More for everyone: The effect of local interests on spending on infrastructure

•I investigate whether politicians which represent narrow geographic interests spend more on local public goods.•I use panel data on representation and policies in the Norwegian regional governments.•I find that investments in roads are lower when politicians represent large populations.•I argue that my empirical application is closer to the theoretical literature than studies on the effects of legislature size.

A central question in political economy is whether decision-making in representative democracies is biased towards local public investments or other types of policies which have locally concentrated benefits. The model by Weingast et al. (1981) predicts that a legislature with members from different areas will spend more in total on local public goods when the geographic constituency of each member is small. I test this prediction using panel data on the 18 Norwegian regional councils, exploiting that the geographic allocation of regional council seats varies considerably over time. Consistent with the theory, I find robust evidence that investments in regional public roads are lower when many council members come from the more populated areas in the region. This gives a more direct test of the prediction by Weingast et al. (1981) than existing empirical evidence, which concerns the relationship between public spending and legislature size. I find similar but less robust results for the maintenance of existing roads.