I was surprised to learn recently that an important research paper that forms the basis of the intellectual underpinnings of the austerity movement is wrong. The data was improperly averaged. And the mistake was discovered by an earnest graduate student who wanted to look at the data. It’s an interesting story and really demonstrates how bias works and the importance of corroboration and it deserves a read. Interestingly, Reinhart and Rogoff who wrote the original erroneous study, never published their data set or coding method.
Partly, I’m interested in how this research by Thomas Herndon will get covered in Postmedia outlets like the National Post, the Calgary Herald and the Ottawa Citizen. On my view, these news outlets have a duty to our democracy to present these kinds of findings to the Canadian public. So far, nothing yet really. Well, the Financial Post published a Reuter’s piece by Edward Krudy. But here’s hoping that Postmedia will do more on this.
But what I’m really interested in is my Google search results. See, I found out about this economics snafu by following a smart fellow on Twitter.
I read the article that Robin Sas linked to and then I did a Google search. And the Google search returned the home page for the Financial Post and the home page for the National Post even though there was not a single reference or a single link to a reference on either home page about “Thomas Herndon”, “Herndon”, “Reinhart”, or “Rogoff”.
That’s a serious problem for Google.
I need to trust that Google is doing a good job indexing webpages and presenting them in my search results. And Google is getting spoofed by Postmedia somehow. And it could hurt Google’s reputation.
How did Google get it wrong?
To investigate further why Google was returning useless results I did a Google site search of the National Post for “Thomas” and “Herndon.” There were four webpages returned, even though none of these pages had any reference to Thomas Herndon. And then I noticed identical text in the page descriptions:
Thomas Herndon spotted errors made by the economists in an influential paper that has been cited globally as justification for slashing spending…
Even though “Thomas Herndon” was not showing up on these pages, or even on the cached pages that Google caches, this text was in the page description. I believe this text was part of an advertisement for a Reuter’s article about Thomas Herndon.
In an age when a lot of Search Engine Optimization is snake oil or worse, black hat, I think the National Post has duped Google into thinking that they have something useful to say when they don’t. Are you listening, Google? Please fix this.