When dedicated investors are distracted: The effect of institutional monitoring on corporate tax avoidance

Abstract

This study investigates whether loosened monitoring from institutional investors affects firm tax planning decisions. We take advantage of shocks to unrelated parts of institutional investors’ portfolios and examine how plausibly exogenous changes in monitoring from institutional investors influence the level of firm tax avoidance. We find that investee firms significantly increase their temporary tax avoidance when there are temporary reductions in the attention of their dedicated institutional investors. Cross-sectional tests show that the tax impact of reduced dedicated investor attention and monitoring intensity is more pronounced when a firm’s information environment is less transparent and when a firm is subject to weaker internal governance. Our findings are robust to alternative research designs.

Introduction

The share ownership structure of a firm is an important determinant of corporate tax planning strategy (Shackelford and Shevlin, 2001, Desai and Dharmapala, 2008, Hanlon and Heitzman, 2010). Many studies in this area provide evidence supporting that ownership patterns affect tax avoidance (e.g., family owners in Chen et al., 2010; insider holdings in Badertscher et al., 2013; managerial equity-based compensation in Desai and Dharmapala, 2006, Desai and Dharmapala, 2009; and institutional investment horizons in Embree and Crabtree, 2012). Khurana and Moser (2013) find that long-term institutional ownership is associated with less tax avoidance. However, most of the studies on share ownership structure and corporate tax avoidance are subject to endogeneity concern as the share ownership structure of a firm could be endogenously determined. This study aims to extend prior literature by exploiting plausibly exogenous shocks to the monitoring strength of institutional investors and identifying the effect of institutional investor monitoring on corporate tax avoidance.

The Scholes-Wolfson framework provides a fundamental building block for empirical tax research, where effective tax planning is a trade-off between marginal benefits and marginal costs of tax avoidance (Shackelford and Shevlin, 2001). The marginal benefits of tax avoidance include increased firm value through tax savings and enhanced shareholder wealth by transferring wealth from the government to a firm’s shareholders (Graham and Tucker, 2006, Frank et al., 2009, Wilson, 2009, Goh et al., 2016), suggesting that shareholders may prefer value-increasing tax avoidance.

However, anecdotal evidence shows that shareholders, especially institutional investors, are concerned with the risks of tax planning at the companies they invest in. As a recent article in the Financial Times points out, large investors fear that the advantage of short-term profitability fueled by tax avoidance “may not be sustainable and could lead to reputational and commercial risks with consumers, governments and regulators” (Marriage, 2014). Academic literature has also emphasized that tax planning may not necessarily increase shareholder value because it could potentially bring in significant costs and risks. The seminal study of Desai and Dharmapala (2006) argues that tax avoidance can facilitate managerial wealth expropriation behaviors, which impair firm value and consequently decrease the wealth of institutional investors. In support of Desai and Dharmapala’s (2006) agency theory of tax avoidance, Chung et al. (2019) document that managers exploit the opacity arising from tax avoidance to extract rent from shareholders by engaging in more profitable insider trading. Relatedly, tax planning activities often increase firm financial complexity and reduce information transparency (e.g., Balakrishnan et al., 2018). The increased information opacity may make it more difficult for institutional investors to interpret the persistence and prospects in the firm’s earnings and cash flows. In summary, institutional investors may suffer from non-tax agency costs in the form of managerial opportunism and information opacity induced by investee firms’ tax avoidance activities.

Being concerned with the potentially significant agency costs and risks associated with tax planning, while institutional investors may allow or even support tax avoidance activities, they also have incentives to monitor their investee firms’ tax strategies to constrain the abuse of tax avoidance.¹ If this argument is valid, when institutional investors’ attention is distracted, the constraints imposed on their investee firms’ tax-avoiding behaviors may be temporarily relaxed as a result of loosened monitoring. The temporarily released restrictions on tax avoidance will potentially lead to an increase in temporary tax avoidance, that is, a temporary deviation from institutional investors’ most desired level of tax avoidance.

In this study, we take advantage of the plausibly exogenous shocks to the attention of institutional investors on a firm and investigate how temporary reduction in the monitoring intensity of institutional investors affects firm temporary tax avoidance. Unlike prior studies that use institutional ownership to infer an institutional monitoring effect² (e.g., Chen et al., 2007, Aggarwal et al., 2011, Boone and White, 2015), we directly capture the changes in institutional monitoring intensity caused by attention-grabbing events in an institutional investor’s portfolio. The idea is that, when institutional investors experience significant information events in some parts of their portfolio, they have to spend time and effort to understand and deal with those events and, as a consequence, pay less attention to other stocks in their portfolio that are not experiencing such material events. The plausibly exogenous variation in institutional investor attention allows us to draw inference regarding the effect of institutional investor monitoring on firm tax avoidance.

This study focuses primarily on dedicated institutional investors (DIIs), based on Bushee’s (2001) classification of institutions, because DIIs are supposed to possess the strongest incentive and competence to monitor their investee firms. To empirically examine whether and how temporarily loosened DII monitoring affects firm temporary tax avoidance, we construct the DII distraction variable (DISTR_DII) in the spirit of the emerging literature (Kempf et al., 2017, Wang, 2021). Specifically, we exploit the plausibly exogenous shocks to unrelated industries held by a firm’s DIIs and examine temporary shifts of DII attention away from the firm towards the part of their portfolios subject to the shocks.³ To illustrate this setting, we use a simplified example as follows. Suppose a DII holds only two companies in the portfolio, Firm A (from the manufacturing industry) and Firm B (from the healthcare industry). Normally, the investor plays a monitoring role in both firms. However, if there is something unexpected happening to the manufacturing industry, such as a large reduction in import tariffs that intensifies the competition from foreign manufacturing rivals, the investor’s attention is likely to shift away from Firm B to Firm A. Because attention is not unlimited, monitoring intensity at Firm B will decrease. More importantly, the attention-grabbing event is plausibly exogenous to the fundamentals of Firm B, enabling us to capture the variation in institutional monitoring and examine how companies such as Firm B in this example react to the loosened monitoring from institutional investors. Furthermore, since in reality a DII can hold many different stocks in its portfolio, we consider the weight of the attention-grabbing industry in the investor’s portfolio to capture the extent to which the investor is distracted. Similarly, a firm usually has multiple DIIs, and hence our measurement also considers the importance of each DII to the firm.

As documented in Kempf et al. (2017), such an institutional investor distraction measure captures a significant reduction in institutional monitoring intensity and is related to a broad variety of corporate actions. The temporarily loosened DII monitoring is likely to result in temporary tax avoidance, which is most likely to take place in the form of deferred taxes. Deferral-based temporary tax avoidance strategies include accelerating deductions and delaying income recognition for tax purposes (Edwards et al., 2016). As such, following the literature (e.g., Chen et al., 2010, Hanlon and Heitzman, 2010, Lennox et al., 2013, Cen et al., 2017, Balakrishnan et al., 2018, Chung et al., 2019), we use two commonly adopted measures as proxies for temporary corporate tax avoidance: cash effective tax rate (CETR) and total book-tax differences (TBTD). These two tax avoidance measures are intended to capture temporary tax avoidance like deferring taxes to the future.

Our sample construction begins with all firms from the Compustat universe over the period 2001–2014. After basic sample screening procedures, we arrive at a final sample of 18,069 firm–year observations for 3,518 unique firms. To examine the relation between temporarily loosened DII monitoring and firm tendency to avoid tax liabilities, we employ a firm fixed effects regression model to further mitigate endogeneity concerns. Our results indicate that temporarily loosened DII monitoring is associated with an increase in temporary tax avoidance of investee firms, after controlling for various tax avoidance determinants and firm and industry-year fixed effects. The increase in tax avoidance is both statistically significant and economically meaningful: for our sample of firms, cash effective tax rate (total book-tax differences) decreases (increase) by, on average, 59 (5) basis points, which translate into 2.3% (5.6%) of the mean, for a one-standard-deviation increase in our DII distraction measure. Our empirical findings thus provide direct and strong support to the literature on the agency cost of tax avoidance, suggesting that DIIs monitor and constrain tax avoidance due to various cost/risk concerns. Temporarily reduced DII monitoring releases restrictions on tax planning, thus leading to an increase in temporary tax avoidance.

To test the robustness of our main findings and establish causality, we use two alternative research designs. First, we conduct a dynamic effect analysis and show that the increased tax avoidance manifests only in the year when a firm’s DIIs are distracted, not before or after. Second, we use propensity score matching (PSM) to assign a low-distraction control to each firm–year observation with a higher level of DII distraction and rerun our main regressions. We find that firms whose DIIs are more distracted exhibit greater tax avoidance relative to their propensity score-matched control firms. An alternative entropy balancing regression analysis yields similar results.

In the cross-sectional analyses, we provide evidence that the impact of loosened DII monitoring on tax avoidance is more pronounced for firms with more volatile stock returns, firms with higher analysts’ forecast error, firms with lower accruals quality, and firms with less voting power controlled by outside directors. These results are consistent with the view that firms are more likely to take responsive actions to reduced DII monitoring on tax avoidance when the information environment is less transparent and when they are subject to weaker internal governance.

Our study makes three major contributions. First, this study contributes to the line of research investigating how ownership structure and corporate governance characteristics affect a firm’s tax avoidance (Chen and Chu, 2005, Khurana and Moser, 2013). We extend Khurana and Moser (2013) with an improved identification exploiting exogenous shocks to the attention of institutional investors and document a plausibly causal impact of institutional monitoring on firm tax avoidance. We provide the first evidence on the effect of DII monitoring on corporate tax avoidance. Our empirical design relies on plausibly exogenous shocks that distract DII monitoring to draw causal inferences. Hence, this study identifies a specific channel, monitoring intensity, through which institutional investors affect corporate tax policies. Second, we add to the literature on the tax implications of different types of institutional investors. Recent studies of Chen et al., 2018a, Khan et al., 2017 show that quasi-indexing institutional ownership newly entering a firm leads to greater tax avoidance. Our findings are opposite to the results documented in the index reconstitution papers. The difference could be potentially due to different objectives of index investors and dedicated investors, or the positive relation between quasi-indexing institutional ownership and firm tax avoidance is driven by factors other than increased institutional monitoring. In addition, some recent studies challenge the use of the index reconstitution setting (Wei and Young, 2017, Gloßner, 2018, Young, 2018). Our study provides a timely reevaluation of the relation between institutional ownership and tax avoidance using an alternative identification strategy. Finally, some recent studies question Desai and Dharmapala’s (2006) manager rent extraction story of tax avoidance (Blaylock, 2016, Seidman and Stomberg, 2017), leaving the question of whether rent extraction and tax avoidance are linked in dispute. However, another strand of studies provides supporting evidence to Desai and Dharmapala (2006) (Atwood and Lewellen, 2019, Chung et al., 2019). The findings of this study contribute to this important but yet uncertain area in the tax literature.

The remainder of this paper proceeds as follows. Section 2 reviews the related literature and develops our hypothesis. Section 3 introduces the sample and presents the research design. Section 4 presents descriptive statistics and the main results. Section 5 sets forth additional tests. Section 6 concludes the paper.