Projects Funded for Jamie Hansen-Lewis


The Impacts of Maritime Emissions Standards

  • Michelle Marcus
  • Jamie Hansen-Lewis


Specific Objectives of the Project:
1. Evaluate the impact of maritime emissions standards on air quality in California and in other coastal states
2. Relate and test the equivalence of ex-ante and ex-post methods of air quality policy evaluation
3. Examine the role of maritime emissions on a range of potential outcomes, such as agricultural yields, worker productivity, and infant health, in California
Project Report/Summary of Results:
1. We found that maritime fuel standards, specifically the 2012 Sulfur Emissions Control Area (ECA), led to reductions in fine particulate matter concentrations (PM2.5). On average, the policy resulted in an 0.35 microgram per cubic meter (3.8%) improvement in PM2.5 (p=0.002). We failed to find evidence that the policy predicted trends in air quality in the years prior to implementation (pre-trends).
The data indicated that air quality improvements differed significantly in locations where the jurisdiction of the policy was less than the full 200 nautical miles. Whereas locations with the full jurisdiction had relatively consistent improvements, areas with partial jurisdiction, which included southern California and Florida, had marked improvements in the initial year of the policy and the year the policy became more stringent, but little improvement in intervening years.

2. Overall, we rejected the equivalence of the ex-ante predictions from the EPA and ex-post observations. For a one microgram per cubic meter planned improvement in PM2.5, we estimated PM2.5 fell 0.45 micrograms per cubic meter on average.
Notably, we detected differences in the performance of the ex-ante predictions between areas with full jurisdiction and areas with partial jurisdiction. For the full jurisdiction locations, we failed to reject the equivalence of the ex-ante predictions and ex-post observations. For the partial jurisdiction locations after the initial years, we rejected the equivalence of the ex-ante predictions from the EPA and ex-post observations.

One explanation for this result is that the ex-ante predictions do not account for behavioral changes in reaction to the policy. Ships are best situated to subvert the policy where the jurisdiction is small because they do not need a large route detour to use highly polluting fuels. For example, within 80 miles of Imperial county, a location of significance for Californian agriculture, ships could continue polluting. These findings indicate that transport model predictions commonly employed to justify pollution regulation work well in a world without behavioral modification but also risk significantly overstating the realized impact of regulation.

3. Thus far, we have examined the impact of the fuel standards on the rate of low infant birthweight and community demographics. We found significant improvements in the rate of low infant birthweight. The policy was associated with an 2.2% fall in the rate low birthweight (p<0.001). We also failed to find evidence that the policy treatment predicted trends in infant heath in the years prior to implementation (pre-trends). In a robustness check, we found that the policy was associated with changes in the demographic composition of mothers. This result indicates that a change in the composition of mothers, rather than air quality, contributed to the improvement in infant health associated with the policy. We are still evaluating the salience of the air quality changes from this policy to determine if the demographic change was a result of the policy or a result of an unobserved factor that coincided with the policy.