The Knight News Challenge is a grant by the Knight Foundation for “breakthrough ideas in news and information”. Between September 8th and 30th of this year, 1.023 entries were submitted by 960 teams or individuals (myself included).
I wrote a scraper that gathers the data from these submissions. Its code and the resulting dataset are available on GitHub. My first curiosity was where the applicants come from. On the submission form there’s a “Location” field, but it’s not clear if it’s related to where the team lives or where the project is going to be applied. It also is a simple textarea, so the data is messy.
To answer this question, I had to go through every entry and figure out what’s the actual location. I used the team’s location, and not where the project will be applied. For example, if a team from the USA submits a proposal for a project about Kenya, I still considered their country as the USA. If the team is distributed, a single proposal can have multiple countries.
I tried to be careful when cleaning the data, but there might have been some mistakes, so take these results with a grain of salt. I also couldn’t determine the location of 7 projects (around 0,7% of the total).
First, let’s see a choropleth map:
The 7 colors are divided by quantiles. As the data is heavily skewed, the values and the colors don’t have a linear relation. Only the USA, with 687 projects, is on the darker band. The following band has 4 countries: UK (51), Germany (30), Brazil (29) and Canada (25). The top 11 countries are:
There are 98 countries in total.
Remember that these numbers represent the number of projects that have at least one team member based in that country, so the total here can be different from the total number of entries (and it is 1181 vs 1023).
I don’t know how the countries’ distribution was in the previous News Challenges, so I don’t have a baseline to compare. Still, it’s interesting to see that, even though more than half of the entries had someone based in the USA in the team, almost every continent is represented in the top 11 countries. The only exception is Oceania, that is quite near, with 10 projects having someone based in Australia.
The data used here is available on GitHub here (unmodified, straight from the News Challenge page), and here (teams’ countries added).