Aneesh Raghunandan
Description: Aneesh Raghunandan, Professor of accounting at the London School of Economics. His research focuses on corporate misconduct, issues in environmental social governance, ESG and auditing. In this episode we talk in depth about the emergence of ESG in financial institutions and its outcomes. Professor Raghunandan explains how ESG funds often underperform standard funds in carbon emissions and profit as well as maintain higher rates of corporate misconduct. We discuss the issues of verification, data manipulation, and transparency in ESG scores and how students can think of ESG in more nuanced ways.
Websites:
Publications:
Financial misconduct and employee mistreatment: evidence from wage theft
Government Subsidies and Corporate Misconduct
Do ESG Funds Make Stakeholder-Friendly Investments?
The Impact of Information Frictions Within Regulators
Related Articles:
Measuring ESG Business Claims v. Actual Behavior
ESG ratings and reports need to be standardized
The dubious appeal of ESG investing is for dupes only
Is There Real Virtue Behind the Business Roundtable’s Signaling?
Videos:
Sustainability Metrics and ESG Investments Guest Lecture
Show Notes:
[0:00:03] Introduction and Background of Professor Aneesh Raghunandan
[0:03:53] Oversight and Regulation of ESG Funds
[0:07:08] The challenge of regulating non-financial performance metrics
[0:14:43] Decentralization and its impact on coordination across agencies
[0:19:16] Understanding the different scopes of a company's carbon footprint and outsourcing
[0:22:03] Challenges in Measuring Scope 3 Emissions
[0:24:00] Tax Arbitrage and Outsourcing Carbon Emissions
[0:30:59] Challenges with Data Accessibility and Standardization
[0:35:55] Requiring Standardized Disclosure: Importance and Trustworthiness
[0:36:30] Enforcing Misreporting for Transparent Company Disclosures
[0:38:53] Challenges with Third-Party ESG Data and Measurement Error
[0:44:18] Market Fix vs Regulation: Carbon Emissions Debate
[0:50:10] Different Perspectives on Company Virtue and Values
[0:52:59] The Complexity of ESG Ratings and Agency Weighting
[0:56:08] Tackling Specific Issues and Disaggregation in ESG Discussions
Unedited AI Generated Transcript
Brent:
[0:03] Welcome, Professor Aneesh Raghunandan. Thank you for joining us.
Aneesh:
[0:06] Yeah, thanks for having me. It's fun to talk about research.
Keller:
[0:09] We'd love to start off by hearing a little bit more about your story.
How'd you got to the London School of Economics?
Aneesh:
[0:14] Yeah, so I mean, I came kind of, you know, I grew up in the States, you can probably tell from my accent.
I studied in the US for my undergrad and my PhD, but I thought it'd be fun to maybe come abroad for a bit.
I never studied abroad in college, and so I figured better late than never.
And so when I looked for academic jobs after grad school, of course I looked in the States, but I looked here too.
And then when this offer came through, I figured it's a good school, it's a great city, I've never lived abroad, why not?
And it's a good department, it's one where I can sort of really have time to do the research I want, how I want it, when I want it, and with the resources and most importantly, the timeavailable to work on it. And so I figured it was a great place to be.
Brent:
[0:55] Yeah, and then before we get jump into what you actually do.
Could you give us a brief explanation of what is ESG and then how did it come about?
Aneesh:
[1:05] Sure. So if you ask any two different people, they're gonna have a different definition of ESG ESG, you know the acronym of course refers to companies environmental social andcorporate governance related actions, And you could sort of imagine what those would be.
The environmental stuff is pretty straightforward as to what it captures and what it doesn't.
The social stuff is where I think we run into a lot of issues where you can sort of call any behavior a company engages in social, whether or not it's profit motivated or whether it'saltruistic.
And then the governance side is really looking at the oversight of these two things and the company's financial performance.
But of course, ESG just refers to, it's a way to almost characterize the actions a company takes in a way that frames these actions around how they affect certain stakeholders, and inparticular, the way the conversation has gone the last few years, non -financial stakeholders, right?
So those who are not investors in or lenders to the company.
Brent:
[2:07] Perfect.
Keller:
[2:07] And can any fund be labeled an ESG fund?
Aneesh:
[2:11] Um, yes, that's actually skipping a couple steps ahead. So basically for context, I have some, I've done some work on ESG funds, which are basically investment products.
I think like mutual funds, exchange traded funds that essentially pool people's money and claim that they're going to use that money to buy a stock portfolio of high ESG or maybeinformally ethical or socially responsible companies.
[2:36] Typically, up until a couple of years ago, there was almost no oversight into whether the way funds marketed themselves actually matched the investment decisions that they made.
And in fact, one of my papers that we might talk about more in the next few minutes or next to the course of today, shows that these funds were taking people's money, charging higherfees on the grounds of promising to do the research on behalf of their clients for who was socially responsible and who wasn't, but then actually just turning around and buying stocks thathad a higher rate of labor violations, a higher rate of EPA and OCEA sanctions, right?
Essentially ones who seem to be treating these stakeholders in the firm worse.
Now, of course, in the past couple of years, for the US audience, the Securities and Exchange Commissioner, the SEC, has started to try to crack down on what you might informally callmisleading marketing, right?
So in that sense, to come back to your big picture question, can any investment mutual fund, for example, call itself ESG -oriented?
[3:42] Five years ago, I would have said probably yes. Now, it's too early to say for sure how much of a crackdown there'll be and how stringently these new laws on the books will beenforced.
Oversight and Regulation of ESG Funds
[3:53] But in theory, now the regulators can come after you for misleading marketing to say, so if say, for example, let me invest your money for you and I will pick socially responsible companies, and you don't, for example, provide a definition of what socially responsible means, if you don't make it clear to the ultimate people whose money it is, how you define socially responsible, what that means, and how you're going to make investment decisions in line with what you say you're going to do, then you can actually be sanctioned.
So we're shifting towards a world in the U .S. where the answer is no.
And that's especially true actually in Europe, right? So Europe actually has much more...
Detailed and kind of advanced regulations on the subject, where without going into all the details, there's actually some regulation around what can and cannot be in the name of aninvestment product.
[4:44] And of course, when you think about how people actually make decisions, we tend to first log on to our brokerage account or bank account and try to look for investment productsthat do what we want.
If you want companies that are small and growing, you would look for a growth fund. If you want companies that are socially responsible, you might look for ESG fund or sociallyresponsible fund.
Right, so of course, what's in a name is hugely important for what people see, what people find, and ultimately what people invest in and buy.
And so in Europe, we've got now much more, you know, kind of understanding that this is how people make decisions, right?
How people like you and I, who may not work in the finance sector, but you know, have a little bit of just our savings that we want to invest.
You know, the EU actually has much stricter standards on what can and cannot be in a name. So the answer in Europe would be largely no, right?
But the answer in the US is, well, historically, yes.
Now it's trending towards no, but it's too soon to say what that'll be, right?
And of course, there's a bigger picture thing here, which we'll get into at some point maybe, which is the issue of greenwashing, which is how do you figure out a fund that markets itselfas being ESG focused, says the right things in its marketing materials, right, so it goes one step further beyond the name, but then quietly doesn't do what it says, right? And that's, that canbe harder to identify, right?
Brent:
[6:09] Yeah, because that's what I was going to ask is whether or not I'm an ESG fund, I say, I define social governance as this, who's to say like they are actually following that?
Like how, is there a certification body? Is there, does the SEC define, all right, these are the metrics we look at for social governance or environmental, or whatever else it may be.
Like how do they certify? Do we take the firm's word?
Aneesh:
[6:37] Yeah, so this is also where things get really tricky, right?
So a good analogy here would be how would you hold firms accountable for financial claims, you know, in the past, versus this new world of ESG claims, right?
And typically, in the past, it was pretty easy to figure out if a firm was doing what it said it did, right?
Because you had kind of one universal metric, which we could agree on, dollars, to measure performance.
And of course, you could argue whether or not we should be living in such a singularly focused setting, but that's the reality of how things work.
The challenge of regulating non-financial performance metrics
[7:09] And so you could easily regulate, OK, if this thing, this affects profits, this affects revenues, you have to be open about how you generate it.
But in this new world where we're taking into account non -financial performance, it's a bit trickier because what's material to one firm may not be material to another, right?
And what really can make the affected stakeholders of a firm's actions better off is going to differ across industries or maybe even within industries, but as a function of how denselypopulated the local area is, right?
There's a lot of things where what matters is very contextual, right?
So which leads itself to what I sometimes like to call a dimensionality problem, right? We have all these metrics, but it's not clear that the same metric matters the same across firms, right?
So it's an issue of, well, if you require a certain quantity to be disclosed, right?
If you say, okay, companies have to say how much they paid in bonuses, for example, to employees, to managers, right? So we can get a sense of pay gaps within a company.
That might be really important in places like retail or places where there's a lot of employees who maybe don't have a lot of bargaining power.
It's probably less important in financial services, right?
So it doesn't really matter if Bank of America is disclosing that given who works there and how they get paid and how much they get paid.
[8:32] Or you think about even things like carbon emissions disclosures.
These are important and this is something that the SEC is actually starting to ask for.
And you can imagine it being hugely important for, if you want to, for example, figure out whether an oil company that claimed to be shifting its business model to be decarbonizing whatwas actually doing this, right?
It's less clear that we need to know Facebook's carbon emissions, right?
In that they're a tech company, they don't really produce tangible goods, right?
And so sure, you know, maybe it's not bad to have the information, but it's maybe we want to know other things about Facebook, and it's more important that they put out these otherthings.
Now, of course, the trade -off here is that if you don't standardize anything, if companies have complete discretion over the metrics they pick, they're gonna cherry pick the ones that makethem look the best.
[9:23] And so this is kind of the struggle in figuring out how to regulate the space in the sense that different things matter and different things are gonna be more impactful on thecommunities and society that firms do, right? And so what do you, you know, what do you kind of mandate?
And this is where you're starting to see guidelines, not regulations, but kind of guidelines that are not necessarily enforceable yet on, okay, if you're in this industry, this is what's material.
If you're in this industry, this is what's material. This is something that there's a body called the Sustainability Accounting Standards Board.
And they've spent a lot of time thinking about this exact thing, right? What matters for who?
And of course, there's mixed evidence on whether companies that.
[10:07] Disclose information in line with the SASB, Sustainability Economic Standards Board, guidance actually do better, or whether doing well by doing good is really a thing.
So you also have to be careful about how you think about the problem, right? Is it one where you care about the financial impact?
Is it a financial risk issue or is it a morality issue?
Neither is necessarily wrong, but it does frame what you should require.
Sorry, there's another thing here as well, which is that we've started to attract a lot of financial professionals, a lot of investors to this space.
[10:46] And a lot of times we often think of, well, if investors are involved, we need the investor protection units, the SEC, for example, in the US, to be the ones who are regulating thestuff.
But in a lot of cases, a lot of the data, a lot of the meaningful metrics that we actually want, they already exist. they just are under the purview of a different regulator, right?
So for example, let's say you invest on behalf of a union pension fund, right?
If you're a union pension fund, you probably care a lot about investing in companies that treat their workers well.
And so you probably care about, for example, workplace safety.
You probably care about who pays a fair wage, right?
But you can actually figure out a lot of these things and learn a lot about these things, not through something the SEC has mandated, or will mandate, or won't mandate, and not evennecessarily through companies', voluntary ESG disclosures, right, these sustainability reports.
But if you just look at, for example, OSHA case statistics, the Occupational Safety and Health Administration, If you look at how many workplace injuries does a company have, howmany workplace safety violations does a company have, how frequently do they get inspected, and of those inspections, how often do they get a clean bill of health?
Or if you look at another branch of the Department of Labor, the wage and hour division, you look at, okay, who seems to be getting in trouble for not paying overtime, for forcingemployees to work through their breaks, something that we call wage theft.
Brent:
[12:12] Yeah.
Aneesh:
[12:13] And so you don't need any new regulation to get this stuff because it's out there.
But you have to think beyond the boundaries of, oh, the SEC should be the one providing it to me.
But there is sort of standardized, well -collected data by an agency that really knows the institutional details, right?
And so why aren't we relying on that stuff more than we already are?
That's kind of an open question and one where, to be honest, I don't know the answer because it surprises me.
Keller:
[12:41] Yeah. So you feel like part of the responsibility falls on the consumer, the investor to do their own due diligence on a topic as opposed to relying on? I think that, well, I think that'spart of it.
Aneesh:
[12:53] But I think what would also be helpful is, you know, a broader shift in the conversation of, okay, if we care about these things, rather than immediately trying to say someone giveme a score or a rating, right?
Think about, okay, what are the most important measurable outcomes we can see, right?
And so this is gonna be different based on what you care about.
If you care a lot about sustainable investing, and if you care about greenness, right?
We have some data on, for example, I'm gonna use largely US examples, just given kind of the audience and given some of my own work, right?
But if, of course, we would like to know more about carbon emissions, and it's a noble goal to get more, but there's already a lot of data out there on, for example, toxic releases, oil spills,and stuff that the EPA already puts out that you could, if you were interested, link up to the ultimate parent companies.
[13:38] And so you might say, well, OK, rather than just using some sort of third -party score, why aren't we looking at this data?
Why aren't we calling out companies that have a lot of oil spills?
Why aren't we going on, if you say you're an investor who cares about this stuff, why aren't you going on the earnings call and asking about a safety incident?
We see this sometimes, but not as much as the amount of, for example, questioning and talk you see about this in context where there's money on the line doesn't match essentially theamount of just straight up talk.
So, you know, it gives you a sense that, yeah, if we just talked about this stuff more, but more importantly, in settings where you have management under pressure, right, maybe youactually start to focus the conversation on stuff that's already out there that already captures some of, not all of, of course, but some of what we're, what we say we're trying to assess onmore rigorously.
Brent:
[14:31] And then do you think the decentralized nature of the government is making it harder to properly access all the different companies and whether or not they're performing orholding themselves to the standards they say they are?
Decentralization and its impact on coordination across agencies
Aneesh:
[14:43] Yeah. So I hesitate to say decentralization is uniformly bad, right?
There's of course, there's a long academic literature on this in accounting finance economics on essentially costs and benefits of decentralization and how much information is containedwithin in a branch versus across branches and what have you, right?
The one thing I will say is that decentralization may make it harder to do two things, right?
One is of course, to coordinate across agencies.
And this is something that I think, based on some of my own work, matters because we do know that companies often respond to financial incentives, by taking essentially what you mightcall bad or.
[15:26] You know, maybe not socially responsible, non -financial activities, right?
So there's work out there, you know, by myself, by others, showing that companies under short -term earnings pressure often respond by skimping on workplace safety costs or employeewage costs, right?
Or, you know, there's, I have other work that looks at decentralization within a single entity, and in this case, OSHA, and finds that, so for a bit of background, OSHA is organized into 50state -level offices, And the offices can talk to each other, but often don't.
And so what happens is each branch has a lot of really good local knowledge.
And it's hard to quantify, but that probably makes them better at their job locally.
So California OSHA probably benefits from having a lot of people who know California industry and California regulations and California standards, and also knowing just locally whoare the most likely to be up to no good.
The problem is if you have a big company that operates in California and Texas, right?
If California OSHA doesn't tell Texas OSHA, hey, look out for this guy.
[16:32] The firm might essentially engage in something we call regulatory arbitrage.
To say, if you get in trouble in California, then I'm gonna clean up my act there, but I'll just do something bad in Texas instead.
And it's not that you go into this saying, I'm gonna do something bad.
It's more just, well, if you have a fixed budget for safety, a fixed budget for some of these things, right?
You just shift around the money, but don't actually invest into fixing the problem, right? And that's, again, I hesitate to say that this means that decentralization doesn't work, but it's just acost of decentralization, right?
That's worth thinking about.
And of course, if you had more data, if you were OSHA, for example, you might think, well, okay, this is a cost, we have to account for this.
Is it still worth bearing this cost because of the benefits of decentralization?
Of course, that's well beyond the scope of what we might talk about today, right?
But a lot of what some of us in this space try to do is just try to understand maybe not so much on net is something good or bad, but trying to better identify what are some costs we hadn'tthought of? What are some benefits we didn't realize?
And how are, for example, corporate financial incentives related to corporate non -financial behavior?
And that actually brings me back to a question you sort of asked super early on, which is how did you get into this stuff?
And I got into this whole ESG research space kind of by accident, because I started from the point of being interested in how do companies' financial incentives affect their non -financialbehavior.
[17:58] And then you start to think about, well, okay, so why are they doing this stuff?
Is there, is it purely financially motivated?
Why do some companies, okay, all companies face earnings pressure from time to time, or all companies face short -term investor pressure.
So why do some companies react by, you know, to put it bluntly, screwing over their employees and others resist the temptation?
And so that got me more into the space of, okay, so on average, the financial incentives seem to shift behavior, but that's an average effect.
But within that average effect, not everyone is the same, right?
And what is it that could lead companies to behave differently?
And of course, one thing leads to another, and you start to think about more of the social, the non -financial incentives, right?
And that's where we start talking about ESG and social responsibility, at least traditionally, and how that might feed in to how, when, and why companies choose to treat their employeeswell or not.
So I really came into this from a perspective of understanding employee treatment, and then of course, you start to think about other dimensions as well.
Brent:
[18:58] Yeah. And then, when you talked about regulatory arbitrage, are companies, or say a parent company, able to outsource their carbon footprint, or regulatory issues to subsidiaries ordifferent states?
And how does that process work?
Understanding the different scopes of a company's carbon footprint and outsourcing
Aneesh:
[19:16] Yeah, so that's something that works in, I guess, different ways depending on who the company is, right?
So, when we talk about, so this is where it might be helpful to bring in some jargon.
So without getting into too much detail, there's essentially three scope.
A company's carbon emissions footprint is divided into three scopes.
[19:36] There's scope one, which is essentially direct emissions from production.
So I run a factory, the factory spews some stuff into the sky, right? The stuff streaming out of that factory is my scope one.
Scope two is essentially emissions that were generated to provide me with like energy, steam, water, heating and cooling, right?
So I buy energy from the electric company, right?
That energy, the electric company had to presumably to burn something or generate some emissions to create that energy, right?
And so the amount of emissions that were caused, quote unquote, by me needing to use that energy would be my scope two.
And then there's something called scope three, which is a catch -all term for everything else.
And this is super hard to define, but when you talk about outsourcing, this is where you start, you might see a shift, right?
So scope three reflects essentially emissions generated both up and down in the supply chain, right?
So there's, you know, one part of scope three involves something called downstream, where I sell you a car, you drive the car, the car takes gas.
[20:39] And so how much you drive the car is going to affect my downstream emissions, because I've essentially enabled you to pollute.
And there's also upstream, which refers to the supply chain.
And this is where, when we talk about outsourcing, this is where it's really going to be more relevant.
So we say, OK, so now I have production. you know, obviously, unless I'm completely vertically integrated, right, I'm going to be buying some raw materials and be buying some partiallyfinished goods, some inventory from suppliers.
[21:10] And those suppliers also generated emissions in producing the whatever it is I'm buying from them.
And so the emissions generated up, all the way up the supply chain up until I buy this, the unfinished products are going to be my scope three, right.
And so you can imagine if If I shift to outsourcing, I'm going to have lower scope 1, higher scope 3.
This, of course, says nothing about some of the other issues we face, which is that sometimes carbon is treated differently.
So carbon is mainly regulated on the grounds of scope one and two.
So there's a bit of arbitrage here.
And this is somewhat understandable, in the sense that a company knows what's going on inside its own factory.
And a company knows how much energy it uses.
And so scope one and scope two are pretty easy to, if not get perfectly through carbon sensors, at least come up with a pretty good approximation of, right?
Challenges in Measuring Scope 3 Emissions
[22:03] So we say that scope one and two are generally pretty measurable.
Scope three is much harder to measure because in order for me to measure scope three, if you guys are my suppliers, I need to know both what's happening in my factory, but also what'shappening in your factory, right?
And I don't have a legal right to force you to tell me that.
And so scope three is often subject to a lot of estimation issues, right?
And this isn't anyone doing anything wrong, it's just companies trying their best with very limited information.
[22:31] And so scope three is often, for example, if you think about carbon taxes and carbon caps and what have you, scope one and scope two are often subject to carbon regulation. Scopethree almost never is.
And even if you think about, for example, disclosure regulations, the UK has had a carbon disclosure mandate for large companies since 2013.
But again, that's only scope one and scope two.
The US, in the US, the SEC is talking about it, and it's maybe it'll pass, maybe it won't.
But even in that case, there will be legal penalties for, if should it pass as it's currently been written, there's gonna be legal penalties for misreporting or not reporting scope one and scopetwo, but not for scope three.
They've created what's called a safe harbor provision on the grounds that it's really hard to estimate, right? So we shouldn't penalize companies for trying.
And so to come back, this is kind of a meandering path and I'm sorry, but to come back to your path of, to your question of regulatory arbitrage.
Well, if scope one is subject to scrutiny and regulation and scope three isn't, right, you can imagine outsourcing has the benefit of shifting scope one into scope three.
So even if you're total, you know, you sell the same total amount of stuff, right?
You sell and, you know, the same stuff is produced and the production process is the same.
You've shifted stuff from scope one to scope three. And in fact, you might've increased emissions because included in scope three is the cost of transportation, right?
And if I have to move a bunch of stuff from a bunch of suppliers, right, that might generate more emissions than just doing stuff in -house.
Tax Arbitrage and Outsourcing Carbon Emissions
[24:00] But if the scope one number goes down, I might actually look good.
The other thing of course is that, and this of course also helps you, for example, if there's a scope one carbon tax, by outsourcing something and having someone else do it, you can avoidthat tax.
So there's a bit of a tax arbitrage angle as well.
And there's also, so pretty much any, unless and until scope three gets regulated, and this is something where, in theory it's great, in practice it's one of these things where, we're not closeto that. There will be these opportunities for what you might call outsourcing carbon emissions in order to make your footprint or your compliance with the regulations look better.
Keller:
[24:45] Yeah. And would emissions from tech firms, would that fall under scope three?
Aneesh:
[24:51] What you would probably see is a lot of scope two, right?
So you need, you know, tech firms are often, you know, running a lot of servers, right? They have a lot of cooling fans and stuff.
So tech firms' main emissions footprint falls in the energy they consume, right?
Scope three for tech firms is generally pretty low because there's not, well, it depends, okay, this depends.
If you're Apple, then of course you have your hardware and that's gonna have a non -trivial scope three component.
If you are Google, okay, so if you're traditional Google, so ignoring the pixel and some of the other stuff it's gotten into, but if you're like the software side, right, you're not gonna havemuch of a supply chain, right?
You just have engineers and code and open source stuff on GitHub doesn't really have an admissions footprint.
You could maybe argue that there's a downstream angle, but even that's harder, right? You'd have to prove that, okay, you're using your phone and charging your phone 40 minutes moreper week because of my app.
And so in some sense, in theory, you could imagine it being there in practice, you won't see this reported because how do you get that level of specificity, right?
[25:57] If you're not using my app, right, if you use some other app, that's still using electricity, right? I'd have to prove that your phone would be off if you weren't using my product, right?And that's, how am I gonna do that?
But so for tech companies, at least for software companies, we'll see a lot of scope two.
Scope one, of course, you'll see a little bit, right, just in the, you know, if they're making some hardware, right?
So most, not all, but a lot of software companies, big software companies, right, the ones you and I can easily buy shares in, do have some hardware production now, right?
It could be chips, could be phones, could be tablets, whatever.
And so you'll see a bit of scope one there, but scope two is probably the biggest one for a lot of tech companies. And this actually is very different from if you take manufacturing, right?
Scope one is way larger than scope two, right? And this is, again, we come back to the issue of what matters for whom, right?
If you care about a tech company's emissions footprint, right, you probably care about scope two more and you should be looking at how much energy they're using.
Right? If you look at, for example, a crypto mining shop, right?
Scope two is where it's gonna be, right? But if you look at a manufacturing firm scope one scope one is a lot more important Yeah, right and then is there a scope one two or three?
Brent:
[27:09] Equivalent for a lot of the like social aspects and what I'm thinking of right now is a lot of the, like Mining aspects that Apple likes like in the Apple supply chain that came up intolight recently and like yeah So that's something where it's not.
Aneesh:
[27:29] It's not codified as neatly, right? So this is something where, if we talk about ESG, a lot of our discussion on emissions, or that whole discussion has kind of been talking about the Epart of it.
When we start to talk about what you might call an ethical supply chain, which is I think where you're going with this, that would fall more under the S.
And of course, you can think about people tend to assess.
[27:52] One of the problems with assessing S relative to E is that there's so many different things you can care about, right? Do you care about the company's direct workforce?
Do you care about what's going on in the supply chain?
Do you care about things like product safety and how consumers benefit?
Do you care about, if you're Facebook, right? Do you care about the amount of disinformation zone?
Right, so there's a lot of different things you can talk about, but there's not really one, there's not really one, I guess, what you might call quantitative metric of how good you are to yoursupply chain.
Of course, what we have in lieu of this are more qualitative certifications, right? to think about fair trade, right?
Which, of course, it's a very specific setting, very specific set of goods, right?
Coffee, chocolate, right? It's not expensive, but it is for companies in that sector, right, certification that they have an ethical supply chain.
And you see similar initiatives in other industries as well.
I think fair trade is probably the most well -known, but I don't think, there's not really one sort of uniform, yes, you are good in your supply chain, right?
The closest we get to that is, you know, sometimes you get these ESG ratings that are put out by commercial vendors that try to assess a company's supply chain performance.
But I'm often deeply cynical of those in that they don't really do their own primary research for the most part.
[29:08] So I would say supply chain is important, but there's not really...
[29:13] So unlike the East, the environmental space, and unlike some aspects of the social space, so for example, it's pretty, maybe not easy, but there are ways to, for example, assess how acompany treats its own employees, right?
We have in the US, we have workplace safety issues. We have a wage and hour issues.
We're starting to see, you know, some other quantities being disclosed.
And of course, no, every metric has its flaws, right? So there's one that's called the CEO pay ratio disclosure, which companies have to put, public companies have to put in their financialstatements that says essentially what is the ratio of the CEO's total pay to the median employee's total pay?
And of course there's only so much you can get from that, right, in that the median employee doesn't tell us a lot about the lowest paid and the most vulnerable, right?
Or in the UK, for example, we have this kind of flagship initiative called, you know, that essentially requires companies to compute something called the gender pay gap and disclose it,right?
But there, again, there's issues in that that metric doesn't account for, for example, differences in seniority.
[30:13] So you can sort of tear into any one metric. But you do have, at least in some aspects of S, there are quantitative metrics.
And you say, well, OK, no one metric is perfect.
But maybe if I take these all together, I can get a sense of a company's labor footprint.
That is harder. You can't really quantify that in the supply chain.
And that's kind of the hardest thing with understanding, yeah, this Apple, you know, Apple needs a lot of minerals, right?
What are the knock -on effects of this? And more importantly, how do we quantify the knock -on effects of this, right?
[30:44] And that, you know, it's one of the hard, and this is where, you know, big picture, one of the hardest things about really holding companies to task is, you know, you can't do that ifyou can't measure what they're up to, right?
Anecdotes are not, anecdotes are great, but anecdotes are not a substitute for sort of systematic data.
Challenges with Data Accessibility and Standardization
Keller:
[31:00] Yeah. And could you talk a little bit more about that aspect of the data and how it is hard to see what really is going on and who has access to the data that these companies areproducing?
Aneesh:
[31:10] Yeah, so there's a couple of different issues here. One is, of course, there's been a big push for, as we talked about earlier, given that different things matter to different companies,there's been a big push for companies to, for example, put out quantitative data in their sustainability reports or to put out quantitative data on their websites on issues that are material tothem, but also leaving it up to the companies to decide on what's material, right?
And so that data, there's two issues. First is it's difficult to audit, right?
Because if you're an auditor, if every company is doing something a little differently, right, how do you, how do you really know?
You don't, you don't develop the expertise you need to say, hey, this, this looks fishy, right?
And the other thing is it's very difficult to, even if the data is public, to compare.
Right, to say, okay, you both say that you have, you know, you have, you both say that you're a good high paying employee.
Oh, but you excluded this many employees. Oh, but you have more contract workers.
Right. And you won't tell me what's going on there. How do I know which one of you is better? Right.
Keller:
[32:09] I just, it's hard to figure out.
Aneesh:
[32:11] And so, you know, without really standardized data, right, you start to get into these issues of, there's a bunch of, there's a lot of self -selection issues and what companies choose toreveal, and importantly, how they choose to calculate it.
So the same statistic on two companies' websites may not mean the same thing.
And so one way to get around this, and it's something that we're seeing increasingly in Europe and to a lesser extent in the US, has been to force companies to report a standardizeddisclosure.
And that, I think, is something that can be more promising, subject to the caveats that, well, again.
[32:52] Any metric will have its limitations, and the same thing may not be as important for one versus another.
But one issue that we faced there, and I've seen this firsthand in the context of studying the aforementioned gender pay gap reporting in the UK, is that, well, you need enforcement.
So I have work showing that in the UK, for context, companies with more than 250 employees—and And this is for -profit companies, charities, universities, government bodies, youname it, right?
So any entity with more than 250 employees, and there's like 10 ,000 of them in the UK alone, and probably way more in the States.
[33:29] Have to reveal, have to produce and disclose publicly some statistics about essentially how much more does the median male, or how much more or less I should say, does themedian male employee in the company get paid relative to the median female?
And of course there's some other stats on bonuses and high -paid employees and low -paid employees you have to disclose, but the gist of it is trying to compare how much does theaverage male get versus the average female, right?
And then trying to put this out there in a standardized form for everyone, then trying to understand, well, what drives this?
The problem is there's almost no evidence that the regulatory body in charge of overseeing this data has bothered to fact -check anything, right?
So one in eight of the disclosures online is mathematically impossible.
And I won't get into all the details of that, right? But basically the fact that companies disclose multiple different figures means that you can reconcile them against each other, right?
I'll give you a super simple analog from the accounting setting, which is, let's say you had a company that said, earnings are positive, but revenues are less than costs, right?
I couldn't tell you which one of those is wrong, right?
Brent:
[34:36] It could be any of the three, but something is obviously wrong.
Aneesh:
[34:38] And we have something similar going on in the setting where I have no private access to the data, but I can just look at this and say, hey, look.
I can't tell you which one is wrong, but one of these numbers is wrong.
And this is just sitting there on the website.
[34:53] And it's a part of what makes me a little cynical about this is that I even, I had a chance to give a service next to a witness in parliament, and I brought this up, and nine monthslater, nothing has happened.
[35:05] So even given the audience that would have been in a position to do something about this, right?
You see, And to be fair, there are likely a lot of resource constraints and enforcement takes time.
And it's, you know, the UK legal system is a bit different from the US, right?
The US tends to be much more fine happy, right, than the UK, which is a much more slap on the wrist culture in terms of enforcement, right? So maybe the slaps on the wrist havehappened.
We haven't seen them and nothing changed. I wouldn't, that, I can't see that level of detail, right?
So it's, there's, it's possible and being sadly unfair to the regulator here.
But at the same time, the bottom line is that if one in eight of these numbers is obviously wrong, what has requiring standardized disclosure, and this stuff costs companies in terms of timeand effort to produce, what have we actually done if the data can't be trusted?
Requiring Standardized Disclosure: Importance and Trustworthiness
[35:55] So the next step to me is really saying, okay, so we want to move to a world of having at least agreeing on maybe a handful of figures that are important enough for, if not everyone,at least a large enough set of firms that maybe we should agree that firms should disclose.
At least maybe big firms, firms for whom it's not prohibitively costly to do so.
So, I'm not saying that the little takeaway restaurant down the street needs to put out a gender pay gap report, but probably a company that's publicly traded on a stock exchange can affordto do that.
Enforcing Misreporting for Transparent Company Disclosures
[36:30] But the next step is saying, okay, well, if we say this is the solution to cherry picking and to companies systematically trying to make themselves look good, then we have toactually make it the solution.
The way to do that is to enforce misreporting, right?
Because if you're not enforcing misreporting, you're basically back to a world of companies releasing whatever they think makes them look best, as opposed to a world where companiesrelease truthfully and we decide for ourselves.
So it makes it hard on the consumers of this data, whether it's employees who want to figure out who care about the fairness of their potential employers, or it's investors who maybeactually want to put their money where their mouth is in terms of whatever social responsibility statements they make.
[37:18] Whoever the audience is, right? And however they make decisions, whether or not they use this information. Of course, they don't need to, right?
But somebody who wants to use this data, right, should be able to trust that it's reliable, right?
And that's sort of the step that's still missing, even in settings where we've been able to pass legislation to force disclosure, right?
And so this is something that I think would also be interesting and something that's a good lesson for the U .S.
Setting as well, right? In the U .S. you don't really have too many of these types of disclosures.
You've got a few, right? But of course, even then, you have something called the EEO -1.
It goes to the Equal Employment Opportunity Commission, or EEOC, right?
But that data actually, companies have to send it to the EEOC, but they don't have to post it publicly, right?
Brent:
[38:06] Interesting.
Aneesh:
[38:07] But again, we don't know if that data is audited. And of course, there's not a great way to... I don't know if you could reverse engineer it in the same way when companies misreport.
But if the EEOC wants to take the next step, and it's to eventually require that to be public to say, hey, look, you want to go find out which firms are committed to diversity, which firms arecommitted to pay equity, which firms just pay better, full stop.
You want to be able to trust that data to be able to say, hey, look, to use it.
And that's where, if this actually comes into play, if we see a larger role for these forms and disclosures companies have to file, we also need to make sure that they're trustworthy.
And that's kind of the next step in getting ESG data to a better place.
Challenges with Third-Party ESG Data and Measurement Error
Brent:
[38:54] Yeah. And then, does your cynicism towards...
Authenticating the published data extend to the third party ESG regulators or some of those boards?
And can you maybe expand a bit on who those people are and what is influencing how they go about collecting that data?
Aneesh:
[39:15] Sure. Yeah. So one thing that I think is interesting and probably consistent is that there's a huge demand for ESG data, right? And it could be big picture, just aggregate ratings ofsocial performance, or it could be specific measures like carbon emissions.
So I'm a green conscious investor, I want to know your carbon footprint, and you won't tell me. What do I do?
And the answer that such investors and other practitioners have turned to is to say, okay, well, I care about this, or at least I want to say that I care about it, but you won't give me data.
I have no legal way to compel you to give me data, so I'm stuck.
What do we do?" Essentially, the demand for information greatly exceeds the supply of information.
This is especially true in the US, where we don't have a lot of disclosure mandates to force the revelation of the information.
What we see is a bunch of information intermediaries, these for -profit companies popping up to try to bridge the gap between demand and supply.
Now, what I don't necessarily want to do is call into account the motives of these companies, right?
[40:26] Especially some of these companies that mainly try to estimate emissions, I think they believe they're doing the right thing.
And I think on average, probably we're better off when people are trying to produce new information.
But with that said, we see that, and I'll cite one example in the context of carbon emissions here.
When estimated data, when essentially data that is estimated by these intermediaries, because companies aren't disclosing, right, when this data doesn't perfectly kind of map on to, orsorry, how do I say this?
When the patterns in this data don't necessarily match the aggregate patterns in sort of the actual data for companies that do disclose and that do reveal information, we start to get intosome issue, we start to induce measurement error.
And in the context, I'll give you an example that I think is particularly problematic in the context of carbon emissions.
So I talked about how there's a lot more demand for corporate emissions data than there is supply of emissions information.
[41:33] So you have vendors, and I won't name any specifically here, that try to estimate this carbon emissions footprint for each company that won't reveal its own.
Now, in theory, this is good. But in practice, what happens if a company is not revealing its information?
Well, it's harder to say, okay, well, based on your business model, I think you two are in the same industry, but I think you're more efficient than he is.
If neither of you is going to tell me that much about your production process, how can I say that?
And so what we see is a lot of reliance on industry averages.
So what we see is that estimated figures tend to bunch much more around industry averages, Whereas if you look at figures that firms actually disclose, right, they're more all over theplace.
You can really identify who are the best and worst within an industry, right, in terms of who's more efficient.
So you and I are both auto manufacturers, but I've figured out how to use much less energy, right?
Or I figured out how to make my production process greener. Right?
If we both disclose our emissions figures, that will be clear.
My efficiency will be clear, right? But if we don't disclose, then all that the data vendor knows is that we make cars.
They might know something about the size of cars, the make and model, and what have you.
But they're probably going to estimate pretty similar emissions figures for us.
[42:53] And more importantly, they're going to estimate those emissions figures as a function of, essentially, financial performance. How many cars do I make? How much do I sell?
How much inventory do I take in? And so what you see is when an estimate of a non -financial metric tends to rely on financial metrics, then what you have is a close correlation betweenthe underlying financial metrics and this estimated non -financial metric.
And this is an issue because, let's say that you ask, do investors care about carbon emissions?
We know for sure that investors care about a company's financial performance.
[43:30] And so if we use the real emissions data, which is based on companies actually measuring their output, you actually find that there's no correlation between, for example, stockreturns and a company's carbon footprint.
If you use the estimated data, which is largely a function of financial metrics, so the estimate is basically just a transformed financial figure, you do see a correlation.
And this is problematic, because if you want to say, well, do investors care?
Right, is the market already doing something about climate change?
Well, you're going to draw different conclusions if you use actual data versus if you use estimated data.
And this is driven a lot not by any nefarious incentives, but just by the limitations that vendors have in trying to figure out what a company's carbon footprint is.
Market Fix vs Regulation: Carbon Emissions Debate
[44:18] And so you have a lot of these issues where it may matter more in some settings than others.
But especially in the case of carbon emissions, It's really important in the sense that, especially in the US, we tend to say, well, here's a problem.
Can the market fix it? Great. If so, leave it alone. If and only if the market cannot fix it, if we say it's a market inefficiency, then maybe we should regulate it.
And of course, you can argue whether that's the right way to view the world, but the reality of the US is that's roughly how regulation in the US works.
It's like, first, let the private sector fix it. If they can't, then we'll talk about regulating it.
And so if we think the market is already punishing high emissions firms, why do we need to put in regulation?
[45:02] But if it turns out that we think the market's doing something and it isn't, that's where we start to get these actual potential negative regulatory implications.
Because we might say, well, if the market is doing something about climate change, great, let it be.
But suppose that we think the market's doing something and it's not.
Then we're accidentally allowing a lot more pollution because we're not trying to regulate the problem.
So it's a case of we need to know the right answer to figure out what we should actually do. If there's a problem, you say, OK, what are companies doing?
[45:32] Is there a private sector solution? Is it working?
If it's not working, is there a way we can regulate it?
But to get to even that second question, you need to know the answer to the first one. And do you see this?
This is probably a pretty extreme example with a specific metric, but I have a broader kind of cynicism about a lot of these vendor -estimated ratings and data points as well.
So one thing that if you look up, if you try to figure out any company's sustainability footprint or how socially responsible they are, one of the first things that you're gonna find issomething called an ESG score or a CSM score.
And what this basically is is saying, let's try to just collect a bunch of data points about the company.
Do you have a policy for anti -harassment? Do you have an anti -discrimination policy?
Do you have, what's your workplace safety record?
What's your carbon emissions? Do you have some sort of policy about how you deal with hazmats?
All sorts of stuff spanning all sorts of environmental and social attributes.
And then what these vendors do is they essentially take all these data points, or as many as they can find, assign some sort of weighting to how important each one is, and come up withsort of a weighted average score.
[46:50] There's a couple issues here. First is a lot of times the weights are a black box, and this is a problem because let's say, again, let's go back to the issue of your union pension fund.
[47:00] Your mandate is probably to care a lot about employee treatment, so to care about wages, safety, harassment, discrimination, like anything that does with employee welfare.
But your mandate probably, and of course, maybe you want to invest in a green way, but your mandate doesn't really cover that.
So you need to prioritize labor issues first and foremost.
And it can be hard to use these scores to tease out the parts that you want, because you're getting these aggregate scores and you don't really know how important these issues that you careabout were to the person coming up with the score, right? So that's one issue with these things.
The other is, and this comes back to the issue of verification, is that there's almost an arms race to collect more and more data points, right?
But a lot of times what you get is, you know, feeding into these ratings is stuff like, do you have an anti -discrimination policy, right?
But there's no verification of, well, do you actually follow it, right? So I can I can have a great policy about anything.
If I don't enforce it, that's not really doing anything.
And so this is where, to come back to the point of relying on hard data.
[48:05] Instead of looking at, do you have an anti -discrimination policy, you might look at, well, have you been punished by the EEOC for discriminating against employees?
And of course, that's going to understate there's a lot of stuff that doesn't get reported to regulators. So you have your own measurement problems there.
But at least you know for sure that companies sanctioned by the EEOC, did something bad to their employees, right?
And so, in some sense, it's a case of, well, where do we get more measurement error and where is it more problematic?
[48:34] And which ones are easier for companies to game in the following sense?
I'm a company, I'm treating my employees pretty badly, right?
I can easily just write a policy that says I'm gonna stop treating them badly.
But still treat them badly, right? Writing that policy costs like a couple hours in lawyer fees, right, and that's it. But I don't actually have to do anything.
But if I'm a company that has a lot of workplace safety issues, I'm gonna show up on the OSHA databases for violations.
And unless, I might write a policy that says I'm gonna start being safe, right?
But if I don't follow that policy, I'm still going to show up in the OSHA data, right?
Only if I actually follow that policy will I stop showing up as a violator, right?
And of course, there's going to be issues where not everyone gets inspected every year, not every inspection catches everything.
And of course you have, whenever you're relying on a government inspection data, right? This is going to be the downsider, which is that not everyone, not everything gets checked orcaught, right?
But at the same time, you at least have hard data on something that we know how to measure and that companies can't really game, right? when the inspector shows up, I can't bribe themto not write me up.
Brent:
[49:40] Yeah.
Aneesh:
[49:40] Right? I hope.
Keller:
[49:43] Yeah.
Aneesh:
[49:43] So it's really an issue of hard data versus soft data and what we should be prioritizing.
Brent:
[49:48] Certainly.
Aneesh:
[49:48] In my view.
Brent:
[49:49] Yeah.
Keller:
[49:49] Yeah. It also seems like it might be smart to try to break it down into each of the E, S, and G, instead of trying to just group it all together.
Aneesh:
[49:56] I think so, right? And that's something where, again, we say which companies are socially responsible.
And that's much more so than financial performance, where we have a single metric, dollars or stock returns or whatever.
Different Perspectives on Company Virtue and Values
[50:10] Which companies are the best in this case is very much in the eye of the beholder.
You are a green investment fund, for example, then you probably care about carbon emissions and toxic releases.
And while you don't want companies to harass their employees, you might weigh the labor issues a little bit lower.
Or you're a union pension fund. The labor issues come front and center.
And of course, while you might say, yeah, lower emissions are good, that's going to be a bit less first order for you, given your mandate, than labor issues.
And so given the dimensionality problems in play here, virtue is much more in the eye of the beholder for investors than it would be when we talk about financial performance. And this iswhere breaking things down a bit helps.
So you're a fund, you're an investor. Maybe just you as an individual investor.
Right? You're really passionate about a certain cause, right?
It helps to understand which companies align with that cause.
You care about one thing, you don't care about these other things quite as much, right?
How do you can, and it would be nice if you could find a company or set of companies or portfolio that you said, yeah, that aligns with my personal values, whatever they may be.
Or I think both, I think two things are great, but I think this is a deal breaker for me and this isn't.
How do you find a company or a set of companies to invest in or to even to work for?
This goes beyond just investing. This goes into, well, do you want to work for a company that you think is doing bad for the world?
[51:37] Some people say, yes, I need a paycheck. Some people say, no, I'm gonna try to, I'm willing to make a personal financial sacrifice to not do that.
Brent:
[51:45] Sure.
Aneesh:
[51:45] Right? Whatever your choice is, you should be able to make that choice.
And this is where it helps to know, OK, what are they doing? Do I care about this?
If I care about this, are they good or bad in this aspect?
And that's easier to do than to say, it's easier to say I care about a specific thing than to say I care about social responsibility.
And it also helps us maybe come up with personal rankings in the sense that most firms are not good at everything.
You have firms, you have, Tesla is a good example here. Tesla is like kind of, at least until recently, was leading the sort of green car revolution, at least in terms of marketing andperception, really getting electric cars to be perceived as something in the mainstream that people wanted.
So you'd say they're very green, but you read the reports from Tesla factories and their labor performance, their labor relations are, suffice to say, not great.
So do you invest in Tesla? Well, it depends on whether you think The latter is a deal breaker.
Or whether you think what Tesla has done to bring green cars into the forefront and make them mainstream justifies the investment.
And of course this is setting aside the financial performance for Elon Musk for a second.
The Complexity of ESG Ratings and Agency Weighting
[52:59] But this is a case where if we look at Tesla's ESG rating, and of course there was a Musk tweet about this last year, where Tesla is rated really highly by some agencies and reallypoorly by others.
And what that comes down to is these agencies quietly deciding they care more about one aspect or the other.
And of course, if you know what these weights are and you know how different rating agencies weight different aspects, then you know almost which rating you care about more andwhich one you trust.
But that takes a lot of work. And agencies aren't necessarily forthcoming about this because they all want you to think they do the best job in capturing a company's overall footprint.
You're not gonna have a company saying, yeah, we're pretty good at this and we don't care about that So you should not use us for that, right?
They want to sell you their ratings, right?
And so it's a case where really focusing on the components and even not just all of E or all of S, but trying to focus on labor relations or focus on product safety or who is, for example, ifFacebook and disinformation, right? That's a consumer issue.
It's not neatly categorized, but it's something you should care about.
And so trying to drill down and focus on the things that you care about, you think are important, requires some degree of disaggregation and also disaggregation of data in a way that iseasily consumable to those of us who don't make a living trying to understand these things.
Brent:
[54:20] Definitely. And then as we kind of wrap up here, what are your thoughts moving forward with the future of ESG and how do we get proper standardization, transparency, betterdata?
Like where do you wanna see this area going?
Aneesh:
[54:37] Yeah, so this is one where it's gonna be complicated, at least in the US, by how politicized it's gotten, right?
And so to me, a lot of the future of ESG is actually moving away from the term ESG and really talking about, well, okay, we've started to talk about stakeholders and we've started to havea conversation on needing to treat stakeholders well.
So a lot of it is saying, okay, who are the stakeholders that we have perhaps been ignoring or underweighting, right?
And how do we, for example, make sure they get their due keep?
How do we make sure that they get treated fairly? And a lot of this really comes down to, I mean, a lot of ESG and a lot of the financial discussion around ESG, really centers on thisnotion that companies should internalize their externalities, right?
[55:22] Where we say, OK, well, companies, for example, a lot of companies have high carbon emissions, and they decide this is a good strategy because they're not being taxed on it to theextent that would force them to go green a lot faster, right?
Or, you know, you look at Walmart who has employees on who basically work full time and then get food stamps, right?
[55:41] And that comes from again, you know, Walmart saying, well, yeah, I don't have to pay you as well because you can also just go apply for SNAP, right?
And if, you know, if you said, if you essentially forced Walmart to pay a living wage or whatever it may be, right? I don't wanna get into debates on specific things, right?
But if if you essentially said, okay, here is how much Walmart paying employees poorly costs the government, right?
How do we recoup that? How do we force Walmart to pay for this externality, right?
Tackling Specific Issues and Disaggregation in ESG Discussions
[56:08] You might have a different conversation, but these conversations are gonna be different for every environmental or social issue that we think is worth fixing.
And, you know, of course, not everyone may think the same issues are first of all problems or need to be fixed, right?
And of course, this aligns a lot with worldviews, political views and have you, right?
But it's easier to, I think, tackle specific issues, right? If you can first say, here's the issue, here's what it costs, right?
Here is the externality is not getting internalized in this context, how do you internalize it?
And that requires some degree of disaggregation, right?
Because it's hard to really say, well, here's the total externality of your ESG performance, right?
That's just, it's gonna be an estimate, no one's gonna believe it.
There's gonna be fights, even more than there currently are.
Brent:
[56:55] Yeah, certainly.
Aneesh:
[56:56] So that's kind of my take on that.
Keller:
[56:58] Makes sense. Thank you for your time today. It's been wonderful.
Brent:
[57:01] It's fun. Thank you.