As the East Coast digs out from the Blizzard of 2016, a recent report from Moody's Analytics estimates the cost of the Blizzard to be between $2.5 and $3 billion dollars. When there are major weather events of this nature, we often see these types of estimates calculated and widely reported in the media.
But what are the types of questions you should be asking (as an everydata consumer) to assess these claims?
First, these types of estimates are often based on what are called "but-for" scenarios—but for the blizzard, what would have been expected to happen with respect to productivity, consumer spending, etc? So, a critical question to ask here is what are the assumptions underlying this particular calculation.
In this follow up article from Forbes, the authors of the calculation provide some helpful context. One of the key components of the calculation is lost retail spending—particularly restaurants and hospitality which is estimated at $860 million. It is not clear, however, if the estimate considers any potential rebound in lost spending from pent-up demand over the next few weeks.
Another key assumption appears to be the closures occurring on Friday and Monday, which directly impede the work week. The approach used here is to assume certain part of these wages are just lost, particularly for Federal government workers. One question I had is how is telework productivity measured in this context? Further, it is not clear over how wide an area the estimated losses are calculated from the press articles.
The study also appears to recognize some form of lost transportation expenses for local municipalities, but again, it is not clear if the demand for public transportation might go up in the aftermath of the blizzard, therefore mitigating some of these losses. It is also not clear how the study factors in the expenditures on the clean up, which the Governor of Virginia estimates at $2-$3 million per hour.
Moody's Analytics is a reputable organization known for thoughtful estimates, so it may well be that their estimate is correct. But, as a consumer, it is helpful to look for the hidden assumptions which are made so you can better understand how these types of numbers are derived.