A further consideration of coalitions under the demand-revealing process

✍ Scribed by Melville L. McMillan

• Publisher: Springer US
• Year: 1983
• Tongue: English
• Weight: 182 KB
• Volume: 40
• Category: Article
• ISSN: 0048-5829
• DOI: 10.1007/bf00118525

✦ Synopsis

Riker (1979)

questions the superiority of the demand-revealing process arguing that nonisolated voters may form coalitions to bribe other voters and so manipulate the outcome. Even if the successful alternative is unchanged, the distribution of the gains and losses will be modified. The Tideman and Tullock (1981) response focuses on the large number case (the only one in which significant disagreement with their position appears) and they demonstrate that the manipulation effort is likely to fail because those bribed have the incentive to freeride, i.e. vote valuations contrary to the agreement because of the possibility of being a critical voter with the burdensome tax consequences. Both contributions relate to bribery where the outcome is complicated by informational, organizational and enforcement problems not to mention legal and moral considerations. The point of this note is that in large number situations a coalition may have the potential to manipulate the outcome without resorting to bribery. This neglected possibility expand,s somewhat the opportunities for successful coalition. Consider the large number example discussed by Riker and by Tideman and Tullock. The situation is outlined in Riker's Table 9 which, as in Tideman and Tullock, is reproduced here for reference as Table 1. In the bribery case these authors discuss, the 100 members of classes 3b and 3c assess themselves $50,000 each for a bribery fund which will be used to pay the 10 members of classes lc and ld $500,000 apiece providing issue x wins. The bribe would be offered along with instructions that each member vote $10 million for x and nothing for y in which case x would win and no one would be required to pay a tax. The Riker-Tideman and Tullock debate turns on the incentives of those bribed to freeride so as to reduce the risk of a tax.

It is easily demonstrated that bribery is unnecessary for the members of 3b and 3c to defeat y and succeed in x winning. A coalition of those persons alone is all that is required. For example, if that 100 agreed among themselves to vote $500,000 for x rather than their true valuation of $200,000, x would win * This note benefitted from the comments of L. S. Wilson and the editor.