Disentangling the Effects of State Ownership on Investment – Evidencefrom Europe
This paper investigates the relation between state ownership and corporate investment and tests theoretical channels of influence. Using a matched panel of 624 European firms, we find that state ownership curtails firms’ responsiveness to investment opportunities. With increasing government ownership, investment becomes more sensitive to internal funds when capital constraints and external financing needs are high, and the government has minimal control of domestic bank assets. State ownership negatively affects investment levelsbut mitigates reductions during the financial crisis. Overall, our evidence suggests that state ownership is associated with inefficient but stability-seeking investment policies and increased levels of capital constraints.
JEL Classification: G32, G31, G32, G38, D92, E22
Keywords: Government policy and regulation, state ownership, corporate investment
1Introduction
Despite multiple privatization waves globally since the early1980s, the state remains an influential shareholder in many of the world’s largest companies. In fact, recent years have brought an increase of state involvement in the private sector in many countries (Borisova, et al. (2015), Bortolotti, Fotak, and Megginson (2015)) that is sometimes described by the media as a revival of state intervention (The Economist (2012)).[1]Against this backdrop, a growing strand of the literature is dedicated to identifying the implications of state ownership on firm outcomes, such as efficiency, profitability, and the cost of financing(e.g., Gupta (2005), Borisova and Megginson (2011), Borisova et al. (2015)).
Fundamental to this debate are firms’ investment policies, which we are interested in for two reasons. On the one hand, investments build up the essential basis for firms’ future profits and firm value. On the other hand, examining firms’ investments behavior allowsfor deeper insights into the question whether and how state ownership affects firm behavior. In the first-best setting corporate investment should be determined by the firm’s investment opportunities, typically measured by Tobin’s Q (Hayashi (1982)). However, Stein (2003) suggests that there are two fundamental sources that might distort firms’ investment policies: agency problems and financing frictions.
State ownership can be associated with both issues in various ways. For example, owing to imperfect monitoring (Vickers and Yarrow (1991), Borisova et al. (2012)) and various political motives, government owners may exhibit empire-building investment patterns, wherein managers overinvest available internal funds in pet projects (Jensen (1986)). Alternatively, state ownership couldalso give rise to managerial “quiet life” behavior characterized by investment inertia (Bertrand and Mullainathan (2003)) and politicalobjectivesto pursuefirm and employee stability (Fogel, Morck, and Yeung (2008), Boubakri, Cosset, and Saffar (2013)).This would imply that state ownership negatively affects both the level of investment and its responsiveness to growth opportunities. From a financing perspective, as Ben-Nasr, Boubakri, and Cosset (2012) argue, the risk of political interference and agency problems may increase the wedge between internal and external capital costs. This additional financing friction reduces investment by firms requiring external funding. On the other hand, the soft-budget constraint theory suggests that state-controlled firms face lower financing constraints because they can obtain external financing more easily than privately controlled firms, owing to their implicit guarantee of government bailout (Borisova and Megginson, (2011), Ben-Nasr, Boubakri, and Cosset (2012)). In sum, it is theoretically unclear if and, more importantly, how having government shareholders affects firms’ capital allocation.
We employ an empirical testing approach to disentangle these conflicting theoretical predictions about the effect of state ownership on corporate investment. Based on a simplified model of investment in the spirit of Stein (2003) that captures agency issues and capital constraints in an intuitive way, we derive testable predictions of the different theoretical arguments relating to state ownership. The resulting predictions focus on the influence of government ownership on the level of investment and its sensitivity to investment opportunities, measured by Tobin’s Q, and to internal funds. We apply this testing scheme to a panel dataset of publicly listed firms in Europe covering the period from 1997 to 2013. Our motivation to study the European setting stems from the fact that it allows us to examine a relatively large number of firms with significant state ownership located in advanced economies and facing fairly homogeneous (market) conditions (Borisova and Megginson (2011), Borisova et al. (2012)). Indeed, an algorithmic data gathering procedure combining various sources of informationallows us to identify 312 firms with significant state ownership in any of our sample yearsand our matched sample covers more than 5,000 firm-year observations.
Our primary findings are threefold: First, our evidence shows that state ownership curtails the sensitivity of investment to Tobin’s Q. This finding is robust to using various measures of investment activity (Capex, Capex plus R&D, employee changes)andalternative measures of investment opportunities that are not directly related to the firm’s stock price, and to employing advanced estimation methods to control for measurement error in Q. Second, investment-cash flow sensitivity (ICFS) increases with the level of state ownership if, and only if, firms are financially constrained and external finance dependent. This relation particularly prevails in countries where the government has minimal ownership of domestic bank assets. Third, government ownership has a negative direct effect on the level of investment and on absolute annual investment changes. It also mitigates capital expenditure cuts during the recent global financial crisis.To ensure that these insights are not prone to endogeneity or measurement error, we additionally implement instrumental variables, dynamic system GMM models (Blundell, et al.(1992), Blundell and Bond (1998)), and higher-order cumulant estimators developed by Erickson and Whited (2000) and Erickson, Jiang, and Whited (2014). We also rule out alternative explanations for the identified relationship between state ownership and investment. Specifically, we control for the influence of stock price informativeness (Chen, Goldstein, and Jiang(2007)), asset tangibility (Almeida and Campello (2007)), and uncertainty levels (Boubakri, Cosset, Saffar(2013)).
Our insights confirm the general notion that state ownership is associated with inferior investment efficiency, as evidenced by its strongly negative effect on investment-Q sensitivity. The underlying mechanism for this suggests that government shareholders promote investment policies that are best characterized as conservative and stability-seeking. State-owned firms invest less and are lethargic when it comes to adjusting capital expenditures to changing opportunities. Government stock ownership also seems to constitute a source of financing friction. One reason for that might be the reluctance of state shareholders to allow their companies to execute dilutive external equity issues, which makes equity-dependent firms’ investment highly reliant on available internal funds.
Our findings contribute to the literature that speaks to the real effects of state ownership on corporate investment. The popular view is that state-owned firms are inefficient because they overspend funds for non-value maximizing purposes. According tothis argument, we would expect a positive link between government ownership and average investment levels, as well as apositive link between investment and cash flow.Our data does not provide any support for this view. Our evidence also refutes the argument that state ownership has a mitigating effect on firms’ financial constraints. Instead, we find that state shareholders can be burdensome for firms that are in need of new capital.Our paper also sheds light on the general empirical investment literature, which still puzzles over why investment decisions are less sensitive to investment opportunities than the neoclassical q theory predicts. We highlight how decision makers’ stability seeking and conservatism can weaken investment sensitivity. Although Kaplan and Zingales (2000) recognize that investment-cash flow sensitivities can be influenced by managerial conservatism, related studies have largely focused on the empire-building type of agency conflicts and financing frictions to explain firms’ investment patterns.[2]
This paper is related to the studies by Chen et al. (2011) and Chen et al. (2014). Both articles suggest that state ownership negatively affects the sensitivity of investment to Tobin’s Q, which serves as a proxy for investment opportunities. We add to that literature in two important dimensions. On the one hand, instead of studying Chinese state firms or global privatization samples, we focus on European firms with significant state-ownership to better understand the real effects of state ownership in a developed market context, where political institutions, their agenda and governanceare substantially different compared to China. On the other hand, and probably more important, our work goes further in that it employs a comprehensive testing scheme – capturing also the level of investment and its sensitivity to cash flow – that allows us to disentangle the theoretical channels of state owner influence. We additionally rule out the possibility that the resultsare driven by varying degrees of stock price informativeness (Chen, Goldstein, and Jiang (2007), Ben-Nasr and Cosset (2014)) or measurement issues with Tobin’s Q (Erickson and Whited (2000)). Other related articles focusing on Chinese state-owned firms include Firth et al. (2012),who find that government ownership causes overinvestment, and Chen et al. (2015), who show that internal capital market allocations are more efficient in privately owned business groups than in state-owned enterprises.
This study’s findings offer important policy implications. While it is widely accepted that state ownership mightnot be optimal from a firm value maximization point of view, a popular argument in favor of it says that it serves common goals, such as economic growth and employment. Our findings cast some doubt on the validity of this argument, since government owners seem to reduce firm-level investment (measured by various empirical proxies) as well as employee growth. This is in line withFogel, Morck, and Yeung (2008), who find (country level) corporate stabilitybeing positively correlated with the degree of state intervention, but negatively related to economic growth.
The rest of the article is organized as follows. The next section presents the theoretical framework, which features our applied model of investment and the different theoretical arguments related to state ownership. Section 3 describes the data and our primary empirical estimation methods. Section 4provides and interprets our main results and presents additional robustness tests. Section 5 concludes.
2Framework and Hypotheses
The fundamental assumption underlying this study is that companies operate in imperfect markets and agency problems as well as financing frictions influence the amount of capital that firmswill allocate to investment projects. The question of interest is if government owners have an influence on whether firms invest efficiently according to their existing investment opportunities or whether they over-/underinvest conditional on other financial factors, such as internal funds.To hypothesize the theoretical effects of state ownership on corporate investment, we proceed in two steps. First, we introduce a heuristic model of investment that captures agency issues and financing frictions in a simple and intuitive way. Second, we discuss various perspectives on government ownership and the way this will affect firms’ investment decisions. This yields a set of testable predictions about the impact of state ownership on, respectively, the level of investment and its sensitivity to investment opportunities and internal funds.
2.1Investment Model
We examine a simple single period model in the spirit of Stein (2003). The idea is that the management of a firm invests in time 1 an amount , which will produce a return in time 2, where represents the firm’s increasing, concave operating profit function.Investment, however, comes withcapital adjustment costs. In the first best case--in the absence of agency issues and financing frictions--the time-1 value of the firm is given by, where is the risk-adjusted discount rate. Managers aiming to maximize the value of the firm will invest the amount that is characterized by the first-order condition. Thereby, the left-hand side of the equationis frequently referred to as marginal. Two things are important here:First, the optimal investment is independent of financing, as internal and external funds constitute perfect substitutes. Second, in case of a quadratic capital adjustment cost function , optimal investment becomes a linear function in marginal. Hayashi (1982) develops a neoclassical investment model in which the value of the firm is a linear function of invested capital, and thus marginal equals average. As the latter – in contrast to marginal – is empirically observable, the model gains empirical content and investment becomes a linear function in average only, i.e.,
While in the first-best setting, managers invest until the marginal product of capital net of compounded marginal adjustment costs () equals , market imperfections can distort investment decisions. To capture managerial biases and financing frictions in the model, we follow Stein (2003) and Aggarwal and Samwick (2006) and augment the objective function along two dimensions.
On the one hand, to capture direct agency costs resulting from managerial biases, we introduce a term, which reflects the additional value of investment perceived by decision makers. We think of decision makers as insidersthat represent the management of the firm presumably acting on behalf of influential blockholders, probably at the expense of other shareholders, which we call outsiders. Insiders’ valuation of the firm is given by In contrast, outsiders’ valuation is . This means that drives a wedge between the value perceived by insiders and the value perceived by outsiders, and canconsequently cause investment distortions as illustrated in insiders’ first-order condition. The sign and extent of this distortion is a matter of the structure of . For instance, modelling empire-building preferences generally requires .[3] Alternatively, quiet life or stability seeking preferences (Bertrand and Mullainathan (2003)) involves an inverse U-shaped climaxing in , which – in a more dynamic view – could be thought of as a function of lagged investment.
On the other hand, to capture financing frictions, we allow for deadweight costs associated with raising external funds . We model these costs by , where is an increasing convex function and measures the degree of financing frictions. Effectively, we assume that value-maximizing managers make full use of internal funds before they approach the capital market and raise external financing, which implies that . The value perceived by insiders (given managerial biases and financing frictions) thus is
/ (1)and insiders’ first-order condition for maximizing with respect to investment reads as follows:
/ (2)From equation (2), we learn that (i) direct agency costs captured by generally distort the linear investment- relation from the first-best setting and (ii) distortions from financing frictions will mainly occur in case optimal investment is larger than internal funds and . Accordingly, firm investment is expected to depend not only on, but also on , which in turn determines a firm’s need for external finance (Fazzari, Hubbard, and Petersen (1988), Kaplan and Zingales (1997)).[4]
2.2State Ownership and Investment: Theoretical Channels of Influence
State ownership canaffect firms’ agency problems as well as their degree of financing frictions and thus their investment policies in various ways. In this section, we outline the major relevant theoretical arguments and organize them according to their association with agency aspects (reflected in ) and financing frictions (reflected inand ). This allows us to structure the various theoretical arguments and to derive testable predictions that help us to disentangle the different channels through which state ownership impact corporate investment.
2.2.1Empire-building and Overinvestment
The first line of arguments posits that state ownership is systematically related to decision makers’ empire-building motives and overinvesting activities (Chen et al. (2011), Firth et al. (2012), Chen et al. (2014)). There are at least two rationales for this view. First, from a managerial perspective, state ownership is generally considered to be associated with weak monitoring of managers (Vickers and Yarrow (1991)). Related to this, Borisova et al. (2012) show that government control is negatively related to corporate governance quality. Moreover, state-owned firms are frequently less exposed to external disciplining forces, such as product market competition or takeover pressures (Chen et al. (2014)). Second, from a political perspective, the state as a major shareholder might have multiple (socio-economic) objectives that promote overinvestment (Firth et al. (2012)). Boycko, Shleifer, and Vishny (1996) suggest that governments are willing and able to subsidize inefficiently high output in order to maximize employment or achieve other socially desirable goals, such as locating production capacities in economically underdeveloped but politically important regions, providing cheap goods and services, and producing unnecessary products (Boubakri, Cosset, and Saffar (2008)). As Megginson, Nash, and van Randenborgh (1994) argue, this logic implies that state-owned firms tend to invest relatively more than firms without state involvement.
These arguments suggest that with respect to government owners, insiders’ marginal additional value of investment is positive (i.e., BI 0), which constitutes a distortion to the investment-q relationship. Further, according to the empire-building concept outlined by Jensen (1986), Stulz (1990), and Stein (2003), the additional benefits from investing depend on the availability of free cash flow. If insiders are cash constrained, they will invest less, because external capital markets will discipline managerial behavior and only provide limited funds to prevent investment in unprofitable projects. As a result, insiders’ investment will be positively related toavailable internal funds, as evidenced by Blanchard, Lopez-de-Silanes, and Shleifer (1994). Under the assumption that state ownership is positively associated with the described agency problems, this argument leads to the following predictions: