“…Moreover, we must recognise that there are limits to what we can learn from historical survey data. Here and in work published elsewhere (Hudson, Lunt et al, 2016) we acknowledge data limitations, including design and sampling, and that responses to survey questions reflect the context of the time in which they were asked, precise wording of the questions presented and the answer options offered (Smith, 1987, 76). In presenting a fourfold argument for there being greater continuities in public attitudes than is typically recognised we concede that our evidence is strongest when pointing to earlier examples of anti-welfare state sentiment and qualified support for social security and social services.…”