Published in


#SwedenInDenial — A coronavirus story

I think at this point most people know that Sweden chose a different path in response to the global coronavirus pandemic. At first, there was widespread condemnation and ridicule, especially in the Twittersphere but also in the general international media. How could a supposedly knowledge-driven country choose to go against the general consensus? In Sweden, however, the reaction was more nuanced.

One of the more popular hashtags that was trending at the time was #SwedenInDenial. You can guess from the name that the people tweeting generally were not happy with how Sweden had responded. The conversation was all over the place though and some people even threw in some racist epithets to spice things up some more (and some things to do with the country’s immigration policy). To provide just a little context, in the earlier days of the pandemic, people of immigrant backgrounds were disproportionately impacted in Sweden. The Järva area of Stockholm, including neighbourhoods like Rinkeby, had the most intense outbreaks in the country.

Reading these tweets, I got the impression that it was more outsiders than Swedes commenting using this hashtag. So, I decided to explore the hashtag and see what was really going on, what people were saying and of course, whether my gut feeling was right.

Using python, I pulled 958 tweets, starting from May 17 and going back to April 29. Then I ran some analysis looking at the time stamps, locations, engagement, frequencies and sentiment analysis to see whether the emotions were positive, negative or neutral. I also limited the language to English for easier sentiment analysis, although some other languages slipped through. I used Power BI to create the charts.

Popular Times of Day

We see that most of the tweets came either early in the morning (0700–0900, 123 tweets) or the afternoon (1400–1600, 139 tweets) which incidentally is when the Swedish Public Health agency held its daily press conference up until mid-June.

We see that most of the tweets came either early in the morning (0700–0900, 123 tweets) or the afternoon (1400–1600, 139 tweets) which incidentally is when the Swedish Public Health agency held its daily press conference up until mid-June.

Where were the tweets coming from?

There were many who had not provided a specific location in their profiles but it was nonetheless possible to discern from their descriptions where they were located (e.g. worked at a particular university or had a flag). In these cases, I manually added the location to the list.

I managed to identify the locations of 617 (64%) tweets. 341 (36%) of the tweeters did not have a location in their profile. These 617 geographically identifiable tweets were made from 38 countries.

From the chart above, we see that the highest number of tweets came from Sweden. More interesting though is that only about half of the tweets came from local tweeters meaning half of the people using the #SwedenInDenial tag were not impacted by the Swedish strategy. In other words, the people most disappointed with the Swedish strategy were living outside Sweden.

Echo chambers?

Of all the tweets (958), 344 (36%) were original tweets, i.e. they were not prompted by anyone. Of the remaining 614 tweets, 522 (54%) were retweets and another 92, that is 19%, were replies made to other users.

Most Engaging Tweets

Who was the loudest?

@KristaVogelber1 was the most prolific tweeter, authoring 27 tweets in the 3-week duration. She sent out a total of 18,600 tweets in the 3.5 years since she joined Twitter in December 2016. Also interesting to note that she is based in Estonia.

Next busiest tweeter in this analysis was @CMewas who joined Twitter just last month (May 2020) and has already managed to tweet 632 times, of which 20 were related to #SwedenInDenial in the 3-week period.

@samted25 sent out 12 tweets included in this analysis and a total of 1,406 tweets since he joined in March 2009.

How polarised was the discussion?

In sentiment analysis, we assess the emotions expressed in texts using polarity and subjectivity. Polarity goes from -1 (negative) to 1 (positive), 0 being neutral. The subjectivity range goes from 0 (objective) to 1 (subjective). I used TextBlob in Python to get an understanding of the sentiments expressed in these tweets. The algorithm cannot pick up all the nuances of human language, especially sarcasm (I tested it).

The average polarity rating of these 958 tweets was 0.06, meaning that the overall sentiment was nominally positive. The average negative polarity was -0.27 while the positive one was also 0.27. The largest proportion (44%) were classed as expressing a neutral sentiment. Do remember that 54% of these tweets are just copies of each other.

Most Common Words

You can get a sense of what people were saying from this word cloud consisting of the 60 most common words used in these tweets.

And finally, the source.

It turns out that the most popular source was the web application, followed by Apple’s devices (iPhones and iPads) and then Android. I would have assumed mobile devices would make up a bigger chunk than 55%. Nothing is to say that isn’t so, people could just be using the web application on their phone.

In conclusion, I was somewhat right in my assumption that most of these tweets were coming from outside the country. I am also fairly confident in drawing the conclusion that the overall sentiment expressed with #SwedenInDenial was quite negative, even if the polarity score would say otherwise.


#SwedenInDenial is a hashtag used by people to bash on the Swedish coronavirus strategy, hate on Anders Tegnell and most of these people were outside Sweden.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
A. Jama

Nomad who calls Stockholm home. I like writing about politics, philosophy, and entrepreneurship. I love discussing “far-fetched” ideas.