognize effort-independent bad news early and good news late (e.g., price level adjustments). If the same agent is retained in both periods, accounting choice is irrelevant because any rent reduction (increase) in the first period is offset by an increased (reduced) rent in the second period. With performance-contingent firing, conservative accounting dominates aggressive and neutral accounting because recognizing bad news shocks early and good news shocks late maximizes the agent’s retention rent, which reduces the cost of providing first-period effort incentives. In the news model, there is no benefit to retaining the same agent in order to wait for a reversal of an understatement of first-period performance. In some cases, it is optimal to recognize both good news and bad news in the same period, but the optimal system continues to be a conservative one in that news is recognized in the first (second) period when the aggregate shock is negative (positive). In an extension, we study overlapping projects that give rise to a multi-task setting. An interior accounting system (neither maximally conservative nor maximally aggressive) can be optimal because extreme systems make one of the over-lapping projects an incentive bottleneck.

Keywords: accounting conservatism, multi-period incentives, incentive spillover, multi-task incentives

JEL Classification: D86, D90

Suggested Citation

Glover, Jonathan C. and Lin, Haijin, Accounting Conservatism and Incentives: Intertemporal Considerations (August 13, 2016). Columbia Business School Research Paper No. 16-56; AAA 2017 Management Accounting Section (MAS) Meeting. Available at SSRN: https://ssrn.com/abstract=2822976 or http://dx.doi.org/10.2139/ssrn.2822976

Jonathan C. Glover

Columbia Business School ( email )

3022 Broadway
New York, NY 10027
United States
212-854-1911 (Phone)

Haijin Lin (Contact Author)

University of Houston ( email )

390F Melcher Hall
Bauer College of Business
Houston, TX 77204-6021
United States
7137437771 (Phone)

Register to save articles to
your library

Register

Paper statistics

Downloads
229
rank
56,117
Abstract Views
1,192
PlumX Metrics