Conclusion

Johnson’s popularity reached a climax in early April after calling the first COVID lockdown. This relationship between lockdowns and public approval is consistent through the timeline. Despite COVID hospitalisations rising through the autumn of 2020, Johnson’s approval rating rises as he enacts the second and third lockdowns. Even when London and much of the South East of England were placed under ‘Tier 4’ restrictions days before Christmas, Johnson’s approval still rose. It appears Johnson enjoys the benefit of taking strong lockdown measures without being seen as responsible for why they are necessary.

Similarly, persistent declines in his popularity seem to be characterised by the loosening of lockdown restrictions. The precipitous drop in approval rating following the first lockdown includes the Barnard Castle scandal which likely compounded the decline. Johnson’s approval rating continued to fall through the summer as lockdown measures were eased and the Free School Meals campaign reached a zenith. This characterises a problem with analysing any politician’s approval rating; though a story may be dominating headlines, it is not necessarily the instructive factor in informing public opinion. Johnson’s approval declines further following ‘Partygate’, however it effectively maintains the trajectory it was on prior to the story.

Data notes and methodology

The project is structured on a timeline of events plotted against a dataset of headlines and approval ratings. Although Boris Johnson has been in power and press before his premiership, due to the difficulty of finding comparable approval ratings and digitally archived headlines, this project covers his career as Prime Minister that began on 19 July 2019 to a self-determined stop date of 24 April 2022.

First and foremost, significant events were found and detailed to construct the timeline. What constituted a ‘significant’ event was judged subjectively, but drawn from Wikipedia entries for Boris Johnson, end of the year recaps by various news organisations, and the Institute for Government’s “Timeline of UK government coronavirus lockdowns and restrictions”.

News headlines were scraped using a third-party news gathering API called Newscatcher API. In order to limit our search the API only drew up headlines which included the words ‘Boris’, ‘Johnson’ or ‘Boris Johnson’ for our given time period. This produced 23,444 results from 1,296 publishers around the world. Data clean-up revealed 16k usable headlines of which 300 had more than five headlines about Boris Johnson, and 175 of which were from British based publications.

The team decided to manually select a sample of headlines for each event and so a subset of British-based media were selected in an attempt to represent a variety of media types (tabloid and broadsheet), and slants (left, right, centrist). The six chosen publications are: Daily Mail, The Guardian, The Mirror, The Telegraph, The Times, Financial Times. Headlines were selected by filtering the resulting dataset for the date(s) of the event, plus 3 days and choosing at least one from each publisher where possible.

The approval ratings are sourced from Britain Elects, UK’s largest poll aggregator, in order to get granular data for each date.