Count Love

Protests for a kinder world.

Where do you get your protest data?

We crawl local newspaper and television sites on a daily basis, and most of our protest data come from these crawls. If you're curious about how an article makes its way from the internet onto the Count Love map, we wrote an introductory piece about our software stack. We've also written an overview of our data, crawler technologies, and natural language models. Our dataset covers protest events in the United States between January 20, 2017 through January 31, 2021.

Our initial protest data for the 2017 Women's March came from the Crowd Counting Consortium, who are continuing to document protests in the United States.

What types of demonstrations do you count?

We count public displays of protest that are not part of “regular business.” We typically do not include awareness events, commemorative celebrations, historic reenactments, fundraising events, townhalls, or political campaign rallies.

Where do you get your count estimates?

We record the most conservative attendance number from the news articles that we link. We interpret “a dozen” as 10, “dozens” as 20, “hundreds” as 100, and so forth. If an article mentions a demonstration but does not include an attendance count, we note the demonstration but leave the count empty.

Why do some of your links lead to error pages?

Unfortunately, the internet is transient. While nothing disappears on the internet, most things don't stay in the same place in perpetuity. When possible, we redundantly link all articles that we've found about a protest event, and that list of links is available through our search page.

Can I use your protest data for...?

Yes! We hope that this site becomes a useful resource for journalists and concerned citizens. If the web interface is not sufficient to compile the information that you need, you can download a full export of our protest events, locations, dates, and counts as a CSV file.

Here are a few organizations that use our data to help construct their own protest datasets:

We also offer an archival version of our data as a GitHub repository: https://github.com/count-love/protest-data.

Our data is available under the Creative Commons 4 Attribution license.

Why are you doing this?

Protests and demonstrations represent one way to communicate to our elected leaders. Yet, it’s easy to lose track of exactly where and when protests took place and how many people participated. Additionally, searching through and visualizing individual records can be quite a daunting task. We hope that keeping a factual record of ongoing demonstrations and making this data more accessible helps citizens, journalists, and politicians make more compelling cases for a diverse, empathetic, and kind country.

Who are you?

We are Tommy Leung and Nathan Perkins, engineers and scientists with a keen interest in civic responsibility and public policy. We started Count Love in catharsis to 2016 and continued active data collection and development through January 31, 2021. We met during overlapping stints at MIT while working on our Masters in Technology and Policy.

We occasionally write about our protest work in other venues:

And a few people have used data from Count Love to tell stories about protests in America:
2020

2019

2018

2017