Abstract
Introduction
This special issue arose from a collaboration between researchers at the University of Manchester and Duke University that was funded by the Economic Social and Research Council [ES/J019453/1]. The project sought to examine the nature of voter contact in the digital age in comparative context. In particular we were interested to understand the extent to which parties and candidates were seeking to mobilise voters during national election campaigns in the US and UK using new communication channels such as email, sms/text messages, and social media. After benchmarking the use of these tactics in national campaigns we also wanted to address the 64 million dollar question of whether these methods actually worked. Does online contact actually increase the likelihood that someone will turn out to vote? And if so, how effective is it in comparison with other more conventional methods? Our interest in researching these questions was triggered by the general growth in use of these tools across both countries and particularly the high profile e-campaign of Barack Obama to reach the White House in 2008.
In the course of conducting our research on the effects of new forms of voter contact we faced some immediate questions about the extent to which internet-based methods can be meaningfully compared with more traditional offline modes. Given the rapid proliferation in the types of online communication channels now available, how far can a simple binary division into online and offline channels really capture this new complexity? How do we take account of the increasingly complex and indirect web of connections through which parties’ online messages can be conveyed and also the extent to which parties themselves can control this? While email is perhaps the closest approximation to the more conventional ‘surgical’ methods that parties use such as direct mail and phone calls, platforms such as Facebook and Twitter are more than simply conduits for content but are themselves immersive new social spaces.
The idea of the internet and particularly social media serving as a new context for mobilisation prompted us to think more generally about how far the field had moved and broadened both substantively and methodologically since the seminal work of Rosenstone and Hansen in the early 1990s. What new forces now impinge on voter contacting and how far do the old channels remain potent? In particular, how have advances in technology affected parties’ strategies and improved their capacity to target voters? What are some of the ‘new’ contextual factors that we might need to consider, particularly beyond expressly political and institutional channels? And finally what new methods are available to best capture this wider complexity of affective forces? To explore these questions we invited leading scholars within the field to reflect on the current and past state of voter mobilisation research as well as to provide us with some fresh empirical evidence about the relevance of context in shaping parties’ GOTV efforts. The papers gathered here represent the fruits of that effort and provide a rich insight into and update on the impact of a variety of technological, social, economic and political influences on parties’ GOTV efforts.
We start by imposing a wide-angle lens on the notion of context and look at the role of time itself and how shifts in the broader political landscape over the past 30 years have influenced the nature of voter contacting. In ‘Mobilization, participation, and political change’, John Mark Hansen provides a masterful review of how shifts in the political environment have created new opportunities and new impediments for voter mobilisation in the US. These include the increasing intensity of competition for control of national government which has made citizen participation more consequential and mobilisation, correspondingly, more attractive. The widening polarisation within the electorate, he argues, has further raised citizens’ commitment to participate and, correspondingly, their readiness to respond to mobilisation attempts. Furthermore on the supply side, the deregulation of finance in US federal elections provided an impetus toward more extensive campaign fundraising efforts and, correspondingly, to broader opportunities for citizens to give. Finally, the shifting technological landscape and particularly arrival of websites, email, and social media, has vastly reduced the cost of both contact – and perhaps even more notably re-contact – with potential voters, donors, and campaign activists, thereby increasing the likelihood of their influence on legislation.
Time also plays a role in the second review article by Green and Schwam-Baird. In ‘Mobilization, participation, and American democracy: A retrospective and postscript’ the authors review how the literature has evolved in its understanding of voter mobilisation from a methodological perspective, and profile the rise of the experimental research agenda that has emerged in recent years. The authors note how the shift toward experiments has reintroduced and popularized an alternative context for research which took researchers back into the ‘field’ to observe voters in their ‘natural habitat’ and measure responses to a variety of mobilizing stimuli. The results have not only reshaped the field of GOTV studies and ensured that those who study elections now have a much clearer sense of which types of mobilisation activities are effective, they have also now spread out across the discipline and supplied a new set of tools to be used in the search for causal inference.
The articles that follow narrow the lens somewhat to more specific aspects of context. New communication and data technologies receive considerable attention with three papers – Aldrich et al., Panagopolous, and Hatch – all of which profile the spread of digital tools and databases within parties and seek to understand the implications of this for the voters and parties themselves. For Aldrich et al. the focus is on the former and particularly on how widely and effectively new social media or web 2.0 technologies – blogs, twitter, social networking sites – are being used for campaign contacting by parties and their supporters in the US and the UK. The article takes on board and develops Hansen’s key point about how the low cost and viral quality of these tools dramatically expands the ‘reach’ and frequency of voter contact and also re-contact that is possible, compared to more traditional methods. Despite finding the US to be significantly more advanced in targeting voters online, the authors find a closer parity in mediated forms of such contact, with the UK electorate reporting healthy rates of online election messages from their friends and families. The impact on voting is seemingly yet to be realized in any major way, with traditional methods still yielding the biggest return on investment in terms of increased likelihood to vote. Closer examination, however, indicates that the effects do vary by age, in that older people in the US do appear to be more receptive to online messages from parties than their younger peers, while younger voters in the UK respond more positively to mobilisation messages mediated through their friends.
Remaining with a focus on the impact of technological change on the quality and quantity of voter mobilisation, Costas Panagopoulos switches attention to look at the supply side of the equation. In ‘Bases loaded: Changing campaign strategies in US elections’ he argues that partisan campaign strategies in US Presidential elections since the turn of the century have been moving to concentrate on preaching to the converted loyal core of supporters rather than expending efforts on pursuing undecided or swing voters. For him this shift is attributable, at least in part, to changes in the technological context in which contemporary campaigns operate that made their first notable appearance in the 2000 presidential election. This cycle, he argues, saw significant advances in microtargeting, e-campaigning, and e-mobilisation strategies and the application of new behavioral sciences related to voter mobilisation. While persuasion through face to face methods remains an option it is a far riskier and resource-intensive approach. The evidence of this increasing switch to focus on base mobilisation in elections, he concludes, can be witnessed through the trend toward an intensifying partisan polarisation in America.
The notion of technology as context for voter mobilisation efforts is revisited in Rebecca Hatch’s contribution ‘Party organizational strength and technological capacity: The adaptation of the state-level party organizations in the United States to voter outreach and data analytics in the internet age’. Rather than looking at the role that context plays in the external efforts at voter communication and outreach, however, she turns the lens inward to examine how parties’ internal technological capacity affects their capacity for undertaking voter mobilisation programs through use of sophisticated data analytics and targeting strategies. Using data from her original 2011 survey of the state party organisations in the US, this research shows that party organisational strength and party technological capacity are interrelated and overlapping concepts and are essential to understanding the parties’ programmatic and voter outreach capabilities.
Pattie and Johnston’s article returns us to the UK and the more practical economic considerations surrounding parties’ efforts to mobilize. In ‘Resourcing the constituency campaign in the UK’ the authors take a close look at British political parties’ campaign capacities and paint a decidedly pessimistic picture. While some local parties are resource-rich, many are not: only half of all Conservative and fewer than one in six Labour and Liberal Democrat constituency parties had an annual turnover in 2010 exceeding £25,000. For most local parties, campaign resources depend primarily on their own local fundraising initiatives – but the yields tend to be low. Even in key marginal constituencies, many local parties are facing increasing difficulties to resource their campaign activities. The longer term prospects for maintaining current levels of local voter mobilisation in the UK, they conclude, are dim and point to the likelihood that a widening section of the electorate will be excluded from future campaign and mobilisation efforts.
The article of Fisher et al. retains the focus on Britain and provides an important follow-up to the work of Pattie and Johnston in confirming how far local campaigning actually does matter in the UK for getting out the vote. In ‘Is all campaigning equally positive? The impact of district level campaigning on voter turnout at the 2010 British general election’ the authors use survey data collected among local party operatives to compare the impact of campaign efforts across the three main parties on levels of turnout. The authors reveal that although resources clearly matter as a baseline for generating activity and intensity of effort, there are also important aspects of partisan outlook, choice of techniques, and the wider electoral context in the UK that influence the extent to which a campaign can successfully mobilize voters.
The work of Schmitt-Beck and Partheymüller in ‘Struggling up the hill: Short-term effects of parties’ contacting, political news and everyday talks on turnout’ turns to look at the impact of wider social forces as drivers of turnout, and directly compares these to the impact of party mobilisation efforts. Taking advantage of the over-time data provided by the 2013 German panel study, the authors compare the impact of social communication to that of media readership and finally partisan messages on likelihood of voting. They measure each information source while controlling for the other two and the panel element of the study allows for a powerful test of the effects of each information source. The results show that media exposure in broadsheet newspapers is the strongest stimulus to turnout, followed by parties’ mediated electioneering efforts. In a final twist, the authors also find that under certain conditions exposure to information sources can demobilize voters.
Finally, in ‘Cascade or echo chamber? A complex agent-based simulation of voter turnout’, Fieldhouse, et al. offer a new and innovative approach to modelling the impact of context on voter decisions. Rather than modelling the impact of individual components of the context in a linear fashion through regression methods, the authors take the ambitious route of simulating the wider world of influences on an individual using an agent-based simulation technique. Using a highly complex model they describe voter behaviour with a wide range of relevant individual level factors such as socio-demographic characteristics, socialisation processes, habit and inertia, ‘rational behaviour’, political interest, social norms, political discussion, and interpersonal and party mobilisation. The direct and indirect effects of campaign mobilisation on individuals and others in their social networks are then simulated, allowing for assessment of how far the effect spreads beyond the primary mobilisation target across the population and whether or not it survives over time. The findings are significant in that second-order mobilisation effects are found to exist, even among those who might have voted anyway. Furthermore these effects are then cemented through helping to stimulate habitual voting and increases in political interest and party affiliation, which reverberate over time, creating long-term spillover effects.
Beyond any substantive contribution that this special issue makes to our understanding of the varying and increasingly complex contexts surrounding voter mobilization, the articles also provide an opportunity to take a more reflexive review of the data and methodological approaches that can be used to explore this important topic. Matching the direction of the literature more generally, the most common approaches used here are statistical analyses of post-election survey data, i.e. they are based on individuals’ recall of having been contacted. While such techniques run into obvious problems of over-reporting and difficulties in isolating genuine causal effects, they remain the only viable source of information for researchers to profile how many people are being contacted during the course of an election, and perhaps more importantly who it is that is being contacted and how. Furthermore the use of panel study data by Schmitt-Beck and Partheymüller shows how over-time survey research can provide an acceptable means of ‘blocking off’ the intrusion of selection effects. Finally, as the work of Fisher et al. and Pattie and Johnston demonstrates, when survey data are joined together with aggregate observations about the wider political and socio-economic context, it becomes possible to provide a richer multi-layered understanding of the individual, organizational, and contextual factors that bring voters to the polls or conversely insulate them from party contact. Of course, as Green and Schwam-Baird point out, the ‘gold standard’ in methodological terms for unpacking causal dynamics are field experiments, and we can only hope to see more and wider application of these methods into new geographic regions beyond the US. Finally, Fieldhouse et al. bring us perhaps closest to the cutting edge of innovative research methodologies in their search for how, when, and why mobilization works. Shaking off the restrictions of the earlier approaches, agent-based modelling assumes a dynamic and complex environment where several social processes occur simultaneously. Somewhat paradoxically, the proximity of this simulation to the real world in which voters live brings home the gargantuan task researchers’ face if they were ever to seek to control and manipulate the effects of those wider social and individual influences. Were this volume to be written 10 years hence, however, we have a sneaking suspicion that serious and successful efforts will have been made to do just this.