Geographical Redistribution With Disproportional Representation: A Politico-economic Model Of Norwegian Road Projects

摘要

Politicians bias public policies to favor particular election districts. According to the traditional common pool model, districts facing low tax shares should receive relatively large government projects. We suggest a swing-voter model where the number of voters on the ideological cut point, lack of party identification and number of district representatives per voter determine project sizes. We analyze the allocation of state road investments in Norway from 1973-1997 exploiting unique data on characteristics of voters, legislative representation and tax prices in 19 election districts. Geographical representation to parliament is biased, mostly due to an ancient constitution. Shares of swing voters and legislative over-representation lead to higher levels of road investments, while high levels of party identification reduce investments.