THE EFFECTS OF POLITICAL FACTORS ON EARNINGS MANAGEMENT
A CROSS-COUNTRY STUDY ON THE EFFECT OF POLITICAL INSTABILITY, CORRUPTION, THE LEVEL OF DEMOCRACY
AND LEFT-WING GOVERNMENTS ON THE OIL AND GAS INDUSTRY
By Obeida Issa
A Thesis
Submitted in Partial Ful llment of the Requirements for the Master of Science in Accountancy
THE EFFECTS OF POLITICAL FACTORS ON EARNINGS MANAGEMENT
A CROSS-COUNTRY STUDY ON THE EFFECT OF POLITICAL INSTABILITY, CORRUPTION, THE LEVEL OF DEMOCRACY
AND LEFT-WING GOVERNMENTS ON THE OIL AND GAS INDUSTRY
Author : Obeida Issa
Studentnumber : S2015055
Date : 30-08-2012
University : University of Groningen
Faculty : Economie & Bedrijfskunde
Study Programme : Master of Science in Accountancy Prime Graduation Supervisor : Dr. R.B.H. Hooghiemstra
Second Graduation Supervisor : Dr. T.A. Marra
Foreword
This master thesis is a result of the research that I carried out at Ernst&Young in Amsterdam during my graduate studies at University of Groningen.
Using this opportunity, I would like to express my gratitude and appreciation to the people who granted me with a full support and understanding.
First and foremost among them is my thesis supervisor, Dr. Reggy Hooghiemstra, who has provided me with a tremendous insight and guidance on a variety of topics.
Special thanks go to Gabro Issa and Lina Issa who made the most generous gesture in my life by their full support at all times. I would like to thank mr. Mary Issa and Ilonca Issa who provided me with their help whenever I needed it.
On behalf of all those mentioned, I would like to use this unique opportunity and wish peace to Syria, the country of my origin, the country that has been an inspiration for me to write this thesis.
Abstract
This paper attempts to tighten the gap between the political economy theory and financial accounting theory by elucidating how political instability, corruption, the level of democracy and left-wing governments have an impact on earnings discretion practices across the
countries within the oil and gas industry. A large body of literature has assumed that the main source of the political costs hypothesis is caused by the company itself while taking the events such as the Gulf War in consideration. Those papers state that the political pressure on
companies is caused by the monopoly of power , their high profits and external privileges (e.g. Watts and Zimmerman; 1986 and Han and Wang (1998)), rather than the political climate of a given company. This paper demonstrates that over the period of 1998 – 2009, together with the changes in political climate, the incentives to manage earnings change as well. By that means, another argument can be added to the political costs hypothesis, namely the political dimension. In this regard, this paper predicts that the current changes in a
political climate caused by the Arabic Spring will influence the choice of accounting policies for oil and gas companies, resulting in earrings management for mineral rich countries in the Middle East and North Africa. In addition to this, the factors affecting the political change such as the resource cures and resource nationalism will result in an equal outcome, yet these effects will not be restricted to the Middle East and North Africa.
Keywords: earnings management, political cost, political instability, corruption, democracy, left-wing.
Table of Contents
1 Introduction 1
1.1 Motivation 1
1.2 Theoretical question & structure 2
1.3 Relevance 2
1.3.1 Financial accounting theory and political economy 2
1.4 Earnings management 4
1.4.1 Country level 4
1.4.2 Industry 5
1.4.3 Political costs hypothesis 5
1.4.4 Earnings management and political cost hypotheses in the oil and gas industry 6
2 Theory 7
2.1 Earnings management 7
2.1.1 Definition 7
2.1.2 Goals and motivations 7
2.2 The political costs hypothesis 9
2.3 Hypotheses 10
2.3.1 Political instability 10
2.3.2 Corruption 11
2.3.3 Level of democracy 12
2.3.4 Left- or rightwing government 13
2.3.5 Correlation among the hypothesis 14
3 Methodology 15
3.1 Sample and data 15
3.2 Earnings management 16
3.3 Sensitivity test 17
3.3.1 Jones 17
3.3.2 Dechow & Dichev as modified by McNichols 17
3.4 Research design 18
3.5 Regression analysis 19
3.5.1 Models 19
3.5.2 Independent variables 19
3.5.3 Control variables 20
3.5.4 Multicollinearity 21
4 Results 22
4.1 Descriptive statistics 22
4.2 Political factors and abnormal accruals 23
4.3 Sensitivity analysis 31
4.3.1 Alternative earnings management models 31
4.3.2 Alternative variables 40
4.4 Summarized results 41
5 Discussion 42
5.1 Conclusion 42
5.2 Limitations & Suggestions for future research 43
6 Appendix 46
7 Bibliography 47
1 Introduction
1.1 Motivation
The oil and gas sector has changed significantly over the last years. The most challenging changes that oil and gas companies face these days are politically related.
According to the 2011 issue of 'Turn risks and opportunities into results', a report based on interviews with oil and gas sector commentators and more than eighty oil and gas executives from fifteen countries, the top four risks identified relate to governmental or political issues.
Specifically, they involve access to reserves, political constraints, and competition for proven reserves. The first risk is primarily caused by a combination of factors such as: political unrest in North Africa as well as the Middle East, high oil prices and the rivalry between the government-backed and opposition parties. The second risk fluctuates around the uncertainty regarding future energy policy of both consuming and producing countries. The top driver behind this risk refers to the sector- specific regulatory pressures. The third risk is cost containment. The interviews indicated that factors which cause this risk are more often legislation restricting flexibility to cut costs and political pressure on organizations when considering cuts. The fourth factor concerns fiscal terms, where “under uncertain market conditions governments tend to make regulatory changes to their benefits” (Ernst & Young, 2011). In addition to the four risks stated above, events like the so-called “Arabic Spring”
have changed the political climate of the resource-rich countries located in North Africa and the Middle East.
Besides these developments, the oil and gas sector is also a subject to “resource nationalism” and “the resource curse”. Although, there is a multitude of different definitions and interpretations of resource nationalism, it is generally used to describe the tendency of societies and governments to assert control over natural resources located within their country. This tendency caused by the theories like peak oil, which made societies and governments more aware of the limits of natural resources and has led many governments to take control of fossil fuel reservoirs, for strategic as well as for economic reasons. Resource nationalism involves two components, the first one has to do with limiting the operations of private international oil companies (IOCs) and the second component with asserting a greater national control over natural resources development (Stevens, 2008). The idea behind the resource curse is that natural resources might be more of an economic curse than of a blessing. Auty (1993) used the term resource curse to describe how countries rich in natural resources were unable to use that wealth to boost their economics and how these countries had lower economic growth than countries without an abundance of natural resources. Ross (1999) explains the resource curse by using three categories. The first one, which is cognitive explanation, refers to a vast amount of resources and creates a type of short-sightedness among policymakers. Societal explanation with resource exports, tends to empower groups that favor growth-impending policies. And the third one, that is the state-centered explanation, refers to a vast amount of resources which tend to weaken state institutions.
Numerous of studies illustrate the broad extend of the resource curse such as: conflicts within societies (Collier, 2003), change in taxation (Bräutigam, 2008), increase of corruption (Koenig, 2004) and worsening human rights (Friedman, 2006). In other words, natural resources lead to resource nationalism and the resource curse, which affects the political climate of a country by creating a political climate with a potential to enhancing political pressure on oil and gas companies. According to the positive accounting theory of Watts and Zimmermann (1978), these developments create incentives for managers to manage their earnings.
1.2 Theoretical question & structure
All these developments indicate that the political climate in which the oil and gas companies operate is changing. While from an accounting perspective a multitude of different questions can be raised, this research will investigate whether these political developments shape the choices of accounting policies by firm managers. A more detailed research question of this paper is:
Do political factors influence the incentives to manage earnings within the oil and gas industry?
The remainder of this study is organized as follows. The next paragraphs discusses the relevance of this study and reviews the previous literature. Section two discusses earnings management (specifically the political cost hypothesis) and develops hypotheses. This is followed by a description of the sample and an outline of research methodology. The empirical findings are discussed in section four. This is followed by conclusions, limitations, and suggestions for future research.
1.3 Relevance
1.3.1 Financial accounting theory and political economy
This paper attempts to tighten the gap1 between political economy and financial accounting. This gap has become of a greater concern for the oil and gas industry since the political climate has changed significantly. Recent developments causing this change in political climate are the Arabic spring, the resource curse and resource nationalism. These developments require both careful and in-depth analyses.
Studies that are consistent with the resource curse show a change in the political climate caused by mineral resources. Ross (2001) has shown that wealth coming from mineral resources not only tends to reduce economic growth and increase the likelihood of civil war (i.e. political instability), but oil causes a great damage to democracy as well, and is not only restricted to the Middle East. Additionally, Tsui (2010) found in his study a causal relation between oil wealth and long-term democratic development. The discovery of 100 billion barrels of oil pushes a country’s democracy level almost by 20 percentages down.
Bhattacharyya and Holder (2010) found, next to the relation between mineral resources and the level of democracy, also a tendency to corruption. Namely, resource-rich countries have the tendency to be corrupted because resource windfalls encourage governments to engage in rent-seeking. An increase in oil rents2 significantly increases corruption and deteriorates state stability (Arezki & Brückner, 2011).
Studies in relation to resource nationalism are scarce. Research focused on the Kazakhstan market shows a tendency of a greater governmental control over its petroleum industry, strict fiscal regime to capture more revenues, and an increase in governmental ownership in major oil and gas projects (Sarsenbayev, 2011). As indicated, various studies find evidence in support of a relation between mineral resources and deteriorating political factors (increase of political instability, increase of corruption, and a decrease of the level of
1 This gap is recognized by Deegan (2002).
2 Oil rent is the difference between the value of crude oil production at world prices and total costs of production.
democracy) (Ross M. L., 1999; Koenig, 2004; Friedman, 2006; Stevens, 2008; Bräutigam, 2008; Tsui, 2010; Bhattacharyya & Hodler, 2010; Arezki & Brückner, 2011). This paper contributes to these political studies by giving a new insight, and by looking at the influences of these factors on firm behavior (i.e. earnings management in response to political factors).
A political instable state, being the result of the Arabic spring, has many causes.
Although these causes are not yet fully determined, it is reasonable to assume that the resource curse and resource nationalism are of a great influence.
As indicated previously, various studies find evidence in support of the theory that natural resources influence politics (Ross M. L., 1999; Koenig, 2004; Friedman, 2006;
Stevens, 2008; Bräutigam, 2008; Tsui, 2010; Bhattacharyya & Hodler, 2010; Arezki &
Brückner, 2011), and thereby the process of decision-making in government. The decision- making concerning the business environment of a state strongly depends on the political spectrum of the government. Many studies contributed to this statement by classifying right- wing governments as those that favor less state- control, and left-wing governments to those that prefer more state- control (Beck et al., 2001). Empirical evidence shows that privatization is associated with right-wing majority of office (Bortolotti et al., 2003). Consistent with this idea, Bortolotti et al. (2001) found empirical evidence that privatization is more likely to be implemented by right-wing governments. In addition to the statements above, another statement consistent with the privatization and right-wing literature can be made about resource nationalism. First of all, oil and gas rich countries with left-wing governments are more likely to nationalize the oil and gas industry. Anecdotal evidence in line with this statement is Argentina's decision to take control of energy company YPF SA, controlled by Repsol YPF SA (a Spanish IOC) (Dittrick, 2012) and in early 2007, Venezuela’s president Hugo Chavez gave some of the world’s biggest oil companies the choice, to turn over majority control of their projects to a state-owned company and remain as minority partners, or face a complete nationalization of operations in Venezuela’s Orinoco River basin (Johnson, 2007). These statements add a new political perspective to this paper, because left- or right- wing governments as shown above, are widely studied in the financial context, but in the earnings management context it has been an overlooked subject. Therefore, this paper contributes to these studies by determining whether left- or right-wing governments affect incentives to manage earnings.
Figure 1. illustrates how political climate changes for the oil and gas industry, and summarizes the relations discussed above.
Figure 1 Conceptual relations examined in knowledge related studies
Factors causing Change in political climate
change in political climate
Resource curse
Resource nationalism Arabic spring
Political Instability
Corruption
Level of democracy
Left-wing
1.4 Earnings management
1.4.1 Country level
The political aspect can be seen as a systematic difference between countries (also known as institutional factors). A large body of literature contributes to institutional factors related to earnings management. For instance, relatively lower levels of earnings management occur in countries with strong investor protection, outsider economies with dispersed ownership, and large stock markets, compared to countries with weak investor protection, concentrated ownership, and less developed stock markets (Leuz et al., 2003). In a similar vein, Francis and Wang (2006) found that abnormal accruals are smaller when country’s investor’s protection becomes stricter (i.e. stronger protection to investors). Other research found that protection of state enterprises by the government reduces the pressure on managers to manipulate earnings, causing lower levels of accrual-based earnings management among state-owned corporations compared to privately held firms (Wang & Yung, 2011; Ding, Zhang, & Zhang, 2007). Furthermore, studies indicate that earnings management is used in order to prevent high taxes (Desai & Dharmapala, 2009; Guenther, 1994). Lara, Osma and Mora (2005) differentiate between civil law and common law. The authors argue that in civil- law countries, managers have incentives to reduce earnings.
Except for the evidence of institutional factors, studies show that managers’ incentives to manage earnings depend strongly on the countries culture (Desender et al. , 2011; Doupnik, 2008). National culture and institutional structure (i.e. investor’s protection) are both important factors for earnings management (Han et al., 2010). Studies focused on institutional and equivalent factors related to earnings management are inexhaustible. This paper attempts to contribute to these studies by using the political factors as a systematic difference across countries, in order to identify a new institutional factor causing managers to manage earnings.
Towards a better understanding of how these institutional factors influence financial reporting, Gray (1988) constructed a frame for the development of accounting systems, acknowledging multiple dimensions influential factors (see figure 2). According to this framework, the particular way in which a country’s accounting system develops, is influenced by the accounting values of accountants, and by the country’s institutional framework, both of which are influenced by the cultural values. The change in political climate as discussed above may have an effect on multiple elements of this framework.
Figure 2: Frame for the development of accounting systems
(Gray, 1988)
External influences Force of nature Trade Investment Conquest
Ecological influences Geographic
Economic Demographic Genetic/hygienic Historical Technological Urbanization
Cultural Dimensions Individualism Uncertainty avoidance Masculinity
Institutional Consequences Legal system
Corporate ownership Capital markets Professional associations Education
Religion
Accounting Values Professionalism Uniformity Conservatism Secrecy
Accounting Systems Authority
Enforcement Measurement Disclosure
A large amount of the earnings management studies is focused on a country level, for instance by only looking at the American market (Healy, 1985), the Canadian market (Magnan et al., 1999), or New-Zealand Market (Navissi, 1999), or by comparing multiple countries together, for example Australia, New Zealand, and the United Kingdom (Black et al., 1998). In order to identify the effect of the political factors, it is important to categorize the countries according to a degree that a given country is a subject to those factors. This paper will contribute to the studies by focusing on a country-level, and by investigating the effect of the political factors worldwide. Which so far has been largely overlooked.
1.4.2 Industry
A large part of the literature focuses on earnings management in a specific industry.
Examples include the printing and publishing industry (McNichols & Wilson, 1988), industries with large amount of research and development expenditures (Dechow & Sloan, 1991), industrial firms (Aharony et al., 1993; Bartov, 1993), business services industry (Erickson & Wang, 1999), the petroleum refining firms (Han & Wang, 1998; Hall ,1993) and a combination of the petroleum refining and the crude oil industry (Hall & Stammerjohan, 1997). This research contributes to these studies by analyzing the oil and gas industry. Han and Wang (1998) researched the oil and gas industry, but with a focus on refining in the United States. The study of Hall and Stammerjohan (1997) has partly the same focus as this paper, with the exception of the fact that this paper will focus on all the oil and gas activities (i.e. upstream and refining activities) worldwide rather than focusing on the refining industry within one or a select number of countries. The focus on the oil and gas industry does not imply that this research is previously performed within another industry, but as argued above, the oil and gas industry is the subject to significant political changes and influences. Having this in mind, it is reasonable to assume that if the political factors affect the incentives to manage earnings, they will be detected within the oil and gas industry.
1.4.3 Political costs hypothesis
In this study, the positive accounting theory (PAT) of Watts and Zimmerman (1978) will be used to answer the question whether the political factors influence the incentives to manage earnings. The predictions made by PAT are largely organized around the three hypotheses; bonus plan hypothesis, debt covenant hypothesis, and political cost hypothesis (Scott, 2011). The political factors as stated above might affect the political cost hypothesis.
Considering the fact that political factors are perceived as the top four risks for the oil and gas companies (Ernst & Young, 2011), it requires a deeper analysis whether these factors influence the political cost hypothesis and to what extent they influence the political cost hypothesis. This means that the political factors and the political cost hypothesis, when merged, provide a unique observation of the effect that political factors have, regarding the accounting policy choices (made by companies' managers) what eventually leads to the effects on earnings. Taking a closer look at the political cost hypothesis, we can find several factors which might affect this hypothesis. This is why the political costs hypothesis has been the subject of a large amount of scientific articles. Scott identifies three examples of the political costs hypothesis. First of all, high earnings may attract media and consumer attention, and causes for political pressure when politicians respond with new taxes or other regulations. Secondly, large companies are seen as environmentally responsible. If they have high earnings, political costs will be magnified. Finally, foreign competition may lead to reduced earnings, as a consequence of lobbying, therefore politicians would respond with grant import protection (Scott, 2011). In this study, I focus on earnings management and political cost hypotheses in the oil and gas industry.
1.4.4 Earnings management and political cost hypotheses in the oil and gas industry
The political costs hypothesis of Watts and Zimmerman (1978) is not a new topic for the oil and gas industry. Hall and Stammerjohan (1997) found a positive relationship between earnings reducing discretionary accruals when oil companies faced a lawsuit with regard to potentially large damages. In an earlier study, Hall (1993) found that refining firms reduced their earnings, using discretionary accruals in periods when gasoline prices were high, and enhanced their earnings, using discretionary accruals, in the periods with lower oil prices.
These findings are consistent with the conclusions of Han and Wang (1998). The results of their research show evidence that refining firms used accruals to decrease their earnings as a response to a sudden increase in gasoline prices during the Gulf War in 1990. In other words, Han and Wang (1998) found a significant relation in high gasoline prices and the discretionary accruals. A non-negligible element during the Gulf War was left out of the scoop, namely the change in political climate. The oil and gas industry is not the only industry having knowledge on external privileges, for example the sudden rise of oil and gas prices. It is also common in the related electricity industry. The study of Gill-de-Albornoz and Illueca (2005) provided parallel evidence with regard to their research on the Spanish electricity industry. The electricity companies engage in earnings reducing accounting policies when the government raises the electricity tariffs, and by so doing, they reduce their political visibility and avoid social objection and demands for a further decrease in tariffs. In the light of the literature on the subjects of oil and gas, and other related industries, there is a correlation between the political influences and changes in the external windfalls.
Political cost studies argue a change in accounting policies from a company perspective (e.g. size (Watts & Zimmerman, 1978))(Watts & Zimmerman, 1978))or a perspective which cannot be directly associated with politics (e.g. external windfalls (Gill-de-Albornoz et al., 2005; Han & Wang, 1998)) (Stolowy & Breton, 2004). This study attempts to investigate the political cost hypothesis from a political perspective by investigating the influence of the political factors, on the level of earnings management by oil and gas companies which have been largely overlooked.
2 Theory
2.1 Earnings management
2.1.1 Definition
A common used definitions of earnings management is the definition of Healy &
Wahlen (1999). They define earnings management as follows: “Earnings management occurs when managers use judgments in financial reporting and in structuring transactions to either mislead some stakeholders about the underlying economic performance of the company or in influence contractual outcomes that depend on reported accounting numbers.”
The context of earnings management is used in a wide spectrum. It includes reporting choices (choice of accounting policies, estimation, presentation, choice of frequency and timing of publication), and real transactions (choices regarding the structure and volume of production and investment activities) aimed at influencing financial reporting to achieve certain objectives (Stolowy & Breton, 2004). Stolowy and Breton (2004) argue that earnings management is a part of accounts manipulation. The difference between accounts manipulation and earnings management is that the accounts manipulation might take place outside the limit of laws and generally accepted standards (i.e. fraud).
2.1.2 Goals and motivations
Stolowy & Breton (2004) based their framework on the understandings of the accounts manipulation, and the possibility that the wealth resulting from information asymmetry that occurs between managers, and other categories of stakeholders, can be transferred from one stakeholder to another. From this agency theory-based perspective, theoretically managers may use earnings management in order to take an advantage of this information asymmetry by using earnings management to increase the remunerations of the managers.
As such, investors are likely to make allowances for it (Dye, 1988). The potential of earnings management to transfer wealth also leads to an external demand for earnings management when existing shareholders want the market to attribute a better value to the firm (i.e. increasing the firm’s value). The increase in firms value results in a potential wealth transfer from new shareholders to the old ones (Schipper, 1989). This process of creating a better value for the firm can also be seen as an action of managers to influence investors’
perceptions of the firm (Degeorge et al., 1999), for example by encouraging investors to buy company’s stock, by increasing a market value of the firm (Kellog & Kellog, 1991). This is an example of the earnings management pattern, also known as “income maximization”.
Earnings management is not only used to create a better value for a company. In some cases, managers have incentives to reduce earnings. When this is carried out by using as many income-decreasing accruals as possible, professional jargon defines it as “taking a bath”.
Studies found that taking a bath occurred while the lower limit of the bonus window could not be reached efficiently. The conclusion drawn, was that it was better to go as low as possible, in order to clean the way for future periods (Healy, 1985). Furthermore, the new management has a tendency to remain very pessimistic about the values of certain assets, and responds by impairing them. The reason for this tendency is that the lower earnings can be blamed on the old management and future income will be free of these charges (Moore, 1973).
The information asymmetry mentioned above is a fundamental problem of financial accounting theory. This problem occurs when it is virtually impossible for shareholders and creditors to observe directly the extent and quality of top manager’s effort on their behalf (i.e.
moral hazard), or when the firm manager will know more about the current condition and future prospects of the firm than the outside investors (i.e. adverse selection) (Scott, 2011). In order to solve this problem, the income measure for a moral hazard needs to be the same as the income measure for adverse selection. Conflicting interest of information between investors (information that enables better investment decisions) and managers (information about their performance in running the firm) makes this problem more difficult to solve, especially if both parties are utility maximizers. The agency theory is typically used to describe solutions to such conflicting interests in the presence of information asymmetry.
Within the agency theory, an agency relationship is defined as a contract under which the principal (i.e. the shareholders and/or stakeholders) engages the agent (i.e. a manager) to perform some services on their behalf which involve delegating a decision-making authority to the agent. The principal can bound divergence from his interest by establishing correct incentives for the agent, and by incurring monitoring costs designed to limit the aberrant behavior of the agent (Jensen & Meckling, 1976). In other words, the shareholders can use contracting to deal with the information asymmetry by controlling manager’s reporting behavior (Evans & Sridhar, 1996).
It is important to recognize, that organizations can be largely described as a “nexus of contracts”, i.e. a set of contracting relationships among individuals (Jensen & Meckling, 1976). The firm will want to minimize various contracting costs associated with these contracts, in order to achieve efficient contracts. Firms will choose accounting policies to attain this goal. Therefore, it is advantageous to give managers some flexibility to choose from a set of available accounting policies, for they can adjust to new or unpredicted conditions. However, this flexibility may lead to opportunistic behavior, and cause managers to maximize their utility, by reducing contract efficiency (Scott, 2011). Watts and Zimmerman (1978) confronted the nexus of contracts and the opportunistic behavior by initiating the positive accounting theory (PAT). The authors state that “PAT is concerned with predicting such actions as the choices of accounting policies by firm managers and how managers will respond to proposed new accounting standards” (Scott, 2011). They use three hypotheses to predict such action of managers. Firstly, the bonus plan hypothesis which states that managers choose accounting policies to redistribute wealth to themselves at expense of another party, when a bonus plan is a part of their remuneration. This would lead to incentives for managers to choose accounting policies which shift earnings from future periods to current period, in order to achieve their bonus (Watts & Zimmerman, 1990). Secondly, the debt covenant hypothesis according to which the companies will use accounting policies to avoid violation of debt covenant, which might cause fines. By creating an incentive for managers, the earnings will shift from future periods to current period. The choice of accounting policies, predicted under circumstances like the bonus plan and debt covenant hypothesis is known as ‘income maximization’. And thirdly, the political cost hypothesis which states that the political sector has the authority to affect wealth transfers between different groups (Stigler, 1971). Corporations are especially vulnerable to these wealth redistributions when certain groups of voters have an incentive to lobby for the nationalization, expropriation, break-up or regulation of an industry or corporation. This provides an incentive for managers to propose such action, by choosing accounting policies to defer reported earnings from current to future periods. (Watts & Zimmerman, 1978). This choice of accounting policies is also known as ‘income minimization’. Such an action, when carried out by the manager signals the stakeholders that the organization is not doing that well as one might expect, and
decreases the organizational pressure from its stakeholders. The political costs hypothesis will be explored in detail in the next paragraph.
2.2 The political costs hypothesis
Before the political costs hypothesis will be furtherly investigated, it is essential to mention that society, politics and economics are inseparable from each other, and the economic issues cannot be meticulously investigated in the absence of considerations political, social and other institutional frameworks, in which the economic activity takes place. This statement is consistent with the political economy theory and the legitimacy theory. The political economy theory explicitly recognizes power conflicts that exist within society and the various struggles that occur between different groups within the society. The legitimacy theory argues that society, politics and economics are connected through so called
“social contracts”3 (Deegan, 2002). In conflicts with corporations, politicians have the upper hand. They have the power to affect the wealth re-distributions of a corporation, by introducing corporate taxes, regulations, subsidies, etc. This power is not only restricted to conflicts between corporations and politicians, but also certain groups of voters (i.e.
stakeholders) may indirectly enforce this power through lobbying (Stigler, 1971). Watts and Zimmerman (1987) argue that managers are aware of the power which lies in the hands of politicians. Their study predicts that managers can fend off such a pressure from politicians (and indirectly from stakeholders) by reporting lower earnings. In connection with that, attention that high profits draw because of the public association, high reported profits and monopoly rents are avoided. Anticipation of the manager on the subsequent decisions of politicians is an act acknowledge within the game theory (Scott, 2011).
By that means, the source of concern Watts and Zimmerman have in mind, are external privileges (e.g. large profits) in association with an abuse of the monopoly power (Milne, 2002) by politicians (and implicitly for the stakeholders as well). Consequently, the main source of the political costs hypothesis is caused by the company itself. This paper investigates whether the political climate influences the accounting policy decisions of a manager. This will be further elucidated on the basis of an example.
Historically, the major oil companies received a widespread public attention when they reportedly had profits. This has been used as an excuse for political actions against them. For example, during the 1990 Persian Gulf crisis, consumers and politicians complained about sudden increase of gasoline prices at service stations. Increasing political costs for oil companies included the revival of federal or state controls on oil and gas prices and the renewal of the windfall profits tax or other form of excessive profits tax (Han & Wang, 1998).
This was also the case in 1980, when new taxes where introduced, as a reaction to unusual high profits reported by oil companies. Han and Wang (1998) investigated sudden increase in oil prices caused by the Persian Gulf crisis. The authors found that oil companies managed their earnings, as a reaction to sudden increase in oil prices. Taking a closer look at this period of time, it would be ignorant to state that the change in oil prices is the only change. Along with a price reaction, the Gulf War also caused a change in the political climate. The authors did not isolate the potential effects of political factors on earnings management, but they also assumed along with other studies that political pressure on companies is caused by the
3 It is worth noting that the theoretical construct of the social contract in not new, having been discussed by philosophers such as Thomas Hobbes (1588-1679); John lock (1632-1704); Jean-Jacques Rousseau (1712-1778). However that this concept, which is used in philosophy and politics literatures, has been embraced within accounting research (Deegan, 2002).
monopoly power and their high profits. Therefore, in my paper I will investigate whether the political costs hypothesis has another, namely a political dimension.
Political influences can be found in governance of a government. Kaufmann et al.
(2010) define it as “the traditions and institutions by which authority in a country is exercised. This includes (a) the process by which governments are selected, monitored and replaced (i.e. the degree of democracy and political stability), (b) the capacity of the government to effectively formulate and implement sound policies (i.e. government effectiveness, left- or right-wing and regulatory quality), and (c) the respect of citizens and the state for the institutions that govern economic and social interactions among them (i.e.
rule of law and control of corruption).” The effect of political influences will be further discussed in the following section due to political related hypothesis, which is inseparably related to the oil and gas industry.
2.3 Hypotheses
As reviewed above, there is limited guidance from previous study regarding the political factors in accounting theory. This paper attempts to incorporate the literature and the formulation of the hypotheses. Therefore, this study could be perceived as exploratory in nature, and perhaps could also become a first step in examining the political factors in the accounting theory.
2.3.1 Political instability
As stated above, the political economy theory argued that society, politics and economics are inseparable from each other. Consistent with this view, organizations are a part of a broader social system. This perspective is provided by the legitimacy theory, and can be directly related to the concept of a ‘social contract’. It is considered that the survival of an organization will be threatened when society arrives to a conclusion that the organization has breached its social contract. In case when the society is not satisfied, this society will effectively revoke the ‘contract’, which has been agreed with the organization (Deegan, 2002). Societal disapproval can be also related to politics. In such a case, it functions as the political instability. The Economist defines political instability as the “High potential of threat posed to governments by social protest or political violence. This high vulnerability to unrest is mostly caused by economic or political distress and can lead to changes in the political system” (Economist Intelligence Unit, 2010). The Arabic Spring provides an excellent example of such a phenomenon. The changes that have occurred among societies of North African and Middle Eastern countries serve as an example of the revoking of the contract between the society and embedded governments. This phenomenon can be explained by using the resource curse (assuming that the countries which are subject to the Arabic spring are oil-rich countries). In the case, when one of the assumptions of the resource curse is that oil-rich countries receive a considerable amount of their income from oil (e.g. oil revenue and oil-rents), the government is depending less on tax income from the society. This higher degree of independence increases the gap between the government and the society, causing the government to be less willing to satisfy the societal demands (Bräutigam, 2008).
As stated above, the legitimacy theory predicts that the society will revoke the social contract, in case when they are dissatisfied. As the Arabic Spring has demonstrated, a reaction to societal discontent with the government results in mass protests, until the embedded governments either quit or satisfy their demands. Oil and gas companies might feel endangered, because society will make no distinction between governments and oil
companies. It will happen for several reasons. For instance, the government is the main business partner of oil and gas industry. According to this, the gap between society and oil and gas companies grows simultaneously with the gap between the society and government.
In addition to that, operations of the world economy are dominated by activities of giant corporations, or multinationals. The largest corporations, which include oil and gas companies such as Exxon Mobile, BP, Royal Dutch Shell, Chevron Texaco, Conoco Phillips and Eni, have annual sales that exceed the gross national income (GNI) of many nations. Therefore, they have been described as sovereign states that have a major impact on international politics and the world economy (Morgan, 2006). For that reason, oil and gas companies will try to avoid attention (i.e. pressure) from the society. Watts and Zimmerman (1978) argue that a way of decreasing or avoiding political pressure is by minimizing the income. This paper posits that the incentives of oil and gas managers to avoid attention by minimizing their earnings increase when political instability of a state increases.
This is consistent with the study of Roe and Siegel (2011), which found that political instability and weak democracy are fundamental setbacks for international organizations.
Vaugirard (2007) found that political instability causes the increase in financial instability.
Vaugirard (2007) noted that political instability is based on an agency problem between governments and society, and that an increasing transparency may suppress political instability. The first hypothesis is:
H1: Ceteris paribus: Political instability will increase managers’ incentives to downwardly manage their earnings.
2.3.2 Corruption
Access to mineral reserves is becoming a bigger concern for oil and gas companies.
Although these companies are very healthy in financial terms, they are finding difficulties to replace their reserves that would ensure future sustainability. This gives a government a more favorable position during the contract negotiations with the oil and gas companies (Ernst &
Young, 2011). It can also result in the abuse of power or resources by the elected government officials for personal gain (e.g. by extortion, soliciting or offering bribes, is known as political corruption (Chinhamo & Shumba, 2007). Many cases indicate that the oil and gas industry is subject to political corruption. For instance, on 27 January 2009, Shell’s regional executive Vice President Ann Pickard, met with the U.S. ambassador in Nigeria. During the meeting, she mentioned that corruption in the Nigerian oil sector was worsening by the day (Ten Albert, 2011). Numerous studies have confirmed this by finding that oil and gas themselves affect the degree of government’s corruption positively (Arezki & Brückner, 2011;
Bhattacharyya & Hodler, 2010).
One of the methods of interpreting the political corruption is by assuming that the governments (e.g. government’s officials) are looking for a certain manner to benefit from corporation’s wealth, and they will do everything to satisfy their personal desires. Politicians use bribes to convince the corporation to pursue the political objectives (Shleifer & Vishny, 1994). Additionally, Ray Vernon (1971) illustrated that once oil was discovered, bargaining power switches in favor of the government which then tries to increase its fiscal take by changing the terms of the original contract. Therefore, this paper assumes that corruption is a part of the political pressure. To avoid this political pressure, corporations will try to signal lower wealth, in order to keep a corrupt government at a distance. This is consistent with Watts and Zimmerman’s (1987) prediction. This paper assumes that the incentives of oil and
gas managers to avoid attention by minimizing their earnings, increase when government’s corruption increases.
This is consistent with the findings of Mathur and Singh (2007), which state that less corrupt states receive more foreign investment. Additionally, Shang Jin Wei (2000) found that a rise in either the tax rate on multinational companies, or the corruption level in a host country reduces inward foreign direct investment.
The second hypothesis is;
H2: Ceteris paribus, Political corruption will increase managers’ incentives to downwardly manage their earnings.
2.3.3 Level of democracy
Democracy is defined as the equality of all citizens before the law and when they have an equal and fair access to legislative process (Dahl et al., 2003). This includes granting the citizens with the power to vote. In a democratic form of government, the political power is decentralized. The Economist Intelligence Unit (2010) states that the opposite of democracies are authoritarian regimes, where the political power is centralized. In such a case the government will try to obtain all the power. In order to understand properly how the level of democracy (i.e. how the power is diffused within a government) influences the political pressure on companies, it is important to define the notion of power. A state possesses military, diplomatic, political and economic powers (Behrman, 1972). All of these refer to the ability to influence the course of an action in one way or another. A multinational corporation is most likely to hold economic and, as noted above, political power. The economic power is partly consistent with the goals of most of the multinational corporations, namely the profit and growth. An example of the political power is served in a case when the Dutch government has granted the political power to Royal Dutch Shell, in order to make the decision whether to remain in Syria, as being an attempt of the Dutch government to boycott the Syrian oil industry (Elsevier, 2011). This paper assumes that when the political power is more centralized, a government does everything to obtain more power by, e.g. sudden introduction of laws and regulations, that aim to restrict multinational corporations. In the case described above, the situation would be different if the Dutch government was authoritarian. As a result, the companies in authoritarian regimes will have more political pressure. Consistent with this assumption, researchers found a relation between foreign direct investment and the level of democracy. Harms and Ursprung (2001) concluded that investors are marginally more likely to invest in countries with a higher level of democracy. Jensen (2003) finds also a significant positive effect of democracy on foreign direct investment inflow. These studies illustrate that multinational corporations are less likely to invest in countries with authoritarian regimes because of higher country risks, e.g. high political pressure. In addition North (1990) and Olson (1993) contend that those in power shape policy to stay in power and amass wealth. In this view, powerful, centralized, closed governments constrain financial development to maintain power and capture wealth (La Porta et al., 2002; Shleifer & Vishny, 1994).
Contradictory to other multinational corporations, most oil and gas multinationals have in some circumstances no choice to avoid investing in countries with authoritarian regimes, as a result to a limited access to mineral reserves (Ernst & Young, 2011). Therefore, the political pressure caused by political regimes is unavoidable. In order to avoid this pressure, corporations will try to signal lower profit, aiming to keep an authoritarian regime government at a distance, by indicate that the company has little power. This is consistent with Watts and Zimmerman’s (1987) prediction. This might lead to incentives for managers
to avoid attention by minimizing their earnings when countries have a lower degree of democracy.
However, there is an inconsistency in this argument. O’Donnell (1978) suggests that foreign direct investors often share a cozy relationship with the political leaders, who protect foreign capital against the pressure for higher wages, stronger labor protection, or less capital- friendly taxation. In accordance with the study of Resnick and Li (2003), the democratic constraints over elected politicians tend to weaken the monopolistic positions of multinational corporations. Their empirical results suggest that by controlling the property rights protection, democratic institutions reduce foreign direct investment inflows. Mathur and Singh (2007) confirmed that theory in their research. Their results suggest that less corrupt countries and less democratic countries receive more foreign direct investment. In addition as noted above, organizations are the part of a broader social system wherein society, government and corporations are connected through social contracts (Deegan, 2002). Democracy gives society the power to vote and thereby, the indirect political power. Therefore, a democracy gives society more power to revoke the social contract when its needs are unsatisfied. As a result, corporations in democratic states are the subject to a high political pressure. In such a case, Watts and Zimmerman (1987) predict that managers will avoid this pressure by minimizing their earnings. Therefore, the third hypothesis will be neutral:
H3: Ceteris paribus, the level of democracy will increase managers’ incentives to managing reported earnings (upwardly or downwardly).
2.3.4 Left- or rightwing government
As noted above, a major difference between either left- or right-wing governments is that the right-wing governments favor less state control, and left-wing governments prefer more state control (Beck et al., 2001; Biais & Perotti, 2002). A large body of literature used this institutional factor to illustrate the influence that politics have on business environment.
For instance, privatization is more common in right-wing governments, rather than in left- wing governments (Bortolotti et al., 2001; Bortolotti et al., 2003; Boubakri et al., 2011).
Other sources provided evidence, that oil and gas itself (e.g. quantity or rents) increase the tendency to limit the operations of private international oil companies and assert a greater national control over the resource development (Stevens, 2008). This tendency is also known as the resource nationalism, which is not merely a phenomenon associated with ‘dodgy’
governments in ‘dark continents’. Canada and Australia are often quoted in the literature as classic examples of countries where the resource nationalism has dominated their politics (Owen, 1988; Uslaner, 1991).
As a result, the resource rich countries with left-winged governments favor more control over the economic environment. Politicians will use their power to transfers wealth for their own favor (Stigler, 1971). According to Watts and Zimmerman (1978), managers will be aware of this and will attempt to avoid the attention of politicians by choosing accounting policies to reduce earnings. The fourth hypothesis is;
H4: Ceteris paribus, left-wing governments will increase managers’ incentives to downwardly manage their earnings.
Political factors
2.3.5 Correlation among the hypothesis
A large number of political literature contributes the relation among the political factors.
For instance, Bhattacharyya and Holder (2010) found, that both the level of corruption as the level of democracy are influenced by mineral resources. Arezki & Brückner (2011) found that corruption and political instability correlated when oil rents increase. Beside these direct relations, as discussed above and as illustrated in figure 1, the political factors correlate due the relation between the resource curse, Arabic Spring and resource nationalism. Figure no. 3 gives an impression on how the political factors correlate and influence earnings management.
The correlation among the political factors might influence the results, therefore it is necessary to determine how each political factor influence earnings management separately.
This will be outlined in the methodology and results section.
Figure 3 Conceptual multicollinearity examined in knowledge related studies
Political Instability Corruption
Level of democracy Left-wing
Earnings management
3 Methodology
3.1 Sample and data
The relation between political factors and abnormal accruals is examined using (1) accounting data from Standard and Poor’s Compustat North-America and Global databases, and (2) political data from The World Bank “The Worldwide Governance Indicators:
Methodology and Analytical Issues” (WGI) and “Database of Political Institutions: Changes and Variable Definitions” (PDI).
The date need to be collected over a long period of time in order to criticize the results from event-studies related to the political costs hypothesis, such as the one found in the research by Han and Wang (1998) that oil and gas companies decreased their earnings as a response to rising oil prices during the Persian Gulf war. The data is collected from Compustat, WGI and PDI from the period of twelve years; from 1998 till 2009 (the data for the Dechow & Dichev model as modified by McNichols is collected from the period of eleven years from 1998 till 2008).
The accounting data consists of firms having a two digit SIC code of 13 oil and gas extraction or 29 petroleum refining and related industries. The firms had to meet data availability requirements in order to be included in the analysis . The sample begins with all the firms with available financial statement and returns history data. Observations are deleted if Compustat reports missing values for sales, total assets, net income before extraordinary items, cash, current assets, current liabilities, or cash from operations (Hribar & Collins, 2002). Data for political instability, corruption and level of democracy hypothesis are gathered from the WGI database. This database covers the definitions of political instability, corruption and democracy with respectively ‘Political Stability and Absence of Violence/Terrorism4’, ‘Control of Corruption5’ and ‘Voice and Accountability6’. This database provides an estimation and a ranking of the political situation of a country. The economics literature as well as the political economy literature do not agree upon the best proxy for these variables. This is why the ranking data7 is used to perform the tests and the estimation data8 is used to perform a sensitivity test (Kaufmann et al., 2010). The EXECRLC item in the PDI database provides data for the left-wing hypothesis. It classifies the left as all these parties which are defined as communist, socialist, social democratic, or as left-wing (Keefer, 2010). Country observations are excluded when WGI and DPI do not reconcile. The final sample consists of 39 countries, 4.981 firm-years with merely negative abnormal accruals in case of the Jones model (5.970 firm-years with positive and negative abnormal accruals), 3.373 firm-years with only negative abnormal accruals for the Modified Jones model (5.970 firm-years with positive and negative abnormal accruals), 2.320 firm years with only negative abnormal accruals for the Dechow & Dichev model as modified by McNichols (4.759 firm-years with positive and negative abnormal accruals), with the complete data needed to estimate equation 5 till 9.
4 Definition ofPolitical Stability and Absence of Violence/Terrorism: “Reflects perceptions of the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means, including politically-motivated violence and terrorism” (Kaufmann et al., 2010).
5 Definition ofControl of Corruption: “Reflects perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as "capture" of the state by elites and private interests” (Kaufmann et al., 2010).
6 Definition of Voice and Accountability: “Reflects perceptions of the extent to which a country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media” (Kaufmann et al., 2010).
7 Percentile rank among all countries (ranges from 0 (lowest) to 100 (highest) rank)
8 Estimate of governance (ranges from approximately -2.5 (weak) to 2.5 (strong) governance performance)
