Back in January I made a very simple case for ethanol, and then I mentioned that I would write a post on why I don’t think that corn ethanol increases the price of food. I started putting together data to make my case, but never had a chance to finish it up. Well I finally did, and so here we go.
As the argument goes, making ethanol out of corn raises food prices because, instead of eating the corn, we’re burning it. But ethanol is made out of yellow corn, which is not used in food for people (that’s sweet corn). So a more persuasive form of the argument is that making ethanol out of corn raises food prices because, rather than feeding the corn to cows, we’re burning it. That raises feed prices, due to competition for corn supplies and thus an increase in the price of corn (ie an increase in the price of feed) which gets passed along to the consumer at the grocery store. When we’re talking about ethanol raising food prices the story is really about corn used (or not) for animal feed.
Here’s the amount of corn ethanol we’ve produced in the US in the past 20 years:
Prior to the 1990’s, basically no one was making fuel ethanol out of corn (or anything, for that matter) in significant quantities. Corn ethanol didn’t really get going until the late 90’s and grew explosively in the past decade. Last year, we made about 14 billion gallons, which is only 1 billion gallons away from the upper limit allowed by the EPA. With such rapid growth, where did all the corn come from and what does that mean for food?
As it turns out, the amount of corn harvested in 2010 was nearly double what was harvested in 1993. So did we plant twice as much corn? No! Instead, crop yields (bushels of corn harvested per acre planted) increased significantly, and the land dedicated to corn production increased by only 20%. Further, the total amount of cropland in the country remained flat over the time period, which suggests that no new land was brought into production:
But was doubling our corn harvest enough to make up for the demand caused by ethanol production? Well, in a word, yes. In 2010, and extra 4.5 billion bushels of corn were converted into ethanol than in 1993. But the corn harvest increased by 6.1 billion bushels over that time frame, leaving an extra 1.6 billion for other uses. And in fact, no major corn consuming sector saw a decrease in the amount of corn they utilized. Here’s the amount of corn used in ethanol production, animal feed, exports, and all other industrial uses (starch, sweeteners, beverage alcohol, etc):
But pay particular attention to the feed category, because as I mentioned that’s the one we care about when it comes to the question of food vs. fuel. More corn is fed to animals after the “ethanol boom” than before it. But what about that downwards slope starting in the mid ‘00s, coinciding with the really rapid part of ethanol’s expansion? Doesn’t that show that actually we’re feeding less corn? Well, no. This is because when you make ethanol out of corn, you also make something called DDGS. DDGS is all the parts of the corn left over that didn’t get fermented away, plus all the yeast cells that did the fermenting. This is important because DDGS is high in protein and fat, which means you can feed it to cows. In fact, DDGS is such a good feed that you can give a cow less DDGS than you would corn to achieve the same result. In practice, about one third of the corn that goes into an ethanol plant winds up as DDGS, and 1 kg of DDGS fed to a cow is equivalent to feeding 1.2 kg of corn. So, in reality, 40% of the corn that goes into an ethanol plant still winds up getting used as feed. We can ferment our corn and eat it too! Here’s what the above graph looks like when you account for the corn returning to feedlot as DDGS (the light blue line, on top):
When we account for the DDGS being returned, feed corn has increased by 25% since 1993! Now that I’ve shown that feed doesn’t face the significant competitive pressures that the food vs. fuel myth implies that it does, what about food prices? As my benchmark food price, I’m using the food component of CPI, over the two decades 1990 – 2010. Remember, the argument goes that ethanol increases the price of corn and thus the price of food so if this is true we would expect to see that the prices of corn and food would increase (at least partially) together. Here’s a comparison of the monthly change in the prices of food and corn over 20 years:
Can you spot a trend? I can’t. There’s no correlation between the two data sets. But monthly price data are noisy, so let’s take a look at the yearly changes:
Here we can see that there is a very weak correlation. A simple linear regression predicts that if corn prices double we can expect to see a 1.6% increase in the price of food, which is less than inflation. Of course the correlation is so weak, and the fit so poor, that the predictive value of the trend line is essentially worthless. However, if an increase in corn prices caused an increase in food prices as the food vs. fuel argument claims, then we should see a strong positive correlation. It’s simply not in the data.
Finally, let’s take a look at the actual changes in food prices. Luckily for me, NPR published a piece on food prices just the other day, so now I don’t have to go digging for the data. Instead, I’ll shamelessly steal theirs:
Between 1982 and 2012, a time span that has seen the amount of fuel ethanol produced increase over 1000 times, food prices have dropped. Significantly. The biggest decreases amongst meat have come from chicken and beef, both of which are fed with corn.
(Data sources: USDA NASS, BLS, FRED)
Read Full Post »