Industrial Policy Handbook

Industrial strategies provide the overarching direction that guides the design and implementation of industrial policies. Part II of this Handbook outlines five key phases that policymakers may consider when developing and operationalising industrial policies, as illustrated in Figure 2. Each phase is addressed in turn, offering practical guidance to support policy design.

• Clarifying the underlying market failure is a critical first step. A well-defined rationale helps identify the most appropriate instrument, ensure that intervention is proportionate, targeted, and aligned with broader strategic objectives whilst providing a transparent basis on which stakeholders can build consensus and support.

• Market failures can be addressed by means other than direct aid, for example, creating a market, coordinating companies, overcoming information asymmetries or other types of measures (e.g. skills and training, regulatory reforms).

• Ex ante assessment of the potential effects of a proposed instrument helps determine whether, and to what extent, it is appropriate to address a given market failure. Such assessment could consider costs and benefits, proportionality, potential trade-offs and limitations, as well as risks of unintended consequences and possible government failures.

• Several instruments may be used to address the same market failure. Recognising potential complementarities amongst instruments to address a given failure can support more coherent policy design.

• International benchmarking is useful to help governments understand different possibilities in the design phase.

Many OECD countries have recognised that young or small and medium-sized enterprises (SMEs) often face difficulties accessing finance, which can limit their growth and innovation potential. This challenge arises from a combination of market failures:

• Information asymmetries – Banks often cannot fully assess the risk of lending to young firms and SMEs because these firms may have short credit histories, limited financial statements, or novel business models.

• Coordination challenges – No single actor can address gaps in financing for emerging firms, making it harder for new industries to grow.

• Nonmarket positive benefits – Lending to innovative start-ups can generate wider benefits for the economy, which are not considered by banks when lending to individual companies.

These factors can result in the financing of less risky projects and moral hazard (borrowers might take excessive risks once they have access to funds).

To address these challenges, many countries include government-backed loan guarantees as part of their industrial strategies. These guarantees help banks lend to young firms and SMEs by sharing some of the risk, making it easier for firms to access the financing they need to grow, innovate, and create jobs. Examples of such schemes include: Growth guarantee (Vækstkautioner) (Denmark), Bpifrance’s Garanties (France), KfW Guarantee Programme (Germany), Credit Review Office’s Credit Guarantee Scheme (Ireland), Accountant General’s Office in the Ministry of Finance - Fund for small and medium-sized businesses with a state guarantee (Israel), MIMIT - SMEs Guarantee Fund (Italy), BMKB’s SME Loans Guarantee (Netherlands), Slovene Enterprise Fund’s P1plus – guarantees (Slovenia), SIDA - Guarantee programme for SMEs (Sweden), British Business Bank’s ENABLE Guarantees (United Kingdom) and Small Business Administration 7(a) loan guarantee programme (United States).

Similar to SMEs, there is often underinvestment in new technologies such as cleantech sectors. The EU’s Clean Industrial Deal targets persistent underinvestment by addressing key market failures such as information externalities, scale‑up barriers, supply bottlenecks, regulatory uncertainty and undervalued nonmarket positive externalities. It deploys a coordinated mix of instruments including risk‑sharing finance through the European Investment Bank and European Investment Fund, streamlined permitting and demand‑creation under the Net Zero Industry Act, manufacturing scale‑up, and supply‑chain security measures under the Critical Raw Materials Act to directly tackle each underlying constraint (further details in the Supporting Material).

• Scope of targeting: Depending on the market failure, policymakers will select whether a wider or narrow targeting is needed. Wider targeting can be administratively simpler but harder to define precisely, and narrower targeting (e.g. by sector) may align more closely with strategic objectives but typically requires more intensive screening, entails higher administrative demands, and risks overlooking opportunities. In addition, narrower targeting can help better tailor interventions to specific market failures, but at the expense of a sector- or technology-bias, making the policy more vulnerable to economic and technological developments.

• Additionality: Often the aim is to target projects that would not proceed without support (i.e. would not be profitable without the intervention). Some strategies, however, do allow partially additional projects (e.g. those that can be expanded or accelerated by public funding). Regardless, evaluating how different instruments interact, and shape overall incentives, is key (as discussed further in Phase 5 of Part II).

• Targeting large firms may support large-scale projects and facilitate diffusion of innovation, but may involve higher risks of windfall gains, lower additionality, possible displacement of smaller competitors, and increased market concentration.

• Supporting smaller firms may generate higher additionality given tighter financial constraints, yet may also raise concerns about implementation capacity, threshold effects, and administrative compliance. An alternative might be to target support to young companies instead.

• Extending aid to firms in difficulty may help preserve high-quality jobs and established networks but also risks wasting public money and may be subject to legal constraints (e.g. EU regulations).

• Some interventions enable non-firms to apply (municipalities, cooperatives, or public–private consortia), with implications for complexity, coordination, and risk-sharing.

Across OECD countries in 2023, the scope of targeting in industrial policy instruments varies widely, even within programmes aimed specifically at SMEs and young firms. Table 2 shows that, across 20 OECD countries, the majority of SMEs instruments apply broad eligibility criteria (“Wider” targeting). However, some instruments are designed to target particular types of SMEs (i.e. operating in a particular sector, technology, region or cluster). For example, targeted tax relief measures for SMEs include small‑supplier thresholds to cope with the harmonised sales tax in Canada or specific excise reductions for small breweries in Finland, the United Kingdom and Korea, according to the Quantifying Industrial Strategies (QuIS) database.

• Countries may use investment incentives (tax expenditures, grants, subsidies, public financial instruments) and improve access to inputs (skills, public R&D, knowledge transfer, infrastructure) to enhance firm‑level performance.

• Framework conditions are also shaped by countries to influence industry dynamics. These include capital markets, labour mobility, tax systems, entrepreneurship, intellectual property/standards, plus competition and trade policies.

• Demand-side instruments can also be deployed to influence market demand, including regulation, public procurement, and awareness‑raising or behavioural change.

• Determining the level of support involves balancing effectiveness and distortion risks. In principle, support may aim to provide the minimum necessary to enable the project to proceed, although identifying this threshold can be challenging in practice.

• Eligible cost base: Policymakers need to define which companies’ costs qualify for support (e.g. wages, equipment, subcontracting). Broader eligibility may increase uptake but also fiscal exposure.

• Payment structures, whether upfront, reimbursed as costs are incurred, or linked to outputs, affect firms’ financing constraints and risk-sharing between the public and private sector.

• Decisions regarding the overall envelope and whether funding caps are binding involve trade-offs between flexibility and fiscal discipline.

• Firms’ expectations regarding the continuity of support may influence long-term investment decisions. Governments may weigh the benefits of policy stability against the need to retain flexibility to adapt to changing circumstances.

• The relevance of public programmes depends in part on whether they offer more favourable risk-sharing, pricing, or certainty than private market alternatives, while remaining adaptable to evolving market dynamics.

Patterns of industrial policy instruments use vary significantly across countries. Evidence from the OECD QuIS database indicates that governments employ a broad and diverse mix of instruments in 2023 (the latest year for which data is available), as shown in Figure 3. The number of policy instruments above the reporting threshold varies widely across countries, between 86 in Israel and 232 in France. Across countries, grants and expenditures constitute the most frequently used instrument category. Notable exceptions include Germany, where loans and loan guarantees are a predominant tool, and Türkiye, Canada and Chile, which rely more heavily on tax expenditure instruments.

Expenditure levels and the scale of financing vary markedly across policy instruments in 2023 (Figure 4). The distribution of support indicates that the majority of programmes—across grants, tax expenditures and financial instruments—operate at relatively small scale when measured as a share of GDP. Grant programmes are generally modest in size, with Finland and Czechia recording the highest average spending per grant, at around 0.045% of GDP, while Korea and Türkiye report the lowest averages, spending less than 0.01% of GDP per grant. Tax expenditure instruments show greater dispersion: the United Kingdom and Israel record the highest average expenditure, at approximately 0.07% of GDP per instrument, whereas Estonia reports considerably lower levels, at around 0.002% of GDP. In contrast, loan and loan guarantee instruments tend to be larger in scale, with Türkiye exhibiting the highest average expenditure per instrument, at 0.32% of GDP, followed by Hungary at 0.16%. By comparison, the United Kingdom and the Netherlands allocate less than 0.001% of GDP per loan or guarantee instrument on average. Venture capital and equity instruments are most significant in Israel, where average spending amounts to approximately 0.045% of GDP; however, more than half of the countries in the dataset—including Latvia, Korea, Canada, Sweden, the United Kingdom, Czechia, Türkiye and Slovenia—allocate less than 0.001% of GDP to these instruments.

As mentioned in Part I, multiple instruments can pursue the same objective. New Zealand provides an interesting illustration. Support for its space and advanced aviation sector combines strategic planning through a national Space and Aviation Strategy, targeted innovation funding via the Catalyst Fund, infrastructure investment linked to the Tāwhaki National Aerospace Centre, and a coherent regulatory framework under the Outer Space and High-Altitude Activities Act 2017. This demonstrates how multiple industrial policy instruments can be coordinated and used to develop emerging industries (further details in the Supporting Material).

• Non-discretionary basis: Allocating support based on predefined, objective criteria can enhance simplicity, predictability and perceived fairness, though it may place pressure on budgets if demand exceeds expectations. It also limits the government’s ability to select the best projects, for instance in terms of additionality or non-market benefits.

• First-come, first-served allocation: Processing applications in order of arrival can help manage budget exposure, yet it makes timing a decisive factor, which may favour firms with greater administrative capacity to apply quickly, and can create uncertainty. It can also favour opportunistic submissions, thereby lowering additionality.

• Selective method: Competitive selection can provide greater control over project quality and strategic alignment, but involves higher administrative and compliance costs, and depends on trust in the effectiveness, independence and credibility of the selection process.

• Process design: Regardless of the allocation method, transparency and timeliness are important to reduce uncertainty and fairness concerns. Collaboration with non-governmental agencies (e.g. commercial banks for financial instruments) may facilitate the selection.

• Requiring more comprehensive proposals can improve selection, while raising preparation costs for applicants and processing costs for administrations.

• Two-stage procedures: A phased approach, such as an open call for short pre-proposals followed by invitation-only full proposals, potentially with financial support for SME applicants, can reduce costs for firms and improve the quality of final submissions but may lengthen overall timelines and increase administrative demands.

• Support can be linked to specific conditions or performance metrics (e.g. environmental outcomes, production targets, innovation outputs). Establishing measurable indicators and credible counterfactuals, however, can be complex.

Across countries participating in the QuIS database, the criteria used to allocate loans and loan guarantees to industry vary significantly (Figure 5). Many governments employ selective criteria, particularly in Canada, Chile, Finland, France, Hungary, Ireland, Slovenia, and Sweden, where support is awarded following detailed assessments of project relevance, feasibility, and expected impact. These approaches typically allow administrations to target resources toward projects that best align with national industrial and strategic priorities. For example, Japan’s Programme for Promoting Investment in Japan to Strengthen Supply Chains affected by COVID-19 and the Ukrainian conflict provide a clear example of firm-level selection criteria, such as contribution to supply chain resilience (e.g. reducing dependence on foreign suppliers); feasibility of the project, including financial capacity and technical soundness; effectiveness and efficiency assessed through cost–benefit measures, as well as expected industrial impact and technological validity(further details in the Supporting Material).

In contrast, countries such as Germany and Korea make greater use of nondiscretionary criteria, applying standardised thresholds (e.g. firm size, sector, or basic compliance requirements) that limit discretion and streamline access. This approach provides predictability for firms while reducing the administrative burden associated with case-by-case evaluation.

A first-come/first-served allocation mechanism remains comparatively uncommon for loans and loan guarantees. Only a small number of instruments follow this model: four in Italy, four in Hungary, and one in the Netherlands. In these cases, funding is awarded in order of application until resources are exhausted. Several of these were temporary SME support schemes introduced during the COVID19 crisis; however, some noncrisis instruments also adopt this approach, including SME financing programmes in Italy, innovation credits in the Netherlands, and guarantees for firms pursuing internationalisation in Italy. These instruments illustrate how first-come/first-served approaches tend to be used for relatively broad-based support programmes where administrative simplicity and rapid deployment are prioritised.

• Assessing demand: Demand for a scheme may be gauged through the number of applications received and their quality, which can also provide insight into the visibility and outreach of the programme; the quality and eligibility of applications may signal whether firms have suitable projects, and in the case of financial instruments, whether the terms offered are attractive relative to market conditions.

• Monitoring the selection process: Implementation of the selection process may be assessed through indicators such as processing times, selection rates, and comparisons between the characteristics of selected and non-selected applicants, helping to determine whether procedures operate as intended.

• Tracking project implementation: Progress in project delivery may be monitored by comparing actual timelines to planned milestones and by establishing clear, measurable performance indicators.

• Assessing beneficiary experience: Firm satisfaction with implementation may be proxied through indicators such as feedback or satisfaction ratings.

• Programme costs: Budget management may be tracked by comparing actual expenditures with initial plans and monitoring deviations over time.

• Additionality is achieved when an intervention stimulates new investment, innovation or risk‑taking that would not have occurred otherwise. Concerns arise when there are signs of opportunistic behaviour, for example, firms accessing support without altering their activities or when windfall profits indicate that public funds have not increased outputs. Crowding‑out effects, where public funding displaces rather than complements private investment, also weaken additionality.

• Signs of distorted competition may include the protection of well‑established firms, reduced business dynamism, and discouraged market entry. Distortions can also arise when support sustains lower‑productivity firms, including SMEs, thereby weakening overall national competitiveness.

• While comprehensive impact evaluations can be resource-intensive and time-consuming, continuous monitoring using available administrative and market data may help identify early signs of additionality, potential crowding-out or windfall gains, and possible effects on competition, such as reduced entry or lowering national competitiveness.

Public evaluations of industrial policy instruments remain rare across OECD countries A few general patterns emerge from the evaluations that do exist: well‑designed R&D tax credits and subsidies can stimulate business R&D and innovation, while policies that strengthen skills, knowledge diffusion and technology transfer are important complements. Evidence on the effectiveness of targeted grants and sector‑specific subsidies is still limited and often inconclusive. Strong framework conditions—particularly competitive markets and open, predictable trade policies—are essential for enabling productive firms to expand and for supporting structural change. Demand‑side instruments can also reinforce supply‑side measures by creating markets for innovation and accelerating technology adoption.

If industrial policies are evaluated, then these can be reformulated to better achieve desired goals. For example, evaluations of France’s Crédit d’Impôt Recherche (R&D Tax Credit) between 2009 and 2021 showed that the 2008 shift to a flat rate of 30% of all R&D expenditure successfully halted the decline in business R&D spending. For every euro of public cost, firms spent at least one additional euro on R&D. However, the effects were highly uneven: significant positive impacts were found for SMEs, while the effects for large corporations were statistically insignificant. Incremental design changes to the Crédit d’Impôt Recherche followed to progressively refocus the instrument on SMEs. SMEs are reimbursed immediately while large firms must wait up to three years; likewise, the policy enforced stricter rules on subcontracted R&D, which impacted large firms more.

In Australia, decades of Automotive Industry Assistance, including tariff protection, production grants, and the Automotive Transformation Scheme, were reassessed through successive reviews, culminating in the Productivity Commission’s 2014 inquiry. The evidence showed that long‑running subsidies could not overcome structural cost disadvantages and generated limited spillovers, even as manufacturers sought additional rescue funding. Acting on these findings, the government withdrew further support, and the remaining manufacturers closed by 2017. The case demonstrates how rigorous evaluation can justify ending ineffective support.

Canada’s Scientific Research and Experimental Development (SR&ED) programme underwent a comprehensive review in 2011, revealing a core paradox: despite one of the OECD’s most generous R&D tax incentives, business R&D intensity remained weak. Evaluators found excessive administrative complexity, and disproportionate benefits for large firms relative to outcomes. In response, the government reduced SR&ED tax credit rates, shifted resources toward direct grant programmes, simplified administration, and later implemented a major 2025 overhaul expanding eligibility to capital expenditures and introducing pre‑claim approval. This illustrates how sustained evaluation can drive long‑term, evidence‑based redesign.

Evaluations of the Netherlands’ Top Sectors Policy in 2016 and 2020 found that the strategy strengthened coordination between business and research actors and reduced fragmentation but also created sectoral silos and overlooked crosscutting technologies such as information and communications technology (ICT). In response, the government established Team ICT in 2016 to embed digital innovation across all sectors, reformed governance of the Eureka Clusters in 2020, and subsequently evolved the wider strategy into a mission-driven framework organised around societal challenges. This case shows how regular evaluation can identify structural blind spots and support strategic corrections in complex, multistakeholder industrial policy systems.