We begin with a straightforward observation.
This condition immediately follows from the definition of social influence structure. If is an influence structure that expresses , it must be the case that , for all and for all . If for some , it is the case that for some , we would have and , which violates the definition of social influence structure. Given any collection, , satisfying the condition in (1), we can define , which expresses it as follows: begin by letting and , and then, for each and i in x, set . If , we are done; otherwise, for each minimal element, x of, , set and . Given that satisfies the condition in (1), this collection of functions will be a well-defined influence structure. It is clear that there are many influence structures which express a given collection of subsets, .
3.1. Increasing Influence Structures
One very useful result that applies to our setting from the general theory of games of strategic complements is the following (see Zhou (1994) [
8] and the introduction):
If is an increasing influence structure, then forms a lattice (with respect to the set containment (⊆) partial order). (2)
(Note: completeness is automatic, because our lattices are always finite)
We present a proof of this statement, as it showcases an argument that frequently arises when working within this classes of games of influence.
If is empty or a singleton, then it is immediately true that it is a lattice. Therefore, suppose that there exist , such that . Note that is nonempty. If for all j, then . If this is not the case, then we can remove the elements , such that . We can proceed iteratively, until we stop at a set, , such that for all . By construction and the fact that is increasing, it is true that for all and, thus, . Furthermore it must be the case that . This will surely be the case, since at no point in the process do we remove any elements in . The reason is that for all given that x and are equilibria and is increasing, so inductively, at each stage of the process, we have that is contained in the set under consideration. It follows that .
To see that exists, suppose it does not. Then, it must be the case that the set has at least two minimal elements, z and . Then, some set, w, , must be an equilibrium: if , then either or . Without loss of generality, suppose that . Then, as z is an equilibrium, it must be the case that and, because and is increasing, we must also have . On the other hand, if , then , because i belongs to either x or , is increasing and both x and are equilibria. Therefore, starting at , we can iteratively remove the elements that prefer not to be active from the set. Due to the fact that is increasing, it will be the case that elements not belonging to the set prefer not to be active at each iteration. Additionally, at some point, before reaching or at , it will also be the case that all elements in the set will prefer to be in the set. This set is the equilibrium, w, that we were looking for, and as and , its existence contradicts the minimality of z and . We conclude that the set, , only has one minimal element; the minimum we were looking for. With an analogous argument, we can show the existence of the meet .
Given a lattice,
, how can we know whether it is expressible by an increasing influence structure? The sparseness condition (SC) in the following proposition is clearly necessary, as it just demands that the minimal restrictions imposed on the influence structure by the sets in
are consistent with it being increasing. The work goes into showing that it is also sufficient.
P
ROPOSITION 1 There exists an increasing influence structure expressing a lattice, , if and only if :
Sparseness Condition: , such that , we have .
Proof of Proposition 1: Necessity: Suppose condition (1) does not hold. That is, assume that there exist and , such that , and . However, then any peer influence structure inducing the game satisfies and , so is not increasing.
Sufficiency: We begin by setting for all i and and, then, modify it in two steps. We first modify it to guarantee that it induces as all elements of reach equilibria, and then, we tweak it in order to remove all unwanted elements, .
(S1) Induce all required equilibria:
For each and , let for .
Note that, in step (S1), we construct an increasing peer influence structure, and by virtue of , each is an equilibrium of the game that it induces. By construction, if , then . Moreover, if , then . To see this, note that to have , it needs to be the case that for some , such that , but by , this would imply , which by assumption is false.
The issue is that
, where
is as defined in (S1), may have other equilibria. To see this in a simple example, consider the influence structure that results from the application of (S1) to the lattice satisfying
depicted on the left in
Figure 1 and note that it actually expresses the lattice shown on the right. (As it generally seems to be the case when attempting to construct games that express a given set of equilibria, it is easy to guarantee that the members of the set are all equilibria of the game, but harder to guarantee that those are the only equilibria.)
We now remove all the unwanted equilibria, and this can be done thanks to the fact that the set in question is a lattice. Therefore, let denote the influence structure constructed in (S1) and , the game that it induces.
(S2) Remove all unwanted equilibria: We remove the unwanted equilibria iteratively and prove that the procedure does not damage any of the desired equilibria by induction.
The base case:
If , we are done. Therefore, suppose there exists . Note that implies for each . By construction, in (S1), can only hold if there is , such that and . Therefore, it follows that:
where , and . (*)
where the strict containment (⊂) follows from the fact that .
Figure 1.
The figure on the right is the lattice expressed by the game induced by the influence structure constructed by applying (S1) to the lattice on the left.
Figure 1.
The figure on the right is the lattice expressed by the game induced by the influence structure constructed by applying (S1) to the lattice on the left.
The key in what follows is that (*) implies that the unique successor of y in is , where the ’s are the elements of described in (*). To see that this is the case, first note that (the smallest upper bound of in ) is well defined by virtue of the fact that is a lattice. The fact that is a successor of y in follows from the fact that if there existed , , then we would have for all , contradicting the fact that is the smallest upper bound of in . is the unique successor, because any other , such that , is an upper bound for and, therefore, .
We will now show that . The definition of only makes adjustments to on sets, . Therefore, if and , then (as ). Therefore, consider some , such that . Any such z must contain , as any set containing , and the only possible difference between and can be on components ( vs. ) involving elements i in . However, implies for any such element, i. Therefore, actually no change was really made to the function in z.
We now show that . By construction of , , and we will show that no other element that was not in can now be part of . As above, we only need to be concerned with sets . If , then, as seen above, and are identical, so the only possible occurrence of a new equilibrium in must involve sets , such that . Note, however, that any such set must lack some element , and for such an element, , and therefore, cannot be an equilibrium of .
The inductive step:
Assume that for , we have , such that for each m, , and . If , then we are done. Otherwise, suppose that there exists . By assumption for all , and therefore, . However, this means that above holds, and we can therefore construct from just as we constructed from . The augmented sequence has the property that for each m, , we have and .
Note that the construction of
, satisfying
in the proof of Proposition 1 only relies on the fact that
satisfies (SC) and, in particular, does not depend on
’s lattice structure. Therefore, the following corollary immediately follows.
C
OROLLARY 1 A set, , is weakly expressible by an increasing influence structure, , if and only if it satisfies (SC). Furthermore, , as constructed in based on and which we denote , is the smallest increasing influence structure weakly expressing (according to the order defined in Section 2). Corollary 1 shows that, as long as is satisfied by a given set of patterns of behavior, there is some increasing influence structure inducing a game whose set of equilibria contains . The fact that is the smallest such influence structure follows from the fact that in (S1) in the proof of Proposition 1, we begin with for all i and x and, then, only make the adjustments that are strictly necessary to guarantee . It is worth pointing out that even in cases in which we are just concerned with whether some binary game of strategic complements weakly expresses a set of behavior patterns, Proposition 1 is very useful. In particular, it tells us that if some satisfying (SC) has indeed been generated by an increasing influence structure, then the complete set of equilibria of the game must be in the set is a lattice satisfying (SC)}. We can thus know which other yet unobserved behaviors we can expect and which behaviors to look out for in order to further narrow down our understanding, by falsifying some of the models in is a lattice satisfying (SC)}.
3.2. Increasing Influence Structures That Admit a Network Representation
An influence structure, , admits a network representation when for each agent i, there exist weights , , and a threshold , such that , if and only if, . Without loss of generality, we will restrict attention to cases in which . Throughout this section, we will also restrict attention to collections of subsets of N, (and lattices ), such that (). (If this condition is violated for some i, this means that either (1) i sets regardless of what other agents do; or (2) i sets regardless of what other agents do. In both cases, we can remove some agents from the game and adjust the thresholds of the remaining ones, obtaining a game on a set, , of agents with an equivalent set of equilibria. If (1) holds, then we can remove i without doing any adjustments. If holds, then we remove i and let for all the remaining agents, and if , then we remove j and continue iterating, until we only have agents with strictly positive thresholds. The argument for restricting attention to and lattices , such that or , is analogous.) We can group all the individual weights in a matrix, W, and let denote the vector of thresholds. As W and fully capture an influence structure, we will directly denote it by , instead of using . The games induced by increasing structures that admit a network representation are well known in the literature as graphical games of strategic complements or games of thresholds. Throughout this section, we refer to the game induced by as a game of thresholds and denote it by . Given a lattice, , we say that it is (weakly) expressible by a game of thresholds if there exist some weights, W, and a vector of thresholds, , such that .
As seen in Example 1, there are some increasing influence structures that do not admit network representations.
E
XAMPLE 1 An influence structure that does not admit a network representation.
Let be an increasing influence structure on , such that , if and only if or . This influence structure does not admit a network representation: The reason is that implies that either or . Similarly, it must be the case that either or . However, this, in turn, implies that at least one of , , or must also be one. ⋄
Example 1 shows that if we want agent 1 to be triggered by groups of agents and , while depicting his incentives using weights and a threshold, then it must be the case that he is also triggered by at least one other group of agents that is not a superset of either of these. In general, this is the only kind of limitation that we encounter when constructing network representations of influence structures: it is straightforward to assign the weights and pick the threshold in order to have an agent prefer to be active when every element in a specified collection of subsets of N is active. What can be difficult and sometimes impossible is choosing them in order to ensure that these are the only triggers.
We now turn to the question of which sub-collection of lattices expressible by increasing influence structures are also expressible by games of thresholds. There are two different kinds of problems that we may face in trying to express a given lattice. The first one is that due to the approximation limitations seen in Example 1, we may gain some equilibria, which we cannot destroy without compromising the equilibria that we do want to include. The other possibility is that none of the approximations expresses supersets of the lattice in question. Example 2 shows that the collection of lattices expressible by games of thresholds is a strict subset of the collection of lattices expressible by increasing influence structures.
E
XAMPLE 2 A set of equilibria of an increasing influence structure that cannot be expressed by a game of thresholds.
Let
be an increasing influence structure on
, such that
,
,
,
,
,
,
,
,
. The set of equilibria of the associated game of influence is shown in
Figure 2. This lattice is not expressible by any game of thresholds. The reason is that
and
are equilibria, and therefore,
. This means that either
or
, so it cannot be the case that
and
are both equilibria. Agent 1 would strictly prefer to join at least one of them. ⋄
Figure 2.
The set of equilibria of .
Figure 2.
The set of equilibria of .
3.2.1. Relation to Monotonic Simple Games
We begin this section with a few definitions from the literature on simple and weighted games that we use in the subsequent analysis.
A simple game on N is a collection of coalitions, with the property that and . A simple game is monotonic if for each pair , such that . (More generally, simple games are usually defined over arbitrary algebras of subsets of N.) The subsets of N that belong to are called winning coalitions, and the subsets belonging to are called losing coalitions.
Note that any given increasing influence function, , for , with the property that and corresponds to the monotonic simple game, , on in which , if and only if . In these terms, an increasing influence structure is just a collection of monotonic simple games.
A simple game, , on N is said to be weighted if there exists a collection of weights , for all j and a threshold q, such that for all coalitions , , if and only if . (Note that all weighted simple games are monotonic by definition.) Taylor and Zwicker provide a characterization of the set of simple games that are weighted.
We say that a sequence of elements,
(not necessarily different), of
can be trade-transformed into a sequence of elements,
(not necessarily different), if
for all
.
D
EFINITION 1 (Trade robustness) A simple game is k-trade robust, if it is not possible to trade-transform k winning coalitions (not necessarily different) into k losing coalitions. The game is trade robust if it is k-trade robust for all values of k.
With the above definition in mind, we can enunciate one of Taylor and Zwicker’s main theorems.
T
HEOREM 1 (Taylor and Zwicker 1992)) A simple game is weighted, if and only if it is trade robust.
Any set of behaviors automatically implies some minimal restrictions on the collection of winning coalitions and losing coalitions of the simple game representing each agent’s influence structure. For example, the lattice shown in
Figure 2 requires
,
,
,
,
,
,
,
,
and
.
Given a collection,
, satisfying
and some
, let
denote the collection of winning coalitions in
describing
i’s influence structure in the game
defined in Corollary 1 of Proposition 1.
therefore satisfies that
, if and only if
,
and
. It also follows from Corollary 1 that
is a monotonic simple game. (Note that
and
, given that we only consider collections
, such that
and
. Coupled with
,
implies
and, therefore,
.) The following proposition, based on Taylor and Zwicker’s characterization, tells us exactly when it is the case that a collection,
, is weakly expressible by a game of thresholds.
P
ROPOSITION 2 There exists a game of thresholds weakly expressing a collection, , if and only if satisfies and is trade robust for each .
Proof of Proposition 2: Sufficiency follows immediately by applying Theorem 1 (Taylor and Zwicker) to the influence structure, .
Necessity: If is weakly expressible by a game of thresholds, then by Corollary 1, we know that it satisfies (SC). By definition, there exists , such that . Define by letting and for each , if for some and otherwise. Then, . To see this, let , and suppose that ; then, by definition of , . If , then . So weakly expresses . Furthermore, for each i, is the simple game generated by and . Since it is a weighted game it follows from Theorem 1 (Taylor and Zwicker) that it must be trade robust. ▐
It follows from Proposition 2 that a necessary condition for a collection of sets
satisfying (SC) to be weakly expressible by a game of thresholds, is that
is trade robust.
E
XAMPLE 3 Going back to example 2, note that is not trade robust, since we can trade-transform the winning coalitions for agent 1, and into the losing coalitions, , . As shown above, is not weakly expressible by a game of thresholds. Note, however, that facing the question of whether , is weakly expressible by a game of thresholds, examining the trade robustness of is not enough, and we need to look at the trade robustness of the simple games, . ⋄
3.3. Games of Thresholds on the Complete Graph
We end by examining the problem of expressing
using games of thresholds on the
complete graph, that is, the graph in which the weights
are the same for all pairs of agents,
. This special case is of interest, because it is equivalent to many models of peer effects in the applied literature (as discussed in the introduction), in which each agent chooses his activity level in response to some measure of mean activity level in the environment. In these models, the agents may have different sensitivities to the environment (in our language, different thresholds), but the environment is the same for everyone, which is analogous to all the weights being equal.
C
LAIM 1 If x and y are equilibria of a game of thresholds in the complete graph, then either or . That is, the set of equilibria must be a chain, and the chain satisfies .
Proof of Claim 1: Suppose that we have two equilibria, x and y. Without loss of generality, assume that , and let . Then, . Therefore, it must be the case that . The fact that the chain satisfies follows from Proposition 1.▐
The sparseness condition (SC) and the strictness condition
of
Section 2 are equivalent in the case of chains and take the simpler form:
, whenever
. The converse of the claim above is also true: any chain that satisfies (SC) is expressible by a game of thresholds on the complete graph.
C
LAIM 2 If is a chain with the property that whenever , then it is expressible by a game of thresholds on the complete graph.
Proof of Claim 2: Let be a chain satisfying the condition of the claim, and without loss of generality, suppose implies . Then, it must be true that . Therefore, let, , , and in general, , , for each . Finally, let , . Let . By construction, the set of equilibria of this game of thresholds is precisely .▐
The simple structure of the lattices that are expressible by games of thresholds on the complete network (see
Figure 3) means that we are able to count the number of different (up to re-labeling of the agents) possible sets of equilibria of these games. In enunciating this counting result, we restrict attention to sets of equilibria that include ∅ and
N.
Figure 3.
An example of a game of thresholds on the complete graphs. The numbers outside the circles are the names of the nodes, and the respective thresholds are shown inside the circles.
Figure 3.
An example of a game of thresholds on the complete graphs. The numbers outside the circles are the names of the nodes, and the respective thresholds are shown inside the circles.
C
OROLLARY 2 There are different chains, which can be expressed as equilibria of games of thresholds on the complete network on the set that include the set N; where denotes the Fibonacci number.
Proof of Corollary 2: Since we are counting the number of different chains up to re-labeling of the agents, two chains, and , are different, if and only if one of them contains a set different in size from all the sets contained in the other chain. Therefore, in what follows, when referring to a given set, we only speak of its size. We can partition the set of expressible chains on agents into two groups: (1) those that contain an element of size ; and (2) those that do not contain an element of size . Each of the chains in (1) corresponds to a unique chain expressible with elements. On the other hand, each of the chains in (2) corresponds to a unique chain expressible with elements. Therefore, if we denote the number of expressible chains on agents by , we have that . Moreover, , .▐
We can compute a conservative lower bound on the collection of structurally different lattices that can be expressed by games of thresholds without restrictions on the weights, by counting the number of different lattices corresponding to different partitions of the set of N agents into completely connected components, each of a size of at least two. (There are no chains on less than two elements satisfying (SC) and the requirement that and .) If we let denote the number of partitions of n into k parts, each of them of a size of at least two, then there are structurally different lattices. Therefore, for example, when , there are only 377 different lattices expressible by games of thresholds on the complete graph, whereas our lower bound for the number of lattices expressible by some game of thresholds is 4,297, slightly over 10 times. When n = 60, there are over 1,000 times as many structurally different lattices expressible by some game of thresholds as there are lattices expressible by some game of thresholds on the complete graph. The lower bound is conservative, since we are only considering what can be expressed ¡by graphs that are the union of complete subgraphs. Claims 1 and 2 show that games of thresholds owe their expressive power to the interplay between thresholds and network structure. No matter how much freedom we have to play with the thresholds, we lose the ability to express anything other than chains, if we restrict all relationships to have the same weight.
A common network representation across applications restricts relations to either have a given positive weight or to have no weight at all (games of thresholds on unweighted, undirected graphs (see Galeotti
et al. (2010) [
24]). As seen in Example 4, this restriction also limits the expressive power of the games of thresholds.
Figure 4.
A lattice, , that can only be expressed by games of thresholds on graphs with weighted links.
Figure 4.
A lattice, , that can only be expressed by games of thresholds on graphs with weighted links.
E
XAMPLE 4 The expressive power added by weights.
The lattice,
, shown in
Figure 4 cannot be expressed by a game of thresholds on a network in which all links have the same weight. To see this, suppose that there existed a game
, in which
or
and such that
. We begin by noting that the agents in
need to be connected, as otherwise, one of them would need to have threshold of zero, but this cannot be the case, since
. Whoever among 2 and 3 is linked to 1 must have a threshold of at most
and, therefore, cannot be linked to
, since, otherwise, at least one of
or
would not be an equilibrium. If
j among 2 and 3 is not directly linked to 1, then
j must have a threshold of at most
w, and just as before, it cannot be linked to
or 7, since, as above, at least one of
or
would not be an equilibrium. We can therefore conclude that neither 2 nor 3 can be linked to
or 7. Therefore
has to be an equilibrium, as well. This contradicts the existence of an unweighted network,
W, and a vector of weights,
, satisfying
. ⋄