Do catalytic converter replacement problems track on-road nitrogen oxide emissions, tons per year per 1,000 residents across states?

A weak hint, not a firm finding: across U.S. states, NHTSA owner complaints about catalytic converter replacement run higher where on-road nitrogen oxide emissions, tons per year per 1,000 residents is higher, after adjusting for the makes sold in each state. This is an association across states, not proof that on-road nitrogen oxide emissions, tons per year per 1,000 residents causes this repair.

How to read it: Each dot is a state. Further right means more on-road nitrogen oxide emissions, tons per year per 1,000 residents (epa nei 2020); higher up means more complaints for this repair than expected (1.0 is average). If the dots climb to the right, reports of this repair rise with on-road nitrogen oxide emissions, tons per year per 1,000 residents (epa nei 2020).

What it shows: States with more on-road nitrogen oxide emissions, tons per year per 1,000 residents (epa nei 2020) tend to sit higher on the chart across 33 states: a pattern in the data, not proof of cause.

• The 10 states with the most on-road nitrogen oxide emissions, tons per year per 1,000 residents (Oklahoma, Oregon, Kansas, Missouri, Alabama, Arkansas, Tennessee, Kentucky, Wisconsin, South Carolina): reports of this repair run 23% above what the makes sold there would predict, and 7 of 10 are above expected. The 10 with the least (New York, Massachusetts, New Jersey, California, Maryland, Michigan, Illinois, Delaware, Florida, Virginia): 1% below, and 5 of 10 are above expected.
• Exceptions: Oklahoma, Arkansas and Tennessee are among the highest in on-road nitrogen oxide emissions, tons per year per 1,000 residents but report below expected; Illinois (28% above expected) and Delaware (88% above expected) are among the lowest but above expected.
• Split the states into thirds: the third with the least on-road nitrogen oxide emissions, tons per year per 1,000 residents runs 1% above what the makes sold there would predict, the middle third 10% above, and the third with the most 20% above. It steps up steadily.
• For on-road nitrogen oxide emissions, tons per year per 1,000 residents, after allowing for the 1562 comparisons we ran, luck alone would produce a pattern this strong about 1 time in 4. The on-road nitrogen oxide emissions, tons per year per 1,000 residents pattern leans on New York: without it, it weakens.
• Weak spot: compare only states with similar share of residents living in non-metro counties and the pattern keeps about 0% of its strength, so part of it may come from share of residents living in non-metro counties instead of on-road nitrogen oxide emissions, tons per year per 1,000 residents.

Where the on-road fleet puts out more nitrogen oxide per resident, emission-control parts are working harder and the places are more likely to enforce emission rules.

Source: EPA 2020 National Emissions Inventory, on-road, joined to NHTSA owner complaints · file downloaded 2026-10-06