Strategy-proof location of public bads in a two-country model

We consider the joint decision of placing public bads in each of two neighboring countries, modeled by two adjacent line segments. Residents of the two countries have single-dipped preferences, determined by the distance of their dips to the nearer public bad (myopic preferences) or, lexicographically, by the distance to the nearer and the other public bad (lexmin preferences). A (social choice) rule takes a profile of reported preferences as input and assigns the location of the public bad in each country. For the case of myopic preferences, all rules satisfying strategy-proofness, country-wise Pareto optimality, non-corruptibility, and the far away condition are characterized. These rules pick only border locations. The same holds for lexmin preferences under strategy-proofness and country-wise Pareto optimality alone.