Each time a Sunday newspaper carries out a survey on the political situation – and we’re having regular ones these days – the reaction is always the same.

Everyone is happy.

Because numbers always are a matter of interpretation and if you look carefully, there is always something positive that can be highlighted.

The glass is always half full, if you want to see it that way.

If, for example, Robert Abela lost a fraction of support, the Labour Party can always claim that it is still way ahead and looks certain of winning the third national election in a row.

And if, for example, the Nationalist Party is continues to trail heavily, it can still argue that Bernard Grech has made inroads when compared to his predecessor.

But we all know that the biggest survey is not done via phone calls.

It comes once every five years, and the next appointment is getting closer.

Social media

The impression one gets when monitoring the social media each time a survey is published is that the people who post their comments are getting instructions from above on what to say.

This is because most of what is written is more or less the same, on either side. It’s as if no one dares to move outside the perimeters of the directives that have been given. If the orders are to focus on a particular angle, then nobody strays from it. And then you get the same image posted and re-posted by different individuals according to their party’s narrative.

It’s easy to imagine a small group of people discussing on what is the best approach after a survey is published, and then sending out the message to a wider cluster which then inundates Facebook and other platforms with similar, sometimes identical, musings. Others just copy and paste, perhaps not even understanding what they are re-posting. So long as it backs the party, then it’s good.

This is how it works, and this is how political parties these days keep up the morale of their supporters. Ministers, MPs, top party officials and the diehards doing their best to boost confidence.

We’re still in the lead, and will win, the Labourites say.

We’re still behind, but we’re catching up fast, the Nationalists reply.

Everyone is happy.

And so it goes, with the public believing without asking questions. Most of the public, anyway.

Accuracy

Surveys are just an exercise to gauge public opinion, earmarked to create a discussion. But they are not perfect.

We all remember how surveys indicated that Donald Trump would not have become United States president in 2016. And yet he did, with all that followed.

The majority of the Brexit polls in 2016 were also wrong. Many of them showed that the Britons were to decide to stay in the European Union but when voting day arrived, the people chose to leave. We’re still feeling the effects of the chaos that this decision had on Europe, and its relations with Britain. (Last month, a poll indicated that the majority of those who voted for Britain to leave the EU are regretting their decision, but there is no turning back now).

Closer to home, the predictions in 2015 were that the Maltese were going to vote to abolish spring hunting. All the surveys that were published pointed in this direction. But, when the time came, the Maltese voted for spring hunting to stay.

Then, in 2017, all surveys showed that there had been a “clear” shift which would have significantly reduced the distance by which the Labour Party would have won the election. Remember, in 2013 Labour had won by 35,107 votes, establishing a record. Surveys in 2017 were showing that the PN would lose by a much smaller margin – around 20,000 or so. What happened is that Labour won by an even bigger margin – 35,280 votes, which Labour increased to 40,000 by not counting in the votes obtained by the Partit Demokratiku candidates who contested on the PN ticket as part of a coalition. Even here, the surveys were wrong.

These are four cases which clearly show that one cannot be too sure that what is indicated in surveys will come to be.

People who conduct surveys always say that there is a margin of error. This margin depends on determined factors, including the number of people who are contacted to share their opinion to unknown strangers over the phone.

But it has happened that this margin of error was not enough, and the real, official results have often resulted in a very different picture of what surveys had previously shown.

Don’t know

One important factor which is often not given due consideration by people who are analysing surveys – and who have every intention to put the party they support in good light – is the number of people who, when asked, do not give an opinion.

These often form a sizeable percentage of the respondents, and their decision not to give an answer skews the outcome of the survey. The higher the percentage of non-respondents, the higher the possibility is that the results of the survey are not accurate.

Some surveys remove the non-respondents from the equation in an attempt to give a clearer picture of the situation. But, again, when this happens the survey “loses” that substantial group of people who, it is claimed, are undecided, but who have an opinion they do not want to share. If they had to give that opinion, the probability is that the outcome of the survey would be different.

People who say they “do not know” do have their own opinion, and will give it at the appropriate time. Discounting them results in a distorted view of the situation.

Lies

There are then others who may feel pressed to give an answer and are untruthful in their replies.

It could be because they believe that their name will be passed on to the political parties. “If they have my number and I’m getting this call, it means that they know who I am and will tell the parties what my answers were”. This is a common argument that is made.

And so they lie.

It could be on anything, not only politics. If, say, a survey is being carried out on which fruit one enjoys eating most, one says bananas when they really prefer strawberries. Sometimes just for the fun of it. And if people can lie on fruit, imagine how more inclined to lie they would be when asked about their political opinion.

Instead of saying “I don’t know” or “I don’t want to answer”, there are people who prefer to give an answer which is different from their true feelings.

Other factors

There are other factors that can influence a survey.

One of them is the sampling. One can try hard to get the views of as wide a spectrum of society – and taking into consideration levels of education, income, social standing, age and more – but a survey cannot capture all the dynamics of the population.

The type of questions asked and the way they are formulated also play a part in the outcome of a survey. Sometimes respondents feel that they are being led to give an answer they do not want to convey. If they get the impression that the survey is being conducted by one particular party – even when this is not the case and the company conducting the survey presents impartial questions – then their answers will most likely veer towards what that particular party wants to hear.

This is more so if they are specifically told that the survey is being carried out on behalf of a newspaper which the respondents believe leans heavily towards a particular party.

People may also be caught in a certain mood which does not reflect their true feelings.

For example, if a Labourite is approached in the week after a request for a permit has been denied, there is a possibility that they will say that they will be voting PN, or not voting. But then, when the election comes and all things are taken into consideration, that person will still vote PL.

Similarly, a Nationalist might be angry at the way the party got entangled in four years of internal strife, and might say to the person conducting the survey that they will not vote or else vote Labour. But then their conscience does not let them tick the red boxes on the day of the election. Neither does it allow them to stay home.

Election Day

This is why what counts is what happens on the day people are asked to vote.

All surveys then become obsolete as it is what people do with their ballot sheet that determines who will govern the country.