Friday 10 April 2009

Zeitgeist via Google Search Insights

I previously noted that we're likely not in an Atlas Moment. While sales of Atlas Shrugged have skyrocketed with the depression, Google Searches show a similar skyrocketing of searches on Karl Marx.

The search there was pretty basic. Let's expand things a bit. First, a straight-up search comparison on Ayn Rand versus Karl Marx isn't entirely fair. Folks might misspell Rand's name, folks might search on "communism" rather than "Marx", and so on. Let's run a more comprehensive comparison.

For Marx, we'll use all of the following terms: "Das Kapital", "Das Capital", "Communist Manifesto", Marxism, Communism, "Karl Marx", "Carl Marx". For Rand, we'll use: "Atlas Shrugged", Objectivism, "Ayn Rand", "Ann Rand", "The Fountainhead", "Virtue of Selfishness", "John Galt". I think that's a reasonably comprehensive list of terms folks would be looking for, but feel free to try your own variations. If you include only Marx, be sure to run the "-Groucho, -Richard, - "Marx Brothers"" and so on. Here's the link to my search. The blue line below shows Marxism while red shows Objectivism. I've also added in an orange line for searches on recession.

The first thing worth noting is massive seasonality in the Marxism search. Any guesses who shows up more in University courses? The seasonality tracks exactly northern hemisphere summer and winter breaks. Ocular least squares suggests a downward trend for both from 2004 through 2007 followed by an increase for both. Ocular least squares also suggests a lot more comovement between the Marxism search and recession than between Objectivism and recession. It looks to me like the increase in the Marxism searches is greater than that for the Objectivism searches. Compared to Marxism, it really doesn't look like an Objectivist Moment.

Some critiques in the comments at my prior post and over at Marginal Revolution wondered about the relevance of search terms or popularity. I'd of course agree. But some libertarians have suggested that the increased sales of Atlas Shrugged may point to an Objectivist Moment; that's also a popularity measure. I'd worry in book sales that online copies of Marx's works are easily found while online versions of Atlas Shrugged are a bit harder to find. Some also have pointed to university syllabus requirements as biasing things. That ought to be a fixed effect across years: I'd say that the growth in either since 2007 points more to effects of the recession. And of course note that Marx still beats Rand even in the middle of semester breaks.

Now, let's try Keynes versus Rand. The Rand search is as before (in red); Keynes (in blue) includes "Keynes" and "Keynesian" then subtracts anybody else named Keynes that shows up otherwise as common searches. Like Marx, Keynes shows a lot of seasonality tied to university schedules. Ocular least squares suggests a strong downward trend for both through mid 2007, flatlining through mid 2008, then a big increase for both from then. Recession isn't included as it throws out the scale.

Let's focus in on the period since January 2008.

Again, I'd suggest looking at growth rather than levels to identify trends. Ocular Least Squares suggests a baseline for Keynes around 40 and for Rand around 55 through mid 2008. The new baseline looks to be about 60 for Keynes and about 80 for Rand: a bigger absolute increase for Rand, but about the same in percentage terms. This is all pretty rough. If the 14 month old weren't about to wake from his nap, I might try downloading the data and running something more than Ocular Least Squares on it. But the eyeball check suggests that if there were an "Atlas Moment", as compared to a "Keynes Moment", I'd say it ran February of 2009 through mid-March, and is now over.

1 comment:

  1. Now that was a lot more interesting... excellent effort, with thanks!!

    What amazes me is how readily that first, inadequate Marx vs Rand plot spread through the Internet and to some newspaper sites.

    I noticed that one can only perform Google Insight plots from January 2004. I wonder how such analysis might turn out done for (some kind of parameters) over 25 year axis. I think an Ocular Regression Analysis would show a steady increase into its present statistical Ball Park.

    I suggest that the very fact that Rand is now so comparable to Marx or Keynes is the actual "Moment" to which you allude. That is, the "Moment" is not that Rand numbers beat Marx numbers, but that Rand is now being compared in that way!

    That is especially significant given that Marx, Hegel, Kant, Plato et al. are all put firmly in university students' minds largely by indoctrination rather than reason. In contrast, anything Rand is actively opposed by sneering & dismissive academics and journalists who often have never read Rand —see the many stories about BB&T's efforts to create a single capitalism course -that uses Rand's ideas- in a very few universities! Academics are horrified.