Generating Big Data in (Micro)Economics

Introduction

The term Big Data refers to unconventionally large data streams from many sources, to a bunch of (not necessarily known) recipients. It might be better for the term to be defined as a co-existence of data streams rather than a data set because of the trouble with its storage and handling as a whole due to its dynamic nature and enormous size. A set of data is ontologically a comprehendible static data frame which promises a possibility of mastering. However, Big Data is closer to be defined as entropy of social universe, unrelated to social entropy concept in sociological theory, and is likely to be remained as abstract and unamenable for a considerable future. It is not viable to categorize the Big Data for limited uses. For that reason, a pinch of Big Data can arguably be appropriate for economic, sociologic or any other kind of disciplinary research demanding high level expertise even for humble quests. It serves as a unified scientific outlet which reaches beyond interdisciplinary efforts. The fundamental aim of this chapter is then to provide a mere perspective for early use of the aforementioned general concept of Big Data in economics until a unified science arrive to tackle it.

Economics has been a theoretical profession with never enough ties to the empirics for a long time. The line between speculation and prediction depends on the perceived soundness of the empirical model. Data oriented economic research papers are cursed with such reliability issues because their claims are of general interest. While there are only a handful of people cares about the actual specifications of the highly-popular CERN experiments, millions get involved when a paper on the inflation dynamics or consumer choice gets distributed. This popularity levies a burden on economists to match the evidence with more lucid models relying on improved data availability. However, one can say that most economists increasingly respond to this interest with enigmatic methods and vague claims to maintain methodological solidity.

Contemporary economists have largely recognized the importance of their works' empirical grounds. When it comes to empirical studies with big data, the gap between main branches of the field such as macroeconomics and microeconomics became vague in favor of the micro. Economic research has been traditionally labeled by the scale/aggregation level of the activity. As the precision and the availability of the data increases, all branches in the literature are methodologically forced to converge towards the lowest level -or higher frequency- of interactions. Among others, microeconomics has been a more data-oriented practice in the field of economics which usually considers agent behavior. Relying on aggregate relations by disregarding underlying dynamics is a suboptimal modeling when the micro data is available. Therefore, introducing Big Data shall favor microeconomics above others as a methodological basis to a more unified economic research. Similarly, Currie et al. (2020) reports that applied microeconomics papers' weight have increased from 55% to 75% in high quality general interest journals recently.

Time stamps in economic modeling can traditionally be classified as ex-ante and ex-post, but the economy is a real-time phenomenon. Unlike contemporary information sciences related to computerized environments, gathering and handling high quality and large data sets which have use for economists is a hard task to manage. For this reason, the literature of economics has been evolving around macro data usage with limited representations, which goes hand in hand with hard and limiting assumptions to provide tractable modeling for colleagues.

Data sources and toolkit of economists have significantly increased during the last decades. These came along with interdisciplinary researchers e.g. physicists or biologists, who manage to apply data analysis techniques when there are plausible similarity of the tasks at hand. While emergence of the complexity economics is offering new perspectives, there have been a wide range of branches occurred such as econophysics and/or econobiology in order to benefit from the progress of other fields.