Data Is Plural

... is a weekly newsletter of useful/curious datasets.

2020.07.01 edition

The coronavirus economy, early voting registrations, COVID clusters, subway accessibility, and twelve men using computers.

The coronavirus economy. Opportunity Insights’ Economic Tracker “combines anonymized data from leading private companies – from credit card processors to payroll firms – to provide a real-time picture of indicators such as employment rates, consumer spending, and job postings across counties, industries, and income groups.” Small business revenue, for instance, appears to be down about 55% in Boston, compared to the beginning of the year. The project’s GitHub page provides aggregated data corresponding to each of the charts and graphs. Previously: Opportunity Insights’ extensive datasets on economic mobility (DIP 2019.06.12). [h/t Matteo Ferroni + Dan Stein]

Voter registration. At least three states publish monthly, machine-readable statistics on new voter registrations: Florida, Virginia, and North Carolina. Maryland and the District of Columbia publish similar data as PDF tables. A recent report by Center for Election Innovation & Research analyzed pandemic-era registrations, using that data plus numbers obtained from Arizona, California, Colorado, Delaware, Georgia, Illinois, and Texas. The analysis found a “steep decline in new registrations,” largely attributable to social distancing measures and DMV closures. To accompany an article on the topic, FiveThirtyEight has compiled CEIR’s counts for those 12 states, for the first five months of 2016 and 2020, into a simple CSV.

COVID clusters. A team of researchers in the UK has been gathering data on COVID-19 “transmission events” that have resulted in clusters of cases, based on various official, scholarly, and news reports. For each of the 250+ events listed so far, the dataset specifies the setting (e.g., “work,” “household,” “religious,” “elderly care”), whether it was indoors or outdoors, geographical location, number of cases involved, and more. The New York Times has also been tracking the location and size of COVID-19 clusters in the US — so far, more than 1,700 clusters with at least 50 cases. The NYT doesn’t provide a download button, but its webpage loads the data from a JSON file. [h/t Kai Kupferschmidt]

Subway accessibility. To better understand accessibility issues in the NYC subway, the Two Sigma Data Clinic has constructed a series of datasets and diagrams describing each station’s elevator connections — from the street to the station, and between various points and platforms within it. Related: The subway system’s official list of elevators and escalators. Previously: Turnstile data from NYC and Chicago (DIP 2017.02.08). [h/t Erin Stein]

Humans using computers. BEHACOM is a dataset that details minute-by-minute usage statistics for 12 Spanish and Italian men “interacting for fifty-five consecutive days with their personal computers in their own way and without restrictions.”