Deep Dive

How far up the creek?

On April 21, 2020, the EPA and the Department of the Army published a final rule in the Federal Register revising the definition of “Waters of the United States.” The Trump administration wanted to streamline the permitting process for miners, drillers, farmers and developers by narrowing the scope of regulation. As of the publication of the rule, only territorial waters and navigable streams, wetlands adjacent to them and certain ponds and lakes were left within the scope of the Clean Water Act. States and tribal areas are still free to enforce stricter rules and a broader definition of wetlands and waterways. But the water permitting data suggests that in the absence of federal regulation, states are not stepping up their efforts.

To reach that conclusion, we concentrated on one economic sector, home building, and tracked the issuing of Clean Water Act permits for construction stormwater over the last few years. We also looked at the information provided about the location of new developments relative to watersheds.

First, we needed to make sure that the water permit data was a good indicator of the overall activity of the industry. To do this, we compared the number of permits to the monthly housing starts numbers published by the US Census.

Water permits (developments) issued and used plotted against housing starts last 5 years

The relationship is not perfect (correlation of 0.63) but it’s strong enough to be useful for answering the question about changes in permitting. (The number of permits is lower than the number of housing units because permits are issued for developments not individual homes.) The relationship between permits issued and used and housing starts means that permits can be used as a model input in predicting the monthly housing numbers. Permits are updated daily. The permits also break down by builder, helping to assess activity and market share by company.

Having established the relevance of the data, we next turned to changes relative to waterways.

Water permits issued and used and ratio of permits that include waterways

Up until the summer of 2019, practically every permit issued included the watershed in which the development was located. A slow erosion then set in, with the portion of permits listing waterways compared to all waterways hovering around 95%. By the end of last year that had slipped to 92%, and by February of 2021 it was 22.5%. Data from the first half of March suggests that it will drop further.


One possible explanation is that builders are now moving into areas where there are no streams or wetlands. A quick look at the geographic location of new housing starts shows that to not be the case. There are shifts as builders react to demographic and price changes, but there is no pronounced shift from wet to dry parts of the country. In the maps below we chose the largest builder, and we organized the data by expiration date, which is typically five years after the issue date.

Permits issued to D.R. Horton by expiry date

As in housing so presumably in other industries. Water permits are being obtained under less rigorous requirements. It will take years to determine whether it matters. It takes time to detect and measure changes in waterway health, and it takes even longer to alleviate negative impacts.