STRATEGIC SHIRKING: A THEORETICAL ANALYSIS OF MULTITASKING AND SPECIALIZATION

• Date: 01 May 2016
• DOI: http://doi.org/10.1111/iere.12166
• Published date: 01 May 2016
• Author: Oliver Gürtler,Jed DeVaro

INTERNATIONAL ECONOMIC REVIEW

Vol. 57, No. 2, May 2016

STRATEGIC SHIRKING: A THEORETICAL ANALYSIS OF MULTITASKING

AND SPECIALIZATION∗

BYJED DEVARO AND OLIVER G¨

URTLER1

California State University, East Bay, U.S.A.; University of Cologne, Germany

We provide a new theory to explain why firms multitask workers instead of specializing them. Workers overperform

in tasks they like and underperform in tasks they dislike to favorably influence future job assignments. Anticipating

this, firms may find it optimal to commit to future multitasking to induce workers to appropriately allocate effort

early in the employment relationship. We show that when the product market is volatile, so that future product prices

are uncertain, the firm’s ability to credibly commit to a multitasking strategy diminishes. This generates a negative

relationship between multitasking and product market volatility, consistent with recent empirical evidence.

1. INTRODUCTION

A fundamental question in the area of job design is whether employers should specialize

workers in certain tasks or whether workers should multitask, meaning they are assigned to

work on multiple tasks.2We present a new theory of multitasking and specialization based on

an aspect of worker behavior that the job design literature has overlooked.

Our theoretical argument has five steps. First is the idea—well recognized since Adam

Smith—that there is a productivity advantage to specialization. Second is the notion that,

over time, an employer learns about the productivities of its workers on each task to which they

are assigned. These observations suggest that the firm should multitask its workers early in their

careers to learn who is good at what. Later in the career, the firm should specialize workers on

the basis of their demonstrated aptitudes on each task. Third, when assigned to multiple tasks

the worker has more than one path to retention. When paid a fixed wage on a one-task job, the

worker must perform the task well enough to avoid getting fired, and there is therefore little

opportunity to manipulate the information conveyed to the employer about task-specific ability.

But if the worker is assigned to multiple tasks, better performance in one can compensate for

worse performance in another. Fourth, workers will allocate effort strategically across those

tasks so as to increase the likelihood of getting assigned to their preferred tasks in the future, i.e.,

the worker underperforms (or “strategically shirks”) on the less preferred tasks and overper-

forms on the preferred ones. Fifth, mindful of such shenanigans, the firm may find it optimal to

commit at the outset of the employment relationship to long-term multitasking. This improves

workers’ incentives early in the career by reducing their expected gains from distorting their

effort across tasks, though at the expense of less efficient future job assignments, given that

workers must be assigned to both tasks even if their earlier performance suggests they should

∗Manuscript received November 2013; revised January 2015.

1We thank two anonymous referees, Hanming Fang (the editor), Martin Farnham, Hideshi Itoh, Fidan Ana Kurtulus,

Hodaka Morita, Jaime Ortega, Suraj Prasad, Anja Sch¨

ottner, Dirk Sliwka, and Mike Waldman for helpful comments.

Please address correspondence to: Jed DeVaro, Department of Economics—College of Business and Economics,

California State University, East Bay, 25800 Carlos Bee Blvd, Hayward, CA 94542. Phone: +1 510 786-7079. Fax: +1

510 885-2660. E-mail: jed.devaro@csueastbay.edu.

2Apart from multitasking, a number of other related terms are used almost interchangeably in the literature (e.g.,

multiskilling, job rotation, cross training, job enrichment, and job enlargement) despite having slightly different mean-

ings (Owan, 2011).

507

C

(2016) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social

and Economic Research Association

508 DEVARO AND G ¨

URTLER

specialize. Multitasking can then be the optimal job design, even after the firm learns a lot about

workers’ task-specific productivities.

The following three examples illustrate strategic shirking behavior. First, consider a fast food

restaurant with two jobs (cashier and cook). On his first day on the job, a worker observes that

the more senior workers are specialized, with some assigned to the cash register and others to

the grill. During the first week, the store manager rotates the new worker between both jobs,

giving him experience on both tasks. Despite being a good cook, the worker prefers to work

the cash register. Realizing that the manager tends to specialize experienced workers, the new

worker is purposely a laggard when assigned to the grill, but when assigned to the cash register

the worker performs exemplarily. Before long, the worker is assigned to the register almost

exclusively.

Second, consider an economics graduate student who works on a team of research assistants

for a professor. The student observes early on that the more senior assistants are specialized

on either of two tasks (SAS programming and editorial work). Despite being a good SAS

programmer the student finds reading and writing more interesting. The student also cares

about impressing the professor, so if he were assigned to only one task he would have to

work hard. But when assigned to both tasks, the student purposely underperforms on the

SAS programming tasks while exceeding all expectations on the writing tasks. Before long the

student is specialized in editorial work.

Third, professors who exhibit poor performance in the areas of service and teaching are

sometimes rewarded with attractive future job assignments. They are not asked to serve on

the most onerous committees because they have successfully created the impression that they

are disorganized, absentminded, and unreliable. Similarly, “for the good of the students,”

professors who teach poorly are rewarded with smaller classes and a smaller range of classes

to teach, allowing them to specialize in the preferred task of research. In some departments,

rules that require an equitable division of teaching and service burdens (i.e., a commitment

to ongoing multitasking) exist and mute the professors’ incentives to strategically shirk. Our

theory demonstrates the potential optimality of such rules.

These examples are all drawn from direct observation and illustrate that strategic shirking is

practically relevant across a range of different worker skill levels and job types.3The first two

examples are drawn from the personal employment history of one of the authors; in both jobs

he purposely allocated effort across tasks in the manner described in our model.4The third is

based on our observations of academic department politics in Germany and the United States

and on our discussions with colleagues throughout the profession. In each example, the worker

can strategically shirk on one task because great performance on the other task proves his future

value to the employer. The worker is paid a fixed wage (either hourly in the first two examples

or a salary in the third) and must demonstrate on at least some dimension that his labor merits

that level of pay.

We develop a dynamic theoretical model that formalizes the preceding ideas. In the two-

period model, two workers are hired into a job with two tasks that must be performed in each

period. Both workers find one of the tasks distasteful. In each period, the workers are paid a

fixed wage, the employer must allocate their time between tasks, and both workers must decide

how much (costly) effort to exert on each task to which they are assigned. At the end of the first

period, the employer observes performance signals for both workers on each task to which they

were assigned. In the second period, the employer allocates the workers’ time across the tasks.

3Other instances of strategic shirking can be found in sports and the performing arts. In efforts to experiment with

alternative assignments that might enhance productivity, athletic coaches and orchestral conductors sometimes rotate

players to new positions. Players might deliberately underperform to avoid making the new assignments permanent,

though they must perform well in their favorite positions to avoid being cut from the team or orchestra.

4In the first example, the worker was 16 years old and had not yet been introduced to economics, illustrating that

the motivations for worker behavior in the model are quite basic and intuitive. In the second example, the professor

explicitly stated that he likes to assign new research assistants to a range of tasks so as to learn their strengths and

weaknesses.