Cornish ethnicity and undercounting: utilising the 2001 England and Wales Census to develop an accurate measurement methodology.

In this paper we outline results from exploratory research which aimed to provide better estimates of self-identified Cornish ethnic group affiliation in Cornwall, a county in the South West of the UK. For the first time the 2001 England and Wales Census recorded a 'write-in' Cornish identity category – however the lack of a dedicated tick-box and limited publicity around the option meant that the result was likely to undercount the true population. We utilise data from a range of sources about Cornish ethnicity to derive weights for the Census result based around statistical means. We subsequently apply these weights to posit a more likely estimate of 25-30% rather than the 6.7% reported. The limitations inherent in such an exploratory model of analysis limit our conclusions, but we feel that our result engages with the methodological literature around small-scale survey measures as well as the ethnicity measurement literature more broadly. We conclude that a group of individuals who share some commonality of response, regardless of the methodological or philosophical problems surrounding the measure, are an interesting topic of study for social science and can inform subsequent analyses of relative position.