Occupational Class and Values: we nowadays need to study both

TCC have used a wide range of segmentation for both its quantitative and qualitative research. This has included standard geo-demographics such as gender, ethnicity, occupational class and geographical region. In work with local government and health clients we also have utilised both Experian’s MOSAIC and Health ACORN. All of these have provided valuable insights.

As much of our work involved behavioural interventions, whether this was around health, sustainability or communications, we increasingly came to the conclusion that geo-demographics could tell you much about “how” people behaved the way they did, but not “why” they behaved that way.

We needed to explore in a far more in-depth way, how needs and motivations that influenced values then led to the behaviour and subsequent opinions that was subsequently measured as an outcome by the other systems. We thus came to the conclusion that we needed additional forms of segmentation to gather that in-depth insight required and recognised that values based segmentation seemed to be the system that most satisfied our requirements.

There are increasing numbers of practitioners working with academic and commercial systems for mapping motivational values. These divide any population or audience according to its dominant needs and motivations at that point in time, rather than other categories such as age, sex, wealth, class, geography, behaviour or ethnicity.

As such these “psychographic‟ forms of segmentation provide a way to test, develop or execute any communication starting from common motivation, rather than communication efforts which start with audiences defined by other categories and then attempt to add the motivating qualities to an “ask” or offer in order to get a particular result, whether in terms of interest, awareness, alignment, engagement or behavioural action.

In recent years such systems have been used in devising communications, analysing social change, marketing, brand strategy, product development, campaign planning and “social marketing” , by many major companies ranging from BMW and CISCO to Unilever; by football clubs and broadcasters and by NGOs ranging from the National Trust to Greenpeace.

However the Public Sector, focused as it is on providing the sort of services, such as education, that in behavioural terms assist people’s ability, has been slower to take up segmentation designed to assist it in supporting their motivation. This is now changing as the new challenges the public services face change themselves.

TCC have used values based segmentation in areas that have included health social marketing campaigns as well as examining in-depth those segments of the public most vulnerable to anti-community cohesion messages, through to researching usage of the Internet by various values groups, where differences are very wide.

Values based segmentation is also being increasingly utilised for looking at public policy issues and is now even advocated by commentators as a key aspect of attempting various types of social and behavioural change.

An example of this is the area of climate change. IPPR and most recently the World Wildlife Fund with their document “Common Cause” have produced reports that utilise values based segmentation and analysis. In response I have blogged that we need to ensure that the effective promotion of intrinsic values is developed further.

However, so far this increasingly utilised form of segmentation has not featured in public opinion polling.

One reason is that occupational class is still seen as the only way to segment some of the drivers for political polling; perhaps because of a residual recognition of the economic class history of the major political parties, despite them both at various times seeking to reach out and recruit support from beyond their traditional occupational class base.

The occupational class definitions used are themselves based on economic classifications that were established in 1921, four years after the 1917 Russian Revolution and five years before the 1926 General Strike at a time when traditional class consciousness was part of the common currency of almost all political discourse, in a way that a majority of people, living in greater social and economic comfort might find somewhat less resonant now.

Once upon a time all politicians talked confidently of class. It seemed a very real thing, almost as physical difference as the difference in age, sex and ethnicity. Class had an intellectual underpinning in applied economic theories, and was crystallised in the very marked tribal politics of left and right. For example in the 1950’s polling surveys showed between 63-75% of those occupationally defined as working and middle classes voting for parties that represented their particular economic interest. However since the 1970’s academics have talked of partisan de-alignment and this has reflected a shift in occupational class membership and voting patterns too. Peter Kellner of the pollsters YouGov recently remarked:

Although much distinctive class difference has dissolved over recent decades, notions of class and socio-economic groups persist as ways of putting people into categories used to formulate public policy, design services and most of all, conduct political debate. Socio-economic stratification, typically seen as a proxy for class, has been the main underpinning of polling and opinion gathering since the 1960s.

However this looks like it might change. Increasingly old classifications are proving inadequate on their own as society gets more complex and old social demarcations dissolve in the light of a wider and more complicated spread of wealth across the generations, educational attainment, globalisation, migration and rapid technological change.

Having worked in the field of values based segmentation for the last few years, we noted that Peter Kellner in a recent article for YouGov recognised the importance of measuring values in politics by further saying:

However despite these very strong words, it is interesting to note that YouGov have yet to publish values based segmentation results in their own polling. One of the reasons for this may be that, unlike occupational classifications – often set by Government or academic institutions – there is currently no widely agreed values set. However this may change as pollsters look at standard typologies such as the World Values Survey that the British Values Survey is based on.

Having spent years collecting this occupational class data, I am certainly not suggesting it now be jettisoned. Indeed, I think it should continue to be collected to show further evidence, to those people, perhaps at either end of the political spectrum, still hoping we might go back to some of the ideological battles of the past, that occupational class is just not the same as it was in the 1950’s when class segregation was more explicitly pronounced.

In the same way that in some countries, information on caste, religion and rural/urban divide is collected, occupational class is sociologically important to record in this country where class has clearly been a big economic and political divider in the past and where we still need to study how it is changing. At the same time Values based segmentation itself can, of course, be mapped across occupational classifications so one can easily gather both types of data.

For example, whilst a higher proportion of the AB socio-economic group (managerial and professional occupations) are outer directed and inner directed  than for the population as a whole, 20% are sustenance driven. For the socio-economic group C1 (clerical non-managerial occupations) the Values split almost mirrors the population as a whole – 40% inner directed, 30% outer directed, 30% sustenance driven.

The implication of this is that a C1 discussion group recruited to understand motivational triggers for a behaviour change campaign, actually results in purely random attitudinal findings. Put another way round a recruiter might as well have recruited randomly from the population as a whole. The results of either methodology have an almost equal likelihood of being the same, yet recruiting C1s to understand certain attitudes is standard practice for many polling organisations.

Some might argue that understanding class is still vitally crucial as it is driven by economics and that has an impact on social status. Values also respond to economics and socio-economic status, because they are based on understanding people’s actual needs – often driven directly by economic factors – and their motivations around those needs. Whilst 20% of people with sustenance and security values are AB in occupational class terms – often with values seeking to “protect what they have secured” – by far the largest group of them are naturally those in DE socio-economic group.

Where values really add to our insight, is that they go further than just a reliance on traditional class based analysis, in explaining why people behave as they do when they do not seem to act according to their economic status or class interests and are influenced by other issues such as fear of migration.

Our current research sees the importance of people’s social networks and type of social capital as having a significant impact on the development and reinforcement of values, however we are at early days in that research so far and will be reporting more on it here in the future.

There are two things that could be done now to take the values based segmentation agenda forward:

1. That quantitative data is collected that lists ID/OD/SD (inner directed, outer directed, sustenance driven) as a standard measurement in the same way that AB/C1/C2/DE is recorded.

2. There should later be a debate across professional polling bodies  such as the British Polling Council to propose a standard way of collecting agreed values based data, similar to the use of national occupational classifications, so the material is of common use to academics and in public policy debates and interventions.

The debate should not be about whether to continue to collect data on occupational class. This will still be of interest to look at long-term trends that build on the valuable data already collected. The debate is now much more about the recognition that in order to understand people’s needs and motivations in a complex, modern society, it requires an understanding of values as well as class.

Charlie Mansell is Research and Development Officer for The Campaign Company. If you want to see what your own primary values set is, why not take the simple Values Questionnaire here.