Single-peaked preferences

Single-peaked preferences
Name	Single-peaked preferences
Field	Social choice theory
Introduced	1960s
Notable	Duncan Black, Anthony Downs, Kenneth Arrow

Single-peaked preferences are a structured class of preference profiles used in Social choice theory and Welfare economics to capture situations where agents have a most-preferred option and decreasing preference for alternatives farther from that peak. They formalize a unidimensional ordering on an axis such as a policy spectrum or a commodity attribute, reducing strategic complexity in collective decision problems and enabling existence and characterization results for aggregation rules and equilibria. Single-peakedness underpins classical results by figures like Duncan Black, Kenneth Arrow, and Anthony Downs and connects to later algorithmic and computational studies by researchers associated with institutions such as MIT and Princeton University.

Definition and formal properties

A preference profile is single-peaked relative to a linear order over alternatives if for each agent there exists a most-preferred alternative (a peak) and preferences decline as alternatives move away from that peak along the order. Formal treatments appear in works by Duncan Black, Kenneth Arrow, and Anthony Downs and are formalized using concepts from Ordinal utility theory and Domain restrictions (social choice). Key properties include the medians voter theorem, the absence of Condorcet cycles under single-peakedness, and the representation of preferences by unimodal utility functions. Mathematical formalizations rely on total orders on the set of alternatives and consider profiles that satisfy the single-peakedness condition for all agents; these profiles are closed under certain operations studied in texts from Cambridge University Press and Oxford University Press.

Examples and intuition

Classic examples place alternatives on a left–right policy line as in analyses by Anthony Downs, where voters have ideal points and prefer policies closer to their ideal. Other canonical settings include choices over tax rates, public spending levels discussed by scholars at London School of Economics, and location models like the Hotelling model used in industrial organization studies at Stanford University. Intuitively, imagine a spatial spectrum with alternatives ordered as in debates at the United Nations General Assembly; each voter's ranking rises to a single peak and then falls. Empirical illustrations appear in case studies involving elections analyzed at Harvard University, preference surveys from Pew Research Center, and laboratory experiments conducted at Max Planck Institute and University of California, Berkeley.

Aggregation and voting implications

Single-peaked domains guarantee the existence of a Condorcet winner—often the median of voters' peaks—which is central to the median voter theorem associated with Duncan Black and formalized by Anthony Downs. Voting rules such as majority rule, studied in contexts at Yale University and Columbia University, exhibit strategyproofness or manipulability characteristics influenced by single-peakedness; for instance, certain strategyproof social choice functions exist on single-peaked domains as shown in work by scholars at Princeton University and University of Chicago. The structure prevents classic impossibility results like scenarios in Kenneth Arrow's theorem from arising when the domain is restricted, a point explored at European University Institute and in research by the Econometric Society.

Extension and generalizations

Generalizations include single-dipped preferences, multi-peaked or double-peaked preferences examined by researchers at University of Oxford, and single-peakedness on trees, circles, or more general graphs studied at Massachusetts Institute of Technology and Carnegie Mellon University. Concepts such as single-crossing preferences, introduced in literature associated with MIT and Princeton University, relate closely and are compared in monographs from Oxford University Press. Multidimensional extensions examine when multidimensional policy spaces retain analogous stability properties, a topic pursued at Stanford University and in collaborative projects with the National Bureau of Economic Research.

Algorithmic recognition and complexity

Determining whether a profile is single-peaked has algorithmic solutions drawing on PQ-tree algorithms and graph-theoretic methods developed by computer scientists at Carnegie Mellon University and University of Toronto. Recognition can be done in polynomial time; related complexity questions involve maximum single-peaked subsequence problems, manipulation and control for voting rules, and hardness results linked to complexity classes studied at Princeton University and ETH Zurich. Implementations and computational experiments appear in software from research groups at University of Washington and in libraries authored by members of Association for Computing Machinery conferences.

Applications in economics and political science

Applications span public choice models from Chicago School of Economics analyses, models of electoral competition in textbooks by Anthony Downs, collective decision-making in parliaments such as the United Kingdom Parliament and legislatures studied at UCLA, and market location problems inspired by Hotelling. Empirical work on policy preference aggregation uses single-peaked assumptions in studies by Pew Research Center, European Commission policy analysis, and research teams at Brookings Institution. The concept also informs mechanism design in auctions and matching markets investigated at Harvard University and Caltech.

Category:Social choice theory