Data Is Plural

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

2020.05.27 edition

Census response rates, more excess mortality data, ranked-choice votes, Shakespeare and Company’s lending library, and seashells.

Census response rates. As it has in the past, the US Census Bureau is encouraging residents to respond to the 2020 Census’s mailed questionnaire, which reduces the need for in-person census-taking. The agency calculates these “self-response” rates all the way down to individual Census tracts, and provides that data as a CSV and via its API. Those datasets, however, only represent the latest numbers, so researchers at the CUNY’s Center for Urban Research have been creating daily snapshots, which they’re also mapping and analyzing for the public. Related: The Center’s FAQ for the data. [h/t Steven Romalewski]

More excess mortality data. Last week’s newsletter featured the data and code that The Economist is using to estimate excess deaths due to COVID-19. Also last week: The New York Times began publishing the data behind its similar-but-different estimates. And earlier this month, the team at the Human Mortality Database launched its Short-term Mortality Fluctuations data series, which provides “user-friendly access to detailed data on mortality by week, sex, and aggregated age group” for more than a dozen countries. [h/t Esteban Ortiz-Ospina]

Ranked-choice votes. The website, built by quantitative analyst Paul Butler, standardizes and visualizes the detailed results of a few dozen elections — in Maine, San Francisco, Santa Fe, and Burlington — that have used ranked-choice voting, where voters can list their preferred candidates in sequential order.

Interwar literary lending. From 1919 to 1941, Sylvia Beach ran Shakespeare and Company, the legendary Paris bookstore. It featured a lending library, whose members included writers such as Gertrude Stein, James Joyce, Simone de Beauvoir, and Ernest Hemingway. Princeton University’s Shakespeare and Company Project has digitized hundreds of the library’s lending cards and logbooks, and has made the data available to explore and download. [h/t Tom Merritt-Smith]

Seashells. A team of researchers has built a dataset that characterizes 29,622 samples of shells from 7,894 water-dwelling species. You can search the shell images online and also download the full dataset.