Are Corporate Tax Reductions Real Benefits under Imputation Systems?

Abstract

Imputation systems integrate corporate and shareholder personal income taxes to alleviate double taxation of dividend income. In this study, we empirically examine whether a corporate tax rate reduction under an imputation tax system benefits shareholders. Using Taiwan as a setting, our analyses indicate that decreasing the corporate tax rate is associated with an increase in dividend payout ratio and foreign investment. Moreover, the increase in dividend payout ratio is even greater for firms that have a higher increase in foreign ownership. Additionally, the market reacts positively to an announcement of a tax rate reduction; specifically, positive stock price reactions are stronger for firms that experienced a greater increase in foreign ownership in response to the tax rate reduction, for firms with greater liquidity constraints and more growth opportunities before the tax rate reduction, and for firms with a bigger decrease in effective tax rates after the tax rate reduction. Overall, we provide evidence that a tax rate reduction is associated with economic impacts and that foreign shareholders appear to be the main beneficiaries of a tax rate reduction under an imputation tax system.

Acknowledgements

We are grateful to Laurence van Lent (editor) and an anonymous reviewer for valuable comments and suggestions.

ORCID

Vincent Y.S. Chen http://orcid.org/0000-0001-8211-2521

Notes

¹An example of a classical system is the US income tax system, which taxes corporations and shareholders separately.

²According to the Organisation for Economic Co-operation and Development tax databases, countries that currently use full or partial imputation tax systems for dividends include Australia, Canada, Chile, Korea, Mexico, New Zealand, and the UK (http://www.oecd.org/tax/tax-policy/tax-database.htm#C_CorporateCaptial).

³Taiwan implemented an imputation system in 1998. From 1998 to 2010, Taiwan's corporate income tax rate was 25% for firms with income equal to or above NT $100,000. To improve the economy and compete with countries such as Hong Kong, Ireland, and Singapore, Taiwan revised its tax policy and lowered the corporate income tax rate from 25% to 17% in 2010. A 2008 press release from Taiwan's Council for Economic Planning and Development noted that

tax reduction is a global trend and that Taiwan's tax rate of 25% was obviously uncompetitive with Hong Kong's 17.5%, Ireland's 12.5%, and Singapore's 18%. To rectify this problem the Executive Yuan will propose a third tax reform to drastically cut business income tax rate … by the end of 2009.

⁴The Australian imputation system requires corporations to keep two-tiered retained earnings accounts: one is a franked account that accumulates earnings taxed at the corporate level that are eligible for imputation credits to shareholders when dividends are distributed from it, and the other is an unfranked account that accumulates earnings that have not been taxed at the corporate level and does not give imputation credits to shareholders if dividends are distributed from it (Chen & Gupta, 2011).

⁵Taiwan-sourced income earned by nonresident taxpayers is subject to a flat withholding tax rate when the income is paid by the company. The withholding tax rate is 20% and has not changed since its inception.

⁶Many of Taiwan's competitor countries lowered their corporate income tax rates during 2008–2011. For example, China, Singapore, Korea, Hong Kong, and Ireland reduced their corporate tax rates to 25%, 17%, 22%, 16.5%, and 12.5%, respectively.

⁷The Statute for Upgrading Industries was intended to promote and upgrade Taiwanese companies in high-tech industries by providing extensive tax incentives, including an investment tax credit, an R&D tax credit, and a five-year tax holiday, to allow Taiwanese companies to invest in designated industries. The abundant tax incentives of the statute caused enormous tax revenue losses. Executive Yuan, Cabinet of the Taiwanese government, proposed to replace the Statute for Upgrading Industries with the Industrial Innovation Act, which essentially eliminates all tax incentives except an R&D tax credit.

⁸It is possible that domestic shareholders do not support an increase in dividend payout ratio because of limited tax benefits. An increase in dividend income may also increase their marginal income tax rates, and the increased tax expenses from non-dividend income can offset a large portion of after-tax dividend income.

⁹Prior to 2010, there were two different rates for foreign withholding tax. For foreign investment approved (not approved) by the Taiwanese investment authority, foreign withholding tax rate was 20% (30%). Since 2010, the Taiwanese government has set foreign withholding tax at a standard rate of 20% regardless of whether the foreign investment receives regulatory approval. We use 20% as a foreign withholding tax rate before 2010 for two reasons. First, in practice, the majority of foreign investment received approval from the authority prior to 2010. Second, this implicit assumption is necessary to ensure that the economic consequences and market pricing discussed in this study result from a corporate tax rate reduction rather than a foreign withholding tax rate reduction.

¹⁰We obtain abnormal stock returns on event dates by performing the SUR model as follows: R_it = α_it + β_i Rm_t + Σ_Tγ _it D_t + ϵ_it where R_it is firm i’s daily stock return at day t. The trading days used in the SUR model include the estimation period of 150 trading days and the event periods. Each event period _T spans the day before through the day after the event. R_mt is the market return at day t. D is the dummy variable, which equals 1 if the day occurs in event period _T and 0 otherwise. The coefficient γ is firm i’s abnormal stock return on event date _T when D equals 1. ϵ is the residual term.

¹¹The Hausman χ² statistic for testing the consistency of random-effects estimation is 383 (p-value < .01), which suggests that the random-effects model may be inconsistent. Accordingly, we report only the fixed-effects estimation results.

¹²The Hausman χ² statistic for testing the consistency of random-effects estimation is 45.77 (p-value < .01), which suggests that the random-effects model may be inconsistent. Accordingly, we report only the fixed-effects estimation results.

¹³We conduct three sensitivity tests to show that a 7% cap on stock price change does not affect our results. First, we find that the percentage of firms that reach the 7% cap is relatively small (less than 1% of the total sample). Second, after removing firms that reach the 7% cap from our sample and rerunning our tests, the results remain quantitatively similar. Lastly, we extend our return period to t + 4, and our results do not change.

Additional information

Funding

We acknowledge financial support from Ministry of Science and Technology, Taiwan (MOST 1012410H004070).