Monday, October 04, 2010

Coups in the News

'Cus even in Madagascar,
we'll find some shack below radar

-Gogol Bordello

Back in early 2009, when things were beginning to turn sour in Madagascar, I remember being frustrated when Google news couldn't feed my information addiction regarding the pending coup. English-language articles that came up from a search of 'Madagascar' were mostly about the release of 'Madagascar: Escape 2 Africa.' Though this would soon be hilariously ironic, I couldn't figure out why this was more news-worthy than a tempest of cyclones, looting and massacres. I still don't know why it didn't get covered as extensively as it might have, but I've been looking at Google trends, and it seems to validate that it was bad coverage, and not my information withdrawal, that sent me into such anxious fits at the internet cafe.

Here's the first plot, showing web traffic and news reports for Madagascar since 2004, to give you an idea of how coverage of the movies compares to the coup. (Click on any of the images below for a larger view.)




The first thing to notice is the relative size of the peaks in overall traffic, at 'A' and 'B,' to the average level of traffic and subsequent coup traffic. Notice that peak A is quite wide -- beginning around the 'Madagascar' release on May 27, 2005 and taking until mid-2006 to return to a normal traffic level. It also has a secondary peak corresponding to the DVD release on November 15, 2005. Since the release of the second film overlapped with the period of turmoil, we'll just have to assume that the traffic peak for the second movie at 'B' has similar traits. This means that when we look at traffic for 2009, the high baseline at the beginning of the year is mostly residual activity from the movie release on November 7, 2008.
Also worth mentioning is that the news coverage, shown on the bottom line, was fairly responsible -- small peaks for the movies relative to the day of the coup.

This next figure we zoom in on 2009. Remember, the high traffic volume at the beginning of the year is residual buzz from the movie.



Here we see some stranger behavior from the news publishers. There's a small hiccup in late January, when the looting and shootings took place, which in my mind were the most intense moments of the period. Yet the greatest news activity only happened when the president peacefully stepped down on March 17.

So far, we've seen that web users care more about movies than Malagasy murders, regardless of how much noise the news makes. We also might hypothesize that the internet-based media mailed this one in. It missed all the real action and mostly covered an event that was pretty much a forgone conclusion in the minds of anyone paying attention. But is this true everywhere? How do other coups stack up?

Honduras suffered a coup in June 2009, just 3 months after Ravolomanana stepped down. I've graphed activity for Honduras and Madagascar in 2009 on the same chart so that we can see the relative intensity of the coverage.



All of a sudden, that huge spike on March 17 looks like a speed bump next to the Honduran Himalayas. This makes sense. To get to Madagascar from the east coast of the United States, you need 19 hours and $5000. Also there are probably more American expats in Honduras and Honduran expats in America than is the case for Madagascar. You can't report what you don't see.

But come on - there were more articles written in
Romanian than in English about the crisis in Madagascar. Romanian! I don't even want to think what that would look like in terms of articles per fluent capita!

Certainly, there's a lot this data can't account for. (Was the coup in Honduras more unexpected and sensational? Is there a disparity of the level of access to telecommunications in the two countries? What effect did World Cup qualifiers have on the traffic?) But I can't help but wonder if there's a certain amount of negligent laziness that determines the news we see and the news we don't. What do you think? And seriously, what's the deal with the Romanians?