Before the UK’s General Election in May 2015, the opinion polls had “herded” into a consensus: the Tories (aka, Conservatives) would have a slight lead over Labour but would not have a majority in Parliament. One of major parties would most likely have to form a coalition with a minor party in order to elect a Prime Minister and run the government. Labour could form a left-leaning coalition with the Scottish National Party (SNP), even if the Tories had a larger share of the vote. No one knew what the Liberal Democrats, who had been in a coalition with the Tories for the previous five years, would do.
But the opinion polls were wrong, very wrong. Instead of being a neck-and-neck race, the Tories won a significant-enough majority and did not need to form a coalition government with anybody else. Clearly, the pollsters had somehow overestimated Labour’s share of the vote and underestimated the Conservatives’ share.
This error may have been less troubling if it had been a one-off event. However, opinion polls had proven to be significantly inaccurate in other important elections, including the 2012 US presidential election, the 2014 Scottish independence referendum, the 2014 US midterm election, and the 2015 Israel election.
Theories as to why the opinion polls were so wrong in the UK general election quickly followed. Some blamed “shy Tories”, or voters who lied to pollsters because they were somehow embarrassed to support the Conservatives. Others blamed “lazy Labour”, or voters who would told pollsters they would vote for Labour but couldn’t be bothered to show up. Or perhaps a large number of voters changed their minds just before the election, and the polls could not catch the sudden shift.
An inquiry by the British Polling Council studied the issue and released their findings last week. In short, the pollsters are the problem, not the respondents.
The Inquiry & What Went Wrong
According to the inquiry, instead of reaching out to a wide spectrum of the British population, the pollsters used cheap and easy over-the-internet polls that reached too many younger voters and not enough older voters. Voters over 65 generally vote for Conservatives but are less likely to use the internet as much as younger voters. Plus, those younger voters who were willing to respond to pollsters’ outreach were more Labour-leaning than the younger voters who ignored the outreach, though those who responded were more likely to vote.
In other words, even though opinion polls are supposed to be based on randomly-selected respondents, those who respond to over-the-internet polls are self-selecting by first being online and then by accepting the invitation to participate in a survey.
Yet there could still be the possibility that the voters had changed their minds quickly just before the election. To address this concern, the British Social Attitudes survey and the British Election Study conducted face-to-face surveys after the May election to ask truly randomly selected respondents if and how they voted. These post-election face-to-face surveys gave results that closely matched the actual election’s result. Meanwhile, internet polling conducted before and after the election gave the same flawed results, meaning that there was no “late surge” or sudden shift. Instead, the pollsters’ sampling methods were off.
The greatest problem with these face-to-face surveys is their cost and length. As Andrew Rawnsley noted in The Guardian, these types of surveys are simply too expensive and too impractical during a political campaign when the news media is measuring the “horse race” between candidates each day. Conducting surveys either over the phone or the internet remains the most practical way to measure voters’ preferences.
Opinion Polls’ Declining Reliability
One of the major headaches pollsters currently face is the declining response rates from landline telephone interviews. Cliff Zukin, president of the American Association for Public Opinion Research, wrote in the New York Times that in the 1970s before answering machines allowed people to easily ignore calls, an 80% response rate was normal. Now, after cell phone usage has surged, the average response rate is 8%. US federal law also requires pollsters to manually call cell phones instead of using an automatic dialer, which can be quite expensive when most of the phone numbers do not work. It is also difficult to know beforehand if the person who may answer on a cell phone still lives in the area that would be voting.
Due to increased costs for phone interviews, pollsters have switched to over-the-internet opinion polls. But Zukin is very doubtful about these techniques for reasons that the UK general election demonstrated. He wrote:
Almost all online election polling is done with nonprobability samples. These are largely unproven methodologically, and… it is impossible to calculate a margin of error on such surveys. What they have going for them is that they are inexpensive to do, and this has attracted a number of new survey firms to the game. We saw a lot more of them in the midterm congressional election in 2014, in Israel and in Britain, where they were heavily relied on. We will see them more still in 2016.
Opinion polls are thus very expensive and time consuming if done face-to-face, reach too few people or are still too expensive if done over-the-phone, and are too inaccurate when done over-the-internet. The British Polling Council’s inquiry has proven how these problems can lead to dramatic failures, and because both the media and politicians have come to rely on polling so much, modern democracy faces a serious challenge.
Problem for Democracy
Opinion polls’ unreliability can affect democracy in multiple ways. First, they can cause voters to behave differently. After the UK’s general election, Liberal Democrats complained that polls showing a neck-and-neck race between Tories and Labour allowed the Tories’ message about the damage a coalition government could do gain more traction with voters. Meanwhile, voters ignored the Liberal Democrat’s message about how a majority Conservative government could be harmful.
This complaint aside, voters and supporters could behave differently if a poll says their candidate is either ahead or behind. Voting may seem like a waste of time when the outcome is obvious, so the election ends up not reflecting how the voters really feel. Or a poor result in an opinion poll could cause donors to stop funding a campaign and force a candidate to drop out of the race completely. Because the candidate’s name isn’t on the ballot, there would be no way of knowing if those polls were accurate or not. This issue could be of particular importance during the current US presidential primary elections.
Second, if polls continue to be inaccurate, politicians and voters will lose confidence in them. Without accurate polling, a politician would have to guess what voters want, enact the policy, and then find out in the next election whether he or she was right. Referencing this concern, Nate Silver, FiveThrityEight’s founder, made a passionate defense for the role opinion polls play in democracy while still admitting the problems that his field needs to address. He wrote, “Polls are essential to understanding public opinion on a host of issues that people never get a chance to vote upon. How do Americans feel about higher taxes on the rich? The Keystone XL pipeline? Abortion? Capital punishment? Obamacare?”
Yet if polling has become more difficult, it means that these policy questions have become more difficult to answer. As the British government now tries to renegotiate the terms of its relationship with the European Union this year, these problems pose a risk for several other nations’ foreign policy
Increased Risk for “Brexit”
In January 2013, Prime Minister David Cameron promised voters that if the Conservatives won the general election in 2015 he would negotiate with the EU for concessions and would then offer Britain an “in/out referendum” on whether the UK would remain part of the EU.
Initially, most analysts assumed voters would choose to vote to stay, but there have been growing doubts over Cameron’s ability to negotiate with the Europeans. He has made blunders while dealing with the EU in the past, and last month the Europeans rejected his demands, even though he promised that if he did not receive them he would campaign for “Brexit”, or Britain’s exit from the EU.
The Prime Minister’s greatest challenge is not negotiating with the Europeans but knowing what concessions the voters back home are willing to accept. Opinion polls would normally be a reliable way to determine this, but the polls before the Scottish independence referendum and the UK general election have shown that they are not reliable. This week’s inquiry may show a way towards reforming the polling industry, but the changes are unlikely to come quick enough, assuming that reforms can even be effective. There is thus a significant risk that Cameron may misjudge his own voters, which could easily lead to Brexit.
The debate over Brexit’s pros and cons has been very contentious in the UK. There is also no clear answer over how the British and European economies would fair after a Brexit, though many economists are concerned.
From a particular right-of-center American perspective, Brexit could be considered a sort of godsend. This view holds that the EU is an overbearing bureaucratic, post-nation-state nightmare holding back democratic national governments. Some on the right, especially religious conservatives, may also see how policies dictated from Brussels could damage their European friends’ priorities. Some on the left may also see how other EU policies could help entrench the economic elites’ hold on European politics.
However, this view is shortsighted, if not also misguided. At least from a US foreign policy perspective, a Brexit would be a net negative. Another article could explain the analysis leading to this conclusion since there is not enough space here. In short, though, Brexit would mean the US would lose a valuable partner for dealing with global issues. The UK would likely lose its “special relationship” with the US if Scotland voted for independence in a second referendum in order to rejoin the EU, which may become inevitable after a Brexit.
The British voters should have their say on a possible Brexit based upon what they believe is in their national interest and how it would affect UK foreign policy without having to worry about US foreign policy. Meanwhile, Americans should still monitor and analyze those factors which would increase the political risk for Brexit. At least one of those factors would be the opinion polls, whose unreliability may have an impact on a close referendum.
Mark Melton is the Deputy Editor for Providence. He earned his Master’s degree in International Relations from the University of St. Andrews and has a specialization in civil conflict and European politics. He earned his Bachelor’s degree in Foreign Language & International Trade from Mississippi College. Prior to moving to DC, he worked as a political science adjunct professor at community colleges in Mississippi. He is Providence’s resident millennial (don’t let the premature salt-and-pepper hair fool you, he’s a millennial).
Photo Source: Pixaby