You are viewing a read-only archive of the Blogs.Harvard network. Learn more.

Newsletter: Problem – Solution Edition


Two pieces caught my attention this week: one sets up the core problem of Facebook in society today. The other sets out a (surprisingly) simple and well-thought-out way to solve it. I wrote about it on Working Paper in more detail, and started a discussion of what it might mean for news organizations, too.

While the two authors didn’t intend to speak to one another directly, they are a part of the same conversation.

Ian Bogost reviews the history of consumer data brokerage, and how we got to the modern advertising system. He describes the core problem today (accurately, in my view), that:

“The real difference between the old and the new ages of data-intelligence-driven consumer marketing, and the invasion of privacy they entail, is that lots of people are finally aware that it is taking place.”

In a fit of link-clicking on Lawfare I ended up at a 2016 Atlantic article by Jack Balkin and Jonathan Zittrain, arguing that a simple legal mechanism to regulate tech companies might be the idea of “information fiduciaries,” or companies that are shielded from some liability if they agree to act in our interest.

Historical context for our problem and an elegant update to a historical idea for the solution.


Hope you had a great weekend,


P.S. Media Mail worked — the library arrived.

“Information Fiduciaries” and News Media


Background on the concept of “Information Fiduciaries” and How It Might Help Align The Interests of Tech Platforms and News Organizations

As many of you know, I’ve been doing a lot of work around and thinking about the interaction between Facebook and news organizations lately. In fairness, who hasn’t?

But I’ve recently stumbled upon a years-old concept from Jack Balkin and Jonathan Zittrain that seems supremely relevant and under-discussed: information fiduciaries.

The concept, first published on Balkin’s blog, and deeply developed in an article for the UC Davis Law Review, is best explained in an Atlantic piece from 2016. (Read the whole thing. It’s worth it. But I’ll excerpt key pieces here.)

The sub-head of the article asks: “Doctors and lawyers are prohibited from using clients’ information for their own interests, so why aren’t Google and Facebook?”

Some background:

__(‘Read the rest of this entry »’)

What Could Tech Actually Do for Journalism


Last week, along with my colleague Yvonne, I gave a talk at the Berkman Klein Center fellows hour (which is actually two hours, don’t ask) summarizing how newspapers made money, how born-digital news organizations do today, and getting the group to help brainstorm about how to fund a BBC Media Action project.  That led to more conversation about what tech companies could do if they really wanted to help journalism.  Not surprisingly, I focused on how they could make the business model even a touch more durable.

A few of these, below:

  • Selling news just as we do any other good/service
    • As digital news organizations are increasingly focused on a single vertical, interest-based advertising might be able to help them grow audience, which can then be monetized directly or indirectly
    • My currently-running experiment is testing ROI on using targeted advertising to grow newsletter subscribers, who will then be solicited for donations
    • Tech companies are great at optimizing advertising for returns; great at matching people who want something with people who make/sell/write about/have it… So what if we ran these kind of experiments at scale?
      • What if there was an advertising product for publishers to hawk themselves to grow their audiences?
  • Measuring what people actually learned from the news they consumed
    • Using a combination of analytics, survey delivery tools, the science of public knowledge testing, and the propensity score matching technique, we can measure how much people learned about a topic, not just if they clicked on a story
    • My previous work at both Northwestern and USC shows this is actually possible — it’s just too labor intensive for most news organizations…
      • That sounds like a job for software!
    • Further, once we have a metric, we can optimize against it… this is again something that tech companies have unique skills because the testing and optimization can happen at Web scale
  • Building productivity tools to let journalists be journalists, not technologists
    • Huge amounts of the total costs borne by news organizations are either paying technologists or large inefficiencies in under skilled employees trying to work with technology
    • Many run on tools developed way before the Internet was a part of everyday life, for which digital output is an afterthought
    • Many born-digital organizations run on G-Suite (SaaS)… And others have already proven that newsroom workflow tools can be a business unto themselves
Log in