Agrar ojämlikhet i Sverige studerad med bevillningstaxeringen som källa, 1870–1920

This Swedish-language investigation concerns the usefulness of the Swedish property tax, introduced in 1862, as a source for studying wealth inequality in the agrarian sector. A one percent sample of all taxpayers in rural areas, following the original taxation lists archived at the National Archives in Stockholm, has been made for four counties in three years: 1870, 1900, 1920. The counties are Malmöhus, a wealthy but polarized area in the south, Älvsborg, a farmer-dominated area in the west, Jönköping, a farmer-dominated area in the centre-south, and Västmanland, an industrialized area in central-northern Sweden. The data encompass 1,477 taxpayers in 1870, 2 499 taxpayers in 1900, and 3 219 taxpayers in 1920. In total, 7 195 taxpayers. The investigation shows that inequality was quite high in the agrarian sector in Sweden in this time. In 1870, the ten percent of taxpayers with the most real estate owned 82.2 percent of taxed real estate in Malmöhus, 53.2 percent in Älvsborg, and 49.7 percent in Västmanland. Inequality declined in Malmöhus and Älvsborg to 1900, when the top decile shares were 62.7 and 43.7 respectively, but rose in Västmanland. In Jönköping, where the sample is too small in 1870, the top decile share in 1900 was 44.0 percent. Inequality was relatively stable from 1900 to 1920: it grew in Älvsborg, but decreased in Västmanland, and was stable in Malmöhus and Jönköping. When all four counties are taken together, it reinforces the picture of especially high inequality in 1870: the top decile’s share measured on the level of all four counties fell from 74.4 percent in 1870 to 59-60 percent in 1900 and 1920. The Gini coefficient for rural wealth in the four counties is estimated as 81.1 in 1870, 70.7 in 1900 and 70.4 in 1920. A decomposition shows that within-county and between-county inequality are both important parts of inequality as a whole. In an appendix, six further counties are added for the year 1900, taking the sample size for this year to 8,053. The results, in terms of inequality, with ten counties are quite similar to those with four. The text discusses the source value of the tax data and how the findings should be interpreted – for example, only those who owned real estate were included, which means that those completely without real estate are not included in the present investigation.