The Effect of Intellectual Property Boxes on Innovative Activity & Tax Avoidance University of Illinois Tax Doctoral Consortium III

Transcription

1 The Effect of Intellectual Property Boxes on Innovative Activity & Tax Avoidance University of Illinois Tax Doctoral Consortium III Tobias Bornemann, Stacie Laplante, Benjamin Osswald 1

2 Overview What? Do IP box regimes firms innovative activities? Which firms respond to IP box regimes? Why? : 18 countries introduce IP box regimes tax rate on specified IP income Several countries have input incentives, e.g. tax credits ØU.S. IP box: cost to taxpayer $9-11 bn. (Merrill et al. 2012) ØCurrent expenditures on R&D tax credits = $11.3 bn. How? Exploit introduction of IP box in Belgium

3 Overview Do patenting activities increase? Patent applications Patent grants Which firms decrease their ETRs? ETR especially pronounced: Domestic firms Subs of MNEs without income shifting opportunities

4 Literature Theoretically, IP boxes = high-powered tax incentives EATR on marginal R&D investments close to zero or even negative (Evers, Miller and Spengel 2015) Empirically mixed evidence: employment, no in fixed assets, limit income shifting (Chen, De Simone, Hanlon and Lester 2017) Attract high-value patents for R&D intensive firms (Alstadsaeter, et al. 2015) Effective for firms with immobile R&D activity (Merrill 2016)

5 Tax Incentives for Innovative Activities Input Incentives Innovative Activities Output Incentives Tax credit Super deduction IP Box = tax rate Patents Trademarks Know-How

6 Empirical Setting Prior literature: cross-country settings Empirically difficult to isolate effect tax rate on income from observable and unobservable intangibles Data not publicly available Our setting: IP box in Belgium in 2008 Clean(er) identification Patent income only Publicly available data Large tax rate : 33.99% to 6.8% All registered taxpayers in Belgium Strict reporting requirements

7 Patenting Activities: Hypothesis 1 Pat it =α+λ 1 Reform+λ 2 BE i + λ 3 Reform BE + θ Y it + ε it (1) IP Box Regime Patenting activities Reform = 1 post 2008 BE = 1 Patent applications, grants and stocks Control variables: firm size, leverage H1: Firms do not increase their patenting activities after the introduction of the IP box regime. 7

8 Patenting Activities: Results Table 5, Panel A: ln(patent Applications) ln(patent Grants) ln(patent Stock) (1) (2) (3) (4) (5) (6) Reform ** *** *** *** 0.067*** 0.051*** (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) BE *** *** *** *** *** *** (0.016) (0.013) (0.013) (0.010) (0.027) (0.023) Reform 0.012** *** 0.027*** *** *** x BE (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) Constant 0.241*** 0.051*** 0.185*** 0.053*** 0.571*** 0.153*** (0.019) (0.017) (0.015) (0.013) (0.032) (0.034) Observations 271, , , , , ,040 Adj. R Controls No Yes No Yes No Yes Industry FE Yes Yes Yes Yes Yes Yes 8

9 Effective Tax Rates: Hypothesis 2(a-c) ETR it = α + β 1 Reform it + β 2 Treat it + β 3 Reform it Treat it + δ X it + ε it (2) ETR it = α+γ 1 Reform it + γ 2 Treat it + γ 3 Shift it + + γ 6 Reform it Treat it γ 7 Reform it Treat it Shift it + δ X it + ε it (3) Income shifting opportunities IP Box Regime Tax payments Tax rate differential Reform = 1 for 2008 and later ETR Control variables: Size, Leverage, Intangibility, ROA, Inventory 9

10 Effective Tax Rates: Hypothesis 2(a-b) H2a H2b Reform x Treatment Reform x Treatment Table 6, Panel A: All Belgian Firms GAAP ETR Cash ETR (1) (2) (3) (4) ** ** (0.009) (0.009) (0.010) (0.010) Table 6, Panel B: BE - Domestic Firms ** * (0.010) (0.011) (0.010) (0.010) Table 6, Panel C: BE - MNEs Reform x Treatment (0.019) (0.019) (0.021) (0.022) Controls No Yes No Yes Ind. Fixed Effects Yes Yes Yes Yes 10

11 Effective Tax Rates: Hypothesis 2c Table 7, Panel C: Belgian MNEs All GAAP ETR Cash ETR (1) (2) (3) (4) Reform *** ** (0.013) (0.012) (0.013) (0.013) Treatment (0.025) (0.024) (0.027) (0.030) Shifting (0.007) (0.006) (0.008) (0.007) Reform x Treatment *** *** (0.027) (0.027) (0.044) (0.060) Reform x Treatment 0.096** 0.121*** x Shifting (0.040) (0.039) (0.053) (0.066) Observations 6,820 6,820 5,456 5,456 Adj. R-squared Controls No Yes No Yes Industry Fixed Effects Yes Yes Yes Yes 11

12 Conclusion We test the effectiveness of IP boxes in patenting activities decreasing firms effective tax rates We find: Patent applications Patent grants Patent stocks Significant in tax rates ETR especially pronounced: Domestic firms Subs of MNEs without income shifting opportunities

13 Next Steps Explore firms entering Belgium after 2007 Address limitations of balanced panel 13

14 Thank you for your attention & Looking forward to Your Questions! 14

15 Backup 15

16 Descriptives Table 3, Panel A: Balanced Panel Belgium - Full Sample N Mean Std. Dev. Q1 Median Q3 ETR 244, Cash ETR 29, Size 244, Leverage 244, Intangibility 244, ROA 244, Capital Intensity 244, Inventory 244, Patent Applications 244, Patent Grants 244, Patent Stock 244, MNE 244, SHIFTING 6,

17 Descriptives Table 3, Panel A - BE Treatment Group Control Group N Mean Std. Std. Q1 Med. Q3 N Mean Dev. Dev. Q1 Med. Q3 ETR 2, , Cash ETR 1, , Size 2, , Leverage 2, , Intangibility 2, , ROA 2, , Capital Int. 2, , Inventory 2, , Patent App. 2, , Patent Grants 2, , Patent Stock 2, , MNE 2, , SHIFTING ,

18 Descriptives Table 3, Panel B: Balanced Panel Germany N Mean Std. Dev. Q1 Median Q3 ETR 26, Cash ETR Size 26, Leverage 26, Intangibility 26, ROA 26, Capital Intensity 26, Inventory 26, Patent Applications 26, Patent Grants 26, Patent Stock 26, MNE 26, SHIFTING 2,

19 Descriptives Table 3, Panel B -DE Treatment Group Control Group N Mean Std. Std. Q1 Med. Q3 N Mean Dev. Dev. Q1 Med. Q3 ETR 6, , Cash ETR Size 6, , Leverage 6, , Intangibility 6, , ROA 6, , Capital Int. 6, , Inventory 6, , Patent App. 6, , , , Patent Grants Patent Stock 6, , MNE 6, , SHIFTING ,

20 Patenting Activities: Results Graphs Patent Applications Patent Grants 20

21 Effective Tax Rates: Hypotheses 2a+b Table 6, Panel A: All Belgian Firms GAAP ETR Cash ETR (1) (2) (3) (4) Reform *** *** *** *** (0.001) (0.001) (0.002) (0.002) Treatment *** (0.008) (0.008) (0.011) (0.010) Reform x ** ** Treatment (0.009) (0.009) (0.010) (0.010) Constant 0.271*** 0.283*** 0.422*** 0.577*** (0.006) (0.007) (0.025) (0.030) Observations 244, ,450 29,779 29,779 Adj. R-squared Controls No Yes No Yes Ind. Fixed Effects Yes Yes Yes Yes 21

22 Effective Tax Rates: Hypotheses 2a+b Table 6 Panel B: BE - Domestic Firms GAAP ETR Cash ETR (5) (6) (7) (8) Reform *** *** *** *** (0.001) (0.001) (0.002) (0.002) Treat *** (0.009) (0.009) (0.014) (0.012) Reform x ** * Treat (0.010) (0.011) (0.010) (0.010) Constant 0.270*** 0.287*** 0.432*** 0.603*** (0.006) (0.007) (0.027) (0.032) Obs. 237, ,630 24,323 24,323 Adj. R Controls No Yes No Yes Ind. FE Yes Yes Yes Yes 22

23 Effective Tax Rates: Hypotheses 2a+b Table 6 Panel C: BE - MNEs GAAP ETR Cash ETR (9) (10) (11) (12) Reform *** *** *** *** (0.004) (0.004) (0.005) (0.004) Treat ** *** (0.017) (0.018) (0.017) (0.019) Reform x Treat (0.019) (0.019) (0.021) (0.022) Constant 0.329*** 0.410*** 0.359*** 0.519*** (0.002) (0.022) (0.002) (0.026) Obs. 6,820 6,820 5,456 5,456 Adj. R Controls No Yes No Yes Ind. FE Yes Yes Yes Yes 23

24 Effective Tax Rates: Hypothesis 2c Table 7, Panel C: Belgian MNEs All GAAP ETR Cash ETR (1) (2) (3) (4) Reform *** ** (0.013) (0.012) (0.013) (0.013) Treatment (0.025) (0.024) (0.027) (0.030) Shifting (0.007) (0.006) (0.008) (0.007) Reform x Treatment *** *** (0.027) (0.027) (0.044) (0.060) Reform x Treatment 0.096** 0.121*** x Shifting (0.040) (0.039) (0.053) (0.066) Constant 0.322*** 0.412*** 0.351*** 0.517*** (0.007) (0.022) (0.008) (0.027) Observations 6,820 6,820 5,456 5,456 Adj. R-squared Controls No Yes No Yes Industry Fixed Effects Yes Yes Yes Yes 24

25 Next Steps ETRs unbalanced sample ETRs firms that occur in sample after 2007 ETR ETR Year Year POSTONLY Firms Other firms POSTONLYpatents Firms POSTONLYpatents Other firms 25