Do auditors constrain intertemporal income shifting in private companies?^*

* Paper accepted by Martin Jacob

Abstract

This study investigates the association between private company auditing and intertemporal income shifting. Using a large reduction in the Finnish corporate tax rate as a strong incentive for income shifting and financial statement data coupled with proprietary information from the tax authorities, we analyse accruals and cost stickiness of small private companies. Our results reveal significant differences in accrual income shifting between audited and unaudited companies, but only among companies that on average could anticipate the tax reduction the most. Further, we find auditors to restrict sticky selling, general, and administrative cost behaviour that we hypothesise is associated with illegal actions. Additional tests expose a nontrivial number of incorrectly unaudited companies which are the ones mostly associated with income shifting. Taken together, our study highlights the effects of audit exemption and the importance of enforcement while also suggesting that the audit process is value adding for the tax authorities.

Keywords:

• audit exemption
• cost stickiness
• earnings management
• tax incentive

Acknowledgement

We thank Martin Jacob and two anonymous journal reviewers, Antti Miihkinen, Ann Gaeremynck, Jesper Haga, and Johan-Erik Fant for their helpful comments and suggestions. We are also appreciative for the comments and feedback from workshop participants and reviewers at the 2016 Workshop on European Financial Reporting, 2016 Nordic Accounting Conference, 2017 European Accounting Association Annual Congress, 2017 American Accounting Association Annual Meeting, 2018 International Accounting Section Midyear Meeting, 2018 Financial Accounting and Reporting Section Midyear Meeting, and 2018 British Accounting and Finance Association Annual Conference. All errors remain our own. This paper has previously been circulated as ‘Earnings management in response to corporate tax reductions and the value of audits’. We thank the Finnish Tax Administration for data access.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

* Paper accepted by Martin Jacob

1 Consistent with existing tax research (e.g. Shackelford and Shevlin 2001), we define intertemporal corporate income shifting as a form of tax planning where companies defer taxable and book income between periods with an underlying incentive to obtain tax savings. We use the terms earnings management and intertemporal income shifting interchangeably throughout the paper.

2 The examination of the administration covers significant decisions, actions taken and circumstances of the company in order to determine whether any member of the board of directors or the managing director is liable to the company or has acted in contravention of any applicable law or the by-laws. If that is the case, the auditor's report should include such a remark.

3 In rankings of continuous measures of book-tax conformity from highest to lowest, Atwood et al. (2010) rank Finland as country 14 of 33 with a sample of public companies while Sundvik (2017) ranks Finland as country 6 of 12 with a sample of private companies. When we calculate the empirical measure for book-tax conformity following Atwood et al. (2010) by year for our sample with the available data the average root mean-squared error is 0.011, which we interpret as a sign of high book-tax conformity. According to the description in Eberhartinger (1999), taxation in Finland depends on financial reporting and all book entries are relevant for taxation. Atwood et al. (2010) furthermore note that single entity company accounts are associated with higher book-tax conformity, and we only examine such accounts.

4 The Finnish Government first announced the corporate tax cut on 21 March 2013 as part of a broader structural reform in the General Government Fiscal Plan (Government proposal 194/2013). The tax bill was finally approved by the Parliament on 16 December 2013 and was effective from 1 January 2014. Besides the large change in the tax rate, new rules limiting the deductibility of interest on intra-group loans came into force in January 2014. Official entertainment costs also changed from being 50% tax-deductible to not deductible at all. Simultaneously as the corporate tax rate was decreased, tax on dividends was increased. However, only minor changes were made in 2014 when compared with the reform in 2005 when a regime with a partial double taxation of dividends (partial relief) replaced the previous system of full dividend imputation (avoir fiscal). Therefore, we do not postulate any effect arising from the change in tax on dividends on our analyses.

5 Small Finnish companies report several other discretionary expenditures under the SG&A item, such as the following period costs: fringe benefits, insurance, office supplies, software, living, entertainment, gifts, auto, cell phone, rent, advertising, research and development, and administration.

6 For example, see Supreme Court of Finland case 2007:102 and Helsinki Court of Appeal case 13/2967.

7 In our main analyses, however, we concentrate on audited versus unaudited companies. In Section 5.3, we perform analyses on the four different groups of companies.

8 The conclusions of (untabulated) analyses of a balanced panel of companies with three fiscal years remain similar to that of the unbalanced panel results presented below.

9 TACC = Change in Current Receivables + Change in Inventories – Change in Current Payables – Depreciation.

10 Interestingly, we observe a nontrivial number of unaudited companies above the size thresholds. We examine these incorrectly unaudited companies closer in Section 5.3.

11 Finnish tax auditors officially conduct audits based on tax risk, with region-based and random-based samples. Based on anecdotal evidence, we reason that the large number of incorrectly unaudited companies (unaudited companies above the audit exemption size thresholds) is a result of low enforcement around the thresholds and ineffective passive monitoring during our period of analysis. The passive monitoring became more effective in 2015 when an online anonymous reporting system of suspected tax evasion was implemented at the tax authority's website. During our period of analysis, individuals wanting to report their suspicions had to pay a visit to the tax office, write a letter or make a toll call, which naturally decreased the number of reports and the effectiveness of the monitoring system.

12 Considering the subsample of December fiscal year end companies, as in the first additional test, the (ununablated) coefficient on the TEST × TAX variable is however positive and statistically significant within subsamples of (2) Mandatory Audited vs. Incorrectly Unaudited (t-statistic 2.15), and (3) Voluntary Audited vs. Incorrectly Unaudited (t-statistic 1.68).

13 The Finnish tax authorities annually disclose the industries that will be subject to higher tax audit scrutiny during the forthcoming period. During 2012–2014, the industries of Construction (12,691 company-years) and Accommodation and food service activities (3041 company-years) were subject to higher tax authority scrutiny.

14 The company fixed-effects regression also alleviates concerns about our tests relying on panel data for the years 2012–2014 that could cause standard errors to correlate within years and across time by company.

Additional information

Funding

Dennis Sundvik gratefully acknowledges the financial support from the Society of Swedish Literature in Finland, the Foundation for Economic Education (Liikesivistysrahasto) [grant number 160312], and Suomen Arvopaperimarkkinoiden Edistämissäätiö.