Ongoing articles
"Taxing workers through the twentieth century"
Abstract: The development of labor taxation is one of the major fiscal features of the 20th century, and one fundamental piece in the growth of contemporary states. This paper asks how different Western countries underwent this transformation. When did average wages start to be subject to direct taxes of different kinds? How were tax payments affected by wage levels and family sizes? Was labor taxation more intense in countries developing more generous welfare states?
I construct and compare series of the tax burden placed on the labor income of workers in France, Spain, Sweden, the United Kingdom, the United States, during the period 1910-1970. The series include (schedular and general) income taxes and social contributions, and are estimated for diverse representative working households, defined in relation to the average wage in the economy. This framework allows connecting with existing series for more recent years at the OECD’s Taxing Wages (which starts for 1979) and other scholarly work.
Results show that labor taxation displayed an inverted U-shape in Sweden and the United Kingdom, together with an increase in latecomer countries, towards significant convergence by the end of the century. Average workers started paying direct income taxes earlier in Sweden, then in the Anglo-Saxon countries (where couples with children were spared for a longer time), and later in Spain. These trends can be related with macroeconomic indicators depicting the development of labor incomes as a tax base: while relative wage levels were higher in Sweden than in the rest of the countries before 1930, the wage share was not – suggesting that there is a room for political explanations along economic ones in understanding the historical development of labor taxation.
"The distributive effects of consumption taxes in early welfare states" (with Oriol Sabaté)
Abstract: This paper presents estimates of the distributive effects of consumption taxes in five Western countries, during the period 1910 to 1970. With this, it aims to contribute to the debate about the welfare state and regressive taxation, by going back to this formative period. We contrast the patterns in five countries representing the different welfare state models: ‘liberal’ (United Kingdom, United States), ‘conservative’ (France), ‘social democratic’ (Sweden), and ‘southern’ (Spain).
Consumption taxes of different kinds (customs, general internal taxes, excises, fiscal monopolies, etc) are obtained from a detailed database (currently under construction). They are then imputed to an income distribution micro-database obtained from previous work (Torregrosa-Hetland and Sabaté, 2021a). The imputations are conducted using the relationship between income and spending in different products, estimated from historical household budget surveys. Distributive effects of consumption taxes are finally combined with those of personal income taxes which we have previously calculated.
"Income tax evasion and inequality in Brazil" (with Monica Calijuri, Carola Pessino, and Andréa Legal)
Abstract: This paper estimates income under-reporting in the Brazilian income tax, using full taxpayer-population microdata from 2019. We follow the econometric approach of Feldman and Slemrod (2007), but modify it according to the characteristics of the Brazilian income tax: instead of charitable donations, we use medical (and educational) expenses as an indicator of taxpayer’s economic capacity. Our study provides the first microdata-based estimation of under-reporting available for a Latin American country.
We obtain three main results. First, income under-reporting of non-labour income in Brazil is considerable, particularly in rents (85%) and urban self-employment (75%) – for movable capital we estimate 27%, unsurprisingly less given that it is subject to lower taxation. Second, average under-reporting of total income is estimated at 16%, decreasing over income levels, which implies a regressive impact. Third, we estimate a tax gap of 44.7% of potential revenue, which corresponds to 2.1% of Brazil’s GDP.
"The effect of public funding on Swedish innovation, 1970-2001" (with Johanna Fink)
Abstract: Public innovation funding has been criticized for crowding out private R&D expenditure and thereby not generating additional innovation. This paper contributes to the debate by analyzing the case of Sweden between 1970 and 2001 and focusing on innovation output, in terms of new products developed and introduced in the market. We use a database of significant Swedish innovations (SWINNO), and link them to both approved and non-approved public funding projects.
Three questions are addressed in the paper. The first concerns the characteristics of firms applying for public funding, within those listed in the innovation database. The second looks at the effects of public funding on the generation of significant innovations, by comparing the outputs of approved projects to those of a comparable control group constructed by propensity score matching from the non-approved projects pool. The third zooms in within SWINNO to see if innovations receiving public funding are of higher quality, which we measure in terms of perceived novelty and the number of trade journal citations.
Our results show that firms with more radical projects and with frequent innovation activity are more likely to apply for public funding. Secondly, the estimated treatment effects show that approved public funding projects are twice as likely to translate into significant innovations. With respect to the third question, we do not find any significant impact. This applies to both public funding access in general and a higher share of the amount applied for.