Archive for November, 2008

The Staff of Life: Wheat Trends in the U.S.

November 10, 2008

UPDATE:  November 09, 2009

I’ve just updated the wheat production data recently published by the U.S. Department of Agriculture.  See the new graphs below.



Whenever I look at a nation-state or a geographic cultural cluster, the first thing I think is, “How do they take care of their basic needs?”  You could do a thought experiment that included a question like, “If this unit of analysis suddenly found itself cut off from the rest of the world, would they be able to survive on their own natural resources?”

Let’s examine the “unit of analysis” thing a little.  In this particular piece, the unit of analysis is the United States as a whole.  But sometimes, the nation-state is not the best one.  Borders are always changing all over the world.  Cultures, on the other hand, tend to persist over longer time periods.  The up-side of nation-states is that they keep lots of detailed statistics.  Cultural groups, however — even though they persist longer than nation-states, often have boundaries that cross national borders, and to get any kind of useful data, you have to send an anthropologist out into the field every time you want a datapoint.  Not practical.  So, you go with what you got.

The other day, I was trying to get trend data on wheat production in the former USSR countries, and fortunately the Russian Republic archived some of the old records from each of the newly (relatively) formed nations.  The data are pretty spotty, though, and I don’t have a Russian specialist ready at hand who knows where all the dusty records are kept.  I still haven’t found the dusty old records at Statistics Canada, but I probably have a better shot at getting trend data from them than anybody other than the U.S.

So, that leaves me with a very nice dataset from the U.S. Bureau of the Census and the U.S. Department of Agriculture.  Most of what follows runs from 1866 to 2008 (the 2008 data points are estimates made by the respective agencies).  There is one series that starts in 1919 and runs to 2008, and that is on the acres planted.  For some reason, this element was either not collected or not reported until 1919.  I don’t know why, because you would think that it would be important to know how much was planted and then how much was actually harvested.  From that you could infer possible causes, like hail storms or dust storms wiping out a planting before it could be harvested.  You will see a stark example of the dust bowl in the first graph.

OK, so here we go…

This first graph runs from 1919 to 2009, and it is the acres harvested as a percent of the acres planted.  It’s quite dramatic to see the results of crops wiped out and never harvested because they were literally buried in dust before they came to maturity.

Wheat harvested as pct of acres planted

This next graph shows the yield per acre of all types of wheat for the period 1866 through 2009

Wheat yield bushels per acre

Beginning in about 1936, the trend line starts to go exponential (at least in the short term).  You might surmise that the sharp jump just after that was the beginning of World War II, and you would probably be right.  You might also surmise that it was the result of farm subsidies from the New Deal, the introduction of large sowing and reaping machines, the introduction of fertilizers and pesticides, cheap fuel, etc., and you might very well be right on all counts.

Now, let’s ask the fundamental question:  For each person in the U.S., how many acres were harvested during this same time period.  So, the two graphs that follow control for the population by putting them on a per capita basis.

First, the number of acres harvested per capita:

Wheat acres harvested percapita

Oh, oh, something is strange here, no?  The yield goes up for every acre but the acres per person goes down.  Let’s see what happens when we ask the next obvious question:  If the apparent efficiency is so high, then the yield per person must go up, no?

Here’s what happened:

Wheat yield bushels percapita

The curve looks for all the world like a very nice sine wave, except for that nasty dip during the depression and dust bowl.  The straight line that runs through the series is the overall average through the entire series.

Keep in mind that none of this takes into account imports and exports.  If I were to attempt to do that, I would very much like to have an economist explain the subtleties, because as a mere sociologist, I haven’t a clue about the mysteries of balance of trade dynamics.

I almost put a regression line through these series, but then thought better of it.  This is the pure raw data and some simple ratios for per-unit production (by unit acre, and per capita).

UPDATE:  November 09, 2009.  Well I finally did succumb to the temptation and put regression lines through them.

Please leave comments in the comments section (just click on “comments” and it will take you to the input page).  I really would like to have your opinion of these trends.