This article was produced in collaboration with Investigate Midwest and Gray Television’s InvestigateTV as part of InvestigateTV’s series, “Secret Subsidies.” Republication is limited to print, digital, radio and non-commercial television outlets.
A U.S. Department of Agriculture program touted as relief for lost trade during the Trump-era trade war with China spent unprecedented amounts of money, bypassed Congressional approval and lacked checks to ensure the payments went to eligible farmers.
The Market Facilitation Program doled out more than $23 billion, more than all other USDA direct payment subsidies combined over the duration of the MFP.
But the USDA failed to make sure the money was going to the farmers who needed it, according to the Government Accountability Office, a nonpartisan government agency.
Over the course of the MFP, the USDA’s own internal auditors found the agency sent more than $800 million to farms “improperly” — including to recipients who didn’t qualify for aid.
To help farmers quickly, the USDA bypassed Congressional approval by funding the MFP through a little-known agency “credit card”: the Credit Commodity Corporation. The CCC, a government-owned entity designed to protect the farming industry, had never been used to fund such a single, large-scale program at the discretion of the Secretary of Agriculture before.
Unlike most farm subsidies approved by Congress as part of the Farm Bill, discretionary CCC-funded programs can be designed quickly and administered by USDA alone, making them more flexible.
But that flexibility came with side effects. It opened the door for industry groups to lobby the USDA directly — and they did so with much success.
And the GAO has criticized the way USDA designed the program.
The agency used an “inappropriately high” baseline to calculate payments, distributed unequal payments for farmers producing the same crop and lacked transparency in its methods, the GAO wrote.
Less than 4% of the funds went to historically underserved farmers including low-income, racial minority, veteran and beginning farmers.
Meanwhile, at least $163 million went to high-income farms and individuals making more than $900,000 per year.
“The MFP was a $23 billion program. So it’s a lot of money to go to about a half a million farmers,” said Steve Morris, director of natural resources and environment at the GAO, which produced three reports about the MFP. “In terms of the oversight that USDA conducted, what we found was that the way they conducted their oversight had a lot of limitations.”
USDA did not respond to multiple requests for comment over several months. The agency did respond to the GAO’s reports, agreeing with some of the findings but disagreeing with the GAO’s criticism of its methodology for calculating payments.
MFP began as a one-time assistance program in 2018 as part of a larger package aimed at addressing the negative effects of retaliatory tariffs on American agricultural products levied by China and others.
But in the summer of 2019, as the trade war continued, the USDA initiated a second round of MFP payments with different eligibility criteria and payment rates.
Lobbyists and industry groups helped decide some of those changes, especially who would benefit from the second round of funds, according to USDA emails obtained via a public records request.
Some policy experts say it created an expectation among farmers for more government help in the future.
“I think, without a doubt, this has set an expectation that whenever there are bumps — or perceived bumps — in the road, that the USDA will automatically trigger out payments in one form or another,” said Johnathan Coppess, an agricultural policy expert at the University of Illinois.
The USDA maintains a detailed internal dataset that includes ID numbers for individuals and partnerships, home addresses, farm addresses, crops grown, acres planted and more. But data available to the public is more vague.
For example, public spreadsheets don’t specify which crop a farmer received a payment for. They also don’t include the physical address of the farm — only the address to which the payment was sent, as well as the county where the farmer applied for MFP.
In an effort to provide some transparency, Investigate Midwest and InvestigateTV turned two dozen USDA spreadsheets into a map of payments made in the Market Facilitation Program.
This map displays the total amount of money received by each recipient over the course of the Market Facilitation Program. This map only displays payments made for non-specialty crops, which account for 94% of the funds spent through the MFP.
To explore this data further, click here. On this full-sized version of the map, you can zoom in and click on dots to see individual payment amounts. Scroll to the bottom of this page to learn more about how we made this map.
‘A lot of people were hurting’
The MFP payments made it easier for U.S. farmers to weather losing the country’s top buyer of corn and soybeans, said Brett Neibling, a soybean farmer and board member of the Kansas Soybean Association.
The impact of the tariffs on producers varied by location and by farm; Neibling had on-site storage that allowed him to wait for better prices and the option of selling his harvest as feed for domestic use. Other farmers didn’t have the same luxuries, he said.
While Neibling’s goal was “trade not aid” — a restoration of export markets instead of government payments — the MFP provided needed income.
“A lot of people were hurting and a lot of people were not going to be able to make it,” Neibling said. “And when they put out the Market Facilitation payment, we were concerned with it, but it was very necessary for us just to be able to help pay for input costs, whether it be for fertilizer, or chemical or diesel fuel to run our tractors.”
Few payments went to needy farmers
In January 2020, Sen. Debbie Stabenow, D-MI, requested that the GAO investigate the MFP, citing the methods used to calculate payments, the lack of correlation between trade damages and payments made and unequal distribution of aid.
The GAO’s three reports found only a small portion — 3.6% — of the funds went to “historically underserved farmers” including applicants who are low-income, racial minorities, veterans and beginning farmers.
The income limit for the MFP was $900,000, but there were exceptions.
If at least 75% of an individual’s income came from farming, they still could qualify for MFP payments. The limits also didn’t apply to farm partnerships — business organizations with multiple members.
According to the GAO, the USDA paid more than $163 million to high-income farms, with much of the aid going to farmers with adjusted gross incomes of more than $900,000 per year.
But the USDA didn’t independently verify whether an individual’s income came from farming, the GAO found. Instead, the agency relied solely on documents provided by attorneys or accountants.
Verifying income limits wasn’t the only example of USDA compliance monitoring that didn’t meet GAO standards.
The USDA conducted compliance spot-checks of 2018 MFP payments, but the GAO determined these checks were “limited in their usefulness” because the agency didn’t fully account for risk when deciding who to check.
While the USDA’s sample size was large, consisting of 35,000 farming operations, the GAO found it wasn’t looking in the right places — the agency didn’t focus on high-risk groups such as farms with no history of receiving subsidies, or payments above the established caps.
“The way they set up their oversight didn’t allow for reliable results,” the GAO’s Morris said. “In some cases, they oversampled populations and in other cases they undersampled, so it was just a lack of consistency in how they went about conducting the reviews.”
GAO found agency staff also didn’t adequately track the spot-checking effort — to the point the agency couldn’t even tell the GAO with confidence how many spot checks were done, making it difficult for auditors to understand the implications of the spot-check results.
Still, the Farm Production and Conservation Business Center, a section of the USDA that reviews agency programs, estimated that nearly $800 million in taxpayer funds were improperly paid out through the MFP, according to GAO’s review of agency financial reports.
The payments were deemed to be “improper” because they were distributed despite lack of evidence to support applications, because the applications were late or incomplete, or because they were not approved by the appropriate agency staff — all issues the USDA’s spot-checking process looks for.
After the second round of the MFP, USDA effectively stopped spot-checking payments, turning its attention towards the Coronavirus Food Assistance Program, according to the GAO.
USDA’s method for determining payments had issues, GAO says
Compliance monitoring wasn’t the only instance where the 2018 and 2019 versions of the MFP diverged — or that government watchdogs found to be lacking.
USDA changed the way it calculated payments from the 2018 to the 2019 version of the MFP, but the reworked system led to more issues.
Another GAO report determined that the method USDA used to set the payment rates for each crop in the second round of the program resulted in total payments to farmers that were higher than the estimated trade damages for that particular crop.
The issue stemmed from USDA using an inflated “baseline” value, according to the GAO.
To determine trade damages, USDA used an established economic model to estimate how exports for each agricultural commodity would be impacted by tariffs. This analysis produced a percentage decline for each crop; then, USDA multiplied that percentage by the baseline.
In the first year of the program, the GAO concluded the baseline, the value of the goods imported by the retaliating country in 2017, was justified.
For example, in 2017, China imported U.S. sorghum valued at $956 million — this was the baseline. USDA estimated that once retaliatory tariffs were imposed, U.S. sorghum exports to China would drop by approximately $314 million in 2018.
USDA then divided the total trade damages by the total production of sorghum in 2017, resulting in damages of $0.86 per bushel.
But when the USDA altered the program in 2019, the agency changed the baseline it used.
In the second year of the program, the USDA looked at the previous decade of crop exports and chose the highest trade value each crop had during that time period.
For example, the value of cotton exports to China varied greatly from 2009 to 2018, from a low of $514 million to a peak of $3.7 billion in 2012.
USDA used the peak value from 2012 to estimate trade damages for 2019.
The baseline value USDA used to estimate the impact of Chinese and European Union tariffs on corn combined the highest value of corn exports from those countries from different years. So, $1.65 billion in 2012 from China, and $349 million in 2018 from the EU, for a baseline value of $2 billion.
The GAO determined this methodology exaggerated trade damage estimates for the vast majority of crops. The USDA disagreed and claimed its process was based on established economic modeling.
The USDA did not publicly explain its reasoning behind choosing this methodology out of several options provided by agency economists.
“As a result, USDA increased its 2019 trade damage estimates in a manner that was not transparent to decision makers and the public,” the GAO report stated.
But the baseline trade damage wasn’t the only departure in methodology between the first and second versions of the program.
For 2019, the USDA made other changes to the way it disbursed the MFP money, including basing payments not on the crop produced, but instead on the geographic location of the acres planted.
For example, in 2018, every corn farmer in the nation qualified to receive $0.01 per bushel of corn produced that year.
That system drew criticism that it unfairly benefitted one crop over another. Soybean farmers, for instance, received the highest payment rate per bushel of all non-specialty crops. The USDA said it was concerned the program could influence farmers’ planting decisions.
The following year, the USDA calculated MFP payments for farmers based on a rate assigned to the county they farmed in, rather than the crop they planted. This meant every farmer in a county with a qualifying crop qualified for the same payment amount per acre.
But the new system traded equity across industries for inequity across regions. Under the new system, Southern and Midwestern states received higher payment rates on average.
In some cases, farmers who grew the same crops in adjacent counties received vastly different payment rates — meaning all things held equal, one farmer would receive more money for the same harvest than his neighbor.
“I think for a lot of farmers, it really did not make a lot of sense to base the payment rate for the Market Facilitation Program on your county,” said Anne Schechinger, Midwest director and food and agriculture economist at the Environmental Working Group, a nonprofit environmental advocacy organization that has criticized the MFP. “A lot of times, you’re going to have similar yields. You’re gonna be growing similar crops. So to base the payment rate on a county really didn’t make a whole lot of sense.”
The GAO found “decoupling” the payments from crops also skewed payments in a way that meant some crops received more aid than estimated trade damage, while others received less than damage estimates.
In response to the GAO’s findings and criticism of the structure and baseline of the MFP, the Office of the Chief Economist for the USDA argued that it was the job of the economist to present policymakers with a range of options for determining the size and distribution of payments — not to direct the policymakers in any one direction.
But the GAO noted that some of the options the chief economist provided — including the methodology that USDA chose for the second round of the MFP — did not meet the agency’s quality guidelines.
Political forces shaped second round of MFP, emails show
On May 1, 2019, U.S. negotiators met with Chinese officials in Beijing to work out a trade agreement that would stop the ongoing trade war and stabilize global markets.
By then, the USDA had already distributed nearly $12 billion in MFP payments to farmers whose crops were directly impacted by retaliatory tariffs, according to Investigate Midwest and InvestigateTV analysis of payment data.
But on May 3, 2019, Chinese officials backed out of the tentative deal.
In response, President Trump announced another increase in tariffs on more that $200 billion in Chinese goods, sending the two countries into another period of intense trade disputes.
Soon after the deal with China fell through, former Secretary of Agriculture Sonny Perdue promised farmers more help was on the way.
“This is Market Facilitation Program number two,” Perdue said in a May 17, 2019 interview with InvestigateTV’s partners at the Gray Television Washington News Bureau. “We learned a lot last year… It will build a lot on that, although we'll look at some of the rough edges that we had last year, and some of the stakeholder comments, and try to take those into consideration and have a better, well-rounded program this year for our farmers.”
Agriculture industry associations quickly took the opportunity to lobby for changes to the MFP that would result in more payments to their constituents, according to emails obtained by Investigate Midwest and InvestigateTV.
USDA staffers forwarded letters from the Minnesota Canola Council and the National Alfalfa and Forage Alliance to Deputy Secretary of Agriculture Stephen Censky. The Almond Alliance and Rolling Plains Cotton Growers also submitted letters to USDA requesting relief under the next round of the MFP.
“As you know, China’s retaliatory tariffs are having an effect on all commodities and all farmers,” Minnesota Canola Council chairman Tony Brateng wrote in a letter. “Though not all commodities have been retaliated against directly, the overall effect on the market has occurred across the entire agricultural sector in the form of depressed prices.”
Lobbyists representing the National Barley Growers Association, U.S. Canola Association, National Sunflower Association, USA Dry Pea & Lentil Council and U.S. Dry Bean Council — crops not directly impacted by tariffs — also sent letters from the organizations to USDA arguing that their farmers should be incorporated into the MFP program.
Perdue’s chief of staff, Joby Young, thanked the organizations for their input and promised to keep them in the loop.
John Gordley, head of the lobbying firm representing the trade organizations for barley, canola, sunflower and dry legumes, wanted a call with the deputy secretary Censky. That call took place on May 21, 2019, according to an email in which Gordley thanked a USDA staffer for setting up the call and continued to advocate for his clients’ inclusion in MFP talks.
Ultimately, the second round of the MFP included payments for the crops that industry groups advocated for through letters to USDA.
Of those crops, only two — cotton and shelled almonds — had been eligible for payments the previous year.
In May 2019, Congress also passed a disaster relief bill that included a provision to remove the cap on the income limit of $900,000 per year for the MFP program, as long as at least 75% of an individual’s income was derived from agriculture.
According to notes from the USDA’s Office of Congressional Relations, removing the income cap “would help utilize unspent funds for pork, dairy, almond, and cherry producers, primarily.”
Funding for the MFP came from an unusual source, started trend
To fund the MFP, the Trump administration bypassed Congressional approval of the program by using the Commodity Credit Corporation, a “wholly-owned government corporation” housed under the USDA that was designed to help stabilize the U.S. agriculture industry.
CCC has the authority to borrow up to $30 billion from the Treasury at a time. The secretary of agriculture has broad, discretionary power over CCC. Using those powers, former Secretary of Agriculture Perdue initiated the MFP.
By law, Congress has to foot the bill, appropriating the amount of money that the CCC borrowed in the previous year.
Before the MFP, the CCC rarely had been used to fund programs outside of the Farm Bill. But after the MFP, the entity has been used to fund two more large-scale USDA discretionary programs without Congressional input, including the Coronavirus Food Assistance program.
Read more from InvestigateTV's reporting on government subsidies in agriculture
“Historically, it's unique in terms of usage of the Commodity Credit Corporation Charter Act authorities,” said Coppess, the agricultural economist at the University of Illinois. ”It's unique in that it first came about in the middle of a Farm Bill discussion but didn't impact the Farm Bill discussion, and it's unique in the sheer amount of money that was paid out.”
But that may change as Congress approaches the next Farm Bill. The Congressional Research Service noted in a March report that the size and administrative differences seen with the MFP may impact consideration of other safety net programs, since USDA’s spending of billions of dollars on discretionary programs in recent years influenced the overall agricultural economy.
Congress can choose to amend the CCC Charter Act to restrict the discretionary spending of the secretary of agriculture, and it did so from 2012 to 2017 after the Obama administration received backlash for a $348 million disaster assistance program.
“I'm trying to find a word that isn't ‘unprecedented’ because everybody seems to use ‘unprecedented’ a lot,” Coppess said. “But this one actually fits very, very squarely within the term ‘unprecedented.’”
Data available to public missing key information
In addition to the GAO, both the USDA Office of Inspector General and the Congressional Research Service have released reports about the inner workings of the MFP. As government agencies, these groups have access to the most detailed data available within the USDA.
Publicly available data, however, is missing important items, including details about partnerships — formalized groups of farm owners.
Partnerships are able to largely avoid payment caps and as a result are some of the highest-grossing recipients of subsidies like the MFP.
An exemption to the Freedom of Information Act that blocks the disclosure of individuals who make up these partnerships was added to agriculture legislation in 2008.
Schechinger, food and agricultural economist* at the Environmental Working Group, said her organization has attempted to determine where exactly the payments went, but the lack of specificity in the data provided to the public makes it difficult for outside observers to study the true effect of the payments made in the program.
“It's really important for taxpayers to know if our taxpayer money is going to these high-income millionaire farmers,” Schechinger said. “Then, on the other end, it's just really hard to know who's actually getting this money.”
A May 2021 InvestigateTV report found that farm partnerships obscure the identities of members and make it difficult to determine if those members are “actively engaged” in farming — a requirement for individuals receiving USDA payments.
Farm partnerships allowed groups of farmers to bypass MFP payment caps because the USDA considers general partnerships to be a group of individuals or entities, not an entity itself, meaning each member is able to receive up to the maximum payment.
A data analysis by Investigate Midwest and InvestigateTV found several examples of partnerships receiving millions of dollars, well above the combined cap on payments, which was $325,000 in 2018 and $500,000 in 2019.
For example, the three partnerships connected to the DeLine family in Missouri — DeLine Farms Partnership, DeLine Farms North and DeLine Farms South — received more than $7.9 million combined over the period of the program.
Robert Serio, an attorney representing hundreds of farm partnerships including the DeLine group, said these partnerships are following the rules as established.
“It's not taking advantage of the rules, because the rules are the rules,” Serio said. “It's just like when I deduct an expense item off of my business. I do that because I'm allowed to. There's a rule that says that. They could take away that rule and I couldn't do that anymore. I don't say that's ‘taking advantage’ of the rule, I say that's ‘complying’ with the rule.”
MFP already setting precedent for other aid programs
As the trade war was coming to a resolution in late 2019, another crisis was quickly approaching — the coronavirus pandemic.
The pandemic sent global markets into a tailspin, prompting another wave of emergency aid for farmers who couldn’t sell their products.
While individuals received stimulus checks from the federal government, so did farms, via the Coronavirus Food Assistance Program.
CFAP not only overlapped with the last year of the MFP, but also took a page from the program’s playbook by using the CCC to fund a large-scale aid program without Congressional approval.
CFAP paid out nearly $30 billion over two years.
The trend has continued into the new administration; a recently announced Biden-led program to promote climate-smart agriculture will use nearly $3 billion in CCC funding.
“That's a third iteration of this Commodity Credit Corporation money that does not need congressional oversight and is just sending money from an administration to farmers,” Schechinger said. “The Market Facilitation Program really was the first instance that set this large precedent — that future administrations can send money to farmers from the CCC for whatever reasons they want to send money to farmers.”
Other aspects of the MFP also carried over to CFAP.
Compared to previous farm aid programs, the GAO found the USDA was more lenient with farmers who misrepresented the number of acres they planted on their MFP applications.
Historically, the USDA would consider farmers who overrepresented the number of acres they planted by 10% or more to be out of compliance. For MFP, that bar quietly moved to 15%.
Instead of reverting back to the 10% level, the next large-scale hardship program for farmers — CFAP — also used 15%.
For example, if the USDA found that a farmer planted 100 acres of soy but reported planting 115 acres on their MFP application, that farmer would have to repay the payments received for the extra 15 acres, but would not face any additional consequences.
Individuals who commit fraud by lying on applications for aid are subject to up to 10 years of imprisonment or a fine not to exceed $10,000.
But the MFP rules didn’t include a clause that gave the Farm Services Agency, the department of USDA that ran the program, the authority to enforce the rules, meaning that even those who were found to be noncompliant with the program requirements, who “showed a lack of good-faith effort,” did not face penalties, the GAO found.
With the current Farm Bill expiring in September 2023, lawmakers could choose to make changes to aspects of USDA initiatives or the CCC Charter Act to increase transparency and oversight of future farm aid programs.
“Cracking down on this in the next Farm Bill would be really great,” Schechinger said, “to make sure that the money is only going to people who actually farm instead of people who are farming the system.”
How we created the map
The data forming the basis of our data analysis came from 24 USDA spreadsheets that contained data for all payments made through agency programs from 2018, the first year of MFP, until 2020, when the last payments were disbursed.
We combined the spreadsheets into a single database and filtered out everything but the MFP. Within the MFP, there were separate rules governing payments for dairy, livestock, specialty and non-specialty crop producers. For simplicity, we decided to focus on non-specialty crops, which received 94% of the money disbursed in the MFP.
This left us with more than 3 million rows of data, with each row describing a single payment issued in the program. MFP payments were issued in several installments, rather than a lump sum, and we wanted to know the total amount paid to each individual over the course of the program.
Using SQLite, a programming language, we grouped the rows by name and address, leaving us with around 754,000 unique name and address pairs, along with the sum of all the payments made to that recipient. This was the data we geocoded and used in our interactive map.
But with such a large dataset behind the map, the tiles forming the map were too large. To trim down the tile sizes, we limited the visibility of data at certain zoom ranges. That means some points are only visible at certain zoom ranges.
*This story originally had the incorrect title for Anne Schechinger, of the Environmental Working Group, in one reference to her in the story. The story has been updated with her correct title. We regret the error.
Top image: photo by InvestigateTV