Big Data describes the information that is frequently more personalised, considerably greater in scale, and frequently collected. Examples include information gathered from smart sensors installed in homes or a collection of Twitter tweets.
Traditional econometric methods frequently perform better in limited data sets than more sophisticated ones. Machine learning techniques, however, excel with massive data sets. To fully utilise Big Data in economics, new analytical techniques are required. Therefore, if researchers and decision-makers wish to utilise new Big Data sources fully, they need to pay attention to recent breakthroughs in machine learning approaches.
In the past, data was exclusively gathered for specified purposes, frequently by a national statistical agency. However, as the world becomes more quantified, even the smallest business now gathers and stores extensive, occasionally personalised data.
A sizable ecosystem of hardware (sensors) and software (apps) is embedded in the enormous sea of “smart” technology, such as smartphones, Wi-Fi-connected appliances, automobiles, and satellites, to collect data. The data flow has grown due to the data avalanche in terms of diversity and speed.
There are numerous new possibilities for transforming previously unstructured data, such as text and satellite photos, into innovative data sets. This change has expanded the scope of the economic question.
How does this data help in economics?
These new findings have an impact on an economic study in several ways. In many disciplines, administrative data with universal or nearly universal population coverage has replaced the reliance on government surveys with relatively small sample sizes.
This change is revolutionary because it enables researchers to establish reliable long-run statistical indices, create novel quasi-experimental study designs, and rigorously evaluate variation in earnings, health, productivity, education, and other indicators across various subpopulations.
The increase in economic activity statistics from the private sector may be even more noteworthy. These data, sometimes gathered through data-sharing agreements with private companies and other times from public sources, can aid in creating more accurate and timely measurements of overall economic statistics.
The Role of Economic Theory
To analyse big data sets with complicated structures, economic theory is crucial. Without a conceptual framework that makes sense, organising and analysing this kind of data can be challenging (or even choosing which variables to create). This is where economic models come in handy. Additionally, more precise assessments of current models and hypotheses that were previously challenging to evaluate are now possible because of better data.
When Scottish philosopher Adam Smith published An Inquiry into the Nature and Causes of the Wealth of Nations in 1776, economics as a field of study truly came into being.
How important is Economics?
Over the past few decades, economic science has changed to place more emphasis on empirical research. The data revolution of the last ten years will likely continue to impact economic research significantly. There are new potentials and problems resulting from economists using newly available large-scale administrative data or private sector data, which are frequently acquired through partnerships with private enterprises.
According to a recent article by CNN reporter Lydia DePillis, over 150 PhD, economists have been employed by Amazon in recent years. After the Federal Reserve, the online retailer employs a lot of economists in the US. The crucial distinction is that these economists are actively involved rather than serving as distant advisors to high management.
How is India adopting this?
Public policy in India is also participating. The Reserve Bank of India is establishing a lab for data sciences. Using information from the railway’s computerised ticketing system and big data analytics, economists at the finance ministry have already mapped our patterns of internal mobility and interstate trade (using preliminary data from the Goods and Services Tax Network).
The IDFC Institute’s work using satellite pictures to illustrate how the density of the built-up area in the Kozhikode Metropolitan Area changed between 1975 and 2014 was referenced in the Economic Survey published in August 2017.
Professor Roberto Rigobon of the MIT Sloan School of Management distinguished between prepared data and organic data in a speech he delivered at RBI in August 2018. The former originates from administrative and survey sources. The latter is “created by individuals who are not aware that they are being polled. It consists of information from your phone’s GPS, web searches, a network of friends, and purchases of goods.
Although it cannot replace planned data, organic data will start to compete with it. According to Rigobon, the primary benefit of organic data is that it is accurate. Instead of recall, it is based on real conduct. Additionally, it is arranged according to behaviour rather than the more conventional methods of location or socioeconomic status. The drawbacks are that organic data might include privacy concerns and is frequently not representative.
Rewind to October 2015. The Indian central bank was fighting high inflation. “Moreover, there has been a comfort on the inflation front—wholesale prices are contracting, GDP (gross domestic product) consumption deflator has been low at around 3%, Retail inflation may be lower than the headline data indicates, given e-commerce merchants are offering low pricing, the committee of economists advising the RBI governor on monetary policy said in a statement issued that month.
Public policy in India could transform as a result of the rising popularity of digital transactions among investors, taxpayers, and consumers. It is unlikely that the more conventional figures obtained from surveys, national accounts, and administrative data will entirely be replaced by these more recent data types.
It is more likely that they will work well together. Big data analytics will become increasingly important to government organisations in the coming years, but the hazards to people’s privacy should not be understated.