The Pros and Cons of Using Big Data in Economic Signals

Big data is being used for a wide range of applications, including economic development evaluations

In most sections of the nation, a fresh wave of covid infections caused by the Omicron strain of the Sars-CoV-2 virus has limited travel once again. According to anonymised data given by Google for India-based mobile phone locations, people spend more time at home than at work, retail outlets, or parks. During the pandemic, such new types of data were extremely valuable in tracking economic activities regularly. Rather than waiting for official statisticians to compile the typical quarterly assessments of activity from across the economy using structured surveys and administrative data, they provide us with an immediate impression of what is going on in the economy. These standard operating procedures might often be too sluggish to grasp a quickly changing circumstance. Various sorts of big data have stepped into the void during the last two years.

Both things have in common that we have evolved as a civilization to the epidemic. Over time, the link between mobility metrics and economic development has changed. The regression results have altered as economic actors have essentially learned to live with the virus, in more technical terms. According to OECD statistics, a ten-percentage-point increase in mobility was linked to a 2.2-percentage-point increment in the third quarter of 2020, but only a 0.9-percentage-point increase in the fourth quarter. That's a significant decline.

It also implies that an analyst using the coefficients and the first and second quarters of 2020 to anticipate the influence of changes in focus on advancing on quarterly economic activity would receive substantially different conclusions than another analyst using the coefficient for the fourth quarter. This fact is significant when attempting to quantify the influence of the fourth wave just on the Indian market, particularly when movement data is a key factor to examine.

Some Other Sorts of Large Data have Their Own Set of Problems

Consider artificial lighting, which is increasingly being utilised as a proxy for business growth by some economists. The data on lights left after dusk comes from a variety of satellites that can detect the brightness of lights produced by people in a certain region. Clouds influence the way data on night lights can be gathered by satellites that hang kilometers over the ground, according to Ayush Patnaik, Ajay Shah, Anshul Tayal, and Susan Thomas of research firm xKDR in a recent working paper. The four researchers have demonstrated that measurements have a downward bias during foggy months and have developed an algorithm to partially rectify this downward bias.

Another issue with big data interpretation is the context where it is read. Because of shutdowns or fear of stepping out, consumer demand has shifted from services to commodities in numerous categories throughout the pandemic months. When commodities travel throughout the country, the e-way invoices created are a highly helpful indicator of economic activity. However, such bills are not required for services. As a result, a broad change in demand from service to commodities will almost certainly result in a greater increase in e-way costs than could be explained by overall economic activity. Similarly, a shift in demand back towards services might indicate that the increase of e-way bills has slowed. This does not imply that the industry has slowed down.

Data analysis is being utilised for a variety of purposes, including economic development analyses. However, there are some advantages and disadvantages to this kind of analysis. Google recently disclosed some anonymised data for India, which was based on the customers' mobile phone locations. Spending more and more time at homes than at work, retail outlets, or parks, according to the research. It implies that people are once again experiencing mobility issues. The new data types are extremely beneficial for following a country's economic activity without having to wait for formal official surveys and monthly estimates. These standardised polls are incapable of comprehending a constantly changing reality. Mobility data, for example, may be used to forecast changes in economic activity.

However, massive data analysis comes with its own set of difficulties. Economists, for example, use the assessment of night lights after sunset as a proxy for business growth. Clouds, on the other hand, appear to obstruct satellites' ability to acquire data on night lighting, according to several study articles. As a result, measurements are low during overcast months. Second, data from e-way bills created during the transportation of products is a highly important indicator of economic activity ahead of time. For services, such e-way bills are not created. As a result of the shift in demand from commodities to services, the number of e bills must be minimal. As a result, it does not indicate a downturn in economic activity.