There is at present a lot happening around the question of ESG measurement. Climate-related metrics are in the spotlight thanks to the Taskforce on Climate-Related Financial Disclosures (TCFD), and broader environmental and social issues are also moving up the agenda. 

But better measurement is only part of what is needed and we can’t afford to wait for perfect measurement before taking other action. Here are five observations on why we have to find ways to manage things that are measured imperfectly or not at all.

1. The Goodhart effect: what gets measured gets manipulated.

The connection between measurement and management is not as strong as popular wisdom has it. We’ve all heard that what gets measured gets managed. But while there’s certainly a connection, this is only a generalisation, not an absolute truth. 

A recent piece from Quartz at Work, for example, argues that “what gets measured gets manipulated”. Measurement, in other words, leads to action – but frequently that action is focused on the metric, not on the underlying reality it is trying to measure. 

This effect, common in social sciences, was neatly captured long ago by the economist Charles Goodhart in his eponymous law, often expressed as “when a measure becomes a target, it ceases to be a good measure”. 

So we can think of the Goodhart effect as being the process by which a measure becomes distorted. And if we are focused only on that measure, then, as the Quartz at Work piece puts it, “fixating on disclosure can perpetuate inequity”. Measurement is good. But if measurement is the only thing we do, it’s only a matter of time before efforts become misdirected.

2. Measurement is difficult.

The weaker the connection between a metric and the thing it is trying to measure, the more likely these distortions become. 

And that connection is weak when it comes to most sustainability measures. In a recent article in the Harvard Business Review, Kenneth Pucker, former COO of Timberland Inc, lists some of the reasons that sustainability measurement is difficult: lack of mandates, lack of auditing, specious targets, opaque supply chains, complexity, confusing information, inattention to developing countries. The result, he argues, is that “the focus on reporting may actually be an obstacle to progress – consuming bandwidth, exaggerating gains, and distracting from the very real need for changes in mindsets, regulation, and corporate behavior.”

Ways to improve the measurement of climate impact are explored in depth in another report, this one by the Portfolio Alignment Team, a recently-formed industry group. It notes that measures fall on a spectrum of sophistication. It is easy, for example, to measure how much of an exposure a portfolio has to fossil fuel companies. But that’s a very crude measure of the climate impact of the portfolio. At the other end of the spectrum would be a degree warming metric such as the implied temperature rise (ITR) of a portfolio. But that’s a black box calculation, only as good as the underlying methodology, and sensitive to the assumptions that are built in to it. These sorts of measure are still early in the development process and as the report notes: “making portfolio warming methods robust will require significant improvements in data and other inputs”. 

The report notes, sensibly, that “there are merits in using a small collection of approaches on the above spectrum as a cross check on the results”. While no single measure is perfect, using several measures together can build a more complete picture. But this is a complex undertaking. 

And if climate impact is tough to measure, consider how much tougher biodiversity impacts are, or human rights or other social impact indicators. We can make best efforts to measure many of these things, but such measures will always be at best imperfect.

3. Don’t assume a can opener.

There’s an old joke that economists are fond of involving a shipwreck, a desert island and a box of canned food, with the punch line being a triumphant proclamation that the economist has solved the problem: “assume a can opener”. 

I’m not endorsing the joke (actuary jokes are much better), but the expression does capture an important point: it’s no use solving everything about a problem except the bit that’s actually difficult. 

In the context of climate (and other ESG issues), that means resisting the temptation to think that imperfect measures are better than they are. Understanding the dynamics of the connection between investment decisions and climate impact is the bit that’s actually difficult. Excellent as the work of the TCFD is, even if their recommendations were fully adopted tomorrow, there would still be an enormous amount of work to be done to answer the question “what is the impact of this entity on climate change”. 

4. What is not measured can still be managed.

Even if it were universally the case that what gets measured gets managed, it wouldn’t follow that it’s impossible to manage things that are not measured.

The nagging feeling that it would be stems from a widespread logical fallacy, which goes by various names – the fallacy of the inverse, denying the antecedent – and which seems to be particularly well-embedded in the human mind. Logically, we know that A implies B is not the same as B implying A; intuitively, that distinction is blurred. Hence we easily make the mistake of thinking that just because better measurement leads to better management, it must follow that better management is only possible with better measurement. But it doesn’t. 

This matters in the context of climate, for example, because, as we’ve just noted, the measurement bit is work-in-progress. Helpful as better measurement is, it’s not a pre-condition for other action.

5. Ultimately, it’s about a governance process not just a metric.

I don’t want to downplay the importance of measurement. Better measurement leads to good things happening. But measurement is not enough.

Best practice goes much further. Better understanding of both the quantifiable and the unquantifiable impacts of investment decisions is needed. And a recognition of those impacts needs to be embedded into the decision-making processes of individuals, of corporations, of legislators, of regulators and of investors. Better measurement is one of the ingredients to make that happen, but ultimately, better management of our environmental impact comes down to a governance process, not just a metric.