The problem is differentiating between those who've given up and who do not want to work (have other means to sustain themselves).

In general, either is fine by me as long we are consistent: they are both proxies for percentage of people needing work and should correlate to a large extent.