Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Trade Shocks and Investments in Corporate Social
Responsibility
Shantanu Banerjeea,∗, Swarnodeep Homroyb and Aurélie Slechtena
aLancaster University, United Kingdom, bUniversity of Groningen
Abstract
This paper investigates the role of stakeholder preference on corporate social respon-
sibility (CSR). We find that Indian firms increase CSR expenses when trade restrictions
are initiated against competing Chinese exports from countries with high stakeholder
preference for CSR but not when these shocks open up other export markets. Capital
investments and R&D of Indian firms increase due to the same shocks, irrespective
of their country of origin. Finally, CSR has a bigger impact on firm value when the
demand shocks originates from countries with higher CSR preference. Collectively, our
results provide empirical evidence of investment motives for CSR.
JEL Codes: D22, O12, M14, G32
Key Words: Corporate social responsibility, Trade, Anti-dumping, India.
Acknowledgement: We thank Janis Berzins, Sudipto Dasgupta, Shubhashis Gan-
gopadhyay, Mohamed Ghaly, Maitreesh Ghatak, Ian Gregory-Smith, Colin Green,
Swami Kalpathy, Maurizio Zanardi, Alminas Zaldokas, and Gijsbert Zwart for their
suggestions on previous versions of this paper. We are also grateful for comments
from the participants at Lancaster University Economics and Finance seminar series,
University of Groningen EEF Seminar series, Baltic Economic Conference, Emerging
1
Market Corporate Finance Conference, Financial Management Association-Asia Pa-
cific Conference, Royal Economic Society Annual Conference, EEA Women in Retreat
and Work, Pension and Employment Group Annual Conference.
* Corresponding author: [email protected]
1 Introduction
In this paper, we attempt to disentangle the investment motive of corporate social respon-
sibility (CSR) from the agency motives. Discretionary corporate spending on philanthropic
causes are substantial and increasing over time. Understanding the motivations of CSR is,
therefore, an important question. There are two competing views of corporate philanthropy.
First, the investment view argues that CSR enhances a firm’s reputation among customers,
workers and other legislators, and is therefore consistent with shareholder value maximiza-
tion (Brown, Helland and Smith, 2006; Besley and Ghatak, 2007). The benefits of reputation
through CSR engagements accrue in the long term (Brown, Helland, and Smith, 2006; Porter
and van der Linde, 1995).1 The alternative viewpoint is that CSR expenses reflect agency
costs arising out of managerial entrenchment (Masulis and Reza, 2015; Tirole, 2001). Man-
agers use this form of discretionary expenses to advance their personal preferences and image,
to the detriment of shareholder value.2
Against this backdrop, we aim to contribute to the debate on the motives of such discre-
tionary expenses. Like any other long term investments, CSR initiatives can reduce short-
term profits but such expenses can be consistent with the objective of long-term shareholder
wealth maximization. However, it is econometrically challenging to isolate the investment
motive due to possible reverse causality in financial performance and CSR expenses. There-
fore, our approach is based on exogenous changes in the determinants of corporate philan-
thropy. Specifically we investigate the firms’response to changes in stakeholder preference
for CSR, triggered by trade-shocks in the export market.
1The benefits of these philanthropic investments may include building brand loyality (Kitzmueller andShimshak, 2012), attracting and retaining high skilled employees (Turban and Greening, 1997), charginghigher rent for buildings (Eichholtz, Kok and Quigley, 2010), and insurance from regulatory monitoring(Baron, 2001; Maxwell, Lyon, and Hackett, 2000).
2We do not examine the causal effects of corporate philanthropy on profitability. Several studies examinethe effect of corporate philanthropy on firm performance (see Margolis, Elfenbein and Walsh (2009) for ameta-analysis of this literature), but it is hard to infer causal evidence of performance effect of corporatephilanthropy due to endogenous associations.
1
Stakeholders’preferences for ethical behaviour varies across countries, and firms in dif-
ferent countries are exposed to different levels of stakeholder preference for CSR (Arora and
Gangopadhyay, 1995. When firms export to foreign markets, they encounter preferences of
the foreign stakeholders (consumers, buyers of intermediate products, and governments). In
a recent study, Newman, Rand, Tarp and Trifkovic (2018) find that Vietnamese exporters to
the U.S. engage in more CSR activities than Vietnamese exporters to China.3 If stakeholders
in certain countries, like the US and the EU, have a higher preference for corporate phil-
anthropy, then firms exporting to these countries will have higher CSR expenses than firms
exporting to countries with lower stakeholder preference for CSR. When positive demand
shocks originate from countries with higher (lower) preference for corporate philanthropy,
exporting firms will increase (not change) their CSR expenditure.4
Our empirical strategy is based on the adjustment of CSR expenses by Indian firms
when faced with an exogenous shock to their export demand. In particular we are interested
to examine if the adjustment of CSR expenses depend on which export destinations the
demand shock originates from. Our empirical tests rely upon data from top five hundred
Indian listed firms for the period 2006—2013.5 Our use of Indian data is motivated by
two important considerations. First, Indian firms face asymmetric preference for corporate
philanthropy in the domestic and the export markets. Indian firms are exposed to relatively
higher stakeholder preference for CSR compared to the domestic market when they export to
destinations like the US and the European Union (EU). According to a report by Bain and
Company (2015), individual and institutional charitable donations form 0.6% of India’s GDP.
3This is related to the cross-country variation in individual philanthropy: India and China’s ranks 106and 144 in the World Giving Index, while US is ranked 2nd, UK 6th and the Netherlands 7th.
4This argument holds if CSR expenses are undertaken as investments. If agency motives drive CSRchoices, then a positive demand shock will increase managerial discretion and, hence, CSR expenses irre-spective of the origin of the shock.
5In 2013, India passed a mandatory CSR regulation whereby listed firms are obliged to donate 2% oftheir net profits to corporate philanthropy, subject to some threshold conditions from 2014. Our sampleperiod ends in 2013 to prevent possible confounding effects of these regulations. Nevertheless, we check therobustness of our results by limiting the sample period to earlier dates.
2
Comparative figures for the US, UK and China are 2.2%, 1.3%, and 0.1%, respectively.6 The
difference in the stakeholder preference for corporate philanthropy in the export markets
is central to the our empirical design. Second, inspite of the low domestic stakeholder
preference, CSR expenses by Indian firms are substantial: the average CSR expense of
Indian firms is about 3% of the net profit. To use a benchmark, the average R&D expense
of Indian firms in our sample is 2.93% of net profit.
We identify cases of antidumping initiations by all export destinations against Chinese
products. We do a product-level matching of Indian exporters to these countries. Anti-
dumping (AD) is a form of trade barrier whereby an importing country (e.g. the US)
can unilaterally impose anti-dumping duties on products exported by firms from another
country (e.g. China) based on evidence that these exporting firms are charging lower price
for their exports than for their domestic sales, and that the dumping practices are injuring the
interests of the domestic producers in the importing country. AD duties are aimed to increase
the price of imported products from the target country and weaken their competitiveness
relative to the domestic producers. However, this also increases the relative competitiveness
of other exporters of the same product.
AD is a commonly used countervailing measure. The number of countries using AD
doubled between 1980 and 2003, with Australia, Canada, the EU, New Zealand and the
US historically being the largest users of this form of trade barrier. Emerging economies
like Brazil, Argentina, and South Africa are also increasingly using AD (Vandenbussche and
Zanardi, 2010).
AD petitions are initiated by domestic producers of the goods and not export market
competitors. Therefore, AD initiations on Chinese exports are unrelated to actions of foreign
competing firms (e.g. Indian exporters). Thus AD initiations on Chinese products are ex-
6This cross-country variation in individual philanthropy is reflected in India and China’s rank of 106 and144 in the World Giving Index, while US is ranked 2nd, UK 6th and the Netherlands 7th.
3
ogenous demand shocks to competing Indian exporters and should not affect their corporate
giving directly. These trade barriers on Chinese products from the U.S. and the EU not
only make Indian exports more competitive, but also increases the exposure of the Indian
firms to customers with a higher preference for corporate philanthropy (Bown and Porto,
2010; Kitzmueller and Shimshak, 2012). Indian firms do not face an increased stakeholder
preference for CSR when AD on Chinese products are initiated from other large exporting
destinations (e.g. United Arab Emirates, South Africa, Brazil, Argentina, etc.)
In recent years, China experienced the largest number of antidumping duty initiation
from a wide cross-section of countries and Indian exporters benefitted significantly in the
majority of these events. The theoretical underpinning of our quasi-natural experiment is
the trade deflection in favour of Indian firms by an exogenous decrease in the competitiveness
of similar Chinese products in the export market. To cite an example, before 2004, Chinese
exports of polyethylene terephthalate (a thermoplastic polymer resin, commonly known as
PET) products was 150% that of India. Following an anti-dumping petition filed against
Chinese PET products by US manufacturers in 2004, Indian exports of PET products to
the US overtook those from China so that by 2008 Indian exports was thrice that of China.
This illustrates how anti-dumping can adversely (positively) affect exports for the affected
(competing) countries.
We find that Indian firms increase their CSR expenses when US and EU initiate anti-
dumping duties against competing Chinese products. We then investigate how charitable
donations of Indian firms change when anti-dumping measures are initiated on Chinese
products by India’s other large trading partners, excluding the US and the EU. For every
year, we identify the top five trading partners of India (other than the US, EU, and China)
and estimate the effect of AD on Chinese firms from these countries.7 We find no significant
increase in charitable donations of Indian firms in response to anti-dumping initiations from
7The following countries are the largest export markets for Indian firms over the sample period: UAE,Saudi-Arabia, Brazil, South Africa, Argentina, and Japan.
4
these countries on Chinese products.
Our research design explicitly takes account of the agency and legacy motives of CSR
by comparing between firms that are closely held through cross ownership within a business
group structure with standalone firms. Coexistence of firms with these two types of owner-
ship structures in India provides us with an ex-ante variation in motives for CSR. Khanna
and Palepu (2000), and Bertrand, Mehta, and Mullainathan (2002) suggest that charitable
donations of group affi liates can reflect agency costs and can be motivated by rent-seeking in
the form of private benefits for the controlling shareholders. On the other hand, corporate
philanthropy can lead to a reputational gain in the product market over a long time horizon
and the lower participation constraint might make such investment more viable for the busi-
ness group affi liates (Bénabou and Tirole, 2010). The CSR spending group affi liates can also
reflect better access to financing through the internal capital market (Khanna and Yafeh,
2007; Gopalan, Nanda, and Seru 2007). These studies indicate that business group firms
are expected to spend more on CSR but they do not provide a clear prediction on how CSR
strategy will differ across the two groups in response to shocks to stakeholder preferences.
We find that business group affi liates, on average, have higher levels of CSR compared
to unaffi liated firms but both types of firms changes their CSR in response to AD shocks in
similar ways. The extent to which firms adjust their CSR expenses depend on the access to
finance and planning horizon. We find that Indian firms’reaction to AD shocks are similar
controlling for these differences. The results also do not differ between firms producing
final goods and intermediate goods. This mitigates the concern that our results are driven
primarily by a subset of firms.
One concern is that the above results do not reflect CSR preferences of stakeholders in
the US and the EU, but is simply an artefact of the relative importance of the US and the EU
as export markets for Indian products. The share of exports of Indian producers to the US
and EU, and other main markets are reasonably balanced: Brazil is just as large as the EU,
5
whilst the United Arab Emirates is the largest export destination for India. Nevertheless,
to address this concern, we examine whether other investments of Indian firms vary by the
origin of the anti-dumping initiation on Chinese products. Autor et al (2017) show that
import competition has a significant impact on R&D expenditure and other investments. If
export-shocks from non-US, non-EU countries are relatively less important, we expect to see
no significant changes in other investments of Indian firms. Our results suggest that Indian
firms increase both capital expenditure, and R&D in response to AD on Chinese products,
regardless of the origin of the export shock. These results corroborate the investment motive
for corporate philanthropy. Additional R&D and capital investments are likely reactions
to meet additional demand but are uncorrelated with stakeholder preference for CSR. In
contrast, firms only increase CSR when they expect to benefit from stakeholder preference
for such expenses. These results reinforce the idea that CSR expenses reflect investment
motives.
Our paper adds to the literature by examining the motives of corporate philanthropy
using the interaction of domestic firms with foreign stakeholders. In a related paper, Masulis
and Reza (2015) focus on the association between corporate philanthropy and the private
preference of the CEO, and the channels through which corporate giving affects firm value.
They show that corporate giving to sources closely linked to the CEO is value-reducing
for the firm, which is consistent with the agency motive of corporate philanthropy. By
contrast, Ferrell, Liang and Renneboog (2016) find no evidence that CSR is a function of
a firm’s agency problem (measured with two ex-ante incentive mechanisms of corporate
governance). We employ a different approach in which we use an exogenous shock to the
exposure to foreign stakeholders and provide evidence in favor of the investment motives of
corporate philanthropy. Our results do not invalidate the agency motive hypothesis as we
cannot disentangle the motive for corporate giving in the absence of a demand shock but we
establish that such expenses respond to investment demand from the export market.
6
The remainder of the paper proceeds as follows: Section 2 describes the institutional
framework for corporate governance in India. Section 3 reviews the main drivers of corporate
philanthropy under the investment and agency motives and develops hypotheses. Section 4
presents the data and key variables used in our empirical analysis. Section 5 discusses the
main results and our robustness checks. Section 6 concludes.
2 Institutional background
The institutional framework for corporate governance in India dates back to 1875 with the
establishment of the Bombay Stock Exchange (BSE). The Companies Act of 1956 governed
the activities of listed firms in India, which was subsequently replaced by the Companies Act
of 2013. Since the liberalization of the economy in 1991, Indian firms have been increasingly
reliant on external sources of finance and the role of government has decreased. There has
been a shift away from the traditional interventionist approach and toward a more Anglo-
American style of governance. Similar in spirit to the 2002 Sarbanes-Oxley Act, the Securities
and Exchange Board of India (SEBI) in 2001 implemented Clause 49 for all firms listed in
the BSE 200 index and subsequently to all listed firms. Clause 49 lays down a range of
governance imperatives for listed firms, from board composition to independence of audit
committee, to enhanced disclosure norms.
The dispersed shareholding pattern, with a relatively small equity ownership of managers
such as in the US or the UK, is not the only form of corporate holding in India. It repre-
sents about 53% of the firms listed in BSE. About 11% of the firms are controlled by the
government (federal and state), and three of the top six Indian firms in 2014 is one of these
public sector firms. Finally, about 36% of the largest Indian firms are parts of diversified
family-owned business groups.
Business group affi liates, often controlled by a family through cross-holdings are common
7
in the emerging economies. The strategic choices and financial outcomes for family firms
are well documented (Shleifer and Vishny, 1997). A common characteristic of these business
groups is the presence and influence of promoters. The term is commonly used to mean
controlling stakeholder, which can be an individual or a family. Promoters collectively hold
about 54% of the shares in business group firms. Consequently, tunnelling of assets can be a
source of ineffi ciency and loss of profitability. Bertrand, Mehta, and Mullainathan (2002) find
evidence that controlling stakeholders in the business groups use their entrenched positions
to benefit at the expense of minority shareholders. However, Khanna and Palepu (2000) find
that affi liate firms of diversified business groups outperform stand-alone firms in the same
industry.
Although Indian business groups share some characteristics of the pyramidal structures
in Japanese keiretsu, several key differences make them unique. Similar to keiretsu, individ-
ual firms within an Indian business group are legally separate entities, they are primarily
responsible to their own shareholders, and their accounts are audited separately. However,
unlike in keiretsu, in which the affi liate firms are connected and coordinated through a
common group-specific bank, firms within an Indian business group are coordinated by in-
terlocked boards and by members of the promoter family, similar to the holding structure of
Korean chaebols (Khanna and Palepu, 2000). A typical Indian business group has dozens
of firms with complex cross-holdings. The complexity of cross-holdings makes it diffi cult to
accurately compute the cash flow rights of the promoters.
The combination of dispersed shareholding, such as in the US and the UK, and an insider-
dominated structure, such as in China and Japan within the same institutional framework
and macroeconomic structure, allow us to compare discretionary spending by firms with
different levels of entrenchments and access to internal capital markets. As shown in the next
section, we will exploit these differences to control for the ex-ante variations in the motives for
CSR. Moreover, market and non-market institutions in India are relatively stable, allowing
8
for results that are comparable with extant corporate social responsibility and corporate
governance literature, which is based predominantly on evidence from US and UK firms
(Sarkar and Sarkar, 2000). The presence of stand-alone firms with dispersed shareholding
and South Korean chaebol-type business group affi liates with complex cross-holdings within
the same regulatory and accounting framework allows us to overcome many shortcomings of
the cross-country comparisons of socially responsibility of firms. At the same time, India’s
financial system bears resemblance to many emerging markets (Gopalan and Gormley, 2013),
which makes our results comparable to the empirical evidence on corporate governance of
other emerging economies.
3 Conceptual framework and related literature
A commonly agreed view on CSR is that it constitutes a form of corporate expenditure.
Therefore an important question is why firms voluntarily incur that cost. Two broad motives
for corporate philanthropy are discussed in the literature. First, although being costly, CSR
could be part of a long-term profit-maximizing strategy. CSR may increase customer’s
willingness-to-pay for a product or create some brand loyalty (Baron, 2001; Besley and
Ghatak, 2007). By spending resources on CSR, firms then make an investment to build
their image and cater to the preferences of certain group of consumers which they expect will
increase sales and profits over time. This first motive, which we will refer to as the investment
motive, is consistent with shareholders’wealth maximization. The second motive of CSR
stems from manager’s personal preference and it is a cost of managerial entrenchment. In
this case, corporate giving is at the expense of shareholder wealth creation.
The implications in terms of financial performance are opposite, depending on which of
the two motives dominates. However, it is diffi cult to isolate the financial returns to CSR
because it is endogenous: high-profit firms can spend more on CSR than low-profit ones. In
9
this section, we first review the main drivers of CSR and how they differ under each motive.
We then explain how we use these differences and an exogenous demand shock to identify
the dominant motive for CSR.
3.1 CSR as an Investment
Under the investment motive, CSR is assumed to increase consumers’willingness-to-pay for a
product and profit-maximizing managers use CSR tactically as part of an overall advertising
strategy designed to promote the firm’s image (Baron, 2001; Besley and Ghatak, 2007). One
main driver of CSR is then consumers’preferences for CSR. Firms selling their product in
a country where consumers (and more generally) are more responsive to CSR should spend
more on corporate giving.
The variability in stakeholder preferences across countries has been attributed to differ-
ences in terms of purchasing power of final consumers (McWilliams and Siegel, 2001; Arora
and Gangopadhyay, 1995). As a result, increased engagement in foreign market where final
consumers have higher purchasing power should lead to more CSR activities from Indian
firms. At the same time, Newman et al (2018) argue that CSR practices are transmitted
through international trade, irrespective of the nature of the product. This implies that
even firms exporting intermediate goods may adjust their CSR practices following an export
shock. As argued by Manasakis et al. (2017), when CSR started becoming widespread,
its further encouragement along the supply chain became a central policy objective in both
the US and the EU. If the exporting firms are part of an integrated supply chain they may
have no choice but to adhere to higher standards in terms of CSR. Additionally, firms from
emerging economies, such as India, are likely to face liabilities of origin from other firms in
the US and the EU; that is negative perceptions about their ability to conduct their busi-
ness responsibly (Marano et al., 2016). Investing in CSR has been found to be an effective
strategy to overcome such liabilities.
10
Additionally, corporate philanthropy can be viewed as a long-term investment, trading
off current profitability with longer-term customer loyalty and corporate legacy. Such social
investments are likely to be facilitated by longer planning horizon, and lower constraints
to investment financing. Business group affi liation could constitute a good proxy for both
planning horizon and access to financing. Oh, Chang, and Martynov (2011) find that long-
term shareholders are more likely to drive corporate philanthropy in South Korean firms.
At the same time, investment financing can be relatively easier for group affi liates because
of the access to internal capital markets (Masulis, et al. 2011; Gopalan, et al. 2007).
Therefore, on balance, business group affi liates may be more willing and able to invest in
corporate philanthropy for long-term reputation building. Banerjee and Homroy (2018) also
show that Indian group affi liates emphasize longer term investments, while stand-alone firms
emphasize shorter-term effi ciency measures. Our empirical strategy exploits these variations
in stakeholders’preferences across India’s main trading partners to provide evidence of the
investment motive for CSR. In particular, the origin of an export shock is likely to be an
important predictor of the firms’response to these shocks. A shock originating from the
EU or the US opens up investment opportunity for firms to cater to higher stakeholders’
preference for corporate philanthropy in the EU and the US. CSR expenses of firms should
then increase. However, if the demand shock originates from export markets with lower
stakeholders’ preference for corporate philanthropy, firms have little motives to increase
their expenses.
3.2 CSR as an Agency Cost
Previous studies have noted that corporate social contributions may fit into the argu-
ment that managerial discretion leads to expenses that reflect their personal preference
(Williamson, 1964; Jensen and Meckling, 1976; Navarro, 1988). In particular, CSR can be
viewed as a preferred expenditure designed to boost the manager’s prestige in the community
11
or, more altruistically, simply as a way to generate some warm-glow effect (Baron, 2008).
The level of these preferred expenditures depends on the managers’entrenchment, which
is typically related to the ownership structure (Jensen and Meckling, 1976). If corporate
philanthropy is a manifestation of the agency problem, then it is likely to be correlated with
firm ownership. The direction of the association between firm ownership and CSR is not un-
ambiguous. It will depend on how ownership structure affects a firm’s level of entrenchment.
To formalize this idea, we use the concept of discretionary profit (Navarro, 1988). Share-
holders define a minimum level of earnings necessary to prevent either a change in manage-
ment by shareholders or a corporate takeover by outside buyers in the market for corporate
control. Any surplus above this minimum threshold is discretionary profit that can be used
to finance managers’preferred expenditures, including CSR. An increase in this minimum
threshold means a lower level of managerial entrenchment and a lower ability of managers
to use CSR for their own private benefits.
Although of a different nature, managers of both types of firms present in India can have
agency motives for corporate philanthropy. Managers of widely-held firms have a lower stake
in the firm and can use charitable donations to further their personal preferences. On the
other hand, business group affi liates with concentrated ownership can have more discretion
in spending if the managers are themselves part of the family. Whilst group-affi liated firms
are likely to have lower agency problems arising out of separation of ownership and control,
family ownership can lead to controlling shareholders engaging in corporate philanthropy
which gives them private benefits at the expense of minority shareholders. In that view,
some papers highlight the greater agency problems among group-affi liated firms (Bertrand
et al., 2002). Managers of widely-held firms should then be characterized by a lower initial
level of entrenchment and lower CSR spending.
From the above discussion, ownership structure and consumers’ preferences for CSR
stand out as two important drivers. Under both motives, group-affi liated firms are expected
12
to spend more on CSR. Comparing CSR spending of group-affi liated and widely-held firms
won’t help us disentangle the motives for CSR. By contrast, changes in consumers’pref-
erences affect CSR spending only under the investment motive. Changes in consumers’
preferences within a country are diffi cult to observe and it can be affected by CSR strategies
of firms. However, as Indian firms also export to other countries, we can use cross-country
variations in consumers’preferences. Arora and Gangopadhyay (1995) show that customer
preference for corporate philanthropy is affected by purchasing power. Moreover, Newman
et al. (2018) find that stakeholders in China seem less concerned about corporate social
responsibility practices than stakeholders in the US or the EU. This seems consistent with
data on cross-country variation in corporate philanthropy: India and China’s ranks of 106
and 144 in the World Giving Index, while US is ranked 2nd, UK 6th and the Netherlands
7th.
Just comparing CSR spending of exporting and non-exporting firms is not suffi cient to
identify the motives for CSR, because exports, as financial performance, may be endogenous.
Our approach to disentangling the agency and investment motives of corporate philanthropy
is to examine the potentially asymmetric impacts of exogenous demand shocks originating
from export markets characterized by different stakeholders’preference for CSR: (1) shocks
originating from the EU and the US (high preference for CSR) and (2) shocks originating
from other main trading partners of Indian firms (low preference for CSR).
Under the investment motive, a shock originating from the EU or the US opens up export
opportunities for firms, but they need to cater to higher stakeholder preference for CSR in
these markets and such expenses of firms should then increase. However, if the demand
shock originates from export markets with lower stakeholder preference for CSR, firms do
not have an investment motive to increase their charitable expenses. Therefore, the origin
of the demand shock is likely to be an important predictor of the firms’strategic response
to these shocks.
13
The difference in terms of ownership structure between both types of firms will affect
the size of the impact of the shock on CSR spending. Given lower financial constraints
and easier access to internal capital markets, business group affi liates are likely to scale up
corporate philanthropy compared to unaffi liated widely held firms. This is our second hy-
pothesis. From an agency standpoint, a positive demand shock should increase profits and
so relaxes constraints on discretionary spending and hence both types of firms can increase
their philanthropic expenses, no matter what the stakeholder preference for corporate phil-
anthropy at the export market is. Moreover, given that managers from group-affi liated firms
are more entrenched, they should have a level of preferred expenditure relatively close to
their optimum. A positive demand shock, whatever its origin, could have a stronger impact
on CSR expenses of widely held firms, which have lower initial entrenchment and are then
ex ante more constrained regarding their preferred expenditures.
4 Data and Key Variables
In this section, we present the data and key variables that we will use to test our hypotheses
and investigate the agency and investment motives for corporate philanthropy.
4.1 Sample
A major challenge to research on corporate strategy in emerging economies is the availability
of reliable and consistent data. India has a mature capital market with internationally
comparable financial information and industry classifications. The main source of our data
is Prowess, maintained by the Centre for Monitoring the Indian Economy (CMIE). The
sample period is from 2006 to 2013. Although data on Indian firms are available before that,
the coverage and the consistency of the data are better 2006 onward. For example, Siegel and
14
Choudhury (2012) note that historical Prowess data had survivor bias, which is corrected for
in later years. Additionally, the Indian Companies Act of 2013 mandates that firms spend
a minimum of 2% of the average net profit made during the three immediately preceding
financial years on CSR. By limiting our sample period up to 2013, we minimize potential
confounding effects of the enforcement of this act from April 1, 2014.8 Our sample consists
of the top five hundred listed firms on the BSE, which represent over 95% of the total market
capitalization. We identify firms that have been listed at least once in BSE 500 within our
sample period, and include all available years for each of these firms. We exclude publicly
owned as they lend themselves poorly to comparison in our context.9 We also exclude firm-
year observations with missing data on ownership, as well as firm performance measures.
Our final sample is an unbalanced panel of 677 firms with 4,143 firm-year observations.
Table 1 presents the summary statistics on firm and board characteristics and philanthropic
expenses. All monetary values are winsorized at 1% levels and expressed terms of US$ as of
the year 2000. Description of key variables is presented in appendix A.
[Insert Table 1 around here]
4.2 Corporate Social Responsibility
We measure corporate social responsibility (CSR) aggregating the annual charitable dona-
tions, expenses in environmental causes and pollution control, and investments on social
infrastructure. Firms report expenses in corporate social responsibility initiatives to the
Securities and Exchange Board of India as part of their financial filings. Corporate social
expenses includes spending on charitable donations, the building, and maintenance of public
8It is possible that Indian firms anticipated this new regulation and, because of that, decided to increasetheir CSR spending before 2013. However, as the anti-dumping initiatives we use in our empirical analysis arespread over the all sample period (2006-2013), this should not qualitatively affect our results. Nevertheless,we check the robustness of our results by limiting our sample till 2012.
9For example, CEOs or Managing Directors of public sector firms are fixed term bureaucratic appoint-ments and the pay is contingent on tenure and rank.
15
services (parks, primary schools, etc.), and expenditures on environmental and pollution con-
trol.10 Our main dependent variable is the log of the linear addition of charitable donations,
environmental expenses, and expenditure on social and community infrastructure. CSR is
zero in 38% of the firm-year observations. The median CSR spending is $ 19,054, which is
about 3% of net profits. While the nominal values of CSR of our sample firms may seem
small relative to studies based on US firms, it is important to view it with respect to the
size of the firm, and other corporate investments. Mean CSR expenses of Indian firms are
of the same order of magnitude as the mean R&D expenses ($ 18,271) , and approximately
50% of that of mean capital expenditure ($ 44,108). In these terms, incurring CSR expenses
are of significant financial and strategic implications for Indian firms.
4.3 Anti-Dumping
Anti-Dumping initiations targeted at Chinese products is an exogenous demand shock to the
competing Indian firms. The data on AD is obtained from the World Bank’s Global Anti-
Dumping database (Bown, 2016), which provides information on all anti-dumping petitions
initiated by each country. Importing countries can choose to initiate an AD case against a
product, a firm, or all firms in a given industry from a particular exporting country. This
can initially lead to the imposition of a primary AD measure. Once dumping and damage
to the domestic industry are established, a final AD measure is imposed on the goods under
investigation. These measures are generally in effect for a three to five-year period. We use
AD initiations as a dummy, and not the imposed measures because the different types of
final anti-dumping measures are not directly comparable.11
The AD measures are designed to make the targeted goods less competitive in the im-
porting country. There is no plausible reason to expect the AD initiations on Chinese firms
10In unreported results, we check the robustness of our results to the individual components of CSR.11For example, AD penalties can be in the form of ad-valorem tax, minimum import price, quotas, etc.
The economic effects of these penalties are not straightforward to compare.
16
by an importing country to be affected by Indian firms. AD petitions are initiated by do-
mestic producers against targeted imports, and not by competitors in the export market.
Our empirical design works through the trade-deflection channel: there is some evidence
that trade barriers imposed against one exporter increase market access to other exporting
countries (Bown and Porto, 2010). Therefore such penalties, irrespective of their exact de-
sign, are likely to impose the intended adverse demand shock on the targeted exporters and
a positive and exogenous demand shock to competing firms from other countries.
We employ a product-level matching algorithm to identify Indian firms affected by anti-
dumping initiations on Chinese products. First, we collect information on anti-dumping
initiations against Chinese products by the United States, and the European Union from
2003-2013 (countries characterized by relatively high stakeholders’preference for CSR). To-
gether, 24% of Indian exports go to the US and EU. We match the Chinese products under
anti-dumping investigation to products sold by our sample of Indian firms.12 We then gen-
erate a binary measure, Anti − dumping − US/EUit−1 ,for Indian firms who export those
products to the US and the EU which are subject to anti-dumping investigation for Chinese
exporters. This gives us the ‘treatment group’of firms who are exposed to the positive de-
mand shock. Within the sample period, we have 73 events of anti-dumping initiations from
the US and the EU on Chinese products which are also exported by Indian firms to the US
and the EU. This gives us 722 firm-year observations of affected Indian firms. The value of
Anti− dumping−US/EUit−1 equals to 1 in only the years for which anti-dumping petition
was under investigation and the years when anti-dumping measures were in effect.
Next, we identify AD initiations on Chinese products from India’s other large export
partners (excluding US, EU, and China). To do that, we look at the top 5 export destinations
of Indian products for every year of the sample period. These countries are United Arab
Emirates (UAE), Brazil, Mexico, South Africa, and Argentina (countries characterized by
12The detailed matching algorithm and the dataset with the matched products are available upon request.
17
lower stakeholders’preference for CSR). Together these countries account for 17% of Indian
exports. We use this to construct a binary measure, Anti − dumping − Othersit−1 ,which
gives us the treatment group of firms who are exposed to a positive demand shock from
non-US/EU export destinations.13 Within the sample period, we have 41 events of anti-
dumping investigations from these countries on Chinese products. These give us 281 firm-
year observations on affected Indian firms. In figure 1, we present a map of India’s export
destinations and the stakeholder preference for CSR in those regions, and in table 2 we
present the volume of India’s exports and the number of anti-dumping initiations against
Chinese products from those countries.
[Insert Table 2 around here]
FIGURE 1 India’s Export Destinations and Foreign Stakeholder Preference for CSRIn this figure, darker shades of grey represent the large export destinations of Indian productsand the red borders show the countries with higher stakeholder preference for corporate phila-.thropy. In our empirical analysis, we attempt to compare the effects of export shocks fromcomparably large export markets but with different stakeholder preference for CP.
13There are 12 instances of Chinese products being simultaneously subjected to AD investigations byUS/EU and India’s other major export destination. We drop these observations to focus only on non-overlapping shocks.
18
4.4 Ownership Measures
We use a binary variable, Bu sin essGroup, to account for differences in ownership structures.
Business groups with a network of public firms, often controlled by a family, is a common
ownership structure in India. Group affi liates are likely to be characterized by concentrated
shareholding, cross-holdings of equity, greater access to an internal capital market, and
potentially higher control of the promoter family. This classification is consistent with the
measures used in the literature on emerging market finance (Khanna and Palepu, 2000;
Bertrand, Mehta, Mullainathan, 2002; Siegel and Choudhury, 2012).
Prowess provides information to accurately identify the shareholders who control a firm
either directly through their own shareholding, or indirectly through cross-holdings. We also
create a variable, %Shareholding − Promoters, which combines the direct shareholding by
promoters and the proportion of shares held by persons acting in concert with the controlling
shareholders. It is a measure of direct and indirect control of a firm by the promoters. Of the
677 firms in our sample, 267 (39.44%) are group affi liates and 410 (60.56%) are stand-alone
firms with dispersed shareholding.
4.5 Control Variables
We use the relevant accounting information from annual financial statements reported in
Prowess, cross-checked with information collected from Datastream using a string-matching
algorithm by firms’names. A firm’s profitability is measured by returns on assets (ROA),
and we control for firm size using natural log of sales.14 Information on board size and the
number of independent directors is collected from Prowess. Following the implementation of
the Clause 49 of SEBI in 2009, the mean proportion of independent directors on the board
14We also check the robustness of our estimates with alternative measures of firm size (total assets) andincluding another measure of firm performance (Tobin’s Q approximated by MTBV).
19
is expected to be at least 50% for all firms in our sample period. We also control for the
export intensity of a firm. Another strand of research suggests that institutional ownership
is positively associated with social responsibility (Shleifer and Vishny, 1997; Sethi, 2005;
Siegel and Vitaliano, 2007). Therefore, we also control for institutional shareholding by
the percentage of equity shares held by financial institutions such as mutual funds, banks,
insurance companies, and pension funds (%Shareholding − Institution).
In panel A of table 3, we compare the key variables for firms with different ownership
structure. Columns 1 and 2 present the mean values of key variables for group affi liates
and stand-alone firms, respectively and Column 3 reports the difference in means. The
mean CP for group affi liates is $19,180; for stand-alone firms, it is $8,675. The mean is
affected by the fact that about 30% of concentrated shareholding firm-years and about
40% of dispersed shareholding firm-years have no charitable donations. Group affi liates
are on average larger than widely held firms in terms of sales revenue and total assets.
However, no statistically significant difference in profitability and performance seems to exist
between business group firms and widely held private stand-alone firms. Also, no statistically
significant differences are evident in board-level characteristics. In panel B of table 3, we
compare the characteristics of firms for firm-years with and without philanthropic expenses.
Firms with charitable donations are, on average, bigger in terms of total assets, have a
higher proportion of exports to sales, and have a higher shareholding of both promoters and
institutional investors. No significant difference emerges in the size and the proportion of
independent directors on the board.
[Insert Table 3 around here]
20
5 Empirical Results
The central focus of our empirical analysis is to study the investment motives of CSR. In this
section, we consider the econometric issues and present the results of our baseline models
and robustness tests.
5.1 Univariate Analyses
5.1.1 Export orientation and corporate social responsibility
One important observation from the univariate analyses is that CSR is higher in export-
oriented firms. Since our exogenous shocks originate from the export markets of Indian
products, we investigate the relationship between the export orientation of firms and corpo-
rate philanthropy more closely using propensity score matching. To compare CSR of business
group affi liates and widely-held firms with similar characteristics, we match firms from the
two ownership categories on sales, ROA, MTBV for each industry and year. We use the
nearest neighbourhood, radius, kernel, and Mahalanobis distance matching methods.
[Insert Table 4 around here]
The results are presented in table 4 for subsamples of firms with high (above the 75th
percentile) and low exports (below the 25th percentile). In both the subsamples, the differ-
ence in CSR is positive and statistically significant at 5% level. However, the difference in
outcomes is much larger in the subsample of firms with low exports. The difference in CSR
of business group affi liates and firms with dispersed holding is mitigated by exports. We
have checked that the smaller difference in outcomes for the high-exports subsample stems
from higher CSR of widely-held firms with high export-orientation.
21
5.1.2 AD initiations on Chinese products and Indian exports
Our empirical design to identify the investment motive relies on the impact of export-driven
demand shocks on CSR expenses. For that design to be appropriate, we need to ensure that
additional necessary conditions are satisfied. First, anti-dumping as a trade barrier needs
to have a significant effect on exports. Second, anti-dumping initiations on Chinese prod-
ucts need to meaningfully affect the competing Indian exports. Vandenbussche and Zanardi
(2010) show that anti-dumping measures depress imports from the targeted country signifi-
cantly. Additionally, Bown and Porto (2010) show that Indian steel manufacturers benefited
in the forms of higher exports and profits when the US and the EU imposed safeguard trade
barriers on Chinese steel imports. They also show that Indian steel manufacturers expanded
existing capacity in response to this positive trade shock. Before systematically testing the
significance of the shock on the export in the treated group, we use two examples from our
data to highlight this point. PET-products exported by Chinese firms were brought under
anti-dumping investigations in 2004 from the US. As shown in figure 2a, Chinese exports of
PET-products were double that of Indian exports in the pre-2004 period. After 2004, Indian
exports of PET to the US overtakes Chinese exports and becomes twice that of Chinese
exports by 2008. Figure 2b shows a similar trade-deflection in favor of Indian products when
an anti-dumping petition was filed from Brazil against Chinese exports of steel-lined pipes
in 2011. These examples highlight the impact of anti-dumping on trade and the deflection
of trade to competing exporters.
22
Figure 2A AD on PET products from US on China and trade deflection to IndiaThis figure presents the time-series variation of Indian and Chinese exports ofPET products to US before and after anti-dumping initiations on Chinese exportsin the year 2004. Post-2004, Indian exports of PET to US overtakes that ofChina’s and becomes twice as large by 2008.
23
Figure 2B AD on Steel Products from Brazil on China and trade deflection to IndiaThis figure presents the time-series variation of Indian and Chinese exportsof steel-lined pipes to Brazil before and after anti-dumping initiations on Chineseexports in the year 2011. Post-2004, Indian exports of steel lined pipe products.overtake that of China
In table 5, we present a triple difference estimate of the effect of AD initiations on
competing Chinese products from the US and the EU on exports of group affi liates and
unaffi liated firms. We find that exports increase significantly in the post-AD periods for
both types of firms, but the difference in the increase is not statistically significant. To
quantify, exports (as a percentage of sales) of Indian firms increase by about 4% following
AD initiations on competing Chinese products by the US and the EU. A similar result
holds for anti-dumping initiations on competing Chinese products from other large export
geographies, where exports of Indian firms increase by about 5.6%. This suggests that the
anti-dumping shock has a statistically significant effect on exports through which it can affect
discretionary expenses of Indian firms.
[Insert Table 5 around here]
24
Third, AD needs to be among the preference set of trade barriers that the importing
countries choose from. The United States and the European Union account for about half
of the global anti-dumping petitions filed (Moore and Zanardi, 2009). Zanardi (2006) shows
that developing countries like Brazil and Argentina are increasingly using anti-dumping as
a trade barrier. Finally, Indian exports to the countries that initiate anti-dumping petitions
against Chinese products needs to be substantial. Exports to the US and EU comprise about
24% of India’s overall exports. The non-US, non-EU countries chosen for the analyses are
the other top 5 export destinations of Indian products and comprise 17% of overall exports.
5.1.3 AD initiation on Chinese products and CSR of Indian firms
Having satisfied these necessary conditions, we compare the CSR expenses of group affi liates
and dispersed shareholding firms, before and after they are affected by AD initiations on
Chinese products. In effect, this gives us the triple difference in mean CSR. In table 6,
we present the difference in 3-year averages of CSR for periods before and after periods
of AD. Unconditionally, there are statistically significant differences in CSR across the two
types of firms. In periods after AD, both types of firms increase social expenses by similar
amounts (these increases are also statistically significant at 1%-level), so that the difference-
in-difference is statistically not significant at conventional levels.
To quantify, the CSR-to-profit ratio of Indian firms go up by 5% (on a baseline of 3%)
following anti-dumping investigations on competing Chinese products by the US and the
EU. In times of increased demand, managers of all firms increase CSR expenses, which we
interpret as preliminary evidence for the investment motives of CSR. We examine these
competing hypotheses using multivariate analyses in the following section.
[Insert Table 6 around here]
25
5.2 Multivariate Analyses
5.2.1 Anti-dumping and Corporate Philanthropy
In this section, we extend the univariate analyses by controlling for a range of factors, and
test for alternative explanations. We regress natural log of CSR expenses on an indicator for
business group affi liates and controls for firm and board characteristics (size, performance,
board size, the proportion of independent directors, etc.) in column 1 of table 7. All results
are for the pooled sample, with the year and industry controls, and robust standard errors
are clustered at the firm levels. The Business Group indicator has a positive and statistically
significant coeffi cient indicates that the group affi liates have higher CSR than unaffi liated
firms. This can be driven by both higher entrenchment through control, and by investment
motives with an aim to build a long-term reputation, financed through an internal capital
market.
Next, we attempt to identify the investment motive using export shocks to Indian firms
originating from anti-dumping initiations against Chinese products from the US and the
EU. These results are presented in column 2, where the main variables of interest are the
indicatorAntiDumping−US/EUit−1 and the interaction term ofAntiDumping−US/EUit−1
with the indicator for Business Group. Anti-dumping shocks have a positive and statistically
significant effect on CSR, and the interaction term is also positive and statistically significant
at 10% level. Therefore, it seems that Indian firms, in general, spend more in charitable
donations when AD on Chinese products increase export market access. Both group affi liates
and unaffi liated firms seems to increase CSR expenses in times of favourable export demand
from the US and the EU. This result provides further support for the investment motive of
CSR.
[Insert Table 7 around here]
26
As the next step to understand the motives of corporate behavior, we examine philan-
thropic expenses of business group affi liates and unaffi liated firms around events of anti-
dumping initiation on Chinese products by India’s other major export destinations. Stake-
holder preference for CSR in India’s other large export destinations is, on average, lower than
that of the US and the EU. Therefore, if the increase in CSR of Indian firms in reaction to
anti-dumping on Chinese products from US and EU were driven by the investment motives
of catering to stakeholder preferences of the export market, such motives will be weaker in
this test. We present this result in column 3. We find no statistically significant association
of Anti− dumping −Othersit−1 with CSR in the pooled sample.
If corporate philanthropy reflects agency cost, and managers increase consumption of
private benefits of corporate philanthropy when the firm’s investment financing is easier, we
will expect to see an increase in CSR expenses irrespective of the source of the AD initiations
on Chinese products. That all firms seem to systematically differ in their adjustment of
corporate philanthropy to the origin of the export-market shock indicates that such spending
is aimed to cater to the demands of the foreign stakeholders.
The results reported in columns 1-3 are cross-sectional estimates of firms that are affected
by AD initiations on Chinese products with respect to firms that are not. In columns 4 and 5,
we compare differences in CSR adjustments between the two groups of ownership structures
following AD shocks. These estimates compare within-firm spells before and after exposure
to the export shock. In column 4, we find that the interaction term, Anti − dumping −
US/EUit−1 ∗ BusinessGroup is positive and only marginally statistically significant. Firms
increase CSR expenses when they are exposed to positive export-demand shock from the US
and the EU but there is no significant change in philanthropic expenses when similar export
shocks originate from other large export destinations.
27
5.2.2 Effect on Firm Value
How does the increase in philanthropic expenses in response to the demand shocks affect firm
value? We examine the value effects of firms that increase corporate philanthropy during
the positive demand shocks compared to firms that do not. In our firm-fixed effects models,
we compare the average market value of firms before and after periods of anti-dumping on
competing products. The results are presented in table 8. Unconditionally, higher corporate
philanthropy is correlated with higher firm value and the value of exporting firms increase
after the anti-dumping period. In addition, the interaction of CSRit−1 and Anti−dumping−
US/EUit−1 is positive and statistically significant. Therefore it seems that firms with higher
CSR have additional value gains from the demand shock. When we estimate a similar model
for the anti-dumping initiations from other export destinations, the parameter estimates and
the interaction of CSRit−1 ∗Anti− dumping−Othersit−1 are not statistically significant at
conventional levels. These results seem to reinforce the investment motives, insofar as the
value gain reflects the investor perception of the philanthropic expenses.
[Insert Table 8 around here]
5.3 Alternate Explanations and Robustness tests
In this section, we summarize a battery of tests to rule out alternative explanations for the
increase in CSR expenses of Indian firms following AD initiations on Chinese products from
the US and the EU but not when the AD initiation is from other large export markets.
5.3.1 Magnitude of the demand shocks
First, it is possible that the increase in corporate philanthropy in reaction to positive export-
shocks from the US and the EU is not related to stakeholder preference for CSR, but because
28
trade shocks from US and EU are significantly stronger than export shocks from other
markets. Therefore it is plausible that firms adjust CSR in reaction to the magnitude of
the shock rather than the nature of it. In addition, if India’s other major export partners
are not heavy users of AD, then the insignificant results for CSR will be an artefact of the
low power of these tests rather than any underlying economic reason. We approach this
issue by examining how firms adjust other investments that are likely to be affected by
positive demand-shocks. Bown and Porto (2010) show that Indian firms expand capacity in
reaction to trade safeguard against Chinese steel products. We estimate the effect on Indian
firms’capital expenditure (Capex), and research and development expenditure (R&D) in
reaction to anti-dumping initiations on Chinese exports by US and EU, and by other large
export-destinations of Indian products. The results are presented in table 9.
[Insert Table 9 around here]
In columns 1 and 2, we present the results for Capex and R&D of Indian firms when
Chinese exports are under anti-dumping investigation from the US and the EU, and in
columns 3 and 4 results when Chinese exports are under anti-dumping investigation from
India’s other large export destinations. We find that AD is associated with an increase in both
Capex and R&D, irrespective of the source of the AD initiation on Chinese products. Firms
expand capacity, and increase R&D in reaction to positive export-market shock, irrespective
of the source of the anti-dumping initiation but they increase CSR expenses only when the
shock originates in US or EU. This indicates a strategic investment motive rather than weak
shocks from the other major importers. Also, if the statistically insignificant association of
CSR with AD shocks from non-US, non-EU export destinations were due to the low power
of the test, we will expect that to affect the regressions with Capex and R&D as well.
29
5.3.2 Origin of the demand shocks
Second, it is possible that the effect of AD shocks on corporate philanthropy is driven by
a single country within the groups due to a spurious association. Therefore, we run our
basic specifications separately for the US and the EU, and Brazil (the heaviest anti-dumping
users in the other export destinations) and rest of the exporting destinations. The results
presented in table 10 show that our baseline results hold in all of these specifications. Of
particular interest is the comparison of the effects on the corporate philanthropy of anti-
dumping initiations from EU and Brazil which are similar in market size for Indian exports.
Whilst anti-dumping initiations from EU on Chinese exports increase philanthropic expenses
of Indian firms, such effects are absent for anti-dumping initiations on Chinese exports from
Brazil.
[Insert Table 10 around here]
We also examine effects on Capex and R&D of anti-dumping initiations on competing
Chinese products by different export destinations separately. The results are presented in
table 11 and show that Indian firms increase Capex and R&D in response to anti-dumping
initiations on Chinese products, irrespective of the source of the anti-dumping initiation. If
we compare similarly sized export markets for Indian products like the EU and Brazil, in
this case, we find similar increases in Capex and R&D of Indian firms. These results are
supportive of the investment motives of corporate philanthropy.
[Insert Table 11 around here]
30
5.3.3 Intermediate vs. Final Goods
It is plausible that differences in the product market brand image provide an alternate
explanation of our results. Firms selling consumer goods, and more visible brands may have
higher corporate philanthropy (Servaes and Tamayo, 2013) compared to firms producing
intermediate goods. If the product category and the ownership structure are correlated,
then our baseline results may be an artefact of the industry classifications. In our baseline
specifications, we use a set of industry dummies to control for this possibility. Further, we test
the difference in means of CSR for firms producing consumer goods, and firms producing
intermediate goods based on the main product category of the firm recorded in Prowess.
This difference is not statistically significant at conventional levels. This is consistent with
anecdotal evidence suggesting that CSR of Indian firms producing intermediate goods have
historically not been insignificant (e.g. Tata Steel, Reliance Petrochemicals, etc.). We present
these results in table 12.
[Insert Table 12 around here]
5.3.4 Additional robustness checks
We finally run a battery of robustness tests. To begin with, we examine a subsample of
cases where AD initiation on Chinese products has led to final anti-dumping duties being
imposed. Only 66% of the AD initiations affecting our sample of firms led to a final AD
measure being imposed. Our main results (reported in appendix B) hold when we restrict
our treated group to these cases only.
Next, we check for mean-reversion in CSR after anti-dumping duties are revoked. We
do this by creating an alternative indicator for anti-dumping which equals to 1 for all years
after an anti-dumping petition was filed against a Chinese competitor of an Indian exporter.
31
If there is mean reversion in CSR, we will see a smaller positive, or indeed no significant,
effect of AD in this specification. Our estimates with this new indicator are similar to that
of our baseline results and do not suggest a reversion to mean, at least in the short term.
This result is consistent with Vandenbussche and Zanardi (2010) who show hysteresis effect
of anti-dumping measures on imports. These results are presented in appendices B and C.
In unreported results, we examine the possibility that institutional shareholding drives
corporate philanthropy (Smith, 1996; Shleifer and Vishny, 1997). From table 3, institutional
shareholding is higher in group affi liates compared to widely held firms. To attenuate this
concern, in our baseline specifications, we control for institutional shareholding. Further,
we partition the data for firms with high (greater than p75) and low (lesser than p25)
institutional ownership. Increases in CSR in response to demand shocks are not significantly
different between the two groups. We do a similar analysis for foreign shareholding, and our
main results persist.
Our baseline specifications contain industry dummies, but in alternate specifications, we
control for the industry competitiveness using the Herfindahl-Hirschman index (HHI). Our
main results remain unaltered. Finally, we test for the robustness of our results to estimation
techniques because not all firms in our sample engage in corporate philanthropy. Using a
censored dependent variable is likely to underestimate the parameter estimates. We check
for the robustness of our baseline results using Tobit regressions. The ordinary least squares
(OLS) estimates are likely to be biased only in the censored region. Therefore, we compare
the Tobit results with the OLS estimates for the subsample of firms with non-zero CP, and
our baseline results are not affected by the choice of estimation technique. These results are
omitted in the interest of brevity.
32
6 Conclusion
We examine the impact of foreign stakeholder preference on the CSR expenses of Indian
firms by exploiting trade deflections towards Indian exporters when AD investigations are
initiated against competing Chinese products. On average export (as a percentage of sales)
of affected Indian firms increase by about 4% following anti-dumping initiations on Chinese
products by the US and the EU, and by 5.6% following similar shocks from other major
export markets. Using firm level CSR data, we find that Indian firms increase philanthropic
expenses when the AD shocks originate from the US and the EU where average stakeholder
preference for CSR is relatively higher. In contrast, when the shocks originate from other
export destinations with lower stakeholder preference for corporate philanthropy, there is no
change in the CSR expenses of Indian firms. The effects of AD shocks on CSR expenses
are economically meaningful, and they are persistent. On average, Indian firms in our
sample increase corporate philanthropy by 5% (on the baseline average of 3% of net profits)
in response to anti-dumping investigations on competing Chinese products by the US and
the EU, and there is evidence of hysteresis in the increase of corporate philanthropy. The
effects are also similar across the ownership structure, i.e. Indian business group affi liates
vs. unaffi liated firms with dispersed shareholding, and similar for firms producing both final
goods and intermediate products.
It is important to highlight that our results do not establish any causal relationship of
corporate philanthropy with profitability, and only examines the strategic motives of CSR.
In addition, we do not have consistent information on the destination of the philanthropic
expenses of Indian firms within our sample period. The destination of these expenses can
be informative as it can shed light on the modes of CSR expenses in response to overseas
stakeholder preference. Following the regulation on mandatory corporate social expenses in
India, such data are likely to be increasingly available, although the regulation itself can
confound the strategic motives for CSR.
33
We also examine other investments of Indian firms in response to AD initiations on
Chinese products. Indian firms increase capital expenditure and R&D in response to AD
against Chinese products, irrespective of the origin of the shocks. Firms invest in capacity
expansion and R&D when there is a positive demand shock from any export destinations
but they increase philanthropic expenses only when such activities are likely to cater to
stakeholder preference. These results are important as they provide empirical evidence in
favor of the investment motives of CSR.
34
References
[1] Albuquerque, R.A., Y. Koskinen, and C. Zhang. (2017) Corporate Social Responsi-
bility and Firm Risk: Theory and Empirical Evidence. European Corporate Gover-
nance Institute (ECGI) - Finance Working Paper No. 359/2013. Available at SSRN:
https://ssrn.com/abstract=1961971 or http://dx.doi.org/10.2139/ssrn.1961971
[2] Ariely, D., A. Bracha, and S. Meier. 2009. Doing Good or DoingWell? Image Motivation
and Monetary Incentives in Behaving Prosocially. American Economic Review 99:544—
55.
[3] Arora, S., and Gangopadhyay, S. 1995. Toward a Theoretical Model of Voluntary Over-
compliance. Journal of Economic Behavior and Organization 28:289—309.
[4] Autor, D., Dorn, D., Hanson, G.H., Pisano, G. and Shu, P. 2016. Foreign competition
and domestic innovation: Evidence from US patents (No. w22879). National Bureau of
Economic Research.
[5] Banerjee, S. and S. Homroy. 2018. Managerial incentives and strategic choices of firms
with different ownership structures. Journal of Corporate Finance 48: 314-330.
[6] Baron, D. P. 2001. Private Politics, Corporate Social Responsibility, and Integrated
Strategy. Journal of Economics & Management Strategy 10:7—45.
[7] Baron, D. P. 2008. Managerial Contracting and Corporate Social Responsibility. Journal
of Public Economics 92:268—88.
[8] Benabou, R., and J. Tirole. 2010. Individual and Corporate Social Responsibility. Eco-
nomica 77:1—19.
[9] Bertrand, M., P. Mehta, and S. Mullainathan. 2002. Ferreting out Tunneling: An Ap-
plication to Indian Business Groups. The Quarterly Journal of Economics 117:121—48.
35
[10] Besley, T., and M. Ghatak. 2007. Retailing public goods: The economics of corporate
social responsibility. Journal of Public Economics 91 (9): 1645-1663.
[11] Bown, C.P. 2016 Global anti-dumping Database, The World Bank, June, available at
http://econ.worldbank.org/ttbd/gad/
[12] Bown, C.P., and G. Porto. 2010. Exporters in Developing Countries: Adjustment to
Foreign Market Access after a Trade Policy Shock," Papers 88, World Trade Institute.
[13] Brown, W. O., E. Helland and J. Smith. 2006. Corporate philanthropic practices. Jour-
nal of Corporate Finance 12 (5): 855-877.
[14] Eichholtz, P., N. Kok, J. M. Quigley, A. Guma, J. Gyourko, M. Kahn, A. Sullivan, M.
Turner, and C. Wolfram. 2494. Doing Well by Doing Good? Green Offi ce Buildings.
American Economic Review 100:2494—2511.
[15] Ferrell, A., H. Liang, and L. Renneboog. 2016. Socially Responsible Firms. Journal of
Financial Economics 122:585—606.
[16] Financial Times (2014) Fortune 500 companies spend more than $15bn on corporate
responsibility. October 12.
[17] Gopalan, R., and T. A. Gormley. 2013. Do Public Equity Markets Matter in Emerging
Economies? Evidence from India. Review of Finance 17:1571—1615.
[18] Gopalan, R., V. Nanda, and A. Seru. 2007. Affi liated Firms and Financial Support:
Evidence from Indian Business Groups. Journal of Financial Economics 86:759—95.
[19] Jensen, M.C., and W.H. Meckling. 1976. Theory of the firm: Managerial behaviour,
agency costs and ownership structure. Journal of Financial Economics 3 (4): 305-360.
[20] Jensen, M. C. 1986. Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers.
American Economic Review 76:323—29.
36
[21] Khanna, T., and K. Palepu. 2000. Is Group Affi liation Profitable in Emerging Markets?
An Analysis of Diversified Indian Business Groups. The Journal of Finance 55:867—91.
[22] Khanna, T., and Y. Yafeh. 2007. Business Groups in Emerging Markets: Paragons or
Parasites? Journal of Economic Literature. 45 (2):331-372.
[23] Kitzmueller, M., and J. Shimshack. 2012. Economic Perspectives on Corporate Social
Responsibility. Journal of Economic Literature 50:51—84.
[24] Manasakis, C., Mitrokostas, E., & Petrakis, E. (2018). Strategic corporate social respon-
sibility by a multinational firm. Review of International Economics, 26(3), 709-720.
[25] Margolis, J., H. Elfenbein, and J. Walsh. 2009. Does It Pay to Be Good and Does It
Matter? A Meta-Analysis of the Relationship between Corporate Social and Financial
Performance. SSRN Electronic Journal :1—68.
[26] Masulis, R. W., P. K. Pham, and J. Zein. 2011. Family Business Groups around the
World: Financing Advantages, Control Motivations, and Organizational Choices. Re-
view of Financial Studies 24:3556—3600.
[27] Masulis, R. W., and S. W. Reza. 2015. Agency Problems of Corporate Philanthropy.
Review of Financial Studies 28:592—636.
[28] Maxwell, J. W., T. P. Lyon, and S. C. Hackett. 2000. Self-Regulation and Social Welfare:
The Political Economy of Corporate Environmentalism. Journal of Law and Economics
43:583—618.
[29] Moore, M. O., and M. Zanardi. 2009. Does Antidumping Use Contribute to Trade
Liberalization in Developing Countries? Canadian Journal of Economics 42:469—95.
[30] Navarro, P. 1988. Why do corporations give to Charity?. The Journal of Business. 61
(1): 65-93.
37
[31] Newman, C., J. Rand, F. Tarp, and N. Trifkovic. 2018. The Transmission of Socially Re-
sponsible Behaviour through International Trade. European Economic Review 101:250—
67.
[32] Oh, W. Y., Y. K. Chang, and A. Martynov. 2011. The Effect of Ownership Structure on
Corporate Social Responsibility: Empirical Evidence from Korea. Journal of Business
Ethics 104:283—97.
[33] Porta, R. La, F. Lopez-De-Silanes, and A. Shleifer. 1999. Corporate Ownership Around
the World. The Journal of Finance 54:471—517.
[34] Porter, M. E., and C. van der Linde. 1995. Toward a New Conception of the
Environment-Competitiveness Relationship. Journal of Economic Perspectives 9:97—
118.
[35] Sarkar, J., and S. Sarkar. 2000. Large Shareholder Activism in Corporate Governance in
Developing Countries: Evidence from India. International Review of Finance 1:161—94.
[36] Servaes, H., and A. Tamayo. 2013. The Impact of Corporate Social Responsibility on
Firm Value: The Role of Customer Awareness. Management Science 59:1045—61.
[37] Sethi, S. P. 2005. Investing in Socially Responsible Companies is a Must for Public
Pension Funds - Because There is no Better Alternative. Journal of Business Ethics 56:
99—129.
[38] Shleifer, A., and R. W. Vishny. 1997. A Survey of Corporate Governance. The Journal
of Finance 52:737—83.
[39] Siegel, D., and D. F. Vitaliano. 2007. An Empirical Analysis of the Strategic Use of
Corporate Social Responsibility. Journal of Economics & Management Strategy 16:773—
792.
38
[40] Siegel, J., and P. Choudhury. 2012. A Reexamination of Tunneling and Business Groups:
New Data and New Methods. Review of Financial Studies .
[41] Smith, M. P. 1996. Shareholder Activism by Institutional Investors: Evidence from
CalPERS. The Journal of Finance 51:227—52.
[42] Tirole, J. 2001. Corporate Governance. Econometrica 69:1—35.
[43] Turban, D. B., and D. W. Greening. 1997. Corporate social performance and orga-
nizational attractiveness to prospective employees. Academy of Management Journal
40:658—72.
[44] Vandenbussche, H., and M. Zanardi. 2010. The Chilling Trade Effects of Antidumping
Proliferation. European Economic Review 54:760—77.
[45] Williamson,O.E. 1964. The Economics of Discretionary Behaviour, Prentice-Hall, New
York.
[46] Zanardi, M. 2006. Antidumping: A Problem in International Trade. European Journal
of Political Economy 22:591—617.
39
Table 1
Descriptive Statistics of Key Variables
This table presents the summary statistics of the key variables used in the
empirical analysis. The sample period is 2006-2013 and all monetary values
are in ’000 US$ as of year 2000.
Variables N Mean p50 Std. Dev p10 p90
Panel A: CSR
Corporate Social Expenses 3,785 19.054 23.622 83.173 0 87.352
Panel B: Export and Antidumping variables
Anti-Dumping-US/EUit−1 4,143 0.161 0 0.367 0 1
Anti-Dumping-Othersit−1 4,143 0.023 0 0.415 0 1
Exports/Sales (%) 4,143 22.538 7.361 42.663 0.000 76.090
Panel C: Ownership Variables
%Shareholding-Promoters 4,143 41.577 51.122 20.847 25.186 74.235
%Shareholding-Institutions 4,143 17.806 15.199 14.496 6.000 37.113
Panel D: Board Variables
Board Size 4,143 9.949 9.000 3.328 6.000 33.000
% Independent Directors 4,143 51.799 50.000 16.181 16.181 92.487
Panel E: Financial Variables
ROA 4,143 0.083 0.077 0.111 0.0006 0.207
MTBV 4,143 1.419 2.360 2.542 0.006 5.278
R&D 3,762 18.271 12.867 19.670 0.258 69.901
CapEx 1,582 44.108 10.029 21.067 0.668 101.68
Sales (/1,000) 4,143 522.592 166.457 229.18 21.401 976.148
Total Assets (/1,000) 4,143 77.121 84.452 264.824 57.100 594.113
40
Table 2
Antidumping (AD) initiations by India’s major trading partners
In this table we present the number of AD initiations on China and India by
India’s large export destinations, and the mean of Indian exports to each
of these countries over the sample period 2006 - 2013.
Countries Indian Exports AD Initiations AD Initiations- AD Initiations-
(’000,00 US$) All China India
USA 286,325.30 445 116 28
EU 25,769.28 288 94 20
UAE 263,704.60 0 0 0
Argentina 4,012.81 218 71 11
Brazil 41,006.48 316 83 16
Mexico 12,302.43 100 48 03
South Africa 36,747.16 99 28 12
41
Table 3Univariate Comparison of MeansIn panel A, we compare performance, size, board characteristics, and CSRexpenses of business group affi liates and unaffi liated firms. The reportedmeans are calculated based on the subgroups of the full sample of 4,143firm years between 2006-2013. In panel B, we compare firms with andwithout CSR expenses. All variables are winsorized at 1% levels.***, **,and * denotes significance at 1%, 5% and 10% levels, respectively.
Panel AVariables Group Unaffi liated
Affi liates Standalones Difference
%Shareholding-Institutions 19.652 15.967 3.685**ROA 0.080 0.089 -0.009MTBV 1.532 1.501 0.031EPS 0.462 0.605 -0.143Sales (/1,000) 697.781 404.465 293.316***Total Assets (/1,000) 114.964 51.611 63.353**Export/Sales (%) 20.781 23.723 -2.942Board Size 10.137 9.195 0.222% Independent Directors 51.020 51.764 -0.744Corporate Philanthropy 19.180 8.675 10.505***
Panel BVariables Mean-No Mean- Difference
CSR CSRReturn on Assets 0.080 0.086 -0.006MTBV 1.553 1.780 -0.227EPS 0.513 0.558 -0.045Sales (/1,000) 667.104 650.722 16.372Total Assets (/1,000) 91.030 68.940 22.09**%Shareholding-Promoters 19.551 49.854 -30.303**%Shareholding-Institutions 16.705 18.518 -1.813**Exports/Sales (%) 17.660 21.924 -4.264**Board Size 9.548 10.210 -0.662% Independent Directors 50.656 50.892 -0.236
42
Table 4Propensity Score Matching ModelsWe match business group affi liates with stand-alone firms with dispersedshareholding using nearest neighbourhood (Panel A), radius =0.1 (Panel B)Gaussian kernel (Panel C), and Mahalanobis (Panel D) matching methods.The variables used in the matching are firm size, firm performance,market-to-book ratio for each industry and year. Firms are split into sub-samples of below and above median exports. ***,**, and * denote significan-ce at the 1%, 5%, and 10% levels, respectively.
Dependent Variable:Mean CSR as % of Net ProfitFirms with Firms withhigh exports low exports
Panel A: Nearest Neighborhood Match
Business Group - Standalone 3.42** 10.19**No. of Observations 2,129 2,017
Panel B: Radius Match (0.1)
Business Group - Standalone 15.27** 18.60**No. of Observations 2,110 2,003
Panel C: Kernel Matching
Business Group - Standalone 9.72** 15.71**No. of Observations 2,132 2,008
Panel D: Mahalanobis Distance Matching
Business Group - Standalone 9.20** 15.08*No. of Observations 1,962 1,998
43
Table 5
Effect of AD on Chinese products on Indian Exports (as % of Sales)
In this table we present univariate results of the differential impact of anti-dump
-ing penalties on exports. We compare the 3-year moving average of exports for
group affi liates and firms with dispersed shareholding, in periods before and after
anti-dumping initiation on competing Chinese products by the US and the EU.
***, **, and * denotes significance at 1%, 5% and 10% levels, respectively.
Anti-Dumping Firm Type Difference
Concentrated Dispersed
Before Anti-Dumping Yes 21.65 25.41 -3.76***
No 20.23 23.77 -3.54***
After Anti-Dumping Yes 25.98 28.57 -2.59**
No 21.67 24.81 -3.14***
After-Before Yes 4.33*** 3.16*** 1.17
No 1.44 1.04 0.40
Table 6
Effect of AD on Chinese exports on CSR (as % of Net Profits) of Indian firms
In this table we present univariate results of the differential impact of anti-dump
-ing penalties on CSR. We compare the 3-year moving average of CSR for
group affi liates and firms with dispersed shareholding, in periods before and after
anti-dumping initiation on competing Chinese products by the US and the EU.
***, **, and * denote significance at 1%, 5%, and 10% levels, respectively.
Anti-Dumping Firm Type Difference
Concentrated Dispersed
Before Anti-Dumping Yes 3.69 2.68 1.01***
No 3.55 2.58 0.97***
After Anti-Dumping Yes 3.86 2.81 1.05***
No 3.57 2.59 0.98***
After-Before Yes 0.17** 0.13** 0.04
No 0.02 0.01 0.01
44
Table 7Effect of Anti-dumping on Corporate Social ResponsibilityIn this table we present the results for the effect of anti-dumping initiations on Chineseexports on CSR of Indian firms who compete in the product market. The dependentvariable is Ln(CSR) in all columns. In column 1 we present the differences between busi-ness groups and unaffi liated firms, in columns 2 and 3 we add the effect of the ADinitiations on Chinese products from US/EU and other importing countries with the intera-ctions of the indicators for AD with the business group indicator, and in columns 4 and5 we report results with firm fixed effects.. Huber-White standard errors clustered atthe firm levels are in the brackets. ****, **, and * indicate significance at the 1%, 5%and 10%levels, respectively.
Dependent Variable: Ln(CSR)(1) (2) (3) (4) (5)
Anti-Dumping-US/EUit−1 0.692*** 0.937***(0.106) (0.082)
Anti-Dumping-Othersit−1 -0.0244 -0.074(0.133) (0.124)
Business Group 0.286*** 0.190*** 0.128**(0.040) (0.052) (0.060)
Anti-Dumping-US/EUit−1∗ 0.477* 0.386*Business Group (0.250) (0.197)Anti-Dumping-Othersit−1∗ 0.285 0.249Business Group (0.208) (0.216)Ln(Sales) 0.486*** 0.311*** 0.519*** 0.168*** 0.231***
(0.035) (0.025) (0.042) (0.0220) (0.030)ROA 0.390*** 0.296*** 0.472*** 0.477** 0.724**
(0.045) (0.011) (0.069) (0.178) (0.265)% Shareholding- 0.002** 0.002** 0.005** 0.005** 0.004**Promotersit (0.0010 (0.001) (0.002) (0.002) (0.002)Board Size 0.087*** 0.087*** 0.075*** 0.004 0.015
(0.013) (0.013) (0.010) (0.007) (0.012)% Independent Directors 0.001 0.001 -0.004 0.002* 0.005**
(0.003) (0.003) (0.005) (0.001) (0.002)Family CEO 0.024 0.024 0.118** 0.024 0.033
(0.037) (0.037) (0.047) (0.047) (0.084)%Shareholding-Institutions 0.017 0.017 0.019 0.007*** 0.009**
(0.013) (0.013) (0.016) (0.001) (0.004)Export / Sales (%) 0.002* 0.002* -0.004 0.003* 0.000
(0.001) (0.001) (0.004) (0.002) (0.000)Year Dummies Yes Yes Yes Yes YesIndustry Dummies Yes Yes Yes No NoFirm fixed effects No No No Yes YesConstant -3.085*** -3.417*** -5.99** -0.293** -0.843**
(0.330) (0.299) (2.41) (0.132) (0.034)Observations 3,762 3,762 3,762 3,762Adjusted-R2 0.388 0.314 0.307 0.245 0.211
45
Table 8Firm Value Effects of CSRIn this table, we present the results for the value effects of CSR expe-
nses after anti-dumping on competing Chinese firms. The dependent
variable is MTBV and we focus on the interactions of the AD indicators
with LnCSRit−1.The full set of control variables are used. Robust
standard errors clustered at firm level are in brackets.***,**, and *
indicate significance at the 1%, 5% and 10% levels, respectively.
Dependent Variable: MTBV
(1) (2)
Anti-Dumping-US/EUit−1 0.224***
(0.056)
Anti-Dumping-Othersit−1 0.109*
(0.058)
LnCPit−1 0.035** 0.023**
(0.016) (0.010)
Anti-Dumping-US/EUit−1∗ 0.143**
LnCPit−1 (0.062)
Anti-Dumping-Othersit−1∗ 0.092
LnCPit−1 (0.071)
Control Variables Yes Yes
Year Dummies Yes Yes
Industry Dummies Yes Yes
Constant 2.119*** 1.656***
(0.049) (0.036)
Observations 3,762 3,762
Adjusted-R2 0.328 0.250
46
Table 9Effect of Anti-Dumping on Other Corporate InvestmentsIn this table we present the results for the effect of anti-dumping initiations
on Chinese products on other firm investments viz. CapeEx and R&D. In
panel A we present the results for AD initiations from US and the EU, and
in panel B we present results for AD initiations from other major export
destination of Indian products. The full set of controls are used. Huber-White
standard errors clustered at firm level are in brackets.***,**, and * indicate
significance at the 1%, 5% and 10% level, respectively.
Dependent Variable Anti-Dumping-US/EU Anti-Dumping-Others
Panel A Panel B
CapEX R&D CapEX R&D
(1) (2) (3) (4)
Anti-Dumping-US/EUit−1 0.575*** 0.300**
(0.147) (0.132)
Anti-Dumping-Othersit−1 0.362** 0.219**
(0.156) (0.105)
Business Group 0.063** 0.008 0.060** 0.010
(0.030) (0.006) (0.029) (0.006)
Anti-Dumping-US/EUit−1∗ 0.014** 0.002*
Business Group (0.005) (0.001)
Anti-Dumping-Othersit−1∗ 0.012** 0.002*
Business Group (0.006) (0.001)
Control Variables Yes Yes Yes Yes
Year Dummies Yes Yes Yes Yes
Industry Dummies Yes Yes Yes Yes
Constant -3.413*** -3.049** -3.417*** -5.990**
(0.299) (0.433) (0.2999) (2.41)
Observations 3,762 1,582 3,762 1,582
Adjusted-R2 0.314 0.307 0.290 0.261
47
Table 10
Effect of AD on CSR for different stakeholder preferences
In this table we present the results for effect on CSR of anti-dumping initiations
on competing Chinese exports from different export destinations. In panel A
we present results separately for the US and the EU, and in panel B we present
the results for Brazil and the rest of India’s large export destinations separately.
The full set of controls are used. Huber-White standard errors clustered at the
firm level are in the brackets. ***, **, and * denotes significance at the 1%, 5%,
and 10% level, respectively.
Dependent Variable: Ln(CSR)
Panel A Panel B
US EU Brazil Rest
(1) (2) (3) (4)
Anti-Dumping-US/EUit−1 0.710*** 0.583***
(0.234) (0.209)
Anti-Dumping-Othersit−1 -0.038 0.022
(0.141) (0.098)
Business Group 0.163*** 0.211*** 0.120*** 0.125***
(0.069) (0.083) (0.041) (0.056)
Anti-Dumping-US/EUit−1∗ 0.484* 0.473*
Business Group (0.249) (0.246)
Anti-Dumping-Othersit−1∗ 0.277 0.289
Business Group (0.224) (0.230)
Control Variables Yes Yes Yes Yes
Year Dummies Yes Yes Yes Yes
Industry Dummies Yes Yes Yes Yes
Observations 3,762 3,762 3,762 3,762
Adjusted-R2 0.303 0.292 0.290 0.275
48
Table11-
EffectofADon
othercorporateinvestmentsfordifferentstakeholderpreferences
WepresenttheresultsforeffectonCPofanti-dumpinginitiationsonCapexandR&D.InpanelAwepresent
thebaselineresultsfortheUSandtheEUseparately,andinPanelBwepresenttheresultsforBrazilandIndia’s
otherlargeexportdestinations.Thedependentvariablesarementionedatthetopofeachcolumns.Huber-White
standarderrorsclusteredatfirm-levelareinthebrackets.***,**,and*denotessignificanceatthe1%,5%
and
10%levels,respectively.
PanelA
PanelB
US
EU
Brazil
Rest
DependentVariable
CapEx
R&D
CapEx
R&D
CapEx
R&D
CapEx
R&D
Anti-Dumping-US/EUit−1
0.512***
0.328**
0.591**
0.286***
(0.234)
(0.155)
(0.253)
(0.117)
Anti-Dumping-Others it−1
0.344**
0.200**
0.378**
0.239**
(0.156)
(0.099)
(0.146))
(0.111)
BusinessGroup
0.070**
0.062**
0.005
0.007
0.060**
0.009
0.012**
0.014
(0.031)
(0.023)
(0.004)
(0.006)
(0.021)
(0.006)
(0.006)
(0.008)
Anti-Dumping-US/EUit−1∗
0.018**
0.003*
0.013*
0.002*
BusinessGroup
(0.009)
(0.002)
(0.006)
(0.001)
Anti-Dumping-Others it−1∗
0.015**
0.001
0.011**
0.003**
BusinessGroup
(0.004)
(0.001)
(0.004)
(0.001)
ControlVariables
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
YearDummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
IndustryDummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Observations
3,762
3,762
3,762
3,762
3,762
3,762
3,762
3,762
49
Table 12Intermediate vs Final Goods
We present the results for the effect of anti-dumping on Chinese products by
the US and the EU (panel A) and by India’s other large trading partners (panel B)
stratified by intermediate and final goods. Huber-White robust standard errors
clustered at firm level are in brackets.***,**, and *indicate significance at the 1%,
5%, and 10% level, respetively.
Dependent Variable: Ln(CSR)
Panel A Panel B
Intermediate Goods Final Goods
(1) (2) (3) (4)
Anti-Dumping-US/EUit−1 0.382*** 0.593***
(0.107) (0.108)
Anti-Dumping-Othersit−1 -0.023 -0.033
(0.048) (0.025)
Business Group 0.187*** 0.125*** 0.192*** 0.138***
(0.059) (0.033) (0.051) (0.030)
Anti-Dumping-US/EUit−1∗ 0.390*** 0.501***
Business Group (0.129) (0.119)
Anti-Dumping-Othersit−1∗ 0.270 0.287
Business Group (0.211) (0.223)
Control Variables Yes Yes Yes Yes
Year Dummies Yes Yes Yes Yes
Industry Dummies Yes Yes Yes Yes
Observations 2,521 2,521 1,241 1,241
Adjusted-R2 0.364 0.340 0.239 0.223
50
AppendixA
DataAppendix.Thistablepresentsthedescriptionsofourkeyvariables.
Variables
Description
Anti-Dumping-US/EUit−1
Indicatorforananti-dumpingpetitionbeingfiledagainsta
ChineseproductbytheUSandtheEU
Anti-Dumping-Others it−1
Indicatorforananti-dumpingpetitionbeingfiledagainsta
ChineseproductbyIndia’sotherlargeexporters
BusinessGroup
Indicatorforbusinessgroupaffiliates
Corporatesocialexpenses
Spendingonbuildingandmaintenanceofpublicservices(parks
primaryschools,etc.),expendituresonenvironmentaland
pollutioncontrol-relatedissues,anddonationstoalocalauthority
oraninstitutionforasocialorhumanitariancause
Export/Sales(%)
Annualexports,scaledbyannualsales
%Shareholding-Promoter
%ofsharesoutstandingownedbyandassociatedwiththe
promoterfamily.
%Shareholding-Institutions
%ofsharesoutstandingownedbyinstitutionssuchasbanks,
insurancecompanies,hedgefunds,andmutualfunds.
BoardSize
Numberofdirectorsontheboard
%IndependentDirectors
Numberofdirectorsclassifiedasindependentnon-executive
directors,scaledbyboardsize.
R&DExpenditures
Researchanddevelopmentexpenses,scaledbytotalassets.
CapitalExpenditures
Annualcapitalexpenditures,scaledbytotalassets.
Ln(Sales)
Naturallogofannualsales
TotalAssets(/1,000,000)
Linearsumofcurrentandfixedassets.
ROA
Netincome,scaledbytotalassets.
MTBV
MarketCapitalisation+BookvalueofDebt
TotalAssets
EPS
NetIncome-Dividends
AverageOutstandingShares,scaledbyclosingprice
51
Appendix B. Subsample of AD petitions that led to duties
In this table we present results for the subsample of AD
initiations that led to a final AD duty. In column 1 we present
results for AD initiations on Chinese products from
the US and the EU, and in column 2 we present results
for AD initiations on Chinese products from India’s other
major export destinations. Robust standard errors clustered
at firm level are in brackets. ***, **, and * indicate significan-
-ce at 1%, 5% and 10% levels, respectively.
Dependent Variable
Ln(CSR)
(1) (2)
Anti-Dumping-US/EUit−1 0.602***
Final Duties (0.198)
Anti-Dumping-Othersit−1 -0.044
Final Duties (0.029)
Business Group 0.170*** 0.081***
(0.051) (0.030)
Anti-Dumping-US/EUit−1∗ 0.436***
Business Group (0.198)
Anti-Dumping-Othersit−1∗ 0.219
Business Group (0.184)
Control Variables Yes Yes
Year Dummies Yes Yes
Industry Dummies Yes Yes
Observations 3,762 3,762
Adjusted-R2 0.260 0.213
52
Appendix C. Test for mean reversion after AD is revoked
In this table we test for mean reversion in AD after the
AD petition has been resulted. We set the AD indicator=
1 for all periods after the AD initiation. Column 1 shows
the results for AD on Chinese products from the US and
the EU and column 2 for AD initiations on Chinese products
from India’s otherlarge trading partners. Robust standard
errors clustered at the firm level are in the brackets.
***, **, and * represent significance at 1%, 5% and 10%
levels, respectively.
Dependent Variable
Ln(CP)
(1) (2)
Anti-Dumping-US/EUit−1− 0.703***
All after Years (0.076)
Anti-Dumping-Othersit−1− -0.037
All after Years (0.058)
Business Group 0.173*** 0.120***
(0.057) (0.038)
Anti-Dumping-US/EUit−1∗ 0.452***
Business Group (0.222)
Anti-Dumping-Othersit−1∗ 0.265
Business Group (0.217)
Control Variables Yes Yes
Year Dummies Yes Yes
Industry Dummies Yes Yes
Observations 3,762 3,762
Adjusted-R2 0.328 0.306
53
54