Syndicate content

Friday links: employment miracles, breakfast, cookstoves, new financial data, and more...

David McKenzie's picture

·         My colleague Leora Klapper and manager Asli Demirguc-Kunt have just released new global data on financial access around the World called the Global Findex, funded by the Gates Foundation. See another post here. They added questions to the Gallup World poll, so have the most comprehensive data yet, which is accessible through this website – I encourage you to explore. They have questions on debit and credit card ownership, loans, savings, insurance, and much more – although my suggestion for the next round is to separate out internal and international remittances which are currently lumped together in the questions.

·         Interesting post on what predicts employment miracles on the World Bank’s MENA blog by Caroline Freund and Bob Rijkers.

·         An interesting description of the impact of blogging and tweeting about her own papers by Melissa Terras at the LSE Impact of Social Sciences blog – results are consistent with Berk and my paper on this.

·         Cookstoves and knock-out punches – Aid Thoughts discusses whether we are too ready to embrace a single study that knocks down a hyped-up development intervention.

·         Nice slides on measurement issues in Cyrus Samii’s latest lecture.

·         The NYTimes Economix blog discusses a new paper claiming to measure the impact of school quality on housing prices – how many other omitted variables can you think of that also are likely to vary with neighborhood apart from schools?

·         The impact of breakfast.

Comments

Submitted by Anonymous on
I disagree with your implied general problem of omitted variables by neighborhood in estimations of the house price capitalization of school quality. In fact there is a history of rigorous empirical work on this topic. I believe that Sandra Black's paper was the first to use an RD with distance to the school district boundary. It is one of the canonical papers for RD in labor econ, was published in QJE and currently shows some 750 citations on Google Scholar. Yes, they did think about unobserved neighborhood characteristics when estimating these models. It is not evident from your comment whether this particular paper uses crappy methods that have been replaced with better tools a long time ago. Your implied criticism is certainly not generally applicable to this literature and your comment does not make that critical distinction.

Black, SE (1999) Do Better Schools Matter? Parental Valuation of Elementary Education. QJE 114, No. 2, Pages 577-599. http://www.mitpressjournals.org/doi/abs/10.1162/003355399556070 The evaluation of numerous school reforms requires an understanding of the value of better schools. Given the difficulty of calculating the relationship between school quality and student outcomes, I turn to another method and use house prices to infer the value parents place on school quality. I look within school districts at houses located on attendance district boundaries; houses then differ only by the elementary school the child attends. I thereby effectively remove the variation in neighborhoods, taxes, and school spending. I find that parents are willing to pay 2.5 percent more for a 5percent increase in test scores. This finding is robust to a number of sensitivity checks.