Forecast COVID-19 cases in the US under different levels of social distancing
An important factor to understanding the evolution of COVID-19 cases is to understand social distancing. Thus, we derived the SafeGraph Social Distancing Index, and plot it at the national and state level. Click here to see how we derived the Social Distancing Index. Higher numbers correspond to greater levels of social distancing; lower numbers correspond to a greater return to normalcy
Click the box below and choose a state to see it’s Social Distancing Index graph.
Derivation of the Social Distancing Index
The social distancing index is calculated in the following way:
1. We take the first principle component of 5 SafeGraph variables to measure the level of social distancing: the percentage of residence staying home, the percentage of residents working at a workplace full time, the percentage of residents working part time, the median duration of time that residents stay home, and the median distance traveled. We orient the principle component do that a higher number means more social distancing.
2. We compute the US average of these daily state-level social-distancing measures (weighted by population) from January up to present day. The minimum of the US social distancing measure appears in late February and the maximum is during the second week of April. We then obtain the US daily social-distancing index by applying the following normalization formula such that on a given date,
US Social Distancing Index = (US Social Distancing Measure – Minimum of US Social Distancing Measure) / Range of US Social Distancing Measure
3. We then obtain each state’s daily social-distancing index by applying the following normalization formula such that on a given date,
State Social Distancing Index = (State Social Distancing Measure – Minimum of US Social Distancing Measure) / Range of US Social Distancing Measure