The article by Jenni Kellokumpu, Leena Savolainen and Simo Pesola (KSP) estimates the size of automatic fiscal stabilisers in Finland from 1993 until 2021. It uses the OECD approach and calculates a budgetary semi-elasticity. The budgetary semi-elasticity measures the sensitivity of the budget balance to cyclical fluctuation, which is measured by the output gap.
KSP provide a careful analysis using standard variable definitions, and the disaggregation of the tax elasticities is in line with previous studies. This makes the article’s estimate of the size of automatic stabilisers comparable to previous estimates based on Finnish data. Hence, it provides interesting and important information about the level of and changes to the size of automatic stabilisers in Finland.
My main comment on the paper is that the authors have produced a solid and important contribution. The results will be very useful in evaluating Finnish fiscal policy.
I have two other comments: a technical one that should be kept in mind when using the results to calculate the structural fiscal balance and another one about the side effects of increasing income tax progressivity to maintain the size of automatic stabilisers.
The technical comment concerns uncertainty regarding the unobservable potential output variable, which is used to measure the output gap variable. The output gap is the difference between the actual and potential output divided by potential output. My comment is a general one, and it concerns all papers that calculate the output gap variable using an approach that leads to an estimated potential output close to actual output. It is an approach that makes the potential output pro-cyclical and, in turn, this results in a structural fiscal balance close to the actual balance. For example, all large observed fiscal surpluses tend to be large structural fiscal surpluses, and the same is the case for deficits.
The pro-cyclicality of the potential output variable meant that the estimated output gap in Finland in 2007 (for example) was small, and the structural fiscal surplus was estimated to be very large, given that the observed fiscal surplus was very large. The implication of all of this for the KSP article is that if the estimated potential output variable has a pro-cyclical bias, then that bias also affects the estimates of the size of the automatic fiscal stabilisers. I want to stress that this comment is not specific to KSP. Moreover, the authors draw attention to this problem. They note that the estimated elasticities are sensitive to how we measure variables that are not directly observable, such as the output gap and equilibrium unemployment.
My final comment concerns the effects of increased income tax progressivity on low incomes in Finland. The paper notes in the concluding remarks that lower income tax rates for low-income individuals have been financed by higher rates at higher income levels. This results in a more progressive taxation that makes a positive contribution to the automatic stabilisers. While tax progressivity helps to maintain the size of automatic fiscal stabilisers, it is important to be aware of unintended consequences. In the last 20 years or so, the earned income tax credit, which is phased out steeply as income rises, has increased significantly, lowering average tax rates but increasing marginal ones. This has markedly increased the tax progressivity for people on low incomes in Finland relative to the other Nordic countries (Figure 1). The marginal tax rates on low earnings in Finland are currently clearly higher than in the other Nordic countries (Figure 2). The high marginal income rates may have contributed to the large rise in part-time employment in Finland in the period 2000–2022 (Figure 3). Moreover, the high marginal tax rates may also have slowed down labour productivity growth.