The point of this post is to show how complex and grassroots public gatherings can be counted and measured. The two main researchers quoted in these Washington Post articles are Erika Chenoweth and Jeremy Pressman, both respected analysts of the details of large-scale civil movements and gatherings. As a data junkie, I have followed this effort with a great deal of interest (and have even counted some of these gatherings in my own town to check others counts) and look forward to more of the analysis of both the methodology and the actual count data. The analysis included not just the number who gathered but who and what was being protested or being supported, where these events were held, what symbols were used, how many arrests were made.
For March 2017, we tallied 585 protests, demonstrations, marches, sit-ins and rallies in the United States, with at least one in every state and the District. Our conservative guess is that 79,389 to 89,585 people showed up at these political gatherings, although it is likely that there were far more participants.
Certainly food and farming systems should note some of the systems used for collection and analysis. For example, the Crowd-Counting Consortium may be something that national entities involved in any grassroots data collection systems like food systems should discuss creating for their own use.
Here is their counting method:
We arrived at these figures by relying on publicly reported estimates of march locations and the number of participants involved in each. We started a spreadsheet and called for crowdsourced information about the location and number of participants in marches. Before long, we had received thousands of reports, allowing us to derive low and high estimates for each event. We carefully validated each estimate by consulting local news sources, law enforcement statements, event pages on social media, and, in some cases, photos of the marchers. When reports were imprecise, we aimed for conservative counts; for example, if observers reported “hundreds” of participants, we reported a value of 200 (“thousands” was 2,000, “tens of thousands” was 20,000, etc).
An example of their public data set.