Who is the most popular government cat?

Every now and again, a story appears in the media about the UK government cats. Most recently, it was reported that the cats may not be fulfilling their primary purpose: catching mice.

But one thing they do seem to be good at catching is fans. According to National Archives records, different government cats have amassed public followings.

The current government cats are: Larry (10 Downing Street), Palmerston (Foreign and Commonwealth Office), Gladstone (Treasury) and Evie and Ossie (Cabinet Office). Peter K Wells, an open data advocate, has compiled data about the government cats going back to the 1930s. The current generation are the first to be represented by Twitter accounts, albeit mostly in an unofficial capacity. This means that we do not have to wait 30 years or so for government documents to be released to find out how popular each government cat is. We can see how popular they are in real time by looking at the respective Twitter profiles.

However, the information in the profiles is often approximated to make it easy to read and it is not possible to view the data simultaneously.

Using Python and Twitter’s own API, I obtained the exact data for the number of followers, friends (the number of accounts the user is following), tweets, likes made and the creation date, as of 1 October 2019:

Government Cat	Followers	Friends	Tweets	Likes	Creation Date
Number10cat	330346	131	17053	28456	2011-02-15 11:00:56
DiploMog	92092	8	229	45	2016-02-17 14:56:00
TreasuryMog	54913	19	19065	72919	2016-10-14 13:47:06
HMCabinetCat	41851	6	17	23256	2016-08-04 07:06:54
PalmerstonFOCat	40453	765	965	1508	2016-04-12 12:30:53
HMTreasuryCat	22024	4	127	140	2016-07-29 09:26:54
PalmerstonCat	10983	32	1007	2996	2016-04-12 22:00:08

@Number10cat is the most popular Twitter account in terms of number of followers and it is also the one that has been around the longest. The rest of the Twitter accounts were created at various points during 2016. It is possible to argue that, over a relatively long period of time, the longer an account has been around, the more followers they are likely to have.

Where accounts were created within a close period of time, there is no clear relationship between the number of tweets and the number of followers. @PalmerstonFOCat and PalmersonCat were both created on the same day, within four hours of each other, and have approximately the same number of tweets, but @PalmerstonFOCat has four times as many followers. But DiploMog, which is the official account for Palmerston, has amassed twice as many followers as PalmerstonFOCat with less than a third of the followers  What’s more, @HMCabinetCat has slightly more followers than than @PalmerstonFOCat but has only tweeted 17 times.

The difficulty of identifying a relationship between the various attributes can be seen in the averages and standard deviations for each attributes.

Attribute	Mean	Median	Standard Deviation
Followers	84666	41851	111385
Friends	138	19	280
Tweets	5495	965	8612
Likes	18478	3022	26737

 

A visualisation of the relationship between the above attributes is available in the full MSc project report and Python code is available on Github. The report includes the following:

• A visualisation of the relationship of the above attributes;
• A network analysis