Jørgen Modalsli

@modals.li

1540 Followers 888 Following

Professor, Oslo Metropolitan University. Economics, economic history https://modals.li

Image of the river, from the city centre.
Female labor force participation in historical census microdata By: Jørgen Modalsli ABSTRACT How reliable is historical microdata? Understanding historical labor force participation is crucial for assessing long-term trends in economic development and intergenerational mobility. Most existing historical studies are, however, limited to men, and little is known about how reliable quantitative historical sources are when studying labor market outcomes for women. This paper documents that the measurement of women's economic activity in the 1910 Norwegian population census had a high level of consistency. There is extensive discussion of measurement issues in historical census reports, micro data can reproduce historical census tables with a high degree of accuracy, and other contemporary reports such as industrial censuses and tax statistics confirm the results found in the census. In addition, a double-enumeration feature of the Norwegian census is leveraged to assess consistency between enumerators, finding no indication that precision in the occupational classification of women is any lower than for men. Some potential sources of downward bias are found in the historical census microdata set provided by IPUMS. Based on the results in this paper, historical census data appears well suited to study economic activity using modern econometric methods, for women as well as men. A slight upward revision of the 39\% female labor force participation in Norway in 1910 might be in order.
This paper presents estimates of income concentration and inequality for Norway using a new comprehensive measure of income, which identifies business income as it is earned by companies rather than when it is paid out as dividends to owners. We assemble several sources of high quality register data that allow us to account for multiple layers of business ownership across all companies between 2001 and 2018. Compared to official statistics, the new measure implies that the share of income attributable to the top 1\% of the distribution more than doubles and the Gini coefficient estimates increase by about 40\%. Our new measure identifies substantial tax regressivity for individuals in the top percentile, a feature that cannot be detected by standard income measures. For instance, while the share of gross income paid in taxes by individuals at the 99th percentile is about 36\% in 2016, the corresponding share paid by individuals in the top 1\% is 19\%.