Discussion of ”Leviathan Inc. and Corporate Environmental Engagement” by Hsu, Liang, and Matos

(1)

Discussion of

”Leviathan Inc. and Corporate Environmental Engagement”

by Hsu, Liang, and Matos

Per Strömberg, SSE & SHoF

SHoF-MFS Conference on Sustainable Finance

August 20-21, 2018

1 11/15/2021

(2)

Motivation

• One of the central questions in governance: is there any conflict between shareholder vs. stakeholder value

• “Doing well by doing good” vs. “Gordon Gekko”

• State-owned enterprises an interesting laboratory

• Increasing in economic importance in “emerging markets” (China, Brazil, Russia – “Leviathan”

• Generally considered badly managed / governed in terms of financial returns

• But what if SOEs are superior in generating value to other stakeholders?

• Interpretation if this is the case?

• There is indeed a conflict between shareholder and stakeholder value (?)

• There is a rationale for state ownership in optimizing total stakeholder value (rather than suboptimizing by maximizing shareholder value?)

11/15/2021

2

(3)

This paper

• Sample of large publicly traded companies (index inclusion) around the world with ESG scores and ownership info

• SOEs seem to indeed have higher “E” ratings

• And also some evidence on higher “S” ratings but not “G”

• Not at the expense of lower Q

• Effect of state ownership on E (and S)

• Only found in emerging markets

• Stronger in oil and gas firms, firms w. lower foreign sales, countries with energy risk and conflicts with neighbors

11/15/2021

3

(4)

My overall view

• Nice data work: high-quality ESG measures, large sample, careful ownership classification (different kinds of govt ownership)

• Many interesting facts.

• Good to challenge “conventional” view on state-owned firms and governance

• Finding that state-owned firms have higher E scores seems robust

• Although some issues regarding interpretation, measures, selection

• Finding that state-owned firms’ CSR are not at expense of profitability / efficiency much less convincing

• External vs internal validity

11/15/2021

4

(5)

Theoretical predictions?

• Shareholder vs stakeholder governance focus interacts with the traditional agency problem of self-interested management

• E.g. Jensen argument for shareholder focus

• Managers chooses action where marginal (monetary and non-

monetary) benefits equals marginal cost, which depends both on governance and technology  ESG actions depend on both

technology and governance, and its interaction

• Strong shareholder governance  manager will focus on ESG issues that have least negative (most positive) impact on profitability

• Weak shareholder governance  manager might focus on ESG issues that yield the highest private benefits

11/15/2021

5

(6)

Implies selection effects

• Cost-benefit of ES(G) will differ across industries

• Decreasing CO2 emissions is less costly for a bank than for an airline

• Decreasing CO2 emissions is less costly for the bank, compared to refraining selling shady tax schemes

• In addition, state ownership depends on industry and general institutional environment

• Natural monopolies, natural resources

• Political system, rule of law

• Decision to privatize vs. not

 Challenging to identify effects of governance on ES(G) in cross- country, cross-industry data.

11/15/2021

6

(7)

Trade-offs…?

11/15/2021

7

(8)

Trade-offs…?

INTRODUCTION / FOUNDER LETTER / / BUILDING BETTER BUSINESSES / SUSTAINABLE DEVELOPMENT GOALS / SUSTAINABILITY IN ACTION / DIVERSITY / MEETING THE UNGC PRINCIPLES

Recognition Of Progress

Leading Tunisian soft discount convenience store chain Aziza, an Abraaj partner company, received the MEDA award in recognition of the successful completion of the Sustainability Innovation Grant Program (SIG). SIG is a prestigious program developed by INFRONT (Impact Investing in Frontier Markets), an innovative public-private partnership between the Government of Canada and three Canadian organizations (MEDA, Sarona Asset Management and MaRS Center for Impact Investing) for sustainable development in frontier and emerging markets.

Aziza Receives the MEDA Award

The Abraaj Group received the highest

accolade under the UN-supported Principles for Responsible Investment’s (PRI) Reporting and Assessment Framework for the third consecutive year. The score is well above the industry standard for private equity, which has been ranked at level B for the past three

reporting cycles.

The Abraaj Group Receives A+ Rating for the

Third Consecutive Year

Tunis, Tunisia HIGHLIGHTS

7 11/15/2021

8

(9)

Normalization of ESG scores

• Thomson Reuters score reflects relative ESG within industry; avg score of 50

• Normalization is a double-edged sword

• No need to control for fact that different industries have different emissions.

• But: Industries do differ in importance of environm. concerns; here, treated alike

• Does not account for governance-industry interactions

• Seems problematic for diff-in-diff analysis (SUTVA violation)

• E.g. Fukushima makes companies increase average environmental investment

• But measure recalibrates this every year to an average of “50”.

• When one group catches up, looks like other group is deteriorating

• (Also: What are pre-trends? Why no firm FEs throughout? ))

11/15/2021 9

37

Table 1. Forbes Top-Ranked Global Companies, 2010

This table presents the average values of state ownership (State_own), the environmental pillar scores (ENVSCORE and sub-categories scores: emission reduction ENER, product innovation ENPI, resource reduction ENRR), social pillar scores (SOCSCORE), and corporate governance pillar scores (CGVSCORE) of the top publicly listed companies in the Forbes Global 2000 list for 2010. The top 10 state-owned enterprises are highlighted in boldface. Country abbreviations are described in Figure 2.

Forbes Rank 2010 Country State_own ENVSCORE SOCSCORE CGVSCORE

ENER ENPI ENRR

1. JPMorgan Chase US 0 92.50 76.57 97.25 87.06 66.48 72.70

2. General Electric US 0 95.06 94.53 97.69 95.05 90.78 94.49

3. Bank of America US 0 77.54 48.28 86.94 80.64 67.41 82.06

4. ExxonMobil US 0 94.19 92.48 94.75 93.17 91.67 86.78

5. ICBC CN 1 87.86 72.09 95.19 85.65 78.27 78.98

6. Banco Santander ES 0 93.21 92.03 87.77 93.30 95.23 89.16

7. Wells Fargo US 0 91.92 93.11 88.13 84.08 59.39 82.47

8. HSBC Holdings GB 0 93.40 93.63 87.41 93.41 86.73 84.91

9. Royal Dutch Shell GB 0 89.69 79.54 89.40 92.34 78.23 87.56

10. BP GB 0 89.86 89.45 75.50 89.25 87.12 83.28

11. BNP Paribas FR 0 93.04 87.99 97.34 90.84 94.07 90.89

12. PetroChina CN 1 57.50 64.25 15.44 75.30 81.13 19.74

13. AT&T US 0 92.71 93.39 88.22 88.37 79.26 91.63

14. Wal-Mart Stores US 0 86.55 69.81 71.89 88.95 75.46 94.06

15. Berkshire Hathaway US 0 9.36 9.39 14.92 8.92 3.75 63.05

16. Gazprom RU 1 81.95 91.28 53.11 79.10 76.46 6.99

17. China Construction Bank CN 1 53.33 34.44 87.36 35.94 81.45 28.92

18. Petrobras BR 1 91.67 90.93 84.42 88.34 93.80 34.01

19. Total FR 0 89.70 77.73 87.75 83.24 83.63 65.24

20. Chevron US 0 90.42 86.96 87.89 82.06 63.51 77.78

21. Barclays GB 0 94.11 90.95 94.89 92.44 93.23 86.60

22. Bank of China CN 1 79.61 37.93 95.50 88.15 82.44 49.77

23. Allianz DE 0 93.50 93.66 88.13 93.40 93.40 78.88

24. GDF Suez FR 1 90.06 92.34 88.28 78.89 95.71 76.96

25. E ON DE 0 91.60 94.91 85.84 84.94 96.59 29.78

26. Goldman Sachs US 0 92.12 78.15 87.37 93.51 53.77 74.37

27. EDF Group FR 1 92.86 84.90 97.53 88.77 96.13 33.16

28. AXA Group FR 0 93.39 85.18 95.44 93.31 94.37 82.90

29. Lloyds GB 1 90.01 92.48 69.86 92.90 93.20 73.90

30. Procter & Gamble US 0 94.69 92.76 97.41 93.50 92.54 81.51

31. ENI IT 1 89.02 83.41 81.75 84.79 96.11 59.61

What do we care most about: a Chinese bank having high

environmental score, or a Chinese oil co having it?

(10)

Normalization, cont.

• Despite this normalization, average scores differ

across industries

• TR uses different industry definitions?

• Suggests important to control for industry-year and country-industry FEs.

• Do this partly in appendix table; qualitatively similar, but much weaker results.

11/15/2021 10

39 Table 2. (continu ed) Panel B: Univaria te Tests by Major Industry Industry Obs. State_own ENVSCORE SOCSCOR E CGVSCORE

All State own=1 State

. ores as well for other sc big effects and might be ental score, for environm meaningless practically are comparisons try Across-indus these measur es as well owned for Do by state- 53.36 52.08 0.00 51.14 57.40 0.066 51.52 Total 28,876 64.80 Utilities 1405 0.256 63.53 62.40 63.32 55.66 0.36 52.13 62.53 0.00 51.95 63.37 0.317 55.43 ns 771 Telecommunicatio 63.00 Technology 1,960 0.021 51.69 58.82 51.46 0.03 51.53 64.61 63.62 48.52 0.00 42.69 0.126 45.48 Oil & Gas 2,061 Industrials 53.83 0.053 59.08 5,610 59.38 0.00 55.40 52.47 Health Care 55.82 50.63 0.00 44.06 20.76 0.010 43.79 1,633 Financials 46.36 0.069 43.23 5,059 43.04 46.02 49.99 0.06 53.55 46.35 0.00 40.79 52.56 0.023 41.05 es 3,992 Consumer Servic 47.15 Consumer Goods 3,370 0.019 61.55 57.76 61.90 0.00 46.95 Basic Materials 54.89 53.39 0.07 55.40 59.84 0.056 55.58 3,015 own=0 p-val (1 - 0) ue

(11)

Selection of firms scored by Thomson Reuters (& MSCI)

• Driven by institutional investor demand  Index inclusion main criteria

• Bias towards largest firms & more developed stock markets  get larger size bias in EMs

• Size is a main determinant of ESG

• Larger firms have more resources for sustainability accounting, CSR, etc.

• Larger fraction of small firms explains why U.S. has relatively low E&S scores?

• Unbalanced sample and different selection of firms in developed vs emerging mkts.

• Simply controlling for size etc will not help

• Seems to call for a matched sample, or at least truncation. E.g.:

• Run regression predicting state-ownership based on size, industry, country

• Remove if P(state)<0.05, >0.95

11/15/2021 11

Firm-year

obs % all firms

# state- owned

firm- observatio

ns

% of state- owned firm- observation

s

US 8 536

30% 26

1.4%

Other developed 16 796

58% 988

52.1%

Emerging 3 558

12% 882

46.5%

Total 28 890

1 896

(12)

Hard to figure out mechanism

• Univariate: SEO effect is only present in developed, who also have higher E-score

• Multivariate + splitting sample: SEO effect is only present in emerging markets

• What is the important control that switches result?

• (Why different splits across tests?)

• Additional regressions suggests further qualifications to result

• Oil & gas, domestic sales, energy security, conflict with neighbor, government political orientation…

• Seem likely to be correlated, but enter one at a time. 11/15/2021 12

3 6

F ig u r e 3 . A v er a g e E n v ir o n m en ta l S c o r es (E N V S C O R E ) o f S O E s a n d N o n -S O E s, p er C o u n tr y

T h i s f i g u r e p r e s e n t s t h e a v e r a g e e n v i r o n m e n t a l s c o r e s ( E N V S C O R E ) o f S O E s a n d n o n - S O E s i n e a c h c o u n t r y . F o r a l l f i r m - y e a r o b s e r v a t i o n s i n t h e S O E g r o u p o r t h e n o n - S O E g r o u p i n e a c h c o u n t r y i n t h e s a m p l e p e r i o d f r o m 2 0 0 4 t o 2 0 1 4 i n t h e s a m p l e p e r i o d f r o m 2 0 0 4 t o 2 0 1 4 , w e c a l c u l a t e t h e i r p o o l e d a v e r a g e i n E N V S C O R E . T h e r e i s n o b a r f o r c o u n t r i e s w i t h o u t S O E .

Developed the diff his higher between SEO vs not, and

higher scores generally.

Country-level stats

sensitive to number of obs. Very small eg for Hungary. More likely to end up in extremes.

3 6

F i g u r e 3 . A v e r a g e E n v ir o n m e n ta l S c o r e s (E N V S C O R E ) o f S O E s a n d N o n - S O E s, p e r C o u n tr y

T h i s f i g u r e p r e s e n t s t h e a v e r a g e e n v i r o n m e n t a l s c o r e s ( E N V S C O R E ) o f S O E s a n d n o n - S O E s i n e a c h c o u n t r y . F o r a l l f i r m - y e a r o b s e r v a t i o n s i n t h e S O E g r o u p o r t h e n o n - S O E g r o u p i n e a c h c o u n t r y i n t h e s a m p l e p e r i o d f r o m 2 0 0 4 t o 2 0 1 4 i n t h e s a m p l e p e r i o d f r o m 2 0 0 4 t o 2 0 1 4 , w e c a l c u l a t e t h e i r p o o l e d a v e r a g e i n E N V S C O R E . T h e r e i s n o b a r f o r c o u n t r i e s w i t h o u t S O E .

Developed the diff his higher between SEO vs not, and

higher scores generally.

Country-level stats

sensitive to number of obs. Very small eg for Hungary. More likely

to end up in extremes.

(13)

High ES in govt firms not at expense of financial performance?

• Huge literature on inefficiencies / underperformance of SOEs

• E.g. Megginson & Netter (2001), Chen et al (2017)

• Paper claims to find evidence to the contrary:

”This paper’s findings are more in line with a ’social view’ that SOEs can be effective in addressing environmental externalities.”

”Importantly, we do not find support for alternative view of state-ownership

(the agency cost view) that SOEs are managed by incapable managers and are captured by politicians to fulfill their political agenda rather than maximizing social welfare.”

”Governments can promote green technology by imposing carbon taxes and providing research subsidies … Alternatively, the state can intervene by

holding ownership stakes in public corporations.”

11/15/2021

13

(14)

How convincing is this argument?

• Evidence to back this up:

• M/B increasing in E, and E*SEO insig.

• G unrelated to SEO

• Not enough new info to change my priors on

efficiency of government firms or optimal climate change policy…

• Would advise authors to tone down this message.

11/15/2021 14

2 2.25 2.5 2.75 3 3.25 3.5

Private Equity Dispersed Shareholders Family, external CEO Managers Other Private Individuals Family, family CEO Founder Government

Average score on 18 management practice questions

Note: Sample of 4,221 medium-sized manufacturing firms. The bottom bar-chart only covers the 3696

firms which have been in the same ownership for the last 3 years. The “Other” category includes venture

capital, joint-ventures, charitable foundations and unknown ownership. Source: Bloom, Sadun, & Van

Reenen (2009)

(15)

How convincing is this argument?

• Evidence to back this up:

• M/B increasing in E, and E*SEO insig.

• G unrelated to SEO

• Not enough new info to change my priors on

efficiency of government firms or optimal climate change policy…

• Would advise authors to tone down this message.

11/15/2021 15

2 2.25 2.5 2.75 3 3.25 3.5

Private Equity Dispersed Shareholders Family, external CEO Managers Other Private Individuals Family, family CEO Founder Government

Average score on 18 management practice questions

Note: Sample of 4,221 medium-sized manufacturing firms. The bottom bar-chart only covers the 3696

firms which have been in the same ownership for the last 3 years. The “Other” category includes venture

capital, joint-ventures, charitable foundations and unknown ownership. Source: Bloom, Sadun, & Van

Reenen (2009)

(16)

Macro vs. micro studies

• Trade-off between external vs. internal validity

• Maybe easier to credibly identify mechanism in a more focused study, looking at one industry (e.g. oil and gas), more direct outcome measures (e.g. emissions), with