The matching method for treatment evaluation does not balance selective unobserved differences between treated and non-treated. We derive a simple correction term if there is an instrument that shifts the treatment probability to zero in specific cases. Policies with eligibility restrictions,...

This paper reviews current issues in youth labour markets in developed countries. It argues that young people aged 16-25 have been particularly hard hit during the current recession. Using the USA and UK as cast studies, it analyses both causes and effects of youth unemployment using micro-data....