MPs, Twitter and gender

A little while ago, I posted about running some Twitter analytics on the accounts of John Key, David Cunliffe and Russel Norman. After some discussion Twitter, questions were raised about the people that Russel Norman was interacting with. This lead to a blog post comparing Russel Norman and Metiria Turei using Twitter analytics. At that time I did some investigation of the tweets I had been collecting, but I didn’t do anything with it, as I didn’t feel I had a solid enough base of data to draw any conclusions from.


So in response to those concerns I collected more data over a longer period. The data I have collected is all tweets to or from the following MPs, Peter Dunne, Russel Norman, Metiria Turei, Nikki Kaye, Judith Colins, Jacinda Ardern and Gareth Hughes. This collection ran from 0500 GMT on 13 March to 0415 GMT on 24 March. (I will embed the spreadsheet below. But this is the link).

These MPs were selected as they have a solid following on Twitter, with a reasonably similar, active level of engagement on Twitter.


From this spreadsheet I have pulled all the tweets sent by the above MPs. (link) This resulted in a spreadsheet containing 541 Tweets (522 unique) covering the same time range.


This collection of Tweets has the following make up:


spreadsheet 1Judith is the most active tweeter, with almost 1/3rd more tweets than the next MP, Peter Dunne, and her tweets making up around 30% of all tweets completed. She also ties for highest number of mentions, and the 3rd= in the number of retweets made, as a percentage. The only MPs with less retweets are Jacinda, who has none and Peter Dunne with 11%. Russel has the highest number of retweets, at nearly 50% of all his tweets.



no rts

When you take out the retweets, things change a lot. Russel Norman drops in the rankings from the 3rd to 4th most active, now barely ahead of Metiria. What strikes me the most is the number of mentions that have vanished. This indicates that most of the mentions by other MPs over this period came about form retweets, not original tweets by the MPs themselves.



However, what sparked this questioning was  the lack of females on the list users being replied to by Russel Norman. One explaination that was suggested was a lower willingness of females to engage in what can be a rather adversarial sphere of public debate on Twitter. I have not passed judgement on that suggestion, but thought the best way to assess this would be to collect tweets from a number of MPs with similar public and Twitter profiles, and compare how they engage.


I have gone through all the Tweets sent, excluding retweets, and looked at all the tweets that were in reply to another tweet. I have then manually labeled all the accounts possible as either male or female. I have done this based on the Twitter profiles, names, details on web addresses linked from the accounts. Now, I assume I will have got some wrong. But due to Twitter profiles not listing sex, there is not a lot I can do about it. I then counted how many tweets they replied to that were male and how many were female, I also counted the total number of tweets they had replied to. I then also counted how many unique accounts of each they replied too. Results are:

Male Female unique male unique female Total tweets replied to
Russel Norman 10 1 8 1 11
Metiria Turei 4 5 4 4 12
Peter Dunne 38 12 24 7 58
Nikki Kaye 2 6 2 5 9
Judith Collins 52 26 43 12 85
Gareth Hughes 24 3 16 3 33
Jacinda Ardern 15 4 8 4 20

The difference between the total of male+female and total tweets is due to the user whose sex I am not able to identify.

It is obvious which MPs are replying to tweets. Of Judith’s 112 Tweets, 85 of them are replies. Both Metiria and Russel have around 1 in 4 of their tweets being replies. Gareth Hughes is sitting at around 1 in 2 tweets being replies.

However, what I was most interested in looking at was the ratio of male to females for replies. I am going to use unique people as the measure, as this negates any issues around people leaving the conversation.

All figures are male to female.

Russel: 8 to 1

Metiria: 1 to 1

Peter: 3.4 to 1

Nikki: 1 to 2.5

Judith: 3.6 to 1

Gareth: 5.3 to 1

Jacinda 2 to 1.


So what we have is Russel as the outlier with the most unique men to unique females with 8 males for every female, Nikki on the other extreme with with 2.5 females for every male. Gareth is also rather high at over 5 to 1 male to female.


From looking at this data, there is something deeper here, as all bar 2 of the MPs (Nikki and Metiria) are replying to more males than females. Is this because there are more males tweeting at them than females? This is something I will do more research into, but I would be interested in hearing my readers views and experiences.

Collected tweets less retweets:

Collected Tweets:

Raw tweets to and from lister MPs:

Tweet visualisation: