Regional payroll tax cuts and individual wages: heterogeneous effects of worker ability and firm productivity

Abstract

This paper exploits a payroll tax reform in Norway and applies matched employer–employee register data on individual wages to study the incidence of payroll taxation. The contribution is to allow for heterogeneous wage effects of payroll tax cuts based on unobserved worker ability and firm productivity (measured by estimated worker and firm fixed effects, respectively). Using the difference-in-difference approach, the estimates show that on average, about 30% of the labor cost reduction is shifted to employees through higher wages and the degree of tax shifting increases gradually during the first 3 years after the reform. Among high-productivity firms, the degree of tax shifting is stable at 40–50% throughout the post-reform period. In low-productivity firms, none of the labor cost reduction is shifted to employees in the short term, but there is a delayed wage response in the medium term. On average, the wage effect of reduced payroll taxes is twice as large in high-productivity compared to low-productivity firms. The difference in wage response is mainly between firms, rather than between workers within firms, although low-ability workers in low-productivity firms miss out on the wage gain from the reduced labor costs. The analysis does not find any robust employment effect of the payroll tax cut, neither in low- nor in high-productivity firms.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2

Notes

  1. 1.

    Other studies using tax variation across firms and states in the USA to estimate tax incidence include Gruber (1994) and Anderson and Meyer (1997, 2000).

  2. 2.

    It should be noted that the analysis estimates short- to medium-term impacts of the reform as it is based on only 3 years of post-reform data.

  3. 3.

    During 2004–2006 tax rates increased gradually in zones 2–4 due to European Economic Area (EEA) regulations. In 2007, payroll taxes were again allowed to differ across municipalities. At the same time, the differentiation was extended to seven regional zones and the determination of tax zone changed from employees’ resident location to the firm’s location.

  4. 4.

    In 2000, Norway consisted of 435 municipalities, and in total, 53 municipalities were affected by the reform; 14 municipalities faced an increase in the payroll tax rate (moving from zone 2 to zone 1), while the remaining 39 municipalities moved to a zone with lower tax rate.

  5. 5.

    As a check of robustness, the analysis is also performed based on all 32 municipalities, and the findings are consistent (documented in Table 8 in Appendix and discussed in Sect. 3).

  6. 6.

    Consistent with the choice of treatment municipalities, the control group is restricted to municipalities that are part of an economic region where all municipalities in the region remain in tax zone 2 during the entire period of study.

  7. 7.

    The employment register separates between three contract types: full-time contracts with at least 30-h work per week, long part-time contracts with 20–29-h work per week and short part-time contracts with less than 20-h work per week.

  8. 8.

    I exclude workers in the primary industries (agriculture, fishing and forestry), as well as public sector workers. Workers above 61 years of age face lower payroll tax rates than other workers and are therefore dropped from the analysis. I also exclude workers below 25 years of age. This gives a dataset of about 270,000 worker-year observations. The tax register gives information on total annual earnings, rather than separate earnings for each work contract. Workers with more than two contracts during a year, as well as workers with one full-time and one part-time contract, are excluded. For workers with two full-time contracts, a maximum of three months of overlap between the contracts is allowed. I also exclude workers whose contract duration is less than 300 days during a year. These restrictions reduce the dataset by about 51,000 observations. Missing data on annual earnings, level of education or industry affiliation, together with exclusion of workers that move between treatment and control municipalities, further excludes approximately 3000 observations. Individuals residing in treatment or control municipalities but working in firms located outside these municipalities are excluded since the estimation of firm fixed effects will be misleading (the majority of workers in these firms are not part of the dataset). This applies to offshore workers as well as workers with long-distance commute to Oslo and other large cities and amounts to about 36,000 observations. Workers that are observed in a single year are excluded, since worker fixed effects cannot be estimated (about 6000 observations). To avoid extreme observations, I exclude the bottom 1% of the wage distribution, which gives the final dataset of 171,991 observations.

  9. 9.

    As a check of robustness, the analysis is also performed based on an alternative control group with sector composition more comparable to the treatment group (documented in Table 8 in Appendix and discussed in Sect. 3).

  10. 10.

    Due to the inclusion of worker fixed effects, the non-interacted treatment dummy is dropped from the regression.

  11. 11.

    Lower payroll taxes can also generate an employment expansion at the municipal level through entry of new firms (rather than increased firm size). Such effects are ignored in the present analysis.

  12. 12.

    Solon et al. (2015) offer a discussion of weighting in regression analysis.

  13. 13.

    The 4.2 percentage point reduction in the payroll tax rate from an initial level of 10.6% corresponds to 3.8% reduction in labor costs: (1.106w − 1.064w)/1.106w = 0.038.

  14. 14.

    In the full dataset, the observations are allocated equally across the 6-year period with 16–17% of the observations in each year. In the subsample with data on hourly wages, the share of observations is down at 12% in 1998, while each of the other years accounts for 15–19% of the observations.

  15. 15.

    A set of tables describing alternative model specifications is available as an external online appendix: https://sites.google.com/site/hildegunnestokke/.

  16. 16.

    The alternative split, where workers in firms with zero fixed-effect coefficient are included in the upper part of the distribution, is documented in external online appendix. The main findings remain.

  17. 17.

    The worker fixed-effect distribution ranges from 4.68 to 8.53 with mean and standard deviation equal to 6.47 and 0.29, respectively.

  18. 18.

    To further check for potential differences across workers, the external online appendix offers comparisons of wage effects of reduced payroll taxes across groups of workers based on observable worker characteristics, notably gender, age and level of education. The broad picture is no significant difference in the estimated wage effects across these worker groups.

  19. 19.

    The estimation results are available from the author upon request.

  20. 20.

    The estimation applies the firm-level data used in Sect. 4, and the average firm wage is calculated from the individual-level data used in Sect. 3. Due to stricter sample restrictions on the individual-level data (to have precise wage measures) compared to the firm-level data, some firm-year observations are dropped due to missing wage data. The estimation is based on a total of 32,408 observations.

References

  1. Abowd, J., Kramarz, F., & Margolis, D. (1999). High-wage workers and high-wage firms. Econometrica, 67(2), 251–333.

    Article  Google Scholar 

  2. Anderson, P., & Meyer, B. (1997). The effects of firm specific taxes and government mandates with an application to the U.S. unemployment insurance program. Journal of Public Economics, 65, 119–145.

    Article  Google Scholar 

  3. Anderson, P., & Meyer, B. (2000). The effects of the unemployment insurance payroll tax on wages, employment, claims and denials. Journal of Public Economics, 78, 81–106.

    Article  Google Scholar 

  4. Bargain, O., & Peichl, A. (2016). Own-wage labor supply elasticities: Variation across time and estimation methods. IZA Journal of Labor Economics, 5, 10. https://doi.org/10.1186/s40172-016-0050-z.

    Article  Google Scholar 

  5. Bennmarker, H., Mellander, E., & Öckert, B. (2009). Do regional payroll tax reductions boost employment? Labour Economics, 16, 480–489.

    Article  Google Scholar 

  6. Borusyak, K. & Jaravel, X. (2017). Revisiting event study designs. Available at SSRN: https://ssrn.com/abstract=2826228

  7. Dale-Olsen, H. (2018). Labour demand and supply changes in Norway following an imposed harmonization of geographically differentiated payroll-tax rates. LABOUR, 32, 261–291.

    Article  Google Scholar 

  8. Gavrilova, E., Zoutman, F., & Hopland, A. O. (2017). How to use one instrument to identify two elasticities. CESifo Working Paper No. 6379.

  9. Gibbons, C., Suárez Serrato, J., & Urbancic, M. (2019). Broken or fixed effects? Journal of Econometric Methods, 8(1), 20170002. https://doi.org/10.1515/jem-2017-0002.

    Article  Google Scholar 

  10. Gruber, J. (1994). The incidence of mandated maternity benefits. American Economic Review, 84, 622–641.

    Google Scholar 

  11. Gruber, J. (1997). The incidence of payroll taxation: Evidence from Chile. Journal of Labor Economics, 15, S72–S101.

    Article  Google Scholar 

  12. Korkeamäki, O., & Uusitalo, R. (2009). Employment and wage effects of a payroll-tax cut: Evidence from a regional experiment. International Tax and Public Finance, 16, 753–772.

    Article  Google Scholar 

  13. Ku, H., Schönberg, U., & Schreiner, R. (2020). Do place-based tax incentives create jobs? Journal of Public Economics. https://doi.org/10.1016/j.jpubeco.2019.104105.

    Article  Google Scholar 

  14. Lichter, A., Peichl, A., & Siegloch, S. (2015). The own-wage elasticity of labor demand: A meta-regression analysis. European Economic Review, 80, 94–119.

    Article  Google Scholar 

  15. Murphy, K. J. (2007). The impact of unemployment insurance taxes on wages. Labour Economics, 14, 457–484.

    Article  Google Scholar 

  16. Saez, E., Matsaganis, M., & Tsakloglou, P. (2012). Earnings determination and taxes: Evidence from a cohort-based payroll tax reform in Greece. Quarterly Journal of Economics, 127, 493–533.

    Article  Google Scholar 

  17. Saez, E., Schoefer, B., & Seim, D. (2019). Payroll taxes, firm behavior, and rent sharing: Evidence from a young workers’ tax cut in Sweden. American Economic Review, 109(5), 1717–1763.

    Article  Google Scholar 

  18. Schmidheiny, K., & Siegloch, S. (2019). On event study designs and distributed-lag models: Equivalence, generalization and practical implications. IZA Institute of Labor Economics DP No. 12079.

  19. Solon, G., Haider, S., & Wooldridge, J. (2015). What are we weighting for? Journal of Human Resources, 50(2), 301–316.

    Article  Google Scholar 

  20. Stock, J., & Yogo, M. (2005). Testing for weak instruments in linear IV regression. In J. Stock & D. Andrews (Eds.), Identification and Inference for Econometric Models: Essays in Honor of Thomas Rothenberg (pp. 109–120). Cambridge: Cambridge University Press.

    Google Scholar 

  21. Zoutman, F., Gavrilova, E., & Hopland, A. O. (2018). Estimating both supply and demand elasticities using variation in a single tax rate. Econometrica, 86(2), 763–771.

    Article  Google Scholar 

Download references

Acknowledgements

I appreciate discussions at the 2015 Norwegian Research Forum on Taxation, the 2015 Meeting of the Urban Economics Association (UEA), the 2016 Meeting of the Society of Labor Economists (SOLE), the 2016 Louis-André Gérard-Varet (LAGV) Conference in Public Economics, the 2016 European Meeting of the Urban Economics Association (UEA), the 2017 Conference of the European Association of Labour Economists (EALE), and the research seminar at Statistics Norway, and in particular comments from Rolf Aaberge, Ådne Cappelen, Tarjei Havnes, Lars Kirkebøen, Andreas Kostøl, Stefan Leknes, Jarle Møen, Jørn Rattsø, Bjarne Strøm, Gaute Torsvik, Mark van Duijn, Russell Weinstein, Jeffrey Zax, the editor, and two anonymous referees. I am grateful for the cooperation of Statistics Norway and funding from the Research Council of Norway (grant number 255509).

Author information

Affiliations

Authors

Corresponding author

Correspondence to Hildegunn E. Stokke.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 710 kb)

Appendix: Additional tables

Appendix: Additional tables

See Tables 8 , 9 , 10 and 11.

Table 8 Robustness of the wage effect of reduced payroll taxes
Table 9 Descriptive statistics for subsample of workers with data on hourly wages
Table 10 Alternative model formulations to check for significant difference in wage effect between groups of workers defined based on firm productivity or worker ability
Table 11 Robustness of the employment effect of reduced payroll taxes

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Stokke, H.E. Regional payroll tax cuts and individual wages: heterogeneous effects of worker ability and firm productivity. Int Tax Public Finance (2021). https://doi.org/10.1007/s10797-020-09647-w

Download citation

Keywords

  • Payroll tax cut
  • Individual wages
  • Matched employer–employee data
  • Heterogeneous effects

JEL codes

  • H22
  • J23
  • J31
  • J38
  • R58