Political Institutions and Cost Stickiness: International Evidence

Abstract

This study explores the association between political institutions and cost stickiness. Prior literature suggests that cost stickiness is driven by the economic implication that pursues firm profits and the agency implication that pursues managers’ benefits. We argue that political institutions affect cost stickiness by constraining government expropriation, which in turn affects the relative importance between these two implications on managerial resource decisions. Using an international sample, we find that strong political institutions are associated with higher cost stickiness. Furthermore, cost stickiness is more likely driven by the economic implication when political institutions are strong and by the agency implication when political institutions are weak, implying that cost stickiness is more likely value-enhancing when political institutions are strong. Mediation analysis reveals that political institutions affect cost stickiness through contract enforcement and corruption control. Our study contributes to the cost behavior literature and the literature about how political institutions affect corporate investment.

Keywords:

• Political institutions
• Cost stickiness

Acknowledgement

We thank Matthias Mahlendorf (the editor) and two anonymous reviewers for their helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We define stronger political institutions as those placing more political constraints on government power, which ensure that governments secure private property rights and comply with policy commitments. With this definition, strong political institutions hinder government officials from seeking rents and thus constrain government expropriation. The term ‘government expropriation’ is not limited to outright seizure and it covers a wide range of rent-seeking activities including solicitation of bribes, overregulation, confiscatory taxation, and reversing a policy or making a regulatory change that breaks commitments or contracts.

2 There are many real cases of government expropriation around the world. For example, in 2004 the Russian government seized Yukos’ assets with the excuse of securing its tax claims against the company. In 2006 the Russian government took over the oil and gas project of Shell by attributing it to environmental concerns. In 2007 Eni of Italy faced disputes with the Kazakhstan government due to new legislation allowing the government to reverse contracts. Also in 2007 Venezuela's leader Hugo Chavez took over the oil projects of ConocoPhilips.

3 The underlying incentive for managers to commit unutilized resources is to pursue firm profits under the economic implication (Anderson et al., 2003) and to pursue personal benefits under the agency implication (Chen et al., 2012).

4 The reasoning for why the economic implication induces greater cost stickiness is outside the scope of our study.

5 Managers may distort resource decisions by cutting resource costs to meet earnings targets (Kama & Weiss, 2013; Weiss, 2010), while this provides another perspective of the agency implication. This perspective should have a second-order effect in the context of government expropriation because high earnings numbers attract attention from governments and make government expropriation more likely.

6 The mechanisms we explore are not exhaustive and there are other mechanisms where political institutions affect cost stickiness. For example, strong political institutions may foster the development of financial markets, which facilitate financing and alleviate managers’ concerns about resource adjustment costs. To highlight the focus of our study, we only explore mechanisms that reflect constraints on government expropriation.

7 Following prior international studies (e.g., Banker, Byzalov, & Chen, 2013; Banker, Byzalov, & Threinen, 2013; Kitching et al., 2016; Lee et al., 2020), we explore operating costs rather than selling, general, and administrative expenses because this not only allows subsequent research to replicate our work but also enhances the comparability of our results with those of prior studies. As presented in Table 3, our inferences remain unchanged if we explore selling, general and administrative expenses.

8 The political uncertainty from elections (Lee et al., 2020) and the economic policy uncertainty (Jin & Wu, 2021) are different from political constraints on government expropriation. This is because a government's policy may have uncertainty due to various reasons such as elections or insufficient capacity to implement the policy, where these reasons do not necessarily capture the uncertainty of policy changes resulting from government expropriation.

9 For example, Glaeser et al. (2004) argue that the indices for the risk of government expropriation or government effectiveness are outcome measures that represent past constraints of the government.

10 The measurement is two-staged. We first regress operating income before depreciation (OI) on the inverse of lagged total assets, current and lagged XOPR (operating costs) from year t−1 to t−14, where OI and XOPR are deflated by total assets. We then multiply the coefficients of lagged XOPR that are positive and significant by a firm's corresponding lagged XOPR and sum their present values assuming a discount rate of 10% to obtain that firm's FV.

11 In these regressions the standard errors of coefficients are clustered by country-year and WLS methodology is used with weights as the inverse of the number of firm-year observations in each country.

12 La Porta et al. (1999) argue that Religion and Ethno are relatively exogenous to government quality so they are ideal control variables. We do not include IRULE because law and order may have some effects overlapping with those of judicial independence, causing it difficult to interpret its coefficient when JUDIND is the dependent variable.

13 Banker, Byzalov, and Chen (2013) report that the difference in cost stickiness between their European sample countries with the least and the most strict employment protection legislation is 11 basis points. Compared with their study, the economic magnitude of our finding is material.

14 In Column (2) of Table 4, the coefficient on DEC*ΔlnSALE changes to positive and this raises the concern that cost stickiness in our sample is driven by those four countries excluded. As a further check, we use the remaining 40 countries to repeat the baseline specification of Anderson et al. (2003) in Column (1) of Table 3 and we find that the coefficient on DEC*ΔlnSALE is negative (−0.037). Following Banker and Byzalov (2014), we estimate this coefficient using equation (1) for each of the remaining countries, where this coefficient is negative for 36 countries and its mean (−0.060) is significantly different from zero at the 1% level (t-statistic = −6.53). These results suggest that cost stickiness in our sample is not driven by specific countries. Based on Balli and Sørensen (2013), the sign of DEC*ΔlnSALE may be changed due to the inclusion of its interaction terms.

15 We include TAE because, like legal institutions, tax enforcement as a well-recognized extra-legal institution has similar effects on disciplining managers (Dyck & Zingale, 2004). However, TAE is time-invariant so it cannot capture the incentives to shift profits across jurisdictions or exploit tax havens (Schenkelberg, 2020; Thomsen & Watrin, 2018) since these incentives are affected by tax regulations that are changed over time. To mitigate this concern we refer to Atwood et al. (2012) by constructing a tax aggressiveness (TG) measure, defined as statutory corporate income tax rates minus firm-level effective rates, where a higher TG represents more tax aggressiveness. We then replace TAE with the country-year means of TG and we find that this does not change our conclusions. The reasoning for this measure is that it can reflect the strength of tax enforcement as Atwood et al. (2012) document that stronger tax enforcement is associated with a lower level of tax aggressiveness.

16 Hofstede's individualism index is relevant to political institutions as countries with higher individualism were more likely to establish political systems with checks and balances (Hofstede et al., 2010). Kitching et al. (2016) find that the effect of individualism on cost stickiness is insignificant, suggesting that individualism satisfies the exclusion restriction.

17 Specifically, Xu and Zheng (2020) find that cash savings from tax avoidance reduce cost stickiness. Their finding implies that in addition to facilitating empire-building, cash reserves or free cash flows also alleviate managers’ concerns about adjustment costs and thus allow them to cut unutilized resources when sales decline. To the extent that paying adjustment costs do not bring private benefits to managers, the reduced cost stickiness induced by CASH and FCF for the high-POLV subsample is more likely economically justified and therefore more consistent with the economic implication. Accordingly, we conclude that the agency implication has a greater influence on resource decisions when political institutions are weak.

18 Interestingly, although insignificant, the coefficient on JUDIND*DEC*ΔlnSALE in Column (4) is positive and this suggests that lower judicial independence is associated with higher cost stickiness. A possible explanation is that lower judicial independence implies that judicial systems are more likely subject to political interferences, in which case courts tend to favor governments so governments have more power in interfering with firm decisions. Because governments face social pressures, they prefer firms to retain unutilized resources since cutting resources may exacerbate social problems such as unemployment. Consequently, lower judicial independence is associated with greater government interferences that in turn force firms to retain more unutilized resources to meet social objectives.