Archiving online election campaigns

Explaing the process behind how we preserve Australia's election websites

As the formal campaign period for the 2013 Australian federal election gets underway, the pace of web archiving work has to move up a notch or two. Collecting election campaign material is a priority and an energising experience for most web archiving programs because election campaigns bring into sharp focus the importance of web archiving, not the least because online election materials represent some of the most ephemeral and interesting of online content. In addition, such intense use of the Internet drives certain platforms or forms of communication and publication that require the web archiving program to deal with the challenge to collect material in innovative or just plainly pragmatic ways. And there is no time to delay, we have to mobilise staff and respond quickly. Fortunately we have considerable experience to draw on.

The National Library’s PANDORA Archive has a long history of collecting election campaign material. Indeed some of the earliest collecting we undertook involved the 1996 federal election. Since that time we have collected content from more than 50 election campaigns including all Australian federal elections since 1996 – that is seven in all including the 2013 election – virtually all Australian state and territory elections since 1999, a number of federal and state by-elections and a small number of local government elections.

The scale of collecting for individual elections has certainly increased – an obvious fact, perhaps. For the 1996 federal election we collected a mere 5 websites, 22 in 1998, 132 in 2001, 271 in 2004, 371 in 2007 and 329 in 2010. It is obvious and unremarkable that the scale of these collections should generally increase since the scale and use and perceived importance of the Internet has increased so significantly over the last 17 years. Also, the tools and workflows available to us to collect content have greatly improved. Resources available to us have not increased so we have had to improve our systems and processes. The fact that our collecting in 2010 was not greater (at least in the number of sites) points, perhaps, to the scalable limit we have reached with our current web archiving infrastructure and processes. Moreover, while our systems and tools and workflows and skills develop and improve efficiencies, the persistent banes of web archiving, those limiting constraints, remain: the technological complexity of the medium and the challenge to harvest it; the seeminly exponential increase in the scale of online publication and communication; and the legal constraints that require the clearance of rights permissions before collecting and archiving can be undertaken.

As with our web archiving endeavour generally, we aim for comprehensiveness but have to settle for best effort. We aim to be disinterested in our assessment and collection of content in order to fully and accurately represent what the online world presents, but we necessarily have to prioritise what we do. This, together with the aforementioned constraints, does tell against the curation of completely satisfactory collections. So, for example, in the 1996 federal election the Coalition Policies website was collected but not the ALP website. This was not a matter of intended bias but an outcome of technological limitation. As I have written of previously, we were originally unable to collect the highly significant Liberal Party site, jeff.com.au for the 1999 Victorian state election, since the rights to do so were not secured before the site disappeared. Through one of those little miracles of digital survival we were able to find a copy of this site over a decade later, so we never give up hope – although it is always preferable to collect the site at the time of its manifestation.

Miracles have been less forthcoming in respect to collecting media and social media and these present form the biggest lacunae in our election collections. With few, but notable exceptions, the major online media outlets have rarely provided us with the permissions necessary to collect their election coverage. It must be admitted that the rights issues in respect to news sites are almost intractably complex. And this is why the extension of legal deposit provisions to electronic resources would be such a help towards achieving the long term preservation of our social and cultural documentary heritage – it would reduce some of the rights complexity in respect to the collecting and preservation process. Social media is equally complicated, not the least because it mixes communication with publication. Even identifing rights holders is a major problem. This is compounded by the strong assertion of rights by the service platforms themselves. So not only are the rights of the content owners to be considered but also those of the platform owners. Look at our collections and you will find some YouTube, a tad of Twitter and no Facebook. The reasons are complex.

So our election collections are unlikely to achieve the comprehensiveness of our ambitions. Nevertheless we do manage to pick up some wonderful records of the frenetic, fascinating and frustrating few weeks that are the election campaign. A couple for former campaign sites that come to my mind and seem apposite right now (with the return of Kevin Rudd and Peter Beattie to the hustings) are the Kevin07 campaign from the 2007 federal election and the ‘Heading in the Right Direction’ campaign by Peter Beattie for the 2001 Queensland state election.

The 2007 federal election campaign put a strong emphasis on video media which certainly set the challenge for us. Collecting YouTube videos embedded in sites is a major challenge for web archiving, firstly in the harvesting and even more so in the rendering in a meaningful way from the archive. Sometimes we have to compromise. We collected the Kevin07 website a day or two after it emerged and did our best to gather (and render) the embedded videos in that first harvest (with a great deal of ‘hands on’ tweaking it must be added). Resources would not allow us to do the same for all of the instances we archived so we focused on collecting the Labor TV website where most of the videos were located. This still required a lot of work although we could employ some scripting techniques in this case to get more outcome for the laborious ’hands-on’ work.

Even the older 2001 Peter Beattie website presented a media format collecting challenge, but one we felt we needed to step up to. That Queensland election campaign included the ‘ear-worming’ and ostensibly very successful campaign jingle ‘heading in the right direction’. These campaign commercials were included on the Peter Beattie website and we had to try and collect them. They were presented in both Quicktime and Windows Media format. When multiple formats are presented resource often don’t allow us to manage to collect all formats. In this case we collected the only the Quicktime format. Such formats will inevitebly present preservation challenges – that is, how do we keep them functioning and accessible as browser technologies change? This will have to remain the subject for a future blog post! Anyway, the campaign commercials (in Quicktime) are still working – I checked – though, be warned, you may have to coax your browser a little.

During election campaigns it is not only campaign material that becomes a priority to collect, but also government, particularly ministerial websites. Websites become less active during the caretaker period which gives us some time to consider the sites we should collect. As was discussed in the previous blog post concerning the recent prime-ministerial leadership challenge, changes in government or even changes within the leadership structure of the government mean government websites will change and we do our best to collect a snapshot of (some of) them before they do transmogrify, perhaps with little evidence left, before our eyes.

The many collections of past online election campaign material curated by the National Library (and partners) can be accessed at the PANDORA Archive.

Add new comment