Since the beginning of the Covid-19 pandemic, big data analytics (BDA) remains a signatory medium in the battle against it. Governments and policymakers alike are yet to leverage on this scalable technology in an attempt to curb the economic effects of Covid-19. The primary objective of this study is to leverage on BDA to identify economic shocks, and propose a strategic solution for economic recovery in ASEAN member states (AMS). The findings of this study suggest that BDA techniques, frameworks, and architectures are effective tools in predicting and tracking economic shocks, as well as in designing and implementing an effective economic recovery plan. This study proposes a guideline to governments and policymakers scrambling for resources and considering different options to start an economic recovery process. Our findings draw a roadmap for complex AMS economies struggling to explore options for economic recovery in the response of Covid-19. This study has extended and confirmed the knowledge and perception about harnessing BDA as a tool as well as proposed a scientific approach to design economic policies. The findings outlined in this study contribute to the pipeline of BDA’s theoretical frameworks which is expected to strengthen its novelty in data science.
The respiratory infection caused by the SARS-CoV-2 virus is known as the Coronavirus Disease 2019 (Covid-19), and was first reported in China’s Hubei (Wuhan) district on 31 December 2019 [
As of 14 August 2020, the ASEAN member states (AMS) had more than 350,035 confirmed cases and 86, 33 Covid-19 related deaths [
The World Bank [
Rewinding a few months, global economies, businesses, and investments were optimistic about the output of growth while entering into 2020, despite lingering fears of the US-China trade war [
The economic shocks of Covid-19 are much more diverse in the AMS, despite the comparatively low spread rate. AMS economies are more vulnerable to economic shocks due to the nature of the economies, structural characteristics of the region, and enhanced dependence on China and other countries for trade and investment [
Nevertheless, the dreadful Covid-19 disease continues to cause havoc with humanity, socio-communal systems, and economies worldwide. Data science powered by machine learning, statistical learning, time series modelling, data visualisation, and their experts holds a key to mitigate the global effects of Covid-19, especially, to revive the economic situation in developing countries. Scientists from around the world have used data science to propose simulation models to predict the behaviour of Covid-19 [
As the enormity of the Covid-19 economic threats are clear, governments appear to heavily rely on data science such as big data analytics (BDA), artificial intelligence, and the internet of things (IoT), to curb these effects. Big data (BD) has the potential to develop a better and confident decision making that may help in achieving operational efficiency, cost reduction and lower risk [
The morbidity effects of the Covid-19 pandemic are diverse, thus governments, health professionals, and policymakers are scrambling resources to minimise the effects, by developing a range of response strategies. Even though BDA is extensively used to diagnose patients, conduct contact tracing and remote working, governments and policymakers have yet to truly explore the full potential of BDA. One of the challenges for developing economies, such as the AMS, is to find an economic solution to lower the shocks of the pandemic. Governments consider advance analytics, such as BDA and artificial intelligence (AI) can contribute approximately US$9.5 to15.4 trillion annually to the economy [
This study has three discrete contributions; firstly, it contributes to the literature on economic policies and implications, by suggesting methods to curb economic shocks related to Covid-19. Secondly, it provides the required tools and methods for policymakers, while considering the initiation of an economic recovery process in AMS. Lastly, our study contributes to the pipeline of BDA knowledge and its versatile potential to initiate an economic recovery process, whereby the data may act as a tool while implementing economic recovery policies.
The rest of the paper is organised as follows; section two briefly overviews the economic challenges in AMS. Section three is analyses of the methods and areas of the economy which are vulnerable to shock, and to which BDA can contribute. Section four outlines the policy recommendations for governments and policymakers and section 5 concludes this paper.
Covid-19 has constrained the economy causing irreversible economic shocks to the global economies. These shocks have escalated uncertainties about economic integration across all AMS. Some of the economic fallouts of Covid-19 in AMS are briefly outlined below.
IMF’s [
Country | Before Covid-19 outbreak | During Covid-19 outbreak | Source | ||
---|---|---|---|---|---|
BRN | 1.5% | Sep 2019 | 2.0% | 3 Apr 2020 | ADB |
KHM | 6.8% | Sep 2019 | 2.3% | 3 Apr 2020 | ADB |
IDN | 5.3% | Official target, Aug 2019 | −0.4% to 2.3% | 1 Apr 2020 | MOF |
LAO | 6.2% | Sep 2019 | 3.5% | 3 Apr 2020 | ADB |
MYS | 4.8% | Official target | 3.2% to 4.2%, −2.0% to 0.5% | Feb 2020- April 3, 2020 | MOF, BNM |
MMR | 6.8% | Sep 2019 | 4.2% | 3 Apr 2020 | ADB |
PHL | 6.5% to 7.5% | Official target | −0.6% to 4.3% | 19 Mar 2020 | NEDA |
SGP | 0.5% to 2.5% | Nov 2019 | −4% to −1% | 26 Mar 2020 | MOTI |
THA | 2.8% | Dec 2019 | −5.3% | 8 Apr 2020 | BOT |
VNM | 6.8% | – | 6.09–6.27% | 9 March 2020 | MOPI |
Source: ASEAN policy brief, 2020
The table shows that GDP growth projections in AMS were positive and optimistic before the outbreak, however, these projections started to deteriorate as Covid-19 unfolded in the region. Even though the infection rate in the region is relatively low, yet the GDP growth projections are negative. This is due to plunging oil prices, pressure on asset prices such as equity and fixed income markets, trade and investment interdependence on hard-hit Covid-19 countries, informal economic nature of AMS, lack of social protection system and high revenue-generating businesses being largely affected by Covid-19.
Covid-19 has an immediate effect on socio-economic lives, and it has disrupted the trade, tourism, and production industries in AMS [
Trade partners | Trade (US$ billion) | Share to ASEAN total (%) | ||||
---|---|---|---|---|---|---|
Total trade | Export | Import | Total trade | Export | Import | |
ASEAN Total | 2,825.3 | 1,436.4 | 1,388.8 | 100.00 | 100.00 | 100.00 |
Intra-ASEAN | 650.7 | 346.5 | 304.3 | 23.03 | 24.12 | 21.91 |
China | 483.8 | 199.0 | 284.8 | 17.12 | 13.85 | 20.51 |
EU | 160.9 | 127.3 | 9.17 | 10.20 | 11.20 | 288.2 |
US | 263.0 | 160.3 | 102.7 | 9.31 | 11.16 | 7.40 |
Japan | 231.7 | 114.8 | 116.9 | 8.20 | 7.99 | 8.42 |
Korea | 161.5 | 60.5 | 101.0 | 5.72 | 4.21 | 7.27 |
Hong Kong, China | 118.3 | 100.2 | 18.1 | 4.19 | 6.98 | 1.30 |
Chinese Taipei | 117.4 | 39.7 | 77.6 | 4.15 | 2.77 | 5.59 |
India | 81.1 | 50.7 | 30.3 | 2.87 | 3.53 | 2.18 |
Australia | 66.2 | 39.2 | 27.0 | 2.34 | 2.73 | 1.94 |
Source: ASEAN policy brief, 2020
Travel and tourism are significant economic contributors in AMS, amounting to 12.6% of the total revenue. The outbreak lead AMS to suspend flights to and from China which had a negative influence on the economies, as Chinese tourists account for nearly 20% of inbound tourists in eight AMS. Cambodia, Philippines, and Thailand are rendered as the most vulnerable regions among other AMS, since the travel and tourism sectors in these countries largely contribute to the GDP, employment, and export revenue [
AMS’ governments have announced fiscal packages to support micro, small and medium enterprises (MSMEs) with a median value of about 3.5% of the GDP [
The deteriorating global growth and persistently low inflation have forced major central banks worldwide to adopt a more accommodative policy stance for the near term [
The escalating trade tension between the U.S. and China had already moderated foreign investments in AMS long before the start of the pandemic. The global trade tension declined the stocks of equity portfolio in AMS by 25.1% as the major investors, U.S. and E.U withdrew. The countries highly affected by the pandemic are the largest holders of equity (U.S. 38.4%; E.U. 27.1%), cumulative debt investment stock (U.S. 22.2%; E.U. 30.6%) and foreign direct investments (U.S. 14.1%; E.U. 18.7%) in AMS. The tight global financing conditions across the world and in AMS will prevail provided the pandemic continues. The stock markets in AMS significantly plummeted from the start of the outbreak, whereas, stock markets in Indonesia, Philippines, Thailand, and Vietnam were exceptionally wiped out. The stock markets in Vietnam dropped by 29.3% however, in Malaysia the downward trend was relatively controlled as it declined by 11.8%.
The exchange rates in Thailand, Indonesia (19.8%), and Singapore experienced the largest depreciation among other AMS. The debt to GDP proportion of most AMS is more than 30%, which indicates the over-reliance on reserves of foreign exchange. The continuous depreciation in the exchange rate will escalate debt payment and increase the risks of debt sustainability. This will increase the borrowing cost and force AMS to turn to multilateral institutions to fill the financing gaps. For instance, Cambodia, Indonesia, Laos PDR, Myanmar, and the Philippines have recently used The World Bank’s Covid-19 Fast-Track and other financial support facilities [
This is a qualitative study based on the secondary sources of data collected after reviewing and systematic evaluation of various documents [
Over the past three centuries, the world has suffered from 10 major pandemics of stronger and lesser magnitude, and Potter [
Catastrophic events such as pandemics and terrorist attacks are epitomised as the black swan events as it results in shock, fear, and panic, and often generates a huge amount of data known as big data. It is an innovative technology that digitally stores a large amount of data and can help in revealing patients’ patterns, associations, and differences [
Additionally, the current pandemic, compared to the past pandemics has occurred in a digitalised and substantially connected world [
Covid-19 has fully exposed the fragility of AMS economies essentially relying global valued chains for cross-border trade and transport. The region is substantially exposed to economic shocks due to over-leaning to supply chains, where as much as 40% of its exports rely on value chains [
UNESCAP [
Simultaneously, most AMS have developed exceptional capabilities to respond to public health emergencies, by readily detecting and reporting epidemics as compared to many developed countries. This reflects the regional commitment to protect and improve public health and lives. Even though AMS has established ASEAN Biodiaspora Virtual Center (ABVC) in response of Covid-19. Yet the mechanism and the effectiveness of visualising BDA, particularly in the context of mitigating economic shocks are largely unexplored. The discussion below elucidates how BDA can be embedded and navigated in the economic systems of AMS to absorb existing shocks and enhance economic resilience.
We propose to mitigate the economic effects of Covid-19 by leveraging on BDA via four main application domains. They are predicting the outbreak during economic shocks, tracking economic shocks proposing solutions and providing policy implications to start the recovery process.
Financial crises constitute diverse economic and social costs therefore, governments and policymakers are largely concerned about the timely detection of early warning signs [
New sources of economic information such as those from private data collection done by search engines, social media platforms, and internet companies are a potential source of real-time data known as BD. The ability of BDA to predict financial crises came into press spotlight when the U.S Federal Reserve accidentally published economic forecasts for the next five years on their website, predicting that the world will not face a recession before 2020 [
Our proposed model predicts the onset of economic shocks in AMS through BDA of economic indicators of GDP growth, trade performance, businesses yield, MSMEs performance, and financial markets performance using a machine learning technique. Studies in the past deployed similar economic indicators; domestic production curve [
BD is a potential resource to track economic shocks spurred by Covid-19, which can help in developing economic policies and decisions. Traditionally, policymakers used to make decisions based on the employment and business activities data, retrieved from domestic surveys of households and businesses. Although BDA has a great value in understanding the economy, there are certain limitations present that are now noticeable in the face of the fast-spreading pandemic [
Recent studies and reports have used different methods of BDA to analyse and recommend policymaking tools to track the economic shocks caused by Covid-19. For instance, [
Every day, billions of searches are plugged into Google, Bing, and other search engines, generating a huge amount of data related to consumer spending. Policymakers can track the spending patterns related to consumer needs and wants by analysing BDA of Google searches on cars, road traffic, energy usage and bookings including those for restaurants and hotel reservations made via sites like
BD has evolved as an essential paradigm to create value and implant growth for industries, businesses and science through large data volume [
Currently, no officially approved solution exists to address and suggest the initiation of an economic recovery process. Policymakers and governments are using a plethora of policies, such as fiscal financial support for various economic sectors, in collaboration with IMF, The World Bank, and the Asian Development Bank (ADB), to ease the socio-economic situation across the world [
BDA is proposed as a pragmatic solution to address GDP growth output, trade, tourism, businesses, and production issues in AMS. However, extraction of real-time data and harnessing the actual value of BDA is a major concern for policymakers. [
We propose a holistic architecture to leverage on BDA as a policymaking tool to initiate the economic recovery process. Firstly, policymakers need to focus on how to turn unstructured heterogeneous data into knowledge. We propose a three-layer BDA architecture for AMS policymakers to consider. These are known as data analysis layer, governance layer, and persistence layer [
The current situation requires AMS to receive data on GDP growth, trade, production, business performance, tourism, supply and demand, and consumer spending patterns from reliable sources. After receiving data, pre-processing is required for validation, authenticity, and extraction of key features of the data. In the context of Covid-19, these key features are data related to the marginalised and vulnerable sectors in the economy which require fiscal financial support. Through a data integration module, policymakers need to ensure the homogenous access of data to the governments. The data preparation module refers to preparing data for analysis according to the format expected and the features of the data to be analysed prior. The analysis module represents the methods used to develop the basis of knowledge and future forecasts. Publishing results in the form of reports, tables, graphs, and charts would help provide a policymaking tool for governments to direct a particular policy for the welfare of Covid-19 affected industry sectors and societies. The governance layer represents the application of policies and regulations by the government to the whole data life cycle. Lastly, the persistence layer supports the other two layers by managing issues related to storage needs. Our three-layer BDA architecture is simple yet an effective policymaking tool that AMS governments and policymakers can adopt to initiate an economic recovery process and to address economic shock issues in the face of Covid-19. Data standardisation, availability, and scalability are the major challenges in the commercial application of BD [
After proposing the initiation of the recovery plan, we embark on how to implement a recovery plan by leveraging on BDA. IoT have been widely used in to create smart decision support system to achieve optimal performance through effective managing, compressing and mining of big data [
Policymakers can employ BDA to lower demand and supply imbalances that will help to reduce the risks of production issues in firms. In addition, the machine learning methods are effective tools to analyse consumer preferences and spending patterns therefore, policymakers need to focus on receiving and gathering relevant information from their data sources. Governments may use BDA, such as Lexian based features, to remove bureaucratic and taxation hurdles and adjust the leaning supply chains for the smooth flow of goods and services. This technique is also applicable to ensure the welfare of marginalised and vulnerable communities in AMS, which are most affected due to Covid-19. Ref. [
The governments and policymakers in AMS are actively identifying resources to start the economic recovery process, while most countries enter into the fifth stage of lockdowns due to the pandemic. The unique characteristics of data science models, particularly BDA can help in predicting and tracking economic shocks which resulted from Covid-19 and act as a medium to pave the way in designing an economic recovery process. Explicitly, this study found that BDA techniques such as machine learning, acquisition of real-time data from social networking sites, three layered data architecture and BDA based Lexian and receiver-based characteristics can help to predict, detect, track economic shocks, and suggest economic recovery plans. These approaches also provide guidelines for governments and policymakers to implement the recovery plans. While BDA is authenticated as a pragmatic solution to start the economic recovery process, there are certain implications to be considered by the policymakers.
Generally, data science systems learn and improve over time due to large data volume. The ideal scenario for the proposed model in this study relies on higher data fidelity and volume. However, to implement the above plans, extensive datasets are not available, for instance, data on how Covid-19 has affected consumer spending patterns, in order to base our projections on the socioeconomic impact of Covid-19 on vulnerable economic sectors of the community. These smaller datasets are unsuitable to deploy machine learning models, due to the distributed nature of the different data sources. This refers to the exigency of improved and automated approaches through common standards and international collaborations to mung and wrangle data which may help in obtaining quick, consistent, and functional outcomes. Besides data availability, there are underlining issues within the data itself especially, ones related to quality and accuracy. For instance, data on firms’ production, business performance, and consumer spending patterns amid Covid-19 may be obsolete by the time it is obtained, curated, and annotated to draw a conclusion. This consequence may lower the effectiveness of a policymaking tool developed based on the analysis of such data. Policymakers may thus need to employ specialised transfer learning models with regional characteristics to devise analytical approaches that can be effective within the data limitations.
The current situation clearly demands the urgency of identifying methods to initiate economic recovery plans and mitigate Covid-19 related economic shocks. The methods and techniques of BDA proposed in this study are successfully applied by different industries under different settings in the past. However, there is a high risk of bias due to exclusion of data of other economic indicators which may severely impact the policies designed for economic recovery in the fast-moving nature of the current crises. Therefore, policymakers need to focus on the inclusion of data from all economic indicators from all of its available sources. Conventionally, outcomes from researches often help governments in strategic decision making and policy development. For example, based on the Lexian technique, a government may introduce a stimulus package for a specific population of society or sector of industry. Therefore, policymakers need to ensure an equilibrium between exigency and evidence-based results to propose a policy.
The proposed model and techniques imply acquisition, sharing, and processing of sensitive data which raise issues of ethics and protection of users’ privacy and security. However, the situation imposed by Covid-19 has escalated a dilemma among policymakers, whether to choose a trade-off between creating solutions or being ethical. Policymakers may use simple techniques to move forward, for instance, ensuring the transparency (user is informed) during data collection, keeping users’ anonymity, and improving the data governance framework.
Policymakers need to focus on multidisciplinary collaboration at regional levels to understand the long-term economic impacts of Covid-19. This will allow collaborating with experts from multidisciplinary fields to push for an internationally acceptable economic recovery plan. The implication of black-box models may provide solutions with superficial protection, which is impossible without engaging economists and experts’ interpretations.
The future remains highly uncertain on the point of the pandemic subsiding therefore, AMS need to start leveraging on digital trade to resume economic frontiers and thus mitigate existing and future production loss. Policymakers need to ensure the provision of secure digital platforms that will create an opportunity and train the public to harness social networking sites (SNS) for businesses, which alternatively can help policymakers in generating data for future policy decisions. The governments need to address the digital trade divide at regional and domestic levels on an urgent basis to improve cybersecurity for a better digital trade within their societies.