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,...

Classical business cycles, following Burns and Mitchell (1946), can be defined as the sequential pattern of expansions and contractions in aggregate economic activity. Recently, Harding and Pagan (2002, 2006) have provided an econometric toolkit for the analysis of these cycles, and this has...

A firm's decision to employ agency workers may be perceived as a replace- ment of directly employed workers or as way to curb union power, which trade unions would oppose. Alternatively, trade unions may encourage the (tem- porary) employment of agency workers in a firm, if they manage to...

We examine the relationships between productivity growth, IT investment and organisational change (∆O) using UK firm data. Consistent with the small number of other micro studies we find (a) IT appears to have high returns in a growth accounting sense when ∆O is omitted; when ∆O is included...

How costly will rising temperature due to climate change be for the U.S. economy? Recent research has used the well-identified response of output to weather to estimate this cost. But agents may adapt to the new climate. We propose a methodology to infer adaptation technology from the...