Simone Romano
OECD
Cyrille Schwellnus
OECD
Simone Romano
OECD
Cyrille Schwellnus
OECD
Luxembourg’s real GDP growth over the past 15 years has largely been driven by the increase in the workforce, especially in finance and business services, while labour productivity has stagnated. To sustainably raise living standards and ease pressures on infrastructure, housing and greenhouse gas emissions, Luxembourg needs to transition to a more broad-based growth model based on innovation and productivity growth. This requires reforms to improve the system of public innovation support, boost workforce skills and strengthen competition. Consolidating coordination of innovation support, making support more mission-oriented and further fostering public-private partnerships would help raise business R&D expenditure from very low levels. Workforce skills would be boosted by strengthening quality standards for training providers and raising the attractiveness of Luxembourg for highly-qualified non-EU nationals. Reducing entry barriers, especially in the services sector, and strengthening disclosure requirements for lobbyists would help to ensure that innovative start-ups can compete with incumbent businesses on an even footing.
Over the past 15 years, Luxembourg’s real GDP growth has been driven by finance and business services, drawing on cross-border workers and immigrants, but the country now faces the limits of this model. Labour productivity growth of the business sector, which is essential to sustainably raise living standards, has stagnated since 2010. The strategy of supporting growth mainly through a rapid increase in employment puts pressure on the housing market, creates strains on infrastructure and GHG emissions from road transport. At the same time, Luxembourg’s financial sector – long a cornerstone of its economy – faces strong competition and its growth in terms of employment may be impacted by increased use of artificial intelligence and other trends, slowing both direct and indirect job creation.
This chapter discusses how Luxembourg could navigate these challenges and sustain its economic dynamism in the long term. Shifting to a more broad-based growth model rooted in innovation and productivity gains requires reforms in the areas of skills and innovation policies, as well as revising regulations that unduly reduce competition. Section 4.1 provides a brief overview of productivity developments over the past 15 years; section 4.2 analyses policies to strengthen innovation, especially in the business sector; section 4.3 discusses policies to address the issue of skills shortages through developing, attracting and retaining the skilled workforce needed to thrive amid rapid technological change; section 4.4 reviews avenues to reduce barriers to competition.
Measured labour productivity in Luxembourg is among the highest in the OECD (Figure 4.1, Panel A). A highly educated workforce; a favourable business environment in terms of regulation, taxes and infrastructure; the strategic location in the centre of Europe; as well as economic openness and political stability contributed to high productivity growth until the mid-1990s. However, labour productivity growth since 1995 has been among the lowest in the OECD (Figure 4.1, Panel B), suggesting that the pre-1990s growth model may no longer be able to deliver the rates of productivity growth experienced in the past. Accounting for issues with the measurement of labour productivity in the financial, real estate and non-profit sectors of the economy does not significantly change this assessment (Box 4.1).
Note: The bars refer to total-economy labour productivity. Labour productivity growth in Panel B is measured as the growth rate of GDP per hour worked. In Panel B, LUX refers to total-economy labour productivity growth, while LUX* refers to Non agriculture business economy excluding real estate and financial services.
Source: OECD National accounts database.
Total economy labour productivity is a standard internationally comparable measure of productivity based on the national accounts, but the unusual structure of Luxembourg’s economy with an unusually high share of the financial sector makes it more difficult to interpret here. The OECD Productivity Database focusses on business sector productivity (excluding real estate), which reduces the reliance on imputed measures. However, financial sector value-added is also partly imputed. From a sectoral perspective, the key issues are the following:
Financial sector: In the system of national accounts, financial services output is measured as the sum of financial intermediation services indirectly measured (FISIM) and fees (for instance on account keeping, credit cards, brokerage, financial advice and asset management). While fees are directly observed, FISIM is estimated by first imputing a reference interest rate and then computing net interest income. The choice of the imputed reference rate can have a significant impact on FISIM. Moreover, estimating FISIM in volume terms – which is the relevant measure for productivity analysis – requires the estimation of prices that are not directly observed to deflate loans and deposits. Further issues include the exclusion of trading profits (for instance capital gains) and other interest (for instance on bonds and derivative products) from the measure of financial sector output in the national accounts.
Real estate sector: Apart from the revenue of real estate agencies, the output of the real estate sector in the system of national account includes rents of tenants and imputed rents of owner-occupiers, whereas the labour input of owners is excluded. This results in very high levels of labour productivity in the real estate sector, since the only input accounted for in value added calculations is the labour input of real estate agents. Fluctuations in rents or imputed rents, for instance due to changes in house prices, can lead to large fluctuations in labour.
Non-profit sector: The output of the non-profit sector (mainly the public administration, education, health and activities of households as employers) is imputed based on the wage bill. This implies that, all else equal, increases in real wages are recorded as labour productivity gains.
Excluding these sectors from the calculation of labour productivity does not fundamentally change the result that productivity growth in Luxembourg has been among the lowest in the OECD over the past 15 years (Figure 4.1, Panel B). Given that differences in productivity growth between the total economy and the non-agricultural and non-financial business sector measures are small and a number of disaggregated productivity measures are available at the total economy level only – for instance regional disaggregations – the remainder of the chapter focuses on the total economy measure.
The slowdown in labour productivity growth has been driven by both declining multi-factor productivity and slowing capital deepening. The decline in multi-factor productivity – which encompasses skills and technological progress – has been particularly striking. While multi-factor productivity growth has slowed across a broad range of OECD countries over the past three decades, the growth rate in the average OECD country remained positive over 2010-22 (Figure 4.2). By contrast, in Luxembourg, multi-factor productivity growth has turned negative, declining by 0.3% annually on average between 2010 and 2022. Even though improvements in skills and technological progress become increasingly difficult to achieve when productivity is already high and economies are close to the technological frontier, declines in multi-factor productivity over several years are unusual. In fact, apart from Greece and Mexico, Luxembourg is the only OECD economy to report negative multifactor productivity growth over 2010-22. The slowdown in capital deepening has been somewhat less pronounced, but the contribution of capital deepening, especially ICT capital deepening, over 2010-22 is nonetheless lower than in other euro area economies.
Note: This refers the non-agricultural business economy excluding real estate and financial services.
Source: OECD Productivity Database.
While Luxembourg’s high level of productivity makes any additional productivity gains more difficult to achieve, its productivity growth performance has been poorer than many similar high-income economies. Economies far below the technological frontier typically experience higher productivity growth than those at or close to the frontier (Ravikumar et al., 2024), as they can raise productivity by adopting technologies developed abroad, whereas economies close or at the technological frontier need to develop new technologies, requiring larger investments in research and development and a more highly-skilled workforce (Aghion et al., 2023).
Given Luxembourg’s relatively small size, comparing its productivity growth performance to other regions with similar initial labour productivity appears to be more informative than comparing it to other countries. This comparison suggests that Luxembourg has underperformed other OECD regions with similar levels of initial labour productivity, which have in most cases experienced significantly higher labour productivity growth (Figure 4.3). This includes a number of high-income US states, such as California, Massachusetts and New York, that may benefit from larger economies of scale than Luxembourg due to their large size, and also a number of European regions, such as the Belgian capital region (“Brussels-Capital”), the French capital region (“Ile-de-France”) or the Irish Eastern and Midland regions. Despite being significantly larger than Luxembourg in terms of population and GDP being close to or at the technological frontier, these regions managed to reach productivity growth rates that are typical of regions at much lower levels of productivity and well above Luxembourg’s.
GDP per worker, USD ‘000 PPP, NUTS2 regions
Note: selected regions labels are AUS-WAU Western Australia, BEL-BRU Brussels-Capital Region, FRA-IDF Île-de-France, IRL-EAM Eastern and Midland, NZ-TKI Taranaki, USA-DC District of Columbia, USA-NY New York, USA-CA California, USA-WY Wyoming, USA-NJ New Jersey, USA-MA Massachusetts, USA-WA Washington, USA-MD Maryland, USA-LA Louisiana.
Source: OECD Regional Database.
Low productivity growth mainly reflects low within-industry growth rather than changes in industry specialisation. For a high-income economy such as Luxembourg, increasing specialisation in services could act as a brake on productivity growth as labour is reallocated from high-productivity manufacturing to lower-productivity services, such as hotels and restaurants and the public sector, due to increasing demand as incomes rise and low potential for productivity growth in these sectors. However, the reallocation component of productivity growth in Luxembourg over 2010-2022 is similar to other OECD countries (Figure 4.4). Reallocation in Luxembourg is mainly explained by increasing employment shares of high-productivity ICT and professional services. By contrast, the average within-industry growth rate – broadly speaking, the average productivity growth rate of productivity across industries weighted by industry value added – has been well below the OECD average.
Shift share decomposition, 2010-2023
Note: “Within industry” denotes contributions from productivity growth within industries, while “overall reallocation” denotes contributions from labour reallocation across industries with different productivity levels and growth rates. 2010-2022 for United Kingdom, Lithuania, Norway, Portugal, Sweden and EU27.
Source: OECD National Accounts database and desk calculations.
To some extent, the negative within-industry contribution to overall labour productivity growth reflects the structure of Luxembourg’s industry specialisation. All else equal, in countries that have relatively large activities with low productivity growth, such as construction or hotels and restaurants, the within-industry component of productivity growth tends to be lower than in countries specialised in industries with typically high productivity growth, such as manufacturing or ICT services. In Luxembourg, the share of manufacturing in total value added was around 5% on average over the period 2010-22 as compared to 17% on average in the rest of the European Union, partly accounting for the shortfall in productivity growth over the period. If Luxembourg had had the same pattern of industry specialisation as the average EU country, annual within-industry productivity growth over 2010-2022 would have been about 0.3 percentage points higher, resulting in a 4% higher level of productivity in 2022 (Figure 4.5). However, even at the same pattern of the average EU country, Luxembourg’s within-industry productivity growth would have been among the lowest in the EU.
Within-industry contribution to labour productivity growth
Low within-industry productivity growth relative to the EU average over the period 2010-22 is mainly explained by low productivity growth in non-financial services and construction (Figure 4.6). Measured productivity growth has been particularly weak relative to EU peers in accommodation and food and in wholesale and retail trade, but also in information and communication services. By contrast, productivity growth in the manufacturing sector has been in line with EU peers, while in transportation and storage it has outperformed EU averages.
Labour productivity, annual growth rates, 2010-2023 average
Note: 2010-2022 for EU27, pending the 2023 release.
Source: OECD National Accounts database and desk calculations.
Innovation and technology adoption are key drivers of productivity growth. Innovation drives the development of new products, services and business models in technologically-leading businesses, which is especially important in high-income economies such as Luxembourg. The adoption of technologies and business models developed elsewhere helps lagging businesses catch up with the technological leaders and might gain even further importance with the spread of artificial intelligence (AI) (OECD et APO, 2022).
Luxembourg’s gross domestic expenditure on R&D, which is a key input into innovation, is among the lowest in the OECD as a share of GDP (Figure 4.7). While public R&D expenditure is in line with other OECD countries, business R&D expenditure is only about one-quarter of that in top performing countries such as Sweden or Japan. To some extent, this is explained by Luxembourg’s specialisation on financial services, which typically have a very low measured R&D intensity. The absence of tax credits for R&D may also play a role, as it weakens incentives to report R&D expenditure (STATEC, 2021). However, business R&D spending is substantially lower than in top-performing countries across a broad range of sectors, suggesting that low business R&D spending does not solely reflect the unusually large size of the financial sector.
Gross expenditures on R&D spending components, 2022
Investment in intangible assets other than R&D is relatively high in international comparison, with new financial products being the dominant type of intangible investment in Luxembourg, in line with the structure of the economy (Conseil National de la Productivité, 2023). While Luxembourg has also witnessed rapid growth in investment in organisational capital and branding, the gap with innovation-led economies remains substantial in some other intangible assets. Investment in ICT equipment is just above the EU average, and the gap with top performing countries is even more pronounced for investment in intellectual property.
Innovation outputs, such as patent applications, lag top-performing countries. Scientific publications have witnessed a significant increase since 2003, when the University of Luxembourg was created. However, they still remain well below the rate of the Nordic countries, the Netherlands and Switzerland (Figure 4.8). Both scientific publications and patent applications are more in line with top-performing countries if expressed as a share of population rather than as a share of total employment, but given that both resident workers and cross-border workers contribute to innovation, normalising scientific publications and patent applications by employment appears more appropriate.
Note: Panel A shows the number of scientific and engineering articles published in scientific journals in the following fields: physics, biology, chemistry, mathematics, clinical medicine, biomedical research. The total number is divided by the workforce in millions. Panel B shows the total number of patent applications lodged by physical persons or legal entities resident in Luxembourg (legal entities resident in Luxembourg include cross-border workers who work for them) divided by workforce in millions.
Source: World Bank.
The share of enterprises that have internally developed product or business innovations is low across a broad range of sectors. Less than 50% of firms operating in the financial sector, which plays a key role in the Luxembourg’s economy, have developed in-house innovations, while the share in top performing countries such as the Nordic countries and Germany is around 70% (Figure 4.9, Panel A). Luxembourg also performs worse than EA average in the industry sector (Figure 4.9, Panel B), with specific examples ranging from the manufacture of metal products, electrical equipment and computers to chemical products.
2020
Adoption of digital technologies is behind EU peers. The EU has set the goal for its “Digital decade” that 75% of all firms could use cloud computing and big data analysis by 2030. While Nordic countries and the Netherlands are already close to the achievement of the objective with regards to cloud computing, Luxembourg hardly reaches 50% of its completion (Figure 4.10, Panel A) and the scenario for the adoption of data analysis is similar. Beyond EU’s “Digital decade”, the lag with the top performing countries and the EU average is confirmed in the use of the internet of things, but also on more established technologies, with the share of firms selling their products online being one of the lowest in Europe (Figure 4.10, Panel B). In terms of the share of enterprises using AI the gap with top performing economies (such as Denmark and Finland) is narrower, but less than 20% of Luxembourg firms are using it, leaving space to better integrate this innovation in productive processes (European Commission, 2024).
2023
Note: Panel A show the country performance in terms of the achievement of the EU’s “Digital Decade” targets regarding the adoption of specific technologies, such as cloud in this case. Panel B show the share of enterprises with 10 or more employees selling online.
Source: EU’s Digital Decade 2024 Report for Panel A, Eurostat for Panel B.
Government financial support is important for sustaining innovation. Public support is aimed at crowding in business investment in R&D and can be direct, with government and public agencies directly funding research projects or institutions. Indirect support, such as tax allowances, exemptions and deductions, effectively subtract R&D investment from the tax base, while tax credits subtract it from tax liabilities. The vast majority of OECD member countries use both forms of support, while Luxembourg relies almost exclusively on direct funding, using only the Intellectual Property (IP) Box as R&D tax incentives (Figure 4.11).
Direct government funding and government tax support for business R&D, 2021
Direct forms of support to R&D and innovation in Luxembourg are provided by several different institutions. The Ministry of the Economy mainly supports businesses’ innovation, offering co-funding for R&D projects, innovation aid for SMEs, and support for young innovative enterprises. The grants can cover up to 25-80% of project costs (Luxembourg, 2024f). The Luxembourg National Research Fund (FNR) is the main funder of public research activities, with funding commitments that can reach up to an overall ceiling of EUR 445 million over the years 2022-2025. The FNR invests public funds and private donations into research projects through grants and other forms of direct funding (FNR, 2024). The “Société Nationale de Crédit et d'Investissement”, a Luxembourg public-sector banking institution that supports companies in their innovation projects, provides innovation loans at a lower-than-market and fixed interest rates (SNCI, 2024). It has also created, together with the European Investment Fund (EIF), the Luxembourg Future Fund 1, which is a EUR 150 million fund that aims to attract venture capital investment to foster the sustainable development of SMEs in Luxembourg’s strategic sectors, such as fintech or cybersecurity (EIF, 2023).
Streamlining the public support for R&D and innovation and establishing a mechanism for coordination of the main actors involved in providing the support would help to improve its effectiveness and link to business. First, having a mechanism that coordinates the public support, for instance an innovation council, would allow to take advantage of synergies between different actors and projects, ensuring that both public and private research efforts are aligned with common national priorities. Second, it would allow to harmonise application procedures and selection criteria, which would increase the capacity to crowd in more business investment, as piecemeal approaches to R&D and innovation support discourage firms from applying, especially smaller enterprises because of the higher costs incurred in following the different procedures (OECD, 2000).
The UK’s effort to better coordinate and streamline its research and innovation provides an example of good practice. In 2018 the UK launched UK Research and Innovation (UKRI), a non-departmental public body that brought together the seven Research Councils, Research England, which was responsible for providing funding to higher education institutions in England, and also the business-oriented innovation agency “Innovate UK”. The idea was to enable a greater focus on cross-cutting issues and improve collaboration between the research base, enhancing the effectiveness of the different research efforts and their impacts. The first years of activity seem to confirm that the creation of UKRI has allowed to tackle cross-cutting problems more effectively than if the nine constituents would have continued to work individually in silos (UK, 2022).
Furthermore, public support could be made more mission-oriented by basing the allocation of fund more on projects rather than institutions, while keeping some continuity of research with a part of recurrent funding. Currently, the government allocates the major part of its R&D expenditure to research institutions, such as the University of Luxembourg, and only a minor part to specific projects, which limits support to private business-led R&D. In the mission-oriented approach, supported projects pursue objectives that are in line with broader and clearly defined national interests, and they typically involve different actors and span different fields. However, policy mission orientation presents some risks, as the selection of goals and the identification of niches of activity is difficult. Moreover, it is also important to clearly determine the timeframe for the support and to put in place an effective system of monitoring and evaluation with the aim to ensure the effectiveness of the support.
Public-private partnerships should be further encouraged by making them a requirement to access public support for public applicants. This would allow to both crowd in more business investment and reinforce collaboration. Furthermore, a part of public support should be made conditional on businesses matching public grants through private R&D and innovation investment. Similar recommendations were made by the 2022 OECD Economic Survey of Luxembourg, but no substantial policy action has been taken to address them so far (Table 4.1)(OECD, 2022d).
RECOMMENDATION |
ACTION TAKEN |
---|---|
Subsidise active on-the-job training schemes targeted to the over-45- year-olds. Expand access to training to help early-school leavers enter the workforce, alongside school system reforms. |
No action taken. Compulsory schooling was extended from 16 to 18 years in 2023, with the measure coming into force in 2026. |
Increase public spending on R&D to match private R&D funding and encourage greater investment by firms. Increase funding to targeted projects by reducing the funds spent on administration. |
No action taken. No action taken. |
Reduce administrative burdens on small firms, notably by streamlining procedures for starting a business |
No action taken |
Source: OECD Economic Survey of Luxembourg (OECD, 2022d)
A more tailored and straightforward approach to R&D and innovation supports could encourage small enterprises to investment more in innovation, technology adoption and overall productivity growth. Simplicity of the tools used to support innovation in SMEs is key for their success. Until recently the application process and supporting documents to be provided to apply for Luxembourg’s financial aid for SME innovation were the same as those for larger enterprises, discouraging small firms (Luxembourg, 2024g). The recent simplification of the participation procedures of the “Fit4” programmes, however, goes in the right direction. The “Fit4” programmes offer to reimburse SMEs the cost of an assessment of needs in terms of innovation, digitalisation and sustainability by an external consultant and the possibility to apply for public co-funding of the investments suggested by the consultant.
Austria and Germany’s innovation vouchers offer similar services to SMEs and their experience strongly confirms the key importance for this kind of tools of being simple and agile, involving low administrative burden (Interreg Europe, 2021). These innovation vouchers allocate funding to enterprises to buy innovation services from knowledge providers, or to recruit an in-house innovation resource or innovation manager, facilitating SMEs’ access to external knowledge. Austria’s evaluation of this tool showed that they have been effectively used by SMEs to address technological development and innovation in a broader sense: 25% of firms that used the voucher then conducted follow-up projects and many participating firms created lasting networks with research organisations, producing as a final result a significant change in innovation behaviour as well as increased R&D expenditures (European Commission, 2021).
Beyond increasing the effectiveness of the existing direct support, Luxembourg could consider introducing more forms of indirect support measures to increase overall business R&D investment (Box 4.2). Public support for business R&D usually requires a mix of direct instruments and market-based incentives, as no single mechanism is able to provide a full range of incentives (OECD, 2023g). While market-based mechanisms such as tax credits tend to increase business R&D at the margin and for innovative projects that are closer to the market, more direct forms of support tend to stimulate research that may not immediately result in new goods or service (OECD, 2020c) (Appelt et al. 2016). Empirical evidence suggests that both direct funding and market-based incentives are effective in stimulating private business R&D investment, with a gross incrementality ratio that is similar for both of them, being of around 1.4, meaning one extra unit of R&D tax support translates into 1.4 extra units of R&D (OECD, 2023g).
Iceland provides a good practice to follow, as it has introduced tax incentives to boost business R&D in 2010 and these have proved to be effective (OECD, 2023d). Iceland’s R&D tax incentive is a refundable volume-based scheme that allows companies to deduct part of the R&D related costs from the corporate income tax. Over the years, it has demonstrated to have a positive impact on total business R&D spending – especially for micro firms – but also on annual firm sales, employment, and average wages. The 2016 reform, which raised the ceiling on total eligible cost and represented a further cost for public finances, has also led to positive results, being effective in increasing R&D spending, especially among small and medium enterprises (OECD, 2023d).
Since the last reform, Iceland’s tax credit rates are 35% for SMEs and 25% for large firms. The credit is refundable, meaning that if a company with an approved R&D project has insufficient taxable profits against which it could deduct the R&D costs, the deduction is converted into a refundable amount and paid out to the company in the following fiscal year. To qualify for the scheme, companies are required to spend a minimum of around EUR 7000 per year per R&D project and need to obtain a certification of validity by the Icelandic Centre for Research. The Centre assesses the R&D content of projects and makes sure that the business plan is well defined and that the staff have the training, education and experience in the area of the proposed project (OECD, 2023d).
The development of successful innovation ecosystems helps strengthen overall innovation. These are networks of firms, higher education and research institutions, government agencies and innovation support organisations, investors and researchers that contribute their expertise, human and financial resources to develop new technologies, products, processes or services that address shared specific goals (OECD, 2023f). Collaborations are often coordinated and funded by government agencies. Successful innovation ecosystems are characterised by cross-pollination of ideas across different fields and, as specialised industrial clusters, attract skilled labour force, specialised business support and venture capital and investors. The interactions and knowledge flows within the ecosystem accelerate the innovation process compared to organisations working in isolation, and the large externalities generated by innovation collaboration justify public support for innovation ecosystems. Luxembourg has already and successfully fostered innovation ecosystems in some specific niches (Box 4.3) and should continue to pursue this approach.
The cleantech cluster aims at fostering innovation, business development and cross-sector cooperation in the field of clean technologies. The cluster supports it members in the generation of new processes, products and services with an ultimate goal of turning the concept of the circular economy into a reality in Luxembourg. Luxinnovation supports the cluster by offering tailored advice on national and European funding opportunities, easy access to key contacts and participation in matchmaking sessions (Luxinnovations, 2024b).
Luxinnovation fosters the development of the healthtech cluster as well, which aims to support the development of sustainable and trusted services and products for the European healthcare market in fields such as digital health, medical devices and software, in vitro diagnostic and neurology. This is meant to favour positive spillovers to other areas where Luxembourg has considerable expertise such as telecommunication (for telemedicine and mobile health), cybersecurity, and artificial intelligence.
The support offered by Luxinnovation is similar in nature to that offered to the CleanTech cluster, but on top Luxembourg has recently launched the initiative “Dataspace 4 Health”, a platform that provides secure access to health data and respects the EU’s data privacy legislation and patients’ rights, as currently health data is often siloed, fragmented, and hence underutilised, limiting the potential for innovation and research. This initiative aims also at helping to build a connected ecosystem by establishing a secure and efficient layer for data sharing across institutions, fostering collaboration and advancements and leveraging AI to improve diagnosis, treatment and prevention of diseases, enhancing patient outcomes and quality of life, and advancing scientific knowledge and discovery (Luxinnovation, 2024a).
The space industry in Luxembourg has witnessed a constant expansion and consolidation, bringing the sector to represent almost 4% of GDP (among the highest ratios in Europe), and counting over 1000 employees, seven research centres and over 60 specialist companies. The origins of this industrial specialisation go back to 1980s, with the creation of the European Society of Satellites (SES), one of the world leading satellite operators, that has given the advantage of the first mover to the Grand Duchy and has favoured its imposition as the main actor in aerospace for over 30 years.
The public sector has played an important role in supporting and nurturing this industrial niche since the beginning. It was a key initiator in the formation of SES in 1985 and has remained a major shareholder since then. The support of the government was key also to enable the company to become Europe's first private satellite operator with the launch of Astra 1A in 1988. In 2005, Luxembourg’s accession to the European Space Agency (ESA) opened the door for national players to access the space market in Europe. In 2017, the Luxembourgish government introduced a legal framework offering legal security to owners of resources from outer space, with a view to becoming a European hub for the use and exploitation of space resources. In 2018, Luxembourg created its own space agency - the Luxembourg Space Agency (LSA) - in order to remain at the cutting edge of the latest mining operation developments. In 2020, the European Space Resources Innovation Centre (ESRIC) has been established in Luxembourg as part of this initiative to create additional opportunities for European and international innovation.
The successful wide-ranging support also includes positive examples of cross-institution collaboration that might serve as examples of good practices to be replicated. Together with a group of private and public investors, the Luxembourg Government has invested in Orbital Ventures, an investment fund focused on providing equity funding for early-stage space companies with ground-breaking ideas and technologies engaged in space activities. Furthermore, as the space industry needs specialised skills and talent, a two-year Interdisciplinary space master program has been launched by the University of Luxembourg in the fall of 2019 and it has been developed in collaboration with the Luxembourg Space Agency (LSA).
The increasing productivity differential observed in Luxembourg between frontier (80th percentile of most productive businesses) and laggard firms (20th percentile) has played a role in explaining slow productivity growth in the last 20 years (OECD, 2022d). Productivity dispersion per se does not have a negative impact if it is determined by fast-growing front-runners, but in Luxembourg widening dispersion is also due to laggard firm productivity performance being in protracted decline since the great financial crisis. Raising the productivity level of laggard firms by favouring a rapid and efficient innovation diffusion and technology adoption among all firms would help foster total-factor productivity growth.
Starting from 2024, Luxembourg has revised the investment tax credit. The aim is to better support Luxembourg companies in their digital and ecological/energetic transformation. The reform has increased the tax credit rate from 8% to 12% for generic investment, introducing a 18% tax credit for investments and operating expenses connected with digital and ecological transformation (Luxembourg, 2024c). Moreover, the minimum threshold for each investment has been scrapped. The tax credit can be carried forward for 10 years.
The introduction of this new tax credit scheme goes in the right direction, but its design could be streamlined. Companies that want to apply need prior approval from the Ministry of Economy, to be obtained through the eligibility attestation submission form that contains several wide-ranging questions aimed at gathering an extensive and in-depth description of the project. Moreover, companies are required to apply for the annual compliance certificate for each fiscal year during which investments have been carried out and granting of a tax credit is requested, presenting supporting documents detailing investments and operational expenses effectively incurred during the period. While conditionality in the concession of tax credits and surveillance of compliance is important to increase the effectiveness of public financial support, an excessively cumbersome procedure might end up discouraging application, especially for smaller firms.
Denmark’s tax credit scheme to promote investment in R&D (Skattekreditordningen) offers a practical example of lighter administrative requirements. The application does not need prior approval and is filed online to the tax authorities together with the annual tax return for the same year. While this reduces compliance costs, it may increase uncertainty for firms as the reimbursement is received only after undertaking the investment. Spain’s system of innovation incentives might provide a solution to this issue. Similarly to Denmark, applications can be filed online together with the corporate tax declaration procedure, but in Spain firms may apply for pre-validation of the application, which are binding for tax authorities at the moment of actual application (and more than 80% of firms do so). A different solution to ease the complexity of application and compliance procedures and encourage smaller firms to apply is provided by Canada, where different sorts of application assistance is available (first-time advisory service, pre-claim project review, assign a contact person that assists in the process) (European Commission, 2014).
The investment tax credit scheme might also benefit from adding further incentives addressed specifically at SMEs, which usually lag behind in their technology adoption and constitute the bulk of laggard firms (OECD, 2020b). Australia’s “Small business technology investment boost” might provide an example. The program has allowed small businesses (less than AUD 50 million of aggregated annual turnover) to benefit from an additional 20% tax deduction (on top of the normal investment deduction rate) in 2022-2023 to support their digital operations and digitalise their operations, encompassing various aspects of digitalisation, from basic hardware and software to more advanced e-commerce and cybersecurity solutions (Australian Taxation Office, 2023).
Access to finance is key to developing innovative products or business models, adopt innovative technologies or invest in intangible capital. The vast majority of Luxembourg firms – 82% over the period 2017-21 – rely on bank financing for their investment projects (Chambre des metiers, 2024). However, bank financing has limitations for small innovative firms and starts-ups at early stages of development that might face challenges because of their limited collateral and track record. Empirical evidence shows that companies financed by private venture capital investors tend to have a greater increase in innovation output (Amess et al., 2016) and that venture capital has a positive causal impact on firms’ growth (Paglia and Harjoto, 2014). However, venture capital investment as a share of GDP in Luxembourg remains relatively low compared to other OECD countries (Figure 4.12).
2023 or latest year available
Governments can have an active role in encouraging venture capital investment and the participation of individuals in the capital of enterprises (business angels). On the supply side, fiscal provisions can encourage this kind of investment. Such incentives are already present in Luxembourg, but they are weaker than in other OECD countries when it comes to individual investors. The ceiling for the tax deduction for individual investment in enterprises is currently set at EUR 5,000 and there is no further incentive to invest in newly created companies. The low ceiling for the deduction does not spur bigger investment, while the lack of further incentives to invest in start-ups does not balance the lower capacity of newly created companies to remunerate investors in the first years.
Luxembourg should remodel its fiscal incentive schemes for private investment in enterprises and start-ups to be more in line with other OECD and European countries. For example, among the tax relief tools available to individuals in the UK who invest in certain companies, the “Enterprise investment scheme” allows to claim a tax relief of 30% on the total amount investment, of up to GBP 1 million. The ceiling is raised to GBP 2 million if at least one million of that is invested in knowledge-intensive companies (UK, 2023). Belgium’s “Tax shelter pour startups” set the amount of deduction at 30% - 45% of the amount invested, with a EUR 100 000 investment ceiling (OECD, 2024b). On the demand side, enhancing financial literacy among managers, especially those of smaller firms, and changing the attitude towards the fear of losing control to the advantage of external subjects tend to improve firms’ capital structure.
The existence of a labour force that is endowed with the right set of skills is instrumental to technology adoption, innovation and, ultimately, overall productivity growth. However, Luxembourg faces widespread skills shortages, as demand for certain skills increasingly outpaces supply (OECD, 2023b). Skills shortages harm productivity growth by preventing the most productive and innovative businesses from expanding. Moreover, the lack of specific skills in the job market may prevent a broad range of businesses from adopting productivity-enhancing technologies and business models, for instance if shortages of IT technicians prevent businesses from adopting digital technologies.
Luxembourg can continue to tackle the issue of skills shortages by further strengthening its capacity of attracting and retaining foreign talents with relevant skills; but this approach has some limits and needs to be coupled with other policy measures. Upskilling and reskilling of the workforce is key, as well as university reforms to better align tertiary education with labour market needs.
As the economy has expanded, employment in Luxembourg has almost doubled over the past 25 years, reaching more than 515,000 workers at the beginning of 2024 (ADEM, 2024b). The workforce has been increasing mainly due to more cross-border workers: while approximately one-quarter of the workforce consists of residents with Luxembourg nationality and another quarter are immigrants (residents with non-Luxembourg nationality), about one half are cross-border workers. 50% of cross-border workers come from France, with the other 50% being evenly split between Belgium and Germany (Figure 4.13, Panel A).
Changing labour demands over the past two decades have led the labour market to become increasingly polarised. This partially reflects the rapid expansion of high value-added services – such as finance, information technology (IT), education, and scientific and technical activities – and robust growth in lower-skilled services – such as construction, and hotels and restaurants (STATEC, 2020). Consequently, employment has increased significantly at the top end of the skills distribution, with a cumulative growth rate for professionals with a tertiary education that has exceeded 100% over the period 2010-2023 (Figure 4.13, Panel B). Over the same period, employment has increased also for low-skilled workers, as the cumulative growth rate for professionals with less than primary or with lower secondary education has exceeded 20%. By contrast, employment for secondary graduates has contracted.
Luxembourg’s resident population has one of the highest shares of tertiary graduates, with about 50% of residents having a tertiary degree, above the OECD average of 40%. This is the case despite weaknesses in the education system, as demonstrated for example by the highest increase among EU member countries of young people neither in employment nor in education or training (NEET) over the last 10 years, or the below average results obtained when the country took part in the OECD PISA assessment (Eurostat, 2024b).
However, even though the share of tertiary graduates is high, Luxembourg faces a mismatch between the field-of-study and the labour market, with tertiary degrees often obtained in fields that are not those in high demand by employers. For example, while shortages are particularly severe in IT and scientific and technical professions (ADEM, 2024a), the share of 25–64-year-olds with tertiary education who studied in the field of science, technology, engineering and mathematics (STEM) is lower than the OECD average (Figure 4.14, Panel A), and the gap is even larger for female tertiary graduates in these fields (OECD, 2023a). This contributes to the rise in the number of unfilled job vacancies, which has been trending up over the past 15 years. While it has come down from its peak in 2022, it still is on a higher level than the pre-pandemic average despite above-average unemployment, posing challenges for total factor productivity growth (Figure 4.14, Panel B).
Note: The graph in Panel B plots the % ratio between the number of unfilled job vacancies and the number of unemployed people registered at the public employment agency. For each monthly observation the graph shows the moving average over the previous year. Data for unfilled job vacancies refer to the number of vacancies notified to the Labour Administration Offices and remaining unfilled at the end of the month. Firms are required by law to declare vacancies at Labour Administration Offices.
Source: Education at a Glance 2022 (Panel A) and OECD Data Explorer • Infra-annual registered unemployment and job vacancies (Panel B).
The public employment agency (ADEM) periodically publishes a list indicating the jobs that are in high demand, but for which very few candidates (or none) are available among the jobseekers registered at the employment agency (ADEM, 2024a). The relevant occupations include finance, IT, engineering, health and personal care, construction, social action and corporate support services. While most of the professions that are currently in short supply require advanced education degrees, such as industrial IT analyst programmer, neuropsychologist or industrial materials engineer, others require specialised skills but not necessarily advanced education degrees, for instance specialised construction jobs or specialised IT technicians (Table 4.2).
Level of qualification |
Field |
Occupation |
---|---|---|
High |
Finance |
Financial Analysis and Engineering; Credit and Banking Risk Analysis |
Information Technology (IT) |
IT Research and Development; Information Systems (IS) Advisory and Project Management |
|
Engineering |
Management and Production Engineering; Management and Engineering – industrial design and studies, research and development; Management and production Engineering |
|
Business Services |
Legal Defence and Advice; Audit Financial and Accounting Control; Management Audit |
|
Care/Education |
Psychology; Early Childhood Education; General Nursing Care |
|
Middle/Low |
Finance |
Bank Customer Management |
Information Technology (IT) |
Information Systems Technical Support |
|
Engineering |
Installation and Maintenance of Industrial and Operating Equipment |
|
Business Services |
Bookkeeping |
|
Care/Education |
Patient Hygiene and Comfort Care Services |
|
Construction |
Roofing Installation and Restoration |
Note: The list contains almost all occupations listed in the original ADEM list but does not contain the specific job titles.
Source: ADEM
The attraction and retention of foreign talent has represented a key and effective way to meet labour and skill demand in Luxembourg. Over the past decades Luxembourg has performed well in attracting foreign professionals, with cross-border employment having grown by 360% since 1994, as compared to 80% for resident employment. However, maintaining the high pace of growth, particularly in the sectors characterised by the most severe shortages, is becoming increasingly challenging. Population ageing will shrink the pool of workers in neighbouring countries available for work in Luxembourg for years to come. Moreover, skilled labour shortages in areas such IT and STEM are a common feature throughout the OECD, pushing up salaries and reducing the relative attractiveness of Luxembourg. Finally, rapid population growth and the surge in cross-border work are putting upward pressure on house prices and straining transport infrastructure.
Luxembourg has recently reformed its preferential impatriés scheme for foreign workers, with new provisions becoming effective from January 2025. This reform aligns Luxembourg’s regime to a number of OECD countries that have introduced preferential tax schemes to attract foreign talent, including, among others, Belgium, France, Denmark, Finland, Italy, the Netherlands, Portugal, Spain, Sweden, and Switzerland. While the evidence suggests that these schemes are effective in attracting skilled workers (Kleven et al., 2020), it is key to design them to avoid beggar thy-neighbours and regressive outcomes.
Luxembourg new impatriés scheme closely resembles those in Denmark and the Netherlands (Table 4.3). Duration of the preferential schemes are similar, being 8 years in Luxembourg, 7 in Denmark and 5 in the Netherlands. The preferential treatment in Denmark is given on the tax rate, substituting a flat income tax rate of 30% for the regular progressive income tax with a top marginal tax rate of 56%, while in Luxembourg and the Netherlands the schemes grant eligible employees a tax exemption of up to 50% of gross annual income for the former and 30% for the latter. In all the three schemes, eligibility conditions dictate a minimum threshold for gross annual salary, even though these limits in Netherlands are somewhat more flexible, having no salary requirement for scientists of qualifying universities or recognised institutions and a lower minimum threshold for workers under the age of 30.
LUX - 2020 scheme |
LUX - Entlaaschtungs-Pak Reform |
DNK |
NLD |
|
---|---|---|---|---|
Remuneration |
Tax exemption for 50% of the impatriation bonus |
Tax exemption for 50% of total gross annual remuneration |
Flat income tax rate of 32.8% instead of the progressive income tax (top marginal tax rate 56%). |
Tax exemption for up to 30% of total gross income (to become 27% from 2027 on) |
Limitations |
Up to a limit of 30% of the fixed annual remuneration (excluding bonuses and benefits in cash/kind) |
Up to a limit of €400,000 (excluding benefits in cash and in kind) |
It is possible to use the scheme only once in a lifetime, but the 7 years can be split into periods |
30% exemption can be applied over a maximum of maximum €233000, including taxed expense reimbursements and employer pension contribution |
Duration |
8 years |
8 years |
7 years |
5 years |
Eligibility conditions |
> Not having been a tax resident in Luxembourg or having lived less than 150km away from the Luxembourg border in the past 5 years. > Earning a gross annual salary of at least EUR 75,000. > No more than 30% of the company’s employees may benefit from the regime. > Having in-depth specialisation |
> Not having been a tax resident in Luxembourg or having lived less than 150km away from the Luxembourg border in the past 5 years. > Earning a gross annual salary of at least EUR 75,000. > No more than 30% of the company’s employees may benefit from the regime. > Having in-depth specialisation |
>Having an annual remuneration higher than the ninety-ninth percentile of the earnings distribution or being a researcher. > Not having been subject to full or limited tax liability in Denmark within the past 10 years prior to using the scheme. |
> Possessing expertise that is scarce in the Dutch labour market. > Being recruited from abroad by a firm withholding payroll taxes in NLD. > Earning a gross annual salary of at least EUR 50000 (exemptions: younger than 30s, PhD researchers and specialising doctors). > Not applicable to Dutch nationals |
The 2024 reform of the preferential tax scheme allows Luxembourg to compete with other OECD countries in the attraction of foreign talent as it aligns the fiscal incentives to those of other countries. However, it will be important to monitor the application of the new schemes to check its fiscal impact, its degree of regressivity and so the overall impact on government revenues and the economy. Both in Denmark and the Netherlands preferential tax rates for high-skilled immigrants are close to being revenue maximising, with revenue losses from lower rates broadly offset by gains from larger tax bases (Kleven et al., 2020) (Timm and Giuliodori and Muller, 2022)).
Road congestion represents a key deterrent for French, Belgian and German talents who live close to the border with Luxembourg and might want to work there without moving to the Grand Duchy. Cross-border workers in Luxembourg spend almost 10 hours per week commuting, with an average duration of a single journey to work being around 58 minutes, despite the relative short distances. As discussed in Chapter 3, decreasing the reliance on private cars for commuting is key to reduce traffic congestion and emissions. To this end, Luxembourg has made public transport free of charge for anyone. However, to be more effective this measure needs to be coupled with the continued expansion of the public transport network, the introduction of highway lanes dedicated to public buses and carpooling vehicles, and the reduction of public transport congestion through the adoption of staggered working and schooling times.
Promoting remote work would be a further effective approach to reduce traffic congestion, improving the quality of daily life and thus increasing the attractiveness of working in Luxembourg. Moreover, hybrid schedules, in which individuals spend a mix of days at home and at work each week, is found to improve overall job satisfaction and reduced quit rates by one-third, without damaging performance (Bloom et al., 2024). However, the bilateral double taxation agreements with neighbouring countries set the tolerance threshold for remote work from abroad at 34 working days per year. If the threshold is exceeded, the salary is not taxed in Luxembourg but in the country of residence (Luxembourg, 2024a). This limits the use of remote working from home as an effective solution for cross-border workers.
As a response, some big firms have opened satellite offices in small centres close to the borders with Germany, Belgium and France, in an attempt to offer more flexibility to their cross-border employees and reduce commuting time without breaching the 34 days threshold. While this solution might be effective in increasing the attractiveness of Luxembourg for cross-border workers, its scope is limited, as small- and medium-size firms might face logistic and financial constraints in applying it regardless of their sector of activity, and for in-person services and manufacturing workers teleworking is not a viable solution.
Beyond the issues with housing and road congestion that affect the quality of daily life in the Grand Duchy, a better integration into society and the labour market of professionals and their families is key to attract and retain foreign talent, as less than half of foreign workers who started working and living in Luxembourg in 2015 were still residents in the Grand Duchy in 2020. Comprehensive programmes offering systematic support to foreign workers’ spouses are not available in Luxembourg and the employment rate of foreign spouses is 16 percentage points lower than the employment rate of the principal migrant (OECD, 2023b). This negatively affects the probability of immigrant talents to come and remain in Luxembourg and constitutes an untapped source of talent.
Since 2023, non-EU accompanying spouses have immediate access to the job market as soon as they arrive, while before they had to apply for the salaried worker residence permit. This is an important change, but to be more effective it should be coupled with a more systematic support aimed at providing information on work opportunities and daily life in Luxembourg, adding to the already existing initiative of the public employment agency ADEM to hold information events. Finland and Estonia provide examples of good practices by including regular workshops and networking events, where spouses can learn about important information on how to settle in the country, its working culture and the availability of public childcare (OECD, 2023b). The programmes also encompass individual work-related counselling and mentoring by peers, including on how to design CVs and cover letters in accordance with the country’s national standards.
Increasing the attractiveness of Luxembourg for non-EU nationals is key to meet specific skill demands. Accessibility in terms of migration policies and admission procedures plays an important role in determining countries’ attractiveness. Non-EU nationals need to go through a structured procedure to come to Luxembourg to carry out a salaried activity for more than three months, after having signed an employment contract with an employer in Luxembourg. The procedure, which involves applications to be sent only on paper, can require from 1 to up to 4 months to issue a temporary authorisation. The waiting time for a long-stay visa that allows to remain in Luxembourg and the Schengen area for more than 90 days (type-D visa) might be even longer, while the median statutory limit in many OECD countries for the same kind of procedures is around 90 days (OECD, 2023b). These features might reduce the attractiveness of Luxembourg for non-EU nationals and create difficulties for employers posting vacancies abroad. Fully digitalising the different steps of visa and work permit processing should simplify and speed up the process, reducing barriers for foreign talent who want to apply to a job in Luxembourg (Box 4.4).
Migration policies have been made more responsive to labour market needs by easing immigration procedures for workers whose skills are scarce and in high demand. In June 2024, Luxembourg transposed in national legislation the new EU directive on the Blue Card. The last update of the directive in 2021 makes some admission conditions more flexible for highly qualified workers. The minimum duration of the work contract is now 6 months (previously 12 months), and the annual salary must be at least equivalent to the average salary (58,968 euros), while before it needed to be 1.2 or 1.5 times the average salary. The directive also imposes a limit of 90 days from the submission of the application to receive a final decision and simplifies the request of family reunification permits for the same duration of the Blue Card permission.
The procedure to recruit a non-EU national has been improved for professionals who do not meet the EU Blue Card criteria. For all vacancies covered by the official list of occupations in high demand, a work permit will be automatically delivered by the employment agency ADEM within 5 working days from the employer’s request, waiving the requirement to perform an assessment by ADEM of whether that particular position could be filled by a jobseeker in the national labour market (OECD, 2023b).
Empirical evidence shows the positive impact of digitalisation on reducing the length of the admission process, with artificial intelligence (AI) tools likely to reduce it even further (EMN-OECD, 2022).
Australia provides an example of good practices in digitalising its visa application system, as it reached a share of more than 93% of visa applications filed online by the end of 2021. The continuous development and improvement of the system over time, improving automated correspondence and assistance services, and the priority given to online applications, which were made mandatory for certain types of visas (work and holidays) in 2019, contributed to this result (OECD, 2023b).
Canada is a good example of introducing AI technologies to increase the effectiveness of the scrutiny process. The introduction of advanced data analytics to identify routine applications for streamlined processing has allowed to better use available human resources, and ultimately speed up the overall process (Canada IRCC, 2024).
Migration policies could also be strengthened to attract non-EU entrepreneurs, as this might favour the attraction of investment, talent and capacity for innovation (Ditsche et al., 2023) (Peroni et al., 2016). Luxembourg has programmes to encourage start-ups and entrepreneurship, also in collaboration with neighbouring countries. However, it does not have any specific legislative framework aimed at attracting start-up founders and innovative entrepreneurs from third countries (EMN, 2019). Portugal, which introduced a start-up visa in 2018 to attract high-potential immigrant entrepreneurs, provides an example of good practices (OECD, 2023c). In the Portuguese system, the applicants need to present an innovative business project and open a startup in collaboration with a Portuguese business incubator. Up to five partners may be included in the application, but the main applicant must be a non-EU citizen or resident. Applications can be filed completely online and IAPMEI, the specialised Portuguese agency for competitiveness and innovation in the scope of the Ministry of Economy, will review them and provide a final response by maximum 30 working days from the submission (IAPMEI, 2018).
Attracting international students and facilitating their transition into Luxembourg’s labour market is an important way to respond to skills shortages and also to attract and retain young foreign talents. The international appeal of the University of Luxembourg has increased over the years since its foundation in 2003, also thanks to a structured inter-institutional strategy that encompassed merit-based scholarships to international students and a national branding campaign aimed at raising the international awareness about Luxembourg’s assets. This has brought international students to represent more than 60% of the total students. However, while measures aimed at attracting talented students are providing good results, there are measures to be taken to maximise the possibilities to retain international talented students as professionals in Luxembourg, since less than 50% of international master and PhD graduates in the period 2014-2019 had a first registered employment in Luxembourg (OECD, 2022a). The support provided to students in establishing connections with Luxembourg’s labour market during their studies should be enhanced. To this end, the authorised number of hours for working alongside studies might be increased beyond the present limit of 15 per week, and targeted career guidance services might be further developed.
Former international students from non-EU countries should be supported to remain in Luxembourg to look for a job following the completion of their studies. To this end, a job-search visa has been introduced. Non-EU students who have completed their diploma, bachelor's, master's, or PhD program and hold a residence permit for students in Luxembourg, or students who have a student residence permit valid in a first Member State and who are in Luxembourg on an educational mobility scheme, can apply for the purposes of finding a job or setting up a business. Following France’s example, the validity period of the job-search visas has been extended to 12 months. However, many other OECD countries, such as Canada, Australia and Germany, allow for a longer duration, from 18 months up to three years.
While the introduction of this visa represents an important and positive measure, some actions need to be taken to improve its effectiveness, as applications have been low so far. First, awareness among students needs to be raised. Second, even though recent cases indicate that the processing period has been shortened from the previous average of 90 days to around 4 weeks, streamlining the application process and reducing permanently the processing time would make it easier for graduates to transition smoothly without gaps in their legal status. New Zealand, with its average processing time for a post study work visa of around 6 weeks and 80% of demands processed within 3 weeks, provides an example of good practices (New Zealand, 2024) and its share of 50% of international students who extended their stay is one of the highest in the OECD (OECD, 2022b). Third, the eligibility might be expanded to include students who have completed short-term or vocational programs, as this could attract individuals with lower levels of education but with skills that are in high demand in Luxembourg. Australia's temporary graduate visa allows for this. Finally, the creation of a dedicated and simplified procedure for candidates who have already received a formal offer of employment from a Luxembourg employers could be envisaged.
Upskilling and reskilling the workforce can be powerful tools to confront skills shortages and raise human capital to the benefit of total factor productivity growth. Luxembourg firms are highly involved in workforce training, with the share of Luxembourg firms providing continuing vocational training courses in line with the shares in neighbouring countries and higher than the EU average (Figure 4.15, Panel A). The same applies to the number of hours spent in continuing vocational training by firms’ employees (Figure 4.15, Panel B) and to the participation in formal and non-formal education and training by adult aged 25-64 (Eurostat, 2024). The existence in Luxembourg of various supporting measures to encourage the participation in learning activities, available to both employees and employers, is a positive step. However, there is ample space to enhance the effectiveness of these measures and improve the quality and the relevance of training activities.
2020
The individual training leave (congé individuel de formation) is a key scheme in Luxembourg to encourage self-employed workers and individuals employed in the private sector to engage in training activities. Each worker has a total of 80 days of training leave during the entire career, with a ceiling of 20 days per two years. When workers take their days of training leave, the employer continues to pay their salary, but the government reimburses the employer for the wage costs as well as the employer’s share of social security contributions. For self-employed workers, the government pays the individual a salary equivalent to their average daily earnings for every day of training leave. While employers can ask for training leave to be postponed guaranteeing the functioning of their company, they cannot oppose it. However, only about 1% of the total workforce in Luxembourg makes use of training leave each year and the take-up is concentrated among young, male, high-skilled employees in the finance and insurance sector, who engage in training courses in finance and accounting (OECD, 2023). This limits the scheme’s effectiveness, as it is used primarily by people who need it least, while low-qualified and older workers, those working in smaller businesses and self-employed workers, who would it need most, are rarely using it (OECD, 2019b).
The need to reform the individual training leave scheme to broaden its impact has been recognised in the 2023-2028 coalition agreement. Enhancing the customisation and targeting of current financial incentives to better support vulnerable groups would allow to increase the effectiveness of the support programmes. For example, additional financial incentives, such as training vouchers, might be provided to cover upfront costs and/or expenses related to training participation. These additional incentives might be targeted to employees with low qualifications and to develop highly demanded skills that are in short supply. More flexibility could also be allowed as to when the 80 days of training leave can be accessed over a career, especially for young low-qualified workers, as taking more training earlier in the career might improve their labour market prospects.
Training offers should be tailored to both workers’ and employers’ needs. Modularising training programmes, as done for example with the Skillsbridges and the Diplom+ programmes (Box 4.5), could foster a wider participation and can be particularly important for employees in micro and small enterprises, as it would allow to reduce the impact of training leave on their working time and so on the functioning of the firm. By contrast, courses could be made more intensive for the long-term unemployed and individuals in need of deep and fast reskilling to be fit for a new job (OECD, 2023).
The Skillsbridges programme, which started in September 2024, provides short-term courses designed to help adults to develop targeted skills that can be immediately applied in practice, with courses covering areas such as artificial intelligence, data analysis, cloud computing, as well as renewable energies and sustainability. The content of the training programmes is selected through labour market analyses and the curricula are regularly updated to ensure that these educational programmes are always up to date with employer needs and new trends in the labour market.
The Diplom+ programme, introduced in 2020, is addressed to secondary graduates holding a general and technical secondary education diploma or technician’s diploma and who are either not sure of their choice of studies or willing to enter the labour market. The programme offers a common core with soft skills (e.g. employability, public speaking), transversal technical skills (e.g. project management), and green skills, as well as optional digital courses and practical training in crafts.
Both programmes are based on a modular architecture, which increases accessibility, participation, and flexibility, with the possibility to attend these courses while working, looking for a job or enrolling in higher education.
Guidance is a key part of encouraging people to improve their skills as across OECD countries the lack of adequate information on skills and learning outcomes has a concrete negative impact on participation to continuing vocational training courses, especially among low-skilled workers (OECD, 2020). Luxembourg established in 2012 the Maison de l’Orientation to offer a coherent and centralised service of lifelong skill development guidance to all individuals, delivered through a national “one-stop-shop”. This includes support to identify people’s capabilities, skills and interests, as well as to inform their professional and training activities, so that individuals become more resourceful in their own career management (CEDEFOP, 2020a). The establishment of the Maison de l’Orientation represents an important step, but there is space to enhance its effectiveness. The approach should be more proactive in engaging low-skilled adults, as they usually make less use of these structures. To this end, skills-development advice might be combined and preceded by awareness raising activities on the payoffs and advantages of reskilling and upskilling for workers. Employers’ and employees’ representatives might be taken onboard to design a coordinated strategy to better reach the targeted groups.
Training quality is important for the outcomes of investment in training and the attractiveness of training for individuals. Currently in Luxembourg’s adult learning system, the Ministry of Education provides accreditation and certification to providers. However, the accreditation process and the selection criteria need to be made more rigorous to improve quality. Currently many employers in Luxembourg are concerned that the training provided to their employees is of low value in terms of addressing their employees’ skills needs (OECD, 2023). Almost 90% of workers declare that a better quality of training would encourage higher participation (CEDEFOP, 2020b).
Setting up a certification agency encompassing relevant ministries, as well as employer and employee representatives could be a first step. Its aim should be to reevaluate from scratch the training providers and increase the quality standards required to grant them the certification as accredited training centres. The certification should be periodically renewed, in order to assure quality standards over time. Introducing a national certification procedure to become an adult teacher and a regular professional development requirement for these teachers would help to ensure and maintain high quality and effective teaching standards. Furthermore, a regular and systematic review process of curricula should be introduced to ensure labour market relevance. Switzerland‘s “eduQua” provides an example of good practices to be followed (Box 4.6).
In 2000 the Swiss Federation for Adult Learning (SVEB) – an umbrella non-governmental organisation representing both public and private institutions, associations, and personnel managers – introduced the quality label “eduQua” to ensure that providers of adult learning meet minimum standards.
In over 20 years since its introduction, eduQua has proved to be very successful in signalling quality of training provision. First, the certification process for the training centres to receive the quality label involves documentation but also on-site visits and yearly intermediate audits, allowing for a more accurate and thorough check of the quality of the formation offered in each centre. Second, the eduQua label certifies the whole institution, and not its individual courses, meaning the quality standards need to be high on each aspect, from curricula to structures and teachers. Third, the certification lasts three years, after which the provider must undergo a renewal, allowing for a continuous verification of quality. Fourth, among the required quality standards, there is also the minimum qualifications for teachers and further training to be followed by teachers themselves. Fifth, the certification process also takes into account the quality of training outcomes, measuring it through feedback from students and labour market performance of past participants. While the whole eduQua system is aimed at ensuring that centres provide high quality formation, this last requirement is aimed at specifically addressing a concern that many employers in Luxembourg have, pointing to reward centres that not only provide high-quality formation, but also teach labour market-relevant skills, allowing for a positive impact on both employees and employers.
Source: (OECD, 2021)
Since its creation, the University of Luxembourg has increased and refined its formal and informal adult learning offer. It provides both bachelor and master programmes for adults that can be pursued on a part-time basis, and a variety of shorter certificate programmes. The University of Luxembourg actively supports the alignment of the higher education offer to labour market needs, as the recent creation of the University of Luxembourg Competence Centre (ULCC) demonstrates. The centre, created jointly with the government, conducts regular studies analysing the skill needs in the labour market and also provides training courses (open online courses) to a variety of sectors. While this is a positive step, the uptake of adult learning courses should be increased. One key difficulty for professionals is the obligation to sign up for a whole bachelor or master programme, as modules are not transferable across programmes. Not providing the possibility to sign up for individual modules can hamper wider participation, as it might constitute a significant obstacle in terms of time and finances needed, ultimately making the adult learning offer less accessible.
To increase participation, module structure and size should be standardised across the whole university and the university should have an interoperable module structure across all the programmes destined to adult learning. This means that modular programmes should be structured to divide the formal education programmes into smaller, self-contained units, with each one of them having its own curriculum and learning objectives. When the module is completed, this grants certified credits or partial qualifications. These credits or partial qualifications can then be tracked in a personal learning portfolio and combined into a full qualification over time, allowing for more flexibility and potentially lowering barriers to participation, contributing also to reducing the drop-out rate (OECD, 2023).
The share of residents who studied STEM disciplines at university is below the OECD and EU averages, while shortages in this field are particularly severe. Moreover, as in other OECD countries, the share of participation and graduation of female students in Luxembourg in these fields is significantly lower than that of male students. Recently, some policy initiatives have been taken to address these issues. “Gender4Stem” is an initiative aimed at combating stereotypes and sparking greater interest in STEM disciplines among girls through the creation of an online platform where educational and awareness-raising materials are available for use by secondary-level teachers. Luxembourg has also joined the network of European Space Education Resource Offices (ESERO), which provides learning material and inspiration for teachers wishing to bring space into the classroom. The Code Club is also active in Luxembourg, a non-profit organisation that involves volunteers, teachers and external professionals who imparts the basics of programming to students of primary and lower secondary school in weekly one-hour programming sessions.
However, a more structured approach is needed, as many students who perform well in science at the high-school level do not consistently pursue scientific degrees afterwards at university level (Mysore, 2018). Together with sparking interest and combating stereotypes, a comprehensive career guidance service should be implemented to increase the uptake of STEM disciplines at the tertiary level. Finland provides an example of good practices, as it integrates career guidance into the education system from an early age, with the Strategy for lifelong guidance ensuring that guidance services are aligned with national education and employment goals (Cedefop, 2021). Finland has one of the highest proportions of bachelor's, master's and doctoral or equivalent graduates in the field of STEM among OECD countries (OECD, 2024). Moreover, schools helping students to understand the relevance and advantages of pursuing STEM-related subjects for their professional future have brought many students to pursue mathematics-related studies exactly because of the usefulness for their career (Kaleva et al., 2019).
Competition is a key engine of productivity growth, driving improvements both within firms and across the broader economy. Economies with more competition-friendly regulations consistently outperform those with more restrictive market environments, achieving higher rates of productivity growth (OECD, 2015). The OECD Product Market Regulation (PMR) indicator suggests high barriers to competition and market contestability (Figure 4.16), with Luxembourg performing poorly with respect to other OECD countries mainly because of barriers to competition in professional services, poor lobbying rules and improvable regulatory processes.
OECD Product Market Regulation indicator, index scale 0-6 from most to least competition friendly regulation, 2023
Barriers to competition in professional services are high compared to other OECD countries (Figure 4.17). Professional services are instrumental to many other business activities, especially in the financial sector. They represent around 8.5% of non-financial employment, a share that is much higher than the EU and euro area averages. In many cases there is only one pathway to access the profession after the relevant university degree, which also requires passing a professional examination, and there is the obligation to be member of the relevant professional association to be allowed to practice the profession. Furthermore, in some cases, such as for notaries, competitive behaviour is hindered by the existence of binding maximum/minimum tariffs and the prohibition of any form of advertising and marketing.
Reducing or making these entry requirements more flexible, while preserving quality especially in the specific case of Luxembourg where the majority of these professionals have studied abroad, would allow to increase competition, encourage productivity enhancing practices and lower prices, to the benefit of all the enterprises that use them in their business. Norway provides a practical example of setting less stringent requirements for architects, that can become even more relevant for Luxembourg as the University of Luxembourg continues to develop and structure further its educational offer. While 5 years of education (3 years bachelor + 2 years master’s) provided at three architect schools in Norway are required to practice the profession and sign projects, there is no mandatory exam to access the profession, nor is membership of one of the relevant professional associations legally required (ACE, 2024). A reform making entry requirements more flexible for civil engineers and architects is underway.
PMR indicator for professional services, index scale 0 to 6 from most to least competition-friendly regulation, 2023
In the retail sector, the requirement of specific authorisations for opening clothing, food and beverage retail outlets increases the administrative burden and discourages the entrance of new businesses. The permission of doing sales promotions only during some specific periods of the year limits entrepreneurial freedom and price competition. Geographical restrictions on where a pharmacy can be located, on the number of pharmacies that can be located in a given geographic area and the obligation that the opening and closing hours of pharmacies are determined by professional associations (OECD, 2024d). Luxembourg should reduce these regulatory barriers in the retail sector to encourage competition. For example, Ireland revoked in 2002 the establishment rules for the opening of new pharmacies. The Netherlands in 1998 prohibited establishment restrictions practiced by the pharmacy association. Norway abolished in 2001 the statutory establishment and ownership criteria for community pharmacies and horizontal and vertical integration were allowed, together with the permission of selling a restricted number of over-the-counter medicines outside pharmacies in 2003 (Vogler et al., 2012).
Complex laws and regulation that are associated with high compliance costs, cumbersome and unclear licensing procedures that require time-consuming paperwork and approvals are examples of how administrative and regulatory burden can concretely make it more difficult and costly for new competitors to enter a market, reducing overall competition. Moreover, these administrative and regulatory burdens often have a disproportionate impact on smaller entering firms, which often are the main drivers of innovation but that may lack the resources and experience to deal efficiently with complex regulatory environments (OECD, 2019) .
Luxembourg has room to increase the quality of its regulatory process and reduce its administrative burden to the benefit of competition (Figure 4.18, Panel A). A regulatory process that fosters market dynamism needs to take into account the impact of new laws on competition and has to produce laws and regulations that are simple to understand and do not hinder entrepreneurial activity. However, in Luxembourg, the assessment of the impact of new laws and regulations on competition and the ability of businesses to innovate is not required (Figure 4.18, Panel B). Also, when such assessment is performed, there are not transparent, shared and clear guidelines on how to do so (OECD, 2024d).
2023
Note: The scale of the indicator in each panel goes from 0 (more competitive) to 6 (less competitive)
Source: Product market regulation | OECD
A sound assessment of the impact and the effectiveness of new rules and laws through both ex-ante and ex-post evaluation is important to ensure that regulation is fit for purpose. To be effective, this assessment needs to be systematic and follow clear and well-established procedures. Flanders and Wallonia provide examples of good practices on public policy assessment. Flanders runs an evaluation unit within its Economy, Science and Innovation Department, which aims to initiate, supervise and monitor public policy evaluations through a transparent evaluation framework. In Wallonia, the Institut wallon de l’évaluation, de la prospective et de la statistique (IWEPS) leads its policy evaluations through a well-established process and each evaluation is described in detail and is published on the IWEPS website (OECD, 2024).
Quality regulation needs to be easy to understand, but in Luxembourg there is no legal requirement for the government to use 'plain language' in the drafting of new primary laws and subordinate regulations. The United States established through the 2010 Plain Writing Act and three different executive orders that regulations should be written in plain and clear language, be simple and easy to understand, with the final goal of minimising the potential for uncertainty and litigation arising from such uncertainty (USA, 2024).
Luxembourg performs well below the OECD average when it comes to regulating lobbying activities, as the rules in place do not fully guarantee the transparency and accountability of these interactions (Figure 4.18, Panel C). The regulatory framework defines lobbying activities and establishes cooling off periods, but only for former members of government, not for other members of Parliament and lobbyists. The only disclosure requirement for lobbyists is to register in a public registry, but there are no sanctions for non-compliance and the register does not disclose the targeted legislation. Finally, members of the highest body of the judiciary and top-tier civil servants are not legally required to officially declare their interest to detect possible conflicts of interest (OECD, 2024c). This regulatory void may play a role in hampering competition, as it can benefit incumbents and larger firms, enabling them to shape rules in their favour at the expense of smaller businesses and new entrants.
Introducing the obligation to make publicly available the identity of the interest groups and public officials that were consulted in each interaction, together with the subject of their meeting, would be a substantial step to guarantee a level-playing field for both new entrants and incumbents. To this end the dedicated public register for lobbyists that was introduced in 2023 should be stregthened by imposing fair, objective, proportionate and dissuasive sanctions in case of non-compliance through disciplinary, administrative, civil and/or criminal processes (OECD, 2024c). France provides an example of good practice. In 2016 the Sapin II Act established that the High Authority for Transparency in Public Life may impose sanctions up to EUR 15000 and one-year imprisonment when legal requirements imposed to lobbyists are not met (European Parliament, 2018). Ireland provides a further example of good practices when it comes to regulate lobby activities, as it imposed disclosure obligations on policymakers and lobbyists to ensure transparency in their interactions, but it also imposed on public officials involved in the regulatory process both conflict of interest rules and a cooling-off period when leaving office (OECD, 2024).
Effective insolvency frameworks are important to reduce barriers to exit markets and thus promote competition, efficiency and productivity growth in different ways. First, they help ensure that non-viable companies exit the market in a timely manner, allowing viable ones to restructure. This reduces the share of productive factors sunk in zombie firms and spurs the reallocation of capital and workers to more productive firms. Second, they promote entrepreneurship by offering a second chance to failed entrepreneurs. Not punishing entrepreneurs for their failure can promote risk taking, experimentation and thus innovation (OECD, 2022).
In 2023, Luxembourg deeply reformed its insolvency framework in line with implementation of the EU Directive 2019/1023 on preventive restructuring frameworks (Box 4.7). The recent reform has allowed Luxembourg to improve its performance in terms of the OECD insolvency indicator, which in 2022 placed the insolvency regime among the worst performing in all the three main subcomponents: treatment of failed entrepreneurs; prevention and streamlining; and restructuring tools (Figure 4.19). The reform has positively impacted all the three pillars as it has addressed many of the problems of the old system. It has introduced an early warning mechanism; repealed outdated and ineffective procedures such as controlled management and introduced new and more agile mechanisms to prevent distressed companies from systematically resorting to bankruptcy. However, more can be done to narrow the gap with the most favorable regimes even further. An example is the treatment of failed entrepreneurs. The new insolvency law has introduced a debt discharge mechanism for good-faith debtors. While this is positive step, further reducing the time to obtain debt discharge from the present 18 months would incentivise entrepreneurship and experimentation even further.
Note: The scores for the three main sub-categories are scaled from zero to one, with lower scores indicating more favourable framework. LUX* refers to the new insolvency regime, after 2023 reform. LUX refers to the old regime.
The law of 7 August 2023 on business preservation and modernisation of the bankruptcy regime entered into force on November 1, 2023, bringing many novelties with respect to the old insolvency framework.
While the previous framework was focused only on commercial entities, the scope of application of the new regime has been broadened, including special limited partnerships, non-trading companies and artisans. Financial firms, pension and investment funds and insurance companies are still excluded.
The preventive arm has been enhanced. Before there was no formal early warning system, while the new framework grants the possibility for the Ministry of Economy to access certain information, including payment defaults, in order to identify companies facing financial difficulties and invite them to discuss how business continuity can be ensured. To improve the detection of companies on financial distress, a special evaluation committee for businesses in difficulties (Cellule d’évaluation des entreprises en difficultés) has been created with members designated by the Minister of the Economy amongst the personnel of some of the main state administrations.
The new regime introduces the figure of the conciliator, who can be appointed at the request of the distressed company to assist the debtor with the reorganisation of all or part of its business.
The reform introduced two new proceedings as alternatives to insolvency: reorganisation by settlement agreement and the judicial reorganisation procedure. The reorganisation by settlement agreement is a voluntary out-of-court proceeding where debtor companies can propose amicable agreements to creditors for reorganising their assets or activities, with or without the assistance of a conciliator, and these agreements, once approved by the court, become enforceable and remain confidential. The judicial reorganisation can be carried out either by amicable agreement, by adopting a plan or through a court-ordered transfer.
Some provisions of the bankruptcy procedures have been amended. Now the bankruptcy procedure can also be initiated by the public prosecutor; fraudulent bankruptcy has been decriminalised as it will now constitute an offence and no longer a crime, with the aim of easing prosecution, and offences of both simple bankruptcy and fraudulent bankruptcy have been extended to also apply to the company’s managers.
The 2023 reform is a major positive step as it aims to foster company reorganisation over liquidation, promoting business continuity. Efficient implementation of the new framework is key to see tangible progress and to be sure this reform simplifies the processes related to companies’ insolvency without risking doing the opposite through heightened uncertainty, longer procedures and administrative complexity. Doing so poses some challenges. First, the new regime provides debtors and creditors with additional options, but this implies that it imposes additional duties and obligations on court officers and other relevant bodies involved in insolvency procedures. Second, the reform also introduces new profiles such as company conciliator and the judicial agent. Third, the introduction of new procedures and professional figures requires a substantial dissemination effort to make sure all stakeholders are aware of the various options to make adequate choices when facing financial distress situations.
In view of this, a viable option to be sure courts are equipped with the right expertise to handle the new procedures and tools to deal with insolvency matters would be to create specialised insolvency courts, equipped with specialised judges and administrative staff, as the US did when it established specialised Federal Bankruptcy Courts (US Courts, 2024). France in 2015 created specialised commercial courts to handle complex insolvency proceedings. As normal commercial courts dealing with simpler insolvency procedures in France, these courts are composed by lay, non-professional judges, who are usually peer-elected former or current company managers or entrepreneurs, as they possess the right expertise to deal with these technical matters (France - Ministere de la Justice, 2024).
MAIN FINDINGS |
RECOMMENDATIONS (key in bold) |
---|---|
Enhancing innovation and technology adoption |
|
R&D support is provided by various institutions, with different application procedures and selection criteria, and the allocation of grants is institution based rather than project based. |
Establish a coordination mechanism to streamline public R&D support, while making it more project-oriented and conditional on public-private partnerships. |
Technology adoption lags behind top-performing countries. |
Streamline the investment tax credit and make incremental digital and technological improvements eligible. |
Firms rely massively on bank financing. Alternative sources of financing, such as venture capital, remain underdeveloped. |
Encourage venture capital supply from individuals through targeted fiscal incentives and enhance financial literacy among managers to increase demand. |
Easing skills shortages |
|
Lengthy visa procedures complicate the hiring of non-EU nationals. |
Fully digitalise the different steps of visa and work permit processing and introduce an ad-hoc visa to attract high-potential immigrant entrepreneurs. |
The uptake of the individual training leave is very low and unbalanced, being lower among low skilled and older workers. |
Enhance the customisation and targeting of current training incentives and develop more proactive guidance. |
Accredited training providers are numerous, and the quality and relevance of courses is not always of the highest standards. |
Tighten quality control by creating a national accreditation agency that encompasses the relevant ministries and the social partners. |
Luxembourg has lower than EU and OECD shares of residents who study STEM disciplines at university. |
Develop a comprehensive career guidance service, integrated into the education system from an early age, showcasing the advantages of choosing STEM-related subjects at university for career’s prospects. |
In recent years, less than 50% of international master and PhD graduates had a first registered employment in Luxembourg |
Extend the limit on authorised number of hours for working alongside studies and streamline the application process for a job-search visa, reducing the processing time. |
Strengthening competition and market dynamism |
|
Barriers to entry in professional services are high A reform to make entry requirements more flexible for civil engineers and architects is underway. |
Review the requirement to be member of a professional association and to pass a professional examination on top of a formal education degree. |
There is no requirement to assess the impact of new laws and regulations on competition nor to use plain language in writing norms. |
Introduce ex-ante and ex-post evaluation of the impact of regulation on competition. Introduce the requirement to use plain language in drafting new law and regulation. |
Lobbying regulations do not guarantee full transparency and accountability. |
Require full disclosure of the identity of lobbyists and public officials for each interaction. |
The insolvency framework has been reformed, but a gap with top performing countries remains. |
Support the new insolvency framework by creating specialised insolvency courts and further reduce the time to discharge debt for good-faith debtors |
ACE (2024), “Norway - Architects Council of Europe”.
ADEM (2024a), “La liste 2024 des métiers très en pénurie”.
ADEM (2024b), “Overview of the labour market”.
Amess et al. (2016), “The impact of private equity on firms׳ patenting activity”.
Aghion, P., C. Antonin, S. Bunel (2023), “The Power of Creative Destruction: Economic Upheaval and the Wealth of Nations”.
Appelt, S., et al. (2016), “R&D Tax Incentives: Evidence on design, incidence and impacts”.
Australian Taxation Office (2023), “Small business technology investment boost”.
BEREC (2018), “BEREC Report on infrastructure sharing”.
Bloom et al. (2024), “Hybrid working from home improves retention without damaging performance”.
Canada IRCC (2024), “Digital transparency: Automation and advanced data analytics”.
Castellani et al. (2016), “The Productivity Impact of R&D Investment”.
Cedefop (2021), “Inventory of lifelong guidance systems and practices – Finland”.
CEDEFOP (2020a), “Career guidance policy and practice in the pandemic: results of a joint international survey”.
CEDEFOP (2020b), “Perceptions on adult learning and vocational education and training in Europe”.
Chambe des metiers (2024), “Enquête sur le financement des nvestissements des entreprises artisanales”.
Conseil National de la Productivité (2023), “Rapport Annuel 2022-2023”.
Ditsche et al. (2023), “How to Attract Start-Ups to Boost National Economies: A German-Georgian Comparison and Case Study”.
EIF (2023), “Luxembourg Future Fund 1”.
EMN (2019), “Les voies migratoires pour fondateurs de start-ups et entrepreneurs innovants au Luxembourg”.
EMN-OECD (2022), “The use of digitalisation and artificial intelligence in migration management”.
European Commission (2024), “Luxembourg 2024 Digital Decade Country Report”.
European Commission (2021), “Study on the effectiveness of public innovation support for SMEs in Europe”.
European Commission (2014), “A Study on R&D Tax Incentives - Annex: Good practice cases”.
European Parliament (2018), “New lobbying law in France”.
Eurostat (2024), “Adult learning statistics”.
Eurostat (2024b), “Statistics on young people neither in employment nor in education or training”.
FNR (2024), “What we do”.
France - Ministere de la Justice (2024), “Les juridictions judiciaires spécialisées en matière civile”.
IAPMEI (2018), “Start Up Visa”.
Interreg Europe (2021), “Voucher schemes: zooming in on innovation vouchers”.
Kaleva et al. (2019), “Students’ reasons for STEM choices and the relationship of mathematics choice to university admission”.
Kleven et al. (2020), “Taxation and Migration: Evidence and Policy Implications”.
Luxembourg (2024a), “5 KEY POINTERS FOR CROSS-BORDER WORKERS IN LUXEMBOURG”.
Luxembourg (2024b), “Tax deductibility”.
Luxembourg (2024c), “Tax relief for investment in a digital transformation project”.
Luxembourg (2024f), “Business incentives and support”.
Luxembourg (2024g), “Financial aid for SME innovation”.
Luxembourg (2022), “Individual training leave”.
Luxinnovation (2024a), Luxembourg launches Dataspace 4 Health”.
Luxinnovations (2024b), “Clean Technologies”.
Mysore, S. (2018), “International Approaches to STEM Education”
New Zealand (2024), “Post Study Work Visa”.
OECD (2024), Education at Glance.
OECD (2024b), OECD Economic Surveys: Belgium 2024
OECD (2024c), Anti-Corruption and Integrity Outlook 2024 – Country Notes: Luxembourg
OECD (2024d), OECD Product Market Regulation (PMR) indicators: how does Luxembourg compare?
OECD (2023a), “Education at Glance”.
OECD (2023b), “OECD Skills Strategy Luxembourg”.
OECD (2023c), “The Missing Entrepreneurs”.
OECD (2023d), “Evaluating the effects of the R&D tax credit in Iceland”.
OECD (2023e), “Public research funding in Sweden”.
OECD (2023f), “Rebooting the innovation”.
OECD (2023g), “The Impact of R&D tax incentives”.
OECD (2023h), “OECD Services Trade Restrictiveness Index (STRI) - Luxembourg”.
OECD (2022), “Enhancing insolvency frameworks to support economic renewal”.
OECD (2022a), “International Migration Outlook”.
OECD (2022b), “Retention and economic impact of international students in the OECD”.
OECD (2022d), “OECD Economic Survey of Luxembourg”.
OECD (2021), Improving the Quality of Non-Formal Adult Learning: Learning from European Best Practices on Quality Assurance”.
OECD (2020), “Strengthening the Governance of Skills Systems: Lessons from Six OECD”.
OECD (2020b), Laggard firms, technology diffusion and its structural and its structural and policy determinants
OECD (2020c), The effects of R&D tax incentives and their role in the innovation policy mix: Findings from the OECD microBeRD project, 2016-19
OECD (2019a), Getting Skills Right: Future-Ready Adult Learning Systems.
OECD (2019b), “Individual Learning Accounts: Panacea or Pandora’s Box?”.
OECD (2015), “The Future of Productivity”.
OECD (2000), “The impact of public R&D expenditure on business R&D”.
OECD et APO (2022), Identifying the Main Drivers of Productivity Growth”.
Paglia and Harjoto (2014), “The effects of private equity and venture capital on sales and employment”.
Peroni, Riillo and Sarracino (2016), “Entrepreneurship and immigration: evidence from GEM Luxembourg”.
SNCI (2024), “About us”.
Ravikumar, Chinagorom-Abiakalam and Smaldone (2024), “Convergence or divergence? A look at GDP growth across richer and poorer countries”.
STATEC (2021), “La performance de R&D et d’innovation des entreprises”.
STATEC (2020), “Polarisation croissante de l’emploi salarié au Luxembourg”.
Timm and Giuliodori and Muller (2022), “Tax Incentives for High Skilled Migrants: Evidence from a Preferential Tax Scheme in the Netherlands”.
UK (2023), “Tax relief for investors using venture capital schemes”.
UK (2022), “Independent review of UK Research and Innovation (UKRI): final report and recommendations”.
USA (2024), “Plain Language”.
US Courts (2024), “Bankruptcy Courts and Cases”.
Vogler et al. (2012), “Impact of pharmacy deregulation and regulation in European countries”.
WMS (2024), “Management matters in an era of disruption”