Social tolerance (proxy)

What this is — and what it is not. This is an *indicative proxy* for social tolerance, not a "racism ranking" and not a definitive measure of discrimination. It is built from a single, named World Values Survey item: the share of respondents who mention "people of a different religion" when asked which groups they would not want as neighbours. WorldScore scores the inverse, so 100 = most accepting (fewest object) and 0 = least accepting.

Why the honest reframe. "Racism" has no reliable, comparable global dataset. Rather than invent one, we surface a specific, sourced attitudinal proxy and label its limits loudly.

Serious limitations (Tier C). - One item (religion-of-neighbour), from surveys run in different years per country (1998–2022) with varying samples — not directly comparable across countries. - Coverage is partial (about 100 countries); everywhere else is "no data". - Attitudes ≠ behaviour or lived discrimination; social-desirability bias affects self-reports.

Treat it as a conversation-starter with explicit gaps, never as an authoritative ranking of tolerance or prejudice.