Am measure measures what is measures. Generally, measures like unemployment are useful in determining trends. Note how most of the unemployment measures are cor

Am measure measures what is measures. Generally, measures like unemployment are useful in determining trends. Note how most of the unemployment measures are correlated. OTOH, it may be wrong (eg) over very long periods. If underlying behavioural changes happen (eg employment seeking behaviour, drifting of underemployment/unemployment ratios) it may miss long term trends. The problems start (almost immediately) when when the measure becomes the definition. Measures are tightly defined, objective and legible. That's attractive. So, we start to define (eg) unemployment as "that which IQ tests measure." This is totally off track, if the measures weaknesses are (they usually are) justified by arguing the measure is just for trend detection. Definition and trend detection are different purposes. In any case, the only way to get a less superficial understanding to define the measure in terms of what it actually measures, not what it's trying to measure. "people trying to find work" is an actually useful thing to pay attention to. It just isn't "unemployment" in the widest sense of the term. Personally, I hate the "index/metric" solution. This is chasing the rabbit down a hole. We can't have a measure that perfectly maps to unemployment (or financial wellbeing, or whatever). Efforts to do so end up measuring something even more abstract and difficult to understand. The alternative is to just use these measures as incomplete. We can look at "job seeker" rates, nonemployment rates, etc. No measure will tell us what's really happening. That's up to us human intellect to determine.