Elizabeth Warren is again at the center of a political controversy. Despite her insistence that she is part Cherokee, based upon “family lore” and her observation that some in her family had “high cheekbones like all the Indians do,” she has failed to produce any concrete evidence to substantiate her claim. We have little to say about that political debacle, but it reminds us of a substantive policy issue where Warren also failed to back up her claims with sufficient proof.

In an op-ed we wrote when Warren was gunning for the top post at the Consumer Financial Protection Bureau, we exposed the poor quality of Warren’s academic work on medical bankruptcies and explained that this research failed to justify its striking conclusion that medical debts account for more than half of all personal bankruptcy filings in the U.S.

We wrote:

"Warren’s findings, based on research of dubious academic standards, exaggerate the magnitude of the problem and affect policymakers’ ability to make informed, unbiased decisions regarding pending legislation. Yet perhaps because of the Harvard label attached to the papers or dismay over the plight of medically distressed households, far too many in Washington seem willing to accept these findings without question."

In addition to the statistical-sampling and econometric issues in Warren’s research, the studies she co-authored found a high incidence of medical bankruptcy because the definition of medical bankruptcy they used was implausibly broad. In her first paper, “medical” included any sort of addiction, including uncontrolled gambling problems. The second study excluded gamblers and addicts but still managed to conclude that more than 60 percent of bankruptcy filings are due to medical reasons, even though only 29 percent of respondents believed their bankruptcies were caused by medical bills.

Personal bankruptcies have increased in recent years, but if medical debts were the cause we should have seen an increase in this measure of debt. We haven’t. The Federal Reserve Board’s Survey of Consumer Finances shows that over the period 1989–2007, “goods and services debts,” which include medical debts, barely changed, from 6.1 to 6.2 percent of all debt.

All this leaves us wondering whether a similar leap of logic is required for us to believe that Elizabeth Warren is a Cherokee Indian. Ultimately, it’s all about how you define things.

— Alex Brill is a research fellow and Aparna Mathur is a resident scholar at the American Enterprise Institute.