Baseline Estimations - Estimation Results

3.6 Estimation Results

3.6.1 Baseline Estimations

Table 3.3 presents our baseline estimation. Following the methodology described in Section 3.5.1, Specification (1) regresses the logarithm of subsidiary intangible asset investment on the firm’s statutory corporate tax rate, while controlling for fixed firm and year effects. In line with our theoretical considerations, we find a statistically significant negative influence that suggests high corporate tax rates at an affiliate to be associated with low intangible asset investment and vice versa. The effect is robust against the inclusion of time-varying country control characteristics in Specification (3) and sales as a firm size control in Specification (5).

However, the subsidiaries’ statutory tax rate may be an imprecise measure for tax incentives on intangible asset location since our hypothesis predicts intangibles to be located in countries with a low tax rate relative to all other affiliates of the corporate group. This is accounted for in Specifications (2), (4), and (6) which regress the level of intangible assets on theaverage tax difference to all other affiliates. Confirming our theoretical expectations of Section 3.2, the results indicate that the average statutory corporate tax rate difference between a subsidiary and other group members exerts a highly significant negative impact on the subsidiary’s intangibles holdings. Quantita- tively, the estimations suggest that a decrease in theaverage tax difference to all other affiliates by 1 percentage point raises the subsidiary’s level of intangible assets by 2.1% (cf. Column (6) of Table 3.3).

Chapter 3 – Corporate Taxes & Location of Intangibles 58

Table 3.3: Baseline Estimations

OLS Firm–Fixed–Effects, Panel 1995–2005

Depend. Variable Log Intangible Assets Log Int./Sales

Explanat. Variables: (1) (2) (3) (4) (5) (6) (7) (8)

Statutory Tax Rate -2.44∗∗∗ -2.38∗∗∗ -2.50∗∗∗ -2.54∗∗∗ (.631) (.663) (.690) (.717)

Av.TaxDiff.toOthers -2.08∗∗∗ -2.01∗∗∗ -2.08∗∗∗ -2.27∗∗∗ (.559) (.584) (.606) (.627) Log Sales .447∗∗∗ .453∗∗∗

(.026) (.026)

Country R&D Exp. .385∗∗∗ .325∗∗∗ .247∗ .178 .094 .041 (.132) (.132) (.140) (.139) (.144) (.143) Corruption Index .007 -.001 .019 .008 .030 .019 (.029) (.029) (.029) (.029) (.030) (.030) Population .013 .010 -.013 -.020 -.025 -.029 (.038) (.039) (.038) (.038) (.038) (.039) GDP per Capita -.028∗∗ _-.028∗∗ _-.025∗ _-.023 _-.021 _-.017 (.013) (.014) (.015) (.016) (.016) (.016) Growth GDPp.Cap. .018 .017 -.000 -.003 -.020∗ -.021∗ (.011) (.012) (.011) (.011) (.012) (.012) Year Dummies √ √ √ √ √ √ √ √ # Observations 44,739 42,994 42,215 40,574 37,112 35,666 37,112 35,666 # Firms 6,619 6,363 6,617 6,361 6,017 5,788 6,017 5,788 AdjustedR2 _.7218 _.7207 _.7274 _.7262 _.7538 _.7531 _.7195 _.7194

Notes: Heteroscedasticity robust standard errors adjusted for firm clusters in parentheses.

∗_, ∗∗_, ∗∗∗ _{indicates significance at the 10%, 5%, 1% level. Observational units are multinational}

subsidiaries, i.e. firms that exhibit a foreign parent which owns at least 90% of the ownership shares. Additionally, to be included in the sample, at least one affiliate of the corporate group has to own intangible assets and at least one has to make positive profits. In (1)–(6), the dependent variable is the natural logarithm (Log) of the level of intangible assets. In (7)–(8), the dependent variable is the log of the ratio intangible assets to sales (Log (Int./Sales)). In both cases, we add a small constant to the initial level of intangible assets to avoid losing observations with zero intangibles by taking the log. An OLS model with fixed firm effects is estimated. Av.TaxDiff.toOthers is defined as the unweighted average statutory tax rate difference between the considered subsidiary and all other affiliates of the corporate group including the parent. AdjustedR2 _{values are calculated from a}

dummy variables regression equivalent to the fixed-effects model.

the affiliate’s sales as a regressor. The coefficient estimate suggests that size positively affects the reported intangible assets where the tax effect on the intangibles variable remains largely unchanged. In Columns (7) and (8), we moreover rerun our baseline model using the ratio of intangible assets to sales as regressand and find comparable results.

Chapter 3 – Corporate Taxes & Location of Intangibles 59

3.6.2 Binary Dependent Variable

In this section, we estimate Equation (3.9) and thus focus on the binary multinational choice whether to locate intangible property at a certain affiliate or not. The results are displayed in Table 3.4. Specifications (1) to (4) thereby present maximum-likelihood estimations of a fixed-effect logit model. The dependent variable is theDummy Intan- gible Assets which is set to 1 if a subsidiary owns intangible assets and 0 otherwise. Since the logit estimation controls for subsidiary fixed effects, many subsidiaries drop out of the estimation since they observe no variation in the status of intangibles-holding vs. non-holding during the observation period. Nevertheless, the estimations still com- prises an adequate number of about 2,000 firms for which information is available for 7.3 years on average.

In Specifications (1) and (3), we regress the binary dependent variable on the subsidiary’s statutory tax rate. The coefficient estimate is negative and highly significant and thus confirms the presumption that a subsidiary’s probability of holding intangible property decreases in the location’s statutory tax rate. Moreover, Specifications (2) and (4) estimate the relation using the average tax difference to all other affiliates as explanatory tax variable. Again, we find a negative effect on intangibles holdings which is statistically significant at the 1% level. Thus, conditioning on country characteristics and firm size, the lower a subsidiary’s statutory corporate tax rate compared to all other affiliates of the same multinational group (including the parent), the higher is its probability of holding intangible assets.30

Nevertheless, the estimation of the fixed-effect logit model critically depends on the assumption of a logistic distribution of the error term. Thus, as a sensitivity check, we moreover estimate a linear probability model with subsidiary fixed effects. The application of an OLS framework thereby has the additional advantage that we make use of all information in our dataset and do not confine the sample to subsidiaries which observe a change over the sample period in the status of intangibles-holding vs. non-holding. The results are displayed in Specifications (5) to (8) of Table 3.4 and are qualitatively equal to the results of the logit model. Ceteris paribus, a reduction of the average tax difference to all other affiliates by 10 percentage points is suggested to raise the subsidiary’s probability of holding intangible assets by 2.1 percentage points on average (cf. Column (8)). As the mean probability of holding intangibles is 55.0%,

30_{The coefficient estimates of a logit estimation cannot be interpreted quantitatively. Moreover, ap-}

plying a logit model with fixed effects makes the calculation of marginal effects impracticable as it requires specifying a distribution for the fixed effects.

Chapter 3 – Corporate Taxes & Location of Intangibles 60

Table 3.4: Binary Dependent Variable

Logit & OLS Firm–Fixed–Effects, Panel 1995–2005 Dependent Variable: Dummy Intangible Assets

Model Logit Fixed-Effects OLS Fixed-Effects

Explanat. Variables: (1) (2) (3) (4) (5) (6) (7) (8)

Statutory Tax Rate -3.36∗∗∗ -4.31∗∗∗ -.271∗∗∗ -.309∗∗∗ (1.14) (1.26) (.102) (.117)

Av.TaxDiff.toOthers -2.71∗∗∗ -2.82∗∗∗ -.218∗∗ -.208∗∗ (1.04) (1.14) (.104) Log Sales .583∗∗∗ .575∗∗∗ .053∗∗∗ .054∗∗∗ (.048) (.048) (.004) (.004) Country R&D Exp. .323 .122 .051∗∗ .039

(.298) (.298) (.025) (.024) Corruption Index .085 .047 .006 .004 (.067) (.068) (.005) (.005) Population -.259∗∗∗ -.287∗∗∗ -.016∗∗∗ -.017∗∗∗ (.082) (.083) (.007) (.007) GDP per Capita -.008 -.015 -.001 -.001 (.039) (.038) (.002) (.003) Growth GDPp.Cap. -.001 -.009 .002 .001 (.029) (.029) (.002) (.002) Year Dummies √ √ √ √ √ √ √ √ # Observations 16,734 16,151 13,164 12,726 44,719 42,974 37,102 35,656 # Firms 2,227 2,150 1,884 1,822 6,619 6,363 6,017 5,788 Pseudo or Adj.R2 .0198 .0197 .0596 .0584 .6724 .6709 .6786 .6763 Notes: Heteroscedasticity robust standard errors adjusted for firm clusters in parentheses.

∗_, ∗∗_, ∗∗∗ _{indicates significance at the 10%, 5%, 1% level. Observational units are multinational}

Av.TaxDiff.toOthers is defined as the unweighted average statutory tax rate difference between the considered subsidiary and all other affiliates of the corporate group including the parent. Adjusted

R2 _{values are calculated from a dummy variables regression equivalent to the fixed-effects model.}

this corresponds to an average increase of 3.8%.

In document Dischinger, Matthias (2010): Multinational Enterprises and Corporate Taxation: An Empirical Assessment of the Location of Assets, Profits and Debt. Dissertation, LMU München: Volkswirtschaftliche Fakultät (Page 63-66)