CORPORATE SOCIAL RESPONSIBILITY AND TAX AVOIDANCE COMPARISON BETWEEN LISTED CONSTRUCTION FIRMS AND REAL ESTATE FIRMS IN VIETNAM45450

CORPORATE SOCIAL RESPONSIBILITY AND TAX AVOIDANCE COMPARISON BETWEEN LISTED CONSTRUCTION FIRMS AND REAL ESTATE FIRMS IN VIETNAM Thai Minh Hanh*, Dao Thanh Binh, Nguyen Thuc Huong Giang

CORPORATE SOCIAL RESPONSIBILITY AND TAX AVOIDANCE COMPARISON BETWEEN LISTED CONSTRUCTION FIRMS

AND REAL ESTATE FIRMS IN VIETNAM

Thai Minh Hanh*, Dao Thanh Binh, Nguyen Thuc Huong Giang

Hanoi University of Science and Technology, School of Economics and Management

ABSTRACT

The relationship between CSR and tax avoidance has been examined and has conflicting results Vietnam economy has been developed at high growth rate for the recent 10 years As an emerging country, construction and real estate industries are the industries which have impressive growth rate Therefore, it is interesting to investigate this relationship in Vietnamese listed firms in these two industries in Vietnam for the period of 2014 - 2017 For construction firms,

we found that environmental CSR firms have less tendency to tax avoidance Foreign ownership weakens the impact of CSR on tax avoidance while the moderating role of state ownership has insignificant result For real estate firms, there is no impact of CSR on tax avoidanc However, with moderating effect of ownership, foreign ownership helps to increase impact of CSR on reducing tax avoidance and state ownership decrease impact of CSR on reducing tax avoid-ance The results of the study help us understand the relationship between CSR and tax avoidance in construction and real estate industries in an emerg-ing country, which brings implications for firms and policy makers

1 NTRODUCTION

Corporate social responsibility (CSR) activities are increasingly drawing the attention

of investors, customers, suppliers, employees as well as governments across the world In this study, we re-examine the CSR and tax avoidance relationship compare the relationship between CSR and tax avoidance on construction and real estate firms They are the two important industries in Vietnam in recent years

In 2017, the number of construction firms was about 67.000 firms, account for more than 20% of the total number of manufacturing listed firms in Vietnamese stock exchanges Because of that, the competitiveness among these firms is quite high In

*Corresponding author.

Email address: hanh.thaiminh@hust.edu.vn

2016, construction products reached 1,089 trillion VND, equivalent to 25% of GDP According to FPTS, housing construction accounts for the largest share followed

by infrastructure, non-residential and industrial construction In the last 30 years, the construction industry has passed through six full acceleration and deceleration cycles, each from 4 to 5 years In recent years, the growth trend shows that the av-erage growth rate of the construction industry is lower The growth rate of 2019 is

at 7.23%, compared to the highest rate of 10.82% in 2015 and 10% in 2016 From

2017, this rate revolves around 8-8.5% The reason is that, the construction industry has passed the growth phase and comes to the restructure period (FPTS, 2018) Relating to real estate sector, according to research from OECD, this sector often regarded as “one of the largest employers in several countries with a significant size

in term of percentage of GDP” This sector also “has linkages with other sectors

of the economy and associated industries, which result in a multiplier effect in its capacity to generate income” (OECD, 2006) In Vietnam, real estate sector 10 years after the bubble crisis in 2008 has a sustainable growth in recent years In 2017, the market grew strongly, achieving impressive growth The real estate revenue increased

by 4.07% which is the highest since 2011, contributing 0.21% to the total increase

of GDP growth of 6.81% that year 2017 and 2018 also welcomed the increase of newly registered and the additional capital of FDI to the sector In the first 9 months

of 2018, real estate contributed to 30% of total FDI in Vietnam The amount of real estate inventory decreases year by year It was more than 70.000 billion dongs at the end of 2015, and reduces to about 30,000 billion dongs at the end of 2017 The development of the real estate sector continuing to attract cash flow into this sector, and many new big projects has been deployed (SBS, 2018)

With the high growth rate and high demand for construction and real estate products, therefore, studying CSR and tax avoidance in construction and real estate firms, comparing the impact of CSR to tax avoidance as well as impact of other factors such as corporate governance, ownership forms to this relation between these two industries should be necessary

We collect and analyze CSR data of a sample of Vietnamese stock exchange listed firms in construction and real estate industry for the period 2014 - 2017 Information

on CSR activities is hand collected from company annual reports Financial statement data are collected from the database of FiinPro The final sample consists of about 90 listed firms (344 firm-year observations) for construction sector and about 50 firms (191 observation) for real estate sector We perform pooled ordinary least squares regression analysis controlling for a variety of firm and year effects The results show

a positive impact of environmental corporate social responsibility activities on tax avoidance Investigating the influence of ownership characteristics, we find that foreign ownership investors weakens the impact of CSR on tax avoidance

One of the contributions comes from our finding that although a large number

of papers study the direct relationship between CSR and tax avoidance, relatively less attention has been paid to examining the interaction effects of ownership structure Theoretical and empirical results in CSR and tax avoidance relationship

in controversial, which emphasizes the need of studying factors that mediate or moderate this relationship

The rest of the paper is organized as follows Section 1 provides an overview of institutional background of Vietnam Section 2 briefly reviews the main theories explaining the relationship between CSR and tax avoidance and presents the hypotheses of the study Section 3 describes the research methodology and the variables A description of the sample and data collection process is presented in Section 4 The empirical results of the study are presented in Section 5 The last section summarizes the main findings and concludes the paper.

2 INSTITUTIONAL BACKGROUND OF VIETNAM

2.1 Corporate social responsibility in Vietnam

Political system and regulation, economic development and culture are important institutional factors that influence CSR practices in Vietnam The country has been gradually transitioning from a centrally planned economy with communist regime towards a market economy while maintaining a single-party government The intervention of the government to the economy is still significant The focus

on the socialist orientation of the economy implies the importance of CSR on government’s goals There are regulatory reforms on several dimensions of CSR for better environment legislation2 and employee rights, that give a positive signal for CSR practices However, up to now, there is a lack of a coherent public CSR policy The major public actor with respect to CSR seems to be Vietnam Chamber

of Commerce (VCCI), which depends on external funding for CSR projects Public initiatives on CSR mainly come from organizations such as UN Global Compact or UNIDO, which are not stable when the project funding finishes CSR regulation has positive progress but with weak implementation capacity Bilowol and Doan (2015) argue that firms have possibilities to engage in socially irresponsible activities due

to ineffective law implementation and high corruption level.3 A 2012 survey by the Vietnam Academy of Social Sciences investigating socially responsible practices revealed that many local companies had not adhered to the minimum standards

of CSR.4 Business misconduct, lack of basic employee benefits and deliberately causing environmental damages were identified as the most common reasons CSR practices, therefore, have both opportunities and challenges

As an emerging economy, Vietnam is viewed as one of the most promising countries

in East Asia with GDP growth rate among the fastest in the world.5 In 2009, Vietnam crossed the GDP threshold to be listed as a Low Middle Income country by the World Bank Economic improvement, better living standards, increased globalization and the pressure from exported-oriented sectors have pushed firms to adopt more CSR practices CSR activities are also enhanced by the establishment of stock markets,

2 National climate change strategy (2011), Green growth strategy (2012), Strategy for promoting renewable energy or other low carbon technologies.

3 During 2008-2013, Vietnam is ranked around 120 out of 175 countries in Global Corruption Report.

4 Ha Noi moi, 2012, Businesses indifferent to social responsibility Retrieved from (http://www.baomoi com/Doanh-nghiep-tho-o-voi-trach-nhiem-xa-hoi/47/8267986.epi).

5 According to 2016 World Bank report, following the implementation of the policy change to market economy, Vietnam’s GDP growth rate increased by an average of 5.5% per year since 1990, 6.4% per year in 2000s and 6.27% in 2015.

the rapid growth of which has led to a greater demand for transparency.6 Firms have now increased pressure to avoid irresponsible CSR and have more motivation for voluntary CSR disclosure Although voluntary CSR information disclosed by Viet-namese listed firms is still at a low level, the stock exchanges are trying to motivate listed firms for more transparency, especially to disclose CSR information.7

CSR practices are also affected by Vietnamese culture that mainly follows Confucianism Morality, sound social relationships, sincerity and justice are the pillars

of Confucian philosophy Confucian leadership is closely related to CSR because its purpose is to bring peace, learning and economic development for both organizational growth and communal well-being (Low and Ang, 2012) Beside Confucianism, firms

in Vietnam are also influenced by the western culture (e.g the United States8) where explicit appreciation of CSR practices is present (Matten and Moon, 2008) CSR engagements have therefore become important for Vietnamese firms and people These national cultural characteristics explain partly the attitudes of customers and managers towards CSR In a global consumer survey, Vietnam belongs to the highest group of Southeast Asian countries with 86% of respondents stating their willingness

to pay extra for products and services that come from companies committed to positive social and environmental impact.9Hieu (2011) finds that a large proportion of managers express a highly positive attitude towards CSR Thang and Fassin (2016) also observe that internal CSR which is related to employee working environment has

a positive and significant correlation with organizational commitment These cultural and historical characteristics also affect the way the board of directors take decisions

on CSR

2.1 Tax avoidance in Vietnam

The definition of tax avoidance can be distinguished with tax planning and tax evasion While tax avoidance is complying law in letter but not in spirit, the objective

of tax planning is to minimize the tax ability by complying law in letter and in spirit, and tax evasion is in another way the general terms for efforts by individuals and entities to evade taxes by illegal means or fraud (Agrawal, 2007) Tax avoidance is taken by taking advantage of loopholes in law and done before tax liability arises Another terminology, tax aggressive, is situated at the lower aggregation level of tax avoidance (Lietz, 2013) It refers to a further delimited scope of tax avoidance actions that are particularly “aggressive” Thus, non-aggressive tax avoidance and tax aggressiveness are often perceived to capture different scopes of explicit tax planning activities

6 Market capitalization increased from US$9.5 billions (9.6% of GDP) in 2008 to US$52 billions (26.8%)

in 2015.

7 The stock exchanges in Vietnam have organized Annual Report Awards since 2008 to motivate listed firms for further transparency Started from 2013, they introduced Sustainability Report Award, which referred to Global Reporting Initiatives 4 (GRI4) In 2015, there was regulation which requires listed firms to disclose CSR in their annual reports.

8 Vietnam has a war history with the USA but the relationship is better now The younger generation born

after the war are influenced by the US culture, education and entrepreneurship (Ralston et al., 2006).

9 AC Nielsen Global Survey of Corporate Social Responsibility and Sustainability, 2015, htttp://www nielsen.com/ sg/en/press-room/2015/sustainability-continues-to-gain-momentum-among-singaporeans- and-southeast-asian-consumers.html

Tax aggressiveness and tax evasion, however, are easy to detect than tax avoidance

In 2017-2018, in inspecting about 90.000 Vietnamese enterprises, the tax paid adjustment amount increased to 19.000 billion dongs, according to the Ministry of Finance of Vietnam

In Vietnam, both construction and real estate companies have to pay Corporate Income Tax, Value Added Tax, Personal Income Tax (on behalf of their employees)

as well as other small taxes Compare to other sectors, regulations concerning construction and real estate activities are quite complicate Real estate activities including industrial real estate development, industrial real estate services, office real estate development, residential real estate development and other real estate development and operation One real estate company can do several activities at the same time, such as land lease, infrastructure and housing construct for transfer, sale

or lease Construction has a value chain from the inputs (machines, materials and equipment) to the step of preparing the project, implementing, using and distributing their products Construction firms often have many employees, many construction sites, many material and equipment warehouses, many kinds of machinery with a very long construction schedule Regulations can be different relating to tax bases and time of tax payment There are also some costs occur before and during project’s implementation, which are quite high, however exceed the limits and will not be deducted as regulated The company has to pay these costs in reality, so it has to find another ways to compensate with these costs, through tax avoidance or tax fraud as an example

The OECD report has found that, for 18 countries that they surveyed, the real estate sector has been identified as an important sector being used to facilitate tax fraud Main behaviors used for violating tax payment such as manipulating the real estate price, hiding or not declaring income or transaction, using nominees and/

or false identities, trusts to hide the identity of the beneficial owner The method

of concealing ownership has three main variations: acquisition through off-shores companies or through a complex structure of ownership, unreported acquisition of properties overseas and use of nominees There was no official figures or statistics

on this problem, but one country like Austria has reported an amount of 70 million Euros of tax fraud related to real estate sector (OECD, 2006)

Of course, tax fraud and tax avoidance are two different definitions, of which tax fraud is illegal when tax avoidance means legally reducing the company’s taxable income However, the boundary between those two is very fragile

In Vietnam, “tax evasion” is an old terminology, not as the same case with the term

“tax avoidance”, which is quite new and has just been use more frequently in Vietnam after there are several foreign tax consulting firms came to Vietnam, at about year

of 1994 While tax evasion is illegal and can be penalized by administrative or criminal sanctions, tax avoidance is considered legal In tax avoidance, the enterprise itself or advised by the tax consultant (which are Big Four and almost other auditing companies) who understand well the tax regulation system and properly apply tax and accounting techniques to minimize the tax paid but not contrary to the law Relating

to corporate income tax (CIT), several particular techniques have been used for tax

avoidance such as (1) choosing the most suitable fixed assets’ depreciation method (register for applying straight line method, adjustable decreasing method, or method

of depreciation by volume, in considering the company’s profit expectation is high or low or the situation when the enterprise is during the period of CIT incentives or not); (2) transfer loses within the 5 years but choosing the fiscal year that the company expected to have substantial profit in order to reduce CIT payable; (3) using bank loan instead of increasing ownership’s equity for the business operations, because interest expenses will be deducted in total reasonable costs of the enterprise when calculating CIT payable (Phuoc, 2016) or (4) transfer the profit from one subsidiary

to another which situates in a country with lower CIT rate in the case of multinational firms With the development of tax consulting service, tax avoidance is becoming more popular However, as it is considered legal, the society accepts it and not condemn on it But in some respects, this influences to the social responsibility of the company

1.3 Corporate governance in Vietnam

Before 1986, Vietnam followed a centrally planned economic regime in which the only major driving force of the economy was the state sector The privatization process started in 1986 witnessed a reduction in the number of enterprises with 100% state ownership, from 5,655 in 2001 to 1,254 in 2013 Yet, the SOEs still contribute to about 30% of the state budget and account for almost one-third of the country’s GDP Nguyen and Dijk (2012) state that government interventions and political connections lead to a differentiation in terms of rights and responsibilities of SOEs and private firms This can cause a differential impact of CSR on the performance

of these two groups of firms

Following the start of economic reforms in 1986, the cancelation of embargo by the United States in 1994 and the country’s inclusion in the World Trade Organization

in 2007, foreign investment in Vietnam has been increasing over the years.10Foreign ownership has also increased, although for stock-exchange listed firms, it was limited to 49% in 2013.11 Multinational corporations, development agencies and other international organizations started to introduce and put pressure for more CSR (Bilowol and Doan, 2015) The increased role of foreign investment as well as foreign ownership and the associated stimulus on CSR practices can also create a differential impact of CSR on firm performance

2 HYPOTHESIS DEVELOPMENT

2.1 Effects of CSR on tax avoidance

The impact of CSR on tax avoidance can be explained through two view of business responsibility Friedman (1962) emphasizes the importance of shareholders’ wealth and profit maximization (e.g through a reduction in tax payments) and discounts the importance of corporate responsibility Therefore, firms are more likely to view

10 Foreign direct investment in 2013 was 5.2% of GDP (World Bank data)

11 Since the implementation of Decree 60/2015/ND-CP issued on 1/9/2015, foreign ownership in listed firms can be maximum 100%.

CSR as a strategic tool for impression management or marketing Because

of less concern on CSR, they have tendency to adopt tax aggressive tax reporting and tax avoidance This may lead to risk of discovery of aggressive tax reporting, which may damage the firm’s reputation and raise public concerns and media pressure (Hanlon and Slemrod, 2009; Wilson, 2009) According

to legitimacy theory, to cover up or divert attention from such negative effects, firm managers may use firms’ CSR activities to meet community and societal expectations (Davis et al., 2016; Gray et al., 1995; Yin and Zhang, 2012).According to Freeman (1984), firms emphasize not only the importance of shareholders but all stakeholders Such firms should take their CSR more seriously and be less likely to engage in aggressive tax avoidance They are expected to pay their fair share of taxes (Bird and Smucker, 2007; Hoi et al., 2013; Lanis and Richardson, 2012)

Empirical studies on the relationship between CSR and tax avoidance include Huseynov and Klamm (2012), Lanis and Richardson (2012), Laguir et al (2016) Lanis and Richardson (2012) found that higher CSR are likely less tax aggressive as according

to them, a company with a higher CSR profile (as measured by its engagement in CSR activities) is expected to be more cautious about undertaking tax aggressiveness activities because this would be inconsistent with its CSR engagements and may counteract the positive effects associated with its CSR activities Huseynov and Klamm (2012) show that the interactions between CSR categories and tax fees affect tax avoidance Firms with strong CSR can have strategy to lower their costs for the benefit of their shareholders, and profitable firms are in a better position to participate in charitable giving, and in some instances, it may be socially acceptable to reduce the tax expense and that is one manner of tax avoidance Laguir et al (2016) concluded that, corporations engage more in CSR activities pay their fair share of corporate taxes and are therefore less tax aggressive in nature According to their empirical test results, tax aggressiveness depends mainly on social and economic CSR dimensions, compared to governance and environmental CSR dimension which have non significant impacts Firms with higher level of CSR social dimension means engaging more in CSR activities of human resources, human rights in workplaces and community involvement were less likely to engage in tax aggressiveness However, firms with higher level of CSR economic dimension develops a culture that promises ethical conduct to stakeholders and this becomes decoupled from the organizational practices that are geared toward improving profits through tax planning activities, which leads to higher level of tax aggressiveness Besides it, to them, larger firms are less likely to engage in tax aggressiveness Larger corporations are likely to disclose their CSR information and tax information in the annual report

as well as having their own CSR report, compared to small corporations which can

be easy to hire their tax avoidance This conclusion however conflicts with some other previous researchs as of Zimmerman (1983), Gupta and Newberry (1997) Lin

et al (2017) found that in regions with lower institutional quality, firms claiming to be socially responsible actually avoid taxes, whereas CSR disclosure in other regions is more aligned with the social responsibility aspect of tax compliance This empirical result is totally fit with the original view of the authors, as they argue that located in

strong regions which has no favoritism and corruption, firms could have more trust

in the government and tend to place a higher value on meeting societal expectations

of responsibility In contrast, firms in weak regions have lower trust degree in the government leads to lower tax moral and they tend to hire their business activities, avoid tax payment and incline to put more weight on economic performance There have been no studies in Vietnam in the relationship between CSR and tax avoidance.Due to inclusive theoretical and empirical results on the relationship between CSR and tax avoidance, we state our first hypothesis as follows:

H1: Firms undertaking more CSR activities experience a lower (higher) level of tax avoidance.

We expect this hypothesis should be similar between construction and real estate firms However, we expect the impact of components of CSR on tax avoidance should be different

2.2 Ownership as a moderating mechanism

Corporate governance comprises a set of mechanisms through which one entity (for example, investors) can protect themselves against expropriation by another (for example, managers and controlling shareholders) Installing a good corporate governance framework helps to reduce diverse costs including agency conflicts

In this study, we focus on the few corporate governance mechanisms that can be important for Vietnamese firms These are share ownerships of foreigners and the government

Foreign ownership moderation

From the side of firms, they interest on foreign shareholders because of their financial resources and their experience such as their superior performance (Ghazali, 2010), higher productivity (Khawar, 2003), higher level of voluntary disclosure (Haniffa and Cooke, 2002) From the investors side, reasons such as access to new markets, opportunities to exploit resources, lower cost of labour, technological progress, policy liberalization, and most importantly, tax incentives, have served as motivating factors for them to go abroad

Huseynov and Klamm (2012) and Lanis and Richardson (2011, 2012) propose consideration of the tax authority as a company stakeholder We argue that tax authorities do act on behalf of the government, which also represents the general public Furthermore, taxes collected by tax authorities are meant for the welfare of citizens as a whole and not for the sole benefit of tax authorities Regarding the issue of organizational legitimacy, firms are expected to be socially responsible to legitimate their existence and maintain their survival within society Thus, they are required to pay taxes as an economic contribution towards the well-being of members

of the society where they operate (Williams, 2007b) Therefore, companies that are deemed to be socially responsible are expected to comply with the tax law, pay fair amounts of taxes and forgo the benefits associated with tax avoidance Coming from developed countries, foreign shareholders require higher CSR standards and more transparent information environment, which can lead to less tax avoidance

On the other hand, investors from other countries when coming to host countries want to exploit opportunities to maximize their profitability and firm value With the complicated legal and accounting framework among countries that firms deal with, they have more chance for tax avoidance In Vietnam, FDI enterprises occupy 45.9% total profit but tax payment amount is the lowest among all economic components (VEPR, 2018) In 2010, tax authority has inspected 575 FDI enterprises and 43 among them have the signal of transfer pricing and 37 have been penalized In 2014, among 870 FDI enterprises have been inspected, 720 have violated Therefore, foreign ownership may create environment for higher tax avoidance.

Due to conflicting theoretical arguments and empirical results on the moderating role of foreign ownership on the relationship between CSR and tax avoidance, we formulate our hypothesis as follows:

H2a: Foreign ownership strengthens the negative (positive) impact of CSR on tax avoidance.

State ownership moderation

SOEs’ equity is almost contributed by the state, and they also have to pay tax for the state Their activities have been audited by State Audit and inspected by State Inspectorate Those are the reasons that make them have less intention in avoiding tax, compare to foreign and private enterprises, of which the profits after tax will run directly to their pockets and their tax paid is only controlled by tax authorities SOEs also can easy take one part of their budget to invest in CSR activities than non-SOEs Empirical results found that corporate governance or ownership structure does not affect tax avoidance (Minnick andNoga, 2010) Others found that SOEs have less tax avoidance than foreign and private ownership (Bradshaw et al., 2016)

Hence, we formulate our as follows:

H2b: State ownership strengthens (weakens) the negative (positive) impact of CSR

(1)Where:

CSR it -1: Corporate social responsibility;

𝐸𝐸𝐸𝐸𝐸𝐸$% = 𝑎𝑎'+ 𝑎𝑎*𝐶𝐶𝐶𝐶𝐸𝐸$% + 𝑎𝑎-𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶$% + 𝜀𝜀$%,

Controls: Size, Leverage, Firm age, Pay-out Dividend, PB, ROA, Year;

εit: Firm-specific errors

The model includes firm-level control variables as well as year fixed effects because

it is very likely that different years have different levels of CSR activities

We test the hypotheses on the moderating effect of ownership factors by estimating the following OLS regression model (Matta and Beamish, 2008; Peng and Wang, 2014):

(2)Where:

CSR it -1: Corporate social responsibility;

OWN it-1: Ownership variables;

Controls: Size, Leverage, Firm age, Pay-out Dividend, PB, ROA, Year;

3.2 Variables

3.2.1 Dependent variable

The dependent variable used in this study is effective tax rate (ETR) ETR is measured

as GAAP ETR which is defined as tax expense divided by pre-tax income (Dyreng

et al., 2008, Minnick and Noga, 2010) We compute three-year average values for ETR because a longer term horizon differentiates tax management activities from isolated actions (Dyrenget al., 2008; Minnick and Noga, 2010) Tax expense includes deferred or accrued taxes, which are determined according to accounting rules, but may subject to earnings management (Hanlon and Heitzman, 2010)

3.2.2 Independent variables

Information on corporate social responsibility is collected from annual reports We assume that the more a firm undertakes CSR activities, the more it discloses CSR information (Clarkson et al., 2008) The main proxy measure of CSR is estimated

by means of content analysis It is a method of codifying written texts such as words, phrases, and sentences into various categories on the basis of selected criteria It assumes that frequency is an indication of the subject matter’s importance (Krip-pendorff, 2004; Neuendorf, 2002) Studies that employ content analysis measure CSR using standalone CSR reports (Dhaliwal et al., 2011, 2012), number of CSR pages in annual reports (Cowen et al., 1987), number of keywords in annual reports and CSR reports (Gamerschlag, 2011) and quantitative and qualitative disclosures (Wiseman, 1982, Xu et al., 2014) Content analysis is also used in other fields such

as risk management (Abraham and Cox, 2007, Linsley and Shrives, 2006) The advantage of content analysis is that it is an objective measure and can be used for replication However, it also has a drawback of using a subjective choice of words to measure CSR The disclosures only provide an indication of what firms say they are 𝐸𝐸𝐸𝐸𝐸𝐸$%= 𝑎𝑎'+ 𝑎𝑎*𝐶𝐶𝐶𝐶𝐸𝐸$% + 𝑎𝑎-𝑂𝑂𝑂𝑂𝑂𝑂$% + 𝑎𝑎1𝐶𝐶𝐶𝐶𝐸𝐸$% 𝑂𝑂𝑂𝑂𝑂𝑂$% +𝑎𝑎2𝐶𝐶𝐶𝐶𝐶𝐶𝑡𝑡𝑟𝑟𝐶𝐶𝑟𝑟 𝑣𝑣𝑎𝑎𝑟𝑟𝑣𝑣𝑎𝑎𝑏𝑏𝑟𝑟𝑙𝑙𝑙𝑙$% + 𝜀𝜀$%

doing, which may be very different from what they actually are doing

We use CSR keywords as unit of analysis These keywords are derived from the core indicators based on GRI framework For CSR keywords, we consider social and environmental dimensions (similar to GRI framework and MSCI index),but exclude corporate governance dimension because we study separately the effect of corporate governance In the empirical analysis, we consider both the single-dimensional CSR measures (social and environmental) and the multi-dimensional (aggregate) CSR measure

Although Vietnamese characters are in Latin system, they are not easy to process with software popular for English characters We use a self-construct software to count Vietnamese keywords In order to use this software, annual reports were converted from PDF or scanned format into Microsoft Word format by another software Following Campbell et al.(2014), we use the natural logarithm of the total number of CSR keywords in the regression analysis

3.2.4 Control variables

In investigating the effects of CSR on tax avoidance, we control for factors that systematically affect tax avoidance Dyrenget al (2008) find that small and high growth firms have higher ETRs We control for these effects with firm size variable, performance variable and book to market variable Firm size is measured by the natural logarithm of book value of total assets (Liu et al., 2015; Harjotoet al., 2015) Performance variables is measured by ROA, which is defined as return on total assets Book to market variable is proxied by the market value of equity divided by the book value of equity (PB)

We control for other firm specific variables, such as dividend payout and leverage Several studies (Chetty and Saez, 2005; Desai and Dharmapala, 2009; Morck, 2005) find that ownership and income tax structure affect a firm’s dividend policy

To control for the potential relationship between dividend policy and tax avoidance,

we use a dummy variable, which equal 1 if firm pays dividend and equal 0 otherwise Leverage is measured as the ratio of total debt to total assets (Barnett and Salomon, 2012; Liu et al., 2015; Harjotoet al., 2015)

4 SAMPLE

The data used in this study are collected from several sources Our initial sample consists of construction firms listed on the two main stock exchanges in Vietnam

(Ho Chi Minh Stock Exchange, HOSE and Hanoi Stock Exchange, HNX).12 HOSE

is the bigger stock exchange with stricter requirements for listed firms, such as registered capital (HOSE: at least 120 billions VND, HNX: at least 30 billions VND), profit before IPO (HOSE: at least two profitable years before IPO, HNX: cumulative profitable before IPO) and stricter disclosure requirements For CSR data, we collect information from annual reports downloaded from the websites of stock exchanges, individual companies and security firms Most of the annual reports are written in Vietnamese; only large firms provide English version

We collect firm-specific data from FiinPro - a large database provider in Vietnam Our sample consists of all Vietnamese construction firms listed during the years 2014-

2017 We begin with firms found at the date of downloading data from FiinPro (August, 2018) We searched for annual reports to collect CSR and ownership information Ownership data is also collected from FiinPro We thus have an unbalanced sample

of 344 construction firm-year observations (82 in 2014, 83 in 2015, 88 in 2016, 91

in 2017) and an unbalance sample of 191 real estate firm-year observations (45 in

We observe that the total number of CSR keywords of the average construction firm during 2014-2017 is 51 (43), which is lower than that of real estate firms (82) This disclosure is much lower than those found in developed countries, such as Germany, which has total CSR score of 129 during 2006-2009 (Gamerschlag et al., 2011) In both countries, the number of environmental CSR keywords is lower than that of social CSR The average number of environmental CSR keywords is only 12 (construction firms) and 22 (real estate firms) in Vietnam compared to 59 in Germany, whereas the number of social CSR keywords is 39 (construction firms) and 57 (real estate firms) in Vietnam and 68 in Germany The number of environmental CSR keywords

is much lower in Vietnam than that of social disclosure Another interesting finding is that the average number of total CSR keywords has slowly increased from 38 for the period of 2008-2013 (Kabir and Thai, 2017) to 51 (construction firms) and 82 (real estate firms) for the period of 2014-2017 This increase is consistent with the global trend on CSR, as observed by Gamerschlag et al (2011)

With regard to ownership structure, we find that the average foreign ownership of Vietnamese listed firms is 0.06 (construction firms) and 0.13 (real estate firms) This value for is reported at 0.08 for 2007-2012 by Phung and Mishra (2015) and 0.12 for the period 2006-2012 of firms listed in Ho Chi Minh stock exchange by Vo

12 The total market capitalization of these two stock exchanges accounts for US$ 52 billions, approximately 26.8% of the total GDP (World Bank Report 2015)

(2015) Average government ownership is 0.22 (construction firms) and 0.10 (real estate firms), compared to 0.25 reported for the period of 2007-2012 by Phung and Mishra (2015)

The market to book value (PB) is 0.78 (construction firms) and 1.05 (real estate firms) on average The average dividend payout is 0.35 (construction firms) and 0.22 (real estate firms) in the sample The average firm in our sample has total assets of 981.2 billions VND (construction firms) and 3,297.65 billions VND (real estate firms) Analyzing firm performance measures, we find that the average ROA is 3% (both construction and real estate firms) The leverage, measured by total debt ratio, of the average Vietnamese construction firm is 64% (construction firms) and 48% (real estate firms)

TABLE 1 DESCRIPTIVE STATISTICS

Panel A Year distribution of construction and real estate firms

Panel B Descriptive statistics - Construction firms

Variable N Mean Sd Min p25 p50 p75 Max Tax avoidance

ETR 342 0.20 0.12 0.00 0.14 0.21 0.25 0.47 CSR variables

PB 344 0.78 0.53 0.14 0.42 0.64 0.95 2.2

SIZE (billion VND) 342 981.20 969.59 52.60 253.00 577.00 1,450.00 3,560.00 ROA 342 0.03 0.03 -0.02 0.01 0.02 0.05 0.11 LEV 342 0.64 0.18 0.23 0.55 0.68 0.77 0.88 Mean values of descriptive statistics are presented along with standard deviations and 25, 50, 75 percentiles ETR represents effective tax rate CSR, CSR_E, CSR_S represent the total number of keywords

of total, environmental and social CSR disclosed in annual reports for the period from 2014 to 2017 FOR and GOV are values of foreign and government ownership PB is market to book ratio; DIV is dividend payout and SIZE is the total assets reported in billion Vietnam dong (VND) ROA refers to return on total assets LEV refers leverage, which is measured by total debt divided by total capital All variables are winsorized

at the 5% level.

Panel B Descriptive statistics -Real estate firms

Variable N Mean Sd Min p25 p50 p75 Max Tax avoidance

ETR 188 0.17 0.12 -0.13 0.11 0.21 0.23 0.40 CSR variables

PB 191 1.05 0.76 0.15 0.41 0.76 1.59 2.69

SIZE (billion VND) 188 3,297.65 4308.86 253.00 623.00 1245.00 4,505.00 15,800.00 ROA 188 0.03 0.04 -0.03 0.01 0.03 0.05 0.11 LEV 188 0.48 0.18 0.13 0.35 0.50 0.64 0.75 Mean values of descriptive statistics are presented along with standard deviations and 25,

50, 75 percentiles ETR represents effective tax rate CSR, CSR_E, CSR_S represent the total number of keywords of total, environmental and social CSR disclosed in annual reports for the period from 2014 to 2017 FOR and GOV are values of foreign and government ownership

PB is market to book ratio; DIV is dividend payout and SIZE is the total assets reported in billion Vietnam dong (VND) ROA refers to return on total assets LEV refers leverage, which is measured by total debt divided by total capital All variables are winsorized at the 5% level.The correlation among variables is shown in Table 2 Among CSR variables, ETR is significantly positively correlated with environmental CSR (construction firms) while significantly positively correlated with total CSR and social CSR (real estate firms) Among the ownership variables, state ownership is positively correlated with ETR (construction firms) For control variables, dividend payout is positively correlated with ETR (construction firms) while all control variables are correlated with ETR (real estate firms) We test for the presence of multicollinearity by calculating variance inflation factor (VIF) The value is around 2, much lower than the threshold 10, indicating that multicollinearity does not appear to pose a problem in our dataset

Table 2 Correlation analysis

Panel A Correlation analysis - Construction firms

ETR represents effective tax rate CSR, CSR_E, CSR_S represent the natural logarithm of the total number of keywords of total, environmental and social CSR disclosed in annual reports for the period from 2014 to 2017 FOR and GOV are values of foreign and government ownership

PB is market to book ratio; DIV is dividend payout and SIZE is the natural logarithm of total assets ROA refers to return on total assets LEV refers leverage, which is measured by total debt divided by total capital All variables are winsorized at the 5% level.

*Statistical significance at the 5% level

Among control variables, market to book ratio and dividend payout are statistically significant in with the expected sign We also find that most of the regression models explain around 5% variations in ETR

Table 3 OLS regressions of ETR on C

(1) ETR(2) ETR(3) ETR