If you own a car, you are not too happy when gas taxes go up: it is more money out of you pocket, however you benefit from a reduction in congestion and lower pollution as long as this tax increase also implies a reduction in gas consumption. If you do not own a car, you would view positively the increase in gas taxes, as the state can now provide more services or reduce other taxes you may be paying. However, some goods with a high share of transportation costs may be become more expensive. Does this reasoning make sense in a political equilibrium, i.e., is the level of gas taxes determined by whether the median voter is a car driver or not?
Fay Dunkerley, Amihai Glazer and Stef Proost show that it does and figure out that in the OECD a car driving median voter leads to a gas tax that is 20% lower than a walking median voter. Of course any such estimate is fraught with endogeneity: the median voter is walking because the tax is high. To overcome this problem, the authors use a dynamic setup that takes into account when a median voter starts to drive.