04 March 2013

Ipsos MORI has measured the public’s trust in various professions since 1983. It is a very simple question that asks whether people trust different groups to tell the truth or not – but it generates huge interest.

And this interest is despite the rank order generally not changing much. We can guarantee that doctors will be at the top: we have measured trust in doctors in the middle of any number of high profile medical scandals, and there is no impact. We can also guarantee that journalists and politicians will be bumping along the bottom together. And that’s the case in our latest survey, where politicians have cemented their position as the least trusted profession.

But there are some big shifts in trust over time – one is the remarkable rise in trust in civil servants. This has gone from 25% in 1983, to 37% in 1993, to 53% now.

Why has this happened? To be honest, I don’t have a completely convincing explanation, and I haven’t been able to find anyone who does. In fact there are a whole range of excellent reasons outlined in studies of trust in government why we might expect trust in UK civil servants to have gone down: improvements in services being banked and discounted, global insecurity making the national civil service seem less effective, constant reform undermining their image for competence etc

But there are three points that do seem plausible.

First, there is almost certainly a contrast effect with politicians. As Lord Butler suggests, high trust in civil servants could be the obverse of low trust in political groups. There is certainly truth in that, but at face value it doesn’t really explain the change: our trust trends show that not a lot has changed in trust in politicians, so why should the contrast effect be greater now?

This measure of telling the truth is only one aspect of trust in politics, and there are others that do suggest there has been a decline in our trust in politicians. For example, between 1983 and 2011 there was a three-fold increase in the proportion of people who say they almost never trust politicians to put the country’s interest before those of their party. We talk too readily about a crisis of trust in politics, but there is something deeply worrying in that trend.

The second explanation is that trust in civil servants increased as they became associated with the optimism around public services in the late 1990s and early 2000s. This period saw the greatest increase in trust in civil servants, and it was certainly the case that the political rhetoric on the importance of public service and investment in them was very positive then Why then isn’t trust going down again now with the focus on austerity? I think this could be explained by a lag in public opinion: I’d say that the next move will be down, as the message that the civil service is a cost to be controlled is increasingly noticed.

The third point relates to the broader cultural context, both reflected in and influenced by the political sitcoms of the different eras. “Yes, Minister” was based on a very clear view of manipulative mandarins, locked in a battle with politicians, where each was equally slippery. In contrast, “The Thick of It” shows officials as fairly neutral, if sometimes inept, servants, with advisors and communications experts pulling the strings.

Of course, there is an important question of what people have in mind when thinking about “civil servants”. We know that understanding of who does what in government is very shaky, and it’s likely that civil servants are a particularly nebulous concept for many people. As data accompanying the Civil Service Reform Plan points out, there are approximately 500,000 civil servants, most of them in low paid, delivery functions. But, while I have no more than anecdotal evidence, I’m pretty sure that’s not what people think of when answering our question, and the most common image is still a vague idea of a central government department official.

More broadly, the language used in these types of questions is important. In previous years we’ve asked about “managers” in local government and the NHS, and each has much lower trust levels than currently seen for civil servants. The term itself may now have advantages that it’s not had in the past. Over 50 years ago, Lord Bridges grouped civil servants with mother-in-laws as similar objects of ridicule – both points seem equally out of date now.

And of course, civil servants are too wise to be too smug about this one trend. A slightly different question on confidence in the civil service from the World Values Survey shows no similar increase in the UK - and we’re only mid-table compared with other developed countries.

But despite all these qualifications and problems of interpretation, in the end the data still suggests that over the last three decades people have increasingly recognised civil servants as worthy of their trust. It doesn’t tell us how much people value that or think it’s important to protect, but it should still give pause to changes that could undermine it.

It is probably true that any decisions affecting the role of civil servants, such as politicians appointing permanent secretaries, are unlikely to affect public opinion – but when trust is at such a premium, we should be very careful taking steps that may eventually put it at risk.


How the question is asked is indeed important: most of the working class people I know use the term "civil servant" to refer to local government as well as central government officials, which would lead to very different conclusions from Mr Duffy's anecdotal evidence.

Another thing to consider is that the private sector has taken a knock in recent years, as the Edelman Trust Barometer reports show (banking collapse, greedy chief executives, collapsing care home companies, Group 4 at the Olympics, media hacking dead children's mobile phones, horsemeat in the ready-meals): this may be a more powerful contrast effect than that of contrast with politicians.