Gender or class - what determines voting? Lessons from expanding the suffrage in early 1900s Norway

Previous literature has found that extending the suffrage to both females and poorer voters increases the supply of public goods. This paper investigates whether the difference in voting between men and women can be explained by differences in income alone, or if there exist gender specific differences in preferences. I exploit two key features of the expansion of suffrage in municipality elections in early 20th century Norway. First, the time at which people gained the right to vote depended on both their gender and their household income. Second, the income threshold for suffrage was set nationally, creating variation across municipalities in the share of new voters following each extension of the suffrage. This variation allows me to estimate separate effects for the change in supply of health personnel following the extension of suffrage to poor men, rich women, and poor women, respectively. I find that the enfranchisement of both poor men and rich women increases the supply of doctors relative to when only rich men had the right to vote. These results are consistent with gender specific preferences for health services to the community.