ESG(+D)?
Bridging the digital rights data gap
Key Takeaways
For responsible tech investment to be feasible, investors need reliable information on how companies manage their impacts on digital and other human rights. Unfortunately, mainstream ESG ratings fail to adequately account for and measure the “S” and “G” risks prevalent in tech.
While there are initiatives attempting to standardize and measure digital rights risks, none offer investors comprehensive, decision-useful data to guide their responsible tech investment practices.
Foundations can bridge this digital rights data gap by supporting closer collaboration among investors and digital rights advocates and experts, campaigns seeking greater accountability from ESG raters, advocacy for greater regulation of ESG data, and the development of investor-focused digital rights standards.
From 2010 to 2020, assets managed using ESG investing strategies by US-domiciled institutions grew from $3 trillion to more than $17 trillion, with one in three US dollars invested now being managed according to ESG principles (1). For ESG investors, the goal is to reallocate capital to companies that are better at managing their risk of adversely impacting people and the planet. In a similar vein, impact investors, whose assets under management are also on the rise (2), seek returns by investing in companies that are trying to achieve desired social or environmental outcomes. Both types of investors use shareholder engagement strategies to achieve their goals.
Overview
The growth in popularity of these responsible investment strategies should be a boon for corporate accountability advocates. In addition to having a wider net of potential campaign partners, the desire to access this deepening pool of capital should, in theory, independently incentivize companies to better manage their risks. The upside of this shift in investor sentiment is even greater for responsible tech advocates, whose adoption of finance-focused strategies has yet to reach the maturity of the climate justice movement.
That said, the realization of this potential depends on responsible investors having the requisite information about a company’s practices and impacts to confidently determine whether it meets their investment criteria. This is where ESG ratings come in. First introduced in the 1980s, these ratings allow investors to readily screen companies on their environmental, social, and corporate governance performance. To evaluate this performance, mainstream ESG ratings agencies (3) rely on their own proprietary methodologies and ESG data gleaned primarily from corporate disclosures, surveys, and in some cases information from external stakeholders (4).
The structure of the ESG financial ecosystem
The proliferation of these ratings should mean that responsible investors are redirecting capital to more socially responsible companies. Yet, large tech companies remain popular with ESG investors in spite of their well documented governance and human rights failures (5). The reason is two-fold: mainstream ESG ratings fail to account for the unique human rights harms and weak governance endemic to the tech sector and have methodological limitations that lend themselves to overestimation of tech sector ESG performance.
While there are numerous disparate initiatives that offer piecemeal standards and metrics to define and measure responsible tech practices, to-date none offer investors comprehensive, decision-useful information that would facilitate effective engagement and screening on digital rights issues. There is thus an opportunity to bolster the leverage of responsible investors interested in promoting greater accountability in the tech sector by increasing transparency and regulation around mainstream ESG products and by redressing the ESG data deficit on tech-specific impacts through investor-friendly standards and metrics that accurately assess the human rights risks of the digital era (6).
ESG’s Embrace of Big Tech: Why Companies Weak on Human Rights Get Strong Scores
Unlike environmental impacts, impacts to human rights and society more broadly are intrinsically harder to measure than the kind of straightforward, numerical rankings upon which investors typically rely. This is particularly true for the tech sector, where “S” impacts range from violations to individual rights like the right to privacy and non-discrimination to broader social ramifications like hate speech and the undermining of democracy. As one responsible tech researcher puts it, “most digital harms lack ‘natural attributes’ that are countable or physically measurable. Further complicating the challenge, there are few widely accepted methods for linking potential or actual digital harms and opportunities to revenue streams or return on investment” (7).
It is then no surprise that responsible investors feel mainstream ESG ratings are not useful tools for measuring the social impacts of tech companies. As one civil society actor who works in tech accountability told us, mainstream ratings agencies are not accurately reflecting the “S” risk of tech companies. Instead, they typically focus on supply chain risks or labor rights issues and fail to capture digital rights challenges. This is why all of the investors we interviewed indicated that while they buy third party data, they largely use it as a high-level screening tool and not as a means of identifying the important issues on which to engage tech companies. Investors told us they either determine engagement issues based on “art and intuition,” rely on their close relationships with civil society partners to help identify focus areas for activism, or try to determine standards of best practices on their own.
The challenge of measuring tech sector social impacts contributes to the underestimation of the sector’s overall social risk relative to others. Benjamin Chekroun, Engagement Specialist at CANDRIAM, a European-based asset manager suggested that “We need a strong coalition on human rights similar to what Climate Action 100+ does on GHG emissions.” But the challenge is in how to weigh the risks of tech companies against the more conventional human rights risks of other sectors. This was echoed by an actor in the impact investing space who told us that, unlike industries with externalities that are more readily measured, it is incredibly difficult to measure the cost to society or the economy of the Big Tech companies, and this is why even responsible investment portfolios are very tech-heavy.
→ What’s not being measured
Digital rights impacts are hard to assess.
Facebook, Volkswagen, and Wirecard all had good ESG ratings from mainstream ratings agencies before starkly negative ESG incidents were uncovered (8). The most fundamental reason companies with harmful business practices receive strong ESG ratings and consequently appear in ESG funds is that, for most mainstream ratings agencies, risk is not defined in terms of a company’s social and environmental impact but in terms of how ESG factors affect a company’s profitability (9). One large asset manager we spoke with said that while they use third-party ESG data, they are aware that this data often excludes important human rights and DEI [diversity, equity, and inclusion] issues because these issues are not considered “material” risks.
While it may no longer be profitable for companies to ignore climate risk as the costs associated with climate change are likely to reach every corner of the economy, there are many other ESG impacts that do not pose immediate financial risk to companies or investors. This is particularly true of digital human rights risks. Technologies like facial recognition and surveillance tools are profitable despite – or even as a result of – their inherent violations to internationally-recognized human rights (10). Absent significant damage to a company’s reputation or legal action following a human rights controversy that makes headlines, human rights impacts may not be viewed by ESG raters as “material” issues for tech companies.
“Surveillance is an area where it’s very lucrative to ignore human rights. From a civil society perspective, campaigning and storytelling is necessary to highlight brand damage,” says Isedua Oribhador, Business and Human Rights Lead at Access Now. “We need to make it clear to companies that even if they make money on these technologies, their brand value will decline. Companies with the majority of their business in ‘B2C’ are more susceptible to this kind of campaigning, so the pressure needs to be even more sustained for ‘B2B’ companies.”
A similar issue arises with respect to the governance component of tech company ESG scores. As discussed in our Shareholder Engagement in Tech report, tech companies commonly employ multi-class share structures that undermine investor oversight and impede good corporate governance. Given the proprietary nature of their methodologies, we are not certain the extent to which mainstream ESG ratings agencies take this issue into account when rating tech companies.
Tech companies benefit when risk is measured in terms of profit, not people.
The data underlying ESG ratings generally come from five distinct sources: voluntary corporate reporting, regulatory filings, media coverage, questionnaires completed by companies, and modeled data (11). This suggests an overwhelming reliance on company self-reporting. According to one civil society actor, the vast majority of ESG ratings criteria are based on public disclosure and are “devoid of any quantitative or qualitative analysis of human impact.” In practice, if a company fails to report its adverse impacts and these impacts are not reported in the media, they will not be factored into the company’s ratings (12). “The challenge with the ‘S’ in ESG is that while some things can be readily measured, other dimensions only emerge through contestation. If there are no activists challenging a company on an issue, it basically doesn’t exist,” says Eli Kasargod-Staub of Majority Action, which organizes proxy campaigns in support of ESG issues.
This is particularly of concern in tech. Most big tech companies have sophisticated disclosures regarding their human rights commitments and processes for mitigating human rights risk. These policies and processes are generally thought of as market leading, with smaller tech companies and companies in other sectors lagging behind. As a result, these large tech companies are more likely to receive higher “S” scores from ESG raters, despite persistent evidence of adverse human rights impacts (see Appendix A).
Isedua Oribhador says, “Big Tech companies have become fairly fluent in the expected human rights narrative; they have all the United Nations Guiding Principles on Business and Human Rights expertise and requirements on paper. Companies are savvy in understanding what they need to do to tick the boxes, but when you look at what they’re actually doing, it’s in complete conflict.”
For example, Meta recently released its first human rights audit, which assesses the company’s approach to managing human rights risks. However, the company’s disclosure failed to include a full assessment of its impact in India, home to Facebook’s largest user base. This omission drew criticism from human rights advocates. Ritumbra Manuvie, an academic who was interviewed as part of the audit, said the disclosed summary was a “cover up of [Meta’s] acute fault-lines in India,” and showed that the company’s “commitment to human rights is rather limited” (13).
Human rights disclosures can obscure actual harmful impacts.
Opportunity for Action: Investor partnerships with civil society
The above analysis reveals that investors lack access to data that tells them about actual company impacts to digital rights. This is why several of the investors we interviewed indicated that they rely on their partnerships with civil society organizations to understand and prioritize the issues on which to engage tech companies. In the short term, until decision-useful standards and metrics exist for gauging tech company social impacts in an investment context, there is an opportunity to help bridge the informational gap by facilitating knowledge sharing between the investor community and civil society organizations representing affected rights-holders. “The investor community should engage more with civil society to get information about what a company is actually doing versus what they say they’re doing,” says Laura Okkonen, Investor Advocate at Access Now. “Big institutional investors don’t have time to start engaging, so we need to put together a shadow index of what civil society sees these companies doing.”
→ What's not being measured well
Mixed signals from “S” and “G” ratings weaken tech company accountability.
As one civil society actor in the responsible investment space told us, “‘ESG’ has become a buzzword that means different things to different people. Without greater standardization to align ESG standards with international human rights norms, anyone can claim to be an ESG ‘expert’ and shape the meaning of this term.” For example, Inclusive Development International (IDI) found that ESG-labeled funds funneled billions of dollars into companies arming, funding, and legitimizing the Myanmar military, which has been recognized as the perpetrator of the Rohingya genocide and a violent crackdown on the country’s pro-democracy movement (14). Because ESG ratings agencies develop their ratings in-house, it is not clear how significantly human rights abuses impact a company’s overall ESG score. According to IDI, “this is by design” (15).
Ratings agencies decide which attributes should be evaluated as part of their scoring procedure in addition to the relative importance of the attributes with respect to final scores. This has led to a low degree of correlation among the scores one company will receive from different ratings agencies (16). And it is particularly true for the “S” and “G” of ESG. There is both less agreement across ratings agencies on the most important issues for these categories and a worse understanding of how to quantify the real impacts of these issues (17). Since “S” and “G” impacts are arguably the most salient for tech sector companies, these companies are more likely to have inconsistent ratings (18).
Selected ESG issuer ratings and issuer credit ratings by sector in the United States, 2019
Note: Sample of public companies selected by largest market capitalisation to represent different industries in the United States. The issuer credit ratings are transformed using a projection to the scale from 0 to 20, where 0 represents the lowest rating (C/D) and 20 the highest rating (Aaa/AAA).
Source: Refinitiv, Bloomberg, MSCI, Yahoo finance, Moody’s, Fitch, S&P; OECD calculations.
In addition to the obvious shortcoming of obscuring the true human rights and governance failures of companies, ratings disagreement has other important consequences. When companies receive mixed signals from ratings agencies on which actions are expected and valued by the market, this reduces the incentive to improve their practices and undermines attempts to link management compensation to ESG performance. It also makes it difficult for investors and other third parties to evaluate the performance of companies, funds, and portfolios, which in turn makes markets less likely to accurately price a company’s ESG risks (19).
Rater bias for climate wins and profitability benefit Big Tech.
One group of researchers found that raters displayed evidence of having been influenced by a “halo effect” whereby a company that receives a high score in one category is more likely to receive high scores in all others (20). There is also evidence of one ratings agency retroactively upgrading the “E” and “S'' scores for companies that performed better in a given year (21). The likely explanation for the upgrade is that investors select ESG ratings providers primarily on the extent to which their ratings predict returns (22). Both the halo effect and profitability biases stand to artificially inflate the ESG scores of large technology companies: these companies outperformed the market in recent years, and their relatively low climate risk often draws attention away from their failures in the social and governance categories.
There is evidence that these biases are reflected in investment decision making: 28 of the largest and best-known ESG funds have invested the majority of their clients’ money in Microsoft, Alphabet, Apple, and Amazon. Meta was also overrepresented in the case of passive funds. The main reason for this preference for tech giants in ESG funds is their relatively small carbon footprint and, to an extent, their profitability: all these funds marketed themselves as ESG funds or socially responsible funds, but they are, for the most part, carbon-free funds. It also doesn’t hurt that these companies have been on a years-long bull run, outpacing market averages and most typical stocks (23).
Opportunity for Action: Tech sector accountability through rater accountability
The tendency for big tech companies to receive inflated ESG scores carries negative consequences. For one, ESG ratings inflation has been associated with future negative ESG incidents (24). This suggests that reducing the ESG ratings inflation associated with large tech companies could be an important factor in improving their practices. Further, the inconsistencies and limitations of the existing ESG ratings undermine the credibility and overall impact of responsible investing as a strategy for change and give fodder to critics of ESG.
There is thus an opportunity to bolster both tech sector accountability and responsible investment practices by engaging mainstream ESG raters to improve their methodologies, particularly in the “S” category and with a focus on digital rights issues. As one civil society actor told us, “challenging ratings agencies like MSCI, Sustainalytics, and ISS is a systemic, high leverage opportunity to address the big problems with responsible investment.”
More than one interviewee suggested to us that it would be worthwhile to organize a campaign around ESG rater accountability that involved multiple organizations in the responsible investment ecosystem, similar to the highly influential coalition that has been engaging the banking sector on climate change. One civil society organization shared that they have already begun trying to find a way to engage with ESG data providers, using the protection of human rights defenders as a wedge issue to expand the human rights-related criteria typically covered in these metrics. Another suggested that greater transparency of ESG ratings would be an effective pathway to holding raters accountable: “We should know what criteria they are using and who has paid for what. ESG raters should be held to the same standards that we apply to the other areas of the financial sector.”
Many interviewees also agreed that there should be greater regulation of ESG data both as a means of ensuring the coherence and credibility of the data but also to entrench mandatory disclosure of ESG-related information. This process has already begun at the SEC, with a specific focus on climate-related data (25). There is also precedent for this in other countries: Colombia and Chile now require disclosure of ESG information according to SASB [Sustainability Accounting Standards Board] standards. A campaign for greater standardization and institutionalization of ESG ratings and standards would both improve ESG rater accountability and bolster ESG investors in the long term.
A Digital Rights Patchwork: Standards and Metrics for Investor Assessment of Tech Sector Risks and Impacts
We set out to identify existing standards, datasets, and tools that attempt to either define or measure responsible technology development and use in practice. As the aim of this broader project is to examine investor-led or finance-focused strategies, we excluded from this overview any standards that are focused solely on government use of technology. We also excluded standards and tools developed by individual technology companies (e.g. Google’s AI Principles) so as to avoid any potential conflict of interest.
The interactive database below contains 90 initiatives that we categorized in terms of type (high level guidance, company rankings, industry standards, investor guides, and tools) and according to topics or issues covered (algorithmic fairness (26), privacy (27), content moderation (28), end user due diligence (29), internet access (30), bias and discrimination (31), international human rights (32), freedom of expression (33), and disinformation (34).
Individual database cells can be expanded by clicking on them, and the data can be sorted by type and issues covered using the “Sort” button at the top.
High Level Guidance
Initiatives in this category either seek to define or problematize a social issue related to technology use or development, offer broad principles for guiding behavior with respect to technology, and/or call for action around the use or development of particular technologies.
→ What’s out there: Overview of existing standards, datasets, and tools
Highlights
Asilomar AI Principles: These 23 principles are guidelines for the research and development of AI. They outline AI developmental issues, ethics and guidelines for the development of beneficial AI and to make beneficial AI development easier. These principles have been signed by 1,797 AI and Robotics researchers and 3,923 others to date.
Fair Information Practice Principles: The Fair Information Practices, also known as the Fair Information Practice Principles (FIPPs), are a set of eight principles regarding data usage, collection, and privacy. They were published in 1980 by the Organization for Economic Cooperation and Development (OECD), and a number of countries agreed on them in principle. Many organizations use them as guidance for how to handle personal data.
Global Network Initiative Principles: The Principles on Freedom of Expression and Privacy provide direction and guidance to the information, communication and technology sector and its stakeholders in protecting and advancing the enjoyment of these human rights globally.
Company Rankings
Initiatives in this category most closely resemble mainstream ESG ratings in that they evaluate the relative performance of specific companies in the technology sector on a subset of issues of social concern.
Highlights
Ranking Digital Rights Corporate Accountability Index: Each year, Ranking Digital Rights evaluates and ranks 14 of the world’s most powerful digital platforms on their policies and practices affecting people’s rights to freedom of expression and privacy. Their Data Explorer is an interactive tool that allows for multiple views of the data collected, including by company and service; by lens, a new view that represents a curated group of indicators that illuminate a specific topic area; and by scores over time.
EthicsGrade: EthicsGrade scores companies on their AI governance. They rate companies on a quarterly basis, and new ratings are added every few days. Companies are graded on a scale of ‘D’ to ‘A.’ The ‘D’ grade is the lowest EthicsGrade provided and signifies a weak level of maturity towards the challenge of ensuring appropriate governance for their technology. Organizations which receive ‘A’ grades have adopted a “civil responsibility” for their actions, they are engaged with stakeholders, and have developed the appropriate muscle to respond to external challenges.
Investigate: Investigate is a tool created and managed by the American Friends Service Committee. The Investigate database includes original research and lists over 200 company profiles that examine each company’s relationship to state violence, including mass incarceration, immigrant detention and surveillance, military occupation, or the border industry.
Industry Standards
We use industry standards to refer to technical guidelines, recommended processes, and risk mitigation tools for individuals and organizations developing or using particular technologies with the aim of reducing adverse social impacts.
Highlights
PAS 440 Responsible Innovation Guide & Framework: This PAS provides guidance for companies that wish to innovate responsibly and to demonstrate their responsible behavior by helping them to structure their thinking and by guiding their actions. There are two main elements of responsible behavior by companies: a) company-level responsibility, applying to organizational behavior in general, and b) innovation-level responsibility, specific to each innovative development or class of developments.
Institute of Electric and Electronics Engineers (IEEE) Algorithmic Bias Considerations: This standard describes specific methodologies to help users certify how they worked to address and eliminate issues of negative bias in the creation of their algorithms, where "negative bias" infers the usage of overly subjective or uniformed data sets or information known to be inconsistent with legislation concerning certain protected characteristics (such as race, gender, sexuality, etc.) or with instances of bias against groups not necessarily protected explicitly by legislation, but otherwise diminishing stakeholder or user well-being and for which there are good reasons to be considered inappropriate.
International Standards Organization (ISO) Information Security Management: ISO/IEC 27001 is an international standard on how to manage information security. It details requirements for establishing, implementing, maintaining, and continually improving information security systems with the ultimate aim of helping organizations make the information assets they hold more secure.
Investor Guides
Initiatives in this category aim to brief investors on important issues related to the social impacts of particular technologies. Some go further in providing methods for assessing social risk in the investment context. While some are initiatives of investors themselves, others come from industry associations and civil society organizations.
Highlights
Sustainability Accounting Standards Board (SASB) Internet Media & Services Sustainability Accounting Standard: This standard sets out expected disclosure and metrics related to sustainability issues relevant to the internet and media services sector, most notably: data privacy, advertising standards & freedom of expression; data security; and intellectual property protection & competitive behavior.
Investor Alliance for Human Rights Salient Issue Briefings: In each salient issue briefing, the Investor Alliance for Human Rights provides investors with an overview of 1) relevant human rights instruments and authorities; 2); the real-world impacts of the information and communications technology sector on the salient issue; 3) an assessment of the business case for engaging on this issue; 4) targeted guidance for investors to consider in their engagement with companies; and 5) practical examples of investor action. Topics covered to date include political participation, conflict and security, freedom of expression, discrimination, and privacy and data protection.
Investors’ Expectations on Responsible Artificial Intelligence and Data Governance: Created by Federated Hermes, this document proposes a structured approach for investors to engage on AI and data governance based on six core principles and sets out a common framework across sectors. This framework is separated into two related strands. The first, taking a risk factor approach, assesses the materiality of issues for companies based on legal, regulatory, and financial outcomes. The second, taking a process-based approach, evaluates the impact of biases that may arise from data input and processes.
Tools
Tool-based initiatives offer step-by-step processes or questionnaires to be followed by the individuals and teams developing and/or auditing the impacts of different technologies in the course of their work with the aim of reducing adverse social impacts.
Highlights
American Civil Liberties Union (ACLU) Algorithmic Equity Toolkit: The Algorithmic Equity Toolkit is a collection of four components designed to identify surveillance and decision-making technologies used by governments; make sense of how those technologies work; and pose questions about their impacts, effectiveness, and oversight. It presents a flowchart with yes/no questions to help identify whether a technology is an automated decision system or a surveillance tool, both, or neither. In addition, it includes a set of open-ended prompts that help to explore potential impacts of these technologies and a set of questions about automated decision systems to ask government employees, elected officials, and vendors.
EthicalOS Toolkit: Created by the Institute for the Future and the Omidyar Network, this toolkit asks technology developers to consider whether the technology they’re building will someday be used in unexpected ways, how they can prepare themselves, and what new categories of risk should be given special attention. The toolkit consists of eight risk zones to help identify the emerging areas of risk and social harm most critical to start considering now; 14 scenarios to spark conversation and stretch imagination about the long-term impacts of tech being built today; and seven future-proofing strategies to take ethical action today.
UNICEF Data for Children Collaborative Ethical Assessment: Developed by the United Nations in partnership with Ethical Intelligence Associates, this tool provides an ethics assessment and safeguarding training for applying the seven Responsible Data for Children Principles in practice. Split into three sections, it supports technology developers in defining ethical issues before a project begins, revisiting existing and highlighting any new ethical issues throughout the duration of a project, and thinking about the consequences of communicating findings and how this could influence stakeholders.
While our analysis has revealed a plethora of initiatives that attempt to either set standards governing socially responsible technology development and use and/or collect data on the social impacts of companies developing and using technology, very few of these offer investors a practical means of evaluating the social risks of the tech companies in their portfolios. Broadly speaking, most initiatives suffer from one or more of the following pitfalls:
Too broad: High level guidance documents are not particularly useful in the investment context as their recommendations are often too broad to be applied without further elaboration around practical expectations for companies.
Compliance difficult/impossible to verify: While industry standards and tools offer more structured and practical guidance, investors nevertheless lack access to the internal company information needed to accurately assess company compliance with these initiatives.
Require specialized technical knowledge: Industry standards and tools often require specialized knowledge of particular technologies, which the average investor likely lacks.
Not dynamic or timely enough: While investor guides are helpful in briefing investors on important issues in the tech sector, they are not dynamic sources of information and become obsolete as technologies evolve and new issues emerge. In addition, these guides stop short of offering timely evaluations of company risk levels, which then leaves this task to the investor.
Not comprehensive enough: Company rankings are more useful in a practical investment context because they offer timely assessments of company risks across different issues. However, the offerings in this space are currently limited, both in terms of the number of companies evaluated and the frequency of evaluation. Further, while the four sets of company rankings we identified cover a broad range of issues when combined, there are still gaps, particularly around AI.
→ What’s needed: Where existing initiatives fall short
Opportunity for Action: Supporting the development of investor-focused standards for tech sector products and services
Despite the limitations of current iterations, standards are nonetheless important for accountability. As one civil society actor put it, “benchmarks and rankings can cause a race to the top because companies care more about scores than they do about controversies.” That said, there are substantial challenges to measuring tech sector impacts that impede the development of investor-friendly standards and metrics in the short term:
The need for qualitative metrics: Several of our interviewees cautioned that metrics measuring impacts to human rights, digital rights, and other social goods should not be overly quantitative since “box ticking” exercises are not effective in practice. VentureESG echoes this sentiment: “The problem with the current implementation [of ESG is that] we are trying to do ESG as we are doing financial metrics, i.e. almost exclusively numerically. This prevents us from capturing nuance. [...] In fact we need to go beyond box ticking and scoring” (35). It is an unfortunate reality that qualitative metrics are often much harder to define, apply, and validate in practice.
The breadth of tech products and impacts: Unlike other sectors for which human rights standards and metrics have more readily been developed (e.g. extractives, apparel, electronics), the “tech” sector encompasses a much wider set of products, services, and business models and, in turn, a wider set of potential adverse impacts. This is compounded by the fact that, to an increasing extent, companies in all sectors are employing tech in potentially harmful ways. The problem is, as one civil society actor succinctly put it, “When you have something as all-encompassing as the tech sector, how could you have standards that tackle issues comprehensively in a long-term investor perspective?” To this point, a few responsible investors told us that the Ranking Digital Rights Corporate Accountability Index is the best available resource for their purposes, but its utility is limited since it only ranks 14 companies and only covers a small subset of relevant issues.
The rapid pace of change: Another characteristic unique to the tech sector is its rapid rate of change, both in terms of the creation of frontier technologies and iterations on existing ones. As one ESG expert we spoke to said, “Tech changes so much, which makes standard creation more difficult. We need more dialogue with standard setting organizations to ensure they’re keeping on top of emerging technologies.”
To address the substantial challenge of reorienting ESG metrics to the digital space, one impact investor recommended pursuing two complementary initiatives: a “better than nothing” standard for digital rights in the short-term in addition to longer-term research into a more substantive way of measuring the human rights and other social impacts of the tech sector. For an example of the latter, a researcher at the University of California, Berkeley's Center for Long-Term Cybersecurity, Dr. Jordan Famularo, is currently convening working groups of experts on different digital rights issues to identify research questions and methods with the ultimate aim of developing ESG reporting standards for these issues. The work is supported by Omidyar Network.
We were advised by an ESG standards creator that standards geared towards sub-industries may be preferable to broad, top-down thematic standards. For example, rather than developing one set of standards or metrics for all tech companies, there is more value in creating separate standards for different types of technologies (e.g. artificial intelligence, social media platforms) and/or different impacts (e.g. harmful content, discrimination, privacy). This was echoed by a responsible investor we interviewed, who added that they would also benefit from materiality assessments, in recognition of the fact that “what’s material to Microsoft is not necessarily material to Amazon.”
There is thus a two-fold opportunity for funders: help bridge the immediate gap by supporting the development of resources investors can use to better evaluate and engage tech companies in the short run and contribute to a longer-term project of convening scholars, civil society, rights-holders, and industry to build effective, investor-friendly digital rights standards from the ground up.
Appendix A
Company
Human Rights Disclosures
Select Controversies
Meta
Alphabet
Australian court to fine Google over misleading Android users about data collection
Arrest of activist raises concerns over the privacy of Google users in India
Saudi dissidents accuse Google of bolstering dictatorship with cloud computing deal
US labor board alleges Google illegally spied on workers before firing them
US lawsuit alleges Google violated privacy law by tracking app users despite their opting out
Apple
Amazon
1. Florian Berg, Kornelia Fabisik and Zacharias Sautner, European Corporate Governance Institute - Finance Working Paper 708/2020, “Is History Repeating Itself? The (Un)Predictable Past of ESG Ratings” (August 24, 2021), p. 1.
2. International Finance Corporation, “Investing for Impact: The Global Impact Investing Market 2020” (July 2021), p. 1.
3. Examples include Sustainalytics, S&P Global CSA, Truevalue Labs/Factset, RepRisk, Vigeo-Eiris, MSCI, ISS ESG Ratings.
4. Dane M. Christensen, George Serafeim & Anywhere (Siko) Sikochi, The Accounting Review, “Why is Corporate Virtue in the Eye of the Beholder? The Case of ESG Ratings” (February 26, 2021), p. 10.
5. Hans Taparia, New York Times, “One of the Hottest Trends in the World of Investing Is a Sham” (September 29, 2022), https://www.nytimes.com/2022/09/29/opinion/esg-investing-responsibility.html.
6. For an overview of digital human rights issues, see United Nations Office of the High Commissioner for Human Rights, “Hub for Human Rights and Digital Technologies,” https://www.digitalhub.ohchr.org/.
7. Jordan Famularo, Medium, “Sustainability Reporting on Digital Harm: State of Play and Future Agenda” (July 21, 2022), https://medium.com/cltc-bulletin/sustainability-reporting-on-digital-harm-state-of-play-and-future-agenda-ca502a15a6f3.
8. Dragon Yongjun Tang, Jiali Yan and Yaqiong Yao, Northern Finance Association, “The Determinants of ESG Ratings: Rater Ownership Matters” (July 18, 2021), https://portal.northernfinanceassociation.org/viewp.php?n=2240017640, p. 1.
9. David Pred and Natalie Bugalski, Business and Human Rights Resource Centre, “Why ESG investing is bad for human rights - and what we can do about it” (March 21, 2022), https://www.business-humanrights.org/en/blog/why-esg-investing-is-bad-for-human-rights-what-we-can-do-about-it/; Taparia, op.cit.
10. Digital technologies are “used to suppress, limit and violate rights, for instance through surveillance, censorship, and online harassment. This is especially true for those who are already vulnerable or have been left behind, or those who are seeking to defend and promote human and civil rights. The digitalisation of our societies has, in many instances, eroded social protections, deepened inequalities, and exacerbated existing discrimination, in particular through the use of technologies such as facial recognition, robotics, digital identification and biotechnology. AI-enabled tools in particular can cause profound harm in the absence of fairness, accountability, explainability and transparency:” United Nations Office of the High Commissioner for Human Rights, “About,” https://www.digitalhub.ohchr.org/about.
11. Berg et al., “ESG Confusion,” op.cit., p. 12.
12. Pred and Bugalski, op.cit.
13. Billy Perrigo, Time, “Facebook Accused of ‘Whitewashing’ Long-Awaited Human Rights Report on India” (July 15, 2022), https://time.com/6197154/facebook-india-human-rights/.
14. Inclusive Development International and ALTSEAN-Burma, “Myanmar ESG Files: How ‘responsible investment’ is enabling a military dictatorship” (March 9, 2022).
15. Inclusive Development International and ALTSEAN-Burma, ibid.
16. Berg et al., “History”, op.cit;Christensen et al, op.cit.; Florian Berg, Julian F. Koelbel and Roberto Rigobon, Forthcoming Review of Finance, “Aggregate Confusion: The Diversion of ESG Ratings” (August 15, 2019); Florian Berg, Julian F. Kolbel, Anna Pavlova and Roberto Rigodon, Social Science Research Network, “ESG Confusion and Stock Returns: The Problem of Noise” (October 12, 2021);Aaron Chatterji, Rodolphe Durand, David Levine and Samuel Touboul, Berkeley Institute for Research on Labor
17. Gibson Brandon et al., op.cit., p. 12-13.
18. This is not to gainsay concerns about technology companies’ environmental impacts, including data centers’ demand for electricity and water resources. With that in mind, the CEOs of Microsoft, Nokia and two dozen other companies formed a European Green Digital Coalition in 2021 committing to become climate neutral or net-zero no later than 2040: European Commission, “Companies take action to support the green and digital transformation of the EU” (March 19, 2021).
19. Berg et al., “Aggregate Confusion,” op.cit., p. 2-3.
20. Berg et al., “Aggregate Confusion,” op.cit., p. 4.
21. Berg et al., “History,” op.cit., p. 3-4.
22. Berg et al., “History,” ibid., p. 3.
23. Fernanda Wenzel, Mongabay, “Behind the buzz of ESG investing, a focus on tech giants and no regulation” (30 April 2021), https://news.mongabay.com/2021/04/behind-the-buzz-of-esg-investing-a-focus-on-tech-giants-and-no-regulation/.
24. Tang et al., op.cit., p. 31-32.
25. The SEC has recently proposed rules and form amendments designed to create a consistent, comparable, and decision-useful regulatory framework for ESG advisory services and investment companies to inform and protect investors while facilitating further innovation in this evolving area of the asset management industry: Securities and Exchange Commission, Federal Register, Enhanced Disclosures by Certain Investment Advisers and Investment Companies About Environmental, Social and Governance Investment Practices: A Proposed Rule by the Securities and Exchange Commission on 06/17/22.
26. Initiatives in this category address ethical issues related to the development and use of algorithms, which encompasses artificial intelligence and machine learning technologies.
27. Initiatives in this category address issues related to the right to privacy.
28. Initiatives in this category address issues related to monitoring and moderating online content that violates pre-set rules or guidelines.
29. Initiatives in this category address how to assess the risk of granting individual prospective users access to a particular tech product or service.
30. Initiatives in this category address issues related to the accessibility of high speed internet.
31. Initiatives in this category address issues related to discrimination against and unequal treatment of persons on the basis of characteristics such as race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth, or other status.
32. Initiatives in this category address how international human rights law norms and standards apply to companies in the tech sector.
33. Initiatives in this category address issues related to the right to receive and impart information and ideas and the freedom to hold opinions without interference.
34. Initiatives in this category address issues related to curtailing the spread of false information online.
35. VentureESG, LinkedIn post in response to Economist article “ESG should be boiled down to one simple measure: emissions” (July 2021), https://www.linkedin.com/posts/venture-esg_esg-should-be-boiled-down-to-one-simple-measure-activity-6957230129395728384-4TfJ?utm_source=share&utm_medium=member_desktop (Last accessed September 30, 2022).
Image Sources
1. US SIF Foundation, "Report on US Sustainable and Impact Investing Trends 2020: Executive Summary", https://www.ussif.org/files/Trends%20Report%202020%20Executive%20Summary.pdf
2. OECD Business and Finance Outlook 2020: Sustainable and Resilient Finance, “Environmental, social and governance (ESG) investing” (2020), https://www.oecd-ilibrary.org/sites/e9ed300b-en/index.html?itemId=/content/component/e9ed300b-en
3. Ibid.
References and Further Reading