Stability of jurisdiction structures under the equal share and median rules (Q2467519)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Stability of jurisdiction structures under the equal share and median rules
scientific article

    Statements

    Stability of jurisdiction structures under the equal share and median rules (English)
    0 references
    0 references
    0 references
    0 references
    22 January 2008
    0 references
    The paper considers \(N\) agents with symmetric single peaked preferences over the unit interval \(I\) of the real line representing alternative choice of locations for public sector projects. Denote the point of the peak of the agent \(i\)'s preference by \(l^{i}\). \ The agents can be partitioned into jurisdictions, each of which has to choose a location \(l\) of its project in \(I\) the cost of which, \(g,\) is exogenously given and assumed to be equally shared within the individuals within that jurisdiction (ES rule). In addition, each agent incurs a linear transportation cost \(d(l^{i},l) =| l-l^{i}|\). Let \(M(S)\) denote the set of median project locations for a jurisdiction \(S\subset \{1,\dots,N\}\) and also assume that the project is located at the mean of the extreme point of the median set (MM rule). The two rules define a hedonic game, i.e., once a jurisdiction \(S\) is formed the total cost for each of its members \(c_{i}(S)\) is uniquely determined. Let \(P=\{S_{1},\dots,S_{k}\}\) be a partition of the set of agents each \(S_{k}\) denoting an individual jurisdiction within the jurisdiction structure \(P\). A jurisdiction \(S\subset N\) blocks \(P\) if \(c_{i}(S) <c_{i}(S_{k(i)})\) for all \(i\in S\), where \(k(i)\) is the number of the jurisdiction in \(P\) that contains \(i\). \(P\) is called core stable if it cannot be blocked. \(P\) is called Nash stable if \(c_{i}(S_{k(i)}) \leq g\) and \(c_{i}(S_{k(i)}) \leq c_{i}(S_{k}\cup \{i\})\) for each \(i\) and all \(S_{k}\in P\). A jurisdiction \(S\subset N\) is consecutive if for all \(i,k\in S,l^{i}<l^{k},\) every agent \(j,\) with \(l^{i}<l^{j}<l^{k},\) also belongs to \(S\). Examples are produced to show that neither Core stable nor Nash stable jurisdiction structures may exist in such hedonic games. However, if the society \(N\) is equidistant, i.e., there exists \(l>0\) \(s.t.\) \(l^{i}-l^{i-1}=l\) for \(i=2,\dots,N\), both Core stable as well as Nash stable jurisdictions exist but they need not be the same; they are quite distinct solution concepts. Nash stable structures, if they exist, are necessarily consecutive. \ By contrast Core stable structures need not be so, however, they are when the society is equidistant. The paper goes on to explore similar issues in quasi-hedonic game structures where more than one median location exists and one of them is arbitrarily selected.
    0 references
    public good
    0 references
    jurisdiction structures
    0 references
    hedonic games
    0 references
    core stable
    0 references
    Nash stable
    0 references

    Identifiers