Welcome to Eureka Street

back to site

Profitless prognostications



One of the minor annoyances of the United States election was how unreliable the polls turned out to be, particularly in some key swing states. Unsurprising, but still annoying for those hoping for a massive repudiation of the Trump presidency. In the large scheme of affairs the failure of the polls to predict is insignificant, but it does raise interesting questions about its implications for public life in the United States and in Australia.

'I voted' stickers (Unsplash)

One of the many reasons given for the mismatch between the polls and the election result had to do with the selection of people who were polled. It was argued that of the people contacted, five out of six either did not answer the call or refused to disclose their voting intentions. This meant that an apparently random group was in fact selective. It represented the minority of people who cooperate with pollsters. If the failure to respond indicates people’s aversion to polling and not merely the pollsters’ inability to contact them, it may also make us question the honesty of those who did respond.

That such a high proportion of people failed to respond to pollsters and their electronic simulacra suggests widespread lack of trust in the value of political polls and a reluctance to participate in them. It may also suggest a more pervasive lack of trust in the political process itself and the perception that it serves only the interests of those who control it, whether these be characterised as the 'Washington Swamp' or as opportunistic populists. People may not see the polls as dispassionate enquiries into truth but as instruments that serve those who commission them. Those who are polled are also used as means to further someone else’s interests.

There are good grounds for this suspicion. Polling is a competitive business operated for profit. The results of the polls are sold to media companies or to political parties, which then adapt their own rhetoric to fit the polls. The polls themselves can also be used to provide media stories favourable to the interests of those who commission them. Because businesses run for profit run the polls, too, they will also try to cut the costs involved in contacting people. From employing local people they move to distant call centres, and more recently to automated calling in which the human contact between the person called and the pollster is minimal or non-existent. As a result those contacted increasingly judge that they, their opinions, and their time are taken advantage of for someone else’s profit. They refuse to provide information about their intentions as part of a broader policy of not responding to any cold caller.

If this is so, the polling process that is designed to give confidence in forecasting the future may paradoxically lead to an exponential loss of confidence and trust. The media lose trust in polls and cease to commission them. Politicians lose trust in voters because they can’t predict their behaviour. Polling companies lose revenue and the capacity to improve their polling. Ordinary people lose trust both in polls and in the political process itself.

The worm at the heart of this collapse of trust is the assumption that information about people’s choices and human goods may properly be bought and sold in a competitive market, and that its collection may safely be entrusted to profit-seeking companies. When economic imperatives control the collection of information, automated processes will inevitably replace face to face conversation. They are much cheaper. People, however, continue to believe that information about important personal choices may be shared only in a conversational context where trust is built. Treating the collection of information as primarily an economic activity will corrode this necessary trust.

This is part of a wider loss of trust in political process and its actors that takes place when competitive individual economic relationships with minimal government participation are thought to guarantee human welfare. The result is harmful for the economy and alienating for people, as has been shown in the response to the coronavirus. Trust in governments and their leaders initially increased when they gave priority to the health of the community over the economic interests of individuals, while supporting people economically in the initial crisis.


'People, however, continue to believe that information about important personal choices may be shared only in a conversational context where trust is built. Treating the collection of information as primarily an economic activity will corrode this necessary trust.'


At the same time, however, the crisis exposed the harm done by previous policies in which governments cut spending on health and other community necessities by contracting them out to profit-making companies. Decision making was left to managers whose overriding criterion of satisfactory service was economic performance. They increased profits by cutting back on food and other essentials, and by relying on poorly skilled and paid casual workers. The catastrophic results for the elderly and for health workers of this abrogation of responsibility by governments can be seen in the evidence given to the Federal Royal Commission into Aged Care and to the Victorian Inquiry into Hotel Quarantine.

These failures inevitably erode trust in government and in political processes where the major parties agree in allowing economic growth to take priority over human welfare, and in accepting that economic growth is best achieved by a competitive market in which individuals seek their own interests. These assumptions spawn programs in which people with human needs are presented with a range of services and demands, but are not listened to. They feel themselves to be part of a system in which they profited from, but from which they do not profit. The result is that, just as people lose trust in polls, they also lose trust in governments and in public services when they are treated as cogs in a machine that does not serve them.



Andrew HamiltonAndrew Hamilton is consulting editor of Eureka Street, and writer at Jesuit Social Services.

Main image: 'I voted' stickers (Unsplash)

Topic tags: Andrew Hamilton, US, election, Australia, polls, politics



submit a comment

Existing comments

Trust:long time to develop. Easy to destroy quickly. Sampling and testing groups of people is a a well established discipline within statistics. Departures from its rigours are often driven by agendas separate from finding out what is going on. And a factor of the discipline is getting to a place with respondents where they feel free to say what they believe rather than what they think the pollster wants to hear. that is why 'unobtrusive measures' are needed. You also need to know how sampling is done. Where do they get their sample numbers? With so many data bases from commerce it is easy to get a skewed group.

Michael D. Breen | 12 November 2020  

I'm sick of all the triumphalist rubbish about the Polls Got it Wrong. Would the commentators please factor in the statistical margins of error around the average results.

Henry Haszler | 12 November 2020  

“Politicians lose trust in voters because they can’t predict their behaviour”. Is this for political manipulation. Hillary Clinton called Republicans “Deplorables”. Trump supporters have been denigrated for 4 years. Why would anyone admit publicly to supporting Trump. Politics has become tribal and if one doesn’t belong to the correct tribe, one is denigrated and worse. There are so many examples of meaningful discussion being shut down, which is power and manipulation. Our society is the poorer for media bias, limited debate and anyone with a different opinion being insulted. It’s anti-intellectual. A cross section of views means a healthier society and a better chance of improvement in every area.

Pansy | 12 November 2020  

I thought of Elvis's song "Suspicious Minds" when reading this. The pollsters may very well categorise complex individuals as fitting a certain dynamic. Which pigeonholes a person somewhat. And no one likes to be thought of as a statistic. Michael Breen's first two sentences sum it up succinctly.

Pam | 12 November 2020  

I believe that Father Andrew's statement, "People may not see the polls as dispassionate enquiries into truth but as instruments that serve those who commission them. Those who are polled are also used as means to further someone else’s interests" is the ring of truth in it. I have answered poll requests for my opinion, however, I found them to have closed, not open questions. So, a person quickly gets to see that slant put on the questions to pigeon hole you, into a slot. Thus I felt that the whole process was a means of mass identification of a narrow number of 'types'; a classification process with no individualism of opinion. If you don't fit the stereotype, you're out/not counted. I totally agree with your conclusions. We become cogs in the wheel, and are treated worse as we get older- neglected and discarded.

john willis | 12 November 2020  

Thanks for the article. Many of us thrive on the comfort of predictability and the ability to plan accordingly; our fragile minds need this to prevent "shock" - even down to the level of a trip to the shop or a municipal tip (who'd ever heard of a city dump closed on Thanksgiving?) Our brains are programmed to enjoy correctly predicting outcomes. Polls are supposed to offer some level of direction but they can also put other attitudes in mind, like inevitability, despondency and anxiety. Does a voter even bother showing up when the forecast is overwhelming for or against their preferred candidate? Aside from the pollsters there are the pundits who use these apparently vague figures in editorial to allege voter reactions to policy or personal performance and even outcomes... not unlike prediction of the stock market for profit, "knowing" the outcome of an election can be a multi-billion dollar industry; somehow Democratically financially influencing that outcome is lawful. If people will donate millions of dollars just to promote their candidate on a whim and notional philosophy how much will they pay to achieve a definite result? Perhaps we shall see...

ray | 12 November 2020  

Andrew Hamilton makes the valid point that there is an inherent trust problem in profit-seeking companies using people’s preferences to provide a media company or political party with intelligence that may not please them or which proves unreliable. Trust is lost by those who purchase the service as well as by the people whose opinions are sought by pollsters. Astute pollsters are transparent about the sample sizes and demographic groups from whom they seek opinions, and are open about margins for error. The phenomenon of ‘shy Trump voter’ was identified by some pollsters leading up to the US election, suggesting that type of voter was reluctant to reveal an intention to vote for Trump while faithfully reciting Trump talking points. It’s a truism to suggest that racecourse punters are not alone in wanting to know what the future holds, and Andrew Hamilton reminds us that money can be made by those who convince people they know the future. The conjurors of conspiracy theories are at an advantage in this field because they speak to audiences that sincerely trust their ‘research’ gives them special insights and knowledge to which the ‘sheep’ who are captives of the deep state are not privy.

Paul Begley | 13 November 2020  

I have little doubt that polling has become part of the 24/7 spin cycle which, sadly, informs much of our politics. This is commercial politics, in which the underlying assumption is that everything can be bought, preferably for as little as possible. It is a sad reflection of what liberal democracy has shrunk to. The continuance of this situation is not inevitable and I think the increasing public cynicism may lead to a change in the way things are done.

Edward Fido | 14 November 2020  

Similar Articles

An American abroad in an election year

  • J O Acholonu
  • 11 November 2020

Being an American abroad during a presidential election means a lot of phone calls and messages back home. It means connecting with other Americans abroad and discussing our plans to vote, our stresses, our anxieties over US politics and ultimately the worthiness of ever going back.


The US presidential election: democracy, threats and transition

  • Binoy Kampmark
  • 11 November 2020

With Joe Biden securing the electoral college votes necessary to win the White House, the concern is whether the transition of power will be one marked by paroxysms of rage and disruption. Donald Trump is promising not to go quietly.