Asymmetrical Information - Megan McArdle

04.26.13

How a Discrimination Settlement Turned into a Bonanza for Fraudsters

Farm loans. Sympathetic administrations. And a judge who didn't worry about false positives.

A New York Times story this morning suggests a pattern of massive fraud and abuse, abetted by a complaisant government, in government settlements that were supposed to compensate for discrimination by the USDA against minority farmers.  

Long story short: black farmers complained that they had been discriminated against when seeking loans from a USDA program.   Those loans are supposed to help farmers buy seed and fertilizer, and float the long period between sowing and harvest.  They sued the Clinton administration, which settled even before they'd opened the discovery process.  This is known as the Pigford settlement, a name which will be familiar to readers of conservative news sites, which have been on this case for some time.

The New York Times reports that the judge initially restricted the settlement to people who had actually been farming during the period covered by the settlement, but some people complained.  What about people who had attempted to farm, but been stymied by their inability to procure government loans? Should they be shut out of the settlement?  So the judge expanded the criteria to include people who'd attempted to farm.  Unfortunately, the government doesn't keep track of people who have maybe thought about farming.  And apparently, it didn't have decades worth of records of the loans they'd denied.  So the government was forced to rely on the word of the applicants.  Here's what happened next:

Delton Wright, a Pine Bluff justice of the peace, recalled what happened after word of the settlement reached his impoverished region: “It just went wild. Some people took the money who didn’t even have a garden in the ground.” He added, “They didn’t make it hard at all, and that’s why people jumped on it.”

Mr. Wright, whose family owns farmland outside Pine Bluff, won his claim. So did two other applicants whose claims were virtually identical to his, with the same rounded handwriting, the same accusations of bias and similar descriptions of damages suffered.

Now 57, with his memory weakened by what he said was a recent stroke, Mr. Wright said he could not recall details of the discrimination he encountered, much less explain the apparent duplicate claims.

But Mr. Cross, the Pine Bluff lawyer, has his suspicions. “It got out of control,” said Mr. Cross, adding that he had filed about 1,500 claims, including Mr. Wright’s and the apparent duplicates. He estimated that up to 15 percent of Arkansas claims were fraudulent.

Claimants described how, at packed meetings, lawyers’ aides would fill out forms for them on the spot, sometimes supplying answers “to keep the line moving,” as one put it.

Even his own staff was complicit, Mr. Cross said; he discovered that four employees had been slipping unverified claims into stacks of papers that he signed. He did not inform the court monitor, he said, because “the damage was done.”

On two floors of the Cotton Annex building in Washington, a 300-member team from the Farm Service Agency reviewed claims before adjudicators rendered their final decisions. In recent interviews, 15 current and former Agriculture Department employees who reviewed or responded to claims said the loose conditions for payment had opened the floodgates to fraud.

“It was the craziest thing I have ever seen,” one former high-ranking department official said. “We had applications for kids who were 4 or 5 years old. We had cases where every single member of the family applied.” The official added, “You couldn’t have designed it worse if you had tried.”

Carl K. Bond, a former Agriculture Department farm loan manager in North Carolina, reviewed thousands of claims over six years.

“I probably could have got paid,” said Mr. Bond, who is black. “You knew it was wrong, but what could you do? Who is going to listen to you?”

Accusations of unfair treatment could be checked against department files if claimants had previously received loans. But four-fifths of successful claimants had never done so. For them, “there was no way to refute what they said,” said Sandy Grammer, a former program analyst from Indiana who reviewed claims for three years. “Basically, it was a rip-off of the American taxpayers.”

The true dimensions of the problem are impossible to gauge. The Agriculture Department insists that the names and addresses of claimants are protected under privacy provisions. But department data released in response to a Freedom of Information request by The Times are telling. The data cover 15,601 African-Americans who filed successful claims and were paid before 2009.

In 16 ZIP codes in Alabama, Arkansas, Mississippi and North Carolina, the number of successful claimants exceeded the total number of farms operated by people of any race in 1997, the year the lawsuit was filed. Those applicants received nearly $100 million.

According to the Times, in 2010 the Obama Administration responded to a similar lawsuit by 91 female and Hispanic farmers--a case that the government seemed likely to win--with a $1.3 billion settlement similar to Pigford.  It will be open to thousands of people who haven't yet sued, and the natural worry is that it will trigger a similar gold rush.  

(Hmmmm . . . I'm a woman.  And I'm related to farmers.  Wonder how hard it would be to file?  We could use some cash to redo the back yard.  I could even put in a vegetable patch, which is like farming, right?)

All government programs face a tradeoff between Type I and Type II error--between false positives and false negatives.  Think of disability benefits.  We give disability benefits for things that are basically impossible for an outside observer to verify, like back pain.  We're pretty sure that at least some of the people who are on disability for their back pain actually could work, but prefer not to.  But we can't tell how many.  

We could tighten up eligibility criteria so that you can only collect disability for an externally verifiable disease, like asbestosis or lost limbs.  But then we'd exclude people who really are in too much pain to work.  

So we've settled on a sort of hybrid system: claims for things like back pain are basically automatically denied, while my understanding is that a claim for something like primary pulmonary hypertension pretty much sails through on the first pass.  Eventually, if you stick with your back pain claim long enough, you'll probably get approved (along with an award for back payments).  The thinking seems to be that if you are willing to stay out of work for four years in order to qualify for disability, you're probably really disabled.  

But this is obviously very imperfect: most people seem to agree that the program is funding at least some malingerers.  Meanwhile, it's causing suffering among people who really are disabled and have to go without income for years while they wait for disability to come through.  The problem is that these things are a direct tradeoff: if you make applications easier, you get more malingering, while if you go after malingering harder, you cause more suffering among the genuinely disabled.  

The judge in the Pigford case seems to have decided that false negatives were the only thing worth worrying about.  As a result, the Times reporting suggests, we got a whole lot of false positives.  And now the administration seems to be making the same mistake with the next round of lawsuits.  

And while I am sympathetic to the plight of folks who wanted to farm, but couldn't get a start, I do think that this is a mistake.  Cutting checks to random people is not a good way to remedy past discrimination in a loan program.  In fact, almost any other remedy seems like it would be better.  Especially since the Times story suggests that there's no evidence that the discrimination underlying the new settlement even happened.