Research
Each piece states what it found, what data it used, and what it could not determine.
Data analysis
Full studies with public government data, charts, and code you can check.
Federal auditors checked four states for autism therapy billing. In all four, nearly every claim they sampled had something wrong. The national data shows the same pattern in every state they have not checked yet.
Medicaid ABA spending grew from $448 million to $2.03 billion nationally between 2019 and 2023. Four state audits found the same documentation failures independently. National data shows why: one billing code has held 72 to 77 percent of spending every year, and the growth came from new entrants, not consolidation.
Read the full paper with charts → Paper 04California published the same year's detention data three times and the numbers differ sixfold
Three documents on the state health department domain all present themselves as the FY2019-20 involuntary detention report. The statewide count is 17,736 in one, 29,117 in the next, and 117,620 in the third. None of them says it revises another.
Read the full paper with charts → Paper 03One in five hospices has no quality score, and half of them are not told why
Medicare publishes a quality score for 5,411 of 6,852 hospices. For the other 1,441 the page is blank. Recent certification explains 587 of them. The rest were open for the entire measurement window, and 735 carry a footnote that does not fit a claims-based measure.
Read the full paper with charts → Paper 02When the auditors said the nursing home numbers were wrong, they were only half right
Federal investigators found 938,000 unsupported RN hours in a single month. That is the accuracy problem. There is also a distribution problem, and they are not the same thing.
Read the full paper with charts → Paper 01Your nursing home's staffing rating is calculated on a Wednesday
The number families use to compare nursing homes is a quarterly average that treats all seven days equally. The daily data shows what that average hides: four days of reduced coverage, every week, at nearly every facility in the country.
Read the full paper with charts →How I think about measurement
Shorter notes on problems I see repeatedly: numbers that hide their own gaps, handoffs nobody owns, automation applied in the wrong order.
What working together looks like
Plain descriptions of how an engagement runs and when I am not the right fit.