Table of Contents Table of Contents
Previous Page  18 /304 Next Page
Information
Show Menu
Previous Page 18 /304 Next Page
Page Background

公司避稅與金字塔結構

18

To test H2, we employ Model (2):

AVOID

it

=

β

0

+

β

1

LAYER

it

+

β

2

TAXH

it

+

β

3

LAYER

it

×

TAXH

it

+

β

4

SIZE

it

+

β

5

MB

it

+

β

6

LEV

it

+

β

7

ROA

it

+

β

8

NOL

it

+

β

9

Δ

NOI

it

+

β

10

STD_ROA

it

+

β

11

Δ

SALES

it

(2)

+

β

12

CASH

it

+

β

13

INTAN

it

+

β

14

RD

it

+

β

15

PPE

it

+

β

16

ADV

it

+

β

17

SGA

it

+

β

18

ln

(

NUM_INVESTEE

)

it

+

YEARdummy

+

INDUSTRYdummy

+

ε

it

In Model (2), we add

TAXH

, which is defined as the number of investees in tax

havens divided by the total number of investees. Because there is no consensus on which

countries are considered tax havens, we employ two lists of tax havens to identify whether

a country is a tax haven.

7

We identify tax havens based on the list of tax havens as in

Durnev, Li, and Magnan (2017) to construct a variable, denoted as

TAXH1

. The list of

offshore financial centers (Tax Havens) comes from the International Monetary Fund

(IMF) and Financial Stability Forum (2000). We also employ the list of tax havens

following Dyreng and Lindsey (2009) to construct another variable,

TAXH2

.

8

The two lists

of tax havens are slightly different. We then interact L

AYER

with

TAXH1 (TAXH2)

as our

experimental variables.

9

Based on H2, we expect that the coefficient on the interaction

term,

LAYER×TAXH1 (TAXH2)

is significantly negative. We winsorize all used variables

at the 1 percent and 99 percent level.

7 While there is not an official definition of a tax haven, the Organization for Economic Cooperation and

Development (OECD) defined a tax haven in 1998 as a jurisdiction which has (1) no or only nominal

taxes, (2) a lack of transparency, (3) laws or administrative practices which prevent the effective

exchange of relevant information with other governments on taxpayers benefiting from the low or no

tax jurisdiction and (4) the absence of a requirement that the activity be substantial.

8 Following Dyreng and Lindsey (2009), tax haven countries are identified as countries that are on at

least three of the four commonly used tax haven lists. See

http://www.globalpolicy.org

on March 4,

2008. The four commonly used tax haven lists come from (1) the Organization for Economic

Co-Operation and Development (OECD), (2) the International Monetary Fund (IMF), (3) the Stop Tax

Haven Abuse Act adopted in the U.S. that targets a large number of tax haven countries, and (4) the

Tax Research Organizations.

9 We also use

NUM_TAX_HAVEN

, the number of investees in tax havens in year t to capture the extent

of the firm operating in tax haven countries, and the results are robust.