Thursday, March 24, 2011

Long Term Crop Prices

The above shows prices for the three main grain crops in the US since 1920, adjusted for inflation (using the CPI).  A few basic points for future reference:
  1. Food crops have been getting gradually cheaper for many years.  With a couple of major excursions due to world events (price drops in the depression, and rises in the second world war and 1970s commodity price shocks), the basic trend has been for food to get cheaper and cheaper. This indicates that despite the growth in world population, improvements in agricultural productivity have tended to outpace increases in demand.
  2. In particular, any possible effect of climate change on agriculture in the future needs to be evaluated in this context: is it likely to be large enough to overcome the historical primacy of technological innovation as the control on prices?
  3. Note that the fact that prices have been low and dropping for decades means that there will be many possible investments to raise production that will not have been made, because prices were too low to justify them (for example, in the northeastern US, large amounts of perfectly usable cropland have actually fallen out of use because it was not economically competitive with better cropland elsewhere).  Therefore, there is likely to be significant scope to increase production in response to any factor that causes a sustained rise in prices.
  4. The recent rises in food prices since the early 2000s are small compared to historical excursions.  But they are still in need of explanation to determine whether they are likely to be a blip, or a more long-lasting trend reversion.
  5. Prices of crops tend to move together.  There are good theoretical reasons to expect this over time, since if crops move too far from their normal relationship, there are farms that can switch from one to another, thus arbitraging prices.  For example, if corn gets too expensive relative to wheat, and farmers believe this will continue, some can switch from growing wheat to growing corn, and this will tend to lower the price of corn and raise the price of wheat.
We can see this last point empirically in the data.  For example, here is a scatter plot of corn vs wheat:

The straight line explains 80% of the variance.  Clearly the prices can vary from one another somewhat in the short term by a few bucks, but large scale long term movements are linked.

And here's soybeans versus wheat, which shows the same basic pattern, but with a slightly less strong linkage:


Robert Wilson said...

During the late 1940's I and other high school kids had reasonably well paid summer jobs at the three large grain elevators near Amarillo TX. We were mostly emptying boxcars with mechanical mules and brooms. I recall some locals a hundred yards or so from the elevators sweeping up wheat spilled from the boxcars. Some farmers were doing so well that they put in landing strips and bought airplanes. I believe that most of the wheat was being shipped to Europe.

Brett said...

Is that why there was such a sharp spike in the prices of those crops in the late 1940s? It would make sense.

The 1970s spike is likely easy enough to explain: a combination of inflation and skyrocketing oil prices.

Unknown said...

In essence, agreed with WiseBass.

For the last 50 years, would not the price of oil explain a very large portion of the variance in the price of wheat, corn, and soy?

That would be consistent with things I've read about the intensive use of fossil fuels in the agricultural production process.

Unknown said...

Thanks for this - insightful. However, I would have expected to see a stronger divergence in Corn vs the others if the 'bio fuel effect' on global food prices is indeed real.

Geoff said...

This might be of interest to you:

Mr. Sunshine said...

Wow. either the CPI revisions are suspect, or my mind is going, because I was alive, legally an adult, married and relatively poor in the 70's, and I really don't remember being shocked at the price of bread then. I am these days.

Lars-Eric Bjerke said...


A paper from the World Bank, A note on rising fuel prices, July 2008 concludes:

“The rapid rise in food prices has been a burden on the poor in developing countries, who spend roughly half of their household incomes on food. This paper examines the factors behind the rapid increase in internationally traded food prices since 2002 and estimates the contribution of various factors such as the increased production of biofuels from food grains and oilseeds, the weak dollar, and the increase in food production costs due to higher energy prices. It concludes that the most important factor was the large increase in biofuels production in the U.S. and the EU.”

There is a similar paper from OECD in 2008, Economic assessment of biofuel support policies. It may be noted that both papers makes a difference between Brazilian sugar cane ethanol and US and EU ethanol.

Unknown said...


wrong question;
Do we want to? that's the right question

we'd be better off aligning group effort to improving our own methodology,not worrying about their trivial needs; that's for them to orient to

Paul Yarbles said...

Your post got me to thinking about the CPI and hedonic pricing.

Let's pretend there are two items in the CPI calculation grains and TVs. Both are equally weighted.

In 2000 grains cost 2 dollars and TVs cost 2 dollars.
In 2010 grains cost 2 dollars and TVs cost 4 dollars.

Now without any fiddling we have an inflation rate of 50% over 10 years.

But using hedonic pricing economists decide that TVs are so much cooler in 2010 than in 2000 that the real price is still 2 dollars. Grains don't have this adjustment because they remain the same. With the adjustment the inflation rate over 10 years is a whopping 0%.

In this case, hedonic pricing actually made a drop in grain prices in constant 2010 dollars go to flat and thereby made a decrease in grain prices go away when charted in constant 2010 dollars. Here are the numbers:

With hedonic pricing we assume 0% inflation and we have:

In 2000 grains cost 2 2010-dollars and TVs cost 2 2010-dollars.
In 2010 grains cost 2 2010-dollars and TVs cost 4 2010-dollars.

Grains remain flat and TVs go up 2 2010-dollars.

Without hedonic pricing we assume 50% inflation and we have:

In 2000 grains cost 3 2010-dollars and TVs cost 3 2010-dollars.
In 2010 grains cost 2 2010-dollars and TVs cost 4 2010-dollars.

Grains go down 1 2010-dollar and TVs go up 1 2010-dollar.

Does this simple example reflect what could be happening in the actual calcualtions of the CPI? That is, could a drop in grains actually look less in inflation adjusted charts because of hedonic pricing of other items?

One other thing about the trends. What is the cost of the final goods made from the grains in terms of average hours worked? This is what most people care about. Not commodity prices. Of course the commodity prices are a big factor in calculating the final costs but many other factors are important too. Maybe the grain producers are getting squeezed. Maybe the wage earners who buy the final products are getting squeezed also.

Anyway, the most important issues are the sustainability of the current agricultural methods, the possibility of alternatives that maintain of expand the current yields, and the damage done to future yields by the current methods. The past trends are great but thinking about these issues fills me with anxiety.

Paul Yarbles said...

Correction to my last post:

With hedonic pricing we assume 0% inflation and we have:

In 2000 grains cost 2 2010-dollars and TVs cost 2 2010-dollars.
In 2010 grains cost 2 2010-dollars and TVs cost 2 2010-dollars.

Grains and TVs remain flat in 2010-dollars.

Unknown said...

Readers may find Julian Cribb's article on Why Farmers Need a Pay Raise interesting. See:

History of prices may not be a good predictor of the future, as population growth goes exponential through 2050. Also fossil fuels, which are key ingredients for fertilizer and pesticides, production is peaking, and we should expect higher prices.

Jay Kimball
8020 Vision

Nick said...

Short-term inflation among especially budget-constrained individuals is where I see much of the pain - even amidst this apparent long-term trend of declining crop prices. We should also remember to consider the impact that falling crop prices have on incomes throughout the farming world.

In my mind, predictability appears to be an important factor, here, and I'd love to compare food security/social unrest measures with a regional-scale price volatility index. I'm guessing something is already out there...?

ghazij said...

Why price crops in 2010 $ and for a long time span ( 1920 - 2010 ) .

Since when is the 2010 $ a reference currency ( i.e a long term , stable value currency in terms
of gold for example ) ? and why should we use the US CPI to derive the long term real value of the current ( 1 year ) USD ?

It seems odd to measure prices in an elastic currency and then to correct for that error of currency value elasticity, by deriving the real value ( fixed over a year ) of that scale-currency thru the use of an inflation -
adjustment gimmick ( i.e use the Price Level hitherto measured in an elastic scale , to get the long term and supposedly fixed real value of that scale )

For example, when it comes to corn , let's not forget the corn - derived ethanol produced as a substitute for gasoline fuel .... and all the price incentives to switch to corn planting .. the morale of the story , is that you can twist the figures to make them say what you want ..... in a fiat- money world

... Money illusion as serious economists would say