At the onset of any crisis, economic policymakers would like to define whether we confront a demand or a supply shock. They would also like to determine the magnitudes and likely persistence of these shocks. Moreover, as a crisis unfolds, policymakers also want to understand how the shocks propagate to the broader economy. In this effort, nontraditional data might provide them with hitherto undiscoverable insights.
The COVID-19 crisis provided a test case for the usefulness of alternative data sources. Nontraditional data – daily point-of-sale card swipe data, consumer sentiment surveys, credit card data, and weekly automotive transactions provided policymakers with a view into economic activity weeks or months before traditional data would become available. They also illuminated household and business activity at a granular level, helping to clarify the mechanisms affecting the pandemic economy. Furthermore, having access to nontraditional data specific to this pandemic-offered policymakers insight into how the virus and associated health policies were evolving.
However, there are pitfalls in using nontraditional data. The absence of a long time series in many of these series hinders seasonal adjustment, makes levels challenging to interpret, and impedes comparisons at a business cycle frequency. These data can also be unreliable because they are nonrepresentative, methodologically inconsistent, highly variable or noisy, or susceptible to discontinuation. Furthermore, the resources to develop the human capital to address these issues are significant—and that is over and above the cost of the data itself.
Even if these nontraditional data sources have limited use during an upswing, it is worth developing them to be prepared for the next crisis. Hence, policymakers should invest in nontraditional data sources to build longer time series of timely indexes to supplement traditional data sources, improve the usability of existing data, validate the granular details that may be available and become important during a downturn, and hone their skills in working with these data.
Granted that it is hard to know what types of idiosyncratic series will be valuable in the future, however, a culture that embraces transparency and data sharing can only help.
Click here to read the original article published by BROOKINGS.
Please give your feedback on this article or share a similar story for publishing by clicking here.
