Weirdness in the Data

There's some weirdness in my data.  Weird can be good; unexpected results are interesting.  Weird also means a lot of re-checking of data.

 

Brazil seems to devote a tiny fraction of its land to agriculture.  Of the years I'm studying (1990-2007) no state has more than half of a percent of its land devoted to agriculture at any point.  At least, this is what the data is telling me.

 

For comparison, Illinois uses about 80% of its total land for agriculture.

 

So, I've been re-checking, and re-checking. My first thought was that I'd mis-read the units on the state areas.  My data had a state with over a billion square kilometers (Pará).  But, that turns out to be true.  Next, I figured that maybe I'd made a mistake in converting from hectares to sq. km. But, I think I got that right, too.

 

The most recent thing I've checked is if the units on the reported areas planted made sense.  Maybe they said 'hectares' when they meant 'sq. km'.  I verified this by looking at tons/hectare.  But, that seems reasonable at least within an order of magnitude (54 tons of sugarcane/hectare seeemed plausible). 

 

So, the next thing to do is to talk to some Brazillians, and see what they have to say.  But, it's making me really curious.