On droughts

1. The drought has missed the corn belt, but the heat waves haven't (making Michael Roberts bullish on food prices).

2. You know a drought is bad when the camels are dying.

Ahmed Mohammad, a Somali camel herdsman, told BBC: "It is a terrible sign when camels start dying because when they start to die, then what chance have sheep, goats and cattle?"

Ethanol fact(?) of the day

I hadn't heard this before, but a colleague told me today that 2/3 of the corn that goes into ethanol comes out in a form that can be used as animal feed (probably DDGS). If that's true, if we use 30% of our corn crop for ethanol it's more like 10% out of the food system, etc.

What crop supply response looks like

Stealing the link and title wholesale from Michael Roberts:

When prices for corn and soybeans surged last fall, Bill Hammitt, a farmer in the fertile hill country of western Iowa, began to see the bulldozers come out, clearing steep hillsides of trees and pastureland to make way for more acres of the state’s staple crops. Now, as spring planting begins, with the chance of drenching rains, Mr. Hammitt worries that such steep ground is at high risk for soil erosion — a farmland scourge that feels as distant to most Americans as tales of the Dust Bowl and Woody Guthrie ballads.

Decoupling of oil price and renewables

Geoff Styles has a post titled "Will $100 oil help renewables?", in which he argues the counterintuitive answer that, "no, not that much." Worth reading in full, but since I like to practice synthesis:

Today, gas predominantly sets the marginal price of power generation, and gas prices have decoupled from oil due to abundant shale gas supply. Transport is minimally electrified, so renewable power cannot yet substitute oil in that sphere. And prices for commodity input often rise along with oil, increasing renewable costs (a.k.a. the "receding horizon").

The first, I totally agree with. The second is broadly speaking true, although paths like CNG, gas-to-liquids and coal-to-liquids become economically viable with high oil prices and could re-strengthen the link between transport and electric power (as could increasing EV penetration over the longer term). The third is directionally true, but not absolute (and not entirely causal). Many second-gen biofuels use waste inputs which are not otherwise traded, so higher oil prices are an unmitigated boon for them. The prices of silicon and corn are often correlated with crude, but probably more because of overall economic growth than because crude drives their price. It will be interesting to see if corn starts to price off of its value as ethanol, as it did back in 2008. Not good for food security, if it does.

Watch the crush spread

Trying to chase down cause and effect in energy and resource markets can be frustrating - it is hard to follow a linear path to a new, coherent equilibrium. Take, for example, Geoff Styles' recent line of thought on the impact of Egyptian unrest on renewable energy.

... since the protests started on January 25, and without any actual disruption in oil deliveries, the price of UK Brent crude... has climbed by around $5 per barrel and now trades solidly above $100.

... [Ethanol and biodiesel] stand to gain if oil prices are driven up by factors that don't also push up the prices of the commodities from which they're made [emphasis mine].

That last bit is critical, and can't be taken for granted. In late 2008 ethanol was clearly the marginal use of corn and corn became priced off of its value in use as ethanol, squeezing margins despite high oil prices. If biofuels come back in a big way, this dynamic is likely to kick in (leading, incidentally, to even higher food prices, not good for most people).

Ethanol subsidies live another year

NOOOOO!!! Despite promising earlier signs, the ethanol tax credit renewal managed to sneak into the tax bill, so it's been extended into 2011 at least. Very disappointing, even though Geoff Styles thinks the subsidy won't last past next year.

Bipartisan support to end ethanol subsidies?

In today's political climate in the U.S., it's rare to find bipartisan support for anything, let alone an eminently sensible idea like ending subsidies for ethanol. (Not to be confused with cutting the cellulosic ethanol RFS requirement for 2011 by 97%, which is unfortunate but necessary given that we haven't actually figured out yet how to produce cellulosic ethanol economically at scale.) It will be interesting to see how the 42 ethanol state senators push back against this one, but it does seem that the shifting balance toward fiscal conservatism makes it both more likely that subsidies could expire and more difficult for them to be re-established once the do, as NRDC's Nathanael Greene points out in the NY Times article.

I'm not close enough to corn futures markets to know to what extent the expectation of lapsing subsidies is baked in, but I'll be very curious to see how world grain prices react if subsidies are allowed to expire. The discontinuity may give us a crude sort of counterfactual to help answer the persistent question of how much biofuels drive up food prices.

Via Michael Roberts, who's also cheering.

Heat and crop yields in 2010

This may be premature, but I'm very tempted to award a gold star for clairvoyance to Michael Roberts, whose research on the effect of heat on crop yields I've blogged here and here. Here he is on August 12:

[Current temperatures] are still rising fast. If this keeps up for a few more days I'd say yields will get hammered.

(to give credit where credit is due, the market may have seen the same thing)

Anyway, fast forward to last week, the October forecasts for wheat, corn and soy are all well below the September forecasts (although USDA puts a bizarrely positive spin on it). As Michael's chart below shows, the September forecast is almost always very good, and revisions are generally upward, not downward, so something funny happened this year, and late heat seems to be a very strong hypothesis.
Michael is investigating the more detailed data and I hope will post any findings as they arise. As I mentioned before, if it was the heat, it will be very valuable to prove and communicate this to stakeholders in agriculture, some of whom have taken very skeptical views toward climate change and its impact on them specifically.

More food price volatility drivers

Upon further reflection on Chris Blattman's rare dud, here are a few more (slightly overlapping) potential drivers of food price volatility in the future. Please note that I'm not saying these will definitely cause higher food price volatility in the future, only that it is very easy to believe that they might.

• Biofuels and bioenergy: An additional source of demand growth - potentially very large - that could keep demand at the very edge of supply capacity.


• Stronger links to energy prices: Energy prices have always been linked to agricultural input costs since the most widely used fertilizer (nitrogen) is generally made from natural gas. More recently ethanol has become at times the marginal buyer of corn (Bruce Babcock at CARD did some nice research demonstrating that ethanol was the marginal buyer from late 2007 to mid 2008; I couldn't find a link to the paper, but all you need to do is look at the high correlation between actual corn prices and break-even prices for ethanol production). If this continues - and it may well, particularly when oil prices are high - then volatility in oil prices will be transmitted to food prices more than in the past.


• Speculation: That old bugbear; I personally am more skeptical about this one, as longtime readers will know, but the former head of IFPRI isn't, and objectively I'm no more likely to be right than he is.
  

Update: Michael Roberts responds and reprimands:

Here Chris seems to talking as much about climate science as economics or politics. He has also stepped onto a pet peeve of mine, common among some economists, which is ascribing personal views as truisms stemming from the branch of social sciences in which one specializes. It's not quite as bad as Steven Levitt pontificating about global cooling, but it reeks of that kind of professional arrogance. If you're an academic and are going to start asserting scientific truisms you need to be more specific about the underlying science.

He also points out a few factors not yet on my list, which I'll paraphrase here:

• Globalization is not irreversible: Think about how surprisingly globalized the world became during 1870-1914, only to regress drastically following World War I and the Great Depression.


• Shifts in comparative advantage due to climate change, which Michael is "convinced" of:
  

  That is, [climate change is] going to shift where things are grown. A lot. It's also likely to change global quantities, but that's hard thing to put a finger on (i.e., model convincingly). With that much change going on, we should worry at least a little bit, and probably a whole lot, about how the kind of turmoil these changes will cause. Loss of comparative advantage is just the kind of thing that brings about bad policy response.

• Uncertainty itself can exacerbate market volatility, and uncertainty about the future is high, not just about long-term commodity prices, but also about the direction of the world economy and shifts in agricultural production due to climate change.

Market believes heat hurts crop yields

The market believes what Michael Roberts' popular paper showed about crop yields - higher temperatures really hurt.

Corn futures rose the most in almost two weeks and soybeans gained on speculation that the recent Midwest heat wave will mean smaller production than the record crops predicted today by the government.

August has gotten off to the second-warmest start since 1960, T-Storm Weather LLC said today in a report. Another forecaster, Commodity Weather Group LLC, said about 25 percent of the U.S. soybean-growing area won’t get enough rain for proper plant development over the next two weeks, and that the dryness could harm a third of the Midwest should rain miss sections of Illinois this weekend, as expected.

“The crops are going downhill rapidly in parts of the Midwest and South,” said Mark Schultz, the chief analyst for Northstar Commodity Investment Co. in Minneapolis. “Our farmers are already preparing for corn yields that may fall 5 percent to as much as 10 percent from earlier field samples.”

Looking for the silver lining, might this influence the farm lobby's stance on climate change?

Via Michael at Greed, Green and Grains.


Update: I'm struggling a bit to square this with USDA predictions of record U.S. corn production... I wonder if those forecasts were based on outdated inputs (like, perhaps, "earlier field samples").

Price rises in non-speculatable food commodities

I recently had a contentious discussion over the extent to which financial speculation caused the food price run-up of 2008 (and prior years). I am skeptical that speculation was a major culprit, and I was unable to remember the specifics of the argument that prices rose similarly in commodities which are not index-traded. For future reference (my own as much as my readers'), I tracked it down in this paper recently released by the OECD. (emphases are mine)

If index fund buying drove commodity prices higher then markets without index fund investment should not have seen prices advance. Again, the observed facts are inconsistent with this notion. Irwin, Sanders, Merrin (2009) show that markets without index fund participation (fluid milk and rice futures) and commodities without futures markets (apples and edible beans) also showed price increases over the 2006-2008 period. Stoll and Whaley (2009) report that returns for Chicago Board of Trade (CBOT) wheat, Kansas City Board of Trade (KCBOT) wheat, and Minneapolis Grain Exchange (MGEX) wheat are all highly positively correlated over 2006-09, yet only CBOT wheat is used heavily by index investors. In a similar fashion, Commodity Exchange (COMEX) gold, COMEX silver, New York Mercantile (NYMEX) palladium, and NYMEX platinum futures prices are highly correlated over the same time period but only gold and silver are included in popular commodity indexes. Headey and Fan (2008) cite the rapid increases in the prices for non-financialized commodities such as rubber, onions, and iron ore as evidence that rapid price inflation occurred in commodities without futures markets. While certainly instructive, the limits of these kinds of comparisons also need to be kept in mind. Bubble proponents have pointed out that commodity markets selected for the development of futures contracts may be naturally more volatile than those commodities without futures markets.

The paper also includes more far more statistical including the latest Granger causality analysis, which is more rigorous, but also in my view less effective than anecdotes like the above in many informal discussions. Their overall conclusion is unequivocally in line with my view that

... at this time, the weight of evidence clearly suggests that increased index fund activity in 2006-08 did not cause a bubble in commodity futures prices.

In related reading, here are previous posts citing Thomas Malthus and Darrell Duffie on the benefits of speculation in food markets, and here is Scott Irwin (one of the OECD paper co-authors) last year on why index speculators didn't break the wheat market.

Dead zones vs. oil spills

While ethanol producers have been quick to spin the Deepwater Horizon tragedy toward their own advantage, NRDC's Nathanael Greene has a quick rebuttal:

The nitrogen runoff from corn grown all along the Mississippi causes a huge dead zone in the Gulf every summer. As this map shows, the dead zone at least as large as the oil spill and it takes a huge toll on the marine life and region's economy every summer. With about a third of the corn crop going to make corn ethanol, it should be clear that more corn ethanol is not a real solution.

He's referring to this image from the NYT, which makes a side-by-side visual comparison easy; in fact, the hypoxic zone looks considerably larger than the oil spill to date.

A hypoxic zone is in some ways not as destructive as an oil spill - it will not cripple fragile marshland ecosystems, for example - but the impact on marine life alone is no doubt harmful for the coastal fishing industry that could be (and once was?).

It is not news that corn ethanol is hardly an environmental angel, but it is worth keeping in the public conscience as the Deepwater Horizon leak continues unabated and the inevitable public backlash builds in strength.

Weather as an instrument, etc.

Michael Roberts explains in a nutshell the econometrics of his fascinating and widely-referenced paper on temperature's worrying effect on corn yields. If you’ve ever wondered why more agricultural economists don’t use weather as an instrument or worry about the fundamental endogeneity of futures prices, this post is for you.

Growing the ethanol market

U.S. ethanol is not cost-competitive abroad, where it competes on an even playing field with Brazilian sugarcane ethanol, so to grow (and reach the RFS mandates) it needs to grow the domestic market:

"Without increasing the blend of ethanol to E15, it will be impossible to achieve the targets set in the Renewable Fuel Standard and there will be no market for cellulosic ethanol," said Jeff Broin, POET chief executive officer. "POET is spending tens of millions of dollars to commercialize the production of ethanol from harvest leftovers but needs E15 to be certain there will be a market for the product."

The clever (but erroneous) framing is that not meeting the ethanol RFS would be the end of the world.

Update: I learn a few things from a comment exchange with Geoff Styles.

R: I recently read Turning Oil Into Salt (which has a lot to like and a lot to dislike - I wrote a rambling review for anyone interested), which would have you believe that the incremental cost of making a car flex-fuel is minimal (in the low hundreds of dollars range above a normal car). If this is true (and I would be interested to hear informed views on whether it is), it would be worth at least looking at the cost/benefit of mandating or incentivizing flex-fuel vehicles, particularly if the alternative to satisfy the ethanol lobby is an E15 ruling that doesn't make economic and/or environmental sense.

Geoff: The Big 3 have already committed to making half their new cars flex-fuel, and with GM and Chrysler beholden to the govt. a 100% target won't be far behind. But even if 100% of all new cars were FFVs starting this year, there would still be many millions of cars vulnerable to damage from higher ethanol blends for many years to come. The problem isn't the FFVs, of which there are already enough to boost E85 sales dramatically. It's making the entire E85 proposition, with its high costs for service station owners and consumers alike more attractive. Does 75% of the energy of gasoline for 86% of its cost sound like a good deal to you?

R: Agree that E15 could be bad for legacy vehicles (I was thinking of FFVs as an alternative demand sink to raising the blending limit to 15%). I think raising the limit would be a bad idea.

And agree that expanding E85 infrastructure will take investment, but the economics of ethanol itself could become attractive enough to justify that investment (and the energy-adjusted price ratio with gasoline has and will continue to fluctuate based on ag supply-demand, crude prices and refining economics).

I don't really like corn ethanol, but if it's a choice between E15 or more FFVs to satisfy the ethanol lobby, the latter sounds more attractive to me.

Geoff: R,
When you put it that way, I agree, though the old saw about horses and water comes to mind. There are sufficient FFVs on the road already to absorb 10x the E85 currently sold. Understanding why they're not buying more of it could be key to figuring out how to make it work on a mass-market level. Is it a critical mass issue on either infrastructure or vehicles, or do consumers just not like the value proposition and attributes?

R: I hadn't realized there were already so many FFVs in the U.S. - sounds like that can't be a constraint now. Do you have any idea how much ethanol E85-equipped stations sell? That might shed some light on whether the issue is midstream infrastructure vs. poor value proposition.

Update 2: ... in which my poor reading skills and attention to detail are exposed:

Geoff: R,
The figure is cited and linked in the posting.

R: So it is :) Good stuff.

Map of the day: Access to markets in Ethiopia

From Roving Bandit, these striking graphs show how close a given point in Ethiopia is to international and domestic demand centers, respectively, in terms of travel time. If you're a farmer, living in a green area means your end markets are very far away, and transport will eat up a big chunk of the end price of whatever you're producing.

Transport times like this give you large import/export parity price wedges like this:

... which obviously leads to a lot of price volatility, which is bad for producers and consumers alike.

(the latter graph is a favorite of mine from the World Development Report 2008 on agriculture and development.)

Turning Oil Into Salt (5): Alcohol as fuel

Given the low incremental cost of making vehicles flex-fuel, the key barrier to the liquid fuels half of the Open Fuel Standard aspiration is sourcing the non-oil-based fuel itself. The authors focus on two in particular, methanol and ethanol.

Their section on methanol is OK, but I still have questions around the feedstock. The throwaway line on biomass that “The raw material is plentiful, and most of the biomass products are currently discarded” is facile and unhelpful, as anyone familiar with the economics of biomass-to-power will know (the upshot is, not many plants are being built in the U.S. because the feedstock isn’t free and thus the economics aren’t great). Coal-to-liquids and gas-to-liquids technology is indeed fairly mature, but the current cost is not great and the GHG emissions (particularly for CTL) are terrible. And while the U.S. has abundant coal, their argument on gas sourcing is inconsistent, as I mentioned before.

Their section on ethanol is way off, and the worst part of the entire book. They correctly rail against ethanol tariffs, but become apologists for ethanol subsidies, calculating the absurd claim that $6 billion in ethanol subsidies reduced the world price of oil in 2008 by enough to save the U.S. alone $60 billion on fuel imports (“ten times the subsidy”). They misleadingly quote the energy requirements of gasoline vs ethanol:

The amount of fossil fuel in mega joules needed to make one mega joule of gasoline is 1.19 versus 0.77 for corn ethanol and 0.10 for cellulosic ethanol.

(That 1.19 includes the mega joule of energy from combusting the gasoline, which is why the wording is misleading.)

Even more flawed, though, is their thinking on the “food vs. fuels myth”:

How could that drastic increase in the price of food commodities from fish to rice possibly be attributed to ethanol? Nobody was growing corn in rice paddies or making biofuels out of fish.

Yikes. Agricultural commodities are largely fungible, like oil, from both the supply and demand sides. Increased corn cultivation reduces the land available to grow other cereals like wheat, and higher prices affect the prices of substitutes like rice as consumers shift their consumption patterns in response. Farmed fish are often fed corn, as are industrially-raised cattle, pigs and poultry, which are meat substitutes for fish. Ethanol is certainly not responsible for all of the rise in food prices, but the most credible estimates link it to ~30% of the rise in grain prices (more for corn and less for wheat and rice.

This lack of understanding infects their analysis of CO2 emissions from indirect land use change and deforestation. Yes, Brazilian sugarcane is not grown near the Amazon, but it displaces commercial soy and corn farming, which in turn displaces lower-value farming and ranching activities and pushes them further into the Amazon. Throughout the world, increasing food production in response to higher prices often means extension onto marginal land (which in turn is often forest). Yes, this effect is hard to measure precisely, but that does not mean it doesn’t exist.

Finally, like many others, they take a rich-world-centric view of food price increases (“farm commodity prices have almost no effect on the retail consumers”). Yes in America, but in poorer countries where food purchases account for >50% of income and raw commodities make up >65% of end food prices, that is not the case – food price rises have a huge negative impact on nourishment, nutrition and overall welfare, which is not something to be casually ignored.

Turning Oil Into Salt (4): Selective thinking on fungibility

The authors recognize that:

Oil is a globally traded, fungible commodity, so stifling U.S. purchases from the Persian Gulf and buying from other regions like Africa would just mean that someone else would buy more from the Persian Gulf with no impact on price and availability, certainly not on oil’s status as a strategic commodity.

This is one of their multiple (and good) arguments against the “Cheney plan” of diversifying America’s oil imports to non-OPEC countries.

Unfortunately, they don’t extend this level of understanding to their analysis of agriculture and biofuels:

How could that drastic increase in the price of food commodities from fish to rice possibly be attributed to ethanol? Nobody was growing corn in rice paddies or making biofuels out of fish.

Yikes. Agricultural commodities are largely fungible from both the supply and demand sides. Increased corn cultivation reduces the land available to grow other cereals like wheat, and higher prices affect the prices of substitutes like rice as consumers shift their consumption patterns in response. Farmed fish are often fed corn, as are industrially-raised cattle, pigs and poultry, which are meat substitutes for fish. Ethanol is certainly not responsible for all of the rise in food prices, but the most credible estimates link it to ~30% of the rise in grain prices (more for corn and less for wheat and rice).

Not only are the crops themselves fungible commodities, some of the resources required for production are fungible and, like oil, fundamentally scarce (think arable land and water). Like diversifying away from Middle Eastern oil, diversifying into biofuels is no free lunch.

Malthus on food speculation

The man who refuses to send his corn to market when it is at twenty pounds a load, because he thinks that in two months time it will be at thirty, if he be right in his judgment, and succeed in his speculation, is a positive and decided benefactor to the state; because he keeps his supply in that period when the state is much more in want of it; and if he and some others did not keep it back in that manner, instead of its being thirty in two months, it would be forty or fifty.

If he be wrong in his speculation, he loses perhaps very considerably himself, and the state suffers little; because, had he brought his corn to market at twenty pounds, the price would have fallen sooner, and the event showed that there was corn enough in the country to allow of it: but the slight evil the state suffers in this case it almost wholly compensated by the glut in the market, when the corn is brought out, which makes the price fall below what it would have been otherwise.

I think Malthus is wrong that any price volatility caused by speculation is benign, but otherwise this is an admirable defense of the beneficial role that much-maligned (today and throughout history) speculation can play in food markets.

The quote comes from Cormac Ó Gráda's excellent book Famine, which I am reading and enjoying, and specifically from his outstanding chapter on the famous Bengal famine of 1943-44.

Eventual Nobel Prize winner in economics Amartya Sen made his name with his 1981 work which made the case that the famine was caused primarily by lacking "entitlements" (e.g., many were too poor to purchase food at escalated prices), rather than by what Ó Gráda terms a "Food Availability Deficit" (FAD). In other words, the famine was an issue entirely of distribution, rather than sufficiency. And moreover, Sen blames the entitlement issues (i.e. high food prices) primarily on speculation, which I hadn't previously realized.

Ó Gráda marshals quantitative and qualitative arguments to make the case that there was in fact a significant Food Availability Deficit in Bengal, despite the strenuous efforts of the governing British to mask this. Such deficits themselves cause price rises, which lead to entitlement issues, but after reading Ó Gráda I'm convinced that, as he says, "the heavy focus of the literature on hoarding is misplaced."

Many views on fuel subsidies

There are a lot of possible angles on fossil fuel subsidies. It's easy to draw a chart like this that makes them seem way out of proportion with renewable fuel subsidies in the U.S.:

But there are also other ways to cut the data, like in proportion to total energy consumption, which give the opposite picture - according to Downstreamtoday.com, fossil fuels receive 71% of subsidies while providing 86% of U.S. energy in 2007 (94% if nuclear is excluded). Oil and gas taxes also far outweigh subsidies, so counting only the subsidies provides a distorted picture.

Geoff Styles argues that the "level playing field" argument is misleading, because of the strange economics of oil and other subsidies for renewables. He calculates the gasoline subsidy on a per volume basis and compares with that corn ethanol; ethanol subsidies are higher, unsurprisingly, but this alone isn't a generalizable argument, because corn ethanol subsidies don't have much support from greenies either.

Enough about the U.S., how about developing countries? According to The Economist, they spend an order of magnitude more than developed countries on fuel subsidies ($310 bn vs. $20 bn):

Ultimately, I find The Economist quite sensible on this issue - they applaud the recent G20 resolution to phase out fossil fuel subsidies (although being appropriately skeptical of the vague "medium-term" timeline) because ultimately, subsidized fuel consumption is both bad for energy efficiency and not particularly effective as an investment in poverty alleviation. And if developing countries eliminate their subsidies, the paying field levels, weakening at least one of the arguments against doing the same here in the U.S.

Heat and ag yield, cont.

Michael Roberts' great paper with Wolfram Schlenker on the effect of detrimental increased heat on crop yields (which I first excitedly blogged here) is getting a lot of (well-deserved) attention.

First, Ezra Klein cites the work in the Washington Post, and, despite a few minor inaccuracies by Klein, Michael is deservedly proud of having cracked into the mainstream media.

Second, the paper was noticed as far away as Brazil, whence came a letter to the editor challenging that Brazil (being warmer and high-yielding) refutes the Schlenker/Roberts conclusion, or at least casts into doubt its generalizability. But their quick-and-dirty analysis seems to corroborate their original findings:

First, they cherry picked the state and the year from Brazil. Mato Grasso is the highest yielding state in Brazil and 2008 was a remarkably good year for them, due to unusually good weather. Other states in Brazil have average yields that are about half those of Mato Grasso.

Second, Mato Grasso yields were higher than average yields in the U.S. as a whole, but not higher than the best yielding states in the U.S.

Third, if we narrow our comparison by looking a particular state—Illinois, the number two yielding soybean state in the U.S.—and also look more closely at the data, Mato Grasso doesn’t look much warmer. In fact, the southern half of Illinois, which has average yields comparable to those in Mato Grasso, also has comparable exposure to extreme heat. (Northern Illinois is cooler and higher yielding than both Mato Grasso and Illinois). This can be seen from careful inspection of the maps below (click the figure for more detail). It turns out that there aren’t any soybeans grown in the hottest part of Mato Grasso. It’s not clear whether the commenters took into account the locations in Mato Grasso where soybeans are actually grown.

Look for more back-and-forth discussion on this interesting area, and the Schlenker/Roberts contribution in particular.