The Owen and the Owen–Banzhaf Values Applied to the Study of the
Madrid Assembly and the Andalusian Parliament in Legislature
2015–2019
Jos
´
e Miguel Gim
´
enez and Mar
´
ıa Albina Puente
Department of Mathematics and Engineering School of Manresa, Technical University of Catalonia, Spain
Keywords:
Cooperative Game, Shapley Value, Banzhaf Value, Coalition Structure, Multilinear Extension.
Abstract:
This work focuses on the Owen value and the Owen–Banzhaf value, two classical concepts of solution defined
on games with structure of coalition blocks. We provide a computation procedure for these solutions based on
a method of double-level work obtained from the multilinear extension of the original game. Moreover, two
applications to several possible political situations in the Madrid Assembly and the Andalusian Parliament
(legislatures 2015–2019) are also given.
1 INTRODUCTION
Introduced in (Aumann and Dr
`
eze, 1974), the notion
of game with a coalition structure gave a new impul-
sion to the development of value theory. These au-
thors extended the Shapley value to this new frame-
work in such a manner that the game really splits into
subgames played by the unions isolatedly from each
other, and every player receives the payoff allocated
to him by the Shapley value in the subgame he is play-
ing within his union.
A second approach was used in (Owen, 1977),
when introducing the first coalitional value, called
now the Owen value. The Owen value is the result
of a two–step procedure: first, the unions play a quo-
tient game among themselves, and each one receives
a payoff which, in turn, is shared among its players in
an internal game. Both payoffs, in the quotient game
for unions and within each union for its players, are
given by the Shapley value.
The same procedure is applied in (Owen, 1982)
to the Banzhaf value and it is obtained the modi-
fied Banzhaf value or Owen–Banzhaf value. In this
case the payoffs at both levels (unions in the quotient
game and players within each union) are given by the
Banzhaf value.
In (Alonso and Fiestras, 2002), the authors sug-
gested to modify the two–step allocation scheme and
use the Banzhaf value for sharing in the quotient
game and the Shapley value within unions. This gave
rise to the symmetric coalitional Banzhaf value or
Alonso–Fiestras value. On the other hand, in (Amer
et al., 2002) was considered a sort of “counterpart”
of the Alonso–Fiestras value where the Shapley value
is used in the quotient game and the Banzhaf value
within unions.
The multilinear extension of a cooperative game
was introduced in (Owen, 1972) and then it was
applied to the calculus of the Shapley value. The
computing technique based on the multilinear ex-
tension has been applied to many values: to the
Banzhaf value in (Owen, 1975); to the Owen value
in (Owen and Winter, 1992); to the Owen–Banzhaf
value in (Carreras and Magana, 1994); to the quo-
tient game in (Carreras and Magana, 1997); to the bi-
nomial semivalues and multinomial probabilistic in-
dices in (Puente, 2000); to the coalitional semivalues
in (Amer and Gim
´
enez, 2003); to the α–decisiveness
and Banzhaf α–indices in (Carreras, 2004); to the
Alonso–Fiestras value in (Alonso et al., 2005); to the
symmetric coalitional binomial semivalues in (Car-
reras and Puente, 2011); to all semivalues in (Carreras
and Gim
´
enez, 2011); and to coalitional multinomial
probabilistic values in (Carreras and Puente, 2013).
The present paper focuses on giving a new com-
putational procedure for the Owen and the Owen–
Banzhaf value by means of the multilinear extension
of the game.
The organization of the paper is as follows. In
Section 2, some preliminaries are provided. Section
3 is devoted to the computation of the Owen and the
Owen–Banzhaf values in terms of the multilinear ex-
Giménez, J. and Puente, M.
The Owen and the Owen-Banzhaf Values Applied to the Study of the Madrid Assembly and the Andalusian Parliament in Legislature 2015-2019.
DOI: 10.5220/0007297000450052
In Proceedings of the 8th International Conference on Operations Research and Enterprise Systems (ICORES 2019), pages 45-52
ISBN: 978-989-758-352-0
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
45
tension. Section 4 contains two applications of these
values to the analysis of the Madrid Assembly and the
Andalusian Parliament (legislatures 2015–2019).
2 PRELIMINARIES
2.1 Cooperative Games and Values
Let N be a finite set of players and 2
N
be the set of
its coalitions (subsets of N). A cooperative game on
N is a function v : 2
N
R, that assigns a real number
v(S) to each coalition S N, with v(
/
0) = 0. A game v
is monotonic if v(S) v(T ) whenever S T N and
simple if, moreover, v(S) = 0 or 1 for every S N.
A player i N is a dummy in v if v(S {i}) = v(S) +
v({i}) for all S N\{i}, and null in v if, moreover,
v({i}) = 0. Two players i, j N are symmetric in v
if v(S {i}) = v(S { j}) for all S N\{i, j}. Given
a nonempty coalition T N, the restriction to T of a
given game v on N is the game v
|T
on T that we will
call a subgame of v and is defined by v
|T
(S) = v(S)
for all S T .
Endowed with the natural operations for real–
valued functions, i.e. v + v
0
and λv for all λ R, the
set of all cooperative games on N is a vector space
G
N
. For every nonempty coalition T N, the una-
nimity game u
T
is defined by u
T
(S) = 1 if T S and
u
T
(S) = 0 otherwise, and it is easily checked that the
set of all unanimity games is a basis for G
N
, so that
dim(G
N
) = 2
n
1 if n = |N|. Each game v G
N
can then be uniquely written as a linear combina-
tion of unanimity games, and its components are the
Harsanyi dividends (Harsanyi, 1959):
v =
T N: T 6=
/
0
α
T
u
T
, where
α
T
= α
T
(v) =
ST
(1)
ts
v(S)
and, as usual, t = |T | and s = |S|.
By a value on G
N
we will mean a map f : G
N
R
N
, that assigns to every game v a vector f [v] with
components f
i
[v] for all i N.
Particularly, the Shapley value (Shapley, 1953) ϕ,
and the Banzhaf value (Owen, 1975) β, are defined by
ϕ
i
[v] =
SN\{i}
1/n
n 1
s
[v(S {i}) v(S)]
and
β
i
[v] =
SN\{i}
1
2
n1
[v(S {i}) v(S)]
for all i N and all v G
N
.
As it is well known, the Shapley value is the
unique value that satisfies:
(i) additivity: ϕ[v+v
0
] = ϕ[v]+ϕ[v
0
], for all v,v
0
G
N
;
(ii) anonymity: ϕ
θi
[θv] = ϕ
i
[v] for all permutation θ
on N, i N, and v G
N
;
(iii) dummy player property: if i N is a dummy in
game v, then ψ
i
[v] = v({i}).
(iv) efficiency:
iN
ϕ
i
[v] = v(N)
The Banzhaf value follows a similar scheme, sat-
isfying the total power property (Owen, 1975)
iN
β
i
[v] =
iN
SN\{i}
1
2
n1
[v(S {i}) v(S)]
for all v G
N
, instead of additivity.
Notice that these two values are defined for each
N. In fact, these values are defined on cardinali-
ties rather than on specific player sets. When nec-
essary, we shall write ϕ
(n)
and β
n
for the Shapley and
Banzhaf values on cardinality n. In both cases, ϕ
(n)
and β
n
induce values ϕ
(t)
and β
t
for all cardinalities
t < n.
The multilinear extension (Owen, 1972) of a game
v G
N
is the real–valued function defined on R
N
by
f
v
(X
N
) =
SN
iS
x
i
jN\S
(1 x
j
)v(S).
where X
N
denotes the set of variables x
i
for i N.
As it is well known, both the Shapley and Banzhaf
values of any game v can be easily obtained from its
multilinear extension. Indeed, ϕ[v] can be calculated
by integrating the partial derivatives of the multilin-
ear extension of the game along the main diagonal
x
1
= x
2
= · ·· = x
n
of the cube [0,1]
N
(Owen, 1972),
while the partial derivatives of that multilinear exten-
sion evaluated at point (1/2,1/2,. .. ,1/2) give β[v]
(Owen, 1975). This latter procedure extends well to
any p–binomial semivalue evaluating the derivatives
at point (p, p,. .., p), as we can see in (Puente, 2000),
(Freixas and Puente, 2002), or (Amer and Gim
´
enez,
2003).
2.2 Games with Coalition Structure and
Coalitional Values
Given N = {1,2,.. ., n}, we will denote by B(N) the
set of all partitions of N. Each B B(N) is called
a coalition structure in N, and a union each member
of B. The so–called trivial coalition structures are
B
n
= {{1},{2},. .. ,{n}} (individual coalitions) and
B
N
= {N} (grand coalition). A cooperative game with
a coalition structure is a pair [v;B], where v G
N
and B B(N) for a given N. Each partition B gives
a pattern of cooperation among players. We denote
ICORES 2019 - 8th International Conference on Operations Research and Enterprise Systems
46
by G
cs
N
= G
N
× B(N) the set of all cooperative games
with a coalition structure and player set N.
If [v; B] G
cs
N
and B = {B
1
,B
2
,. .. ,B
m
}, the quo-
tient game v
B
is the cooperative game played by the
unions or, rather, by the quotient set M = {1, 2,..., m}
of their representatives, as follows:
v
B
(R) = v(
[
rR
B
r
) for all R M.
By a coalitional value on G
cs
N
we will mean a map
g : G
cs
N
R
N
, which assigns to every pair [v;B] a vec-
tor g[v;B] with components g
i
[v;B] for each i N.
If f is a value on G
N
and g is a coalitional value
on G
cs
N
, it is said that g is a coalitional value of f iff
g[v;B
n
] = f [v] for all v G
N
.
The Owen value (Owen, 1977) is the coalitional
value Φ defined by
Φ
i
[v;B] =
RM\{k}
T B
k
\{i}
1
mb
k
1
m1
r
1
b
k
1
t
[v(Q T {i}) v(Q T )]
for all i N and [v;B] G
cs
N
, where B
k
B is the union
such that i B
k
and Q =
S
rR
B
r
.
It was axiomatically characterized in (Owen,
1977) as the only coalitional value that satisfies the
following properties: the natural extensions to this
framework of
efficiency
additivity
the dummy player property
and also
symmetry within unions: if i, j B
k
are symmetric
in v then
Φ
i
[v;B] = Φ
j
[v;B]
symmetry in the quotient game: if B
r
,B
s
B are
symmetric in [v;B] then
iB
r
Φ
i
[v;B] =
jB
s
Φ
j
[v;B].
The Owen value is a coalitional value of the Shap-
ley value ϕ in the sense that Φ[v; B
n
] = ϕ[v] for all
v G
N
. Besides, Φ[v; B
N
] = ϕ[v]. Finally, as Φ is de-
fined for any N, the following property makes sense
and is also satisfied:
quotient game property: for all [v; B] G
cs
N
,
iB
k
Φ
i
[v;B] = Φ
k
[v
B
;B
m
] for all B
k
B.
The Owen value can be viewed as a two–step al-
location rule. First, each union B
k
receives its payoff
in the quotient game according to the Shapley value;
then, each B
k
splits this amount among its players by
applying the Shapley value to a game played in B
k
as follows: the worth of each subcoalition T of B
k
is
the Shapley value that T would get in a “pseudoquo-
tient game” played by T and the remaining unions on
the assumption that B
k
\T leaves the game, i.e. the
quotient game after replacing B
k
with T . This is the
way to bargain within the union: each subcoalition T
claims the payoff it would obtain when dealing with
the other unions in absence of its partners in B
k
.
The Owen–Banzhaf value Γ (Owen, 1982) fol-
lows a similar scheme. The resulting formula par-
allels that of the Owen value given above with
the sole change of coefficient 1/mb
k
m1
r

b
k
1
t
by
2
1m
2
1b
k
. This value, which is a coalitional value
of the Banzhaf value β, does not satisfy efficiency, but
neither symmetry in the quotient game nor the quo-
tient game property. The bargaining interpretation is
the same as in the case of the Owen value by replac-
ing everywhere the Shapley value with the Banzhaf
value.
3 A NEW COMPUTATIONAL
PROCEDURE
In this section we give a new computational procedure
to calculate the Owen and the Owen–Banzhaf values
in terms of the MLE of the game. First of all we need
the following lemma.
Lemma 3.1. Let [v; B] G
cs
N
, B = {B
1
,B
2
,. .. ,B
m
} a
coalition structure in N. Then, the allocations given
by Φ and Γ to players belonging to a union B
j
can be
obtained as a linear combination of the allocations to
unanimity games u
T
, where T = V W , V B
j
and
W 2
B\B
j
.
Proof. Each game v G
N
can be uniquely written as
linear combination of unanimity games
v =
T N: T 6=
/
0
α
T
u
T
,
where α
T
= α
T
(v) =
ST
(1)
ts
v(S).
For all i B
j
, by linearity, Φ
i
[v;B] =
T N: T 6=
/
0
α
T
Φ
i
[u
T
] and it suffices consider unanim-
ity games u
T
with
T = V A
i
1
A
i
2
. .. A
i
p
, V B
j
,
{i
1
,i
2
,...,i
p
} M \ { j},
/
0 6= A
i
q
B
i
q
, q = 1, ..., p.
According to the definition of the Owen value it is
easy to check that the allocations to players in B
j
only
depend on the allocations in the unanimity games de-
fined on inside coalitions in B
j
and entire unions out-
side B
j
. That is,
Φ
i
[u
T
;B] = Φ
i
[u
V A
i
1
A
i
2
...A
i
p
;B]
= Φ
i
[u
V B
i
1
B
i
2
...B
i
p
;B].
The Owen and the Owen-Banzhaf Values Applied to the Study of the Madrid Assembly and the Andalusian Parliament in Legislature
2015-2019
47
Analogously for the Owen–Banzhaf value.
Notice that the number of unanimity games of this
form is (2
b
j
1)2
m
with b
j
= |B
j
| and m = |M|.
Proposition 3.2. Let B = {B
1
,B
2
,. .. ,B
m
} be a coali-
tion structure in N. Fixed a union B
j
,
(i) The allocation to a player i belonging to B
j
in
a unanimity game u
T
, T = V B
i
1
· ·· B
i
h
, V B
j
and {i
1
,...,i
h
} M \ { j}, given by the Owen value Φ
is
Φ
i
[u
T
;B] =
1
h+1
1
v
i T
0 i / T
where
1
h+1
and
1
v
are the induced coefficients of ϕ
(h+1)
and ϕ
(v)
, respectively.
(ii) The allocation to a player i belonging to B
j
in
a unanimity game u
T
, T = V B
i
1
· ·· B
i
h
, V B
j
and {i
1
,...,i
h
} M \{ j}, given by the Owen–Banzhaf
value Γ is
Γ
i
[u
T
;B] =
2
h
2
(v1)
= 2
(h+v1)
i T
0 i / T
where
1
2
h
and
1
2
v1
are the induced coefficients of
β
(h+1)
and β
(v)
, respectively.
Proof. (i) For i T , we have
Φ
i
[u
T
;B] =
RM\{ j}
1/m
m 1
r
SB
j
\{i}
1/b
j
b
j
1
s
[u
T
(Q S {i}) u
T
(Q S)]
where Q =
[
rR
B
r
, b
j
= |B
j
|, and s = |S|.
Only u
T
(Q S {i}) u
T
(Q S) does not vanish
for coalitions R such that {i
1
,...,i
h
} R M \ { j}
and for coalitions S such that V \ {i} S B
j
\ {i}.
Then,
Φ
i
[u
T
;B] =
m1
r=h
m 1 h
r h
1/m
m 1
r
b
j
1
s=v1
b
j
v
s v + 1
1/b
j
b
j
1
s
=
1
h + 1
1
v
In case of i 6∈ T , all marginal contributions u
T
(Q S
{i}) u
T
(Q S) vanish.
(ii) For i T , we have
Γ
i
[u
T
;B] =
RM\{ j}
2
(m1)
SB
j
\{i}
2
(b
j
1)
[u
T
(Q S {i}) u
T
(Q S)]
where Q =
[
rR
B
r
, b
j
= |B
j
|, and s = |S|.
Analogously to the previous case, only u
T
(QS
{i}) u
T
(Q S) does not vanish for coalitions R such
that {i
1
,...,i
h
} R M\{ j} and for coalitions S such
that V \ {i} S B
j
\ {i}. Then,
Γ
i
[u
T
;B] =
m1
r=h
m 1 h
r h
2
(m1)
b
j
1
s=v1
b
j
v
s v + 1
2
(b
j
1)
= 2
(h+v1)
In case of i 6∈ T , all marginal contributions u
T
(Q S
{i}) u
T
(Q S) vanish.
Example 3.3. On the set N = {1,2,3,4,5,6, 7}, let
B = {{1,2,3},{4, 5},{6}, {7}} be a coalition struc-
ture.
We will obtain the allocations to players i B
1
ac-
cording to Φ for the unanimity games u
{1,2,4,6,7}
and
u
{1,2,4,5,6,7}
. They are
Φ
i
[u
{1,2,4,6,7}
;B] =
1
4
·
1
2
, for i = 1, 2 and
Φ
3
[u
{1,2,4,6}
;B] = 0,
In a similar way and according to Lemma 3.1, for
u
{1,2,4,5,6,7}
we obtain
Φ
i
[u
{1,2,4,5,6,7}
;B] =
1
4
·
1
2
, for i = 1, 2 and
Φ
3
[u
{1,2,4,5,6,7}
;B] = 0,
Notice that the allocations in both games are the same
because coalitions {1,2,4, 6,7} and {1,2, 4,5,6, 7}
intersect the same unions B
2
, B
3
and B
4
.
The computing technique based on the multilinear
extension has been applied to many coalitional val-
ues: to the Owen value in (Owen and Winter, 1992);
to the Owen–Banzhaf value in (Carreras and Magana,
1994); to the quotient game in (Carreras and Mag-
ana, 1997); to the coalitional semivalues in (Amer
and Gim
´
enez, 2003); to the Alonso–Fiestras value
in (Alonso et al., 2005); to the symmetric coalitional
binomial semivalues in (Carreras and Puente, 2011);
and to the coalitional multinomial probabilistic val-
ues in (Carreras and Puente, 2013). In next theorems
we present a new method to compute the Owen and
the Owen–Banhaf values by means of the multilinear
extension of the game.
Theorem 3.4. Given in N a game with coalition
structure, [v;B] G
cs
N
, B = {B
1
,B
2
,. .. ,B
m
} coalition
structure in N, the following steps lead to the Owen
value of any player i B
j
in [v;B].
1. Obtain the multilinear extension f (x
1
,x
2
,. .. ,x
n
)
of game v.
ICORES 2019 - 8th International Conference on Operations Research and Enterprise Systems
48
2. For every r 6= j and all h B
r
, replace the variable
x
h
with y
r
. This yields a new function of x
k
for
k B
j
and y
r
for r M\{ j}.
3. In this new function, reduce to 1 all higher expo-
nents, i.e. replace with y
r
each y
q
r
such that q > 1.
This gives a new multilinear function denoted as
g
j
((x
k
)
kB
j
, (y
r
)
rM\{ j}
) –the modified multilin-
ear extension of union B
j
4. After some calculus, the obtained modified multi-
linear extension reduces to
g
j
((x
k
)
kB
j
,(y
r
)
rM\{ j}
) =
V B
j
W M\{ j}
λ
V W
kV
x
k
rW
y
r
5. Multiply each product
kV
x
k
by
1
v
and each
product
rW
y
r
by
1
w+1
obtaining a new multi-
linear function called g
j
.
6. Obtain the partial derivative of g
j
with respect to
x
i
evaluated at point (1,.. ., 1) and
Φ
i
[v;B] =
g
j
x
i
(1
B
j
,1
M\{ j}
).
Proof. Steps 1–3 have been already used in (Owen
and Winter, 1992), (Carreras and Magana, 1994),
(Puente, 2000), (Freixas and Puente, 2002), (Alonso
et al., 2005), (Carreras and Puente, 2011) and (Car-
reras and Puente, 2013) to obtain the modified multi-
linear extension of union B
j
. Step 4 shows the mod-
ified multilinear extension as a linear combination of
multilinear extensions of unanimity games. Step 5
weights each unanimity game according to Proposi-
tion 3.2 so that step 6 gives as usual the marginal
contribution of player i and his allocation Φ
i
[v;B] is
obtained.
Theorem 3.5. Given in N a game with coalition
structure, [v;B] G
cs
N
, B = {B
1
,B
2
,. .. ,B
m
} coalition
structure in N, the following steps lead to the Owen–
Banzhaf value of any player i B
j
in [v;B].
1. Obtain the multilinear extension f (x
1
,x
2
,. .. ,x
n
)
of game v.
2. For every r 6= j and all h B
r
, replace the variable
x
h
with y
r
. This yields a new function of x
k
for
k B
j
and y
r
for r M\{ j}.
3. In this new function, reduce to 1 all higher expo-
nents, i.e. replace with y
r
each y
q
r
such that q > 1.
This gives a new multilinear function denoted as
g
j
((x
k
)
kB
j
, (y
r
)
rM\{ j}
) –the modified multilin-
ear extension of union B
j
4. After some calculus, the obtained modified multi-
linear extension reduces to
g
j
((x
k
)
kB
j
,(y
r
)
rM\{ j}
) =
V B
j
W M\{ j}
λ
V W
kV
x
k
rW
y
r
5. Multiply each product
kV
x
k
by 2
(v1)
and
each product
rW
y
r
by 2
w
obtaining a new
multilinear function called g
j
.
6. Obtain the partial derivative of g
j
with respect to
x
i
evaluated at point (1,.. ., 1) and
Γ
i
[v;B] =
g
j
x
i
(1
B
j
,1
M\{ j}
).
Proof. Analogously to the previous theorem.
Example 3.6. Let us consider the 4–person weighted
majority game v [65;48, 37,27, 17] and the coali-
tion structure B = {{1,4},{2}, {3}}. We will com-
pute Φ[v;B].
The set of minimal winning coalitions is
W
m
(v) = {{1,2},{1, 3},{1, 4},{2,3,4}},
and the multilinear extension of v is given by
f (x
1
,x
2
,x
3
,x
4
) = x
1
x
2
+ x
1
x
3
+
x
1
x
4
+ x
2
x
3
x
4
x
1
x
2
x
3
x
1
x
2
x
4
x
1
x
3
x
4
.
Notice that players 2 and 3 became null in the
quotient game and, by the quotient game property,
Φ
2
[v;B] = Φ
3
[v;B] = 0 and it is not necessary to
compute the corresponding modified multilinear ex-
tensions g
2
and g
3
.
Steps 1–4 in Theorem 3.4 give the modified multi-
linear extension g
1
of union B
1
:
g
1
(x
1
,x
4
,y
2
,y
3
) = x
1
y
2
+ x
1
y
3
+
x
1
x
4
+ y
2
y
3
x
4
x
1
y
2
y
3
x
1
y
2
x
4
x
1
y
3
x
4
,
Step 5 leads to g
1
.
g
1
(x
1
,x
4
,y
2
,y
3
) =
1
2
x
1
y
2
+
1
2
x
1
y
3
+
1
2
x
1
x
4
+
1
3
y
2
y
3
x
4
1
3
x
1
y
2
y
3
1
2
1
2
x
1
y
2
x
4
1
2
1
2
x
1
y
3
x
4
Step 6 yields
Φ
1
[v;B] =
2
3
Φ
4
[v;B] =
1
3
.
4 TWO APPLICATIONS TO THE
POLITICAL ANALYSIS
Example 4.1. We consider here the Madrid Assembly
in legislature 2015–2019.
Four parties elected members to the Madrid As-
sembly (129 seats) in the elections held on 24 May
2015. The seat distribution of the parties are as fol-
lows.
The Owen and the Owen-Banzhaf Values Applied to the Study of the Madrid Assembly and the Andalusian Parliament in Legislature
2015-2019
49
Table 1: Classical measures of power in the Madrid Assembly, legislature 2015–2019.
(a) (b) (c)
(–) (R) (L) (–) (R) (L) (–) (R) (L)
1. PP 0.5000 0.6666 0.0000 0.7500 0.7500 0.0000 0.7500 0.7500 0.0000
2. PSOE 0.1666 0.0000 0.3333 0.2500 0.0000 0.2500 0.2500 0.0000 0.3333
3. Podemos 0.1666 0.0000 0.3333 0.2500 0.0000 0.2500 0.2500 0.0000 0.3333
4. C’s 0.1666 0.3333 0.3333 0.2500 0.2500 0.2500 0.2500 0.2500 0.3333
1: PP (Partido Popular), conservative party: 48
seats.
2: PSOE (Partido Socialista Obrero Espa
˜
nol), mod-
erate left–wing party: 37 seats.
3: Podemos, radical left–wing party: 27 seats
4: C’s (Ciudadanos), Spanish nationalist liberal
party: 17 seats.
Under the standard absolute majority rule, and as-
suming voting discipline within parties, the structure
of this parliamentary body can be represented by the
weighted majority game
v [65; 48,37, 27,17].
Therefore, the strategic situation given by means of
the set of minimal wining coalitions
W
m
(v) = {{1,2},{1, 3},{1, 4},{2,3,4}},
shows that players 2, 3 and 4 are symmetric in v, and
the multilinear extension of v is
f (x
1
,x
2
,x
3
,x
4
) =x
1
x
2
+ x
1
x
3
+ x
1
x
4
+
x
2
x
3
x
4
x
1
x
2
x
3
x
1
x
2
x
4
x
1
x
3
x
4
.
A main feature of the Madrid Assembly issued from
the elections was the absence of a party enjoying ab-
solute majority, so a coalition government was ex-
pected to form. We will not try to give here a full
description of the complexity of the Madrid politics.
We wish only to state that the politically most likely
coalitions to form, and the corresponding coalition
structures to the analysis of which we will limit our-
selves, were clearly the following:
PP + C’s, the “right”–wing majority alliance:
B
R
= {{1, 4},{2},{3}}.
PSOE + Podemos + C’s, the “left”–wing majority
alliance: B
L
= {{1}, {2,3,4}}.
We would like to analyze these two situations. Of
course, our main interest will center on the strategic
possibilities of party 4 (C’s), whose position is crucial
in the two–alternative scenario we are considering.
A classical approach to study the problem would
consist in using either (a) the Shapley value and the
Owen value, (b) the Banzhaf value and the Owen–
Banzhaf value, or (c) the Banzhaf value and the sym-
metric coalitional Banzhaf value, in order to evaluate
the strategic possibilities of each party under both hy-
potheses. The results are given in Table 1, where (–)
means no coalition formation, (R) means that PP +
C’s forms, and (L) means that PSOE + Podemos +
C’s forms.
According to (a), C’s gets the same profit in both
alliances. The same conclusion is obtained according
to (b). Instead, according to (c), C’s would strictly
prefer joining PSOE and Podemos instead of PP.
Moreover, by symmetry, the power of C’s when there
is not a coalition formation coincides with the power
of PSOE. According to (a), when the “right”-wing al-
liance is formed, the outside parties are reduced to a
null position and the power of C’s increases regard-
ing to the initial power in v. The same happens when
the “left”–wing alliance is formed
As we have seen, in the present Legislature, stud-
ied here, in order to form a government coalition
the role of C’s was crucial. Thus, C’s was faced to
the dilemma of choosing among either a a “left”–
wing majority coalition with PSOE and Podemos or a
“right”–wing majority coalition with PP, which was
finally formed in 2015.
Example 4.2. We consider here the Andalusian Par-
liament (legislature 2015–2019).
Five parties elected members to the Andalucia
Parliament (109 seats) in the elections held on 22
March 2015. The seat distribution of the parties are
as follows.
1: PSOE (Partido Socialista Obrero Espa
˜
nol), mod-
erate left–wing party: 47 seats.
2: PP (Partido Popular), conservative party: 33
seats.
3: Podemos, radical left–wing party: 15 seats
4: C’s (Ciudadanos), Spanish nationalist liberal
party: 9 seats.
5: IULV–CA, Coalition of eurocommunist parties,
federated to Izquierda Unida, and ecologist
ICORES 2019 - 8th International Conference on Operations Research and Enterprise Systems
50
Table 2: Classical measures of power in the Andalusian Parliament, legislature 2015–2019.
(a) (b)
(–) (R) (L) (N) (–) (R) (L) (N)
1. PSOE 0.5000 0.3333 0.6666 0.6666 0.7500 0.5000 0.7500 0.7500
2. PP 0.1666 0.1666 0.0000 0.0000 0.2500 0.2500 0.0000 0.0000
3. Podemos 0.1666 0.3333 0.3333 0.0000 0.2500 0.5000 0.2500 0.0000
4. C’s 0.1666 0.1666 0.0000 0.3333 0.2500 0.2500 0.0000 0.2500
5. IULV–CA 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
groups : 5 seats
Under the standard absolute majority rule, and
assuming voting discipline within parties, the struc-
ture of this parliamentary body can be represented by
the weighted majority game
v [55; 47,33, 15,9, 5].
Therefore, the strategic situation is given by
W
m
(v) = {{1,2},{1, 3},{1, 4},{2,3,4}},
so that players 2, 3 and 4 are symmetric in v, player 5
is null and the multilinear extension of v is
f (x
1
,x
2
,x
3
,x
4
) =x
1
x
2
+ x
1
x
3
+ x
1
x
4
+
x
2
x
3
x
4
x
1
x
2
x
3
x
1
x
2
x
4
x
1
x
3
x
4
.
As in the same case as the Madrid Assembly, a coali-
tion government was expected to form. We will ana-
lyze the following alliances:
PP + C’s, the “right” alliance: B
R
=
{{2,4}, {1},{3}, {5}}.
PSOE + Podemos, the “left”–wing majority al-
liance: B
L
= {{1, 3},{2},{4}, {5}}.
PSOE + C’s, the “neutral”–wing majority al-
liance: B
N
= {{1, 4},{2},{3}, {5}}
To study the problem we will use either (a) the Shapley
value and the Owen value and (b) the Banzhaf value
and the Owen–Banzhaf value, in order to evaluate the
strategic possibilities of each party under the three
hypotheses. The results are given in Table 2, where
(–) means no coalition formation, (R) means that PP
+ C’s forms, and (L) means that PSOE + Podemos
forms and (N) that PSOE + C’s forms.
In general, we can conclude that the formation of
a two-person coalition block is favorable for its mem-
bers and, especially, for the one that was obtained a
fewer number of seats.
5 CONCLUSIONS
We have obtained the allocations to the players ac-
cording to solution concepts modified by coalition
structures following a double-level procedure based
on the multilinear extension of the game. The two
levels are (i) the modification of the multilinear ex-
tension according to the quotient game and (ii) the
weighting of each product in the modified multilinear
extension according to the solution concept that we
want to compute.
This procedure has been effective for the compu-
tation of allocations according to the Owen value and
the Owen–Banzhaf value by means of simple steps.
In this way, the calculus in Section 4 of several situ-
ations in two territorial Spanish Parliaments has been
easy to compute.
ACKNOWLEDGEMENTS
Research supported by Grant SGR 2017–758 of the
Catalonia Government (Generalitat de Catalunya)
and Grant MTM 2015–66818P of the Economy and
Competitiveness Spanish Ministry.
REFERENCES
Alonso, J. M., Carreras, F., and Fiestras, M. G. (2005).
The multilinear extension and the symmetric coalition
banzhaf value. Theory and Decision, 59:111–126.
Alonso, J. M. and Fiestras, M. G. (2002). Modification of
the banzhaf value for games with a coalition structure.
Annals of Operations Research, 109:213–227.
Amer, R., Carreras, F., and Gim
´
enez, J. M. (2002). The
modified banzhaf value for games with a coalition
structure: an axiomatic characterization. Mathemat-
ical Social Sciences, 43:45–54.
Amer, R. and Gim
´
enez, J. M. (2003). Modification of semi-
values for games with coalition structures. Theory and
Decision, 54:185–205.
The Owen and the Owen-Banzhaf Values Applied to the Study of the Madrid Assembly and the Andalusian Parliament in Legislature
2015-2019
51
Aumann, R. J. and Dr
`
eze, J. (1974). Cooperative games
with coalition structures. International Journal of
Game Theory, 3:217–237.
Carreras, F. (2004). α–decisiveness in simple games. The-
ory and Decision, 56:77–91.
Carreras, F. and Gim
´
enez, J. M. (2011). Power and poten-
tial maps induced by any semivalue: Some algrebraic
properties and computation by multilinear extension.
European Journal of Operational Research, 211:148–
159.
Carreras, F. and Magana, A. (1994). The multilinear exten-
sion and the modified banzhaf–coleman index. Math-
ematical Social Sciences, 28:215–222.
Carreras, F. and Magana, A. (1997). The multilinear ex-
tension of the quotient game. Games and Economic
Behavior, 18:22–31.
Carreras, F. and Puente, M. A. (2011). Symmetric coali-
tional binomial semivalues. Group Decision and Ne-
gotiation, 21:637–662.
Carreras, F. and Puente, M. A. (2013). Coalitional multino-
mial probabilistic values. European Journal of Oper-
ational Research, forthcoming.
Freixas, J. and Puente, M. A. (2002). Reliability impor-
tance measures of the components in a system based
on semivalues and probabilistic values. Annals of Op-
erations Research, 109:331–342.
Harsanyi, J. C. (1959). A bargaining model for the cooper-
ative n–person games. In Contributions to the Theory
of Games IV, pages 325–356. A.W. Tucker and R.D.
Luce (Editors), Princeton University Press.
Owen, G. (1972). Multilinear extensions of games. Man-
agement Science, 18:64–79.
Owen, G. (1975). Multilinear extensions and the Banzhaf
value. Naval Research Logistics Quarterly, 22:741–
750.
Owen, G. (1977). Values of games with a priori unions. In
Essays in Mathematical Economics and Game The-
ory, pages 76–88. R. Henn and O. Moeschelin (Edi-
tors), Springer-Verlag.
Owen, G. (1982). Modification of the banzhaf-coleman in-
dex for games with a priori unions. In Power, Voting
and Voting Power, pages 232–238. M.J. Holler (Edi-
tor), Physica-Verlag.
Owen, G. and Winter, E. (1992). Multilinear extensions and
the coalitional value. Games and Economic Behavior,
4:582–587.
Puente, M. A. (2000). Aportaciones a la representabilidad
de juegos simples y al c
´
alculo de soluciones de esta
clase de juegos (in spanish). In Ph.D. Thesis. Techni-
cal University of Catalonia. Spain.
Shapley, L. S. (1953). A value for n-person games. In Con-
tributions to the Theory of Games II, pages 307–317.
H.W. Kuhn and A.W. Tucker (Editors), Princeton Uni-
versity Press.
ICORES 2019 - 8th International Conference on Operations Research and Enterprise Systems
52