Working Paper: NBER ID: w23966
Authors: Nicholas J. Sanders; Ryan Sandler
Abstract: Personal automobile emissions are a major source of urban air pollution. Many U.S. states control emissions through mandated vehicle inspections and repairs. But there is little empirical evidence directly linking mandated inspections, maintenance, and local air pollution levels. To test for a link, we estimate the contemporaneous effect of inspections on local air quality. We use day-to-day, within-county variation in the number of vehicles repaired and recertified after failing an initial emissions inspection, with individual-level data from 1998–2012 from California’s inspection program. Additional re-inspections of pre-1985 model year vehicles reduce local carbon monoxide, nitrogen oxide, and particulate matter levels, while re-inspections of newer vehicles with more modern engine technology have no economically significant effect on air pollution. This suggests emissions inspections have become less effective at reducing local air pollution as more high-polluting vehicles from the 1970s and 1980s leave the road, and provides an example of how the social efficiency of programs can change under improving technologies. We also estimate the importance of station quality, using a metric devised for California’s new STAR certification program. We show re-inspections of older vehicles conducted by low quality inspection stations do not change air pollution, while inspections at high quality stations have a moderate effect on pollution concentrations, which suggests the potential for ineffective monitoring at low quality inspection stations. We find little effect on ambient ozone levels, regardless of station quality or vehicle age.
Keywords: vehicle emissions; smog checks; air pollution; regulatory programs; California
JEL Codes: Q52; Q53; Q58
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Inspection-driven repair of a faulty vehicle (L62) | Reduction in local air pollution (Q52) |
| Repairing and reinspecting 1,000 vehicles manufactured in 1985 or earlier (L62) | Decrease in ambient CO levels (Q54) |
| Repairing and reinspecting 1,000 vehicles manufactured in 1985 or earlier (L62) | Decrease in ambient NOx levels (Q52) |
| Reinspections at high-quality stations (L15) | Moderate reduction in pollution concentrations (Q53) |
| Reinspections at low-quality stations (L15) | No significant impact on air quality (Q53) |
| Smog check program benefits declining over time (Q52) | Diminished effectiveness of inspections (L15) |
| Inspection of older vehicles (L62) | Improved air quality (Q53) |
| Inspection of newer vehicles (L62) | Little to no effect on air pollution (Q52) |
| Smog check program (R48) | Little to no effect on ambient ozone levels (F69) |