The BBC’s Great British Class Survey (GBCS) generated an unprecedented amount of media interest when the first study from the survey was published in April 2013. Since that time, many sociologists, journalists, and members of the public have responded to both the press coverage and to the actual research. Last week, a special issue of The Sociological Review was released, showcasing further research conducted using the data, as well as a number of engagements with the GBCS team’s work. Although I was not an author on the original paper, I have found the GBCS to be an exciting data source and I believe in the work I and others have conducted using it. Given the controversy that still surrounds the project, I want to take this opportunity to clear up a few myths or misunderstandings of the GBCS data and project. I hope this might prove helpful to those conducting or engaging with GBCS-based research.
Myth: The ‘GBCS’ is the ‘Class Calculator’ posted on the BBC’s website in April 2013 that gave respondents a ‘class’ based on their answers to just a few questions.
Fact: The ‘Class Calculator’ is essentially an internet quiz, based on but by no means identical to the model reported in Savage et al 2013, or the data collected by the BBC. GBCS stands for ‘Great British Class Survey’ and refers to the survey conducted by the BBC, hosted on their Lab UK website from January 2011 through July 2013. The actual survey had far more questions, with many more possible answers, than the ‘Class Calculator’. (‘GBCS’ also, for maximum clarity, should ideally not be used to refer to the first article based on the data; it is a dataset from a survey). GBCS is also sometimes used to refer to the supplemental survey conducted by the firm GfK, which asked the same questions, but of a sample of 1026 meant to be representative of the UK population. Both these datasets are now available through the UK Data Archive, and were introduced in a blog post I co-wrote with Mike Savage.
Myth: The GBC S team proposed their seven new classes to take the place of all heretofore existing measures in class analysis, especially the NS-SEC.
Fact: Some press coverage probably implied that, but the 2013 article itself as I read it makes no such claim. Instead, the article reported the results of a latent class analysis based on measures of respondents’ economic, social, and cultural capitals; respondents were grouped into seven ‘classes’ based on their relative scores on each of these measures. There are many reasonable critiques to be made of the particular measures and methods used in the 2013 analysis, of course, but I do believe these sorts of resources are absolutely relevant to people’s class position. However, neither I, nor, to my knowledge, any other scholar involved with the GBCS has ever said that these seven classes ought to replace all other possible measures of class in the UK, or that occupation-based measures such as NS-SEC are irrelevant or ought never to be used. Many of us have happily used NS-SEC in our research and will likely continue to do so.
Myth: Because the GBCS is not a random-sample or representative survey, it cannot be used to say anything worthwhile about class (or anything else) in the UK.
Fact: The GBCS is most certainly not a representative, let alone random-sample, survey. That means that standard tools of statistical inference are not applicable to analyses based on it. This does not by any means make it unusable as a data source. I believe that this non-representative data can be used for worthwhile research in (at least) three ways:
- If non-representative, non-random-sample small scale qualitative studies can provide useful insights for sociologists (and I certainly believe they can), so can the 300,000-respondent GBCS. We can think carefully about what is and is not likely to be generalizable based on our knowledge of how the respondents compare to the population, just as we do when we conduct qualitative research. This is further aided by the availability of the GfK survey, which asked the same questions to a more representative, non-self-selected sample of respondents.
- The ‘skew’ of the respondents relative to the UK population reveals something interesting about who did (and did not) find out about, and think it worth participating in, a study on class in the UK.
- Multiple regression and related methods can essentially help ‘control’ for the skews. While the mean value of any variable in the GBCS is almost certainly different from that for the population as a whole, the relationships between variables will often be similar. For example, we argued in our ‘Class Ceiling’ paper based on the GBCS that there was no reason to expect that the relationship between GBCS respondents’ class origins and their pay in elite occupations was different from the relationship for others in elite occupations; and indeed, we found a substantively identical relationship in the gold-standard nationally representative Labour Force Survey.
I continue to work with GBCS data alongside projects using more traditional surveys (and other methods), and I believe the data have a lot to offer those who are interested in class and inequality in the UK. There are many interesting and worthwhile conversations to be had about issues arising from the GBCS data collection, the initial and subsequent articles, and the relevance of Bourdieusian approaches to class; the special issue engages many of these directly, and I look forward to the ongoing discussion.
Daniel Laurison is a Post-Doctoral Fellow in the Sociology Department at the LSE who studies class, inequality and politics in the US and the UK.