Based on a question I received from a reporter, we looked into how the shopping patterns in Manhattan shifted for Hurricane Sandy the storm period (October 28 to 31). We compared shopper traffic to comparable stores in upper, mid, and lower Manhattan during the storm period and during a control period the same days of the previous week.
This data set comprises 65,507 shopping trips across 25 stores. We found:
- Lower Manhattan traffic during the storm period was down 90% from the control period.
- Midtown traffic was down 19% during the storm period.
- Upper Manhattan traffic was down 50% during the storm period.
While the data exclusively tell us what happened and not why it happened, the idea that traffic had moved out of the most affected areas to equivalent stores in less affected areas is consistent with the measured results. We see that retail activity almost disappeared in lower Manhattan. The 50% drop off in upper Manhattan might reasonably represent the general effect of the storm on all residents. But if a percentage of lower Manhattan residents traveled uptown for critical needs during the storm, we would see a smaller dip in the region closer to the area where power outages and flooding occurred.