8 /304
公司避稅與金字塔結構
8
Lisowsky (2010) uses only 267 tax shelter-year observations between year 2000-2004 in
the U.S. The small sample size may limit its generalizability to other firms, or other
countries. Third, Lisowsky (2010) cross-reference Exhibit 21 of 10K to identify the
subsidiaries in tax havens. In the U.S., the Securities and Exchange Commission rules
demand subsidiary disclosure in Exhibit 21 of 10K only when subsidiary operations are
"significant", using the 10% threshold based on assets, investment, or income. Lisowsky
(2010) suffers from selection bias by ignoring subsidiaries who do not meet “significant”
criteria. Thus, our study will fill the gap by taking advantage of a comprehensive
subsidiary data in Taiwan to explore the relationship between tax avoidance, investment
layers, and tax havens.
2.3 Literature on Tax Avoidance and Agency Costs
Traditional view is held that tax avoidance is value enhancing to shareholders
because tax avoidance can reduce the wealth transfers from shareholders to the
government. This view is associated with the underlying premise that the interests of
shareholders and managers are aligned and a firm makes the tax reporting decisions
without agency considerations but consider only tax rates, the probability of tax avoidance
being detected, and penalties (Allingham and Sandmo, 1972).
Recently, there is another stream of research exploring the association between tax
avoidance activities and the agency problems between shareholders and managers (e.g.,
Desai, Dyck, and Zingales, 2007; Desai and Dharmapala, 2008; Chen and Chu, 2005;
Crocker and Slemrod, 2005; Slemrod, 2004). Using the principal-agency model, Slemrod
(2004) and Chen and Chu (2005) incorporate the non-tax cost considerations and find that
separation of ownership and management can lead to corporate avoidance behavior that
does not maximize shareholders’ benefits, but reflect the private interest of the manager or
the controlling shareholders. In particular, Desai and Dharmapala (2006) argue that the
crucial characteristics of tax avoidance activities are complexity and obfuscation.
Controlling shareholders can exploit the proprietary and obfuscation nature of tax
avoidance to mask rent extraction, such as earning management, related-party
transactions, selective information disclosure, and etc., which reduces firm value (Desai
and Dharmapala, 2009). Consistent with this perspective, Desai and Dharmapala (2009)
and Wilson (2009) document that tax avoidance (Tax Sheltering) increases firm value only
for well-governed firms. Likewise, Desai and Dharmapala (2009) find that the effect of
tax avoidance on firm valuation is positive only for firms with lower level of anti-takeover