Public goods, as my economist friends like to point out, are under provided.

Today’s example from the federal wrecking ball is National Weather Service radiosonde launches. Here’s Daniel Cusick at E&E (the key bits aren’t behind their paywall):

NOAA is “temporarily reducing” weather balloon flights from six National Weather Service offices in the Great Lakes and Mountain West, citing insufficient staffing in weather forecast offices.

In a Thursday release, NWS said it would launch only one data-collecting weather balloon daily from Aberdeen, South Dakota; Gaylord, Michigan; Grand Junction, Colorado; Green Bay, Wisconsin; North Platte, Nebraska, and Riverton, Wyoming.

The agency previously canceled weather balloon flights from Rapid City, South Dakota and Omaha, Nebraska, according to Axios, which first reported on the additional balloon cancellations. The reduction in flights began Thursday and will continue indefinitely, NOAA said.

The National Weather Service Employees Organization said last week that weather balloon flights also had been “canceled or made intermittent” at offices in Alaska, Maine and New York.

Daily radiosonde launches from National Weather Service sites across the country, coordinated with similar launches at the same time around the world, provide critical data input to weather forecast models. While satellites and other data sources play an increasingly important role, the tried and true twice-a-day weather balloon launch provides the vital skeleton on which our weather forecasts depend. A study by Britain’s Met office clearly showed that, when you take away the weather balloon data, forecast accuracy declines.

There’s something important going on here that we need to think carefully about. The decisions being made to slash US government spending right now seem focused entirely on the cost side, without considering the benefit side. It is possible that there is a cost-benefit argument that the costs of the staff making the balloon launches does not justify the benefit of improved forecasts (or the costs of degraded forecasts). But clearly no such analysis has been done here.

I’m trying to guard against my visceral anger at the wrecking ball crews, which could easily lead me to reflexively defend current all federal spending against the actions of the Trump-Musk autocratic shithousery. That would be intellectually sloppy on my part, so I’m trying hard not to fall into that trap.

But I spent years writing about the way we forecast the weather, and the benefits it provides. The benefits of good weather forecasts vastly outweigh the costs of collecting and analyzing the data. The National Weather Service and its counterparts around the world are a stunning bargain.

Thanks again to Inkstain’s supporters, who make this quasi-journalism possible.