Among the many interesting presentations that I attended at the recent Annual Conference of the N.Z. Association of Economists was one by Frances Krsinich, from Statistics New Zealand.
The paper that Frances gave was titled "Price Indexes From Online Data Using the Fixed Effects Window Splice (FEWS) Method". Here's the abstract:
"Automated web-scraping of online data gives potential for timely and high-frequency price indices. But online data has minimal information on the characteristics of products with which to apply quality-adjustment techniques such as hedonic regression.
Statistics New Zealand has been investigating the use of a fixed-effects approach which implicitly controls the price-indices for al price-determining characteristics by fitting product-specific intercepts. This 'fixed-effects window splice' (FEWS) index is equivalent to a fully-interacted time dummy hedonic index.
We will discuss the results of applying this approach to 15 months of daily online consumer electronics price data, shared with us MIT's Billion Prices Project."
I was really impressed with the work that Frances described, and it's clear that Statistics New Zealand is way ahead of the curve when it comes to the innovative use of online data in the compilation of official statistics. Apparently, they have some serious work underway in collaboration with the folk at the Billion Prices Project, so there should be some interesting and innovative results in the pipeline.
Meanwhile, you can learn more about the work by Frances, and her co-author Jan de Haan, in their paper that is "forthcoming" in the Journal of Business and Economic Statistics.