Introducing proportionality to Thailand’s regulatory impact assessment framework

2. The proportionality test for Thailand

The current situation in Thailand

The government of Thailand has made it a priority to reform its GRP framework, including for ex ante RIA and ex post review, in order to move towards OECD best practice. The OECD is carrying out a parallel Regulatory Policy Review of Thailand which shall analyse and produce recommendations for Thai GRPs. The motivation for this project stems from the fact that the OCS, in its statutory role as a ROB, has been receiving high numbers of primary bills and draft secondary legislation from ministries and agencies, for RIA quality review and legal check:

• In 2022, there were 31 primary bills and 224 pieces of draft secondary legislation.

• In 2023, there were 15 primary bills and 221 pieces of draft secondary legislation.

In the absence prioritisation, or proportionality test, it is difficult for the OCS to plan how much scrutiny to give a draft law or regulation, and how much time the Cabinet of Thailand should dedicate to discussing them. Furthermore, it is difficult for ministries and agencies to plan for how much detailed analysis that ex ante RIAs, or even ex post reviews, should contain. An overuse of RIA may create the typical “paralysis by analysis” phenomenon or, conversely, spread constrained resources across an excessive number of initiatives. The OECD has found that international experience suggests that full RIAs might not constitute more than 5-10 percent of the overall RIAs carried out by the government in a given year (OIRA, n.d.[9]). Rather than being viewed as another bureaucratic step, a proportionality test is a tool that is designed to efficiently allocate analytical resources to policy proposals with the highest impacts. A well implemented proportionality framework can help administration reduce the amount of full RIA conducts and produce higher quality full RIA in the narrow circumstances where in depth analysis is required.

The Act on Legislative Drafting and Evaluation of Law (2019) implemented Section 77 of the Constitution, by legislating for ex ante RIA, stakeholder engagement and ex post review of draft laws. The Act stipulated a generalised scope of application of RIA for all draft primary legislation. Whilst secondary legislation was initially excluded from the scope of RIA, a Ministerial Regulation in 2022 extended GRPs to certain subordinate regulations that have direct effects on people’s lives. However, Thai government ministries as well as the OCS do not presently have access to tools or methodologies to help them understand which draft laws or regulations will have significant impacts on businesses or the lives of citizens early on in the policy cycle so that analytical resources can be efficiently allocated. A proportionality test will provide such a tool.

The 2019 Act further specifies that the ex post review requirements do not apply to limited-time expired legislation or those relevant to symbolic representations, such as those for academic accreditations, neither laws relating to the restructuring of government ministries. Section 29 also states that “Legal Codes” nor “other laws as prescribed in the ministerial regulation” are subject to ex post assessment. The current ministerial regulation under this clause exempts (1) palace laws, (2) laws concerning the administration of religious organisations, (3) laws concerning the personnel management of government officials, and (4) laws concerning immunities and the protection of international organisations and meetings in Thailand.

Furthermore, section 34 of the 2019 Act requires that all applicable laws must be evaluated every five years or if: 1) there is a petition from citizens or organisations to review the law and the responsible state agency agrees; 2) the state agency receives a recommendation from the Law Reform Commission; or 3) other cases as prescribed in ministerial regulation. The completed ex post review report is published in the central system and is shared with the Law Reform Commission. The Thai legislative process could benefit from more robust forward planning, the current practice lacks the necessary time for ministries to carry out ex ante RIA and comprehensive stakeholder engagement at a sufficiently early point in the legislative process. The procedures of setting the Government agenda and kicking off the regulatory process rest mainly with individual regulators. While stakeholders that the OECD met with noted that they understand the procedural aspect of law making in Thailand, some stated that they do not have a strong sense of government priorities. They noted that not all planned government and ministerial initiatives are announced on systematic and timely manner. Stakeholder said that they are sometimes unaware of the purpose of a given law in the Thai context other than that it may emulate similar rules in other nations.

Proper forward planning would provide an opportunity for regulatory management tools to better inform the development of legislative proposals and avoid a natural bias many governments tend to have toward legislative solutions for policy problems. As mentioned previously, most OECD countries have instituted some form of forward planning in their rule-making processes.1 These processes give ministries a chance to plan where they should channel their analytical resources and ROBs the opportunity to focus their scrutiny on high-impact legislation.

A further element of concern is the way in which urgent and emergency procedures are invoked to advance regulatory dossiers. If recourse to such procedures is not robustly monitored by a regulatory oversight body, it may undermine the application of established due process standards, predictability, and accountability.

Implementing a proportionality test would provide Thailand an opportunity to enhance the quality of the RIA reports by focusing its analytical efforts on proposals that have most significant impacts on people, businesses, and create greater costs to public administration.

Recommendation for a proportionality test

To improve the implementation of RIA in Thailand, the OECD recommends the Thai administration to introduce a triage system to ensure of the proposed study of a given regulation is proportional to its estimated impacts. A triage system would ensure the proper allocation of resources and time needed to support the impact assessment. With the introduction of this triage system, we would expect to improve the quality of the RIA by achieving a more efficient and effective use of internal resources. The implementation of this system would imply the following steps:

• Identify the selection criteria – define the threshold for which the estimated qualitative and quantitative impacts a proposed policy must reach in order for an in-depth RIA and the commensurate resources to be required.

• Establish the selection procedure based on the defined criteria – define the process by which the above criteria will be assessed, including the responsible entity (e.g., the ministry enforcing the law) and the standard if evidence required, which should not be too onerous at the triage stage.

Given this policy will be nascent in the Thai administration and - based on other work the OECD has conducted with Thailand - the administration may also lack robust analytical capacities in many of its ministries and agencies.2 As the Thai administration matures in its application of RIA and a triage system, it should consider implementing higher evidentiary standards and requiring more depth of analysis. In the short term the OECD recommends that the Thai administration focus its limited analytical capacity on only the highest impact policies. As such, in the short term, the Thai administration should not conduct RIA on law or regulations triaged to be low impact (stage 2.a below). In the medium to long term, as the Thai administration builds its analytical capacity the OECD recommends it begins to conduct a “light” RIA on low impact proposals (stage 2.b below).

Following the implementation of a triage system, the OCS in its role as ROB who verifies the assessment by the respective ministry, should be able classify the RIA requirement for a given policy under one of the following categories:

1. An exemption from RIA may be allowed (e.g. national security proposals).

a. No RIA, the triage document itself can act as a “light” RIA to provide some insight into the estimated effects.

b. A “light” RIA will be able to provide better insight into the effects of a proposed regulation that is estimated to have impacts below the designated threshold.

2. A “full” RIA might be required due to the estimated high impact of the proposed regulation. Proposals that undergo “full” RIA should also be prioritised for comprehensive ex post review.

The “light” RIA would refer to the basic evaluation that corresponds to what is currently being carried out within the framework of the Thai RIA framework. A “full” RIA would be a more complex evaluation that could in time aspire to include more intensive RIA methodologies, for example, cost-effectiveness analysis or cost-benefit analysis of legislative proposals. An overview of this triage process is illustrated below in Figure 4and discussed in more detail in the following sub-sections.

Figure 4. Overview of a RIA Triage System

Source: Author’s elaborations.

Criteria for an exemption from RIA

This paper suggests some qualitative criteria to assist ministries in performing an initial “triage” of legislative proposals, to help to identify proposals that could be exempted or excluded from a “full” RIA requirement. Some of the exemptions are already stipulated in the 2019 law and others have been added based upon OECD best practice.

Exemptions from conducting RIA would be admissible in the following examples of legislative proposals:

• Emergency Decrees: These are currently listed as an exemption in the 2019 Act and are issued only in the case of an emergency of necessity and urgency which is unavoidable and for the purpose of maintaining national or public safety or national economic security or avert public calamity. The Act states that an evaluation of these decrees should take place within two years from the date such Emergency Decree comes into force. This is currently listed in the 2019 Act.

• Proposals that are sensitive to national security. This is currently listed in the 2019 Act.

• Proposals that do not have effects on people. This is currently listed in the 2019 Act.

• Changes in governmental internal organisation or structure that will not imply the creation of new bodies or administrative obligations. This is currently listed in the 2019 Act.

• The reorganisation of existing legal texts, or technical changes to existing laws or regulations (with no new obligations created for people or businesses). This would be a new exemption.

• The direct transposition of international treaties with no optionality in implementation methodology into the domestic framework. This would be a new exemption.

• Government budgetary proposals that are not creating any new regulatory or legal obligations on people or businesses. This would be a new exemption.

The qualitative criteria would indicate situations where it may be possible to exclude a proposed law from the obligation to conduct RIA. The procedure for this exclusion should be accompanied with brief analysis from the ministry. The OECD recommends that the OCS should assess and evaluate the rationales for excluding a RIA, and should be able to adopt one of three positions:

• accept the exclusion;

• accept the exclusion subject to an ex post evaluation after two years;

• or, recommend rejecting the exclusion.

If a draft law is exempted from conducting a RIA, the OCS should work with the Law Reform Commission to ensure that the systematic implementation of the existing provisions in the 2019 Act, which stipulate that in the case the law was passed as an emergency decree, an evaluation should be carried out two years after its date of enforcement.

Proportionality rule: Economic, environmental and social impacts

If the exemption criteria are not met, a triage questionnaire should always be developed by a policy team. This questionnaire would form the key tool for implementing the proportionality process by focusing on estimates relating to the potential economic, social, and environmental impacts of a draft regulation as well as the timeline over which these effects are expected to be felt, to determine if a “full” RIA is required.

An example of a template for this questionnaire is set out in Table 1 and an overview of how the triage process would fit into the Thai regulatory procedure is depicted in Annex A. This questionnaire asks a series of questions based on a set of clear criteria to serve to determine the likely impact of a draft regulation, whilst also providing a basic understanding of the rationale and expected impacts of all regulatory proposals. As the OECD understands, Thai Ministries and agencies must submit their legislative proposals to the Office of Cabinet at an early stage in the policy cycle. This would likely be the best stage at which the OCS would have an opportunity to review and approve the triage document, including whether the policy is exempt, and if the appropriate level of analysis is proposed by the responsible ministry. For added clarity, OECD best practice considers RIA to be a tool to help select the best policy tool to address a given issue – at this early triage stage the responsible ministry, the Office of Cabinet, and the OCS should be open to various regulatory and non-regulatory options to address the policy issue at hand (OECD, 2020[2]).

The answers to the triage questions should be based on readily available information, not in-depth analysis (unless it has already been done for other reasons). The level of detail and specific information required in a full RIA is not required at triage stage. Based on the responses of the triage questionnaire, the responsible ministry will self-assess whether a full RIA needs to be developed. Responses to the triage questionnaire must be shared with the OCS, who will evaluate the responses and determine whether the responses are based upon a sound rationale, and whether the ministry’s proposal for an exemption or their proposed depth of analysis is appropriate. The ministry may choose to proceed with an in-depth impact study should the triage questionnaire indicate, at an early stage, that a full RIA is necessary.

For each of the three policy areas on the triage questionnaire, policy teams will evaluate the expected effects of a proposal. If the proposal has no impact in the given policy area, they can mark the box “No/NA” box. If the proposal does have an effect in the given policy area, practitioners should rank the effect in the respective area on a scale of 1 (low impact) to 4 (high impact). The use of the word “impact” in this document refers to both positive and negative impacts. For the purposes of the triage document and in order to manage resources efficiently, the impact (whether positive or negative) should be described as an estimate using any quantitative figures or readily available data. In depth study should only be required in later stages of RIA, when a full RIA is deemed necessary).

• No Impact/Not Applicable: The proposal may have no impact or is not applicable to the areas covered by the question. In these cases, no description of the impacts is required.

• Low (1-2): The proposal may have minimal impacts on the areas covered by the question (e.g. routine or administrative in nature, generally acknowledged as acceptable to the public, and would have negligible impacts on public health and safety, the environment, business, government, etc.). Moderate impacts of a law are expected to be short term.

• High (3-4): The proposal may have significant impacts on areas covered by the question (e.g., is highly controversial, would represent a very significant change to the status quo, heavily impact public health and safety, the environment, business, government, etc.). Moderate impacts of a law are expected to be long term.

As it gains more experience in reviewing triage documents and eventually see proposals through to the full RIA process, the OCS may consider issuing guidelines for quantitative or qualitative thresholds for scoring (1-4) in a given policy area.3 Providing such guidance will likely be more difficult for strictly qualitative indicators but through experience the OCS can work with ministries to better define qualitative thresholds and provide some examples it has seen in its experience.

The OECD recommends that the responsible ministry be required to do a “full” RIA if the overall score is 7 (or higher) out of 12, or if any given criteria see the maximum impact (4). In the case of non-exempt policy issues that do not reach the threshold for “full” RIA, step 2.a (no RIA) in the near term or 2.b (“light” RIA) in the medium or long term should be engaged as discussed above.

As the regulatory oversight body, the OCS should play an important role in reviewing and approving the criteria the criteria selected by the ministry. It is understood that at the triage stage, ministries are using their best estimates based on experience and easily accessible data to estimate the impacts in each section, however, this process requires oversight to ensure standards are applied evenly across the administration. While OCS will not have the subject matter expertise to thoroughly assess each document, it should have the authority to request the available evidence that was used to support the triage questionnaire and meet with ministries to seek clarification when necessary.

Finally, the OCS, in its capacity as ROB, should conduct regular reviews of the proportionality process, to determine whether it is operating effectively. These reviews could contain information like:

• How many RIAs were undertaken in a given time period;

• How many RIA exemptions were granted in a given time period;

• Under what categories were the exemptions granted;

• How many “full” vs “light” RIA were conducted (after stage 2.b is implemented); and

• The average impact quantitative impact of policies that have undergone “full RIA”.

The first such review could take place two years after the initial implementation of the process and provide the Cabinet with a report on the effectiveness of the proportionality framework and whether further reforms are required.

Applying the proportionality test to ex post reviews

For new laws and regulations published by the Thai administration after it implements a proportionality test to its RIA framework, the same proportionality standard for ex ante RIA should apply to ex post assessments. Policies that were deemed to be high impact and subject to “full” RIA at the initial triage stage should also receive comprehensive ex post reviews. Given that these policies would have been subject to “full” RIA before becoming law, there should be measurable objectives already set out for ease of comparison. Furthermore, data requirements and collection strategies to facilitate ex post review should have been outlined at the ex ante stage which would further facilitate ex post review (OECD, 2020[2]).

Given that all policies must be reviewed every five years according to the 2019 Act, policies that are initially deemed to be low impact may not have benchmarks and data for ease of comparison, especially as the Thai administration remains under step 2.a outlined above (no RIA for low impact policies). While Thai RIA practitioners should use best efforts and available data to quantify effects of a given regulation ex post, it is still important to apply a proportionality principle and not allocate scarce resources to low impact legislation. In the case of policies that were initially deemed low impact but have had unintended or larger impacts than initially estimated, strategies involving stakeholders such as those discussed in Box 3 above can prove useful.

While the above two scenarios deal with ex post reviews where the ex ante RIA was conducted after the implementation of a proportionality standard, there of course remains a significant stock of regulation subject to ex post review that did not undergo a proportionality triage. In this scenario the OECD recommends that the responsible ministry or agency complete the proportionality triage as it would for an ex ante RIA but with a view to ex post review. This triage document can then be submitted to the OCS for review and approval on the level of ex post review to be conducted. Given that ministries and agencies should have ample time to complete and submit this triage to the OCS for approval (e.g., it could do so several months before the five-year review is triggered by the 2019 Act), it should be unlikely to this process presents a barrier to agencies meeting their statutory review requirements. This same process would also be helpful in the case of emergency regulations deemed exempt from RIA which, as recommended above, should be reviewed two years after implementation (in this case Section II of Table 1 would not apply).

Table 1. Example of Triage Questionnaire Template

SECTION I: OVERVIEW
Problem Describe the problem(s) and demonstrate why government intervention is needed. Objectives State the objectives for government intervention in concrete terms and its broader policy context. Description A preliminary description of the regulatory action(s) under consideration.
SECTION II: Exemption
Does the issue being analysed fall into one of the RIA exempt categories?4 If yes please explain which exempt category the issue falls into and why. If not exempt, please continue with Section III below.	Yes	No
SECTION III: ESTIMATED IMPACTS Economic Impacts
If a regulatory proposal is expected to have no economic impacts or implications (e.g. economy, business including administrative burden and duplication, consumers, competition and internal trade) or is not applicable, it receives a No/NA rating. If a regulatory proposal is expected to have minimal economic impacts, it receives a low rating; if it is expected to have significant impacts, it receives a high rating. Practical issues to be considered reflecting economic impacts Does the policy create regulatory burdens on businesses or citizens? Does the policy create regulatory burdens that will have adverse or positive impact on small and medium sized businesses? Does the policy facilitate competition or create new barriers to market entry? Does the policy enhance or cause harm to consume welfare e.g. price levels? Does the policy discriminate against or otherwise impede foreign investment and foreign ownership or foreign supply of services?	No/NA	1-41
Describe the expected impacts or provide a brief justification of why there are none.
Environmental impacts
If a regulatory proposal is expected to have no impact on the environment or is not applicable, it receives a No/NA rating. If a regulatory proposal is expected to have minimal impacts, it receives a low rating; and if it is expected to have significant impacts (e.g., damaging or protecting a sensitive ecosystem from irreversible harm or damage), it receives a high rating. Practical issues to be considered reflecting high environmental impact: Does the policy affect the emission of ozone depleting substances (CFCs, HCFCs etc.)? Does the policy affect the emission of greenhouse gases sinks and reservoirs? Does the policy affect our ability to adapt to climate change and to deploy mitigation measures? Does the policy allow to accomplish with environmental damage reparation duties? What are the consequences for energy usage and mobility, waste and emissions into the air, soil and surface water and also for use of the available physical space? What are the impacts of the proposed regulation on circularity? Does the policy involve trade-offs with other resources or ecosystem services, considering the full supply chain and all environmental impacts? (e.g. using less land but more water to produce the same quantity of crops) Does the policy encourage substitution of high-impact resources by resources with less impact on the environment? Does it affect protected or endangered species or their habitats or ecologically sensitive areas?	No/NA	1-4
Describe the expected impacts or provide a brief justification of why there are none.
Social impacts
If a regulatory proposal is expected to have no social impacts or implications (e.g., changes to people’s way of life, culture, community, political systems, well-being, personal and property rights, fears and aspirations or raises ethical concerns) or is not applicable, it receives a No/NA rating. If a regulatory proposal is expected to have minimal impacts it receives a low rating; if it is expected to have significant impacts, it receives a high rating. Practical issues to be considered when considering high social impact: Does the policy affect inequalities and the distribution of incomes and risk of poverty? Does the policy affect the access to and quality of social protection benefits, including social services of general interest, particularly for those subject to social exclusion and from disadvantaged backgrounds?	No/NA	1-4
Describe the expected impacts or provide a brief justification of why there are none.
Overall impacts
If the policy issue does not fall into an exempt category a “full” RIA is triggered if:2 The cumulative score of the three categories is 7 or greater (out of 12); or Any one category receives a score of 4.	Total

1. As it gains more experience in reviewing triage documents and eventually see proposals through to the full RIA process, the OCS may consider issuing guideline quantitative or qualitative thresholds for respective scores (e.g. for estimated economic impacts: 1 = <2M Baht; 2 = 2M – 10M Baht; 3 = 10M – 35M Baht; 4 = >35M Baht).

2. In the event a policy issue is not exempt but does not trigger a full RIA it should follow the step 2 (2.a in the near term or 2.b in the medium to long term), outlined above in the section: Recommendation for a Proportionality Test.

Source: Author’s elaboration.

Notes

← 1.  The various challenges facing the implementation of GRPs in Thailand, including lack of forward planning, are analysed in detail in the OECD Regulatory Review of Thailand (OECD, 2025[10]).

← 2.  For more information see the Regulatory Policy Review of Thailand (OECD, 2025[10]).

← 3.  For example, for estimated economic impacts: 1 = <2M Baht; 2 = 2M – 10M Baht; 3 = 10M – 35M Baht; 4 = >35M Baht, a similar mechanism could work in other quantifiable areas such as emitted tons of CO₂.

← 4.  The OCS may consider inserting a list of exempt categories once this list is finalised, it may further consider refining the justification question or requesting evidence as they gain experience with this process.