European Journal of Operational Research 207 (2010) 878885

Contents lists available at ScienceDirect

European Journal of Operational Research

journal homepage: www.elsevier.com/locate/ejor

Decision Support

Incentives and individual motivation in supervised work groups

Arianna Dal Forno, Ugo Merlone *
Department of Statistics and Applied Mathematics, University of Torino, Corso Unione Sovietica 218 bis, Torino I-10134, Italy

a r t i c l e

i n f o

a b s t r a c t
This paper introduces and analyzes a model of supervised work group where subordinates decide how to exert their effort in complementary tasks while the supervisors decide incentives. Incentives may be a combination of individual and group-based ones. The optimality of incentives is analyzed when considering two different cost functions for subordinates. The two cost functions describe different individual motivations; comparing the resulting effort allocations and production optimality, we can relate them to different organizational theories. Our results provide a measure of how motivation among subordinates may affect production and incentives. Furthermore, the optimal incentives schemes are examined in terms of Adams equity theory. 2010 Elsevier B.V. All rights reserved.

Article history: Received 23 April 2008 Accepted 17 May 2010 Available online 25 May 2010 Keywords: Organization theory Production Organizational behavior Incentives Individual motivation

1. Introduction According to Boudreau et al. (2003), the elds of operations management and human resources management are intimately related, yet they maintain distinct perspectives. Among the examples which try to merge these perspectives, Bordoloi and Matsuo (2001) applied control theory to deal with workforce planning taking into account also worker learning and controls for the risk. Another example is Gendron (2005), where a store scheduling problem with constraints deriving from union representatives and human resources personnels was approached and solved. Recently, Boudreau et al. (2003) examine how human considerations affect classical operation management. In their conclusion, they highlight the research challenges and opportunities of bringing the human resources and operation management together. In this paper, we explore this line of research and try to integrate human considerations in optimal incentive problems. The Moral Hazard literature approaches multi-agent relationships in different ways. For example, the joint production models provide interesting insights in terms of income distribution among the agents, see for instance Alchian and Demsetz (1972) and Holmstrm (1982). Another relevant aspect is the comparison between centralized and decentralized structures as far as contracting goes. For example, the literature provides conditions under which the delegation of the supervisory task, i.e. decentralizing,

* Corresponding author. Tel.: +39 011 6705753; fax: +39 011 6705783. E-mail address: merlone@econ.unito.it (U. Merlone). 0377-2217/$ - see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.ejor.2010.05.023

is benecial; to obtain a rst analysis of the advantages and disadvantages of delegation the reader may refer to Macho-Stadler and Prez-Castrillo (1997). One aspect usually neglected in the economic literature is the role of individual motivation; while, in psychology, motivation is a concept that has been discussed extensively. According to Spector (2003) work motivation theories are most typically concerned with the reasons, rather than ability, that some people perform their jobs better than others. Steers and Black list the stages the evolution of management thought concerning employee motivation has passed through. They are the traditional model, the human relations approach and, more recently, the human resources model. In particular, this newer approach also assumes that different employees want different rewards from their jobs, that many employees sincerely want to contribute, and that employees by and large have the capacity to exercise a great deal of self-direction and self-control at work (Steers and Black, 1994, p. 139). To this extent the case of Kyocera Corporation is striking. As it concerns the reward system, Kyoceras founder Kazuo Inamori, in a booklet describing his philosophy, writes We dont think in terms of individual rewards. We dont buy individuals loyalty with monetary incentives or titles. Rather, we believe that individuals who are endowed with superior capabilities should contribute their capabilities for the good of the entire group. (for a discussion of Kyoceras organizational culture the reader may refer to Bylinsky (1990)). This example is contrasted by pay incentives used at Lincoln Electric, where most employees are paid on a piece-rate system (the Lincoln Electric Company has been described in several case studies by business schools, see for instance Fast and Berg (1975)). Several comparisons between incentive programs have

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

879

been presented in the literature. Among others, Weiss (1987) provides an empirical comparison between individual wage incentives and group incentives examining the effects in terms of motivation and quit rates. Following the joint production approach, we consider a modular model of hierarchical organization. Specically, we concentrate mainly on pyramidal structures. This particular structure is widespread and, consequently, both the economic (see Beckmann, 1988, for a formal analysis) and simulative literature (for instance, see Glance and Huberman, 1994) nd interest (for an analysis of the different approaches to pyramidal structures see Merlone, submitted for publication). In our model, the organization consists of two heterogeneous agents interacting in a supervised work group with a CobbDouglas production function. In the literature, the distinction between team and work group is hazy, nevertheless we will follow Spector (2003). As a consequence, we will refer to a supervised work group as, in our case, we consider interchangeable subordinates. In this paper, we assume that individuals with different motivation may have different cost functions and analyze the incentive problem the supervisor faces, in order to maximize her own prot. In particular, we analyze how the optimal solution for the supervisor varies when considering two different cost functions for subordinates. Cost function plays an important role for the agent. While in Dal Forno and Merlone (2007) only piecewise constant cost function was considered, in this paper we also consider strictly convex costs. Usually economics assumes that costs are increasing and marginally increasing; the other kind of cost function in some cases may model more realistically some situations. We refer in particular to situations like those described in Smith (1977) where employees expend their discretionary effort (for a discussion about antecedents of discretionary effort and its consequences on performance the reader may refer to Bailey et al. (2001) and Sutton (2007)). In this sense, the two different cost functions can be interpreted in terms of motivation; while the piecewise constant cost function may be appropriate when subordinates have high self-efcacy and are highly motivated, the other cost function seems to be more appropriate for individuals who are mainly interested in monetary incentives. It is well known (Zhou, 2002) that, with regard to the principal-agent theory, in general it is difcult to derive analytical solutions; therefore, even in scholarly contributions, strong assumptions are usually required. This tradeoff between analytical tractability and extensive simplication is acknowledged by other authors approaching organization complexity (see Ethiraj and Levinthal, 2009). Nevertheless, the two different cost functions we analyze can be related to different cultures in the organizational structure we consider. As a consequence, by comparing the production outputs under the two different cost function assumptions, our analysis allows us to compare productivity under different organizational cultures and to measure what the cost of the culture in terms of production is. The two different cost functions, and the related organizational cultures, are particularly interesting when considering changes in competition induced by emergent countries such as China (for a discussion on the role of culture in the economic style of China, the reader may refer to Herrmann-Pillath (2005) and to Lum (2003), for an analysis of labor conditions in the same country). The structure of the paper is the following. In Section 2, we present the theoretical model. Sections 3 and 4 summarize and analyze the optimal incentive problem with the two different cost functions in question, describing how these are related to the subordinates motivation. In Section 5, optimal incentive schemes and outputs are compared and the results are interpreted in terms of organizational culture consequences on productivity. Finally, Section 6 is devoted to conclusions and further research.

2. The model As in Dal Forno and Merlone (2007) and Dal Forno and Merlone (2009), we consider a model of supervised work group in which a supervisor (acting as principal) and two subordinates (acting as agents) cooperate. Agent i allocates his effort li with the partner, and the effort ui with the supervisor. The joint production function for agents 1 and 2 is C(u1 + u2)a(l1 + l2)b, where C 2 R is a constant factor,1 and a, b 2 (0, 1) are, respectively, the output elasticity with respect to the joint effort with the supervisor and with the partner. As a consequence, the agents have to decide both how much effort to exert, and how to partition it in the two complementary tasks.2 Agents bear a cost for effort: agent is cost function ci : R2 ! R will be denoted with ci(ui, li); cost functions are private information. Furthermore, each agent can observe the level effort his partner provides with him, but not the one which is provided with the supervisor. Conversely, the supervisor can only observe the joint output and the effort each agent provides with her. The supervisors prot is a share c 2 (0, 1) of the supervised work group production minus the incentives she pays to her subordinates. In the following, we assume that the output is sold on market at unitary price and the production and sharing constants C and c are such that Cc = 1; this is not restrictive, it simplies the notation, and allows us to simply consider monetary payoffs. Finally, agents retribution consists of a xed wage w > 0 plus a performance-contingent reward; we assume that the xed wage is sufcient to meet basic needs, in terms of the hierarchy of needs theory (Maslow, 1970), physiological needs and needs of safety; in economic terms we say that the participation constraint is met. Proposition 1. The gross production (u1 + u2)a(l1 + l2)b is maximized if and only if the aggregate efforts are allocated proportionally to the output elasticities.

Proof. The result follows from combining and rearranging the rst order conditions of the problem
u1 ;u2 ;l1 ;l2

max u1 u2 a l1 l2 b :

In fact, from

au1 u2 a1 l1 l2 b 0;
bu1 u2 a l1 l2 b1 0;

it follows

u1 u2 a : l1 l2 b

Condition (3) is necessary in order to maximize the production of the supervised group. When either as the result of misaligned incentives or as lack of coordination between subordinates this condition is not met, then the effort allocation is not efcient. The performance-contingent reward is a linear incentive bt on the joint output of the team and a linear incentive bi on the effort each agent exerts with the supervisor. Therefore, the problem can be formalized as a bilevel programming problem:
bt ;b1 ;b2

max 1 2bt u1 u2 a l1 l2 b b1 u1 b2 u2 ;

such that, given the incentives bt, b1, b2, subordinates solve

max
u1 ;l1 u2 ;l2
1 2

w bt u1 u2 a l1 l2 b b1 u1 c1 u1 ; l1 ; w bt u1 u2 a l1 l2 b b2 u2 c2 u2 ; l2 ;

max

We recall that R is the set of positive real numbers; the case C = 0 is trivial. From the functional form of the production function it is immediate to observe that the two tasks are not additive; for a discussion the reader may refer to Spector (2003).

880

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

where we have dened:          i: index of subordinate, i = 1, 2; ui: effort subordinate i exerts in the task with the supervisor; li: effort subordinate i exerts in the task with the colleague; a: output elasticity with respect to the joint effort with the supervisor; b: output elasticity with respect to the joint effort between subordinate; ci(ui, li): cost function for subordinate i; w: xed wage; bt: linear incentive on the joint output of the team; bi: linear incentive on the effort subordinate i exerts with the supervisor.

out goldbricking. This kind of cost function is not completely new to the principal-agent literature; for instance, Holmstrm and Milgrom (1991) assume that incentives are necessary to encourage agents to work beyond a limit they would take pleasure in working. In our case we assume that this limit coincides with their physical capacity. Finally, observe that this assumption can also be interpreted in terms of the self-efcacy theory (Bandura, 1982), assuming that both subordinates have high self-efcacy and are motivated to put in as much effort as they can. 3.1. The agents problem Starting from the subordinates information set and assuming coordination and commitment between agents, problem (5) reduces to:
u1 ;u2 ;l1 ;l2

For the sake of simplicity we assume w = 0, this is not restrictive. Examining the form of the problem, it is rather immediate to predict the behavior of rational supervisor and subordinates in this interaction situation. It is a nite dynamic game with perfect information, with supervisor moving rst, and then subordinates, acting simultaneously after observing the incentive. This game has a proper subgame starting from the information set of the subordinates. Therefore, there exists a set of subgame perfect Nash equilibria which equals the set of Nash equilibria that can be derived by backward induction. In fact, given any feasible incentives scheme (bt, b1, b2), the subordinates will play a Nash equilibrium3 (ui, li) of the subgame. Knowing the fact that subordinates will play a Nash equilibrium in the subgame, the supervisor will maximize her prot by choosing the optimal incentive scheme. In the following sections, we will analyze how different assumptions on the subordinates cost functions may determine the optimal incentive scheme. 3. Piecewise constant costs In the literature, some factor or factors that keep agents from working innitely hard are usually considered. For example, Wageman and Baker (1997) propose several mechanisms considering effort becoming increasingly either unproductive or unpleasant. In this paper, as common, we assume that the unpleasantness of the work increases with respect to the effort. Different cost functions may be considered and the functional form may reect different underlying assumptions. While economics usually considers mainly rational agents, other approaches to work group dynamics take into account other aspects, such as norms and team commitment. Industrial and organizational psychology have proposed different theories of motivation to explain individual behavior in the organization (for a rst survey the reader may refer to Spector (2003)). Nevertheless, several other aspects underlie group and team behavior; the role of norms is well documented empirically (see Coch and French, 1948 and Roy, 1952, for example) and it is commonly assumed, for work groups, to dictate how much each person will produce (see Spector, 2003). A rst cost function we consider is, as in Dal Forno and Merlone (2007), the following:

max u1 u2 a l1 l2 b

sub ui li 6 i c

for all i = 1, 2. Corollary 2. Consider the production function (u1 + u2)a(l1 + l2)b, with piecewise constant cost functions (6). Then, any effort allocation such that

a c u1 u2 ab 1 2 ; c
b l1 l2 ab 1 2 ; c c

maximizes the production. Proof. The result follows trivially from rst order conditions of problem (7), assuming the constraint on individual aggregate effort given by capacities. h Therefore, there is a continuum of solutions to the considered problem (7), nevertheless, a rather natural effort allocation is the one that can be interpreted as focal in the sense of Schelling (1960):

ui ; li

a  b  c; c ab i ab i

for all i = 1, 2. Effort allocations such that (8) holds, will be called efcient, in fact they satisfy also condition (3) which is necessary to the production maximization. 3.2. The supervisors problem The supervisors problem is to design a linear compensation scheme for the subordinates that induces them to use their capacity to maximize the team output. The supervisor can observe the efforts ui the subordinates exert with her and the team output. Each subordinates compensation is si = biui + bt(u1 + u2)a(l1 + l2)b, where, we recall, bi is the incentive given to subordinate i for his individual effort with supervisor, and bt is the incentive given to them for the team output. We assume that the supervisor declares the incentives and then the subordinates decide their efforts in order to maximize their wage. The supervisor has to solve the following problem:
bt ;b1 ;b2

& ci ui ; li

c 0 if ui li 6 i ; c 1 if ui li > i :

First observe that this cost function is non-decreasing with respect to the aggregate effort. It assumes that each subordinate has a physical capacity i under which effort has zero cost, or, alternatively, c that at some exertion level the effort becomes unpleasant enough to lead the individual to conclude that it is not worth working any harder independently of the reward. In this case, we assume that each individual knows his individual capacity and uses it with3

max 1 2bt u1 u2 a l1 l2 b b1 u1 b2 u2 :

10

When considering fully rational agents the solution is obvious. Since any individual incentive given to agents gives a suboptimal effort allocation, and null team output incentive makes for subordinates any allocation optimal, the optimal solution is

In the next sections, we will examine also the equilibrium selection problem.

8 > bt e > 0; < b 0; > 1 : b2 0:

11

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

881

Some comments about this incentive scheme are in order. While it seems to be appropriate for agents with the same capacities or a low number of repetitions, for agents with different capacities interacting over a long period of time some problems may arise. For example, when considering equity theory (Adams, 1965), it is rather evident that different capacity individuals will nd themselves in inequitable situations, and will experience dissatisfaction and emotional tension, that they will be motivated to reduce. In order to improve the distributive justice (see Spector, 2003, for a discussion about equity theory and distributive justice) a different incentive scheme may be more appropriate in the long run. 4. Strictly convex costs Following Holmstrm and Milgrom (1991), we suppose the effort in the two tasks is perfectly substitutable in the agents cost function. Formally we assume that

Lemma 3. Given an incentive scheme (bt, b1, b2), if subordinates cost functions have form (12), then an optimal effort allocation u1 ; l1 ; u2 ; l2 must be such that

c01 u l1 c02 u l2 : 1 2

16

Furthermore, optimal incentives must be such that individual incentives are identical, that is

b1 b2 :

17

Proof. Since we assume that the effort allocation constitutes a Nash equilibrium, rst order conditions can be derived from (14) and (15); they are:

ci ui ; li ci ui li :

12

8 a1 b l1 l2 b1 c01 u l1 0; > abt u u 1 2 1 > > > a b1 > < bbt u u l l c01 u1 l1 0; 1 2 1 2 a1 b > ab u u > l1 l2 b2 c02 u2 l2 0; > t 1 2 > > a : b1 c02 u l2 0: bbt u u l1 l2 1 2 2

18

As it is common, ci is assumed differentiable, increasing and marginally increasing. As mentioned above, while this cost function is closer to the classical modeling of behavior of the rational individual, we may assume this kind of cost function when agents motivation consists only in economic reward and neither social norms nor other motivational theories determine individual effort. It is not difcult to assume that this kind of cost function may be realistic in organizations where individual objective consists only in the maximization of the individual payoff. Another interpretation of this cost function is as the result of high-powered incentives. According to Steers and Black (1994), individual incentives may, at times, lead to employees competing with one another; in our case we assume that the internal competition, driven by high-powered incentives, may result as this individual perception of cost of effort (for a discussion on high-powered incentives, see Encinosa et al., 2007). As a consequence, we can think that this kind of cost function is the result of a different culture when compared to the piecewise constant case. In this case also, we rst analyze the agents problem and then the supervisors one. Finally, we observe that some of the results we present in the following can be generalized to more general cost functions. 4.1. The agents problem In this case, it would be rather unrealistic to assume commitment and coordination between subordinates, rather we will assume that agents solve problem (5) with cost function (12)

Combining the second and the fourth equation it is immediate to obtain

c01 u l1 c02 u l2 : 1 2

19

For the second part, combining the rst and the third equation in (18) it is immediate to obtain

b1 c01 u l1 b2 c02 u l2 : 1 2
Finally, by Eq. (19) we obtain

20

b1 b2 :

21

Following this result, incentives schemes will be written using the shortened notation (bt, b1, b2) = (bt, b). As mentioned in Section 1, it is well known that principalagents models often require strong assumptions to derive analytical results. This model is not an exception; in the following, as it is common (see for instance, Holmstrm and Milgrom, 1987; Schattler and Sung, 1997; Encinosa et al., 2007), we assume a quadratic cost function

ci ui ; li di ui li 2 ;

22

where di, i = 1, 2 are positive constants. We also assume a + b = 1. In this case, a result concerning proportionality of allocated efforts can be proved, as well. Theorem 4. Given an incentive scheme (bt, b), with b 0, there exists a unique equilibrium effort allocation u ; l1 ; u ; l2 such that 1 2

max
u1 ;l1 u2 ;l2

bt u1 u2 l1 l2 b1 u1 c1 u1 l1 ; bt u1 u2 l1 l2 b2 u2 c2 u2 l2 :
a
b

max

13

u q u l1 qk1 ; l1 1 q u l1 1 qk1 ; 1 1 1 u2 q u2 l2 qk2 ; l2 1 q u2 l2 1 qk2 ;

k1

23
As a consequence, given the incentives (bt, b1, b2), subordinates nd an effort allocation which is a Nash equilibrium, i.e., u ; l1 ; u ; l2 1 2 must be such that where : u 1
l1 ; k2

u 2

l2

and 0 < q < 1.

u ; l1 1

arg max bt u1

a u l1 2

b l2

b1 u1 c1 u1 l1

14

Proof. First we will show that such an effort allocation is an equilibrium allocation. Since the objective functions are concave, rst order conditions are also sufcient and the optimal solution must satisfy (18):

and

a b u ; l2 arg max bt u u2 l1 l2 b2 u2 c2 u2 l2 : 2 1

15

While for the piecewise constant cost function it was immediate to derive the effort allocation, in this case some preliminary considerations are in order.

8 > abt qa1 k k a1 1 q1a k k 1a b 2d1 k 0; > 1 2 1 2 1 > > a > < 1 ab qa k k 1 qa k k a 2d k 0; t 1 1 1 2 1 2 > ab qa1 k k a1 1 q1a k k 1a b 2d k 0; > 2 2 > t 1 2 1 2 > > : a a 2d2 k2 0: 1 abt qa k1 k2 1 qa k1 k2 24

882

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

Combining the second and the fourth equation it is immediate to obtain d1 k1 d2 k2 , that is

k2

d1 k ; d2 1

25

8 > abt u1 u2 a1 l1 l2 1a 2d1 u1 l1 0; > > > < bb u u a l l a 2d u l 0 t 1 2 1 2 1 1 1 > abt u1 u2 a1 l1 l2 1a 2d2 u2 l2 0; > > > : bbt u1 u2 a l1 l2 a 2d2 u2 l2 0:

33

so we can discard the third equation as redundant. By simple algebra we obtain

Combining the rst and the second equation of (33) it is immediate to obtain

8 > abt qa1 1 q1a b 2d1 k 0; > 1 < 1 abt qa 1 qa 2d1 k1 0; > d > :k 1 k : 2 d2 1
Putting together the rst two equations and rearranging

u1 u2 26

a
1a

l1 l2 ;

34

which, substituted in the rst equation of (33), yields

k1 u1 l1

aa bt ; 2d1 1 aa1 aa bt : 2d2 1 aa1

35

analogously, we nd

8 a q1a > abt 11a 1 abt q a b 0; > q 1q > < a qa k 1abt2d11q ; > 1 > > d1 :k k :
2 d2 1

k2 u2 l2 27

36

The focal allocation can be found, as before, as

(
1 abt 1qa b; it is
qa

u ak1 ; l1 1 ak1 ; 1 u 2 a
k2 ; l2

Now consider function f q abt

1qa

q1a

1 ak2 :

37

easy to prove that there exists a unique q* 2 ]0, 1[ such that

abt

1 q 1a

q1a

qa 1 abt b 0; 1 q a

28

since function f is continuous in ]0, 1[. Furthermore, it holds

8 a q1a > lim abt 11a 1 abt q 1b b 1; < q!0 q 1q > lim ab 1q1a 1 ab : t t 1a
q!1 q
1qa

qa

29

Remarkably, effort allocation (37) is the analogous of the focal allocation (9) we found for piecewise constant costs in Corollary 2. In this case also, given the aggregate efforts the subordinates put forth, the production is maximized. Furthermore, we observe that q denes what fraction of the agents efforts is made with the supervisor. By the implicit function theorem we can prove that, as expected, this fraction increases as b increases. In fact, a greater individual incentive reallocates a greater effort with the supervisor. Corollary 6. The fraction q of effort made with the supervisor is increasing with respect to the individual incentive b and its derivative is

b 1;

and

f 0 q abt b1q bbt

b1 1a q 1aqa 1qb

q1a 2

30

a1q1b qa1 1bqa 1qb

1q1b 2

< 0:

q0 b

q2a 1 q1a : a1 abt

38

Thus, since, as b 0, q* is unique, there exists the unique effort allocation

Proof. Eq. (28) can be written as

8 a1 a 1a > u 1abt q a ; l 1abt q 1q ; > 1 1 > 2d1 2d1 1q > < a1 a 1abt q 1abt q 1q 1a ; 2d2 > u2 2d2 1q a ; l2 > > a > 1abt q : k 1abt qa ; k2 2d 1q a : 1 2d 1q a
1 2

f q; b 0;
where

39 1 q1a

31

f q; b : abt

1a

1 abt

qa b; 1 qa

40

Theorem 4 does not claim that, with positive individual incentives, all effort allocation equilibria are characterized by efforts being proportionally allocated between the two tasks. Rather, it claims that among the proportional effort allocations, i.e. those satisfying condition (3), one and only one is an equilibrium. In the following corollary, we analyze the effort allocation with null individual incentives. Corollary 5. Given an incentive scheme (bt,0), if subordinates cost functions have form (22), then any allocation such that ak1 aa bt 1 ak2 is an equilibrium allocation, where k1 : ; 21aa1 d1 aa bt k2 : a1 . Furthermore, there exists a focal effort allocation

implicitly denes q as a function of b. By the implicit function theorem

q0 b

of =ob of =oq

1
a1abt q2a 1q1a

q2a 1 q1a : a1 abt

41

As both 0 < a < 1 and 0 < q < 1 the thesis follows. h This result proves that when individual incentives are both positive, the effort allocation no longer satises condition (3) and has several implications when considering how the supervisor must determine incentives optimally. 4.2. The supervisors problem As for the piecewise constant function, the supervisors problem is to determine the incentive scheme maximizing her own payoff as in (10). Before investigating whether individual incentive b increases supervisors prot, we characterize the optimal incentives when b = 0. To this extent, we consider the aggregate prot

8 a1 aa bt < u a abt ; l ; 1 1 21a 1 d 21aa2 d

1 1

21a

d2

aa bt : u aa1 bt ; l : 2 2 21aa1 d 21aa2 d

2 2

32

Proof. Under the assumptions b = 0 and a + b = 1, conditions (18) become

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

883

u u 1 2

a 1a 2 2 l1 l2 d1 u l1 d2 u l2 : 1 2

42

i 1 abt d1 d2 h 1 2bt q2a 1 q12a bqa1 1 qa : 2 d1 d2 50

That is, the supervisors prot is proportional to

Theorem 7. With incentive schemes (bt, b) where b = 0, the aggregate prot (42) is maximized if and only if it is completely shared between the two agents; by contrast, in order to maximize supervisors prot, half of the aggregate prot must be given to the supervisor and the remaining equally shared between the agents. Proof. To maximize the aggregate prot the following program must be solved:

1 2bt q2a 1 q12a bqa1 1 qa :

51

Now recall that in Corollary 6 we proved that q is a function of b. As a consequence, we can consider the supervisors prot4 as a function of the individual incentive b

max
bt 61 2

u 1

a u l1 2

1a l2

d1 u 1

2 l1

d2 u 2

2 l2 :

pb 1 2bt q2a b1 qb12a bqa1 b1 qba :

52
Now derive p with respect to b using expression (38) we found in Corollary 6, substitute b = 0 and recall that q(0) = a,

43

Taking into account the focal effort allocation, as given by Eq. (32), the program (43) becomes

aa1 bt d1 d2 max bt 61 21 aa1 d1 d2 2

d1
that is

"

#a "

aa bt d1 d2 a2 d d 1 2 21 a aa b
!2
t

#1a

p0 0 a

1 2bt a 1 2bt aa1 a 1 aa a : bt bt 1 aa

53

aa b

!2 d2

21 aa1 d1

21 aa1 d2 "

44

Since bt < 1/2, the supervisors prot increases as b increases. As a consequence, incentive schemes (bt, b) where b = 0 cannot be optimal. h The consequences of this theorem are striking. In fact, this result proves that, with quadratic costs (22), the individual incentives are necessary. Nevertheless, by Corollary 6, it follows that individual incentives increase the fraction of effort made with the supervisor. Therefore, with positive individual incentives the effort allocation does not satisfy condition (3), and gross production is not maximized. In the following section, we will relate the assumption underlying the different cost functions we have considered to the managerial literature. 5. Comparing the two cost functions In Section 3, we found that the optimal incentive scheme resulted in null individual incentives, and that any efcient allocation (8), maximizes the production. Furthermore, in Section 4 we found that, also with quadratic cost functions, when individual incentives are null the resultant effort allocation (32) is optimal. Putting together these results with Theorem 8 thesis, it results that given the aggregate efforts exerted by subordinates, the production is not maximized. In other words, assuming that the resulting capacities k1 and k2 for subordinates with quadratic cost functions are the actual capacity constraints for piecewise constant cost functions, then the production would be higher since in this case the effort allocation would be efcient. Note that in this case we consider gross production because, otherwise, the different incentives for quadratic cost functions would make a comparison between net productions trivial. This result can be interpreted as a further example of systems that pay off for one behavior even though the rewarder hopes dearly for another (Kerr, 1975); the reason here is that the members of the work group possess different goals and motivation. The contrast is even starker when we assume that the whole incentive mechanism was designed by top management in order to introduce pay incentives for the supervisor and delegating her the authority to motivate subordinates. When the two different cost functions are interpreted in terms of motivation, these results may provide some interesting insights. Within industrial and organizational psychology the study of employee motivation represents one of the most important topics in the discipline (Jex, 2002). A variety of theories of human motivation have been developed over the years (for a review the reader may consult Spector, 2003); we may assume that the two cost
4

max
bt 61 2

aa bt d1 d2 a aa bt a 1 a1a a1 d d 1 2 21 a 21 aa1

#2

d1 d2 : d1 d2 45

1 Since 0 < a < 1 and dd1d2 > 0, the original program (43) is equivalent d2 to

max bt
bt 61 2

bt ; 2

46

and the optimal solution is bt = 1/2, that is, all the prot is shared between the two agents. For the second part of the theorem, let us consider the supervisors prot (10) when the incentives are null

a 1a max1 2bt u u l1 l2 ; 1 2
bt 61 2

47

that is

max1 2bt
bt 61 2

aa bt d1 d2 a a 1 a1a : 21 aa1 d1 d2
d1 d2 d1 d2

48

This problem, since 0 < a < 1 and

> 0, is equivalent to

max1 2bt bt ;
bt 61 2

49

and the optimal solution is bt = 1/4. This proves the thesis. h This result shows that the agents are the residual claimant (see for instance, Varian, 1992) to the output produced if and only if the supervisor distributes it completely to the agents. Hence, having the supervisor in charge to decide how to share the output does not maximize the work group production. Finally, we prove that, under realistic conditions, non-null individual incentive is necessary for supervisors prot maximization. Theorem 8. When bt < 1/2, incentive schemes which maximize the supervisors prot are such that individual incentives are non-null, i.e., b > 0. Proof. Substituting the effort allocations (23) described in Theorem 4 in the objective function of the supervisors problem (10), we obtain

For the sake of brevity, the proportionality factor is omitted.

884

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

functions we examined are the result of different motivations the subordinates in the work group share. While the quadratic cost function seems to be appropriate for individuals who are interested just in monetary incentives, the piecewise constant cost function, as we mentioned in Section 3, may be appropriate when subordinates have high self-efcacy and believe they are capable of accomplishing tasks and motivated to put forth effort (Bandura, 1982). In this sense, the fact that with the quadratic costs individual incentives are necessary, together with the fact that optimal production is not achieved, may highlight some limitation of the scientic management approach, which, among other aspects, emphasized the provision of economic reward for high performance (Daft and Noe, 2001). When considering that, according to Weick (1969), any system consists of several causal relationships, some direct and others inverse, we may assume that, at least in part, the organization may have some effects on the individual motivation of the employees. This aspect is related to Theory X/Theory Y (McGregor, 1960); in fact, according to this theory, the organization management approach determines how subordinates behave and is determined by the belief of supervisors about their subordinates. According to Theory X, the worker is viewed as being indifferent to the organizations needs, lazy and unmotivated; in this case one approach might be incentivizing the effort, and in this case, the quadratic cost function would be appropriate. On the other hand, according to Theory Y, subordinates are capable and not inherently unmotivated or unresponsive to organizational needs, rather it is the responsibility of the manager to create organizational conditions so that subordinates can achieve organizational goals through achieving their own goals; in this case, assuming that individual motivation does not consist of economic rewards, the piecewise constant cost function can be appropriate. Another interpretation of the differences between the two cost functions, may be found taking into account organizational commitment. Among the different denitions of commitment, Mowday et al. (1979) view organizational commitment as consisting of three components: (1) an acceptance of the organizational goals; (2) a willingness to work hard for the organization; and (3) the desire to stay with the organization. We can therefore consider the piecewise constant functions appropriate for describing individuals highly committed to the organization, since they tend to use their whole capacity. In fact, quadratic costs, may describe a lower commitment situation. On the other hand, the research on withdrawal behaviors relates to employees not being at work when scheduled or needed (Spector, 2003); therefore, we can assume the piecewise constant cost functions realistic in organizations where these counterproductive behaviors do not occur, while by contrast, the quadratic cost functions may be more appropriately assumed in organization where employee attendance is incentivized. A last possible interpretation we present, relates to the work of authors like Herman (1973) and Smith (1977), and, more recently, Sutton (2007) who examined the relationship between work attitude and such work-related behavior as job performance; in particular Sutton (2007) interprets impaired organizational performance as the result of dysfunctional leaders. In this case, piecewise constant and quadratic cost functions, might describe different work attitudes, the second being the result of dysfunctional leadership; in fact we found that with the latter cost function the effort is not allocated efciently. Finally, for both cost functions, the optimal incentives are independent of agent capacity. This result is quite interesting since, given the fact that individual incentives are identical, subordinates with different capacities may perceive inequity in the sense of Adams (1965). With piecewise constant cost functions, when individuals compare ratio of outcomes to input, those with higher

capacity may experience underpayment while those with lower capacity on the contrary, may experience overpayment. Vice versa, as it concerns the quadratic cost function this phenomenon may be not so evident as the individual incentives, at least partially compensate the supervisor exerting higher effort.

6. Conclusion and further research In this paper, a simple interaction scheme has been proposed and analyzed. The subordinates are called to allocate their effort in two interdependent tasks. This interdependence is twofold: rstly, in each of the two tasks the performance depends upon the efforts of both subordinates; secondly, the overall performance depends on both tasks. Given the tasks and the incentive mechanism, the optimal incentives depend on the cost functions of subordinates. When considering the analytical results we obtain from the human resources management perspective several insights may be provided. In fact, interpreting the cost functions in terms of subordinate motivation, it emerges that when individuals are not sensitive to monetary incentives, and rather share a norm according to which they use their capacities, then there exists an incentive scheme such that the effort is efciently allocated. By converse, when individuals are interested only in gain-sharing, no such a scheme exists. In Wageman and Baker (1997) it was shown that the interaction between task and reward interdependence made it difcult to provide effective guidance in solving organizational design problems; in this paper we xed the task interdependence and analyzed the interaction between reward interdependence and individual motivation. In our case simple pay practices, such as delegating the manager to decide incentives, may be ineffective and, under some conditions, even counterproductive. In fact, the outcome is that having this sort of self-managed work group with a gain-sharing incentive scheme has several drawbacks; in our example the efciency of the effort allocation is contingent to the motivation of the subordinates. In terms of incentive design, we found that, in the case of quadratic costs, the supervisor incentive should be kept separated from the work group prot. In fact, when the supervisor is in charge of deciding how to share the output, the work group aggregate prot is not maximized. Furthermore, also incentives schemes that may be optimal in the one-shot iteration may exhibit drawbacks when the interaction is repeated over time. We provided several interpretations of the two different cost functions. Independently of the one we choose, what is relevant here is both the effort allocation efciency and the cost of incentives which are peculiar to each of them: clearly, a situation with piecewise constant costs is preferable both in terms of efciency and cost of incentives. Furthermore, the difference in terms of production and cost of incentives may provide a quantitative measure of the differences between organizational cultures. Finally, Kidwell and Bennett (1993) suggest that the individual predisposition to withhold effort may be the result of the combination of the environment, the organization and the group; in this paper we assume that the diverse propensity to withhold effort is formalized by the two cost functions we consider. When considering this propensity as opposite to exerting the discretionary effort, we show that when the discretionary effort is not exerted, the introduction of individual incentives does not allow for the efcient allocation of effort. This way we can give a formal example of the damage done by dysfunctional leaders according to Sutton (2007). In further research, we will analyze some dynamical aspects of the interaction. For example it is evident that, while having

A. Dal Forno, U. Merlone / European Journal of Operational Research 207 (2010) 878885

885

piecewise constant function increases the production and reduces the amount of paid incentives, at the same time it may lead individuals with higher capacity to experience underpayment inequity. It would be interesting to examine the effort allocation dynamics when individuals try to reduce the perceived inequity, and also to understand whether, in this case, individual incentives may become useful to limit the perception. Furthermore, it would also be interesting to allow several subordinates to interact in a supervised group, extending these results both analytically and experimentally.

References
Adams, J.S., 1965. Inequity in social exchange. In: Berkowitz, L. (Ed.), Advances in Experimental Social Psychology. Academic Press, New York, NY, pp. 276299. Alchian, A.A., Demsetz, H., 1972. Production, information costs, and economic organization. American Economic Review 62, 777795. Bailey, T., Berg, P., Sandy, C., 2001. The effect of high-performance work practices on employee earnings in the steel apparel and medical electronics and imaging industries. Industrial and Labor Relations Review 54 (2A), 525543. Bandura, A., 1982. Self-efcacy mechanism in human agency. American Psychologist 37, 122147. Beckmann, M.J., 1988. Tinbergen Lectures on Organization Theory, second ed. Springer, Berlin. Bordoloi, S.K., Matsuo, H., 2001. Human resource planning in knowledge-intensive operations: A model for learning with stochastic turnover, European Journal of Operational Research. Boudreau, J., Hopp, W., McClain, J.O., Thomas, L.J., 2003. On the interface between operations and human resources management. Manufacturing & Service Operations Management 5 (3), 179202. Bylinsky, G., 1990. The hottest high-tech company in Japan. Fortune 121 (1), 8288. Coch, L., French Jr., J.R.P., 1948. Overcoming resistance to change. Human Relations 1, 512532. Daft, R.L., Noe, R.A., 2001. Organizational Behavior. South-Western Publishing, Mason, OH. Dal Forno, A., Merlone, U., 2007. Incentives in supervised teams: An experimental and computational approach. Journal of Social Complexity 3 (1), 3752. Dal Forno, A., Merlone, U., 2009. Individual incentives in supervised work groups. from human subject experiments to agent based simulation. International Journal Internet Enterprise Management 6 (1), 422. Encinosa, W.E., Gaynor, M., Rebitzer, J.B., 2007. The sociology of groups and the economics of incentives: Theory and evidences on compensation systems. Journal of Economic Behavior & Organization 62, 187214. Ethiraj, S.K., Levinthal, D., 2009. Hoping for a to z while rewarding only a: Complex organizations and multiple goals. Organization Science 20, 421. Fast, N., Berg, N., 1975. The Lincoln Electric Company. Harvard Business School Case, #376-028. Gendron, B., 2005. Scheduling employees in Quebecs liquor stores with integer programming. Interfaces 35, 402410. Glance, N.S., Huberman, B.A., 1994. Social dilemmas and uid organizations. In: Carley, K.M., Prietula, M.J. (Eds.), Computational Organization Theory, Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 217239.

Herman, J.B., 1973. Are situational contingencies limiting job attitudejob performance relationships? Organizational Behavior and Human Performance 10, 208224. Herrmann-Pillath, C., 2005. Culture, economic style and the nature of the Chinese economic system. Contribution to the China Roundtable of The International Economic Association. Holmstrm, B., 1982. Moral hazard in teams. Bell Journal Economics 13 (2), 324 340. Holmstrm, B., Milgrom, P., 1987. Aggregation and linearity in prevision of intertemporal incentives. Econometrica 7 (2), 302328. Holmstrm, B., Milgrom, P., 1991. Multitask principal-agent analyses: Incentive contracts, asset ownership, and job design. Journal of Law Economics & Organization 7, 2452. Jex, S.M., 2002. Organizational Psychology. Wiley, New York, NY. Kerr, S., 1975. On the folly of rewarding a while hoping for b. Academy of Management Journal 18 (4), 769783. Kidwell, R.E., Bennett, N., 1993. Employee propensity to withhold effort: A conceptual model to intersect three avenues of research. Academy of Management Review 18 (3), 429456. Lum, T., 2003. Workplace codes of conduct in China and related labor conditions. Key Workplace Documents, Federal Pubblications, Cornell University, ILR School. Macho-Stadler, I., Prez-Castrillo, D., 1997. An Introduction to the Economics of Information. Oxford University Press, Oxford. Maslow, A.H., 1970. Motivation and Personality, second ed. Harper & Row, New York, NY. McGregor, D.M., 1960. The Human Size of Enterprise. McGraw-Hill, New York, NY. Merlone, U., submitted for publication. Organizations as allocators of human and material resources. Mowday, R.T., Steers, R.M., Porter, L., 1979. The measurement of organizational commitment. Journal of Vocational Behavior 14. Roy, D., 1952. Quota restriction and goldbricking in a machine shop. The American Journal of Sociology 57 (5), 427442. Schattler, H., Sung, J., 1997. On optimal sharing rules in discrete- and continuoustime principal-agent problems with exponential utility. Journal of Economic Dynamics & Control 21, 551574. Schelling, T.C., 1960. The Strategy of Conict. Harvard University Press, Cambridge, MA. Smith, F.J., 1977. Work attitudes as predictors of attendance on a specic day. Journal of Applied Psychology 62 (1), 1619. Spector, P.E., 2003. Industrial and Organizational Psychology, third ed. Wiley, New York, NY. Steers, R.M., Black, J.S., 1994. Organizational Behavior, fth ed. Harper Collins, New York, NY. Sutton, R., 2007. The No Asshole Rule. Sphere, London. Varian, H.R., 1992. Microeconomic Analysis. Norton & Company, New York, NY. Wageman, R., Baker, G., 1997. Incentives and cooperation: The joint effects of task and reward interdependence on group performance. Journal of Organizational Behavior 18, 139158. Weick, K.E., 1969. The Social Psychology of Organizing. Addison-Wesley, Reading, MA. Weiss, A., 1987. Incentives and worker behavior: Some evidence. In: Nalbantian, H.R. (Ed.), Incentives, Cooperation and Risk-Taking, Rowman & Littleleld, Lanham, MD, pp. 137150. Zhou, X., 2002. A graphical approach to the standard principal-agent model. Journal of Economic Education 33 (3), 265276.