This past winter was unusually harsh, and it’s showing up in the data. We are tracking GDP growth of just 0.5% for the first quarter, a sharp step down from 3.4% averaged over the second half of last year. To be sure, part of the weakness reflects an inevitable slowing following unsustainably rapid inventory-building over the second half of last year. But the unusually harsh winter played a role, too. In past work, we identified the effects of the harsh winter by keying in on the unusually cold winter temperatures. But we were unable to study the effects of snowfall because an appropriate time-series on national snowfall, to our knowledge, did not exist … until now.
We have compiled a snowfall database and developed several measures of unusual snowfall to help us understand its role in the weak Q1 data. The snowfall database is built-up from daily measures of snowfall by weather station from the National Climatic Data Center, aggregated into monthly measures by county, and weighted up to the national level using counts of population by county from the 2010 Census. Shown in the chart is our population-weighted measure of unusually high snowfall. It is a population-weighted average of snowfall by county in excess of the norm, where the norm varies by county and calendar month. The idea here is that normal amounts of snowfall likely affect economic activity, but those effects are adjusted out of the data in the seasonal adjustment process. It’s only unusually high amounts of snowfall whose effects remain in the seasonally adjusted data.
Using these data and our model of weather effects in monthly GDP, we estimate that elevated snowfall this winter, mainly in January and February, reduced first-quarter GDP growth by 1.4 percentage points. Furthermore, assuming normal snowfall in April, May, and June, the boost to second-quarter growth reflecting a return to a normal level of GDP is 1.6 percentage points. Given our current forecast of 0.5% GDP growth in the first quarter and 3.8% growth in the second quarter, a “normal weather counterfactual” would put Q1 GDP growth at 1.9% and Q2 growth at 2.2%, not a bad profile given the effects of the sharp decline in inventory investment noted above.