Hand seeding innovation

I my conversation with agronomy professor Bill Raun while doing interviews for Unnaturally Delicious, he informed me of a major agricultural problem which I'd never before heard about.

“He estimates that 60 percent of maize in the developing world is planted by hand. That’s more than seventy one million acres on Earth where poor, often subsistence, farmers use long sticks to poke a hole in the ground and drop in three or four kernels of corn before moving a foot or so and repeating the process again—and again and again. This is imprecise, backbreaking work, and potentially deadly.

According to Raun, many of these rural farmers have access to high-quality hybrid seeds, but the seeds have been pretreated with fungicides and insecticides. The treatments protect the vulnerable seedlings from insects and disease, but chemically treated seeds weren’t meant to be routinely handled by the farmer.”

Here are a few photos he took of hand-seeding in action.

Fortunately, he's got a potential technological solution (see also his web page)

“Raun again teamed up with engineers to create what he calls a GreenSeeder—a handheld device that can be loaded with seed and reliably deliver a single seed with each poke in the ground. Raun optimistically estimates the device could boost yields by 25 percent, resulting in $2 billion of extra revenue to the developing world if farmers abandoned their wooden sticks in favor of his mechanical poles.”

Precision Agriculture

Chapters 6 and 8 in Unnaturally Delicious were two of the most enjoyable to write because I got to learn about some of the amazing things going on in my own town and university.  

Chapter 6 focuses on the story of David Waits, who created SST Software, a company focused on data management solutions for farmers, particularly geo-spatial information.  While most farmers are today accustomed to seeing colorful yield maps showing which areas of the field are providing more and less grain, most food consumers have no idea of how much information and complexity goes into running a modern farm.

They key for farmers is to combine yield information with spatially-lined information on varieties planted, soil characteristics, etc. so that more precise decisions can made, for example, on fertilizer applications (to help prevent nitrogen runoff and increase yields).  

I write:

“Today SST Software is one of the leading agricultural suppliers of geographic decision support tools in the world. SST houses data on more than 100 million acres of farmland in twenty-three countries from Australia to Africa. At the heart of the operation is software that consists of relational databases that link information about the use of farm inputs to geographic identifiers and to site-specific information about soils, moisture, and much more. Some farmers can use SST’s software directly themselves, but given the size and complexity of today’s farms, the rapid pace of technological change, and the expertise needed in entomology, agronomy, and economics, many farmers rely on consultants to help make management decisions. As a result, SST’s biggest clients are crop consulting companies like Crop Quest and Servi-Tech and seed and chemical suppliers like Monsanto and Helena Chemical. These companies often work with farmers to send to SST information on soil samples, pesticide and fertilizer applications, yields, insect scouting reports, and seed varieties planted. The companies use these data to make site-specific fertilizer or planting recommendations. For example, based on his company’s agronomic models, an adviser might use SST software to identify which areas of a field should receive which kinds of fertilizer and in which amounts—a recommendation that can be sent electronically to a variable rate spreader that communicates with satellites to determine when and where to apply which mix of fertilizers. Given the high cost of seed, new variable-rate planters are also coming on the market. A thumb drive loaded with a recommendation from SST can be plugged into a planter, which can plant two different corn hybrids at different seeding rates and at different depths throughout the field. SST doesn’t make recommendations; it provides the mechanism for translating an agronomist’s recommendation into an action plan.”

I had the chance to play around with their software a bit myself, and it is truly amazing how many possible decisions a modern commercial farmer today has to make - something many critics of modern agriculture scarcely comprehend.  Here's what I had to say after talking about different decisions on seed choice and fertilizer choice:

“Each field might have 100,000 possible items of information linked to it in a given year, and that doesn’t begin to count the combinations of management decisions the farmer might have to make when mixing a particular type of seed with a particular soil type and a particular fungicide.”

Later in Chapter 8, I also talked about Bill Raun, an agronomy professor at Oklahoma State University (more on him later) who is also working on precision agriculture applications.  Along with some engineers he created the GreenSeeker, which is:

“is a handheld device that senses the color of the plants’ leaves and, along with other information—such as the date the crop was planted—provides a recommendation of how much nitrogen to apply to satisfy the plants’ needs. The sensors can also be placed atop a tractor or fertilizer applicator so that, as a farmer drives through a corn or wheat field, the amount of nitrogen applied changes in response to what the GreenSeeker is “seeing.” Today several companies make the GreenSeeker and related technologies commercially available to farmers all over the world. Adoption has been limited by the cost of the technology, lack of precision, and the low cost of alternatives, like simply applying a uniform rate of nitrogen throughout portions of the field before planting.”

Bill has a lot of amazing pictures at his website, here is one with several GreenSeeker's attached to a fertilizer applicator.

How bad, really, is meat eating?

Last week, Scientific American published a piece on how to get people to eat less meat.  Apparently, the science is settled and we now only need to come up  with the right "messages." 

There are an awful lot of apocalyptic pronouncements about the adverse effects of meat eating on the environment.  In a widely viewed TED talk, Mark Bittman likens meat production to a nuclear explosion and says it is leading to a "holocaust of a different kind," pointing directly to impacts on climate change.  As another example, Bill Maher, comedian and host of an HBO talk show, has written, “But when it comes to bad for the environment, nothing—literally—compares with eating meat. . . . If you care about the planet, it’s actually better to eat a salad in a Hummer than a cheeseburger in a Prius.”   I could, quite literally, provide dozens of these sorts of quotes from well known journals, writers, actors, etc, but I think you get the point.  The overall message is pretty clear: we should become increasingly more vegetarian.  

Let's take a look at one of the outcomes people are most worried about and where externality is relatively clear: climate change.   I went to the EPA's calculations, and surmise the following carbon equivalent impacts for the US attributable to beef cattle, swine, and poultry production during the year 2014 (MMT is million metric tons):

• beef cattle: 116.7 MMT C02 (from digestion) + 4 MMT C02 (from waste management) = 119.7 MMT C02;
• swine: 1 MMT C02 (from digestion) + 22.4 MMT C02 (from waste management) = 23.4 MMT C02; and
• poultry: 0 MMT C02 (from digestion) + 3.2 MMT C02 (from waste management) = 3.2 MMT C02.

Elsewhere, the EPA suggests using using a social cost of carbon of $36/metric ton of C02 (assuming a discount rate of 3%) for cost benefit analysis and rule making.  Multiplying the C02 impacts above by the price tag of $36 implies the following total carbon costs in 2014.

• beef cattle: $4.3 billion;
• swine: $842 million; and
• poultry: $115 million.

That seem like a lot, but keep in mind that we also eat a lot of meat.  Data from the USDA suggests that in 2014, farmers/ranchers in the US produced 24.32 billion pounds of beef, 22.86 billion lbs of pork, and 44.98 billion lbs of poultry.  Putting the carbon costs on a per-pound basis (i.e., dividing total carbon cost by pounds produced), suggests the following.

• beef: $0.177/lb;
• pork: $0.037/lb; and
• poultry: $0.002/lb.

I don't know about you, but those don't seem like enormous costs.  Let's think about it a different way.  Suppose you wanted to "internalize" the the impacts you're having on climate change by altering how much beef, pork, and poultry you buy.  To do this, take the price you see at the grocery store and add about $0.18/lb to the price of beef, $0.04/lb to the price of pork, and less than a penny to the price of poultry, and act as if these were the prices actually being charged.  Would you change your behavior much based on such price increase?  If not, we'd could say the climate impacts are relatively small.

I saw the following from the economist Bob Lawson on twitter this weekend.

If you have ever said "externality" but don't understand that the optimal amount of pollution >0, then you're doing it wrong.

— Bob Lawson (@doclawson12) April 17, 2016

The key isn't to have zero greenhouse gas impacts, but rather to to make sure you're taking into account the cost of those impacts.  For the case of beef, that means acting as if the price were about $0.17/lb higher.  Do that and you can shop away, guilt free (well, at least the guilt of carbon impacts).

P.S.  USDA data suggests the retail price of beef in 2014 was around $5.60/lb.  An $0.18/lb price increase would represent about a 3% increase in the price of beef.  Assuming the own-price elasticity of demand is, say, -0.6, this price increase would lead to a 1.9% reduction in the quantity of beef demanded. So, to internalize the carbon impacts of beef eating, Americans would reduce beef consumption by 1.9%.   It's not zero but its a far cry from vegetarianism.  

Food Demand Survey (FooDS) - April 2016

The latest edition of the Food Demand Survey is now out.

From the regular tracking portion of the survey:

• Willingness-to-pay for all food products was down this month, with the largest drop occurring for hamburger;
• The was a sizable uptick in consumers' anticipation of price increases for beef, pork, and chicken and a slight reduction in planned purchases of these items;
• Expenditures on food away from home were up about 5%; and
• We added "cancer and meat consumption" to the list of, now, 18 times for which we track awareness and concern.

Three new ad hoc questions were added this month.

The first set of questions was meant as follow-up to a survey Politico recently conducted of "food experts" (I was a participant in their survey).  Politico asked the following question to the experts: "Are the presidential candidates doing a sufficient job in the campaign discussing the future of food policy?" A whopping 97% said "no".  

I posed a related question to the respondents of FooDS.  Rather than just asking about the issue as a stand alone question, I put food an agricultural policy in the context of other issues candidates spend their time talking about.  In particular, participants were asked: “Are the presidential candidates spending too much or too little time discussing each of the following issues?”  

A list of nine issues was provided, which included “food policy” and “agricultural policy.”  Only about 6% of respondents thought too much time was being spent on the two issues.  44% and 46% thought about the right amount of time was being spent on agricultural and food policy, and 50% and 47% thought too little time was spent on agriculture and food, respectively.  

Immigration policy was the only issue for which more respondents thought candidates were spending too much vs. too little time.  Except for food and agricultural policy, the largest fraction of respondents thought the candidates were spending the right amount of time on the other issues.  

Secondly, we asked another question - this time exactly as it was asked in the Politico food-expert survey.  In particular, participants were asked “Should the government’s role in regulating the US food system be more active, less active, or the same?”  Here, our respondents lined up closely with Politico's food experts:  

Over half of the participants (59%) believe that the government should become more active in regulating the US food system, while less than 13% of participants believe the government should be less active in regulating the US food system.  This is consistent with other research that suggests consumers tend to be rather "statist" when it comes to food policy.

Finally, based on a suggestion from Jason Winfree at University of Idaho, who passed along an article about the (sometimes unjustified) negative perceptions of frozen food,  respondents were asked whether they agreed or disagreed with a list of nine statements related to the tastiness, affordability, and health of fruits and vegetables that are either fresh, frozen, or canned.  

In terms of taste, fresh rated higher than frozen, which was rated higher than canned.  All three had a mean score above 3, meaning respondents were more likely than not to agree that all three types of fruits and vegetables were tasty.  In terms of affordability, the ranking was exactly reversed with canned being perceived as most affordable and fresh least affordable (although all three were far about the mean of 3, implying most consumers though all three were affordable.  Finally, perceived health lined up almost exactly with perceived tasted: fresh was perceived as healthier than frozen which was perceived healthier than canned.

Farm subsidies by state

This post by Kevin Patrick at farmdocdaily alerted me to an interesting data visualization tool created by the USDA Economic Research Service (see the bottom of this web page for a three different tools).  

I used the "get to know your state" tool and pulled up Oklahoma.  I was a bit surprised to see that in 2014, total government payments to the state were over $1 billion while at the same time net farm income was $2.8 billion.  If I'm reading this right, it implies 36% of net farm income in OK in 2014 was government payments, and Oklahoma ranked 2nd in government payments while ranking only 23rd in sales.  Thinking this might be an aberration, I pulled up 2013, and then "only" 19% of net farm income in Oklahoma was government payments.  

Looking at the supposed "big government", California (which has the most ag sales of any state), only 1.5% of its net farm income was from government payments in 2014.  

The "farm income atlas" tool lets you plot government payments (and other components of income) for each state. Here's total government payments for 2014. 

There's a lot more available in the tools.  Check it out yourself.