On Farms Big and Small, Putting Blockchain to the Test
10.05.2018

Earlier this spring, 210 people across the U.S. were sickened from eating E. coli-laden greens, resulting in five deaths. Because it can take weeks for E. coli to produce symptoms, the Food and Drug Administration didn’t immediately know there was a problem.

Once discovered, it took a week for the outbreak to be traced to Yuma, Arizona. Immediately, food safety agencies began warning consumers not to eat lettuce from the region. The problem is, 90 percent of all leafy greens sold during certain times of the year are grown in Yuma. A VP of food safety for California grower Taylor Farms told the Wall Street Journal, “Trucks all across the country were dumping romaine.”

Despite warnings to customers and retailers to throw out any Yuma-grown romaine, it’s so difficult to trace food after it leaves the farm that the outbreak was only considered officially over months later. The harvest season had ended, the CDC wrote, so the contaminated lettuce that made people sick was no longer be available.

Most of the lettuce that was thrown out was probably perfectly fine. During a major 2006 spinach recall, the FDA found out after two weeks that the bad spinach had only come from one area of one farm—though not before the price and amount of spinach purchased nationwide dropped to half of the pre-recall levels. Sickened customers still had the bags of spinach they’d purchased but the barcodes weren’t much help since they didn’t show where every lettuce leaf came from.

Some hope it can provide more transparency about where food or crops come from, who grows them, and what pesticides were used. Others see blockchain as a way to streamline regulation to make trade easier in our global economy.

With food safety scares like the killer romaine lettuce in recent memory, companies are positing that tracking products with blockchain will make it quicker and easier to trace infected foods to their source. That means less illness and less food wastegood for consumers and retailers alike. Some hope it can provide more transparency about where food or crops come from, who grows them, and what pesticides were used. Others see blockchain as a way to streamline regulation to make trade easier in our global economy.

Those are a lot of hopes to pin on a technology that’s essentially a ledger of transactions that can’t be tampered with. Yet everyone from cotton farmers in Haiti to Walmart, America’s largest retailer, are being seduced by the possibilities.

When footwear and clothing maker Timberland and the Smallholder Farmers’ Alliance (SFA) got involved with Haiti in 2010, it was part of a mission to help reforest the island. Timberland executives saw the success of the program and started wondering whether it could move from a charitable relationship with Haitian farmers to something more commercial. In addition to the social good, as a “least developed country,” Haiti can also trade with developed countries duty and import free. It would be a win for both Timberland and Haiti.

In the last few years, more than 36 major brands, including Timberland, have pledged to use 100-percent sustainable and/or organic cotton by 2025 or earlier, and many are trying to find places to source that cotton from. Up to 70 percent of the global cotton supply is grown by smallholder farmers, who often use pesticides and genetically modified varieties. And because there can be so many middlemen between farmers and buyers, a lot of sustainable cotton gets lost in the shuffle. A report by Impact Farming on export potential for cotton in Haiti found that as much as 83 percent of cotton “as a consequence of poor tractability is mixed in and ‘lost’ with regular cotton.” While Timberland was interested in Haiti’s cotton, the social impact marketing angle also held a lot of appeal. But you can’t tell that story if you can’t track whether the cotton was ethically sourced.

While Haiti has not grown cotton commercially since the 1980s, that has more to do with politics and the rule of dictators than Haiti being a bad place to grow cotton. “The specific issues that led to the demise of cotton are gone,” says SFA co-founder Hugh Locke.

“Cotton did not stop because the land couldn’t produce it,” says Timote Georges, the other SFA co-founder who was born and raised in Haiti. “Farmers have good memories about cotton.”

If all went well with cotton production, Timberland had guaranteed to purchase up to one third of Haiti’s supply. But the SFA was being stretched thin. “The whole thing was getting too big to manage as a nonprofit,” Locke said. An early project of planting trees had turned into an accidental “smallholder supply chain corporation,” as Locke describes it. “The moment you do that, you need to have a handle on data.”

For most smallholder farmers, including those in Haiti, farming wasn’t about tracking expenses versus income but whether you had enough to buy food and send your kids to school. Timberland and the SFA needed to not only make it easy for farmers to track their operating costs but also for Timberland to track product quality and origins.

To figure out the best way to do that, the SFA turned to Columbia’s School of International and Public Affairs, where students did an analysis of existing data management systems to find whether anything currently in use could match the needs of both farmers and buyers. Blockchain technology, which was gaining buzz from much bigger players in the industry as a solution for supply chain problems, was considered from the very beginning.

The agricultural industry, especially for farmers, is one with tight margins. The difference between a good year and a disastrous one can hinge on many factors outside farmers’ control: attacks from pests, a drought or unseasonably late snowstorm. The more streamlined the supply chain becomes and the more precisely farmers can water or fertilize their fields—whether growing food, cotton, or some other crop—the more money they can make. Sylvain Charlebois, professor of food distribution and policy at Dalhousie University says that blockchain could “‘Uberize’ the agri-food sector by eliminating middlemen and lowering transaction fees.” This could lead to faster payment and fairer pricing.

“Ideally, some of this transparency could potentially stand in for certification.”

There’s also hope that blockchain could eventually be a stand-in for expensive certification programs—giving farmers the premiums of an organic label without the cost to sign up. Most certification systems including organic and fair trade need inspections to make sure people are doing what they say they are on paper. Some small farms in the United States already find that major certifications like USDA organic are too onerous—financially and otherwise—and rely instead on selling at places like farmers markets, direct to consumer. That way instead of customers reading a label to find out how the farmer grew their produce, they can just look their farmer in the eye and ask. If blockchain systems started sharing that information with customers, perhaps it could create that same sense of trust digitally.

“Ideally,” says Atlanta McIlwraith, Timberland’s community engagement manager, “some of this transparency could potentially stand in for certification.” It’s too early to say for sure. One problem is that all that information people are so gleeful about blockchain’s potential for collecting, might just be more than people want to hear. “If you lead with the facts, you alienate people who don’t have time to deal with all of that,” she says. “I would expect us to lead with stories and powerful images and then, you know, allow people to dig deeper.” In other words, for circumventing certifications using nothing but transparency to work, people would have to take the time and have the knowledge to read all that information and know what it means. And in all likelihood, whether shopping at REI or Whole Foods, customers just won’t be willing to stop and read up before adding an item to their cart.

“We’re generating more data about a given crop than anyone can absorb,” says Locke. Companies will have to figure out the best format to give that information to their customers and how much of it to make available. “But that’s not our problem. Our challenge is to generate the most data we can and have it available for anyone who wants it.”

Last year Walmart joined forces with nine companies—Dole, Unilever, and Tyson Foods among them—to test how to apply blockchain to their existing supply chains. IBM is doing the technical work for the project, hoping to sell subscriptions to a blockchain-meets-the-cloud management system. For Walmart, which recently required all its leafy greens suppliers to implement blockchain technology by September 2019, its interest is more practical than idealistic. Walmart has 12,000 stores around the world, and works with tens of thousands of suppliers. Tracing items when something goes wrong is a logistical nightmare.

“The food system has one major Achilles heel,” Frank Yiannas, Walmart’s vice president of food safety said earlier this year at MIT’s Business of Blockchain conference. “A lack of transparency in the food system.” Not because customers want to know where their food comes from or because Walmart can better market a mango to a customer if it tells people that their fruit was grown by Jorge from Sinaloa. Transparency in the food chain is important because when E. coli-laden romaine lettuce was sickening people, the most food safety agencies could tell consumers was not to eat romaine grown in Yuma, Arizona. 

Yiannas describes the project with IBM as building a “blockchain ecosystem,” allowing companies to access multiple data points in the supply chain instead of the drawn-out game of telephone that food traceability often looks like today. In a trial where Walmart traced mangos to the source, it took six days, 18 hours, and 26 minutes to get that information by traditional means; using blockchain it happened in only 2.2 seconds. 

Which is not to say that blockchain is fail-proof. Like most technology, blockchain is only as idealistic as the people or organizations harnessing it. “Blockchain is a team sport,” Yiannas noted at the MIT conference. Not only does it require a lot of players to get the game going, it only works if those players don’t cheat.

“Blockchain can be sexy and interesting but the pre-work needs to be done so the automation at farm level and processing can be done correctly.”

“Blockchain is fantastic in terms of creating that accounting paper trail,” Mitchell Weinberg, founder of food fraud and prevention firm Inscatech, told The New Food Economy. But it doesn’t show you how much pesticide is actually being used or prove that a strawberry is really coming from the farm it says. It’s one thing to use blockchain to track a digital currency—another to track an object that exists in the real world. “There’s just no validation of the information that’s being put in. It’s only as good as the person who’s entering the information,” Weinberg said.

The world of food and farming is already swimming with labels—organic, local, non-GMO, gluten-free, made in the USA—and fraud is still rampant. Olive oil is often cut with inferior products made from anything from corn to walnut oil. One study found that as much as 47 percent of the fish served in sushi restaurants was mislabeled. More organic food is sold every year than is actually grown. Transparency only works if companies want to be transparent.

Getting smaller operations on board with blockchain is also a challenge, says Robert Johnson, the faculty advisor for the Columbia project that helped SFA settle on blockchain as a long-term data management strategy for cotton. Getting the farmers to a point where they can input the kind of transparent data companies seem excited about is a multi-step process.

For a digital system to be convenient and easily readable—whether blockchain or otherwise—there was some behind the scenes work that had to take place first.

Many organizations working with smallholder farmers are not using data in an efficient way, Johnson says, and that’s before adding blockchain to the mix. “Most of the instances where you’re seeing blockchain deployed, it’s by a business or large farm or organization,” Johnson says. There’s a reason why Walmart is among the first major companies to look into blockchain. “They have the technical systems and data already in place,” Johnson says. “Blockchain can be sexy and interesting but the pre-work needs to be done so the automation at farm level and processing can be done correctly.”

As it’s currently being used, blockchain in agriculture is more akin to companies inputting information into a shareable Google doc with limited editing privileges than groundbreaking technology.

But it’s still new and the story doesn’t have to end there. Even if all blockchain were able to do is prevent food safety outbreaks, it would be a worthwhile investment in technology. SFA’s project, if successful, will show that blockchain in agriculture isn’t just for big companies. With over 500 million smallholder farms in the world, most of them in developing countries, allowing people who have been sized out of global markets to take part is an important goal.