I just went and fact-checked myself via Google Trends. It does seem like the term had similar amount of web search traffic in late-2005 across all categories,

I just went and fact-checked myself via Google Trends. It does seem like the term had similar amount of web search traffic in late-2005 across all categories, compared to now. I didn't expect that at all. However, "News Search" only goes back to January 2008, so it's difficult to tell if 2005 also had similar news coverage. But the graph since 2008 definitely shows a massive increase: https://trends.google.com/trends/explore?date=all_2008&geo=US&gprop=news&q=yield%20curve https://trends.google.com/trends/explore?date=all_2008&geo=U... So I think I might be mostly wrong. (Of course, this wouldn't mean there's no observer effect... just that it would be at least partially baked-into our current understanding of how it works.) Edit: Hm. This seems to change a lot depending on the specific term used. "Yield curve inversion" makes it sound like people only started searching for this term in the last few years, and not really back in 2005. But changing it to "Yield curve inverted" shows the 2005 spike again. So now I'm just discounting Google Trends as a source, either way.