Firms and social responsibility: A review of ESG and CSR research in corporate finance

Author links open overlay panelStuart L.Gillan^aAndrewKoch^bLaura T.Starks からの、ESG投資市場急拡大に関する引用です。（以下）

Interest in ESG/CSR from the corporate perspective is indicated by the Governance & Accountability Institute which, in 2018, reported that 86% of S&P 500 firms released sustainability or corporate responsibility reports compared with just under 20% in 2011.² Investor interest in ESG/CSR is highlighted by the fact that in 2019 alone, 300 mutual funds with ESG mandates received a total of $20 billion in net flows, which was 4 times the 2018 total.Moreover, there are currently more than 3000 institutional investors and service providers that have signed onto the Principles of Responsible Investment (PRI), an agreement to incorporate ESG/CSR issues into their investment analysis and decision-making processes. Assets under management for these investors has increased from $6.5 trillion in 2006 to over $86 trillion in 2019.⁴

ESGとCSRの違いについては、ガバナンスを明示的に組み込んでいるかどうかの点を挙げています。

n. ESG is an acronym developed in a
2004 report by 20 financial institutions in response to a call from Kofi Anon, Secretary-General of the United Nations7 As it implies,
ESG refers to how corporations and investors integrate environmental, social and governance concerns into their business models. CSR
traditionally has referred to corporations’ activities with regard to being more socially responsible, to being a better corporate citizen.
One difference between the two terms is that ESG includes governance explicitly and CSR includes governance issues indirectly as they
relate to environmental and social considerations. Thus, ESG tends to be a more expansive terminology than CSR

また以下のように、ガバナンス構造が、環境や社会問題への企業の対応にどう影響するかも論じられています。

our focus on the environmental and social aspects of ESG/
CSR allow us to examine the interrelationships between a company’s governance structure and its environmental and social activities.

市場やその企業が位置する国の特性が、ESG/CSR行動に大きな影響を与えることも示されている。

In fact, Cai et al. (2016)
and Liang and Renneboog (2017b) provide evidence that country characteristics appear to be quite important in explaining firm’s
ESG/CSR activities.

Cai, Pan, and Statman show that variation across countries is associated more
strongly with country factors than firm characteristics. The authors also provide evidence that economic development, law, and culture
play a role in these differences.

結局、法律で定められているかが最も企業のCRS・ESG対応を説明するという先行研究も。

Liang and Renneboog (2017b) conclude that legal origin is the strongest predictor of firms’ ESG/CSR adoption and
performance, more-so than political institutions, regulations, social preferences and a firm’s own financial and operational
performance.

インダストリータイプの影響も大きい｡

Firms’ ESG/CSR practices also tend to have a strong industry component. Using six of the KLD categories (omitting the corporate
governance category), Borghesi et al. (2014) find some industries, such as consumer goods and computer hardware, have aboveaverage scores and other industries, such as aircraft and petroleum and natural gas, have below-average scores. The industry effect
is so significant that it is common for researchers to use ESG/CSR scores demeaned by industry, rather than the raw scores themselves.
In fact, while some rating providers score firms relative to the universe of firms that the data provider covers, others use industry
benchmarking when constructing measures of firms’ ESG/CSR practices.

興味深いのが、クリーンエネルギー産業は環境スコアを高め、ソーシャルスコアはそうでもないのに対し、伝統的エネルギー産業などは環境スコアではなくSのソーシャルスコアで高める傾向にあることだ。また財務状況が芳しくない企業はEもそう積極的になれないことも示されている。また規制強化やレポーティング要件の変化を経験した国際企業で働いたディレクターが役会に入ると、その経験・知識が移転されることも示されている。（そらそうだろうという気もするが。）

Iliev and Roth (2020)examine the influence of boards on ESG/CSR. Specifically, the authors study U.S. directors who serve on the
boards of international firms exposed to changes in environmental and social regulations and reporting requirements. The rationale is
that such directors are likely to transmit information about these changes to the U.S. boards on which they also serve, and the information transmission provides a basis for causal inference. The central finding is that the U.S. firms that have directors exposed to the
changes in regulations and reporting requirements experience an increase in ESG/CSR performance (using the MSCI KLD scores) of
some 4.4%. The authors also note that the changes are concentrated in environmental rather than social performance, and that
environmental improvements are primarily observed in so-called ‘clean’ industries. In contrast, the authors also document that firms in
‘dirty’ industries improve their social but not their environmental performance (which the authors interpret as evidence of greenwashing), and that firms that are financially weaker are

これは面白い発見なのだが女性がCEOだったりボードにいるとCSRスコアりが高くなるのはよくある論議だが、CEOが結婚していたり娘がいると、これもスコアの違いに反映されるという先行研究もある。

Other studies provide evidence that firms’ ESG/CSR activities are associated with variation in CEO and board characteristics. One
demographic variable that is commonly found to be significant is gender. For example, Borghesi et al. (2014) report that U.S. firms
with women as corporate leaders or board members have significantly higher ESG/CSR scores. In addition to finding that S&P 500
firms with female CEOs have more socially responsible corporate practices, Cronqvist and Yu (2017) examine the gender of the CEO’s
children. The authors propose that CEOs’ daughters may exhibit stronger “other-regarding” preferences than their sons, and that CEOs
with daughters may internalize the preferences of their children. The authors find that executives with a daughter are employed by
firms with ESG/CSR scores about 9.1% higher than the median firm’s rating, and the magnitude equals about one quarter of the effect
of the firm’s CEO being a woman. Hegde and Mishra (2019) conclude that firms with married CEOs have significantly higher aggregate

他にも面白いのが、CEOが若ければよりCSRに熱心になり、また自信満々のCEOは逆にCSRに熱心でないという結果だ。これは、CSR活動が”パフォーマンスのリスクヘッジ”として使われている可能性を示唆するという。

Two other significant CEO attributes that have been documented as being linked to firms’ ESG/CSR profiles: CEO age and CEO
confidence levels. Borghesi et al. (2014) conclude that younger CEOs are significantly more likely to lead firms with higher ESG/CSR
scores. McCarthy et al. (2017) report a negative relationship between CEO confidence and firm ESG/CSR performance. They suggest
that this result is attributable to firms’ ESG/CSR scores having a hedging component and overconfident CEOs being less likely to hedge

経営者や創業者、株主がリベラル政党支持であれば、CSR活動とスコアが高まる傾向も示されている。

Several studies consider the political stance of the corporate leaders, but arrive at different conclusions. Di Giuli and Kostovetsky
(2014) provide evidence that U.S. firms with CEOs, directors, and founders that make larger donations to Democratic (rather than
Republican) candidates tend to have higher ESG/CSR scores. The authors also hypothesize that the political views of the firm’s
stakeholders (employees, suppliers, shareholders, customers, and regulators) can affect firms’ ESG/CSR choices. Using geographic
clustering of political views as sources of exogenous variation in the firm’s internal and external political environ

ただし同じ研究で、違う結果が出ており、その結論は定まっていない。

Using the same source of variation as Di Giuli and Kostovetsky (2014), Borghesi et al. (2014) offer an opposing view. They conclude
that, after controlling for CEOs who donate to both Democrats and Republicans, the relation between CEOs who donate to Democrats
and their firms’ ESG/CSR scores is insignificant. It is unclear what is causing the differences in results. They could be driven by the
presence of non-CEOs in the Di Giuli and Kostovetsky sample (which includes directors and founders) since the Borghesi, Houston, and
Naranjo sample is limited to CEOs; they could be driven by the differences in sample periods, Borghesi, Houston, and Naranjo data
covers the years 1992 to 2006 and the Di Giuli and Kostovetsky covers the years 2003 to 2009; or they could be driven by the differences in sample firms as the underlying KLD sample firms increase substantially over the period

他にも興味深いのは、CEOの報酬は、CSRに金を使い過ぎると下がるというものだが、これは研究によっては無関係と出ている。

Using the same source of variation as Di Giuli and Kostovetsky (2014), Borghesi et al. (2014) offer an opposing view. They conclude
that, after controlling for CEOs who donate to both Democrats and Republicans, the relation between CEOs who donate to Democrats
and their firms’ ESG/CSR scores is insignificant. It is unclear what is causing the differences in results. They could be driven by the
presence of non-CEOs in the Di Giuli and Kostovetsky sample (which includes directors and founders) since the Borghesi, Houston, and
Naranjo sample is limited to CEOs; they could be driven by the differences in sample periods, Borghesi, Houston, and Naranjo data
covers the years 1992 to 2006 and the Di Giuli and Kostovetsky covers the years 2003 to 2009; or they could be driven by the differences in sample firms as the underlying KLD sample firms increase substantially over the period

他にも、機関投資家の株式保有がESGスコア低下につながるとの研究もあるが、これは２０１４年のものであり、現在の実態に当てはまらない可能性がある。またこれと反する研究結果も存在する。

Institutional investors are by far the largest owners of equity securities in the United States, the country on which many of the
studies focus. As mentioned earlier, examinations of institutional investor ownership patterns can help us understand shareholder
preferences with respect to ESG/CSR, as well as how owners might affect ESG/CSR through their role as monitors. Thus, the relationship between institutional ownership and firms’ ESG/CSR activities has generated a large number of studies, not all of which agree
on the form or sign of the relationship. In addition, they use different measures of firms’ ESG/CSR profiles.
Regarding associations reflective of shareholder preferences, Borghesi et al. (2014) find that institutional ownership is negatively
related to firms’ ESG/CSR scores and Gillan et al. (2010) provide evidence that during their sample period when firms improve their
ESG scores, institutional ownership actually declines. Several studies, however, have shown that this relationship is nuanced.

機関投資家の中でも、公的年金が投資家の場合、社会的に無責任な会社は持たれない傾向にあることが示されている。

nvironmental performance.
Two studies focus on categorizations of institutional investors and conclude that the particular type of institutional investor is
important for understanding shareholder preferences. Hong and Kacperczyk (2009) show that socially constrained institutions (e.g.
pensions) have a distaste for socially irresponsible stocks. As a consequence the authors argue and provide evidence that these stocks
tend to be disproportionately held by less constrained institutions (e.g. mutual funds and hedge funds). Researchers have also categorized mutual fund ownership according to the portfolio manager’s political stance. Hong and Kostovetsky (2012) find that mutual

またポートフォリオマネジャーが民主党に寄付している場合、社会的に無責任な会社への投資が減る傾向が示されている

また投資家のエンゲージメント活動によりESGスコアが上昇したり、環境に有害な活動が減少したことが示されている。

r Hoepner et al. (2019) find that the
successful engagements on environmental issues are accompanied by subsequent reductions in downside risk. Similarly, Naaraayanan
et al. (2019) examine the unexpected focus on environmental issues by the New York City Pension System’s Board Accountability
Project and, using plant-level data, provide evidence that target firms subsequently have lower environmentally harmful emissions.

なお投資家の投資期間が長くなれば投資先企業のESGスコアが高まる傾向にあり、ESGスコアが高ければ業績が悪くても持ち続けられる傾向があることも示されている。ただESGやCSRスコアと機関投資家の関係は研究によって異なり、おそらくダイナミックなため将来期間が変われば結果が異なる可能性も示唆されている。

Several studies examine the importance of the investor’s horizon, including both how it affects and is affected by ESG/CSR. Kim
et al. (2019) employ difference-in-differences estimations to examine whether ESG/CSR scores increase when an institution changes its
investment horizon to long-term and conclude that the scores increase after the change. Consistent with this, Gloßner (2019) uses leadlag regressions of long-term investors, primarily blockholders, and concludes that they have a positive relationship with firms’ ESG/
CSR scores. In contrast, Starks et al. (2019) argue and provide evidence that long-term institutional investors are attracted to firms with
higher ESG/CSR profiles rather than such institutions influencing the firms’ choices directly. Supporting this result, they also find that
the long-term institutional investors are more patient with short-term poor performance of high ESG firms than other firms in their
portfolio.

またファミリー企業かそうでないかで、USやスウェーデンではファミリーオーナーシップがある企業の方がESG.CSRスコアが高いが、東アジアでは正反対の結果が出ており、これは地域差か、データ差か、分析手法の違いかが指摘されている。

In consideration of the agency issues involved in ESG/CSR choices and whether these choices benefit shareholders, Abeysekera and
Fernando (2020) hypothesize that family firms and non-family firms differ in their agency issues regarding ESG/CSR. The authors
expect that management of family firms would align their ESG/CSR choices with shareholder wealth maximization due to the lack of
diversification by controlling families. However, the authors also point out that an alternative hypothesis exists as conflicts can arise
between family owners and minority shareholders, in which case the family firms could make decisions that do not result in wealth
maximization for all shareholders. The authors examine their contrasting hypotheses through empirical tests in which they measure

また国営企業だった会社は、政府の影響があるのでCSRスコアが高くなる（中国ではそれは観察できなかった）という研究もある。

Some argue that environmental or social issues should be managed by governments rather than corporations because of governments’ superior abilities to handle the issues. In particular, state-owned firms may be better positioned to deal with the market failures
and externalities caused by such issues as has been hypothesized by Hsu, Liang, and Matos (2018) and Hart and Zingales (2017). On the
other hand, governments and the firms they own could have other incentives that discourage ESG/CSR. The empirical evidence tends
to be more consistent with the first hypothesis, although it is not uniformly so. Hsu, Liang, and Matos consider a sample of firms across
45 countries that includes publicly-owned companies with majority government ownership. The authors show that state-owned firms
are more engaged in environmental and social issues than other firms, a result concentrated in energy firms and firms in emerging
economies. They also use a difference-in-difference analysis and find that following the 2009 Copenhagen Accord, state-owned firms
improved their environmental performance more than other firms. Consistent with this view, using Asset4 Environmental and Social
scores as measures of ESG/CSR and a sample of publicly traded firms from 41 countries that privatized from state ownership, Boubakri
et al. (2019) find that before their privatization, the privatized firms have higher ESG/CSR scores in aggregate, and on both the
Environmental and Social dimensions, than other publicly listed firms.11 They further find that state ownership and the political
environment of the country are influencing factors in this relationship. On the other hand, examining one country’s state ownership
(China) McGuinness et al. (2017) provide evidence of a nonlinear relationship between state ownership and corporate ESG/CSR
profiles. The association is negative at lower state ownership levels and positive when state ownership is higher. Future research might
reconcile these somewhat conflicting results.

ESG/CSR using the KLD environmental rankings. They conclude that in decisions about environmental investments, U.S. family firms
are more responsible to shareholders than are the non-family firms. Consistent with this view, Gillan, Sekerci, and Starks (2020) focus
on a sample of Swedish firms and find evidence suggesting that family firms cater to investor demand for environmental (but not
social) investment. The results of these two papers are somewhat counter to those of El Ghoul et al. (2016) who examine a sample of
publicly traded firms from nine East Asian economies. In their sample, the family-controlled firms have lower ESG/CSR performance,
which the authors argue is consistent with the expropriation hypothesis of family control (the conflicting interests between family
owners and minority shareholders). The authors also provide evidence that these family-controlled firms with lower ESG/CSR performance have greater agency problems and reside in countries with weaker institutions. One issue that remains unanswered is
whether these differences in results are due to differences in geography or differences in data and methodology.