The problem is not that tools such as online polls inaccurate because they are not necessarily so; the problem is that such poll results are conflated and treated as being representative of dynamics obtaining in the ‘real world’.
“Offline Voters are the blind-spot of e-campaigns”
Anecdotal evidence and academic studies have reinforced the view that new communication technologies including social media can drive political action and influence voter behavior. E-campaigning describes the deployment of new communication tools and platforms (such as email, social media, Whatsapp, websites and blogs) to campaign. Whilst new communication technologies provide opportunities for political parties and candidates to engage voters, the hype around technology’s presumed power to impact electoral outcomes has been misleading.
By design, e-campaign strategies target potential voters online but by default, the results of metrics, polls and other predictive tools tend to be viewed uncritically as reflecting the views of offline voters as well. Political parties that get giddy at the sight of social media’s vanity metrics (i.e being overly impressed with having huge numbers on social platforms without considering quality of engagement or impact) often get lulled into a false sense of security, wrongly assuming that their social media numbers signal overwhelming support at the polls. In many instances, post-election defeat stuns political parties or candidates that are hugely popular on social media because online polls would have indicated that they would win.
The problem is not that tools such as online polls are inaccurate (because they can plausibly mirror the wishes of those who participate in them); the problem is that such poll results are conflated and treated as being representative of dynamics obtaining in the ‘real world’.
E-campaign tools, such as online polls and viral hashtags, measure the activities and capture the preferences of online voters but those same results cannot be used to understand, strategize on and anticipate the activities or preferences of offline voters (in this case voters who do not have regular access to the Internet). Offline voters are the blindspot of e-campaigns, particularly because confirmation bias triggers immediate acceptance of favorable digital polling results, metrics and data. To avoid the pitfalls of euphoria induced by misinterpretation of e-campaign outcomes, the following factors must be foregrounded.
Start with the ‘real world’ data
Elections do not happen in a vacuum. Before getting excited about the outcomes of e-campaign monitoring, evaluation and polling tools, parties must start with the real world data, in order to put online support into proper context. For instance, real world data indicates that in Africa, the voting majority is rural-based, whilst the digital divide is pervasive because Internet
penetration in rural areas is still limited. Consequently, a Facebook poll that shows a candidate is the most popular one reflects the exclusive views of those who own internet-enabled devices, who can afford data or have access to wifi, who have the digital literacy to use the Internet and who have enough interest in politics to participate in a Facebook poll. Compared to the offline voter (such as the rural-based majority), the online pollees are arguably in the minority.
Consider the ‘real world’ divisions
The question of who can access e-campaigns invites some reflection upon real world divisions that determine who can use the Internet. For instance, class is a factor in terms of Internet access when considering affordability of mobile data or Wi-Fi. Urban populations are not homogenous; therefore many people may not have the resources to be online frequently making social media a space that is largely dominated by the middle class, the elites and also the diaspora-based communities (who are highly active online, driving conversations about their home countries). Additionally, age is a factor insofar as participating in online activity of any kind. Parties seeking to appeal to young voters (who are mainly from families with means) will likely find them online, whereas older voters may not be keen to use those platforms. Similarly, gender can also be a factor in terms of who can be online and how they engage, particularly in highly emotive subjects such as political polls, hashtags or discussion forums.
Reflect on voting status
Elections are about voting and voting is guided by specific rules that determine who can vote, how, when and where. E-campaigns can sometimes attract legions of fans; have hashtags that go viral and metrics that indicate that certain parties or candidates are very popular. However, as long as those metrics capture the views and preferences of individuals that are not registered voters then any predictions made about the outcome at the actual polls will be entirely misinformed. So before a party or candidate celebrates winning an online popularity poll, they should consider whether those who took part in the poll are registered voters or not – and whether they will actually go out and vote on polling day.
Contemplate the proximity of online users
Related to the voting status of online users (i.e whether they are registered or not) is the issue of proximity and willingness to participate in the actual voting. In many African countries, Diaspora-based communities have largely supported opposition parties, often regarding ruling party leadership with hostility and holding them responsible forcing them to leave their homelands to eke out a living in foreign lands. These Diaspora-based communities are usually a formidable presence online because they dominate conversations and are bold in criticizing governments (at little personal risk given that they are far away) whilst unabashed in supporting opposition parties. Since e-campaign tools such as online polls do not disaggregate in terms of whether respondents are in based the country or not – it is possible that users who are not based in the country, who are not registered voters and who (even if they
were registered) are highly unlikely to go and vote participate in these polls and skewer the results.
Any elation over the outcome of online polls or the adulation a party or candidate receives online must be weighed against the context of an offline majority who may feel and decide otherwise on polling day. E-campaign strategies are highly effective in engaging voters who are online, but to reach the voters that do not access Internet; political parties and candidates must consult real-world data, rely on evidence-based insights as well as prioritize connecting with grassroots structures and responding to their lived realities.