1. Introduction

Data have always been fundamental to economic research. Conventional economic analysis relies extensively on national accounts, household surveys, labour-force statistics, business surveys, trade statistics, price indices, and administrative records. These sources have provided the empirical foundation for understanding economic growth, inflation, employment, productivity, consumption, investment, poverty, and international trade. However, the digital transformation of economic activity has created an unprecedented expansion in the volume and variety of information available to researchers and policymakers. Digital payments, e-commerce transactions, social media activity, mobile-phone records, online job advertisements, satellite imagery, financial-market data, web searches, and administrative databases can now generate information at frequencies and levels of detail that were previously difficult or impossible to obtain [1]. This transformation has contributed to the emergence of big data analytics as an important component of modern economic research. Big data generally refers not simply to large datasets but to data characterized by dimensions such as volume, velocity, variety, and complexity. Its significance for economics arises from the possibility of observing economic behaviour more rapidly and at greater granularity than conventional statistical systems allow. Real-time transaction data, for example, can provide indications of changes in consumption before official statistics become available, while online job vacancies can provide timely information about labour-market demand [2]. The growing availability of alternative data has also changed the methodological possibilities of economic research. Machine-learning algorithms can identify complex patterns in large datasets, natural language processing can extract economic information from textual sources, and predictive models can process large numbers of variables that may be difficult to incorporate into conventional econometric models [3]. These tools can complement traditional statistical methods and provide new approaches to forecasting and classification.

The relevance of big data became particularly evident during periods of economic disruption. Conventional statistics may be released with substantial delays, whereas high-frequency digital indicators can provide more immediate information about changes in mobility, spending, employment, production, and business activity. During the COVID-19 pandemic, researchers used mobility data, electronic transactions, online searches, and other alternative indicators to monitor rapidly changing economic conditions [4]. Such applications demonstrated the potential value of real-time data for policy institutions facing rapidly evolving economic circumstances. Nevertheless, the increasing use of big data also raises important methodological and ethical questions. Large datasets are not necessarily representative of entire populations, and digital data may systematically exclude individuals with limited internet access or digital participation. Algorithmic models may reproduce existing biases, while correlations identified in large datasets should not automatically be interpreted as causal relationships [5]. Privacy, data ownership, cybersecurity, and responsible data governance are equally important considerations. This review examines the major applications of big data analytics in contemporary economic research and policy development. It focuses on how new data sources and analytical technologies are changing economic measurement, forecasting, policy evaluation, and decision-making while considering the limitations that must be addressed for responsible and scientifically rigorous application.

2. Big Data and the Transformation of Economic Research

The traditional economic research process generally depends on structured datasets collected through surveys, censuses, administrative systems, and statistical agencies. Although these sources remain essential, they often involve substantial time lags and relatively limited frequency. Big data provides alternative sources that can supplement these established systems with more frequent observations [1]. Financial transactions represent one important source. Credit-card purchases, electronic payments, banking records, and digital-wallet transactions can provide information about consumption patterns and economic activity at high frequency. Similarly, online prices can be collected continuously, creating opportunities for more timely measurement of price movements and inflation [6].

Digital platforms also generate information concerning employment and business activity. Online job vacancies can reveal changes in labour demand across occupations and geographical areas, while online business listings and digital sales data can provide information about firm activity. Satellite imagery can further contribute to economic measurement by providing information concerning agricultural production, construction, urban expansion, nighttime economic activity, and environmental changes [7]. The value of these sources lies not only in their volume but also in their frequency and granularity. Economic researchers can increasingly study economic processes at daily, weekly, or even real-time intervals and examine differences between localities, demographic groups, industries, and firms.

3. Big Data in Macroeconomic Forecasting

Macroeconomic forecasting is one of the most important applications of big data analytics. Governments and central banks require timely estimates of inflation, output, consumption, employment, investment, and other indicators to formulate monetary and fiscal policy. Traditional macroeconomic statistics are often published with delays and may subsequently undergo revisions. Alternative data can reduce this information gap. Online prices, financial transactions, mobility indicators, electricity consumption, shipping activity, and search behaviour can provide early signals of changing economic conditions [6].

Machine-learning models can process large numbers of predictors simultaneously and identify nonlinear relationships that may be difficult to capture using conventional forecasting approaches. However, predictive accuracy should not be confused with economic explanation. Econometric theory remains important for interpreting relationships, identifying mechanisms, and evaluating policy effects [3]. The combination of traditional indicators with high-frequency alternative data can therefore produce more flexible forecasting systems. Such systems may be particularly valuable during economic shocks, when historical relationships between conventional variables may become unstable.

4. Big Data and Labour-Market Research

Labour markets have been substantially affected by digitalization, and online data provide new opportunities for monitoring employment dynamics. Online job vacancies contain information about employer demand, required skills, wages, locations, and occupational requirements. These data can reveal changes in labour demand more rapidly than conventional employment surveys [8].

Big data can also help identify emerging skills and changes in occupational structures. Natural language processing can be used to analyse job descriptions and identify increasingly demanded technical, cognitive, and interpersonal skills. This information can support education and workforce-development policies by identifying gaps between skills supplied by workers and skills demanded by employers. However, online vacancy data may not represent the entire labour market. Informal employment, small businesses, occupations recruited through personal networks, and workers with limited digital access may be underrepresented. Researchers must therefore combine digital labour-market indicators with conventional employment statistics.

5. Big Data in Consumer and Household Economics

Consumer behaviour generates enormous amounts of digital information through online purchases, electronic payments, search activity, reviews, and digital-platform interactions. These data can provide insights into consumption patterns, price sensitivity, product preferences, and responses to economic shocks. Real-time transaction information can help policymakers monitor changes in household spending. During economic crises, such information can provide early evidence of declining consumption or changing expenditure patterns [4]. Retailers and policymakers can also use spatially detailed data to examine differences between regions and communities. Nevertheless, digital consumption data may be affected by selection bias. Individuals who do not use digital payment systems or online shopping platforms may be systematically different from those who do. Consequently, digital consumer data should be interpreted within the broader demographic and socioeconomic context.

6. Financial Markets and Risk Analysis

Financial markets are particularly suitable for big data applications because they generate enormous quantities of high-frequency information. Market prices, transaction records, financial news, corporate disclosures, analyst reports, and social-media discussions can all be incorporated into analytical models. Machine-learning techniques can identify patterns in financial data and support risk assessment, portfolio analysis, fraud detection, and market surveillance. Natural language processing can analyse financial news and corporate reports to identify information potentially relevant to market expectations [9]. Big data can also support financial stability monitoring. Financial institutions and regulators can use large-scale transaction data to identify unusual patterns and potential systemic risks. However, financial models based on historical patterns can perform poorly during structural breaks or unprecedented crises. Strong governance and human oversight are therefore essential.

7. Big Data and Public Finance

Governments generate extensive administrative data through taxation, social security, customs, public procurement, business registration, and welfare programmes. Integrating these sources can improve tax administration, identify fraudulent claims, target public expenditure, and evaluate social programmes [10]. Machine-learning methods can identify unusual patterns in tax returns or procurement transactions, helping authorities prioritize cases for further investigation. Data integration can also improve the targeting of social assistance by identifying eligible populations and monitoring programme outcomes.

However, administrative data are collected primarily for operational purposes rather than research. Differences in definitions, missing information, inconsistent records, and interoperability problems can complicate integration. Strong institutional data governance is therefore necessary.

8. Big Data in Agricultural and Regional Economics

Big data analytics has considerable potential for agricultural economics. Satellite imagery, weather observations, soil data, farm records, market prices, and mobile information can be combined to monitor agricultural production and environmental conditions. Remote sensing can provide estimates of crop conditions, land use, drought exposure, and vegetation health. These data can support agricultural forecasting, crop insurance, food-security monitoring, and disaster assessment [7]. At the regional level, satellite imagery and geospatial data can help measure infrastructure development, urbanization, economic activity, and regional inequalities. Such information can support spatially targeted development policies.

9. Artificial Intelligence and Machine Learning in Economic Analysis

Artificial intelligence and machine learning have become increasingly important in big data analytics. Traditional econometric models often require researchers to specify functional relationships between variables. Machine-learning methods can identify complex nonlinear patterns and interactions with fewer assumptions about functional form [3]. Supervised learning can be used for prediction and classification, while unsupervised learning can identify clusters and hidden structures. Natural language processing allows researchers to convert textual information into quantitative indicators. Deep-learning techniques can analyse complex data such as images, text, and high-dimensional time series. Despite these advantages, machine learning has limitations. Highly predictive models may be difficult to interpret, and algorithmic results can reflect biases contained in training data. Economic policy decisions require understanding not only what is likely to happen but also why it happens and how policy intervention may change the outcome. For this reason, machine learning should complement rather than replace economic theory and causal inference.

10. Big Data for Policy Evaluation

Policy evaluation is another major area of application. Governments need reliable evidence to determine whether programmes achieve their intended objectives. Administrative records, digital transactions, geospatial data, and platform information can provide detailed evidence about programme participation and outcomes. Big data can improve policy evaluation by allowing researchers to examine heterogeneous effects across regions and population groups. It may also permit more frequent monitoring than traditional surveys. However, causal inference remains a central challenge. Large datasets can reveal statistical associations but cannot automatically establish causality. Researchers must therefore combine big data with appropriate identification strategies, including natural experiments, randomized interventions, difference-in-differences approaches, instrumental variables, and other econometric techniques.

11. Big Data and Crisis Management

Economic crises require rapid information. Conventional indicators may become less useful when economic conditions change rapidly, particularly because they are often released with substantial delays. Big data can provide near-real-time indicators of economic activity. During the COVID-19 crisis, researchers used mobility data, credit-card transactions, employment information, and online activity to monitor changes in economic behaviour [4]. These applications demonstrated how alternative data can support rapid policy responses. The same principle can apply to natural disasters, financial crises, food-price shocks, and geopolitical disruptions. High-frequency data can help policymakers identify emerging problems and allocate resources more quickly.

12. Ethical, Privacy, and Governance Challenges

The expansion of big data analytics creates significant ethical challenges. Personal financial transactions, mobility records, online behaviour, and digital communications may contain sensitive information. The use of such data requires appropriate privacy protections and governance mechanisms [5]. Data security is another major concern. Centralized databases can become targets for cyberattacks, while unauthorized data sharing may expose individuals and organizations to financial or social harm. Algorithmic bias is equally important. If datasets systematically underrepresent certain populations, predictive models may generate unequal outcomes. Policymakers must therefore examine data representativeness and algorithmic performance across demographic and socioeconomic groups. Transparency is also essential. Citizens should have appropriate information about how their data are collected and used, particularly when automated systems influence access to public services, credit, employment, or social benefits.

13. Integrating Big Data with Conventional Economic Methods

Big data should not be viewed as a replacement for traditional economic statistics. National accounts, household surveys, administrative data, and established economic indicators remain essential because they provide standardized and theoretically grounded measures. The most productive approach is likely to be integration. Traditional datasets can provide representative benchmarks, while alternative data can provide higher frequency and greater granularity. Econometric models can establish causal relationships, while machine-learning techniques can improve prediction and identify complex patterns. Such integration can create a more comprehensive evidence infrastructure for economic policy. Statistical agencies, universities, financial institutions, and governments can collaborate to develop data systems that combine conventional and alternative sources while maintaining rigorous methodological standards.

15. Conclusion

Big data analytics is transforming modern economic research by expanding the scale, frequency, and diversity of information available for economic analysis. Digital transactions, online platforms, administrative records, satellite imagery, financial data, and textual information provide new opportunities to measure economic activity, forecast changes, understand behavioural patterns, monitor labour markets, evaluate policies, and respond to crises. The principal advantage of big data lies in its ability to complement conventional economic statistics with more timely and granular information. Machine learning, artificial intelligence, and natural language processing further expand analytical capabilities by allowing researchers to process complex datasets and identify patterns that may otherwise remain difficult to detect. Nevertheless, the availability of more data does not automatically produce better economic knowledge. Questions concerning representativeness, data quality, privacy, algorithmic bias, interoperability, and causal inference remain fundamental. Predictive models must be interpreted carefully, particularly when they are used to guide public policy. The future of economic research therefore lies not in replacing conventional economics with big data analytics but in developing an integrated methodological framework in which economic theory, econometrics, official statistics, administrative records, and alternative digital data complement one another. When supported by strong governance, ethical safeguards, methodological transparency, and appropriate analytical techniques, big data analytics can substantially improve the evidence base for economic policy and contribute to more timely, responsive, and inclusive decision-making.

References

[1] Einav, L., & Levin, J. (2014). Economics in the age of big data. Science, 346(6210), 1243089. https://doi.org/10.1126/science.1243089

[2] Varian, H. R. (2014). Big data: New tricks for econometrics. Journal of Economic Perspectives, 28(2), 3–28.

[3] Mullainathan, S., & Spiess, J. (2017). Machine learning: An applied econometric approach. Journal of Economic Perspectives, 31(2), 87–106.

[4] Chetty, R., Friedman, J. N., Hendren, N., Stepner, M., & The Opportunity Insights Team. (2020). How did COVID-19 and stabilization policies affect spending and employment? A new real-time economic tracker. Journal of Economic Perspectives, 34(4), 145–166.

[5] Varian, H. R. (2018). Artificial intelligence, economics, and industrial organization. In A. Agrawal, J. Gans, & A. Goldfarb (Eds.), The Economics of Artificial Intelligence: An Agenda. University of Chicago Press.

[6] Cavallo, A., & Rigobon, R. (2016). The billion prices project: Using online prices for measurement and research. Journal of Economic Perspectives, 30(2), 151–178.

[7] Donaldson, D., & Storeygard, A. (2016). The view from above: Applications of satellite data in economics. Journal of Economic Perspectives, 30(4), 171–198. https://doi.org/10.1257/jep.30.4.171

[8] Hershbein, B., & Kahn, L. B. (2018). Do recessions accelerate routine-biased technological change? Evidence from vacancy postings. American Economic Review, 108(7), 1737–1772. https://doi.org/10.1257/aer.20161570

[9] Gentzkow, M., Kelly, B., & Taddy, M. (2019). Text as data. Journal of Economic Literature, 57(3), 535–574.

[10] Athey, S., & Imbens, G. W. (2019). Machine learning methods that economists should know about. Annual Review of Economics, 11, 685–725.

[11] Brynjolfsson, E., Hitt, L. M., & Kim, H. H. (2011). Strength in numbers: How does data-driven decisionmaking affect firm performance? SSRN Electronic Journal.

[12] Goldfarb, A., & Tucker, C. (2019). Digital economics. Journal of Economic Literature, 57(1), 3–43.

[13] Einav, L., & Levin, J. (2014). Economics in the age of big data. Science, 346(6210), 1243089. https://doi.org/10.1126/science.1243089

[14] World Bank. (2021). World Development Report 2021: Data for Better Lives. World Bank.

[15] OECD. (2021). OECD Digital Economy Outlook 2020. OECD Publishing.