## REINVIGORATING THAILAND’S STRUCTURAL TRANSFORMATION

## Source details

**Canonical URL:** [REINVIGORATING THAILAND’S STRUCTURAL TRANSFORMATION](https://www.imf.org/-/media/files/publications/cr/2026/english/1thaea2026002-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2026/english/1thaea2026002-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2026/english/1thaea2026002-source-pdf.pdf.json)

---

### Background
- Structural transformation defined as reallocation of labor from lower- to higher-productivity sectors.
- Key stylized facts for Thailand:
  - “approximately one-third of Thai workers are still employed in agriculture.”
  - Employment in industry and services is relatively low for Thailand’s income level.
  - Within services, a disproportionately large share of workers are employed in low-productivity services (trade services and other services); employment in high-productivity services (including finance and business services) remains relatively low.
- Imbalance between sectoral labor and output shares:
  - Industry and high-productivity services account for the majority of value-added.
  - Low-productivity services and agriculture contribute substantially less to value-added but employ larger shares of labor.
  - Result: a wide and growing labor productivity gap across sectors.
- Implication: scope for productivity gains by reallocating workers to higher-productivity sectors.

### Assessing Thailand’s Structural Transformation (measurement and patterns)
- Measurement approach:
  - Covariance between sectoral employment share and labor productivity used to quantify structural transformation (following IMF (2013) and Klyuev (2015)).
  - Covariance term defined (chapter) as:
    퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶 = �(푠푠
    푖푖
    −푠푠̅
    푛푛
    푖푖
    )
    (
    푝푝
    푖푖
    −푝푝̅
    )
  - Simplified identity presented: �(푠푠
    푖푖
    −푠푠̅
    푛푛
    푖푖
    )
    (
    푝푝
    푖푖
    −푝푝̅
    )
    = �푠푠
    푖푖
    푝푝
    푖푖
     −  
    푛푛
    푖푖
    �푠푠
    푖푖
    푝푝̅ 
    푛푛
    푖푖
    =푃푃− 푃푃
    �
  - Interpretation: covariance > 0 when more productive sectors employ larger shares of labor; covariance < 0 when less productive sectors employ larger shares of labor.
- Data and classifications:
  - Data: UNU-Wider Economic Transformation Database (ETD) covering 1990–2018 for cross-country comparisons.
  - Covariance computed for both 3-sector (agriculture, industry, services) and 4-sector (agriculture, industry, low-productivity services, high-productivity services) classifications.
- Cross-country evolution:
  - Most selected Asian countries exhibit negative covariance historically but many improved over time.
  - Korea and Malaysia: covariance close to or above zero.
  - Chinese mainland: substantial improvement approaching Korea and Malaysia levels.
  - Vietnam: dramatic gains.
  - Thailand: progress in early 1990s, but momentum slowed since mid-2010s; covariance plateaued and slightly declined in recent years.
  - Sectoral disaggregation matters: Thailand’s covariance relatively high in 3-sector view but significantly lower in 4-sector view, revealing an internal services imbalance.

### Drivers of Structural Transformation (empirical model and main results)
- Empirical setup:
  - Dependent variable: scaled covariance term.
  - Estimation: cross-country panel fixed-effects with year dummies.
  - Baseline independent variables included: Log GDP per capita and its square; Log population; old-age dependency ratio; average years of schooling; Human Capital Index; trade-to-GDP ratio; FDI flows as % of GDP; terms of trade index; REER; agricultural raw materials exports share; vulnerable employment share; domestic credit to private sector as % of GDP; Political Corruption Index (V-Dem, rescaled ×100).
  - Policy variables lagged by one year (except GDP per capita, population, old-age dependency ratio).
  - Robustness checks: Labor Market Regulations index; Gini Index; Internet access; disaggregated education enrollment ratios; Rule of Law replacing Political Corruption Index.
- Key qualitative findings:
  - Log GDP per capita: strongly positive; quadratic term negative — improvement with income but diminishing returns.
  - Log population: strong positive — larger population associated with greater labor allocation efficiency.
  - Old-age dependency ratio: positive correlation.
  - Terms of trade and trade openness: positive contributions to higher covariance.
  - REER appreciation: associated with lower covariance.
  - FDI inflows: not statistically significant (heterogeneous effects).
  - Average years of schooling: positive and significant across specifications.
  - Human Capital Index: not statistically significant.
  - Vulnerable employment (informality): negatively correlated with covariance (significant in some specs).
  - Share of agricultural exports: negatively associated with covariance.
  - Private credit-to-GDP ratio: positively correlated with covariance.
  - Corruption (rescaled Political Corruption Index): negatively associated with covariance across all specifications.
  - Labor market regulation (alternative spec): better regulation associated with higher covariance.
  - Education detail: primary and secondary enrollment more strongly correlated with covariance than tertiary enrollment.
  - Internet access: positive coefficient but not statistically significant.
  - Rule of Law: stronger rule of law positively associated with structural transformation.

### Thailand-specific decomposition of recent slowdown
- Method:
  - Use estimated model to generate fitted covariance and compare with actual covariance.
  - Compute changes in driver values between 2010–2014 and 2014–2018, multiply by estimated coefficients to obtain contributions to covariance change.
- Observed shifts (2010–2014 vs 2014–2018):
  - Private sector credit growth slowed markedly in the latter period.
  - Progress in trade openness reversed.
  - REER depreciated.
  - Share of vulnerable employment declined.
  - Terms of trade improved.
  - Additional changes: acceleration in old-age dependency ratio; reversal in FDI growth; slower growth in agriculture export share; more moderate increase in average years of schooling; deterioration in governance (higher corruption); slower population growth.
- Net effect:
  - These changes produced increasing downward pressures on Thailand’s structural transformation since the mid-2010s, contributing to the observed plateau and slight decline in the covariance term in recent years.

### Findings and policy-relevant implications
- Empirical findings:
  - Thailand exhibits an inefficient labor allocation: agriculture and low-productivity services employ large labor shares despite low value-added contributions.
  - Structural transformation in Thailand slowed since the mid-2010s; slowdown linked to shifts in credit growth, trade openness, governance, population dynamics, and other factors.
  - Education (especially primary and secondary), trade integration, credit access, and governance quality empirically linked to stronger structural transformation.
- Policy implications emphasized:
  - Facilitate labor reallocation toward higher-productivity sectors to boost aggregate labor productivity.
  - Strengthen trade openness and competitiveness (address REER appreciation pressures) to support tradable sectors.
  - Improve access to private credit to enable investments that shift labor into higher-productivity activities.
  - Address informality (vulnerable employment) to enable better labor mobility and productivity gains.
  - Invest in education—particularly primary and secondary attainment—to support structural transformation.
  - Improve governance and reduce corruption; strengthen rule of law to remove impediments to labor reallocation.
  - Consider labor market regulations that facilitate efficient movement of labor across sectors.

### Quantitative evidence (selected regression coefficients)
- Appendix I, Table 1 (4-Sector Covariance), selected estimates (Driscoll–Kraay SEs):
  - Log GDP per capita: 7.7069 *** (1.4912)
  - Square of Log GDP per capita: -0.4075 *** (0.0783)
  - Log population: 2.5526 *** (0.2912)
  - Old-age dependency ratio: 0.0768 *** (0.0142)
  - Terms of trade (lagged): 0.0039 *** (0.0008)
  - REER (lagged): -0.0017 ** (0.0007)
  - Average years of schooling (lagged): 0.1549 *** (0.0390)
  - Domestic credit to the private sector to GDP (lagged): 0.0039 *** (0.0010)
  - Political corruption index (lagged): -0.0045 *** (0.0014)
  - Share of agricultural raw material exports (lagged): -0.0132 ** (0.0063)
  - Share of vulnerable employment (lagged): -0.0088 *** (0.0020) in Spec 2
  - Observations: 865; Countries: 44; Within R^2: 0.624
- Appendix I, Table 2 (3-Sector Covariance), selected estimates:
  - Log GDP per capita: 6.5674 *** (1.3976)
  - Square of Log GDP per capita: -0.3408 *** (0.0727)
  - Old-age dependency ratio: 0.0597 *** (0.0128)
  - Terms of trade (lagged): 0.0028 *** (0.0007)
  - Domestic credit to the private sector to GDP (lagged): 0.0028 *** (0.0006)
  - Share of agricultural raw material exports (lagged): -0.0115 ** (0.0042)
  - Observations: 865; Countries: 44; Within R^2: 0.624

---

### Digital Cross-Border Payments and QR Linkages — Thailand and ASEAN

### Regional and institutional context
- G20 Roadmap for Enhancing Cross-border Payments addresses frictions: high costs, slow transaction speeds, limited accessibility, lack of transparency.
- ASEAN responses:
  - ASEAN Regional Payment Connectivity (RPC): link national payment systems for integrated cross-border payment ecosystem.
  - ASEAN Local Currency Transaction Framework (LCTF): promote local currencies for cross-border transactions to reduce costs and speed transactions.

### Thailand’s progress in digital payments
- Penetration and usage:
  - Mobile phone ownership in ASEAN: over 80 percent.
  - E-money account ownership in ASEAN: about 20 percent.
  - In Thailand:
    - close to half of the adult population owns an e-money account.
    - about 80 percent of adults own financial accounts.
    - 55 percent of adults own a debit card.
    - 50 percent of adults have made digital payments.
- Growth of non-physical payments in Thailand:
  - Use of non-physical payments increased annually by more than 75 percent on average during 2019-24.
  - PromptPay adoption: average fast payment transactions through PromptPay rose more than eight-fold from below 40 per person per year to almost 350 during 2019-24.
- Cross-border linkages leadership:
  - Thailand established first QR payment linkage with Japan in 2018 and later with eight other economies.
  - April 2021: Singapore and Thailand established first cross-border fund transfer connectivity.
  - Thailand is a first mover in linking digital payments across borders among ASEAN peers.

### ASEAN-wide developments and scale
- Private fintech investment:
  - Fintech funding to ASEAN surged since 2014, peaked in 2021, then declined thereafter.
  - Singapore is a key hub, followed by the Philippines and Indonesia.
- E-money usage examples:
  - Malaysia: total value of e-money payments accounted for more than 8 percent of annual GDP.
  - Singapore: about 300 e-money transactions per person in 2024.
- Asia’s share of global cross-border payments:
  - Asia accounted for 32 percent of global cross-border payments in 2024 and is expected to account for 37 percent in 2032.
  - Total volume of cross-border payments in Asia expected to increase from 12.8 billion in 2024 to 23.8 billion in 2032.
- Remittance costs and speed:
  - Average cost of remittances in East Asia and Pacific economies declined from 6.7 percent in 2021Q1 to 5.8 percent in 2025Q1.
  - Asia lags other regions on percentage of cross-border payments credited within one business day.

### Bilateral linkages, multilateral initiatives, and CBDC experimentation
- Types of linkages:
  - QR payment connectivity: real-time merchant payments by scanning QR codes across borders.
  - Fund transfer connectivity: instant remittances and fund transfers using simple identifiers (mobile phone numbers or national IDs).
- QR growth and usage:
  - 2024 examples:
    - QR payments in Thailand increased by more than 300 percent in 2024 compared with 2023.
    - QR payments in Malaysia increased by 550 percent in 2024 compared with 2023.
    - Inbound payments through all seven QR linkages in Thailand in 2024 were about five times that in 2023.
- Project Nexus:
  - Collaboration: BIS Innovation Hub, Bank of Thailand, Bank Negara Malaysia, Bangko Sentral ng Pilipinas, Monetary Authority of Singapore, Reserve Bank of India.
  - Proof of concept completed in 2022; comprehensive blueprint announced July 2024.
  - Nexus Global Payments incorporated in Singapore aiming to go-live by 2027.
- CBDC status:
  - Majority of ASEAN countries launched CBDC research since 2016; several moved to experiments and pilots since 2020.
  - No ASEAN country has officially launched CBDC for public use as of data presented.
  - Singapore and Thailand most advanced in development stage.
- Project mBridge (wholesale CBDC initiative):
  - Participants include Bank of Thailand, Central Bank of the UAE, Digital Currency Institute of the People's Bank of China, Saudi Central Bank, Hong Kong Monetary Authority.
  - Six-week pilot in 2022 involved 20 commercial banks across four jurisdictions, executing over 160 real-value transactions.
  - Project reached Minimum Viable Product (MVP) stage in mid-2024; ongoing development and real-value transactions continue subject to participating jurisdictions’ conditions.

### Benefits for SMEs and tourism linkages
- SME adoption:
  - Bank of Thailand survey: 96 percent of SMEs had adopted digital payments accounting for over two-thirds of their payment value.
  - Digital payments enable SMEs to build digital financial footprints and provide transaction data to assess creditworthiness.
  - PromptBiz (introduced August 2023) supports SMEs to access finance using PromptBiz trade transactions as supplemental documents.
- Tourism:
  - Intra-ASEAN tourists in 2023 accounted for 42 percent of total visitors, up from 36 percent in 2019.
  - Tourism accounted for 8 percent of the region's GDP and 12 percent of employment in 2023.
  - Expanded cross-border digital payment use could have significant positive effects on the region’s service sector.

### Risks, fragmentation, and multilateral efficiency
- Financial crime and AML risks:
  - Rapid adoption introduced significant financial crime and money laundering risks, acute in cross-border transactions due to regulatory inconsistencies and instant payments.
  - Criminal techniques: micro-structuring, identity theft, use of mule accounts; risk heightened by scam centers and transnational criminal operations from high-risk countries.
- Fragmentation:
  - Multiple bilateral linkages create a highly fragmented system, complicating interoperability and increasing resource costs.
- Multilateral approaches:
  - Project Nexus and hub-and-spoke models can simplify infrastructure and resolve interoperability challenges more efficiently than many bilateral linkages.

---

### Empirical Analysis: Cross-Border QR Payments to Thailand (econometric model and results)

### Model specification and data
- Dataset:
  - Payments from seven economies: Cambodia, Hong Kong SAR, Indonesia, Laos, Malaysia, Singapore, and Vietnam to Thailand.
  - Period: 2020-2024, monthly aggregate; 60 months; total observations: 420 (by volume and value).
- Dependent variables: volume and value of cross-border QR payments at bilateral economy-pair level.
- Key explanatory variables:
  - FX_volit: volatility of local currency vs US dollar and vs Thai Baht.
  - FX_git: monthly depreciation of local currency vs Dollar and Baht.
  - flightit: number of flights per month (proxy for tourists).
  - Monthit: months the linkage has been established.
  - Commer_bankiT: number of commercial bank branches.
  - Credit_cardiT: number of credit cards.
  - CiT: macro indicators (GDP and CPI).
- Model (log specification):
  - QRit = β1 Monthit + β2 flightit + β3 FX_volit + β4 FX_git + β5 Commer_bankiT + β6 Credit_cardiT + CiT + εit

### Key empirical findings (summary)
- FX volatility and depreciation:
  - Higher local currency-USD volatility associated with higher QR payments (value) and weaker effect on volume.
  - Local currency-THB volatility does not appear to affect QR payments materially.
  - Local currency depreciation against the US dollar and the Thai Baht associated with lower QR payments (value and volume).
- Traditional banking substitutes:
  - Number of commercial bank branches and number of credit cards tend to be negatively associated with QR payments in most regressions — QR payments substitute for traditional banking and card payments when access to traditional services is lower.
- Adoption and tourism drivers:
  - Month of implementation and log(number of flight) are positively associated with both value and volume of QR payments.
- Robustness:
  - Results robust when currency depreciations excluded and when volatilities against USD and THB included separately.

### Selected estimated coefficients (table highlights; N 420)
- R^2: 0.76, 0.77, 0.80, 0.81, 0.78, 0.83 (columns (1)–(6)).
- Local currency-USD volatility:
  - Column (1): 82.43***
  - Column (2): 35.06*
  - Column (3): 56.25**
  - Column (4): 7.168
  - Column (5): 3.118
  - Column (6): 19.26*
- Local currency-THB volatility:
  - Column (1): 0.852
  - Column (2): 2.939
  - Column (3): 5.345
  - Column (4): 2.356
  - Column (5): 0.422
  - Column (6): 2.587
- Local currency-USD depreciation:
  - Column (1): -0.264***
  - Column (2): -0.317***
  - Column (3): -0.208**
  - Column (4): -0.126**
  - Column (5): -0.215***
  - Column (6): -0.133**
- Local currency-THB depreciation:
  - Column (1): -0.142
  - Column (2): -0.160
  - Column (3): -0.179*
  - Column (4): -0.132**
  - Column (5): -0.107*
  - Column (6): -0.129**
- Number of commercial bank branches:
  - Column (1): 0.00653
  - Column (2): -1.966***
  - Column (3): -1.615***
  - Column (4): -0.0159
  - Column (5): -0.977***
  - Column (6): -0.771***
- Number of credit cards:
  - Column (1): -0.182***
  - Column (2): 0.173
  - Column (3): 0.217
  - Column (4): -0.0989***
  - Column (5): -0.358***
  - Column (6): -0.292***
- Month of implementation:
  - Column (1): 0.111***
  - Column (2): 0.240***
  - Column (3): 0.0691*
  - Column (4): 0.0442**
  - Column (5): 0.186***
  - Column (6): 0.0689***
- Log (number of flight):
  - Column (1): 0.961***
  - Column (2): 1.432***
  - Column (3): 1.062***
  - Column (4): 0.478***
  - Column (5): 0.806***
  - Column (6): 0.576***
- Fixed effects:
  - Year FE: included in columns (2)–(6) as marked.
  - Country FE: included as marked.

### Policy conclusions and recommendations (selected)
- Scalability and interoperability:
  - Bilateral cross-border arrangements are not scalable; pursue multilateral hub-and-spoke approaches (e.g., Project Nexus) to enhance efficiency and interoperability.
- Financial inclusion and SME support:
  - Leverage digital payment data to expand financial inclusion — examples: “Your Data” project; National Credit Guarantee Agency (NaCGA); virtual banks using alternative data.
  - Encourage SMEs to engage in regional trade and utilize cross-border payment arrangements via awareness campaigns and incentives.
- Oversight and AML/CFT:
  - Implement robust oversight and reporting; apply risk-based AML/CFT measures with simplified measures in lower-risk scenarios to promote inclusion.
  - Harmonize data protection laws across ASEAN; develop a dedicated framework for cross-border supervisory, law enforcement, and oversight coordination.
- Network expansion and user uptake:
  - Promote broader participation from banks and financial institutions to expand network and promote QR payments as viable alternative for unbanked and underbanked groups.
  - Communicate the low-cost and real-time nature of local currency settlement to position services as tools for financial resilience during currency volatility.

*Source: IMF staff calculation, chapter “Reinvigorating Thailand’s Structural Transformation.”*

### 1.    Asian Countries: Income and Sectoral Employment Shares ___________________________ 3

### REINVIGORATING THAILAND’S STRUCTURAL TRANSFORMATION

### Background
- Structural transformation defined as reallocation of labor from lower- to higher-productivity sectors.
- Thailand follows common development trajectory but with idiosyncrasies:
  - The share of employment in agriculture remains persistently high relative to countries at similar income levels; “approximately one-third of Thai workers are still employed in agriculture.”
  - Employment in industry and services is relatively low for Thailand’s income level.
  - Within services, a disproportionately large share of workers are employed in low-productivity services (trade services and other services); employment in high-productivity services (including finance and business services) remains relatively low.
- Imbalance between sectoral labor and output shares:
  - Industry and high-productivity services account for the majority of value-added.
  - Low-productivity services and agriculture contribute substantially less to value-added but employ larger shares of labor.
  - Result: a wide and growing labor productivity gap across sectors.
- Implication: scope for productivity gains by reallocating workers to higher-productivity sectors.

### Assessing Thailand’s Structural Transformation
- Measurement approach:
  - Covariance between sectoral employment share and labor productivity used to quantify structural transformation (following IMF (2013) and Klyuev (2015)).
  - Covariance term defined in the source as:
    퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶퐶 = �(푠푠
    푖푖
    −푠푠̅
    푛푛
    푖푖
    )
    (
    푝푝
    푖푖
    −푝푝̅
    )
  - Simplified identity in the chapter:
    �(푠푠
    푖푖
    −푠푠̅
    푛푛
    푖푖
    )
    (
    푝푝
    푖푖
    −푝푝̅
    )
    = �푠푠
    푖푖
    푝푝
    푖푖
     −  
    푛푛
    푖푖
    �푠푠
    푖푖
    푝푝̅ 
    푛푛
    푖푖
    =푃푃− 푃푃
    �
  - Interpretation: covariance > 0 when more productive sectors employ larger shares of labor; covariance < 0 when less productive sectors employ larger shares of labor.
- Data and classifications:
  - Data from UNU-Wider Economic Transformation Database (ETD) covering 1990–2018 for cross-country comparisons.
  - Covariance computed for both 3-sector (agriculture, industry, services) and 4-sector (agriculture, industry, low-productivity services, high-productivity services) classifications.
- Evolution and cross-country patterns:
  - Most selected Asian countries exhibit negative covariance values historically, but many have improved over time.
  - Korea and Malaysia: covariance close to or above zero.
  - Chinese mainland: substantial improvement approaching Korea and Malaysia levels.
  - Vietnam: dramatic gains.
  - Thailand: progress in early 1990s, but momentum slowed since mid-2010s; covariance plateaued and slightly declined in recent years.
  - Sectoral disaggregation matters: Thailand’s covariance is relatively high in 3-sector view but significantly lower in 4-sector view, revealing an internal services imbalance.

### Drivers of Structural Transformation (Empirical Approach and Results)
- Empirical model:
  - Dependent variable: scaled covariance term.
  - Estimation: cross-country panel fixed-effects with year dummies.
  - Baseline independent variables (country fundamentals and policy-related):
    - Log GDP per capita (constant PPP, WDI) and its square.
    - Log population size; old-age dependency ratio (WDI).
    - Average years of schooling (UNDP) and Human Capital Index (Penn World Tables).
    - Trade-to-GDP ratio; FDI flows as % of GDP (WDI); terms of trade index (WDI); real effective exchange rate (REER) (WEO).
    - Agricultural raw materials exports as share of merchandise exports (WDI).
    - Vulnerable employment as share of total employment (WDI).
    - Domestic credit to private sector as % of GDP (WDI).
    - Political Corruption Index (V-Dem), rescaled by multiplying original values by 100.
  - Policy variables lagged by one year to mitigate endogeneity; contemporaneous values retained for GDP per capita, population, and old-age dependency ratio.
- Robustness checks / augmented specifications:
  - Labor Market Regulations index (2024 Fraser Economic Freedom of the World Dataset).
  - Gini Index (WIID).
  - Internet access (share of population with internet access, WDI).
  - Disaggregated education enrollment ratios (primary, secondary, tertiary, UNESCO).
  - Rule of Law (World Bank WGI) replacing Political Corruption Index.
- Key regression findings (qualitative summary from source):
  - Log GDP per capita: strongly and positively correlated with covariance; quadratic term negative — improvement with income but at a diminishing rate.
  - Log population: strong positive coefficient — larger population associated with greater labor allocation efficiency.
  - Old-age dependency ratio: positive correlation (consistent with association of aging with lower agriculture value-added share and higher services value-added share).
  - Terms of trade and trade openness: contribute to higher covariance (trade integration supports structural transformation).
  - REER appreciation: associated with lower covariance (reduced competitiveness hampers tradable sector development).
  - FDI inflows: not statistically significant (heterogeneous effects across sectors).
  - Average years of schooling: positive and significant across specifications.
  - Human Capital Index: not statistically significant.
  - Vulnerable employment (informality): negatively correlated with covariance (statistical significance in some specifications).
  - Share of agricultural exports: negatively associated with covariance.
  - Private credit-to-GDP ratio: positively correlated with covariance (credit access facilitates labor allocation).
  - Corruption (rescaled Political Corruption Index): negatively associated with covariance across all specifications.
  - Labor market regulation (alternative specification): better regulation associated with higher covariance.
  - Education detail: primary and secondary enrollment more strongly correlated with covariance than tertiary enrollment.
  - Internet access: positive coefficient but not statistically significant.
  - Rule of Law: stronger rule of law positively associated with structural transformation.

### Understanding the Recent Slowdown (Thailand-specific decomposition)
- Method:
  - Use estimated model to generate fitted covariance and compare with actual covariance.
  - Compute changes in driver values between 2010–2014 and 2014–2018, multiply by estimated coefficients to obtain contributions to covariance change.
- Observed shifts between 2010–2014 and 2014–2018:
  - Private sector credit growth slowed markedly in the latter period.
  - Progress in trade openness reversed.
  - REER depreciated.
  - Share of vulnerable employment declined.
  - Terms of trade improved.
  - Additional changes: acceleration in the old-age dependency ratio; reversal in FDI growth; slower growth in agriculture export share; more moderate increase in average years of schooling; deterioration in governance (higher corruption); slower population growth.
- Net effect:
  - These changes produced increasing downward pressures on Thailand’s structural transformation since the mid-2010s, contributing to the observed plateau and slight decline in the covariance term in recent years.

### Findings and Policy-relevant Implications (drawn from chapter evidence)
- Findings:
  - Thailand exhibits an inefficient labor allocation with agriculture and low-productivity services employing large labor shares despite lower value-added contributions.
  - Structural transformation in Thailand slowed since the mid-2010s, with the slowdown traceable to shifts in credit growth, trade openness, governance, population dynamics, and other factors.
  - Education (especially primary and secondary), trade integration, credit access, and governance quality are empirically linked to stronger structural transformation.
- Policy implications emphasized by the analysis:
  - Facilitate labor reallocation toward higher-productivity sectors to boost aggregate labor productivity.
  - Strengthen trade openness and competitiveness (address REER appreciation pressures) to support tradable sectors.
  - Improve access to private credit to enable investments that shift labor into higher-productivity activities.
  - Address informality (vulnerable employment) to enable better labor mobility and productivity gains.
  - Invest in education—particularly primary and secondary attainment—to support structural transformation.
  - Improve governance and reduce corruption; strengthen rule of law to remove impediments to labor reallocation.
  - Consider labor market regulations that facilitate efficient movement of labor across sectors.

*Source: IMF staff calculation, chapter “Reinvigorating Thailand’s Structural Transformation.”*

### 17.      These factors collectively influenced Thailand’s structural transformation. Beyond

### Thailand: Structural Transformation — Drivers, Policy Implications, and Findings

### Drivers of Structural Transformation and Recent Contributions
- Collective influences slowing structural transformation:
  - Slower private sector credit expansion.
  - A weaker rise in average years of schooling.
  - Decelerating population growth.
  - Reduced trade openness.
  - Deteriorating governance.
  - These factors contributed to a lower covariance term, thereby slowing structural transformation relative to the previous period.
- Factors mitigating the slowdown:
  - Population aging.
  - A decline in vulnerable employment.
  - Improved terms of trade.
  - REER depreciation.
  - These mitigating factors supported labor reallocation toward more productive sectors.
- Sectoral drag:
  - The agricultural export share continued to exert a negative impact, though its drag diminished, resulting in a modest reduction in its effect.

### Policy Implications and Recommendations
- Trade and market integration:
  - Accelerate trade liberalization, including deeper integration into global and regional markets, to facilitate labor reallocation toward more productive sectors.
- Education and skills:
  - Invest in education, particularly at the primary and secondary levels, to build a workforce capable of transitioning to higher-productivity employment.
- Finance and credit access:
  - Enhance financial inclusion and expand access to credit for productive investments to broaden opportunities and improve labor allocation efficiency.
  - Reinforce reforms with complementary measures to broaden financial access for firms and foster a more diversified financial structure to improve capital allocation efficiency.
- Governance and informality:
  - Reduce corruption and enhance the rule of law to create a predictable environment conducive to efficient labor reallocation.
  - Reduce informality through expanded social insurance coverage to support integration of informal workers into formal, higher-productivity sectors.
- Labor market reform:
  - Improve labor market flexibility by streamlining hiring and dismissal procedures and clarifying labor regulations to lower barriers to formal employment growth.

### Lessons from Regional Peers (select examples)
- Korea:
  - Human capital investment underpinned growth.
  - Average years of schooling increased from 5.4 in 1970 to 12.7 in 2023.
  - STEM capabilities enhanced via flagship research institutions and targeted R&D initiatives (e.g., Brain Korea 21).
  - Vocational training strengthened through Meister high schools co-designed with industry partners.
  - Employment Insurance System supported worker adaptation to technological change and industrial restructuring.
- Malaysia:
  - Inclusive finance via multi-year strategic frameworks (Financial Inclusion Framework (2023–2026)) with clear objectives and measurable targets.
  - Policy goals include expanding financial access to all subdistricts, increasing e-payment transactions, enhancing SME financing ecosystem, doubling microinsurance/microtakaful subscriptions, and raising insurance/takaful penetration.
  - Outcomes: nearly all subdistricts now have access to basic financial services via agent banking and mobile banks; operationalization of digital banks and launch of the DuitNow real-time retail payment platform broadened financial access.
- Vietnam:
  - Structural transformation propelled by strong integration into global trade following Đổi Mới reforms.
  - Policy actions: Foreign Investment Law, WTO accession, and 17 Free Trade Agreements in effect.
  - Recent government restructuring and continued basic education efforts aim to enhance efficiency and human capital, though skill mismatches persist.

### Key Analytical Conclusions
- Structural transformation drivers:
  - Both structural factors (income levels, demographic trends) and policy-related factors (trade, education, credit access, governance, labor markets) critically influence structural transformation.
- Thailand-specific dynamics:
  - Decline in informal employment and more favorable terms of trade have supported recent labor reallocation.
  - Deceleration in credit provision, limited progress in educational attainment, reversal in trade liberalization, and deterioration in governance contributed to slower structural transformation.
- Policy priorities to reinvigorate transformation:
  - Advance trade liberalization; invest in education and skills; enhance financial inclusion and access to credit; reduce labor informality; improve labor market flexibility.

### Quantitative Evidence (selected regression coefficients and results)
- Appendix I, Table 1 (4-Sector Covariance), selected estimates (with Driscoll–Kraay Standard errors in parentheses):
  - Log GDP per capita: 7.7069 *** (1.4912) in Base Spec.
  - Square of Log GDP per capita: -0.4075 *** (0.0783) in Base Spec.
  - Log population: 2.5526 *** (0.2912) in Base Spec.
  - Old-age dependency ratio: 0.0768 *** (0.0142) in Base Spec.
  - Terms of trade (lagged): 0.0039 *** (0.0008) in Base Spec.
  - REER (lagged): -0.0017 ** (0.0007) in Base Spec.
  - Average years of schooling (lagged): 0.1549 *** (0.0390) in Base Spec.
  - Domestic credit to the private sector to GDP (lagged): 0.0039 *** (0.0010) in Base Spec.
  - Political corruption index (lagged): -0.0045 *** (0.0014) in Base Spec.
  - Share of agricultural raw material exports (lagged): -0.0132 ** (0.0063) in Base Spec.
  - Share of vulnerable employment (lagged): -0.0088 *** (0.0020) in Spec 2.
  - Observations: 865 in Base Spec; Countries: 44; Within R^2: 0.624.
- Appendix I, Table 2 (3-Sector Covariance), selected estimates:
  - Log GDP per capita: 6.5674 *** (1.3976) in Base Spec.
  - Square of Log GDP per capita: -0.3408 *** (0.0727) in Base Spec.
  - Old-age dependency ratio: 0.0597 *** (0.0128) in Base Spec.
  - Terms of trade (lagged): 0.0028 *** (0.0007) in Base Spec.
  - Domestic credit to the private sector to GDP (lagged): 0.0028 *** (0.0006) in Base Spec.
  - Share of agricultural raw material exports (lagged): -0.0115 ** (0.0042) in Base Spec.
  - Observations: 865 in Base Spec; Countries: 44; Within R^2: 0.624.

### Conclusion
- Thailand’s structural transformation has slowed but remains incomplete.
- Targeted, well-coordinated policies promoting labor movement toward its most productive uses could unlock substantial productivity gains and boost economic growth.

*Source: IMF staff calculation.*

### 3.      Recent developments in digital cross-border payments in ASEAN have aligned with

### 3.      Recent developments in digital cross-border payments in ASEAN have aligned with

### Alignment with global and regional initiatives
- The G20 Roadmap for Enhancing Cross-border Payments serves as a framework to address frictions: high costs, slow transaction speeds, limited accessibility, and a lack of transparency.
- ASEAN regional response: ASEAN Regional Payment Connectivity (RPC) and ASEAN Local Currency Transaction Framework (LCTF).
  - RPC objective: link national payment systems seamlessly to create a more integrated, efficient, and user-friendly cross-border payment ecosystem.
  - LCTF objective: promote the use of local currencies for cross-border transactions to reduce costs, speed up transactions by eliminating reliance on third-party currencies, and strengthen financial resilience.

### Thailand’s progress in domestic and cross-border digital payments
- Digital payment uptake and financial inclusion:
  - Mobile phone ownership in ASEAN: over 80 percent.
  - E-money account ownership in ASEAN: about 20 percent.
  - In Thailand: close to half of the adult population owns an e-money account.
  - About 80 percent of adults in Thailand own financial accounts.
  - 55 percent of adults in Thailand own a debit card.
  - 50 percent of adults in Thailand have made digital payments.
- Growth of non-physical payments in Thailand:
  - Use of non-physical payments increased annually by more than 75 percent on average during 2019-24.
  - PromptPay adoption: average number of fast payment transactions through PromptPay rose more than eight-fold from below 40 per person per year to almost 350 during 2019-24.
- Cross-border linkages leadership:
  - Thailand established its first QR payment linkage with Japan in 2018 and later with eight other economies.
  - April 2021: Singapore and Thailand established the first cross-border fund transfer connectivity.
  - Thailand is a first mover in linking digital payments across borders among ASEAN peers.

### ASEAN-wide digital payments developments and drivers
- Private sector fintech investments:
  - Fintech funding to ASEAN surged since 2014, peaked in 2021, then declined thereafter.
  - Singapore is a key hub for private investment funding, followed by the Philippines and Indonesia.
- E-money adoption and usage:
  - In Malaysia, total value of e-money payments accounted for more than 8 percent of annual GDP.
  - In Singapore, each person made about 300 transactions through e-money in 2024.
- Central bank roles and national systems:
  - Fast payment systems and QR payment standards were developed to ensure interoperability and catalyze adoption.
  - Examples:
    - Bank of Thailand: PromptPay — universal, low-cost real-time payment system linking bank accounts to national IDs or mobile phone numbers.
    - Bank Indonesia: QRIS — single national QR standard to eliminate fragmentation.
    - Vietnam: State Bank supports VietQR but has not mandated nationwide use.
- Domestic fast payments and QR growth:
  - QR payments are a relatively small component of fast payments but exhibit significant growth.
  - Philippines example: QR payments increased by 467 percent in volume and 372 percent in value in 2024 compared with 2023.

### Cross-border payments outlook and scale
- Asia’s share of global cross-border payments:
  - Asia accounted for 32 percent of global cross-border payments in 2024 and is expected to account for 37 percent in 2032.
  - Total volume of cross-border payments in Asia expected to increase from 12.8 billion in 2024 to 23.8 billion in 2032.
- Cost and speed characteristics:
  - Average cost of remittances in East Asia and Pacific economies declined from 6.7 percent in 2021Q1 to 5.8 percent in 2025Q1 (source: World Bank Remittance Prices Worldwide database).
  - Asia lags other regions on percentage of cross-border payments credited within one business day.

### Bilateral linkages, multilateral initiatives, and CBDC experimentation
- Types of linkages established in ASEAN:
  - QR payment connectivity: enables travelers to make real-time merchant payments in another country by scanning a QR code.
  - Fund transfer connectivity: facilitates instant remittances and fund transfers across borders using simple identifiers like mobile phone numbers or national IDs.
- Growth in QR cross-border payments:
  - 2024 example: QR payments in Thailand increased by more than 300 percent in 2024 compared with 2023.
  - 2024 example: QR payments in Malaysia increased by 550 percent in 2024 compared with 2023.
  - Inbound payments through all seven QR linkages in Thailand in 2024 were about five times that in 2023.
- Project Nexus (multilateral hub-and-spoke approach):
  - Collaboration among BIS Innovation Hub, Bank of Thailand, Bank Negara Malaysia, Bangko Sentral ng Pilipinas, Monetary Authority of Singapore, and Reserve Bank of India.
  - Proof of concept completed in 2022; comprehensive blueprint announced in July 2024.
  - Nexus Global Payments incorporated in Singapore aiming to go-live by 2027.
- CBDC developments:
  - Majority of ASEAN countries launched CBDC research since 2016; several moved to experiments and pilots since 2020.
  - No ASEAN country has officially launched CBDC for public use as of the data presented.
  - Singapore and Thailand are the most advanced in their development stage; others at preparation, proof-of-concept, prototype, or pilot stages.
- Project mBridge (wholesale CBDC multilateral initiative):
  - Participants include Bank of Thailand, Central Bank of the United Arab Emirates, Digital Currency Institute of the People's Bank of China, Saudi Central Bank, and Hong Kong Monetary Authority.
  - Six-week pilot in 2022 involved 20 commercial banks across four jurisdictions, executing over 160 real-value transactions.
  - Project reached Minimum Viable Product (MVP) stage in mid-2024; ongoing development and real-value transactions continue subject to participating jurisdictions’ conditions.

### Benefits for SMEs and tourism linkages
- SME adoption and benefits:
  - Bank of Thailand survey: 96 percent of SMEs had adopted digital payments accounting for over two-thirds of their payment value.
  - Digital payments enable SMEs to build digital financial footprints and provide transaction data to assess creditworthiness.
  - PromptBiz (introduced August 2023) supports SMEs to access finance using PromptBiz trade transactions as supplemental documents for invoice factoring or other financing.
- Tourism implications:
  - Intra-ASEAN tourists in 2023 accounted for 42 percent of total visitors, up from 36 percent in 2019.
  - Tourism accounted for 8 percent of the region's GDP and 12 percent of employment in 2023.
  - Expanded cross-border digital payment use could have significant positive effects on the region’s service sector.

### Risks, fragmentation, and policy-relevant considerations
- Financial crime and AML risks:
  - Rapid adoption introduced significant financial crime and money laundering risks, acute in cross-border transactions due to regulatory inconsistencies and speed of instant payments.
  - Criminal techniques include micro-structuring, identity theft, and use of mule accounts; risk heightened by scam centers and transnational criminal operations from high-risk countries.
- Fragmentation challenges:
  - Multiple bilateral linkages create a highly fragmented system, complicating interoperability between systems and jurisdictions.
  - Bilateral linkage buildup is resource-consuming and replicates processes for each new partner.
- Multilateral approaches and efficiencies:
  - Project Nexus and similar hub-and-spoke models can simplify infrastructure and resolve interoperability challenges more efficiently than many bilateral linkages.

### Empirical findings: Thailand-specific drivers of cross-border QR payments
- Payment mix and trends for inbound payments to Thailand:
  - Share of cross-border ATM withdrawals has declined in recent years and remains lower than pre-pandemic levels.
  - Credit card and debit card payments on average account for about 60 percent of total inbound payments.
  - The remainder comprises digital payments, including e-money and QR payments through bilateral cross-border fast payment linkages.
- Correlates of bilateral-linkage QR payment volumes:
  - QR payments from partner economies are positively correlated with tourism receipts from that economy.
  - Economies with higher relative transfer costs via traditional commercial bank channels have higher QR transactions.
  - Linkages with a higher number of participating banks are associated with higher QR transactions.
  - Economies geographically further from Thailand seem to have higher QR transactions in Thailand.
  - These relationships could not be quantified in an econometric model due to data availability limitations.

*IMF staff summary based on the provided chapter content.*

### 21.      An econometric model is used to estimate the effect of FX depreciation, FX volatility

### 21.      An econometric model is used to estimate the effect of FX depreciation, FX volatility and alternative channels of payments on the volume and value of cross-border QR payments

### Model specification and dataset
- Dataset: payments from seven economies (Cambodia, Hong Kong SAR, Indonesia, Laos, Malaysia, Singapore, and Vietnam) to Thailand, period 2020-2024, monthly aggregate level, covering 60 months with a total of 420 observations, both by volume and value.
- Dependent variables: volume and value of cross-border QR payments at the bilateral economy pair level.
- Explanatory variables:
  - FX_volit: Foreign exchange risk measured first as the volatility of local currency against the US dollar, and second as the volatility of local currency against the Thai Baht.
  - FX_git: Monthly depreciation of the local currency against the Dollar and the Baht.
  - flightit: Number of flights in each month as an approximation for tourists visiting Thailand from each partner economy.
  - Monthit: Number of months the linkage has been established (to account for familiarity).
  - Commer_bankiT: Number of commercial bank branches from the IMF financial access survey (country-level, time-varying).
  - Credit_cardiT: Number of credit cards from the IMF financial access survey (country-level, time-varying).
  - CiT: Macroeconomic indicators (GDP and CPI) from the latest IMF WEO database.
- Model (log specification for value and volume):  
  QRit = β1 Monthit + β2 flightit + β3 FX_volit + β4 FX_git + β5 Commer_bankiT + β6 Credit_cardiT + CiT + εit
- Notes: i = partner economy, t = month, T = year (some variables annual frequency).

### Key empirical findings (summary)
- Higher local currency-USD volatility is associated with higher cross-border QR payments (value) and affects volume less strongly.
- Higher local currency-THB volatility does not appear to affect cross-border QR payments.
- Local currency depreciation against the US dollar and the Thai Baht is associated with lower QR payments (both value and volume).
- Number of commercial bank branches and number of credit cards tend to be negatively associated with QR payments in most regressions, suggesting QR payments substitute for traditional banking and card payments when access to traditional services is lower.
- Month of implementation and log(number of flight) are positively associated with both value and volume of QR payments.
- Results remain robust when currency depreciations are excluded and when volatilities against the USD and THB are included separately.

### Table of estimated coefficients (selected results)
- Sample size: N 420 (for all regressions).
- R^2: 0.76, 0.77, 0.80, 0.81, 0.78, 0.83 (columns (1) through (6) respectively).
- Dependent variables by column:
  - (1) Log (value)
  - (2) Log (value)
  - (3) Log (value)
  - (4) Log (volume)
  - (5) Log (volume)
  - (6) Log (volume)

- Local currency-USD volatility:
  - Column (1): 82.43***
  - Column (2): 35.06*
  - Column (3): 56.25**
  - Column (4): 7.168
  - Column (5): 3.118
  - Column (6): 19.26*

- Local currency-THB volatility:
  - Column (1): 0.852
  - Column (2): 2.939
  - Column (3): 5.345
  - Column (4): 2.356
  - Column (5): 0.422
  - Column (6): 2.587

- Local currency-USD depreciation:
  - Column (1): -0.264***
  - Column (2): -0.317***
  - Column (3): -0.208**
  - Column (4): -0.126**
  - Column (5): -0.215***
  - Column (6): -0.133**

- Local currency-THB depreciation:
  - Column (1): -0.142
  - Column (2): -0.160
  - Column (3): -0.179*
  - Column (4): -0.132**
  - Column (5): -0.107*
  - Column (6): -0.129**

- Number of commercial bank branches:
  - Column (1): 0.00653
  - Column (2): -1.966***
  - Column (3): -1.615***
  - Column (4): -0.0159
  - Column (5): -0.977***
  - Column (6): -0.771***

- Number of credit card:
  - Column (1): -0.182***
  - Column (2): 0.173
  - Column (3): 0.217
  - Column (4): -0.0989***
  - Column (5): -0.358***
  - Column (6): -0.292***

- Month of implementation:
  - Column (1): 0.111***
  - Column (2): 0.240***
  - Column (3): 0.0691*
  - Column (4): 0.0442**
  - Column (5): 0.186***
  - Column (6): 0.0689***

- Log (number of flight):
  - Column (1): 0.961***
  - Column (2): 1.432***
  - Column (3): 1.062***
  - Column (4): 0.478***
  - Column (5): 0.806***
  - Column (6): 0.576***

- Constant (C):
  - Column (1): 4.936***
  - Column (2): -20.71
  - Column (3): -14.43
  - Column (4): 1.398*
  - Column (5): -0.346
  - Column (6): -0.614

- Fixed effects:
  - Year FE: Y in columns (2), (3), (4), (5), (6)
  - Country FE: Y in columns as marked in table.

(Note: Standard errors in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.)

### Conclusions and policy discussion (selected points)
- ASEAN progress: Significant progress in developing domestic and cross-border digital payments, supported by coordination, supervision, standardization, investment, and private-sector innovation.
- Scalability challenge: Current web of bilateral cross-border arrangements is not scalable; multilateral approaches are being explored to enhance efficiency, interoperability, and reduce complexity.
- Leveraging payment data for inclusion:
  - Thailand initiatives mentioned: "Your Data" project; establishment of National Credit Guarantee Agency (NaCGA); virtual banks using alternative data to assess creditworthiness of underserved groups.
  - Recommendation: Encourage SMEs to engage in regional trade and utilize existing digital cross-border payment arrangements through awareness campaigns and incentives.
- Oversight and reporting:
  - Implement robust oversight and reporting mechanisms.
  - Ensure risk-based application of AML/CFT measures, including simplified measures in lower-risk scenarios to promote financial inclusion.
  - Push for greater harmonization of data protection laws across ASEAN.
  - Enhance collaboration between authorities and financial institutions to improve digital and financial literacy and cybersecurity preparedness.
  - Develop a dedicated framework for cross-border supervisory, law enforcement, and oversight coordination to facilitate rapid information sharing and combat financial crime.
- Benefits of local currency usage and regional connectivity:
  - Empirical evidence suggests usage of cross-border QR payments is positively associated with higher volatility of the local currency against the US dollar.
  - QR payments may serve as an alternative to traditional banking and credit card payments.
  - Possible positive correlations between QR payments and tourism receipts, the cost of traditional cross-border bank transfers, and the number of banks participating in the linkages.
- Strategies to promote trade integration via cross-border QR payments:
  - Provide widespread explanations to potential users of the low-cost and real-time nature of local currency settlement, positioning the service as a tool for financial resilience during periods of local currency volatility.
  - Actively encourage broader participation from banks and other financial institutions to expand the network and promote QR payments as a viable alternative, particularly for unbanked and underbanked groups.

*IMF — Thailand chapter (analysis of cross-border QR payments, econometric model and policy discussion).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2026/english/1thaea2026002-source-pdf.pdf_