Why are there no census measures of mental health?

At the moment a random sample is considered the best way to be
representative. That's not quite true. It's the cheapest way. Stratified
samples, matched studies and the like are all much better but more
expensive ways to get closer to a truly representative sample and
experiment.

A census is the most representative 'sample'. This is done once a decade
in the UK. There are no measures of mental health, well being or
happiness included as far as I am aware.

The data would be the start of making a very useful dataset. The
Bhutanese do a 5 day census which is considerably closer to a full
research project to determine the government's success on the measure of
Gross Domestic Happiness. They do that every year as well. Why? They
measure what matters I guess.