Abstract
This paper discusses how firms use “big data” and the role and challenges for economists when getting involved in big data research. The firms’ success stories have taken advantage of building the biggest databases, using the best extraction tools, and using the fastest algorithms for data analysis and management. Although there are already great examples of economists entering big data research, analysts involved at present are mostly statisticians and computer scientists. The challenges economists face lie in computation, publication, and replication when using proprietary big data. The opportunities for economists lie in modeling and designing frameworks for analyzing large observational data panels, as well as developing empirical designs and strategies that are motivated by a model framework; in this way, economists can guide large-scale big data experiments toward identifying causal effects, rather than just correlations.
Highlights
Big data is becoming a common term in industry, media and academia
The opportunities for economists lie in modeling and designing frameworks for analyzing large observational data panels, as well as developing empirical designs and strategies that are motivated by a model framework; in this way, economists can guide large-scale big data experiments toward identifying causal effects, rather than just correlations
According to an executive survey of twelve hundred companies in Asia, Latin America, North America, and Europe, 47% of the companies had no big data initiatives and 40% stated they still used a hunch or their “gut” feeling to make decisions. Among those who do engage in big data initiatives, the average reported return on investment (ROI) on big data for 2012 was 46%
Summary
Big data is becoming a common term in industry, media and academia. Since the 1990s, the so called three V’s Volume, Velocity, and Variety - have defined big data, as more and more data are created, originating from a large variety of sources, such as virtual data, that have been exploding. This paper turns to examples of successful examples where academic research has already used big data originating from firms It discusses the resulting research possibilities, and how data-sharing agreements have evolved into creating unique data and empirical settings to answer research questions of interest to economics. It highlights the limitations of using datasets originating in firms, and discusses the open challenges researchers deal with when using proprietary big data; these challenges include computation, publication, and replication. If the data are unstructured (for example, originating from video, or consumer feedback, or comments), other challenges for big unstructured data management emerge before those data can be used to gain insights useful for strategy
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
