2014 Hansen Lecture
Danny Pfeffermann, University of Southampton, UK; Hebrew University of Jerusalem,
Israel; and Central Bureau of Statistics, Israel
“Methodological Issues and Challenges in the Production of Official Statistics”
Discussants: Lawrence Brown, Wharton School, University of Pennsylvania
John Eltinge, Bureau of Labor Statistics
Abstract: The big advancement in technology, coupled with the constant rise of 'big data', and yet increased demand for more accurate, more detailed and more timely official data with tightened budgets, places enormous challenges to producers of official statistics across the world. In this presentation I shall discuss the major challenges as I see them and occasionally offer ways of dealing with them. Examples of such challenges include the potential use of big data with integration of computer science; increasing data accessibility but maintaining privacy and confidentiality; possible uses of data obtained from web-panels; addressing the problems of response burden and nonresponse; mode effects and proxy surveys; integration of statistical and geospatial information; and future censuses based on administrative data and small area estimation. Based on my extensive experience in the academia, researching and consulting on topics related to the work of National Statistical Offices (NSOs), and serving as the National Statistician of Israel, I shall also confront the question of how well universities train students to work at NSOs. Considering that NSOs are among the largest employers of economists and statisticians, and official statistics is the most frequently encountered statistics for most people, I strongly believe that the academia should include topics applied in the work of statisticians and economists working at NSOs in their teaching and research programs. As advocated in my presentation, these topics are just other facets of statistical inference.