This study has been prepared within the UNU-WIDER project Inclusive growth in Mozambique-scaling-up research and capacity implemented in collaboration between UNU-WIDER, University of Copenhagen, University Eduardo Mondlane, and the Mozambican Ministry of Economics and Finance. The project is financed through specific programme contributions by the governments of Denmark, Finland, and Norway.
We begin by developing a characterization of teams' strategies (extent of offense or defense) using principal component analysis. This is used to estimate the relationship between a team's probabilities of scoring and conceding goals and its chosen strategy. Knowing that relationship, it is then possible to derive a team's optimal strategy, and to study how this varies in different situations (such as playing at home or away). A comparison between optimal and actual strategy reveals that teams appear to adopt more defensive strategies than is optimal. A notable feature of our study is that we model a team as choosing a strategy at the start of each match and also at half time, thereby incorporating a dynamic element.
How jobseekers set their earnings expectations is central to job search models. To study this process, we track the evolution of own-earnings forecasts over 18 months for a representative panel of university-leavers in Mozambique and estimate the impact of a wage information intervention. We sent participants differentiated messages about the average earnings of their peers, obtained from prior survey rounds. Demonstrating the stickiness of (initially optimistic) beliefs, we find an elasticity of own-wage expectations to this news of around 7 per cent in the short term and 16 per cent over the long term, which compares to a 22 per cent elasticity in response to unanticipated actual wage offers. We further find evidence of heterogeneous updating heuristics, where factors such as the initial level of optimism, cognitive skills, perceived reliability of the information, and valence of the news shape how wage expectations are updated. We recommend institutionalizing public information about earnings.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.