KDI - Korea Development Institute - RESEARCH - Research Areas

Summary

Research Monograph KOR The Impact of Artificial Intelligence on the Labor Market and Policy Implications

This report provides an overview of the recent advancements in artificial intelligence (AI) and examines its potential to automate jobs in the Korean labor market. Through an analysis that merges expert assessments of AI capabilities with the skill profiles of occupations in Korea, it is found that, as of 2023, 38.8% of current jobs are at risk of having over 70% of their tasks automated. Predictions for 2030 suggest that nearly 90% of tasks across 90% of existing job types might be subject to automation, if the current task structure should remain as it is.

Nevertheless, the technical feasibility of automation does not directly translate to an imminent risk of automation. As of the end of 2021, only 2.7% of private firms with more than ten employees had adopted AI according to official statistics. When considering the differential AI adoption rates by firm size, it is estimated that about 10% of the workforce is exposed to AI.

In pursuit of a more nuanced understanding of the actual risk posed by automation, this research delves into the impact of increased AI exposure within local labor markets. The findings suggest that while AI's introduction does not meaningfully affect total employment, it tends to lower earnings, with women's average earnings disproportionately affected. More granular analyses indicate that an increase in AI exposure, driven by plausibly exogenous factors, has primarily resulted in diminished employment and earnings opportunities for younger workers at the nascent stages of their careers.

Drawing from these findings, this report proposes policy recommendations to navigate the challenges posed by AI in the labor market effectively.

Summary

Research Monograph KOR Digital Transformation Strategy for Digital-Based Growth

COVID-19 has caused many changes socially and economically. While closing their workplaces and stopping face-to-face meetings, companies expanded non-face-to-face activities such as working from home, webinar. In addition, consumers switched activities such as education, daily necessities purchase, food delivery, and fitness to online or mobile methods. COVID-19 caused a rapid change in the existing supply and demand system, and the changed pattern of economic activity is more likely to continue in the future.

The rapid transformation of lifestyle in the global pandemic situation was possible thanks to digital technology that has continued to develop over the past 30 years since the spread of the Internet in the 1990s. Digital technology has become a general-purpose technology that causes fundamental changes in economic, industrial, and social environment, including education, jobs, and lifestyles, beyond the level of limited use for specific products or services.

In particular, digital technology has undergone a rapid commercialization due to the expansion of infrastructure for high-speed transmission and computation of large amounts of data, the emergence of data-based applications, and the simultaneous development of hardware technology. The demand for digital technology has soared due to COVID-19, and it has emerged at the forefront of our society.

This phenomenon, also called digital transformation, is felt as a rapid and overall change in society. However, Companies has been feeing the impact of digital transformation before Covid-19. Advances in digital technologies such as the Internet of Things, Cloud, Big Data, and Mobile have provided companies with opportunities to innovate through new business models, while causing fundamental changes in the way companies work.

Digital transformation is expected to be positive for productivity improvement and economic growth by promoting the emergence of new products and services and enhancing work efficiency. In addition, digital transformation will promote the expansion of digital trade, thereby expanding export opportunities for products and services through digital. On the other hand, while the risk of cyber security is expected to increase due to high dependence on digital technology, there are also concerns that automation and digitalization of work may have a negative impact on employment.

Accordingly, this study examines tasks for digital transformation and policy directions for stable transition to a digital society in six dimensions: corporate, government policy, regulation, job, education, and global environment.

Chapter 1 identifies the key elements and expected effects of a company's digital transformation, and presents the patterns and levels of digital transformation of domestic companies. Chapter 2 examines policy challenges for digital transformation of SMEs so that SMEs can grow into small and medium-sized enterprises by using digital to increase productivity and strengthening innovation capabilities. Chapter 3 preemptively examines all problems related to the emergence of new industries using digital technology to seek ways to improve regulations that help the smooth growth of new industries. Chapter 4 quantitatively analyzes whether industrial robots, a specific form of automation, replace domestic jobs. Chapter 5 discusses the transition to digital education using artificial intelligence and its tasks as an alternative to solving the problems of our public education. Chapter 6 analyzes the current situation centering on digital trade, a new economic opportunity in the digital era, and examines recent trade issues.

This report summarizes the policy directions that can minimize the side effects that may be caused by digital tansformation while using digital transformation as a stepping stone for the growth of the Korean economy.

First of all, in order to produce a good performance for digital transformation of domestic SMEs, it is necessary to diagnose the readiness of SMEs for digital transformation and provide necessary policy support to SMEs while conducting consulting on what digital technologies and capabilities to be equipped with. In addition, policy integration and coordination is needed so that the government's various digital transformation policies for SMEs can be linked to each other.

Second, it is necessary to clarify regulations in order to harmonize newly emerging technologies and industries with the existing system in the process of transition to a digital society. Regulatory reform has been recognized as deregulation, but the current rapid digitization requires clarification of regulations, preemptive modification and timely provision of regulations, and advancement of regulatory quality.

Third, the trend of robot increase is expected to continue due to continuous digital transformation and industrial structure reorganization, so it is necessary to protect the vulnerable, centering on mid-skilled jobs in the manufacturing industry. It is necessary to expand vocational education so that workers can move smoothly to areas where labor demand is expanding, and to help higher education acquire robot utilization skills and complementary skills.

Fourth, the educational gap should be narrowed by using artificial intelligence. It is analyzed that non-face-to-face education due to COVID-19 has decreased academic ability and widened the educational gap. In order to solve this problem, policy efforts to improve the quality of education, such as strengthening learning feedback, spreading customized education at each student level, and introducing a continuous academic diagnosis system, are considered to be feasible through transition to digital education using artificial intelligence.

Finally, in order to expand the market through digital trade, digital trade statistics should be advanced, digital trade agreements should be expanded, systems should be established to preemptively respond to overseas data regulations, and support measures should be prepared to resolve digital trade disputes for SMEs.

Summary

Policy Study KOR A Study on Effective Innovation Policy: Focusing on Vaccine R&D

Vaccines can be an effective means to cope with infectious diseases from a social point of view, but investment by private firms in the development of new vaccines could be suboptimal. First, this study briefly discusses a typical vaccine development process and the characteristics of a vaccine product to explain the causes of this market failure.

Next, this study analyzes the optimal use of ‘push incentives’ and ‘pull incentives’ to enhance R&D incentives for a vaccine development firm. More specifically, here, push incentives can lower R&D costs and alleviate the firm’s financing constraint (regardless of R&D result); pull incentives can raise the level of reward from a successful R&D result. This study shows that (i) it is optimal to use only push incentives if the firm faces severe financing constraints, and the social planner’s budget is relatively small; (ii) It is optimal to use both push and pull incentives if the firm faces severe financing constraints, and the social planner’s budget is relatively large; (iii) It is optimal to use only pull incentives if the firm does not face any financing constraints; (iv) In cases where the rewards that the firm can obtain from a successful R&D result are quite limited (i.e., the market size is small), providing push incentives alone is not sufficient to enhance the company's R&D incentives.

Lastly, after examining vaccine R&D support policy in Korea and a few other foreign countries, this study provides relevant policy implications, including the circumstances under which providing pull incentives play an important role, the need for a vaccine innovation policy for harmonious use of push and pull incentives, and institutional improvements required to use pull incentives more effectively.

Summary

Research Monograph KOR Job Creation Strategy in an Knowledge-based Economy

Job creation is now at the top position in the Korean government’s agenda. But the means for achieving this goal as well as general understanding on the issue lag behind Korea’s rapid transition to a knowledge-based economy. Direct job creation constitutes an important part of government policy, and the business calls for labor market flexibility. This reports shows that Korea’s job creation mechanism has shifted from one of a manufacturing economy’s to that of a knowledge economy, and knowledge has been the major driving force of job creation since the 2000s. Manufacturing employment has peaked as early as in 1990, and job creation thereafter has been fueled by the growth of knowledge intensive service industry jobs as in any advanced economies. If we presume two sectors in an economy, tradable and non-tradable, a knowledge intensive service job in the tradable sector creates twice as many jobs in the non-tradable sector as a manufacturing job. The knowledge intensive jobs themselves are created by innovation in manufacturing as well as in services and by business profits. Higher service to manufacturing jobs ratios, the growing share of investment on intangibles, concentration of jobs in the Seoul metropolitan area, all such major changes are phenomena in a knowledge-based economy. Job creation has been active with a rapid transition to a knowledge economy and has slowed down with its weakening.

The nontradable sector is larger in numbers of jobs, but the tradable sector is the engine and the determinant of quality of jobs in an economy. Hence the focus of this report is on the latter. Paradoxically the nontradable sector is less related to the economy’s competitiveness, which allows rooms for labor market policies in the sector.

This report covers firm productivity and knowledge capital, local job creation multipliers, patents and firm values, knowledge property right protection, and university-company research collaboration issues. But this list is not at all comprehensive and further research is called for in order to set up a knowledge economy jobs strategy, in areas such as place-based policies, R&D strategies for job creation, et cetera.

Summary

KDI FOCUS Smart Policies for Smart Factories

□ Smart factories have been found to enhance the productivity of the overall production process, proving to be an appropriate policy as a key strategy for domestic manufacturing innovation. However, to maintain its effectiveness even when the scale and recipients of policy support are expanded, there should be improvements in the policy details and action plans. Above all, making factories smart requires not only the adoption of relevant technologies, but also proper management practices and organizational operations necessary to utilize the new technologies. The government should place the policy priority on helping firms to align workers‘ incentives with the technological change. It is also necessary to smartize the governance system of the overall manufacturing innovation strategy which include smart factories.

- The scale of the smart factory fostering policy, a key element of Korea's manufacturing innovation strategy, is gradually expanding. It is time to check the effectiveness of the policy.

- This study provides policy implications by measuring the level of smartness at the factory-level, analyzing economic effects from enhanced smartiziation, and examining factors that can facilitate the smartization.

- Smart factory is a factory in which workers and equipment exchange information and thus can make better decisions.

- Recent technological progress is drawing much attention to smart factories, but adopting the technology is a way of helping factories become smarter, and not directly transforming them into a smart factory.

- The level of smartization in a factory can be measured based on their levels of system interconnection and data sharing and utilization.

- According to the analysis, the level of smartization in Korean firms is low overall with a huge gap between factories.

- Smartization is found to have varying effects on factories’ performance depending on the type of production process, but in general, smartization helps to significantly improve productivity.

- The US-China trade conflict could directly/indirectly reduce Korea's exports, which in turn, could lead to a fall in GDP.

- The nature of smartization is a fundamental transformation of the manufacturing system, using data as a proxy, and not replacing human labor through automation.

- Automation is likely to curtail labor demand while smartization is found to have different effects on different occupations. In particular, the demand for technicians specialized in process management is not expected to decrease.

- Responsive measures should be developed to support occupations that will likely experience a contraction in employment amid factory smartiziation.

- The adoption of technologies for smartization can be helpful only when it is well aligned with workers’ incentives.

- Technological and organizational factors that determine the level of smartization are complementary, meaning that it would be difficult to smartize by focusing on just one factor.

- To make the policy work more effectively, companies should be provided with management training and consulting support so that they can prepare for technological change. It is recommended support programs for consulting, training and human resource development are systemized.

- The government’s support system needs to be smartized to enhance the efficiency of support and to provide companies with more practical information.

- Effective educational and retraining programs should be provided, targeting occupations that will likely face a decrease in demand, through cooperation between large enterprises and SMEs and between suppliers and users.

- A public-private-academic network platform holding real authority should take the driving role in developing manufacturing innovation strategies and responding to the coming changes.

Summary

Research Monograph KOR Smart Factory: Economic Impacts and Policy Implications

Smart manufacturing, also known as Industry 4.0, represents the digital transformation that modern manufacturing is undergoing. Rather than employing individual technologies for traditional automation, firms try to make the whole manufacturing process more integrated and flexible to meet fast-changing demand. Collecting and utilizing data plays a crucial role in the integration of production.

As digital capabilities become more critical in firm survival and industry competitiveness, major countries put forward strategies to promote and support firms' digitization. The Korean government has also pursued policies to encourage firms, small and medium-sized in particular, to adopt smart manufacturing. The goal is to have 30,000 smart factories by 2022.

Despite its importance and difference from traditional automation, there has been little effort to understand the economics of smart manufacturing. This report is the first comprehensive study to measure the level of smart manufacturing, examine its economic impacts, and identify the determinants of adopting the smart factory. Based on the concrete quantitative evidence, this study provides policy suggestions for the effective digital transformation of manufacturing.

From a technical point of view, a smart factory is defined as "Digitization and networking of all processes, products, and resources." As the smart factory integrates the factory system to digitize and network all production activities―and thereby go beyond simple automation―it can improve the efficiency of the production process, flexibly cope with breakdowns and sudden changes in supply and demand, and even help to develop new products and predict future demands.

We first measure the level of the smart factory by two factory-specific characteristics: (1) the degree in which all the production activities in the factory are interconnected systemically (system integration) and (2) the degree to which the factory collects, shares and utilizes data for decision-making (data share and use). To collect necessary information, we surveyed about 1,000 domestic plants in the Korean manufacturing industry. From the survey result, we find that the sample factories were generally low in the smart factory level, but there was a modest improvement between 2015 and 2017. We also find a significant gap between factories in both levels and change.

Chapter 2 examines the effects of factory smartization on the three key performance indicators (KPI) of the factory: productivity, cost efficiency, and product variety. For productivity, we find that daily production increases when the smart factory level increases. We find similar effects on other KPIs, which are heterogeneous by the type of production process. For example, we find a significant reduction in the lead time only from continuous-process factories, while we observe a reduction in the defect rates, a cost-efficiency indicator, only from assembly line-process factories. For product variety, we find the most substantial positive effect from the batch (and job shop) process.

Chapter 3 explores the employment effect of factory smartization and compares it to that of automation. While economists find negative impacts of individual technologies, such as robots and AIs, engineers argue that smartization, a system-wide change distinct from automation, creates new labor demand for constant reconfiguration and training of workers. This chapter tests the hypothesis by examining the relationship between planned smartization and predicted labor demand by worker groups (production workers, technical engineers, and office workers). The results support the engineers' prediction. While automation is expected to decrease the demand for all three classes, improvement in the production process is not likely to decrease the demand for technical engineers. Smartization in even greater scale - integration of all business activities - is expected to have no adverse effects on all three worker groups. Having said that automation and smartization are different in their job impact, this chapter also emphasizes that there is considerable heterogeneity across firms and workers. Production workers and office workers, young and old workers, and workers at low-level smart factories are those more vulnerable to production smartization. Policymakers should encourage firms to provide them with proper retraining and reallocation program.

Given the effects of the factory smartization on performance and employment, it is important to understand what factors determine the level of smartization. Chapter 4 investigates the determinants including the adoption of technologies relevant to the smart factory with a focus on the role of complementary organizational features. We find that there is a complementarity between human resource management practices and technology adoption in driving smartization. Technology adoption leads to smartization, and such effect magnifies disproportionately with the level of structured human resources management. We also find other magnifiers, such as the presence of an ICT division and CEO's willingness to upgrade the manufacturing system.

The complementarity between technology and organizational practices suggests the need to reconsider the current policy direction. First, before providing support, there needs to be an accurate diagnosis of the target firm's organizational practices, needs, and expectations for the effects of the smartization. This “pre-consulting & post-support” approach will be more effective than merely helping firms to introduce specific technologies. Second, we propose to build a platform that provides one-stop services for SMEs. Third, we make suggestions to make the government's support system smart (government smartization) to increase the effectiveness of the support.

Chapter 5 concludes this study by proposing a plan to reform the governance for manufacturing innovation. The key message is that the private sector should play a central role in both policy-making and implementation process for policies to be effective in this transforming digital era. We suggest to organize a council by benchmarking the Germany’s Industrie 4.0 platform. The council is a network of representatives from business, science, government ministries and trade unions. The council is expected to present the economy- wide digitization strategy. We also argue that cooperation between SMEs and large corporations is crucial in pursuing smart manufacturing in the Korean business environment. We discuss possible policy measures to facilitate their cooperation and make the environment more favorable to SMEs.

Summary

Research Monograph KOR Korea’s Road to an Innovation Economy: Enhancing Productivity through Structural Transformation and Institutional Reform

This study starts with two premises: first, business enterprises are at the center of the innovation economy and second, entrepreneurial activities are most effective within a well-functioning innovation ecosystem. Accordingly, the analyses are comprised of certain aspects of the innovation activities of business enterprises as well as the systemic aspects of the Korean innovation system (KIS). Policy implications are also presented for Korea’s innovation-driven growth strategy including enhancing the innovative capacities of business enterprises and upgrading the support systems surrounding them.

The first, analytical part of the study deals with some of the key aspects of business enterprises’ innovation activities, such as the patterns of business innovation, the relationship between industrial organizations and business innovation and that between M&A activities and business performance. The section also introduces the results of a management survey, which is the first one of its kind in Korea. The second, systemic part contains two chapters on Korea’s innovation system: one on the distinctive nature of the KIS and the other on its evolutionary process. Additionally, in regards to the importance of human resources as a permanent and ultimate source of innovation, one of the following chapters makes a cross-country comparison between adults’ skills and another deals with Korea’s re-training system.

Despite the necessity of the government’s involvement in building an effective innovation ecosystem, government intervention is not straightforward. Governments face the same uncertainties as the market― although often with much more intensity―caused by the rapid changing nature of techno-economic paradigms. In this regard, the fourth industrial revolution presents a prime example. Policies will be most effective when business enterprises respond to the government’s initiatives with their own investments. The list of such policies includes: creating open, competitive market environments; pursuing a two-thronged labor policy of strengthening labor competence and enhancing flexibility in the labor market and; redirecting science and technology policies in a way that enhances the linkage between innovation actors such as business enterprises, universities, and government-funded research institutes. The government should also strive to improve system efficiency with consistency over longer-terms.

Summary

Policy Study KOR Measuring Social Returns to R&D and Two Countervailing Spillovers

Corporate R&D investment creates additional economic value in the process of product development and technology transfer to related companies. The ripple effect of corporate R&D investment has been studied in various ways. It is known that there is a positive spillover effect; although the size of R&D investment returns varies depending on the type of technology, industry across country and time.

The purpose of this study is to accurately estimate the returns of R&D investment and the magnitude of the spillover effect of Korean manufacturing companies. Specifically, this study analyzed the private and public returns to R&D and spillover effects of 6,172 firms in industries where R&D investment was more than 1% of total private R&D investment in 2016. In particular, overestimations of the spillover effect of R&D was prevented by reflecting not only the positive effect of R&D but also the negative effect caused by competitive factors among firms.

The results of the analysis on corporate financial and patent data for 2007-2016 revealed that the average for the social and private returns of domestic R&D investment was approximately 14.24% and 11.77%, respectively―making the spillover effect about 2.47%p. The industries with high R&D spillover effects are the Manufacture of pharmaceuticals, medicinal chemical and botanical products, Manufacture of medical, precision and optical instruments, watches and clocks, and Manufacture of electrical equipment. The industries with small R&D spillover effects are the Manufacture of other transport equipment, and Manufacture of fabricated metal products, except machinery and furniture. As a result of examining the differences in social return by characteristics of firms, it was found that the larger the employment size, the larger the patent registration, and the higher the R&D investment ratio to the sales, the greater the spillover effect of R&D.

Korea's private R&D support policy is characterized by policies for future growth engine industries and small and medium-sized enterprise (SME) support policy. It is highly desirable if technologies and products with high social and economic impact are selected and fostered through such policies. However, if the goals and targets of R&D support are set too broadly without considering the R&D capabilities or influence of individual firms, the effectiveness of the policy may be diminished as it would be indistinguishable from non-R&D SME support policies.

In order to enhance the effectiveness of the R&D support policy, it is necessary to return to basics and raise the amount of knowledge production which is below the social optimum level, and to concentrate resources on more rigorous industries and technologies. This should not only be based on the criteria analyzed in this study, but should also begin with a more precise understanding of the spillover effects of R&D support.

Summary

KDI FOCUS Government R&D Support for SMEs: Policy Effects and Improvement Measures

□ The government’s R&D grant for SMEs has risen to 3 trillion won a year, placing Korea second among OECD nations. Indeed, analysis revealed that government support has not only expanded corporate R&D investment and registration of intellectual property rights but has also increased investment in tangible and human assets and marketing. However, there has been a lack of improvements found in the value added, sales and operating profit because the recipient selection system, which relies solely on the qualitative assessments of technology experts was ineffective. Nevertheless, if a predictive model was properly applied to the system, the causal effect on the value added could increase by more than two fold. Accordingly, to develop such a model, it is important to focus on the economic performance rather than the technical achievements. Also, more policy experiments should be conducted on small firms and a phased approach for R&D financing should to be adopted (① grant → ② equity investment → ③ loan).

- Korea’s R&D subsidy for SMEs has risen to approx. 3 trillion won a year, making Korea the second largest spender among OECD nations.

- The SBIR, the main R&D support program for SMEs in the US, is structured in three phases.

- The US case shows that the small lump sum grants for a large number of small firms is more effective than large funds for a few mid-sized firms.

- The Korean government’s corporate R&D support mostly targets medium-sized development activities, not small-scaled exploratory research.

- Current corporate R&D support which is evaluated by the number of registered patents and publications must be reformed.

- The scalability of intangible assets is as important as the economies of scale of tangible assets.

- In all indicators, recipients exhibited much better performance than nonrecipients at the time of support.

- In most performance indicators, recipients stand lower than non-recipients in terms of growth. The former even posted negative growth in operating profit and R&D investment.

- This study estimated the causal effects using the two-step approach which integrates the nonparametric matching method and parametric regression model.

- The government’s R&D support contributed to SMEs’ debt and equity financing, and firms expanded their investment in capabilities/ assets such as intellectual properties, relational assets, tangible assets and human capital.

- Using the funding, firms invested more in capabilities/ assets, but it did not lead to improving the value added, operating profit and sales growth.

- Firms with high growth prospects were selected as recipients in a smaller proportion while those with low growth prospects were selected as recipients in a larger proportion, making the value added growth of recipients lower than the average.

- Estimation of the heterogenous causal effects on the value added increments found a positive effect only in the top four deciles.

- If the support given to those with negative treatment effect was redistributed, this could double the positive effect.

- In keeping pace with the flexibility in corporate R&D practices, the government needs to explore an operating system in which an active exchange of feedback takes place between R&D experiments and market verification.

- This study suggests reforming the existing recipient selection practice which is solely based on the qualitative evaluation by technology experts, and promoting the use of the predictive model and a phased expansion of policy experiments.

- Evaluations should target economic performance, not publications, IP rights and R&D amount, and a selection model should be developed to optimize the evaluation results.

- In accordance to risks associated with the respective stages of R&D and commercialization, the government needs to choose the most suitable financing methods among grants, equity investment and loan support.

- More free contests should be offered.

Summary

Policy Study KOR Impact of International Trade on Firm’s Innovation and Business Portfolio

This study empirically analyzes how international trade affects the technological innovation of domestic incumbent firms. While international trade is divided into exports and imports, the research focuses on whether it increases the innovation incentives of local firms or instead reduces the incentives for innovation when competition between imports and domestic products intensifies in the domestic market. According to the literature, it is relatively well established that expansion of export markets promotes innovation activities of domestic firms, but the effects of import competition are shown differently among the analyzed countries. This calls for a specific investigation of the Korean case.

The firm-level intensity of innovation is measured through the number of patent applications per employee, representing that the more patents are applied by a firm, the more innovative it is. Next, this study divides import competition into two country groups: one from developed countries and the other from developing countries. The relationship between the level of import competition for each group and the intensity of innovation activities is then investigated.

The analysis finds that Korean firms' innovation activities are significantly affected by the competition between the two groups. Specifically, the import competition from developing countries is shown to promote domestic innovation, but innovation from developed countries tends to reduce innovation incentives. This result unveils two different motives (i.e., escape-competition incentive and Schumpeterian force) that exist between market competition and innovation and thereby supports Aghion et al. (2005, 2009). In particular, firms that have little technological gap with imports from developing countries, including China, are found to be the most sensitive to competition with developing countries, showing that the escape-competition incentive is working well. In other words, firms are responding to import competition according to market mechanisms in Korea.

The findings have important implications for policy-making. Import competition from developed countries occurs mainly high-tech and high- value-added industries that our economy is aiming for. In these industries, market competition has made it difficult for domestic firms to step up as leaders by weakening their innovation incentives. This implies that appropriate government support is necessary, at least concerning technological innovation. Of course, further research is needed specifically on which supporting policies are the most appropriate. However, at least it would be desirable that the government support should be provided in a way that encourages corporate innovation incentives without distorting competition in the market.

SITEMAP

HOT ISSUE

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary