Anne Lauringson
OECD
Strengthening Active Labour Market Policies in Korea
Connecting People with Jobs
5. Digital employment services in Korea
Abstract
Digitalisation is an integral part of strategies for public employment services (PES) in Korea. As such, digital channels have become the core of PES to support jobseekers and employers and the digital solutions are innovative, using state‑of-the‑art technologies. To continue this successful pathway, clearer frameworks and long-term principles could be established to manage risks associated with advanced technologies and to establish systematic evaluation and monitoring of digital solutions. Korea needs also to carefully design the concept of channel management in PES and support the accessibility of digital services to ensure that jobseekers with lower digital skills or other constraints to using digital channels would not be left behind. The future one‑stop shop of digital PES services could be further augmented by profiling jobseeker skills and interests, facilitating online training, assisting employers in filling bottleneck vacancies and increasing interactivity and support to users of the digital platform.
5.1. Introduction
This chapter assesses the digital infrastructure to deliver employment services in Korea, focusing above all on the recent advancements and future plans, and providing comparisons and examples from other OECD countries to help Korea further continue its successful pathway of digital transformation. Digital systems in public employment services (PES) have become the key enablers to effectively and efficiently support jobseekers, people at risk of job loss and employers. The digital leap was particularly fast during the COVID‑19 pandemic and the positive trend in better digital services has since continued, offering opportunities for mutual learning across PES in OECD countries.
Korea has been one of the OECD countries that has made a particularly impressive progress in public digital services, driven by the Korean Government that aims at the world’s best digital government infrastructure relying on interconnected data and integrated platforms and services, offering high value‑added evidence and data-based innovative public services for the citizens and businesses. Such high prioritisation of digital public services is likely one of the reasons why Korea is placed on the number one position of the OECD 2023 Digital Government Index, well ahead of other OECD countries. The OECD Digital Government Index benchmarks the efforts made by governments to establish the foundations necessary for a coherent, human-centred digital transformation of the public sector (OECD, 2024[1]).
Similarly to the public sector in Korea more generally, the progress in the digital transformation of PES has been remarkable. The digital transformation relies on detailed strategies and has resulted on advanced digital services enhanced by Artificial Intelligence (AI). The progress is continuing to further integrate data, platforms and services and better tailor digital services to jobseekers’ and employers’ needs.
This chapter looks first at the strategies supporting digitalisation in PES (Section 5.2), followed by a discussion on channel management (Section 5.4) and digital tools specific to PES (Section 5.5).
5.2. Digitalisation strategy and resources
Digitalisation is thoroughly embedded in the strategic thinking of employment services in Korea. The high-level vision of digital employment services is addressed by the Ministry of Employment and Labour (MOEL), while the dedicated digitalisation strategy of the Korea Employment Information Service (KEIS) guides the digital developments in further detail. Compared to the other OECD and EU countries, Korea is very advanced in strategic management of PES digitalisation. Only about a quarter of other PES systems have a dedicated digitalisation strategy, another quarter have included some aspects of digitalisation in the general PES strategy and half of the systems do not address the PES digital transformation explicitly in their strategies (Brioscu et al., forthcoming[2]).
5.2.1. The Ministry of Employment and Labour has established a comprehensive strategy to improve employment services
MOEL has an overarching role to design and govern the provision of active labour market policies (ALMPs). As such, MOEL sets the key objectives for digitalisation within the overall strategy of ALMPs called “The Plan to Improve Employment Services”. The current three‑year strategy took effect in January 2023 and sets the objective to change the concept of the provision of ALMPs by i) shifting the focus of employment services toward supporting labour market integration of jobseekers, ii) reorganising the system to deliver online and offline employment services, and iii) creating an ecosystem of employment services that responds to changes on the labour market. To achieve this objective, MOEL sets 12 key tasks centred around four areas – support to jobseekers and working poor, services for employers, capacity and service‑focus of employment services, and co‑operation with private employment services.
Digitalisation of employment services is addressed throughout MOEL’s three‑year comprehensive strategy. The different aspects of PES digitalisation are most thoroughly discussed in the area of capacity and specialisation of employment services. Under this strategic area, one of the key tasks foresees enhancing digital and online employment services. This task consists of three main elements:
Establish an online employment centre (Employment24) – an integrated portal that allows the users to get information on, apply to and report on different employment services and other ALMPs (e.g. training) and subsidies (e.g. unemployment benefits). The portal should benefit from linking data from different registers, such as the National Tax Service, to decrease the need for additional documents and data from the users and increase user-friendliness. The portal would be supported by an enhanced chatbot to cover all PES services and provide tailored guidance. The AI-based job matching service within the portal will provide further customised job recommendations based on enhanced analysis of the up-to-date data on occupations, training and certifications.
Advance non-face‑to-face employment services through the national job portal WorkNet. WorkNet will be made more user-friendly and will be used for a higher volume of non-face‑to-face counselling than before. The information on vacancies will be exchanged with other job portals to mediate more vacancies. Monitoring wage information will be strengthened for better labour market information. WorkNet will be fine‑tuned by establishing a systematic process to collect feedback from the end-users.
Automate simple repetitive tasks, such as payment of various benefits and subsidies to improve efficiency.
The key dimensions concerning digitalisation under the area of capacity and specialisation of employment services, are additionally mentioned under the other three strategic areas as relevant. Regarding support to jobseekers and working poor, the strategy addresses comprehensive digital career management services and online counselling. To improve services to employers and job matching, the strategy discusses the necessary advancements in the digital job matching platform Worknet and co‑operation with private job portals to exchange vacancies via open Application Programming Interfaces (APIs). To strengthen the co‑operation with private employment services, MOEL aims to strengthen data exchange with them (jobseeker data on employment, wages, training, unemployment benefits etc.) and contract out tailored services based on data analysis.
All in all, MOEL’s strategy is a comprehensive and concise high-level concept which aims to significantly improve and modernise the PES business model by setting clear objectives, tasks, targets and timelines. MOEL recognises the digital backbone of delivering employment services rightfully as key to ensure effective, efficient and user-friendly services, and thus highlights the main areas for improvement in digitalisation for the future PES in Korea.
5.2.2. The Korea Employment Information Service has developed a dedicated digitalisation strategy
The Korea Employment Information Service (KEIS) needs to support MOEL in implementing the PES strategy, particularly in terms of ensuring the digital resources to deliver employment services. Ensuring the operation and development of the digital infrastructure for PES and private providers is one of the two key tasks of KEIS. Furthermore, even the second key task of KEIS, conducting research and analysis on ALMPs, is highly dependent on the digital infrastructure and data management in PES.
To guide the task concerning digital developments, KEIS launched a five‑year strategy called “Mid- to long-term development plan for the next-generation digital employment service platform” in December 2022 (following the previous strategy for 2018‑22). In addition to the strategy of MOEL and the digitalisation projects highlighted there, the strategy of KEIS considers the ambitions regarding digitalisation of the Korean Government in its goal setting. The Korean Government aims to implement the world’s best digital government infrastructure where all data are interconnected, providing an integrated and customised platform to solve social issues and create new value to the citizens and businesses, and laying the foundation for research and data-based innovations in public policies. Through advancements in digitalisation, the Korean Government’s objectives for 2023‑27 also aim to increase the effectiveness of ALMPs (and create jobs), strengthen employment services and customise employment support for youth.
Based on the progress made in the digitalisation of the Korean PES system in the previous years, the strategies of MOEL and the government, as well as best practices from other countries, KEIS strategy outlines a clear strategic vision, objectives, principles and 13 tasks centred around four key areas for the next five years (Box 5.1). The key strategy components focus above all around increasing personalisation of digital employment services to better meet the (changing) needs of the clients, and achieve that by using efficient, user-friendly and innovative digital solutions that incorporate cutting-edge technologies.
Box 5.1. The digital strategy of the Korea Employment Information Service sets a clear vision, objectives and tasks for 2023‑27
Key strategic elements in the “Mid- to long-term development plan for the next-generation digital employment service platform”
Vision: Intelligent platform to support employment that guides career choices throughout life for all citizens
Objectives:
Minimising information gaps on employment services among jobseekers to better support employment opportunities and enhance social inclusion.
Strengthening the dissemination of labour market information to support career management over lifetime.
Increasing the utilisation of employment services.
Target principles:
Providing user-friendly and convenient employment services depending on the customer’s age and stage in life.
Applying the “once‑only” principle to data collection from the clients of employment services.
Providing employment services in different ways so that clients can choose the method that suits them best.
Making data on jobseekers’ employment and services available for private employment services.
Tasks:
Data-driven employment services: 1) Expand linking data from external registers and strengthen data analysis, 2) Expand the availability of open data 3) Advance data analytics by collaborating with academia and private sector.
Intelligent, innovative and customised employment services: 4) Introduce proactively customised innovative services for jobseekers, 5) Enhance AI-based recommendation services for jobseekers, 6) Provide customised services for employers, 7) Strengthen “smart” administration of employment services, 8) Introduce digital dossiers and certifications for jobseekers and employers.
Increased utilisation of multi-platform employment services: 9) Improve the environment for using digital employment services (e.g. improving client-centred user experience), 10) Implement the possibility for the users to access employment services wherever they need (e.g. smart call centre), 11) Pilot and adopt new technologies for the digital platform of employment services.
The support system for digital employment services: 12) Redesign the work processes in the Employment Centres to support innovation and change, 13) Improve the support system to provide digital employment services (staff training, change management).
The 13 tasks above a further broken down to 35 specific activities, together with detailed timelines and budgets.
Source: KEIS (2022[3]), “차세대 디지털 고용서비스 플랫폼중장기 발전 방안” (“Mid- to long-term development plan for the next-generation digital employment service platform”).
5.3. The digitalisation strategy of the Korea Employment Information Service should define frameworks and long-term principles, while implementing digitalisation should be supported with agile action planning
The digital strategy of KEIS complements well the strategy to improve employment services by MOEL. While MOEL’s strategy includes high-level references to key digital improvement needs intertwined with other fields of improvement, the KEIS strategy covers digitalisation in PES systematically and in detail. The strategy of KEIS covers additional digitalisation needs compared to those in the strategy of MOEL aiming at a more comprehensive plan for work. For example, in addition to the digital platforms directly used in client services (e.g. user interfaces and operational databases), KEIS discusses other key parts of the digital infrastructure like the data warehouse, software for data analytics, the system of data quality management, the solution to make data available for researchers and the use of cloud solutions. In addition, this five‑year planning goes into great detail to provide the technical descriptions and frameworks for the activities to reach the vision and strategic objectives (combining the strategic elements with action planning). The planning does not only cover what needs to be done, but to a degree also how the tasks should be carried out. Furthermore, the strategy proposes a monitoring framework, outlining indicators with assigned target levels regarding take‑up rates, user numbers, cost and time savings in administration, etc.
KEIS could consider to more clearly separate its long-term strategic concept regarding digitalisation and the more detailed action planning that could be made more agile. The current five‑year planning aims to cover both planning dimensions simultaneously, leading to in total 99‑page document covering aspects from vision to descriptions close to the detail of technical specifications of different digital tools. The taken approach is related to the waterfall project management concept used in KEIS for digital projects, which is a front-loaded method relying on careful planning and detailed documentation in the initial stages of the projects. Separating the long-term strategy and detailed action planning could enable KEIS to introduce more agility in its project management and modern agile methodologies in digital developments without losing sight on the strategic vision and objectives. In such case, action planning could be updated more frequently, e.g. yearly. While the strategic concept would cover long-term objectives and principles, action planning could be yearly revised to cover the next e.g. three to five years ahead (a rolling action plan). Nevertheless, both parts of planning the digital transformation need to remain fully aligned and even in the same document if the long-term strategy and more agile action planning are clearly distinguished.
A long-term strategy of KEIS could also aim at covering the underlying relevant frameworks and principles more comprehensively than the current document. For example, the current planning does not specifically address system and data security (the latter only indirectly for example in the context of data anonymisation and pseudonymisation processes). The strategy should also consider systematic frameworks to ensure digital tools have high value‑added and user-friendliness, for example involving end-users more systematically in the development process (currently only PES staff involved in the planning and inception phases, but no involvement of jobseekers and employers). The strategy could discuss the chosen development methodologies, aiming at modern agile methods. The strategy should lay the foundation of systematic monitoring of digital developments, as well as evaluating digital tools in terms of their effects to the users and the labour market.
As Korea puts a particular emphasis on enhancing its digital solutions with AI and other cutting-edge technologies, it is particularly important that frameworks are created to manage the associated risks of such tools, e.g. due to ethics, biases, trustworthiness, accountability, transparency, fairness, data protection and system security (Brioscu et al., forthcoming[2]). For example, the PES of Belgium (Flanders), France, Germany and the Netherlands that are strongly investing in harnessing AI, have also put in place dedicated frameworks to minimise negative effects and risks incurred by digital technology (Pôle Emploi, 2022[4]; Scheerlinck, 2020[5]), while some PES use the national regulation, cross-policy guidelines and committees to ensure ethicality of their AI tools (e.g. Australia, Canada, Colombia, Finland, Luxembourg, New Zealand). The PES of Flanders launched the Ethical Board comprising of experts on AI from the PES and external experts and academics in the field of AI and digitalisation. The Ethical Bord gives independent and external advice to the PES on AI, profiling and automated decision-making. The ethics and risk management of AI in the PES of Flanders is further ensured by a framework for model risk management and privacy impact assessments (Scheerlinck, 2023[6]). The PES in the Netherlands has adopted the Ethical Charter and established the Ethical Committee to guide their development of AI tools, has a dedicated team to guide the digital transformation in the PES, and interacts with a dedicated national committee (BIT Review Committee) that advises the government and other public sector institutions on larger digital projects. MOEL and KEIS need to consider whether cross-policy regulations and institutions advising on ethical AI in Korea could be sufficient to ensure an appropriate risk management for the digital solutions in PES, or additional dedicated frameworks need to be still set up.
KEIS and MOEL need to continue co‑ordinating tightly also in the future to ensure that their strategies are fully aligned, including regarding the digital transformation of PES. As such, also any additional overarching frameworks to be included in the KEIS digital strategy in the future should be co‑ordinated with MOEL as relevant.
5.3.1. The budget for digitalisation has enabled to cover key needs
As MOEL governs the system of employment services, it also needs to ensure the budget for the system, including for its digitalisation. MOEL assess the budget needs in accordance with the strategy (“The Plan to Improve Employment Services”) and negotiates it within the state budget discussions led by the Ministry of Strategy and Finance. KEIS implements digital developments within the provided budget under the supervision of MOEL.
MOEL and KEIS do not find the budget to be among the main challenges in adopting and implementing digital technology, although the budget is considered to be somewhat limited to be able to cover all relevant digitalisation needs. Nevertheless, the budget has been so far sufficient to make a lot of progress over the past years in digitalising the PES system, including harnessing AI technologies (see Section 5.5). Hence, although the labour authorities in Korea feel less confident in having a sufficient digitalisation budget (53% of OECD and EU countries assess that their digitalisation budget for PES is sufficient or even ideal, while 37% along with Korea see some shortcomings, and 11% have serious concerns),1 their progress in digitalisation does not refer to serious budgeting issues.
KEIS is in charge of implementing PES digitalisation in-house, employing for this task 250 employees, which is thus its biggest strand of work (another 150 employees are in research and analytics, and 50 in administration). Nevertheless, lack of experience and expertise in data analytics and AI technologies is perceived to be a main challenge in adopting new digital solutions, very similarly to the majority of ALMP systems in the OECD and EU countries (62% of countries assess the lack of skills to be the main hurdle in digitalisation, being the most common concern across countries).2 The PES system in Finland has addressed this issue by contracting major digitalisation projects out to private providers with expertise in AI technologies (OECD, 2023[7]). The PES in Belgium (Flanders) has developed the relevant expertise by prioritising the digital transformation and advanced data analytics in-house, as well as by co‑operating other (including private) partners (OECD, 2024[8]). The PES in Estonia co‑operates tightly on data analytics with academia and on new technologies with private IT companies, aiming to improve the relevant skills over time as well (Leinuste, 2021[9]). Hence, the aim to c‑operate with academia and other partners on data analytics and piloting new technologies over time (these are included in the 13 core tasks that KEIS has set to itself), will likely further build the staff skills and capacity in KEIS to and overcome the current constraints to digitalisation.
5.4. Digital services and channel management
Driven by the comprehensive strategic approach to digitalisation, the digital transition has been one of the key dynamics changing the delivery of employment services in Korea over the past years. The Korean labour authorities are making utmost efforts to be at bar with the PES in other the OECD countries in digital services. As a result, digitalisation has become an integral part of the PES business model, and all key services for jobseekers and employers are now available digitally in Korea.
5.4.1. Digital services for jobseekers have become widely available in Korea
Services that are commonly available for jobseekers via online platforms in the OECD and EU countries, are also digital in Korea (Table 5.1). Jobseekers can use online channels for such essential activities like registering with and providing information to the PES, looking for and applying for vacancies, creating job application documents, report and manager their participation in ALMPs (receive information on ALMPs, apply to them, choose a provider, provide feedback). The online platforms in Korea assist jobseekers also in career management, although these will be developed to be still more comprehensive in the coming years (as described in the strategy, see also Section 5.5). Many of these functionalities are not simply digit(al)ised, but aim to provide jobseekers with user-friendly and effective support, often using AI technologies.
Table 5.1. All key services for jobseekers can be provided digitally in Korea
Online solutions for jobseekers, people at risk of job loss and citizens in the OECD and EU countries in spring 2023
Note: Online solutions exclusively for unemployment benefits are not included. Answers from 40 OECD and EU countries and regions regarding 66 different online platforms for jobseekers, people at risk of job loss and citizens that were operational in spring 2023 or soon to be launched.
Source: OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023.
In addition, some Employment Centres provide jobseekers an opportunity to conduct virtual mock job interviews to prepare them for the actual job interviews later on. The AI-based mock interview system launched in 2021 converts the answers of jobseekers into text and assesses the interview characteristics, particularly by frequently used words, providing feedback to the jobseeker on which dimensions could be improved. Such feedback helps jobseekers to prepare for actual face‑to-face interviews, as well as AI-based job interviews that have been increasingly adopted by larger companies in Korea. In addition, the PES system supports virtual job interviews between jobseekers and employers via their platform, which was introduced as a response to the COVID‑19 crises but has stayed relevant even afterwards.
The digital services do not currently offer an opportunity for jobseekers to test their skills (e.g. to have a better understanding for their possible career paths or showcase the skills to employers). Also, virtual training participation is not possible via the online platforms in Korea. Both of these latter functionalities are also less common in the PES in other countries, although are increasingly becoming more popular.
Facilitating access to online training via the PES platform could further help Korea deliver jobseeker services as a one‑stop shop aimed at in the strategies of MOEL and KEIS (see also Subsection 5.4.1).For example, the PES in the Netherlands has a long experience in providing online training courses to jobseekers via platform werk.nl, both as courses to take by jobseekers independently at their own pace, as well as online training involving a coach, including webinars to improve job search skills (networking, drafting CVs and cover letters, excelling at job interviews) or prepare jobseekers for self-employment. The PES in Dutch-speaking Flanders (Belgium) and French-speaking Wallonia (Belgium) double as training providers and offer a wide range on online courses that can be also taken by the jobseekers in the bilingual capital Brussels. Also in other PES, offering online courses has been used more widely, particularly since social distancing due to the COVID‑19 pandemic disrupted face‑to-face training provision in 2020 (then 76% of PES in OECD and EU countries moved training online and 70% introduced new online courses, (OECD, 2021[10])). Also, many online learning platforms made their content then available for the jobseekers registered with PES, including for massive open online courses (MOOCs), and some PES have continued the co‑operation with such external providers to some scale.
Understanding jobseeker skills well can contribute to better targeting of training (including online training), more effective job matching and career services, as well as more personalised support to jobseekers overall. For example, the German PES launched the platform New Plan in late 2020 relying on three pillars: 1) modular tests on different topics and competencies, 2) tailored suggestions for career changes based on test results and other information (experience of jobseekers with similar profiles), 3) further training opportunities to support career progression. The testing pillar includes different categories depending on the purpose of testing, including quick competency tests each taking a couple of minutes on topics like motivation, desire to learn, self-efficacy, objective‑orientation, time‑management and decision-making (Bundesagentur für Arbeit, 2022[11]). The Latvian PES is looking into developing similar tests, as well as potentially in-depth tests for jobseeker digital skills, to better target training and other support to jobseekers. For Korea, similar tests could be useful to feed into the Job Care services and channel management (see more in Subsection5.4.1).
5.4.2. Digital services for employers focus on job matching
Similarly to the PES systems in other OECD countries, the range of digital services available for employers is slightly more limited than for jobseekers in Korea (Table 5.2). Nevertheless, also employers can conduct most of their interactions with the PES digitally, such as upload and advertise their vacancies, look for employees and exchange any relevant data with the PES. In addition, Employment Centres can facilitate virtual job interviews between employers and jobseekers. Furthermore, also the digital services for employers aim at efficiency and user-friendliness in Korea. For example, since 2019, the digital systems suggest details for the vacancy postings automatically, so that employers do not need to insert information from scratch. Strengthening automatic and user-friendly processes further for employers, such as regarding processing grants for employers, is being currently developed (within the integration of digital platforms to Employment24, see Subsection 5.4.1).
Table 5.2. Employers in Korea have a digital channel to advertise their vacancies, look for staff and exchange relevant information with the PES
Online solutions in PES for employers in the OECD and EU countries in spring 2023
Note: Answers from 40 OECD and EU countries and regions regarding 57 different online platforms for employers to support the delivery of active labour market policies that were operational in spring 2023 or soon to be launched.
Source: OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023.
To further facilitate matching labour market supply and demand through digital services for employers, Korea could consider developing a digital tool to assist employers in designing job vacancies. The AI-based tool Action Recrut launched in 2020 aims to help employers that have difficulties in finding staff by identifying those vacancies that the model predicts difficult to fill, as well as assess the appeal of a vacancy posting (an attractiveness score based on criteria set in the vacancy, local labour market situation etc) (Brioscu et al., forthcoming[2]). The information generated by the tool helps PES counsellors to reach out to the employers proactively and discuss the solutions to find the right skills quicker (Pôle Emploi, 2023[12]). In addition, the French PES has developed an AI-based tool DUNE/LEGO to ensure that the vacancies mediated by them are fully compliant with legislation and do not discriminate e.g. by age or gender, so that AI could drop fraudulent vacancies form the job matching platform altogether, and counsellors could reach out to the employers that have issues in their vacancy postings to align these with legislation and validate these vacancies. A similar AI-based tool to identify vacancies with discriminatory elements has been launched also by the Swedish PES in 2023 (Brioscu et al., forthcoming[2]). Combining tools to improve vacancy postings regarding their compliance with legislation and labour market situation also in Korea would simultaneously help employers find the staff they need, as well as support jobseekers find good quality jobs.
5.4.3. Channel management needs a careful concept, so that no‑one is left behind
For higher efficiency and user-friendliness, it is vital to offer digital channels to provide PES services for employers. Employers need to be able to conduct as much of their interactions with PES automatically and digitally (self-services) as possible, decreasing the need to devote PES staff for such standard processes as validating and uploading vacancies or selecting matching jobseekers for vacancies. The digital self-services need to be accompanied by sufficient user support (online resources, chatbots, video calls, phone, in person) to help in case of unexpected errors and anomalies, as well as guide employers with lower digital capacity if needed. A stronger involvement of PES staff is needed to network with employers, solicit employers to use PES services, and provide tailored support – e.g. job crafting, job carving, tailored training (OECD, 2021[13]). Indeed, the local and regional offices of MOEL recognise the need to intervene face‑to-face in case a company has structural challenges to find the staff they need, although the engagement with employers needs to be further strengthened (see Chapter 3).
Achieving the optimal offer of digital self-services to support jobseekers is more complicated, as the same approach does not fit all jobseekers. The core parts of PES support to jobseekers are counselling services that are usually provided throughout the unemployment spell and sometimes even beyond. Counselling services can include different components to help the jobseeker during the labour market integration pathway, such as advice on job search strategy, motivation for job search, identification of needs for additional support, monitor job search efforts etc. Counselling services can have significant positive effects in helping jobseekers to enter employment and not revert to inactivity (Vikström, Söderström and Cederlöf, 2021[14]; Suárez, Cueto and Mayor, 2014[15]) and be a cost-effective way to support jobseekers (Card, Kluve and Weber, 2018[16]). Nevertheless, the effects of counselling are dependent on how these services are delivered (Schiprowski, 2020[17]; Bolhaar, Ketel and van der Klaauw, 2020[18]; Huber, Lechner and Mellace, 2017[19]; Vikström, Söderström and Cederlöf, 2021[14]; Behncke, Frölich and Lechner, 2010[20]). Very importantly, the evaluations have shown that digital counselling arrangements have lower positive effects than face‑to-face counselling (O’Leary et al., 2021[21]; Vervliet and Heyma, 2022[22]), particularly concerning the more vulnerable groups (Schmidt and Mitze, 2023[23]).
The Korean PES system does not have an explicit concept to guide jobseekers to different service channels. The national level MOEL foresees offering the key services (e.g. registering with the PES, counselling, career guidance) via digital and non-digital channels, leaving the choice entirely up to the jobseekers themselves and leading to some self-selection of vulnerable groups to face‑to-face and phone services due to their lower digital skills or limited access to necessary equipment. The local and regional offices of MOEL recognise the need for face‑to-face counselling for vulnerable groups more clearly, to facilitate interactions between counsellors and vulnerable jobseekers, create a trustful environment to receive comprehensive and truthful information on jobseekers’ needs, provide in-depth counselling, better customise the services and resolve the more difficult employment barriers these people face. As such, the Employment Centres guide vulnerable groups somewhat more actively to face‑to-face services to be able to actively intervene and resolve the more complex employment barriers.
The Korean labour authorities need to develop a concept to manage service channels to jobseekers optimally and homogenously across Korea, considering the type and objective of the specific services, as well as the target groups. A digital service that is offered complementary to other services, can have positive effects on jobseekers’ labour market outcomes, particularly if the digital service offers added value compared to services with similar objectives offered offline, for example as the digital service can rely on extensive data analytics to provide knowledge‑based tailored advice (OECD, 2022[24]). Thus, such services could be promoted for all groups of jobseekers. This can be the role for example for the existing and planned career management tools in the Korean PES, in case the basic career counselling services remain available both digitally and face‑to-face. If the digital channel was to fully substitute the in-person channel, it would save costs for the PES on staff, but potentially lead to increases in other costs for the PES and the society, as well as prevent reaching a more inclusive society.
The concept of channel management in Korea, as well as the PES strategy more generally need to aim at sufficient human resources being available for in-depth counselling, both in terms of the number and qualification of counsellors. Replacing face‑to-face channel for a digital only approach would be particularly problematic regarding in-depth support to more vulnerable groups, who need more motivational and encouraging support in addition to addressing their complex employment barriers. The labour market outcomes of these groups can improve if counselling is delivered not only face‑to-face, but also by the same counsellors over the unemployment period (assigned case managers) who believe in their clients’ chances to integrate to the labour market (Rosholm, Sørensen and Skipper, 2019[25]). The counsellors need to master the counselling techniques, as well as be able to make the best of digital tools in the PES available to them. Thus, the Korean PES should aim at creating synergies between counsellors and digital solutions, and not view online and offline services entirely in isolation. Counsellors need to access advanced evidence‑driven digital solutions to increase synergies between human and artificial intelligence, as well as be trained and supported to use these solutions.
Simultaneously with increasing the availability of digital service channels, the PES in Korea needs to increase the accessibility of such channels. Supporting jobseekers to gain relevant digital skills and access equipment to use digital services, would enable jobseekers to access richer support from the PES, as well as contribute to their job-search possibilities more widely. Strengthened access to digital services is particularly important for services that have low efficiency and value added via face‑to-face counselling, such as general job mediation services. Abilities to use the digital channels of PES for job search would potentially further translate to use other digital job search portals. Face‑to-face job mediation and placement would be, however, relevant in cases when counsellors need to solicit and support particularly vulnerable people in direct contact with employers.
Accessibility to digital PES services gained particular attention across OECD countries in 2020‑21 due to the COVID‑19 crises and a sudden need to develop new and adjust the existing digital services (OECD, 2021[26]). In addition to adjusting (online) training to consider and increase participants’ digital skills, PES supported jobseekers with laptops and internet access if needed (e.g. Brussels (Belgium), Canada). To support the digital transition and decrease the digital divide, many PES have further prioritised strengthening digital skills among jobseekers. The Spanish Recovery, Transformation, and Resilience Plan (RRP) was developed in response to the economic and social impact of the COVID‑19 pandemic. One of the objectives of this plan is to support digital transformation in accordance with the “National Digital Skills Plan”, and it includes several major training programmes to increase digital skills among jobseekers and employed with low digital skills. The Estonian PES prioritises ICT training for any jobseeker lacking in digital skills, and enables the use of digital PES services with the relevant equipment on PES premises (OECD, 2021[27]). The major ICT training programmes in Greece (aiming to prepare jobseekers for jobs in ICT sector, rather than only providing basic digital skills), have been evaluated to significantly increase participants’ employment prospects, earnings and upward occupational mobility (OECD, 2024[28]). Neglecting jobseeker digital skills, particularly in channel management can negatively affect PES performance and increase long-term unemployment. Australia launched a new jobseeker portal Workforce Australia in July 2022, directing jobseekers assessed to be job-ready to this platform rather than face‑to-face support. The Australian approach has received critique as employability assessments neglect digital literacy, and many of the jobseekers directed to the online platform indeed lack digital skills, do not get support in increasing digital skills, nor revert to face‑to-face counselling on their own initiative, and potentially suffering longer unemployment duration because of this (Parliament of Australia, 2023[29]).
Successfully prioritising digital channels and increasing their take‑up as foreseen in the strategies of MOEL and KEIS, require that these channels have high performance and user experience, which can only be achieved if end-user insights are systematically taken into consideration throughout the development process (OECD, 2022[24]). User insights can be particularly useful and feasible to collect in the initial planning stage (for the overall purpose and design), when testing the prototype or later stages before adoption, as well as after adoption by setting up a channel for continuous feedback from customers. Yet, Korea does not systematically involve (representatives of) jobseekers and employers in developing digital channels or tools. Involving employers and jobseekers in developing digital tools and channels is somewhat less common and less systematic among PES in the OECD countries than involving PES staff (e.g. counsellors), as it requires more planning and effort. While 95% of PES involve their staff in the development of digital infrastructure, less than half of PES involve employers and jobseekers, mostly in the planning or testing stage (49% of PES involve jobseekers and 44% involve employers).3 As a good example, Sweden has increased the involvement of end-users in developing digital tools and channels to be able to implement its digital first strategy (OECD, 2023[30]). The digital first strategy in the Swedish PES includes such principles as focusing on customer needs to create real value and benefit for the society, showing customers how digital services and data-driven methods create value, strengthening PES ability to create seamless and value‑generating customer experience, and focusing on customers, their accessibility and inclusion while prioritising digital channels. In the French PES, testing the solutions with real users in real use case scenarios and environment is systematically embedded in the development processes to ensure high added value and user experience in their digital tools (Mogollon, 2021[31]).
5.5. Advanced technologies to deliver employment services in Korea
As modernising digital infrastructure, including via advanced technologies, has been prioritised in the strategies of MOEL and KEIS for the past years, great progress has been made in adopting innovative tools using state‑of-the‑art technologies for the core employment services. As such, Korea is among the 45% of OECD countries that are using AI within the digital infrastructure of the PES (Brioscu et al., forthcoming[2]).
Currently, AI is used in the Korean PES to facilitate job matching and career counselling services, which are also the services that are most commonly enhanced by AI in PES in other OECD countries. As Korea has the ambition and resources to further invest in advanced technologies, AI, including generative AI, could be harnessed to further increase the user-friendliness, effectiveness and efficiency of digital PES. In addition, Korea needs to invest in systematically evaluating the effectiveness and cost-effectiveness of its digital solutions.
5.5.1. Recommending jobs and career pathways are supported by advanced digital solutions
Matching jobseekers with suitable vacancies was the first major digital service in the Korean PES that was boosted with AI technologies in 2018‑20. As the first step, the AI-based job matching service TheWork enabled connecting jobseekers with suitable jobs by analysing vacancy postings and resumes using job dictionaries (ontology). The job dictionary uses machine learning to collect information from public vacancies and generate an up-to-date ontology of occupations and their requirements regarding certificates, training etc.
As the second step, KEIS has enhanced the AI-based vacancy matching algorithm to additionally consider the job search behaviour of jobseekers (click data in the job matching portal by similar previous users) to take into account which kinds of vacancies the jobseeker is more likely to be interested in (Figure 5.1). Additionally, the jobseeker (or job counsellor) can filter job matches by specific criteria like the region or wage.
Figure 5.1. The AI-based job matching tool in the Korean PES relies on job descriptions and job search behaviour data
Illustration of algorithms used in the job matching tool developed by the Korean Employment Information Service
Source: Kim (2023[32]), “AI & Public Employment Services in KOREA” in the International Conference on AI in Work, Innovation, Productivity, and Skills, 30 March 2023.
A similar approach of using competency-based matching enhanced with job search behaviour data (but also by the jobseeker themselves, not only previous users) and enabling additional filters is used by one of the other digitally more advanced PES in the OECD, the VDAB in Flanders (Belgium) (Scheerlinck, 2020[5]), and soon by the PES in Luxembourg (Baer, 2023[33]). The AI-based job matching tools in other OECD countries harness AI above all to facilitate (competency-based) matching and facilitate the application of skill taxonomies within the vacancy matching tools (e.g. Canada, France, soon Luxembourg). The job matching platform in Finland Job Market uses Natural Language Processing (NLP) to identify relevant competencies in the vacancy postings and job application documents, in addition to facilitating using the skills taxonomy (European Skills, Competencies, Qualifications, and Occupations classification, ESCO) and matching jobseekers and vacancies (OECD, 2023[7]; OECD, 2024[8]). Korea could analyse whether additionally using NLP could enable taking into account supplementary relevant competencies in the job dictionary (particularly soft skills), or enable detecting skills and competencies easier in jobseekers’ documents and further decrease the administrative burden on job counsellors.
The job dictionaries are additionally used in the new AI-based tool called Job Care to provide career management services, which was launched in 2021 and has been rolled out across Korea in 2023. Job Care uses also information in the jobseeker’s resume and compares this with the information in the job dictionaries and on similar jobseekers in the past, to suggest occupations for job search, possible career pathways and training if upskilling or reskilling is relevant for the desired occupations. Job Care can be used by jobseekers independently, or by counsellors within in-depth counselling sessions. Fine‑tuning and improving of Job Care continues, aiming to involve additional relevant data from other registers and data sources, as well as incorporating generative AI (such as ChatGPT) to facilitate user friendliness of the tool for jobseekers.
In the future, Korea aims to include information on jobseekers’ aptitude, interests and knowledge in the job matching and career recommendation services. Indeed, better understanding jobseeker interests can better tailor career recommendations to them. For example, the PES in Flanders (Belgium), VDAB, uses the AI-based tool called Orient to help jobseekers better understand their job preferences and the job matching platform and counsellors make thus better suited recommendations to the jobseekers (VDAB, 2023[34]; De Blauwe, 2021[35]). Orient asks jobseekers simple questions regarding their preferences for tasks and working conditions taking only about 10 minutes, as the AI-algorithm optimises the questions depending on jobseeker answers, decreasing the time to answer the questions like this down from 45 minutes (Radix, 2020[36]). After completing the short questionnaire, Orient provides a list of occupations ranked by their match to jobseeker skills and considering up-to-date information on labour demand, and links these to open vacancies (OECD, 2024[8]; Brioscu et al., forthcoming[2]).
In addition, MOEL and KEIS could consider improving the functioning of the Job Care platform by feeding in more information on jobseeker skills. Only regarding skills that are detectable in jobseekers’ previous employment history and education provided in their CVs can easily overlook some of the skills and competencies that jobseekers have and do not enable to validate the possession and the level of the skills. Certificates for skills are also likely only covering a fraction of actual skills, particularly regarding soft skills that are becoming increasingly important in labour market matching (OECD, 2023[37]). Korea could consider how to improve the coverage and validation of jobseeker skills that they can use for tailoring (digital) services, for example by learning from competency tests adopted by the German PES (see Subsection 5.3.1)
5.5.2. Counsellors’ needs drive the development of digital tools to support face‑to-face counselling
In general, counsellors in the Employment Centres are well supported by the digital infrastructure to deliver their (offline) services to jobseekers and employers (Table 5.3). The user interface for counsellors in WorkNet enables key functionalities, such as jobseeker registration, job matching and career counselling (Job Care application integrated to WorkNet), profile jobseekers and develop the individual action plans to support their labour market integration pathways, and share information on jobseekers with employers. Only functionalities to support the provision of ALMP measures are somewhat more limited in WorkNet than in user interfaces for job counsellors in other countries. Furthermore, the provision of training to jobseekers is supported by another platform – HRD-Net – via which jobseekers can apply for training and provide their feedback, and training providers can receive information on the training participants, as well as provide data on their participation back to the PES.
Table 5.3. The provision of active labour market policies is not fully integrated in the digital platform to support job counsellors in Korea
Online solutions to support staff in public employment services in the OECD and EU countries in spring 2023
Note: Online solutions exclusively to administer unemployment benefits or manage back-office functions not concerning job counsellors are not included. Answers from 40 OECD and EU countries and regions regarding 46 different digital platforms to support staff in public employment services that were operational in spring 2023 or soon to be launched.
Source: OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023.
In addition to covering the key functionalities, the user interface of WorkNet for counsellors aims at continuous improvement of user friendliness. Counsellors are the only group of end-users that are more systematically involved in the development process of digital tools by KEIS, including in the inception and planning phase, as well as testing and piloting. The needs of counsellors are continuously checked when reorganising and improving WorkNet and implementing new data exchanges with external registers. The counsellors have a channel to continuously post any issues they find in WorkNet (functionalities, speed, needs for additional menus or options etc.) and KEIS has a large, dedicated team to address the issues notified, making the continuous improvement process an exemplary practice for other countries.
5.5.3. Jobseeker profiling could be enhanced to support counsellors’ work, channel management and personalise services to jobseekers
To provide sufficient information for the job matching and career management services via Job Care, jobseekers need to provide their information via the dedicated WorkNet user interface for them when they register with PES. Alternatively, counsellors collect the relevant information during the face‑to-face counselling sessions in Employment Centres, and insert the information in WorkNet (above all personal information, education, preferred occupations and other preferred criteria for the new job, optionally information on work experience, qualification, training, skills etc.).
Job Care enables profiling jobseekers by employment aspirations and to some degree by qualification and skills, but is not a jobseeker profiling tool in the narrow sense of assessing jobseeker general employability to target ALMPs and manage PES resources (see more on jobseeker profiling tools in OECD (forthcoming[38]; 2024[8]; 2018[39]; 2020[40])). Nevertheless, Job Care enables to identify training needs, and the information collected on jobseekers helps counsellors to suggest also other types of ALMPs and steps to take in the individual action plan, although in the Korean ALMP system jobseekers need to apply to different ALMPs rather than are strictly referred to by counsellors.
A more comprehensive jobseeker profiling takes place for those jobseekers that participate in the pilot programme called Jobseeker Advancement Guarantee Package (JAGP), launched in August 2022. The JAGP programme aims to provide jobseekers with more tailored support and thus analyses the employment barriers in more detail. The counsellor uses a dedicated questionnaire in the first counselling session to understand the jobseeker’s professional, economic, and psychological situation in-depth. This pilot programme should be evaluated to generate knowledge on the effectiveness of the more intensive support to jobseekers, as well as of the in-depth information collection and profiling. If the effects of collecting additional information on jobseekers’ labour market outcomes is positive, KEIS could consider augmenting the current digital solution to collect additional information (questions to jobseekers on voluntary basis, triggered by specific replies to mandatory questions, accessing data from external sources) to inform better the support and advice provided to jobseekers.
Even more crucially, Korea should aim to understand better jobseekers’ digital skills to make more informed decisions when targeting digital and face‑to-face channels, and refer them to ICT training. While some jobseekers have certificates for the ICT skills that assure their ability to be able to use PES online services independently, it would be important for the PES to detect those jobseekers that indeed need face‑to-face support early on and prevent them becoming long-term unemployed. For example, the French PES co‑operates with a non-profit organisation Pix that has developed an online platform to test different types and levels of digital skills simulating real-life situations in the digital world (OECD, 2024[8]; Banca and Plard, 2023[41]). In addition to understanding jobseekers’ training needs on digital skills to facilitate their access to job search tools and equip them with the skills needed by the labour market, such knowledge‑based testing enables to provide the jobseekers with relevant certificates (OECD, 2020[42]). Such skill assessment and certification to make jobseeker skills transparent for the jobseekers and employers can support jobseekers getting employed and achieve higher wage (Carranza et al., 2022[43]).
MOEL and KEIS could consider whether to co‑operate with a provider of digital skills testing, integrate these tests within their digital infrastructure (even if only via linking to another platform), trigger referrals to these tests if jobseekers lack relevant certificates, and use the test results for profiling jobseekers for improved channel management and targeting training. While the digital skills of a jobseeker have not been tested or certified, referring them to counselling in the Employment Centres would be advisable.
5.5.4. Streamlining and integrating digital employment services in the new platform Employment24
As the next major step regarding the digital infrastructure for employment services, MOEL and KEIS aim at a better integration of different digital platforms used by Employment Centres, jobseekers and employers. Currently, WorkNet is the main platform to support delivering employment services, to which also TheWork and Job Care are connected to. Additional platforms are developed by KEIS to manage unemployment insurance (ei.go.kr) and training (HRD-Net). Moving towards a new more integrated digital system initiated in 2021 called Employment24 will streamline the different platforms, tools and data exchanges, providing a one‑stop shop of digital services for jobseekers and employers. The new system would enable a more tailored and comprehensive response to jobseeker and employer needs, providing support regarding employment services, training and benefits via the same system. For the PES staff, a more integrated system would enable faster administrative processes and more automation.
Employment24 will aim also at strengthening data exchanges with other registers and increase data validity and integration to further support the concept of one‑stop shop, and increase automation and efficiency. MOEL and KEIS have made significant progress in establishing data exchanges with other registers over the past decade. By now, data are exchanged with many central administrative agencies and public institutions (e.g. information on basic livelihood benefit recipients and welfare recipient of the Ministry of Health and Welfare, business registration information of the National Tax Service, information on the insured of Occupational Health and Safety Insurance of Korea Workers’ Compensation and Welfare Service), local governments and private companies to provide better employment services to jobseekers and recruiting companies. The efforts are continued to establish relevant data exchanges for example with the Ministry of Education and additional data exchanges with the Ministry of Health and Welfare, and solutions to comply with data protection legislation are sought for this purpose. In addition, MOEL and KEIS aim to streamline how data from different sources are used and processed, as well as apply data validity frameworks to establish a single source of truth. The improved data exchanges are expected to feed into both smarter services for jobseekers and employers, as well as monitoring, evaluation and evidence‑based policy making.
Indeed, MOEL and KEIS need to continue their efforts to establish data exchanges with other registers to avoid duplicating information collection from jobseekers (e.g. on education, qualification, certificates, employment history etc) and employers. Accessing such data from an administrative register would make the processes more efficient for the clients and counsellors, as well as help to address the data validity concerns. For example, the Estonian secure data exchange system X-road applying once‑only data collection principle is estimated to save the country about 1 400 years or working time and the digitalised public services about 2% of GDP annually (Nordic Institute for Interoperability Solutions, 2023[44]; PricewaterhouseCoopers, 2019[45]). The Estonian PES exchanges data with 30 registers to receive all data existing in other administrative registers directly from these, e.g. on employment, education, entrepreneurship, qualification etc (OECD, 2023[30]). These data exchanges enable the Estonian PES to develop smart services, as well as have an advanced system to monitor and evaluate ALMPs. In 2019, Estonia launched a digital tool called MALLE to automatically evaluate its main ALMPs and visualise the results, using the rich near-live administrative data from different registers and statistical computing software (OECD, 2022[46]), see Figure 5.2. At the moment, KEIS invests a lot of (human) resources to conduct regular monitoring and evaluation of all individual ALMPs. Developing a system similar to the Estonian MALLE could streamline regular monitoring and evaluation efforts and invest the KEIS resources to meet the more sophisticated needs for evaluation, such as designing randomised controlled trials for new policies and approaches or develop evaluations beyond specific ALMPs, such as digital tools (see next section).
Figure 5.2. Data exchanges with many external registers enable the Estonian PES to conduct counterfactual impact evaluations automatically
The technical set-up of the automatic digital tool to conduct counterfactual impact evaluations of ALMPs in the Estonian PES
Note: PES – public employment service. ALMPs – active labour market policies.
Source: Printsmann (2023[47]), “Automated tool for impact evaluation of active labour market measures: linking administrative data” in the EC-OECD Technical Workshop: Evaluating labour market policies through the use of linked administrative data, www.oecd.org/els/emp/ALMP_impact_evaluation_EC-OECD_workshop_9Oct.pdf.
To further increase the value added of Employment24, Korea could aim to strengthen the information used in the platform on vacancies available on the labour market. Information is already exchanged with some private job search portals like Job Korea, Saramin and Incruit. Some other OECD countries (e.g. Austria, Canada, Germany, Norway, the Netherlands and Switzerland) use web crawling / scraping in addition to dedicated data exchange channels to pool vacancies across country. The PES in the Netherlands is co‑operating with a partner company providing web scraping service (Textekernel) and over one‑third of all vacancies in the job mediation portal of the Dutch PES were collected via this service. Using this approach, the PES in the Netherlands is able to enrich its pool of vacancies, thus increasing the opportunities for jobseekers to find employment via a digital one‑stop shop. In addition to receiving vacancies from 110 online job mediation (Busson, 2018[48]), the French PES is using an AI-based tool called “La Bonne Boîte” to reveal a “hidden job market” since 2015. This tool identifies companies that are likely to hire even if they have not yet published a job vacancy, based on an algorithm that predicts hiring trends using extensive data on French companies. The French job matching platform considers these potential vacancies together with public vacancies to offer more opportunities for jobseekers. The evaluation of “La Bonne Boîte” has shown that this tool has indeed enriched jobseekers’ job search chances and led to increased employment rates (Behaghel et al., 2022[49]). Adopting similar practices – web scraping and predicting vacancies – could also improve the value added of Employment24. A higher coverage of vacancies on the labour market would provide more opportunities for jobseekers using Employment24, as well as enrich the information in the job dictionaries that support the vacancy matching and Job Care career management algorithms.
To further increase user experience of Employment24, it would need to aim at intuitiveness, good user support and interactivity. One way of achieving this is introducing a chatbot within the platform to guide the users, which is also considered to aim at in the strategies of MOEL and KEIS. Currently, the PES in Austria, Greece, Iceland and Norway have introduced chatbots for enhanced user experience, as well as decrease administrative burden for PES staff to answer certain types of (often recurring) questions (Brioscu et al., forthcoming[2]). Similar bots using NLP can be introduced for written communication with jobseekers and employers. Such email-bots are being for example developed in the PES of France and Luxembourg, and would be a logical step for Korea to further decrease administrative burden.
5.5.5. Monitoring and evaluating digital tools and services needs to be systematic
Digital tools and services need to be systematically monitored and evaluated similarly to other services and measures provided by PES to ensure that these tools and services perform well and to continuously improve them. Monitoring take‑up numbers to understand usage and accessibility and collecting user-feedback to pinpoint issues with design and functioning are critical to fine‑tune the digital solutions. In addition, the effects of the digital solutions should be rigorously evaluated to understand whether these are helping jobseekers and employers (Brioscu et al., forthcoming[2]; OECD, 2022[46]). When feasible, the digital solutions should be evaluated using counterfactual impact evaluations (CIE) to credibly determine the labour market effects of such tools (comparing the results for those exposed to the solution and a control group to credibly establish a causal link between the solution and the results). As a gold standard, the evaluations should be conducted using an experimental design (randomised controlled trials) in the framework of piloting these solutions. Depending on the digital solution, evaluating labour market effects might not be always feasible or primary concern, and other or additional aspects might be important, such as time savings in administrative processing. In addition, the effects (benefits) of the digital solutions need to be compared to their costs to generate knowledge on the returns of investments into digitalisation (OECD, 2022[46]).
Currently, Korea generally does pilot its new digital solutions before rolling them out nation-wide, particularly if the end-users of the solution are PES counsellors. Such pilots are used above all to collect qualitative insights on the usefulness and usability of the new solutions to determine whether to continue adopting the solution and which adjustments would be relevant. The feedback on pilots is collected via user satisfaction surveys and (focus group) interviews. Regarding AI-based tools, piloting follows four specific stages: 1) development of AI model and initial performance assessment, 2) verification with developers and system operators, feedback collection from external consultants, 3) end-user testing and beta tests, 4) algorithm comparison tests on all jobseekers. Once, a digital solution has been adopted, its monitoring focuses on take‑up numbers, and occasionally assessing the tool performance or quality (e.g. collecting user feedback). Counterfactual evaluations of digital solutions used in the Korean PES have not been conducted so far.
Generally, the practices of piloting and monitoring digital solutions in PES in Korea resembles the practices in most OECD countries. Close to two‑thirds of PES in the OECD countries pilot at least some of their digital solutions. Monitoring of digital solutions focuses above all on take‑up numbers, end-user experience and general tool performance (81%, 69% and 61% of PES respectively), while only a third of PES has looked at cost efficiency or efficiency incurred by digital solutions (31% and 39% respectively).4 CIEs of digital solutions in PES are even rarer, but their relevance is being increasingly acknowledged and more of such evaluations have been conducted over the past few years. The PES in the Netherlands has focused above all on evaluating the effectiveness of its digital channels (compared to face‑to-face channels, see Vervliet and Heyma (2022[22])), the German PES has evaluated its digital tools for career orientation and job recommendations (tools with a similar objective as Job Care in Korea), the PES in Finland has evaluated the effects of its AI-based job matching platform (Räisänen, 2023[50]; Räisänen, 2022[51]), and the VDAB has studied the effects on job finding rates of online tools to enable registration, provide information, support follow-up actions and resources for e‑learning. In a few other countries (e.g. France, Denmark, Sweden) some tools have been evaluated by external researchers or academia (see e.g. Behaghel et al. (2022[49]), Altmann et al. (2022[52]), (Barbanchon, Hensvik and Rathelot (2023[53])). The Spanish PES developed an online tool Send@ in 2020‑21 to support job counsellors in recommending career paths and providing job search advice (OECD, 2022[24]). The OECD and the European Commission (Directorate General for Structural Reform Support)5 supported the Spanish PES to design a randomised controlled trial for Send@ and evaluated its effects on labour market outcomes of jobseekers counselled with the help of Send@. The positive effect of Send@ on jobseekers’ entry to employment has facilitated the Spanish PES to encourage the counsellors to use Send@, as well as further fine‑tune it.
MOEL and KEIS need to introduce the evaluation of the effects of digital tools systematically in the development and adoption process of such tools. Considerations for ensuring tool performance and avoiding negative side‑effects need to be taken into account already in the initial phases of designing new solutions. For example, the digital solutions should not be treating some population groups unfairly, distort labour market functioning or substantially alter competition between jobseekers for job opportunities (Altmann et al., 2022[52]; Behaghel et al., 2022[49]). Before launching, the tool needs to undergo thorough testing and piloting. In addition to testing and piloting action undertaken by KEIS now, the new digital solutions (beta versions) should be piloted using randomised controlled trials whenever feasible and thereafter rigorously evaluated. As only the careful designing is not guaranteeing performance and avoiding negative side‑effects, the evaluations should aim to look at different key impacts that such tool can cause. For example, career recommender tools like Job Care need to lead to better labour market outcomes of jobseekers (e.g. in terms of employment probability, employment stability, job quality, wage, career progression etc.) while avoiding locking any groups into low-quality jobs (e.g. women or migrants if recommendations are based on historic labour market data only) or directing jobseekers to occupations where competition for jobs is already high. In addition, such evaluations need to look at the effects on sub-groups and potentially other break-downs to ensure that the new solutions work for all those that the group is targeting. In addition, further CIEs using quasi‑experimental design can be relevant after the full roll-out of the solution as the effects can differ in large scale. CIEs need to be accompanied by process evaluations to further pinpoint needs for fine‑tuning the digital solutions.
5.6. Conclusion
Digitalisation is an integral part of strategies for employment services in Korea. The high-level vision of digital employment services is addressed by MOEL and finely detailed by KEIS several years ahead, aiming at a continuous improvement of digital services similarly to public sector in Korea more generally. Going ahead, the authorities could more clearly establish long-term principles and frameworks (for example on managing risks associated with AI and evaluating digital tools) and medium-term action plans to implement digital transition.
Enabled by comprehensive strategic approach to digitalisation and sufficient resources for key needs, digital channels for employment services have become the core of PES in Korea, and all the key services for jobseekers and employers are now available digitally. To continue this successful pathway, MOEL and KEIS need to carefully design the concept of channel management and support the accessibility of digital services, to ensure that jobseekers with lower digital skills or other constraints to using digital channels would not be left behind.
Great progress has been made in adopting innovative tools using state‑of-the‑art technologies in the Korean PES and a continuous improvement process is envisioned. In addition to better tailoring and integrating support in the PES digital platform already being implemented, additional improvements in the platform and tools could be considered learning from other OECD countries. These concern for example strengthening profiling jobseeker competencies and interests for better career management, facilitating online training to further support the one‑stop shop for digital support to jobseekers and assisting employers in designing job vacancies to attract the right talent. Korea could also consider augmenting the coverage of vacancies in the PES portal by web scraping and predicting hiring needs, and enhancing user experience and interactivity of PES platform by harnessing generalised AI. In addition, Korea needs to invest in systematically evaluating the effectiveness and cost-effectiveness of its digital solutions to generate evidence for fine‑tuning and further continuous improvement.
References
[52] Altmann, S. et al. (2022), “The Direct and Indirect Effects of Online Job Search Advice”, IZA Discussion Papers 15830, Institute of Labor Economics (IZA).
[33] Baer, I. (2023), Algorithms to facilitate job matching in Luxembourg, Digital tools for matching jobseekers and vacancies: international workshop, http://oe.cd/5ny.
[41] Banca, M. and J. Plard (2023), The Pix Project, https://www.oecd.org/els/emp/Digital_tools_PIX_07032023.pdf.
[53] Barbanchon, T., L. Hensvik and R. Rathelot (2023), “How can AI improve search and matching? Evidence from 59 million personalized job recommendations”, Working paper, http://rolandrathelot.com/wp-content/uploads/LeBarbanchon_Hensvik_Rathelot_Jobbtips_20230722.pdf.
[49] Behaghel, L. et al. (2022), “Encouraging and Directing Job Search: Direct and Spillover Effects in a Large Scale Experiment”, Banque de France Working Papers, https://publications.banque-france.fr/sites/default/files/medias/documents/wp900.pdf.
[20] Behncke, S., M. Frölich and M. Lechner (2010), “A Caseworker Like Me – Does The Similarity Between The Unemployed and Their Caseworkers Increase Job Placements?”, The Economic Journal, Vol. 120/549, pp. 1430-1459, https://doi.org/10.1111/j.1468-0297.2010.02382.x.
[18] Bolhaar, J., N. Ketel and B. van der Klaauw (2020), “Caseworker’s discretion and the effectiveness of welfare-to-work programs”, Journal of Public Economics, Vol. 183, p. 104080, https://doi.org/10.1016/j.jpubeco.2019.104080.
[2] Brioscu, A. et al. (forthcoming), New Dawn for Public Employment Services: Service delivery in the age of artificial intelligence, OECD Publishing, Paris, https://doi.org/10.1787/dee339a8-en.
[11] Bundesagentur für Arbeit (2022), Möglichkeiten testen [test possibilities], http://web.arbeitsagentur.de/sete/testuebersicht/ (accessed on 26 February 2023).
[48] Busson, A. (2018), Working with different employer groups.
[16] Card, D., J. Kluve and A. Weber (2018), “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations”, Journal of the European Economic Association, Vol. 16/3, pp. 894–931, https://doi.org/10.1093/jeea/jvx028.
[43] Carranza, E. et al. (2022), “Job Search and Hiring with Limited Information about Workseekers’ Skills”, American Economic Review, https://doi.org/10.1257/aer.20200961.
[35] De Blauwe, M. (2021), Oriënt: Use AI to serve business needs, VDAB.
[19] Huber, M., M. Lechner and G. Mellace (2017), “Why Do Tougher Caseworkers Increase Employment? The Role of Program Assignment as a Causal Mechanism”, Review of Economics and Statistics, Vol. 99/1, pp. 180-183, https://doi.org/10.1162/rest_a_00632.
[3] KEIS (2022), 차세대 디지털 고용서비스 플랫폼중장기 발전 방안 (“Mid- to long-term development plan for the next-generation digital employment service platform”).
[32] Kim, H. (2023), AI & Public Employment Services in KOREA.
[9] Leinuste, K. (2021), Decision support tool „OTT“ for employment counsellors in the Estonian PES, https://www.oecd.org/els/emp/PES-Digital-Oct2021-Estonia.pdf.
[31] Mogollon, M. (2021), Pôle emploi incubator methodology, https://www.oecd.org/els/emp/PES-Digital-Oct2021-France.pdf.
[44] Nordic Institute for Interoperability Solutions (2023), X-Road, https://x-road.global/the-17-goals/.
[21] O’Leary, C. et al. (2021), “Effects of Job Referrals on Labor Market Outcomes in Brazil”, Economia LACEA Journal, Vol. 21/2, https://doi.org/10.31389/eco.228.
[1] OECD (2024), 2023 OECD Digital Government Index, OECD Publishing, Paris, https://doi.org/10.1787/1a89ed5e-en.
[28] OECD (2024), Impact Evaluation of Training and Wage Subsidies for the Unemployed in Greece, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/4b908517-en.
[8] OECD (2024), Modernising Latvia’s Public Employment Service through Digitalisation, OECD Publishing, Paris.
[30] OECD (2023), Addressing the legal and IT challenges of data exchange to support contracted-out employment services in Sweden, https://www.oecd.org/els/emp/Sweden_Challenges_Data_Exchange.pdf.
[7] OECD (2023), Evaluation of Active Labour Market Policies in Finland, OECD Publishing, Paris, https://doi.org/10.1787/115b186e-en.
[37] OECD (2023), OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market, OECD Publishing, Paris, https://doi.org/10.1787/08785bba-en.
[46] OECD (2022), Harnessing digitalisation in Public Employment Services to connect people with jobs, OECD Publishing, Paris, https://www.oecd.org/els/emp/Harnessing_digitalisation_in_Public_Employment_Services_to_connect_people_with_jobs.pdf.
[24] OECD (2022), Impact evaluation of the digital tool for employment counsellors in Spain: SEND@, OECD Publishing, Paris, https://www.oecd.org/els/emp/FinalReport-EvaluationOfSEND.pdf.
[27] OECD (2021), Improving the Provision of Active Labour Market Policies in Estonia, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/31f72c5b-en.
[13] OECD (2021), New ways to reach out to employers, OECD, http://oecd.org/els/emp/Employers.pdf.
[26] OECD (2021), OECD Employment Outlook 2021: Navigating the COVID-19 Crisis and Recovery, OECD Publishing, https://doi.org/10.1787/5a700c4b-en.
[10] OECD (2021), “Scaling up policies that connect people with jobs in the recovery from COVID-19”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/a91d2087-en.
[40] OECD (2020), Statistical profiling tools for effective targeting and tailoring of services to jobseekers in Slovenia – Workshop Summary, https://www.oecd.org/els/emp/Statistical_profiling_Slovenia.pdf.
[42] OECD (2020), “The potential of online learning for adults: Early lessons from the COVID-19 crisis”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/ee040002-en.
[39] OECD (2018), Profiling tools for early identification of jobseekers who need extra support, Policy Brief on Activation Policies, OECD Publishing, Paris, http://oecd.org/els/emp/OECD-Profiling-policy-brief-DEC-18.pdf.
[38] OECD (forthcoming), Handbook on labour market transformations, Edward Elgar.
[29] Parliament of Australia (2023), Rebuilding Employment Services. Final report on Workforce Australia Employment Services, https://www.aph.gov.au/Parliamentary_Business/Committees/House/Workforce_Australia_Employment_Services/WorkforceAustralia/Report.
[12] Pôle Emploi (2023), Action recrut’ : 3 ans après sa mise en place, le dispositif permet aux entreprises bénéficiaires de diminuer d’environ 15% les abandons de recrutement et de réduire leurs délais de recrutement d’environ 2 jours, https://www.francetravail.org/accueil/communiques/action-recrut--3-ans-apres-sa-mise-en-place-le-dispositif-permet-aux-entreprises-beneficiaires-de-diminuer-denviron-15-les-aband.html?type=article (accessed on 24 January 2024).
[4] Pôle Emploi (2022), Pôle emploi se dote d’une charte pour une utilisation éthique de l’intelligence artificielle, Communiqué de presse, https://www.pole-emploi.org/accueil/communiques/pole-emploi-se-dote-dune-charte-pour-une-utilisation-ethique-de-lintelligence-artificielle.html?type=article.
[45] PricewaterhouseCoopers (2019), Estonia – the Digital Republic Secured by Blockchain, https://www.pwc.com/gx/en/services/legal/tech/assets/estonia-the-digital-republic-secured-by-blockchain.pdf.
[47] Printsmann, P. (2023), Automated tool for impact evaluation of active labour market measures: linking administrative data., https://www.oecd.org/els/emp/ALMP_impact_evaluation_EC-OECD_workshop_9Oct.pdf.
[36] Radix (2020), Radix and VDAB reinvent orientation test with Artificial Intelligence, Radix, http://radix.ai/cases/flemish-orientation-test-faster-and-more-accessible-than-ever-thanks-to-collaboration-between-radix-and-vdab/ (accessed on 26 February 2023).
[50] Räisänen, H. (2023), How to approach the ex ante employment effects of the Job Market matching platform?, https://www.oecd.org/els/emp/Session3_Job_Market_Matching_Platform.pdf.
[51] Räisänen, H. (2022), “Miten Työmarkkinatorin työllisyysvaikutusta voidaan arvioida?”, Työpoliittinen aikakauskirja, Vol. 1, https://julkaisut.valtioneuvosto.fi/bitstream/handle/10024/163904/TAK_1_2022_web.pdf?sequence=1&isAllowed=y.
[25] Rosholm, M., K. Sørensen and L. Skipper (2019), What Affects Job Prospects? The Employability Indicator Project, Væksthusets Forskningscenter, http://www.vaeksthusets-forskningscenter.dk.
[6] Scheerlinck, K. (2023), Ethical Board of VDAB and its role in AI Governance, https://www.oecd-events.org/ai-wips-2023/session/8d837eb4-cbbc-ed11-9f73-6045bd8890e4.
[5] Scheerlinck, K. (2020), AI & Advanced Analytics @ VDAB, https://www.oecd.org/els/emp/AI_presentation_VDAB.pdf.
[17] Schiprowski, A. (2020), “The Role of Caseworkers in Unemployment Insurance: Evidence from Unplanned Absences”, Journal of Labor Economics, Vol. 38/4, pp. 1189-1225, https://doi.org/10.1086/706092.
[23] Schmidt, T. and T. Mitze (2023), “Crisis and the welfare state: the role of public employment services for job placement and the Danish flexicurity system during COVID-19”, Cambridge Journal of Regions, Economy and Society, Vol. 16/1, https://doi.org/10.1093/cjres/rsad002.
[15] Suárez, P., B. Cueto and M. Mayor (2014), “Effects of public employment services on labor transitions: an analysis for the Spanish case”, International Journal of Manpower, Vol. 35/7, https://doi.org/10.1108/IJM-05-2014-0109.
[34] VDAB (2023), Orient, http://orientatie.vdab.be/ (accessed on 26 February 2023).
[22] Vervliet, T. and A. Heyma (2022), Langetermijn effecten van persoonlijke dienstverlening in WW [Long-term effects of personal services in WW], SEO Economisch Onderzoek, https://www.seo.nl/wp-content/uploads/2022/09/2022-57-rapport-Experiment-WW-deel-3-20220516-DEF.pdf.
[14] Vikström, J., M. Söderström and A. Cederlöf (2021), “What makes a good caseworker?”, IFAU Working papers.
Notes
← 1. OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023; the share of countries calculated based on those countries that responded to the question.
← 2. OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023; the share of countries calculated based on those countries that responded to the question.
← 3. OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023; the share of countries calculated based on those countries that responded to the question.
← 4. OECD questionnaire on digitalisation in Public Employment Services to support the provision of active labour market policies launched in March 2023; the share of countries calculated based on those countries that responded to the question.
← 5. The support was provided within the framework of the project “Implementing a new approach to the management of statistical and analytical information in the Spanish labour and social security administration” in 2020‑22. This project was funded by the European Union via the Technical Support Instrument, and implemented by the OECD, in co-operation with the Directorate-General for Structural Reform Support of the European Commission.