On Tuesday, July 23, we joined our partners at Environment America Research & Policy Center in publishing “Safe for Swimming?” For the report, we analyzed roughly 200,000 water testing records from more than 4,500 testing sites at U.S. beaches and coastal areas. Our analysis highlighted the prevalence of fecal contamination at beaches across the country.
In the hours after the report was released, we began to receive questions about our data. Our investigation into the answers revealed that we made a mistake in our analysis. We reissued the report with corrections to address that first error on July 23. However, when we discovered another problem on July 24, we decided to remove the report and associated content from our website until we could be sure of its accuracy. We recommended that Environment America and its affiliates do the same. (The mistakes we made were technical in nature; for more detail, see below.)
Since then, we have reviewed and reanalyzed the data thoroughly. The revised version of the report has been published and is available here. The most consequential errors in our original report were at the local level, especially in the Great Lakes states.
Frontier Group’s initial report may have misinformed you about contamination frequency at your local beaches. We are deeply sorry for that, and for the resulting distraction from the environmental hazard we sought to bring to light.
The revised analysis shows a change of less than 2 percent in the total number of beach sites affected by fecal contamination nationwide, and no changes in the vast majority of states. The report’s conclusion remains the same: contamination in our waters is a widespread problem. Smart public policies and infrastructure investments can help to address it.
But for most of us, the overall health of America’s beaches is a somewhat abstract concern. The concrete question is: How is the water at the beaches that we, our neighbors and our families visit? And that’s where, in some places, we got it wrong.
We see our work as an opportunity to contribute to the advancement of democracy by providing the news that we all need to hear, backed by solid data. Our mistakes last week created confusion, rather than the clarity we strive for. We are grateful to those who reached out to us with the concerns that led us to correct the errors. And we appreciate the opportunity to set the record straight.
We thank you, as always, for your interest in and attention to our work, and we welcome your questions.
 Specifically: We applied incorrect thresholds for bacterial contamination to two types of testing results in the original data obtained from the National Water Quality Monitoring Council: 1) tests using the qPCR method (used primarily in IL and WI); and 2) single sample test results for E. coli. These errors led to over-reporting of the number of days with potentially unsafe water at beaches, particularly in the Great Lakes states. The revised report also includes two changes in how the data are presented: reordered lists of beaches in several states that reflect a revised approach to breaking “ties” between sites with the same number of potentially unsafe days, and improved location information for beaches in some states, especially Wisconsin. These adjustments result in lists of beaches having changed for some states in this revised version of the report – even those for which the underlying data remains the same.