Wednesday, July 22
Thursday, July 23
Improving schools depends on attracting high-caliber teachers as well as increasing retention, both made possible by appealing to teacher preferences. Since teacher preferences cannot be estimated from traditional choice records, I deploy a discrete-choice experiment in a setting where teachers have reason to reveal their preferences. This generates four main findings: (1) I calculate willingness-to-pay for a series of workplace attributes including salary structure, retirement benefits, class size, performance pay, and time-to-tenure; schools can improve efficiency by shifting compensation into vehicles with greater WTP-to-cost ratios. (2) Highly rated teachers have stronger preferences for schools offering performance pay, which can be used to differentially attract and retain them. (3) Using preferences, I simulate how a school would structure compensation to maximize teacher welfare, teacher retention, or student achievement. Under each criterion, the results suggest that schools underpay in salary and performance pay while overpaying in retirement. Finally, (4) the most valued factor is having a principal who supports teachers with disruptive students; this single attribute is worth a 17-percent salary increase and it substantially reduces teacher aversion to low-achieving schools.
Does management matter for how plants weather times of economic crisis? Using firm survey data that was collected in Spain just prior to the Great Recession in 2006, we describe clusters of management practices (“management styles”) using unsupervised machine learning. We establish a positive correlation of a management style associated with rigidity and pronounced hierarchies (``Style 2") with performance prior to the crisis starting in 2006. Even after taking into account firm survival, this correlation turns negative during the financial crisis. Further results suggest that Style 2 correlates with relatively higher holdings of non-liquid assets and lower employee turnover. This is indicative that a management style emphasizing rigidity and classic hierarchies allows firms to exploit economies of scale in times of a boom but hinders them to adjust to rapidly deteriorating economic times.
How has the job of the CEO changed over time? We study this question using a rich corpus of CEO job descriptions providing in depth specifications of skills and capabilities requested by several thousands of firms around the world over the course of 20 years. We classify the free-text descriptions using a novel machine learning approach, which allows us to map sentences found in the documents with well-defined categories of skills and capabilities provides in the O*NET database. We find wide differences in the demand for skills across industries, and a stark increase in the demand for social skills (i.e., skills that facilitate interactions among people, persuasion, and team work more generally) over operational skills (i.e., skills that relate to the execution of set operational and financial goals, measurement of performance, and controls over physical and human resources). The increase in the demand for social skills is larger in companies with greater IT investments, suggesting a possible complementarity between specific leadership characteristics and the adoption of new technologies.
Happy Hour