The weaponisation of surveys

Polls Don’t Just Measure Opinion — They Aim to Shape It

By now, Maltese voters should have learned a hard lesson: the polls that are published don’t just reflect public opinion, they can also distort it.

For years, surveys by Vince Marmara and MaltaToday have set the political agenda, competing with lotto and Super5 draws for the growing obsession with numbers.

Carefully-selected headlines splash figures, parties and pundits spin the numbers, and citizens start believing that the ‘public mood’ has already been decided, and shift their perspective accordingly.

However, as the June 2024 elections amply proved, those polls can be spectacularly wrong.

The problem is not simply accuracy. Margins of error, sample selection and interpretation, to mention but a few, all contribute to the high subjectivity of the results that we are bombarded with.  

The deeper issue is, however, the way in which these polls are used, or rather misused, to shape narratives, energise supporters, discourage opponents and sometimes, quite often in fact, even manufacture consent and build mirage-like perceptions that turn out to be illusions.

From Snapshot to Weapon

In theory, polls should be a snapshot of where voters stand. In practice, they are anything but neutral. Academic research shows that polls can nudge people by as much as 10 to 15 percentage points, depending on how results are reported. 

When surveys paint one side as the overwhelming favourite, many undecided voters fall into line. Others keep quiet, afraid their view is unpopular — what communication scholar Elisabeth Noelle-Neumann called the “spiral of silence.”

In Malta, where the political scene is small and polarised, the effect is magnified. A single poll can dominate the national conversation for weeks, or at least until the next ‘draw’.

Some notable local failures

Let me give a few examples from the past few years, when polls were catastrophically wrong, and when they influenced decisions:

In Labour’s 2020 internal leadership race and election, Vincent Marmarà’s polling had predicted a comfortable lead for Chris Fearne. At times, he had predicted Fearne’s support as being high as 68%, with Robert Abela trailing at just under one-third of the vote.

The actual voting told a different story, with Abela emerging victorious, going from the poll-predicted underdog to winner in less than a week.  This huge misfire underlines the fact that even polls from local figures like Marmarà, who enjoys prophet-like status among the Maltese, especially among his loyal Labour followers, can be weaponised to create narratives of inevitability, ephemeral narratives that disappear once real votes are cast.

Five years ago almost to the day, in August 2020, MaltaToday had published a survey purportedly held among a stratified sample of the PN’s tesserati. This was after the PL contest above, and was held to gauge the tesserati’s preference between Adrian Delia, Bernard Grech and Roberta Metsola, in the run-up to the inevitable PN leadership contest that was looming. The results that they published gave the following indications of preference:

• Bernard Grech: 63.1%;

 • Roberta Metsola: 39.5%;

 • Adrian Delia: 23.3%.

These results impacted heavily on the choice of a main candidate to run against Delia.  Metsola, then an MEP, stepped aside, perhaps to avoid splitting the vote. Grech contested Adrian Delia and the rest is history. It remains a mystery, considering subsequent real electoral results, where those figures had come from, and why they were used to achieve that outcome.

The proverbial cherry on the cake was the case of the 2024 election results which turned out to be a major shock for Labour. Both Marmarà’s polls and MaltaToday’s surveys suggested one outcome – another PL landslide in the MEP and Local Councils elections.  Reality delivered another, dealing a strong blow to the reliability of these polls.

The backlash in 2024 went beyond statistics. Disappointed Labour Party agents in the counting hall were heard commenting, “This was another ‘partnership’, referring to the 2003 claims by Alfred Sant that the referendum had been won by the anti-EU Membership side, and some internal dissent was also evidenced.  Outside the counting hall, voters felt manipulated. After weeks of being told a story through polls, they discovered the numbers that had been flashed in bright lights didn’t match their lived reality.

What does this tell us?

These were not just local hiccups. They echo mistakes worldwide.

Remember Brexit? Most surveys said ‘Remain’ would edge out ‘Leave’. The UK left the EU on the basis of the opposite outcome in the referendum.  

Or the 2016 US election, when polls across America got it so wrong, totally underestimating Donald Trump’s support. Again, the polls didn’t just get it wrong; they influenced how people thought about the contest.

Unfortunately, people have a short memory, and lessons are not learnt. Bernard Grech, one of the main architects of the 2024 victory was pressured, and maybe even personally influenced, by survey results to resign from the Party leadership. How could he have slumped so badly and quickly in public popularity, when nothing in the political arena could have been a causative factor for it.

He was not embroiled in any scandals or wrongdoing of any kind. In my humble opinion, that casts some serious doubt on the motivation and manipulation of these particular surveys. The desired end result seems to have been achieved, with the PN caught up in a leadership race, while the PL is gearing up for general elections.

Not Mirrors but Megaphones

Polls are often treated like mirrors of society. However, they are acting, or rather being used more like megaphones, amplifying certain narratives while silencing others.

This isn’t new. Over one hundred years ago, Walter Lippmann, an early 20th-century journalist and political commentator, and much more contemporarily Noam Chomsky, a modern linguist and critic of media power, both argued that elites, including media houses and their owners, use tools like polls and mass communication to “manufacture consent”, by subtly shaping the public’s opinions.  

In Malta, we see it when politicians, and their mouthpieces, seize on favourable polls to declare “the people have spoken” long before a single vote is cast, and sometimes prematurely expressing themselves on the results, even before they are made public, indicating prior knowledge.

What Can be Done?

Does this mean that we should ban polls? Of course not. They are part of the democratic process. But just like kitchen knives, they are only relevant if used properly for the purpose for which they were designed. 

We need to consider them critically and with scepticism on some key areas.

Margins of error matter, especially in a small pool like Malta, where one single reply in a small sample can skew the result, and where a 3% swing can decide an election.

Full transparency on the methodology is a must; who was asked, how where they chosen and how were the questions formulated?

Was there a focus group involved in the preparation or was it just an individual’s whimsical drafting?

The timing is also crucial. Polls close to elections carry extra influence, but aren’t necessarily more accurate, in fact rather the opposite.

As citizens, we should demand more transparency from pollsters, much more caution and responsibility from the media who use these polls to boost readership and followings, and we should demand humility from politicians who succumb to the temptation of trumpeting numbers as gospel truth, when it suits them best.

A Call for Critical Readers

We cannot turn our clocks back, at least not for more than one hour on the last Sunday of October. Polls are here to stay. But if 2024 taught us, or rather confirmed, anything, it’s that we Maltese voters cannot take them at face value.

The danger is not just being misinformed. It’s being subtly steered by being told what ‘everyone else’ is thinking, we start to ignore what we ourselves think.

Polls will always be both tools of knowledge and instruments of influence. The only defence is a critical and informed public, who is wary and always willing to read beyond the numbers chosen for the headlines.

Martin Bugelli holds a BA (Hons) in English from Goldsmiths College, University of London, and an MA in Mass Communications from the University of Leicester. His Masters dissertation had analysed the impact of printed media on Maltese politics.