Ownership and Governance of State-Owned Enterprises 2024

Annex A. Assessment methodology and sample of jurisdictions

Sources of data and information

Information in this publication draws on existing data collected by the OECD Secretariat from 2022 to 2024. Some of the information was collected in earlier stocktaking work, which was updated along a data verification procedure in 2024. The data verification procedure allowed jurisdictions to review and comment on information available. If necessary and justified with evidence and reform efforts, jurisdiction-specific data was updated to reflect the most recent national de jure and, where possible, de facto SOE landscape.

In total, 59 jurisdictions were surveyed in the data verification exercise, building on information collected through prior surveys undertaken to support past OECD publications: Argentina, Australia, Austria, Azerbaijan, Belgium, Brazil, Bulgaria, Canada, Chile, China (People's Republic of), Colombia, Costa Rica, Croatia, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, India, Indonesia, Ireland, Israel, Italy, Japan, Kazakhstan, Korea, Latvia, Lithuania, Luxembourg, Malaysia, Mexico, Morocco, Netherlands, New Zealand, Norway, Peru, Philippines, Poland, Portugal, Romania, Saudi Arabia, Slovak Republic, Slovenia, South Africa, Spain, Sweden, Switzerland, Thailand, Tunisia, Türkiye, Ukraine, United Kingdom, United States, Uzbekistan, and Viet Nam. The sample size per question or indicator, however, varies depending on the availability of information.

In addition, this publication benefits from OECD work on competition and market regulation conducted in the context of the 2023 OECD Product Market Regulation (PMR) indicators, which aim to measure regulatory barriers to firm entry and competition. A focus is placed on information available that aligns with recommended outcomes as provided in Chapter III of the OECD Guidelines on Corporate Governance of SOEs covering “State-owned enterprises in the marketplace”.

This report includes a benchmarking exercise of national corporate governance frameworks on disclosure and transparency practices of SOEs, and the responsibility of SOE board of directors. The exercise benchmarks jurisdictions against the SOE Guidelines in those areas, covering all EU Member States and several OECD member countries (Australia, Canada, Chile, Colombia, Costa Rica, Israel, Korea, New Zealand, Norway, Switzerland, Türkiye, United Kingdom), and non-member economies (Brazil, Philippines, Ukraine). The information received was assessed in line with a methodology developed by the OECD to support a comparative analysis across the surveyed jurisdictions. The methodology aims to benchmark the level of alignment of national practices with best practices. The level of alignment is evaluated for the policy, regulatory and legislative environment, as well as actual practices bearing on the governance of SOEs. The data from the surveyed jurisdictions is based on self-reporting and is complemented by desk research where necessary. Some self-reported answer items, evidence permitting, were changed by the OECD Secretariat to reflect the national situation as accurately as possible.

The purpose of this methodology is to allow for an “at a glance” approach to present comparative information on national practices of state ownership against best practices advocated by the SOE Guidelines across reporting jurisdictions. The benchmarking exercise further strengthens and deepens the knowledge of the Working Party on State Ownership and Privatisation Practices of national practices and serves as input to future comparative reports that the Working Party may develop. The methodology allows for visual representation of large quantities of qualitative information by way of a heatmap.1 In this way, the comparisons can support implementation of the SOE Guidelines as the internationally agreed standard with recommendations to governments on how to ensure that SOEs operate efficiently, transparently, and in an accountable manner.

Classification of country assessments

The methodology benchmarks national practices into four categories in order of diminishing level of alignment: (1) Fully aligned with best practice; (2) Fully aligned across parts of the SOE portfolio; (3) Partially aligned with best practice; and (4) Not aligned. The description of the four categories is as follows:

• Fully aligned: Implementation of the best practice through legislation, corporate governance codes, and/or company by-laws, to the effect that the national practices of SOEs match the outcomes recommended by the SOE Guidelines and are comparable to those applicable to private listed companies. De facto implementation requires evidence of SOE practices, which is disclosed and/or demonstrated.

• Aligned across parts of the SOE portfolio: Implementation of the best practice, as suggested in the category above, however only applicable to a materially smaller subgroup of the portfolio of SOEs. Such a subgroup can be listed SOEs or selected large SOEs, for example. De-facto implementation requires evidence of SOE practices, which is disclosed and/or demonstrated.

• Partially aligned: Evidence of efforts to implement the best practice, which goes in the direction of the SOE Guidelines’ recommendations. However, it achieves an outcome that is materially different from the recommended outcome. For example, a material difference between the full outcome recommended in the 2015 SOE Guidelines and those of national practices would be that national practices for SOEs are weaker than the comparable requirements for listed companies.

• Not aligned: No evidence of efforts, de jure or de facto, in the national practices that would lead towards the recommended outcome.

The methodology uses two concepts to place countries and their SOE practices into categories explained above. The concept of materiality to determine information material to the achievement of the category; while the concept of equivalence is used to ensure that diverse national practices that achieve the same outcomes are placed as peers in the same category. In this way the methodology is in line with the spirit the SOE Guidelines as an outcomes-based instrument. In other words, “it must be recognised that no one size fits all and different legal and administrative traditions may call for different arrangements. The SOE Guidelines are therefore outcomes-based, meaning that it is the role of governments to decide how to achieve the outcomes that they recommend.”

As national practices differ, the methodology uses the concept of equivalence to assess them. The concept of equivalence is used to compare national practices placed within the same category based on the comparable outcomes they achieve irrespective of the way they achieve it. It thus inserts an outcome-focused component into the assessment methodology. For example, a jurisdiction may achieve the same disclosure and board practice standards through means of respecting non-binding (or “comply or explain”) codes as one that implements binding legislature.

To differentiate between national practices and their level of implementation of the SOE Guidelines, the methodology uses the concept of materiality in differences. The principle of materiality is used to ensure objectivity and consistency. A material difference is one where implementation explicitly departs from the SOE Guidelines in terms of outcomes. The visual below gives a short explanation of the various categories a respondent’s answer to a survey question may qualify under.

Figure A A.1. Classification of national practices

The information collection and assessment process

The methodology draws on information from a sample of OECD members and participants of the Working Party of State Ownership and Privatisation Practices who were surveyed as part of a stock-taking exercise in collaboration with the European Commission’s Directorate-General for Economic and Financial Affairs (DG ECFIN). This information serves as the baseline of the assessment. Further to the information gathering process, Secretariat assessments were based on the methodology of the benchmarking exercise outlined in this Annex. To assess country practices and how they relate to applicable legal or regulatory frameworks, the Secretariat allocated a numerical indicator, representative of a category of alignment with the SOE Guidelines, to each of the responses to the 62 questions covered by the original questionnaire distributed to the sample of countries.

Each question of the survey draws a connection to each of the Recommendations or sub-Recommendation under Chapters of the SOE Guidelines dealing with disclosure and transparency practices of SOEs, and the responsibility of SOE board of directors. Assigning a numerical indicator to each question enables to weigh the assessed categories across each Recommendation and sub-Recommendation. The results of the Secretariat assessments thus detail an assigned category per recommendation, which is underpinned by the assessment of multiple question items latent to it. The figure below summarises the process of categorisation and weighing to arrive at a given category.

Figure A A.2. Process of data gathering and processing to arrive at a final categorisation of country practices

The way in which the numerical indicators and categories are assigned differs slightly depending on the question item, but always apply the categorisation detailed above. For example, to assess responses to question 21, which asked about whether the ownership entity regularly produces a report on the activities and performance of the state-owned enterprise sector (an “aggregate report”), the following assessment procedure was considered to arrive at the final numerical category (Figure A A.3).

Figure A A.3. Assessment process example for question 21 of the questionnaire

Note: Category 3 is assigned if the online inventory only has limited information.

Survey design and data collection

The survey used to collect data is extensive and constructed from 62 multiple-choice and open-ended questions. Some of the questions are lengthy and rich in detail. As a result, answers provided by authorities occasionally overlook aspects of the question item or were not answered in a way that easily offered comparability across jurisdictions. The survey’s questions also necessitated the quoting of legal or other available sources, though the scope of application of said legal or other types of requirements often varies across the overall population of SOEs in a country, and as such the policy, legal or regulatory frameworks quoted must be understood to sometimes cover only parts of the SOE portfolio.

The implementation of the assessment methodology requires sufficient information to identify the correct category a jurisdiction may fall under. In total, information on 41 jurisdictions has been collected for the comparative heatmaps featured in Tables 3.1 and 4.1. In certain instances, survey responses were incomplete and could only be supplemented if information was readily available through desk research available to the Secretariat. Therefore, the stock of high-quality and analytical information available varies across survey jurisdictions. The Secretariat thus invited jurisdictions to complement any missing information that may affect the final categorisation of country practices in April 2023, October 2023, and March 2024.

As part of the assessment, information was sourced from self-reported information by the relevant authorities. This was complemented (when available) by information sourced from various OECD information gathering exercises to verify and expand the knowledge base on national practices. Due to differences of interpretation of OECD questionnaires, sometimes, the answers provided did not fully align with the Secretariat’s assessment of individual country practices. Where the Secretariat assessment may differ from responses provided, jurisdictions were invited to provide additional clarifications to support an informed assessment of national practices. It should, however, be noted that due to limited resources, a third-party verification of the data was not possible.

Secretariat assessments

Every question of the survey fed into the overall assessment of the relevant SOE Guidelines’ recommendations. Results of the question-level assessments were then aggregated to get to an assessment result on the recommendation-level. The aggregation exercise required assigning a numerical score to each category of alignment, which allowed to compute the average result per recommendation. This means that the denoted category of alignment per recommendation is a direct outcome of the weighting exercise. Assessments of a given recommendation are thus supported by several survey questions, however, the amount of questions indicating a given recommendation varies.

The data gathering exercise resulted in an extensive amount of survey response data, clarifications, and additional sources used to form a holistic information base. As mentioned before, the surveyed jurisdictions follow different legal, regulatory and administrative traditions, and employ different ownership arrangements and frameworks that shape the institutional landscape and context for SOEs. For this reason, comparisons across jurisdictions are challenging, as comparisons will not fully reflect diverse systems and approaches to corporate governance of the state-owned sector. While the principle of equivalence in outcomes ensures that different approaches are deemed equal in their alignment with a particular outcome, assessing pathways towards such an outcome also requires nuance and consideration of these differences of approaches across jurisdictions.

Note

← 1.  Such representations are widely utilised in the literature, especially related to monitoring of international practices as in the following two examples: (i) OECD / FATF uses a heatmap to illustrate the consolidated assessment of its country members implementing the FATF Recommendations (FATF, 2024_[31]); while (ii) the BIS uses a heatmap to illustrate its monitoring of country adoption of the Basel III standards (BIS, 2024_[32]).