Issue deep-dive: Education
Tuesday, May 14, 2019
Tuesday, May 14, 2019
A deep dive of sustainability topics with Ethic's lead data scientist
Issue deep-dive: Education
Tuesdays with Travis is a collection of monthly interviews with our data science lead, Travis Korte, that explores the complexities of expressing values through data.

In this month’s Tuesdays with Travis, we focus on a cause that is both popular among clients and complex to capture in public-market data. That topic is education.

Are common methods of measuring education among public companies the best tools to track impact? If we consider education to be about informing the public, what data do we have available to measure truth? This is where finance meets philosophy.  

Comments have been edited and condensed for clarity.

How do you choose the data that best captures a topic as vast as education?

The traditional institutions we think about when we think of education—colleges, universities, K-12 programs—are largely run by private, non-profit or government-funded entities. To measure public-company involvement, we consider a variety of approaches.

One way is to look at business involvement in the education sector—companies in the education technology space, for example. But we’re not convinced that that’s an informative criterion for sustainability. Look at for-profit education companies in the U.S. over the past couple of decades: a lot of these companies create enormous social costs, but we’d miss that if all we look at is involvement in the education sector. Involvement, in and of itself, isn’t a good enough indication that the company is contributing to better educational outcomes.

We’ve also seen education measured through donations, such as sponsoring research labs or departments of universities. We’re skeptical of that. If we took donations into account, it would be an easy way for a company to improve its ESG ratings using just money, making no actual changes to its internal processes. We don’t think that’s a reliable metric of a company’s commitment to education.

Because those lenses don’t make a lot of sense to us, we’re thinking about several approaches, including internal company training, such as education of employees, and media integrity.

Media integrity is an educational issue because people use media to inform themselves about the world. So we want to look at the extent to which companies have leverage over the information people consume, whether they be traditional media companies, social media companies, internet service providers (ISPs), or various other kinds of information gatekeepers. And we want to evaluate the efforts these companies are making and the impacts they’re having in improving the quality of information they’re spreading.

You mentioned media integrity, which strikes me as a very thoughtful way to think about education as expressed in public markets. How do you measure that concept?

We’d look at both media companies and those that are advertising through those channels. In both cases, we’re finding ways to capture the quality of the information that’s being channeled to the public. If you’re a broadcaster with no controls around the way you advertise or create editorial content, for example, we have no reason to think you’re vetting the information you’re transmitting. We don’t think that makes for an ideal educational resource.

If the content is disingenuous or misleading to the public, how do you determine which company shoulders the most blame? Is it the author, the advertiser, the platform, or some other actor?

It’s a really hard question. This is also something that government agencies and policymakers around the world are trying to figure out: where does blame lie? Our view is that media is an ecosystem with a lot of entities contributing in a lot of different ways, and meaningful change requires action from all the players.

So we’re looking all up and down the communications chain, not just at the content producers themselves. Since this is an active field of policy and an active critique of mass communications in general, I think it’s important to expose that subtlety to clients. There will be clients who say that their number one issue is ‘fake news’ and whoever is responsible, at whatever level, needs to clean it up. Another client may want to look further up the chain at the ISPs and search companies. Yet another may say that these upstream companies are protected legally from the downstream content and we should only look at the content producers. We have our house views on this, but we recognize that there are certain judgment calls in terms of what kinds of companies clients want to target.

If we’re measuring education by the quality and validity of information produced, how do you distinguish between education and truth?

It’s funny to ask the data scientist questions about truth. Working with measurement and statistics, you’re constantly reminded that everything is a proxy. In the best case, the thing you’re measuring is adjacent to the thing you’re actually trying to understand. It’s hard enough to settle on the criteria to say what makes a “quality education,” what makes a “reliable media channel,” but it’s even harder to try to reconcile those criteria with what we can observe, what we can measure. So we have to use proxies. We don’t necessarily care about the exact language in your advertising policy, per se, but we believe that it’s a proxy for the quality and reliability and independence of the content you’re producing. So that’s where we start.

My hope is that we’ll continue incorporating new data and new lenses for thinking about these problems over time. Luckily, a lot of people in academia and civil society are doing a lot of thinking about ways to measure media quality, and so we already have a few approaches we’re excited to test out.

For companies whose business depends on user interaction with information, I’d imagine there’s sometimes a conflict of interest between engagement and accuracy?

Yeah, media companies will probably need some encouragement to come around on this. But in the same way that we’re asking manufacturers to pollute less, or technology companies to better protect their customers’ data, the ESG community is asking media companies and advertisers to do things they wouldn’t have done on their own. We’re exerting pressure. There hasn’t been a strong enough market incentive for social media companies, for example, to improve the quality of their discourse. This is partly because of concerns that controlling toxic content on these platforms might reduce engagement—and revenue, by extension. But that’s what ESG investing is for: we live in a world where it’s not always in companies’ interests to reduce their social costs, and we want the investment community to help tip the scales and provide direction.

There’s also a dearth of information on the extent to which these channels are being gamed or abused. What would be really helpful is more standardized and comprehensive forms of reporting for companies that control communication channels. Is the content really getting better? Did the steps this company took actually improve things? I said earlier that everything is a proxy, but some proxies are better than others, and we don’t have much to go on for social media companies, ISPs, and other important players. That’s something that should be concerning to investors and the general public alike.

Key Takeaways

  • A public company’s donations to or basic involvement in the education sector aren’t always strong indicators of their commitment to education. Lenses such as media integrity or employee training could better capture whether or not a company makes education a top priority.
  • There are many companies involved in the production and dissemination of information. Clients should consider which parts of the knowledge chain matter most to them.
  • There hasn’t been an obvious market incentive for the social media companies to improve the quality of their discourse. Investors can help change that.
Sources and footnotes

Travis Korte is the Sustainability Research and Data Lead at Ethic. Previously, Travis organized civic-minded technologists at Hack for LA and advised a wide range of clients on data science, data policy, and quantitative methods. You can follow him on Twitter at @traviskorte.

We made a thing!