A private rule of thumb: The accuracy and general informational content of news reports that "[something] costs the US economy N billion dollars every [time period]" is inversely proportional to square of the N dollar figure. If something supposedly cost the US economy 5 billion dollars (over whatever time period), the concept is probably quite meaningful and also probably quite accurate. If something supposedly cost the US economy 50 billion, be prepared to doubt the methodology and to ask whether the concept might be slightly fishy. If something supposedly cost 500 billion, it's almost certainly total nonsense -- or worse than nonsense, it's a fallacy, because it has been dolled up so it can pass for a "fact".
The surprising fact is that you can account for the degree of speciousness/spuriousness of the media assertion just with the dollar figure -- you don't also need to correlate the period of time mentioned, nor do you need mythical quantification of the weirdness of the concept.
So the next time you hear someone report something like "stuttering costs the US economy 232 billon dollars a year", or child abuse, or grafitti, or pagers, don't say I didn't warn you