STUDY ON THE LONG-TERM INCENTIVE MECHANISM
OF THE LARGE-SCALE DREDGING PROJECT
Bin Zhou and Zigang Zhang
Huazhong University of Science and Technology, Wuhan 430074, Hubei Province, China
Keywords: Dredging project, Principal-agent, Long-term income, Incentive mechanism.
Abstract: Through the principal-agent theory and game theory, this article has established the long-term income
model of the large-scale dredging project, which has obtained the solution of the long-term incentive model,
analyzed the impact of the dynamic consistency as well as pledge and negotiating cost to incomes of the
principal and agent by means of increasing different constraint conditions. Furthermore, the study also
shows that the long-term incentive model can provide the agent with stronger incentive.
1 INTRODUCTION
The incentive is essential for the management, and
the principal-agent theory is widely applied for the
analysis on the incentive problem. And, the
principal-agent theory deems the management
problem as the fact that how the principal designs
the incentive mechanism to seduce the agent to take
behaviors optimum to the principal from his own
interests. In 1981, Lazear put forward the game
theory, and he thought that the large salary gap can
reduce the monitoring cost, seduce efforts of the
agent, and highly motivate the consistent interests of
the principal and the agent. (Lazear E, Rosen S.,
1981). Furthermore, Holmstrom & Milgrom put
forward the output share incentive mechanism that
the purely selfish and risk-neutral principal shall
employ the agent with the jealousy and pride
preference and risk avoidance. (Holmstrom B,
Milgrom P., 1987). And, Aoki pointed out that, the
long-term employment can motivate the agent to
accumulate human capitals special for the enterprise.
(Aoki, M.,1988). Also, Yong Zhang established the
two-stage model for the manager, that is the long-
term and short-term income incentives as well as
obtained solutions and analyzed relevant
conclusions. (Yong Zhang. 2004). The study of
Debing Ni revealed that the optimal sharing
proportion will increase along with the increased
effort cost of the agent while reduce along with the
increased expected growth rate of the market price
and effort output. (Debing Ni and Xiaowo Tang.,
2005). Besides, Zongjun Wang obtained the optimal
income combination of the manager through the
long-term and short-term incentive models.
(Zongjun Wang, Chongshuai Qian, and Tian Xia,
2008). Lijun Li studied the problem of how to
motivate the producer to reduce costs as well as
pointed out the cost difference between the
asymmetric information and the symmetric
information, namely the incentive cost; only if the
agent shares the cost-saving income, can the
principal realize his expected income; in the premise
of incomplete and asymmetric information of the
dredging project. (Lijun Li, Xiaoyuan Huang, etc.,
2003). Bin Zhou put forward the relationship
between the agent’s incentive coefficient and the
project cost; that is, the higher the agent’s incentive
coefficient, the lower the project cost; in addition to
the above, it also established the incentive
mechanism based on the equity preference. (Bin
Zhou, Zigang Zhang, 2010).
Above scholars have studied the income of the
manager from different views; however the actual
income of the agent will be restricted by numerous
kinds of factors. This article, from the long-term
incentive view, has corrected relevant study
assumptions, increased the long-term incentive
constraint and the agent’s capability constraint, and
compared the impact of the long-term and short-term
incentives to the agent’s income to make the model
be close to actual conditions and obtain
corresponding conclusions.
574
Zhou B. and Zhang Z..
STUDY ON THE LONG-TERM INCENTIVE MECHANISM OF THE LARGE-SCALE DREDGING PROJECT.
DOI: 10.5220/0003591905740580
In Proceedings of the 13th International Conference on Enterprise Information Systems (PMSS-2011), pages 574-580
ISBN: 978-989-8425-56-0
Copyright
c
2011 SCITEPRESS (Science and Technology Publications, Lda.)
2 MODELING
In order to facilitate the study, following
assumptions are introduced:
Assumption1: Both the principal and the agent
pursue maximizing own benefits.
Assumption 2: In a certain period, the agent’s
current income will be related to that of previous
period, and his ability will be enhanced along with
the increased service year.
Assumption 3: The agent’s capability can be
applied to pursue the short-term income and long-
term income, and the required negotiating cost for
signing one contract is
r
.
Supposing the payment made by the principal to
the agent is
i
N
i
Ni
vky =
,
Ni ,...,2,1
=
; i refers to the
stage
i project, and the total of the stage
i
project is
i
Q
cube (earthwork); the bid award price of the
unilateral dredged soil of the stage
i
project is
i
x
yuan/cube; the project agent’s oil consumption
and material consumption of the stage
i
are
i
t
yuan/cube, with the fixed cost of
i
c
0
yuan/hour; the
income of the stage
i
project is
i
N
v
yuan/cube; the
principal’s income of the stage
i
project is
i
N
v
1
yuan/cube; the payment made by the principal to
the agent of the stage
i project is
i
y
yuan/cube.
When the agent’s effort cost is
)(ac
i
yuan/hour,
ai
c
refers to the agent’s payroll in the competitive
market; the agent’s income of the stage
i project is
i
N
v
2
yuan/cube; when the maximum rated hourly
output of the engineering ship applied by the agent
is
m
p
cube/hour, the output when the project income
is zero; namely the critical output is
0
p
cube /hour.
According to above assumptions, the model to
maximize the principal’s income is as follows:
∑∑
=
NN
i
N
i
N
i
N
vyvvMax
11
1
)]([
(1)
0])([
1
0
1
=
N
iii
N
i
N
ctxpv
(2)
=
N
i
i
N
N
i
N
acvyv
11
2
0)]()([
(3)
N
i
N
k
1
1
2
,
i
N
i
N
kk
+1
,(
m
pp = )
(4)
=
N
di
i
N
N
i
N
Nyacvyv
11
2
)]()([
0
))(
2
1
(
2
1
pp
pp
y
dm
d
d
=
λ
,
48.0=
d
λ
(5)
Ni ,...,2,1
=
In the model, the objective function (1) formula
means the principal’s utility function of the stage N
project; constraint (2) means the total utility function
of the stage N project; constraint (3) means the
agent’s utility function of the stage N project; when
the agent pursues maximize his own utility, the
formula is objective function, and the formula (1)
which is larger than or equals to 0 is the constraint
condition; the (4) formula is the rigid constraint
condition, when the agent enhances the hourly
output to the rated output, the game pricing incentive
coefficient is 1/2, and the incentive coefficient of the
stage N project will not be larger than the sum of N
short-term game incentive coefficients; or else, the
principal is inclined to sign the short-term contract.
According to the assumption 2, when the agent is
engaged in some work for a long time, his capability
will be improved gradually; therefore, his income
shall be increased correspondingly; the constraint (5)
expresses that, the income of the long-term contract
signed by the agent shall be higher than that of the
short-term contract;
d
y refers to the income of the
short-term game pricing formula.
3 MODEL SOLUTION AND ITS
ANALYSIS
3.1 Solution to Maximize the
Principal’s Income in Stage N
Firstly, give up the constraint (4) and the constraint
(5), and directly solve the problem of maximizing
the principal’s income in the stage N
Establish Lagrange function and obtain the
agent’s reaction function in the stage N:
Derive the
λ
,,...,,
21 N
NNN
kkk
in turn and
eliminate
λ
, including
1
1
=
i
N
i
N
i
N
i
N
k
v
v
k
Ni ,...3,2=
(6)
STUDY ON THE LONG-TERM INCENTIVE MECHANISM OF THE LARGE-SCALE DREDGING PROJECT
575
The principal knows the agent’s reaction
function in the stage N; then, the principal’s optimal
incentive coefficient in the stage I is as follows
aNa
N
N
N
N
N
NNi
ccvkvkv ...)1...()1(
1
1
1
1
1
+++=
(7)
Substitute the (7) formula by (6); make the first
derivation to
1
N
k
, and make the derivative as 0,
including
0
1
1
pp
c
N
k
N
N
=
(8)
As
p
k
N
1
>0 ,
2
1
2
p
k
N
<0, when
m
pp =
, make
ω
=
1
N
k
; that is, when the agent increases the output
to the rated one of the ship, the corresponding
incentive coefficient will reach the maximum
value
ω
, including:
)(
))(
1
(
1
0
0
1
pp
pp
N
N
k
m
N
=
ω
(9)
Viewing from the (9) formula, the
ω
value shall
be larger than 0, which shall be related to the
validity of the contract signed between the principal
and the agent.
Making
1
1
+
=
N
ω
(10)
))(1(
)(1
0
0
1
ppNN
pp
N
k
m
N
+
=
0
pp >
(11)
Such formula meets the monotonic increasing
requirements of the incentive coefficient
k to the
hourly output
p
as well as demands that the second
derivative shall be smaller than 0. As
0
)1(
1
2
1
<
+
=
NN
k
N
fails to meet conditions of the
assumption (2), this formula shall be converted.
Supposing that the agent increases the hourly output
to
m
p in whole N cooperation periods of the
principal and the agent, the agent’s optimal incentive
coefficient in the stage I is
1
1
1
+
=
k
N
. In addition to
the above, with the increased cooperation period, the
incentive coefficient will increase till
2
1
=
N
N
k
in the
last period. During N stages of the contract, the
agent’s incentive coefficient will be increased by
1
for each increasing cooperation. Therefore, the
agent’s optimal incentive coefficient shall be as
follows:
Table 1: Calculation Sheet of the Principal’s and Agent’s
Income when p=p
m..
i
1 2 3 N
i
N
k
1
1
+
N
N
1
1
1
N
2
1
According to above reasonings, the (11) formula
can be rewritten as
3.2 Solution to Maximize the Agent’s
Income in Stage N
Firstly, give up the constraint (4) and the constraint
(5), and directly solve the problem of maximizing
the agent’s income in the stage N. here, the agent’s
income is the objective function, and the principal’s
income is larger than 0, which is the participation
constraint.
Establish Lagrange function and obtain the
principal’s reaction function in the stage N,
including:
Derive the
λ
,,...,,
21 N
NNN
kkk in turn and
eliminate
λ
, including
)1(1
1
1
=
i
N
i
N
i
N
i
N
k
v
v
k
Ni ,...3,2=
(13)
The agent knows the principal’s reaction
function in the stage N; then, the agent’s optimal
incentive coefficient in the stage I is as follows
Substitute the (13) formula by (3); make the first
derivation to
1
N
k
, including:
)(
1
1
0
2
1
ppN
c
N
k
N
N
=
(14)
As
p
k
n
1
>0 ,
2
1
2
p
k
n
<0, when
m
pp =
,
'
1
N
k will be
maximum; therefore,
0
ppc
mN
=
; that is, when the
))(2)(1(1
1
0
0
ppiNiN
pp
iN
k
m
i
N
++
+
=
(12)
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
576
agent increases the output to the rated one of the
ship, the corresponding incentive coefficient will
reach the maximum value, including:
)(
1
1
0
2
0
1
ppN
pp
N
k
m
N
=
(15)
0
21
32
1
>+=
k
N
, which meets the constraint
(2); that is, during the contract period, the agent’s
incentive coefficient will be increased along with the
increased working time. Accordingly, the (15)
formula can be rewritten as:
0=
i
N
k
1=i
)(
)(1
1
0
2
0
ppi
pp
i
k
m
i
N
=
Ni ,...3,2=
(16)
3.3 Solution of the Long-term Incentive
Model
When both the principal’s income and the agent’s
income are optimal, the optimization solution can be
obtained without the constraint (4) and the constraint
(5). Compare the (12) and (16) formulas, the agent’s
incentive coefficient will meet
0,0
2
2
<
>
i
k
i
k
i
N
i
N
; that
is, the agent’s incentive coefficient will be increased
along with the increased working time. The long-
term cooperation will be good to the agent; however
the principal’s long-term optimization is not the
same as the agent’s long-term optimization;
accordingly, the feasible solution of the long-term
optimization is between the principal’s long-term
income optimization solution and the agent’s long-
term income optimization solution. Viewing from
the figure 1, we find out that the feasible solution is
located in the area between line I and line III.
However, after increasing the constraint (4) and
constraint (5), the principal’s income optimization
solution (12) formula and the agent’s income
optimization solution (16) formula can’t meet the
constraint (4) constraint (5); therefore, solutions
meeting all constraint conditions of the long-term
incentive shall be obtained
. According to the
constraint conditions (4), when
m
pp =
,
121
...
NN
N
N
N
N
kkkk >>>>
, supposing
1
])1(1[
N
i
N
kik
φ
+=
Ni ,...,2,1=
, then
]
2
1
1[2
1
1
φ
+
N
k
N
(17)
According to the agent’s incentive coefficient
solution when the principal income is optimal, when
m
pp
=
:
1
1
1
+
=
N
k
N
(18)
Making
]
2
1
1[2
1
1
1
φ
+
=
+
N
N
,
1=
φ
; therefore,
when
m
pp
=
,
1+
=
N
i
k
i
N
, then, when
m
ppp
0
,
))(1(
)(
0
0
ppNN
ppi
N
i
k
m
i
N
+
=
(19)
Figure 1: Comparison Chart of the Agent’s Incentive
Coefficient.
1
=
k
—Total incentive coefficient of the project
2
1
=k
—Agent’s incentive coefficient line in the
dynamic game pricing formula (
m
pp =
)
rk =
2
1
—Agent’s incentive coefficient after
deducting the negotiating cost in the dynamic game
pricing formula (
m
pp
=
)
I---Agent’s incentive coefficient line when the
long-term contract agent is optimal (
m
pp =
),
2
11
1
ii
k
i
N
=
,
Ni ,...2,1
=
II---Agent’s long-term incentive coefficient line
(
m
pp
=
),
1+
=
N
i
k
i
n
,
Ni ,...2,1
=
III---Agent’s incentive coefficient line when the
long-term contract principal is optimal (
m
pp =
),
iN
k
i
N
+
=
2
1
,
Ni ,...2,1
=
STUDY ON THE LONG-TERM INCENTIVE MECHANISM OF THE LARGE-SCALE DREDGING PROJECT
577
3.4 Comparative Analysis on Three
Solutions
3.4.1 Long-term and Short-term Dynamic
Consistency
Determine the agent’s incentive coefficient based on
the maximization of the principal’s long-term
income; namely the (12) formula.
Without considering discounts, suppose that the
agent tries to increase the output of the ship to the
rated one during the cooperation period, the sum of
incentive coefficients of three stages shall be
12
13
2
1
3
1
4
1
3
1
=++=
i
N
k
and agent’s total incentive
coefficient sum during 30-year career shall be
12
130
.
However, if the agent and the principal sign the 30-
year contract, the agent’s 30-year incentive
coefficient sum shall
be
9.2
2
1
3
1
...
30
1
31
1
30
1
++++=
i
N
k
. Therefore, the
longer contract period will be more beneficial to the
principal; however, for the agent, the higher
expectations for the future, the shorter-term contract
will be.
Determine the agent’s incentive coefficient based
on the maximization of the agent’s long-term
income; namely the (16) formula or the comparison
between 3-period and 30-period. Without
considering discounts, the incentive coefficient sum
of three-period contract is:
36
13
9
1
4
1
0
3
1
=++=
i
N
k
;
that of ten 3-period contracts within 30 years is
36
130
,
and that of 30-period is:
2
30
900
869
841
811
...
9
5
4
1
0
30
1
>+++++=
i
N
k
.
Obviously, the long-term contract is more
beneficial to the agent. Under such condition, the
principal will choose to sign the contract with short
cooperation period. Therefore, under these two
conditions, the principal and the agent are in the
bargaining game process; namely, the principal and
agent are dynamically inconsistent in the long-term
and short-term incomes, see figure 2 [A,B,C].
I, II—Incentive coefficient line of
3,2,1,,3 === ippN
m
;
III—Incentive coefficient line of
NippN
m
,...2,1,,30 ===
Figure 2A: Comparison Between the Long-term and
Short-term Incentive Coefficients in the Principal’s Long-
term Income Optimization.
I, II—Incentive coefficient line of
3,2,1,,3 === ippN
m
III—Incentive coefficient line of
NippN
m
,...2,1,,30 ===
Figure 2B: Comparison Between the Long-term and
Short-term Incentive Coefficients in the Agent’s Long-
term Income Optimization.
I, II—Incentive coefficient line of
3,2,1,,3 === ippN
m
III—Incentive coefficient line of
NippN
m
,...2,1,,30 ===
Figure 2C: Comparison the Long-term and Short-term
Incentive Coefficients in the Long-term Incentive Model.
Determine the agent’s incentive coefficient based
on the long-term incentive model solution; namely
the (19) formula. No matter how long the contract
period is, when the agent tries to increase it to the
rated output
m
p
, it can certify that the incentive
coefficient sum meets
=
N
i
N
N
k
1
2
; namely the (19)
formula meets the dynamic consistency of the long-
term and short-term incomes of the principal and the
agent.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
578
3.4.2 Comparison between the Long-term
Income and the Short-term Income
As the long-term optimization solutions of the agent
and the principal can’t meet the dynamic
consistency, it is not a kind of stable solution, which
will change along with both parties’ negotiating
skills. When the principal takes the priority, he
hopes to sign the long-term contract based on his
long-term income optimization; however, the agent
prefers to the dynamic game pricing. When the agent
takes the priority, he hopes to sign the long-term
contract based on his long-term income
optimization; however, the principal prefers to the
dynamic game pricing. Accordingly, the analysis on
the dynamic game pricing and long-term incentive
model solutions will be more significant. Taking the
3 stages as the example, the negotiating cost will not
be considered and the agent’s calculation data in the
3 stages are the same.
The income of 3 contracts signed by the agent
according to the short-term game pricing coefficient
is:
dpctxp
pp
pp
v
m
m
p
ppp
m
])([]
)(
))(3
2
3
(
2
3
[3
0
)(
2
1
0
0
2
00
=
+
λ
The agent signs a 3-period contract based on the
long-term incentive model:
)(12
)(
3
0
3
pp
ppii
k
m
i
=
3,2,1,3 == iN
dpctxp
pp
pp
v
m
m
p
ppp
m
i
+
=
)(
4
1
0
0
0
3
1
3
00
])(][
)(2
)(
2[
dpctxp
pp
pp
vv
m
m
p
ppp
m
i
])([]
)(
)(44.0
2
1
[3
0
)(
2
1
0
0
2
3
1
3
00
=
+
0])(][
)(2
)(
2[
)(
2
1
)(
4
1
0
0
0
00
00
>
+
+
+
ppp
ppp
m
m
m
dpctxp
pp
pp
3.4.3 Negotiating Cost r
Costs are required for facts that the principal
searches for the agent and the agent signs the
agreement with the principal; compared with the
scale advantages of the principal, the proportion of
the agent’s negotiating cost of its own income will
be higher. As for the agent, if there is no negotiating
cost
r
, the incentive coefficient of N 1-period
contract is the same as the sum of the incentive
coefficient of the N-period contract. When the
negotiating cost
0>r , and
Nrrk
N
i
N
)
2
1
(
1
>
, the
income of the N-period contract will be higher than
that of N 1-period contracts. As for the principal,
though the negotiating cost proportion of the income
is not high, the long-term contract will be more
beneficial.
3.4.4 Pledged Capital w or “Hostage”
When calculating from the figure 1,
==
N
N
ai
i
N
i
Nai
N
i
N
i
N
dtcvkdtcvkw
2/
2/
1
)(])
2
1
[(
The income of the agent’s first half of the career
which is deducted by the principal has been
compensated in his later half of the career, which
can be deemed as the investment or savings made by
the agent to the principal. Due to the pledged capital,
the agent’s and the principal’s goals are further
harmonized, and if the agent be fired because of
effortless, or lead to the principals’ fewer income;
the agent will also have corresponding loss. And, the
longer the worktime is, the larger the corresponding
loss will be. Accordingly, the long-term contract
incentive to the agent is larger.
4 CONCLUSIONS
From the standpoint of the long-term incentive, the
article has studied the income of the agent of the
large-scale dredging project, established the long-
term incentive model, obtained the optimization
solutions of the principal’s long-term income and the
agent’s long-term income; in addition to the above,
it has obtained the long-term incentive model
solution through increasing various kinds of
constraints (long-term incentive constraint, agent’s
capability constraint, and short-term income
constraint ). The main conclusions cover: I. As for
the principal’s long-term optimization solution and
the agent’s long-term optimization solution; while as
the long-term incentive model solution can meet the
dynamic consistency requirements, it is a kind of
stable solution; II. In the long-term incentive model,
the agent’s long-term income is higher than the
short-term income, which has provided the agent
with stronger incentive; III. If consideration is not
given to the negotiating cost, the sum of the
coefficient of the short-term game price equals to
STUDY ON THE LONG-TERM INCENTIVE MECHANISM OF THE LARGE-SCALE DREDGING PROJECT
579
that of the long-term incentive model. If there is the
negotiating cost, the coefficient sum of the long-
term contract is larger than that of the short-term
contract; accordingly, the long-term contract will be
beneficial to the principal and the agent; IV. As the
existence of the pledge, the objection of the agent
and the principal will be more gradually-consistent.
the agent’s incentive shall be further strengthened;
V. The long-term incentive coefficient is similar to
the seniority pay while the seniority pay also has
differences. The classification of the long-term
incentive coefficient is the sharing proportion of the
project income; thus the principal shall also assess
the agent’s hourly output. When the income of the
project under the management of the agent is lower,
the principal shall not increase the agent’s incentive;
namely, coefficient by years, that is, the interior
promotion system shall be established. After
increasing the constraint, the long-term incentive
model will be more practicable and convenient
operation and application. And, the disadvantage is
that, as the article fails to give the agent capability
classification incentive mode, which shall be further
studied.
REFERENCES
Lazear E, Rosen S. 1981, Rank-order Tournaments As
Optimum Labor Contracts [J]. Journal of Political
Economy, 89(5): 841
864.
Holmstrom B, Milgrom P. 1987, Aggregation and linearity
in the provision of intertemporal incentives [J].
Econometrica, 55(2): 303
328.
Aoki, M., 1988, Information, Incentives and Bargaining in
the Japanese Economy, Cambridge University Press.
Yong Zhang. 2004, Discussion of the Manager Long-term
and Short-term Income Incentive [J]. Management
Engineering Journal, 18(3): 125
127.
Debing Ni and Xiaowo Tang. 2005, Paincipal-agent
Model of the Hard Decision Sharing System of
Agent’s Hard Decision Flexibility Cutting [J].
Management Science Journal, 8 (3): 15
23.
Zongjun Wang, Chongshuai Qian, and Tian Xia. 2008
manager long-term and short-term incentive income
model and its optimization study [J] Management
Engineering Journal, Vol.22, No.1,113
116.
Lijun Li, Xiaoyuan Huang, etc. 2003, Application of the
Paincipal-agent theory in the Reduced Cost [J].
Industrial Engineering and Management, 4(4): 55
58.
Bin Zhou, Zigang Zhang, 2010, Pricing the Cost of
Dredging Engineering Based on the Dynamic Game,
Vol.2 No. 184,
© IEEE.
Bin Zhou, Zi Gang Zhang, 2011, Study on Incentive
Mechanism of Large-scale Dredging Project Based on
Fairness Preference [J]. Advanced Materials Research,
Vol.219-220:144
150.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
580