A quick trip down memory lane for some historical perspective on the impact of the internet on political campaigning with this piece that I wrote for an academic email newsletter (the Political Marketing Group Newsletter) after the 2003 Brent East by-election. The lessons still look extremely relevant – making the point that principles of good campaigning and communication last as individual technologies, programs and companies come and go.
Winning Brent East: did the internet matter?
When Sarah Teather won the Brent East Parliamentary seat for the Liberal Democrats in the by-election on September 18th 2003, it was one of the party’s most dramatic election victories.
A previously rock-solid Labour seat – former Parliamentary seat of London Mayor Ken Livingstone when he was an MP – became a Liberal Democrat gain with a swing of 29%.
Although the Liberal Democrats started using email extensively back in the Internet’s early days of the 1990s, it was the Romsey by-election in 2000 which first saw a concerted approach to using the Internet to win public elections beyond simply using email as a convenient internal communications tool.
The Romsey by-election both saw a website for the public and the use of email lists for mobilising volunteers to go and help in the constituency. The Brent East by-election was the first major opportunity to revisit the potential of the Internet in a large by-election campaign after three years in which the Internet had continued to grow in importance, hype and penetration.
So how did the Internet shape up? In summary – it was an effective tool for mobilising volunteers and improving internal organisation but is still of only limited effect in reaching out to floating voters. Given the apparently similar lessons from Howard Dean’s bid for the Democratic Presidential nomination, this divergence between the internal and external impacts of the Internet may well soon become the received wisdom.
In Brent East, what it meant was that regular emails to party members and helpers around the country were used to encourage people to come and help. Email made it easy to provide information cheaply and quickly and allowed little flourishes such as providing links to online maps to make it easy for people to work out where to go. The ease of responding to emails also meant it was easier to engage in conversation with people who might need a little extra persuasion or information to come. If they had been posted information, they might not have got round to replying with questions (e.g. “are we serious about winning – do you really need my help?”) but with email the reply button is always just to hand.
Many of these emails replaced what would have been done by post or phone in the past. This meant the impact of the internet was evolutionary rather than revolutionary – doing similar things as before, but cheaper, quicker and with more impact. Particularly striking was the impact on the organisation of volunteer telephone canvassers, which was done almost completely over the internet. Whilst in the past nearly everything had been done by post, with some faxing and very little email, in Brent it was nearly all email, some faxing and very little post.
Fundraising over the Internet was also significant, with email-based appeals raising around 25% of the by-election budget. There was some displacement effect from posted appeals, but email appeals proved their worth in allowing immediate needs for extra cash to be quickly turned into appeals and responses.
Emails were also used to communicate with voters who opted in to receiving them. Again, the particular benefit of email was its speed – such as in reminding people just before the deadline for postal votes that they could still apply for one. Some casework came in by email, and it was also solicited through the website (see below). The most convenient aspect of this was the ability to easily forward on the casework to the relevant official or department for action.
The public website, http://www.sarahteather.org.uk (still available, though updated regularly since the election) had a three-fold purpose. First, the existence of a frequently updated and professional website both provided information to journalists doing their background research before visiting the constituency. By its very existence it also helped to reinforce the message that this was a by-election campaign the Liberal Democrats were serious about winning.
Second, it provided a similar function for members thinking about coming to help or donating. And thirdly, it provided a source of information for voters in the constituency.
As with email though, the number of floating voters interacted with was relatively small compared to the number of votes needed to be swung to win the election.
The number of hits on the website continued a pattern seen in previous elections with the peak of traffic being on the day after polling day. This poses an interesting question for academics looking at political parties and their use of the internet. Such studies need an implicit or explicit model of what information people are after by which to judge the internet performance of political parties.
But how well do any of these models explain this consistent pattern of a peak of traffic just after the election? If these models cannot explain this pattern easily, then how suitable are they to use in evaluating the use of the internet?
It also raises questions about what criteria should be used to rate party websites – for example, speed of updating after polling day is rarely, if ever, used.
The other question for academics raised by the Liberal Democrats’ experience in Brent East and similar campaigns is that of the relative importance of email and the web. Most academic studies have concentrated on the web, understandably so as it is more amenable to outside inspection – particularly when compared with internal member-only emails. Yet, if the choice had had to be made between having email or having a website, the former would have been picked. Although websites get the greater study, email would be harder for us to live without.