The digital revolution in Asia and its macroeconomic effects

Saadi Sedik, Tahsin et al. Working Paper The digital revolution in Asia and its macroeconomic effects ADBI Working Paper Series, No. 1029 Provided in Cooperation with: Asian Development Bank Institute (ADBI), Tokyo Suggested Citation: Saadi Sedik, Tahsin et al. (2019) : The digital revolution in Asia and its macroeconomic effects, ADBI Working Paper Series, No. 1029, Asian Development Bank Institute (ADBI), Tokyo This Version is available at: https://hdl.handle.net/10419/222796 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/igo/ ADBI Working Paper Series THE DIGITAL REVOLUTION IN ASIA AND ITS MACROECONOMIC EFFECTS Tahsin Saadi Sedik, Sally Chen, Tarhan Feyzioglu, Manuk Ghazanchyan, Souvik Gupta, Sarwat Jahan, Juan Manuel Jauregui, Tidiane Kinda, Vipichbolreach Long, Elena Loukoianova, Alexandros Mourmouras, Masahiro Nozaki, Simon Paroutzoglou, Cormac Sullivan, Jiae Yoo, and Longmei Zhang No. 1029 October 2019 Asian Development Bank Institute The Working Paper series is a continuation of the formerly named Discussion Paper series; the numbering of the papers continued without interruption or change. ADBI’s working papers reflect initial ideas on a topic and are posted online for discussion. Some working papers may develop into other forms of publication. The Asian Development Bank refers to “China” as the People’s Republic of China. Suggested citation: Sedik, T. S., S. Chen, T. Feyzioglu, M. Ghazanchyan, S. Gupta, S. Jahan, J. M. Jauregui, T. Kinda, V. Long, E. Loukoianova, A. Mourmouras, M. Nozaki, S. Paroutzoglou, C. Sullivan, J. Yoo, and L. Zhang. 2019. The Digital Revolution in Asia and Its Macroeconomic Effects. ADBI Working Paper 1029. Tokyo: Asian Development Bank Institute. Available: https://www.adb.org/publications/digital-revolution-asia-macroeconomic-effects Please contact the authors for information about this paper. Email: sc[email protected] Tahsin Saadi Sedik, Sarwat Jahan, Juan Manuel Jauregui, Tidiane Kinda, and Masahiro Nozaki are senior economists; Sally Chen is the resident representative for Hong Kong, China; Tarhan Feyzioglu and Alexandros Mourmouras are division chiefs; Manuk Ghazanchyan, Souvik Gupta, and Jiae Yoo are economists; Vipichbolreach Long is a special appointee; Elena Loukoianova is deputy division chief; Simon Paroutzoglou and Cormac Sullivan are research analysts; and Longmei Zhang is a deputy resident representative for the People’s Republic of China, all at the International Monetary Fund. The views expressed in this paper are the views of the author and do not necessarily reflect the views or policies of ADBI, ADB, its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms. Working papers are subject to formal revision and correction before they are finalized and considered published. Asian Development Bank Institute Kasumigaseki Building, 8th Floor 3-2-5 Kasumigaseki, Chiyoda-ku Tokyo 100-6008, Japan Tel: +81-3-3593-5500 Fax: +81-3-3593-5571 URL: www.adbi.org E-mail: [email protected] © 2019 Asian Development Bank Institute ADBI Working Paper 1029 T. S. Sedik et al. Abstract Asia is at the forefront of the digital revolution, which promises a radical transformation of the global economy, and indeed of society itself, while at the same time threatening substantial disruptions and dislocation. For instance, workers worry that robots will make them obsolete, while financial supervisors are concerned about the risks to financial stability posed by the latest fintech innovations. At the same time, digitalization may well be a key driver of productivity growth and improved welfare. This paper focuses on whether the digital revolution in Asia is driving growth or disrupting it. Given the diversity of digital innovations and the sparsity of data, the paper uses a multipronged approach to analyze the digital economy in Asia. Keywords: fintech, digital innovations JEL Classification: G21, MI3, O33 ADBI Working Paper 1029 T. S. Sedik et al. Contents 1. INTRODUCTION AND MAIN FINDINGS .................................................................... 1 1.1 Key Findings .................................................................................................... 1 1.2 Striking the Right Balance ............................................................................... 3 2. ASIA’S DIGITAL LANDSCAPE ................................................................................... 3 2.1 Defining and Measuring the Digital Economy .................................................. 4 2.2 Supply ............................................................................................................. 4 2.3 Use of Digital Technologies ............................................................................. 5 2.4 Innovation ........................................................................................................ 9 3. ASIA’S GROWTH: FROM PERSPIRATION TO DIGITAL INSPIRATION ................ 10 3.1 Asia’s Rising R&D Intensity ........................................................................... 10 3.2 Results .......................................................................................................... 11 4. AUTOMATION AND THE FUTURE OF WORK IN ASIA .......................................... 13 5. E-COMMERCE AS A NEW ENGINE FOR GROWTH .............................................. 15 5.1 Evidence from Firm-Level Data Highlights the Benefits of E-commerce for Productivity ............................................................................................... 16 5.2 Platforms Can Magnify the Benefits of E-Commerce but also Raise Competition Issues ........................................................................................ 18 6. DIGITALIZATION OF FINANCE IN ASIA .................................................................. 18 7. DIGITALIZATION TO STRENGTHEN PUBLIC FINANCE ........................................ 21 7.1 Taxation: Opportunities and Risks ................................................................ 21 7.2 Improving Social Safety Nets with Digitalization ............................................ 24 8. THE ROLE OF POLICIES ......................................................................................... 25 8.1 Policies to Facilitate Technological Advances ............................................... 25 8.2 Fostering E-Commerce ................................................................................. 25 8.3 Policies to Manage the Transition and Reduce Inequality ............................. 26 8.4 Digitalization of Finance ................................................................................ 26 8.5 Policies to Strengthen Public Finance ........................................................... 27 REFERENCES ..................................................................................................................... 28 ADBI Working Paper 1029 T. S. Sedik et al. 1 1. INTRODUCTION AND MAIN FINDINGS The digital revolution is underway. While digitalization and automation are not new, they have accelerated in recent years, and a new wave of innovation—triggered by advances in artificial intelligence, robotics, computing power, and cryptography, as well as the explosion of big data—is reshaping the global economy. More so than during past periods of innovation, including the spread of personal computers in the 1980s and the rise of the internet in the 1990s, today’s technological advances are multiple and overlapping, creating synergies and accelerating outcomes. The digital revolution is affecting all sectors, with a far-reaching social and economic impact. The new technologies are general purpose, with the potential to transform the global economy, boost productivity, and fundamentally alter the way we live and work, much as the steam engine and electricity did. But in the process, these technologies may also cause substantial disruptions and dislocations. This paper focuses on whether the digital revolution in Asia is driving growth or disrupting it. 1.1 Key Findings First, Asia has been at the forefront of the digital revolution, though with heterogeneity across the region: • There are Asian players leading nearly every aspect of digitalization, while at the same time some economies are lagging behind. In fact, the region’s economies have the highest dispersion in terms of the adoption of digital technologies—not surprising given that Asia covers the entire income spectrum. Nonetheless, at any given income level, Asian economies are at the frontier relative to their global peers; moreover, digitalization is accelerating even for relatively poor Asian economies. • Automation via industrial robots is one area in which Asia is clearly at the forefront, although it is limited to a few Asian economies. With Asia being the “factory to the world,” it is perhaps to be expected that a full two-thirds of the world’s industrial robots are employed in the region. The use of robots has accelerated since 2010. The People’s Republic of China (PRC) is now the single biggest user, accounting for some 30% of the market; further, in 2016, the PRC, Japan, and the Republic of Korea each employed more robots than the US. But this is not just because production volumes are high in Asia. Robot density (the number of industrial robots per 1,000 workers) is high and rising fast in several Asian economies, attesting to their rapid and extensive adoption. Indeed, the Republic of Korea and Singapore are the global leaders in robot density, followed by Germany and Japan. Finally, Asia is a leader not only in the use of robots, but also in their production—Japan and the Republic of Korea are the world’s top two producers, with market shares of 52% and 12%, respectively. • E-commerce and fintech are other areas in which Asia leads. For instance, the PRC accounted for less than 1% of global retail e-commerce about a decade ago, but that has grown to more than 40%, and the penetration of e-commerce (as a percentage of total retail sales) now stands at 15%, compared to 10% in the US. E-commerce penetration is lower in the rest of Asia, but it is growing fast, particularly in India, Indonesia, and Viet Nam. In terms of fintech, Asian economies have made significant progress, in many cases leapfrogging into new types of technology. For example, in 2016, mobile payments made by individuals ADBI Working Paper 1029 T. S. Sedik et al. 2 for consumption purchases totaled $790 billion in the PRC, 11 times the size of such payments in the US. Asia has also been a leader in cryptoassets, including initial coin offerings. Finally, some small states in the region have even been approached by private investors to adopt cryptoassets as the legal tender, raising serious legal and regulatory concerns. A second key finding is that Asia has already benefited immensely from digitalization. This paper finds that the diffusion of technological innovation has been the key driver of growth in per capita gross domestic product (GDP) in Asia over the past two decades, with digital innovation alone accounting for nearly 30%. The digital component of GDP, proxied most narrowly by the share of the information and communication technology (ICT) sector, is relatively large in many Asian economies—Asia is home to seven of the world’s top 10 economies in terms of the ICT share of GDP. The sector has also been growing substantially faster than overall GDP—twice as fast in India and Thailand, and nearly four times as fast in Japan. Digitalization has also boosted the productivity of nonICT sectors. Innovation in Asia is tilted toward the digital sector, further highlighting its potential to boost future growth. Third, e-commerce has the potential to support growth and rebalance economies. For consumers, e-commerce may translate into better access to a wider range of products and services at lower prices, ultimately boosting consumption. For firms, e-commerce could also provide new business opportunities and access to larger markets and may thus support investment. The econometric analysis shows that participation in online commerce is associated with a more than 30% increase in total factor productivity at the firm level in Asia. Innovation, human capital, and, to some extent, access to finance seem to be behind online firms’ stronger performance. Finally, the paper finds that firms engaged in e-commerce also export 50% more, relying on their skilled labor force and capacity to innovate. Interestingly, e-commerce seems to be especially beneficial for small firms in Asia. Fourth, digitalization presents opportunities for improving public finance in Asia. Government adoption of digitalization can, by improving reporting of transactions, increase value-added tax (VAT), tariffs, and other revenue. The analysis indicates that if Asian economies were to move halfway to the global frontier, import-VAT revenue could rise by 0.6% of GDP. Digitalization can also improve the efficiency of public spending, including via the targeting of social assistance, by reducing inclusion and exclusion errors. More generally, digitalization can improve public financial management systems. Fifth, the paper finds the impact of robots on employment depends on country-specific conditions. Using an approach pioneered by Acemoglu and Restrepo (2017a), the paper analyzes the impact of robot usage on employment across a large sample of economies in Asia, Europe, and the Americas. Contrary to some observers, the paper finds no evidence that robots destroy jobs on net—that is, the productivity-enhancing (and thus job-creating) effects of industrial robots have offset the displacement effect (that is, the destruction of old jobs). Restricting attention to Asia, however, there is a slight negative impact on overall manufacturing employment, and particularly so in certain heavily automated sectors like electronics and automobiles. Furthermore, like others, this paper finds that workers with medium-level education are more vulnerable to displacement than those with either low or high education levels. Interestingly, in Japan, with its aging population and declining labor force, increased robot density in manufacturing is associated not only with greater productivity, but also with local gains in employment and wages. Japan’s experience suggests that other Asian economies facing similar demographic trends in the future, such as the PRC, the Republic of Korea, and Thailand, may also benefit from automation. ADBI Working Paper 1029 T. S. Sedik et al. 3 Finally, the paper finds that economies with a greater propensity for technological leapfrogging have also tended to see declining traditional financial infrastructure, particularly bank branches. Unlike US tech companies, Asian tech giants, especially in the PRC, have become key providers of financial services, putting competitive pressures on traditional financial institutions. Neither the opportunities nor the challenges related to digitalization have yet become fully apparent. Some economists have questioned the ability of technological progress to keep propelling the economy forward, arguing that the low-hanging fruit have mostly been picked, and further advances will become increasingly difficult. Others argue that the new technologies are not widely diffused, complementary innovations and production processes that will boost productivity have not been fully developed, occupations may need to be redesigned, and the capital investments required to implement new technologies have not yet been made. It is worthwhile recalling that it took more than 2 decades for electricity to substantially increase productivity. 1.2 Striking the Right Balance While the digital revolution is inevitable, the outcome—utopian or dystopian—will depend on policies. To realize the potential of the digital revolution, comprehensive policies and fresh thinking are needed. For policymakers, the first hurdle is to accept that the digital revolution is inevitable. Policy responses will need to strike the right balance between enabling digital innovation and addressing digitalization-linked risks. Policy priorities differ across Asia (and the world), as economies’ initial conditions are different. Policies to harness digital dividends include revamping education to meet the demand for more flexible skill sets and lifelong learning, as well as new training, especially for the most adversely affected workers; reducing skill mismatches between workers and jobs; investing in physical and regulatory infrastructure that spurs competition and innovation; and addressing labor market and social challenges, including income redistribution and safety nets. But considering the inherent global reach of these technologies, regional and international cooperation will be key to developing effective policy responses. Policies to soften the labor market impact of new technologies can improve welfare. The more willing society is to support the necessary transition and those who are left behind, the faster the pace of innovation that society can accommodate, while still ensuring that the outcomes improve welfare, with all members better off. With the right policies, the digital revolution could be a new engine of growth and prosperity for Asia and the world. This paper first surveys the Asian digital landscape. It then revisits the debate on the sources of growth in Asia, focusing on the role of digital innovation. The paper then turns to analyzing four specific topics: automation and the future of work; e-commerce as a new engine of growth; digitalization of financial services; and digitalization to strengthen public finance. The final section concludes with a discussion on policy challenges. 2. ASIA’S DIGITAL LANDSCAPE Asia has made significant strides in the digitalization of consumption, production, and innovation. While the PRC has been the global trendsetter in many aspects of digitalization, many economies in Asia have advanced significantly. Nonetheless a digital divide still exists, with only a select few economies adopting digitalization at the highest level of sophistication. The impact of digitalization has also been far-reaching, with fintech already starting to impact traditional banking, e-commerce supplanting smaller businesses, and governments adopting digitalization to improve public finance. ADBI Working Paper 1029 T. S. Sedik et al. 4 2.1 Defining and Measuring the Digital Economy The digital economy can be defined in a narrow or broad sense. The narrow definition refers to the ICT sector only or the “digital sector,” including telecommunications, the internet, services, hardware and software, etc. The broad definition includes both the ICT sector and parts of traditional sectors that have been integrated with digital technology, often called the “digital economy.” The lack of a generally agreed-upon definition of the “digital economy” or “digital sector” is a hurdle to measuring both concepts. In the future, as digitalization penetrates an increasing number of activities and sectors, the boundaries between the digital and physical worlds will be blurred, and the entire world economy may be considered to be digital. Reflecting different definitions, there are a range of measures of the digital economy; unsurprisingly, these provide very different size estimates.1 In addition, there are also many blended indices that include the enabling conditions for driving digitalization (such as ICT infrastructure and mobile penetration) and indicators for certain digital industries (such as e-commerce transactions). For the purposes of this paper, the analysis mainly uses the narrow definition because of data availability. However, the paper also uses other measures to capture developments in specific areas such as robotics and ecommerce. A blended index has also been created for the purpose of the paper. 2.2 Supply Asia’s digital sector is growing, as detailed in Figure 1. Not surprisingly, advanced economies such as Taipei,China; the Republic of Korea; and Japan have a large ICT sector. During 2005–15, the ICT/GDP ratio in Taipei,China almost doubled from 9.3% to about 18% of GDP, while the Republic of Korea’s increased from 7.5% to about 10% of GDP. Major economies that are members of the Association of Southeast Asian Nations, such as Malaysia, Singapore, Thailand, and the Philippines, are also among the global leaders in the ICT sector. Asia is the world’s largest supplier of ICT goods2 and services, accounting for about half of global exports. Within Asia, the PRC’s contribution is about 60% (including Hong Kong, China) and has contributed to more than half of ICT export growth over the past decade, followed by the Republic of Korea. Most of the contribution comes from goods exports, but services are starting to gain momentum (Figure 2). 1 See Zhang and Chen (2019). 2 ICT goods exports include computers and peripheral equipment, communication equipment, consumer electronic equipment, electronic components, and other information and technology goods. ADBI Working Paper 1029 T. S. Sedik et al. 11 in the Jones (2002) and Fernald and Jones (2014) framework, and since the creation of new ideas is ultimately a function of population, it stands to reason that Asia, with its large and fast-growing population, should also increasingly contribute to global knowledge. R&D efforts, whether measured in terms of expenditure or number of researchers, have risen globally in recent decades. Asian economies have seen especially rapid growth in R&D, particularly in the PRC; the Republic of Korea; and Taipei,China. The R&Dexpenditure-to-GDP ratio and the share of researchers in total employment (R&D intensity) are both higher in Japan; the Republic of Korea; and Taipei,China than in, for example, the US. Asia, however, still has scope for growth—R&D intensity in the PRC, for instance, has more than doubled since 2000, but is still at relatively low levels. Asia’s R&D intensity in the digital sector and associated patents have increased even faster, but important heterogeneity exists. The share of researchers working on the digital sector ranges from 18% in the PRC to 23% in Australia; 34% in Japan; 46% in the Republic of Korea; and 73% in Taipei,China, as against an OECD average of 30% (Figure 10). Figure 10: Share of ICT Researchers (% of total number of researchers) ICT = information and communications technology, ISIC = International Standard Industrial Classification, PRC = People’s Republic of China, UK = United Kingdom. Sources: Organisation for Economic Co-operation and Development, Research and Development Statistics: Business Enterprise R-D Expenditure by Industry – ISIC Rev. 4; and International Monetary Fund staff calculations. 3.2 Results While earlier literature found that factor accumulation was the key driver of Asia’s growth in the 1960s, 1970s, and 1980s, this section finds that TFP growth (or technological progress) explains most of the economic growth over 1995–2016, though the results vary across countries. In advanced economies, which are closer to the global frontier and have older populations with greater human capital, factor accumulation played a more limited role than in emerging and developing economies. ADBI Working Paper 1029 T. S. Sedik et al. 12 Increases in human capital contributed 11.9% to per capita income growth on average, with the contribution ranging from –0.5% in New Zealand to 27% in Singapore. For some economies in the region, especially those that were affected by the Asian financial crisis, the process of capital deepening made a negative contribution to per capita income growth, ranging from 78.6% in Thailand to 7.2% in Indonesia (Figure 11). Figure 11: Sources of Economic Growth, 1995–2016 (percentage points) ICT = information and communication technology, PRC = People’s Republic of China. Sources: Organisation of Economic Co-operation and Development; Fernald and Jones (2014); and International Monetary Fund staff estimates. More interestingly, the analysis for this section finds that innovation in the digital sector contributed to around 28% of per capita growth over 1995–2016, with contributions ranging between 12% (Singapore) and 49% (Thailand). Since the analysis uses the narrow definition (based on the OECD framework), the share of researchers working on the digital sector could be underestimated. Thus, the contribution of the digital sector to per capita growth could be higher if a broader definition of digitalization were used. In addition, these estimates do not capture the overall contributions from the digital sector to growth, since those from digital capital stock could not be estimated due to data availability issues across economies. Looking forward, the digital sector will likely be an even more important driver of growth in Asia. Indeed, assuming current trends continue, innovation in the digital sector could account for 36% of Asia’s economic growth within 15 years. This section has estimated the contribution of the digital sector to Asia’s per capita growth over the past 25 years. Technological progress is found to have been the main driver of Asia’s per capita growth, and digital technological progress is especially important, accounting for between 12% and 49% of per capita growth. The next four sections dig more deeply into specific aspects of the digital revolution, starting with automation and the future of work. ADBI Working Paper 1029 T. S. Sedik et al. 13 4. AUTOMATION AND THE FUTURE OF WORK IN ASIA This section analyzes the impact of robot usage on employment across a large sample of economies in Asia, Europe, and the Americas. The analysis finds no evidence that robots destroy jobs on net. Restricting attention to Asia, however, there is a slight negative impact on overall manufacturing employment, particularly in certain heavily automated sectors. Furthermore, the analysis finds that workers with medium-level education are more vulnerable to displacement than those with either low or high education levels. Interestingly, in Japan, with its aging population and declining labor force, increased robot density in manufacturing is associated not only with greater productivity, but also with local gains in employment and wages. Automation, like other technological changes, brings both opportunities and challenges. By reducing costs and improving productivity, it may boost economic growth at a time of lackluster productivity growth and demographic headwinds. But the fear is that it may disrupt labor markets in transition as it takes over tasks and makes traditional jobs obsolete. One of the most discussed examples of automation technologies is the use of industrial robots. In 2016, there were about 1.8 million industrial robots—machines that are automatically controlled and reprogrammable to perform physical, productionrelated tasks—operating in the world, and their use has been growing at double-digit rates in recent years (International Federation of Robotics 2017). More importantly, they are becoming more flexible, safer, and cheaper. Many economies in Asia have been at the forefront of automation using industrial robots. More than half of the estimated operational stock of industrial robots is in Asia (1 million units out of a total of 1.8 million units in 2016). These robots are used almost exclusively in manufacturing, with automotive manufacturing being by far the most automated subsector. In several Asian economies, the rise of industrial robots in recent years has also been driven by their use in the manufacturing of computers and electronics. Automation can have two opposing effects on employment. On the one hand, robots may displace jobs, as they replace human labor and reduce labor demand directly. But on the other hand, they may also increase labor demand by boosting productivity and facilitating expanded production (Acemoglu and Restrepo 2017a). Furthermore, the employment impact of industrial robots may also indirectly reach across industries, as a productivity boost in one sector may have positive spillovers across supply chains, thus raising total production and income in the overall economy (Autor and Salomons 2018). The analysis for this section finds a negative impact of robots on manufacturing employment in Asia, but not in the world overall (Figure 12). Following the framework of Acemoglu and Restrepo (2017a), the analysis finds that robot penetration is not significantly associated with net employment losses in a sample of 14 manufacturing subsectors in 40 economies in Asia, Europe, and the Western hemisphere for the period 2010–14. This suggests, contrary to some observers’ worst fears, that the job-creating productivity effect of automation might have offset the displacement effect even at the industry level. When restricting attention to Asia, however, the analysis finds that the increased use of robots is associated with lower employment growth. One more robot per 1,000 employees is associated with a 0.26 percentage point decrease in employment growth in manufacturing sectors. The negative employment effect estimated for Asian economies is driven by highly automated sectors and economies, such as manufacturing of automotive components, plastic and rubber products, and electronics, where robot density was relatively high already in 2010 and has been increasing rapidly since then. This suggests that as automation intensifies, the job displacement effect may start to outweigh the productivity effect at least in the short run ADBI Working Paper 1029 T. S. Sedik et al. 14 at the sectoral levels; a critical mass of robots may be needed before the impact becomes apparent. Also, it is important to note that employment data do not capture jobs created outside the sectors (for example, companies providing robotics repair and maintenance services). Figure 12: Estimated Effect on Employment Growth, 2010–2014 (percentage points, associated with one more robot per 1,000 workers) Note: Figure is based on regressions of the changes in employment on the changes in robots per 1,000 employees during the period 2010–2014. The left three charts are based on 14 manufacturing subsectors in 40 countries, and the right chart is based on countries for which education breakdown of employment data is available. Intermediate education refers to workers with upper secondary and post-secondary non-tertiary education. Bars show the estimated total effects calculated based on the estimate coefficients for each specified group in the horizontal axis. Error bars refer to the 95% confidence interval: ** p<0.05. Sources: International Federation of Robotics; World Input-Output Database; International Labour Organization; and International Monetary Fund staff calculations. The impact of automation, however, depends on country-specific conditions. For example, in Japan, whose demographics dictate a declining labor force, increased robot density in manufacturing is associated not only with greater productivity, but also with local gains in employment and wages. Specifically, panel regressions using estimated prefecture-level robot density show that Japanese prefectures with higher exposure to robots had higher productivity and employment growth. The analysis for this section finds that those prefectures more exposed to robots have sizable positive effects on local labor market outcomes as well as productivity—an increase of robot density by 1% corresponds to a 15% increase in TFP growth for all samples, and of 6% in a manufacturing subsample.4 In addition, employment growth is also positively correlated—a 1% increase in robot density leads to a 0.2% increase in employment growth. Japan’s experience suggests that other Asian economies facing similar demographic trends in the future, such as the PRC, the Republic of Korea, and Thailand, may also benefit from automation. Automation has an uneven impact on employees with different skill levels (Figure 12, panel 4). Automation will render many jobs obsolete, and many will be created and changed. Jobs that are most susceptible to automation tend to involve routine and manual tasks, most prevalent in manufacturing. Those jobs have traditionally been performed by workers with mid-level skills or in the middle of the pay scale (Autor, Levy, and Murnane 2003). Several studies have documented that the use of industrial robots has a negative impact on middleor low-skilled workers, with little effect on high-skilled jobs (Graetz and Michaels 2015). The analysis here also supports an uneven 4 For more details, see IMF (2018). ADBI Working Paper 1029 T. S. Sedik et al. 15 impact: penetration of industrial robots is negatively related to employment growth for workers with secondary education, while there is no significant relation for those with higher education. For workers with upper-secondary education (for example, high school), a standard deviation increase in robot penetration at the economy level (equivalent to about 0.12 more robots per 1,000 employees in an economy over the period of 2010–14) is associated with a decrease in employment of about 0.24 standard deviations (or about a 0.01 percentage point decrease in employment in the sample) on average across the sample economies. The challenge is how to manage the transition. Automation will help increase productivity (Graetz and Michaels 2015), and, as noted above, it may be necessary in the face of population aging. Acemoglu and Restrepo (2017b) found that economies with more pronounced demographic changes tend to invest more in automation technologies, and that helps mitigate the potential negative effect of aging on productivity and output. The challenges with automation, however, involve supporting those who are more vulnerable to changes and in need of a transition to new jobs. The analysis suggests that automation-induced labor market changes may already be happening in some highly automated sectors in Asia. As automation intensifies, there will be a bigger transition necessary, and more workers may need new jobs, especially those who are less skilled. It is thus imperative to provide training and retraining opportunities to help workers adapt and acquire skills that will be in demand. Policies that help create more flexible labor markets, such as active labor market policies, can help absorb employment displacement related to automation. Neither the opportunities nor the challenges have become fully apparent, as robots have not yet been widely used. As with past technologies, productivity effects await complementary innovations. For example, to boost productivity, firms need to redesign production processes and occupations. As these changes are slow, the impact of automation on productivity may even follow a “J-Curve,” that is, productivity may even decline before it ultimately increases (Brynjolfsson, Rock and Syverson 2017). 5. E-COMMERCE AS A NEW ENGINE FOR GROWTH E-commerce can support growth. The econometric analysis shows that participation in online commerce is associated with a more than 30% increase in TFP at the firm level in Asia. Innovation, human capital, and, to some extent, access to finance account for online firms’ better performance. Finally, the analysis for this paper finds that firms engaged in e-commerce also export 50% more, relying on their skilled labor force and capacity to innovate. Interestingly, e-commerce seems to be especially beneficial for small firms in Asia. E-commerce can boost private consumption and investment. For consumers, e-commerce may translate into better access to a wide range of products and services at lower prices, ultimately boosting consumption. Two studies by McKinsey in the PRC and Indonesia highlight that e-commerce generates new consumption. In the PRC, one study shows that out of $100 in internet spending, close to 40% represents incremental (new) consumption, while the remaining 60% is diverted from traditional offline retail channels (Dobbs et al. 2013). In Indonesia, about 30% of online commerce spending is new consumption, capturing previously untapped needs (Das et al. 2018). For firms, e-commerce could also provide new business opportunities and access to larger markets, supporting investment. E-commerce has great potential to improve labor and capital productivity, including for small and medium-sized enterprises. Fast-growing cross-border e-commerce is also ADBI Working Paper 1029 T. S. Sedik et al. 16 gaining traction, bringing greater potential to increase participation in regional and global value chains and support international trade. The empirical literature on the impact of ecommerce on firm activity is limited, but existing evidence suggests an overall positive effect on firm performance. 5.1 Evidence from Firm-Level Data Highlights the Benefits of E-commerce for Productivity This section provides a novel analysis of performance differences between firms engaged in e-commerce and other firms. It relies on World Bank Enterprise Surveys (WBES) and uses a comprehensive sample of developing economies, including several Asian economies during 2006–12. The WBES data include information on firms’ inputs and outputs as well as various characteristics of firms such as age, size, foreign ownership, and export status. Firms with online activities differ on many fronts from other firms. Evidence from the WBES suggests that firms engaged in e-commerce activities tend to have a more educated labor force and better access to finance, and they innovate more than other firms. For instance, a larger portion of online firms, relative to other firms, introduced new products or processes, used technology licensed from a foreign company, spent on R&D, or acquired internationally recognized quality certifications. Possibly reflecting the above factors, e-commerce firms tend to enjoy higher sales, value added, stock of capital, and exports than non-e-commerce firms. Firms with online activities have higher labor productivity. A first look at labor productivity, defined as the ratio of value added to the number of employees, highlights that firms with online activities (sales or purchases) have higher labor productivity (Figure 13). In Asia, firms engaged in online activities seem to have sizably higher labor productivity—on average 50% higher than other firms. Figure 13: Labor Productivity (average, $’000) Note: Labor productivity is the ratio of value added to the number of employees. Sources: World Bank, Enterprise Survey; and International Monetary Fund staff calculations. ADBI Working Paper 1029 T. S. Sedik et al. 17 Firms with online activities, including small firms, also have higher TFP. To capture a more complete picture of the performance differential between firms with online activities and other firms, this section analyzes those differences in TFP. Comparing the distribution of TFP between the two groups confirms that firms with online activities have higher productivity, particularly in Asia. Interestingly, e-commerce seems to be especially beneficial for small firms in Asia. Controlling for firms’ characteristics confirms the results presented in the analysis here (Figure 14, panel 1). The suggestive evidence that firms (including small and mediumsized ones) involved in e-commerce are more productive holds after controlling for several firm characteristics (age, size, foreign ownership, and export status) that are also known to affect performance. Consistent with the earlier evidence presented here, the potential impact of e-commerce on firm productivity seems to be greater in Asia than in other developing regions. Innovation, human capital, and to some extent access to finance seem to support online firms’ greater performance. The higher productivity of firms with online activities seems to occur through their more highly skilled labor force, faster pace of innovation, and, to some extent, better access to finance, which allows these firms to deliver products and services with internationally recognized quality certification. E-commerce firms also export more, relying on their skilled labor force and capacity to innovate (Figure 14, panel 2).5 Firms with e-commerce activities generate a larger share of sales revenues from exports, particularly in Asia, highlighting the potential of ecommerce to promote cross-border trade. A better-skilled labor force and a higher quality of products seem to support higher exports by firms with online activities. The role of skill premia in supporting export activities seems particularly important in Asia. Figure 14: Estimated Impacts of E-Commerce Participation on Productivity and Exports Notes: These figures illustrate coefficients and confidence intervals from two firm-level estimations: (a) the impact of e-commerce participation on total factor productivity controlling for firms' age, size, foreign ownership, and export status; and (b) the impact of e-commerce participation on the share of exports in total sales controlling for firms' size, age, and foreign ownership. The error bars refer to the 95% confidence intervals around the estimated coefficients. For Asia, the estimated coefficients imply that participation in e-commerce is associated with more than a 30% increase in total factor productivity and an increase in the share of exports to total sales by about two units, corresponding to a 50% rise. ** p<0.05; *** p<0.01. Sources: World Bank Enterprise Surveys; and International Monetary Fund staff calculations. 5 For more details, see Kinda (2019). ADBI Working Paper 1029 T. S. Sedik et al. 18 E-commerce, therefore, has the potential to support growth and economic rebalancing by boosting consumption and supporting new industries, especially smaller firms in Asia. 5.2 Platforms Can Magnify the Benefits of E-Commerce but also Raise Competition Issues Platforms can create positive externalities, including through network effects. Platforms have great potential to amplify the economic benefits of e-commerce. In addition to increased competition within the market and pressure to lower prices, including through reduced search costs, a broader geographical reach of suppliers, and savings in supply chain management, platforms bring about additional advantages through network effects. As illustrated in the section above, firms with online activities also have better access to finance. For instance, Ant Financial Services Group, an affiliate company of the PRC Alibaba group, collects information from Taobao, an e-commerce platform that is a subsidiary of Alibaba, to extend the credit frontier to firms not served by traditional banks. By enabling small and medium-sized enterprises to access advanced ICT infrastructure, data centers, applications, and processes usually available to the most productive firms, platforms can further help firms boost their productivity. A higher number of providers or customers using a platform tends to enhance its efficiency, including through using big data to better customize products and services, attracting more providers and customers (same-side network effect). Platforms can also raise competition issues. While e-commerce can provide various benefits, economies of scale and exclusive access to information platforms pose anti-competitive concerns, particularly when e-commerce platforms become large. Network effects also make it challenging for retailers and vendors to switch platforms, reinforcing their market power and exacerbating the risk of anti-competitive practices. Overall, the development of a platform economy has brought significant benefit to consumers, but it also poses many challenges as new issues emerge. Designing the proper policy response remains an open question, especially in the areas of taxation, competition, and data privacy. As a dynamic area of economic development, further research and regulatory experiments would be needed to establish a formal framework for the platform economy. 6. DIGITALIZATION OF FINANCE IN ASIA Fintech can support growth and poverty reduction by strengthening financial development, inclusion, and efficiency. Fintech also poses risks to the financial sector, however. While the use of fintech in Asia is heterogeneous, the analysis for this section finds evidence of convergence. It also finds that fintech is positively associated with financial inclusion yet demonstrates that it also has a potentially disruptive impact on traditional financial services. In Asia, digitalization of finance has been growing faster than the global average. Three of the five economies identified as having the highest rate of fintech adoption globally are in the region (the PRC, India, and Australia). Fintech activities are widespread and have grown rapidly in frontier economies such as Mongolia and Bangladesh, as well as in emerging markets such as Malaysia and Thailand. The growth of fintech activities in Asia has been fueled by a dramatic rise in funding. Since 2010, investments have picked up, led by the PRC, but also in Southeast Asia by Singapore, Malaysia, and Thailand. ADBI Working Paper 1029 T. S. Sedik et al. 19 Cumulative fintech equity funding reached about $28 billion in 2017, with two-thirds of that growth captured by the PRC. However, the development of fintech has not been uniform. Economies have adopted a wide range of technologies based on consumer needs, level of development, regulatory stance, and existing financial and technological infrastructure. For example, while mobile payments have grown rapidly in the PRC, Australia has instead experienced growth in contactless card payments, building on existing infrastructure and experience with the use of cards for secure payment. Similarly, several economies have not developed mobile money products that operate by monetizing pay-as-you-go phone credit, as “postpaid” monthly phone contracts have become standard (replacing prepaid phone credit). The empirical work shows evidence of convergence (Figure 15). Using data on digital payments between 2014 and 2017, the analysis finds economies catching up to the frontier of universal access to digital payments. Economies with low levels of digital payment in 2014 have significantly higher growth rates over 2014–17. This initial evidence of convergence is surprising given the wide underlying heterogeneity in the technologies and business models used. Figure 15: Change in Share of Population Using Digital Payments, 2014–2017 (percentage points) Sources: World Bank, Global Findex database 2017; International Monetary Fund, World Economic Outlook; and International Monetary Fund staff calculations. The econometric evidence indicates that digital financial services can boost financial inclusion (Figure 16). These results are particularly relevant for Asia, where nearly 30% of the population still lacks access to even a basic savings account. Furthermore, given the existing evidence that greater inclusion in the financial system has positive effects on growth, poverty, and inequality, there is potential for greater adoption of mobile technology for financial inclusion to translate into positive macroeconomic outcomes. In addition to direct benefits, fintech has complementary benefits given its role in facilitating other digital activities such as e-commerce. ADBI Working Paper 1029 T. S. Sedik et al. 20 Figure 16: Technological Development and Financial Inclusion (marginal impact of increase, after controls, including country fixed effects) Source: International Monetary Fund staff calculations. The econometric results also highlight the potentially disruptive nature of fintech innovations (Figure 17). Specifically, they suggest that economies with a greater propensity for technological leapfrogging in moving to cellular technologies have also tended to see falling levels of traditional financial infrastructure, particularly bank branches. Specifically, there is a negative association between the adoption of new technology without widespread adoption of prior technology and traditional financial infrastructure. This is particularly pronounced in Europe and the Western hemisphere. However, in Africa, as well as the Pacific, the picture is more mixed, and technology may complement traditional means of financial service delivery, even after controlling for relatively lower levels of income and cellular access. The empirical results are supported by developments at the country level, where many economies in the region have seen an increase in digital banks and a corresponding decline in their physical presence. Figure 17: Leapfrogging and Financial Infrastructure (marginal impact of increase in leapfrogging variable on bank branches per 100,00 adults by geographic region) Sources: International Monetary Fund, Financial Access Survey, World Bank, World Development Indicators; and International Monetary Fund staff calculations. However, fintech also faces challenges in promoting economic development or financial inclusion. Much of the use of fintech has replicated patterns seen in the use ADBI Working Paper 1029 T. S. Sedik et al. 27 Meanwhile, promoting lower barriers to entry while maintaining a level playing field becomes a growing issue with the rising dominance of large firms. Regulations should allow for more competition and further reduce the costs of financial intermediation, while helping solve some problems of the current banking environment, such as the too-bigto-fail issue. In addition, regulation could encourage low leverage among new market participants from the beginning, which would allow for the sustainable growth of the industry and improve discipline, while addressing risks arising from anti-money laundering/combatting the financing of terrorism and cybersecurity threats. Harnessing digital dividends requires a strong cybersecurity framework. 8.5 Policies to Strengthen Public Finance Policy actions can transform risks into opportunities. Digitalization also allows for an increased monitoring of business transactions that would otherwise be informal. This possibility of data collection is particularly evident in P2P platforms when they replace decentralized informal activities. Moreover, the development of P2P platforms can even present an opportunity for governments to pass legislation requiring the withholding of funds related to transactions. The withholding can be established for income, goods and services, or value-added taxes applicable to sellers. This withholding already seems straightforward for indirect taxes. Better data sharing is possible with the increased adoption of digital technologies. The OECD and the G20 have established an automatic exchange of information of nonresident financial accounts. Other useful measures include the establishment of international registers of asset ownership and shareholders, which allows for taxation of capital income on a residence rather than a source basis. A combination of information on assets and capital incomes would allow for the introduction of dual income tax systems under which capital income and wealth would be linked under a single schedule, creating a synthetic capital income tax. ADBI Working Paper 1029 T. S. Sedik et al. 28 REFERENCES Acemoglu, D., and P. Restrepo. 2017a. Robots and Jobs: Evidence from US Labor Markets. National Bureau of Economic Research (NBER) Working Paper No. 23285. Cambridge, MA: NBER. ———. 2017b. Demographics and Automation. National Bureau of Economic Research (NBER) Working Paper No. 24421. Cambridge, MA: NBER. Autor, D., and A. Salomons. 2018. Is Automation Labor-Displacing? Productivity Growth, Employment, and the Labor Share. National Bureau of Economic Research (NBER) Working Paper No. 24871. Cambridge, MA: NBER. Autor, D., F. Levy, and R. Murnane. 2003. The Skill Content of Recent Technological Change: An Empirical Exploration. Quarterly Journal of Economics 118(4): 1279–333. Brynjolfsson, E., D. Rock, and C. Syverson. 2017. Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics. National Bureau of Economic Research (NBER) Working Paper No. 24001. Cambridge, MA: NBER. Dobbs, R., Y. Chen, G. Orr, J. Manyika, M. Chui, and E. Chang. 2013. China’s Retail Revolution: Online Shopping as a Catalyst for Growth. McKinsey Global Institute. Fernald, J. G., and C. I. Jones. 2014. The Future of US Economic Growth. American Economic Review, AEA Papers & Proceedings 104(5): 44–49. Graetz, G., and G. Michaels. 2015. Robots at Work. IZA Discussion Paper No. 8938. Bonn: Institute for the Study of Labor. International Federation of Robotics. 2017. World Robotics, 2017. International Monetary Fund (IMF). 2016. Virtual Currencies and Beyond: Initial Considerations. Staff Discussion Note No. 16/03. Washington, DC: IMF. ———. 2018. Japan – Macroeconomic Implications of Automation. Country Report No.18/334. Washington, DC: IMF. Jones, C. I. 2002. Source of U.S. Economic Growth in a World of Ideas. American Economic Review 92(1): 220–238. Kinda, T. 2019. E-commerce as a Potential New Engine for Growth in Asia. IMF Working Paper (forthcoming). Korinek, A., and J. E. Stiglitz. 2018. Artificial Intelligence and Its Implications for Income Distribution and Unemployment. In The Economics of Artificial Intelligence: An Agenda, edited by J. Agrawal, J. Gans, and A. Goldfarb. Chicago: University of Chicago Press. Krugman, P. 1994. The Myth of Asia’s Miracle. Foreign Affairs 73(6): 62–78. Leite, P., T. George, C. Sun, T. Jones, and K. Lindert. 2017. Social Registries for Social Assistance and Beyond: A Guidance Note & Assessment Tool. World Bank Social Protection and Labor Discussion Paper No. 1704. Washington, DC: World Bank. Michaels, G., A. Natraj, and J. Van Reenen. 2014. Has ICT Polarized Skill Demand? Evidence from Eleven Countries over Twenty-Five Years. The Review of Economics and Statistics 96(1), 60–77. ADBI Working Paper 1029 T. S. Sedik et al. 29 Organisation for Economic Co-operation and Development (OECD). 2016. Skills for a Digital World. Background Paper for Ministerial Panel 4.2. Paris: OECD. Schneider, T., G. H. Hong, and A. V. Le. 2018. Land of the Rising Robots. Finance and Development 55(2): 28–31. United Nations Conference on Trade and Development (UNCTAD). 2017. Robots, Industrialization and Inclusive Growth, Beyond Austerity: Toward Global New Deal. Geneva: UNCTAD. Young, A. 1992. A Tale of Two Cities: Factor Accumulation and Technical Change in Hong Kong SAR and Singapore. NBER Macroeconomics Annual 7: 13–54. Yang, Z., Y. Shi, and H. Yan. 2017. Analysis on Pure E-commerce Congestion Effect, Productivity Effect, and Profitability in China. Socio-Economic Planning Sciences 57: 35–49. World Bank. 2016. World Development Report: Digital Dividends. Washington, DC: World Bank. Zhang, L. and S. Chen. 2019. China’s Digital Economy: Opportunities and Risks. International Monetary Fund (IMF) Working Paper No. 19/16. Washington, DC: IMF.