## wpiea2024207-print-pdf

## Source details

**Canonical URL:** [wpiea2024207-print-pdf](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024207-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024207-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024207-print-pdf.pdf.json)

---

### Accession process and nature of commitments
- Accession to the WTO is the outcome of negotiations and entails a series of reforms that can fundamentally transform the economy of the acceding member.
- Commitments are reflected in Protocols of Accession and are enforceable under the Dispute Settlement Mechanism.
- Implementation of reforms often starts years ahead of the accession date; accession is not set until a candidate has met all necessary conditions.
- Negotiations can take several years and differ across applicants in the number of meetings, length, and set of issues discussed.
- The study exploits heterogeneity in accession proceedings (timing of meetings, number of issues raised and answered) to explore the relationship between reforms and economic growth.

### Channels through which WTO accession may affect growth
- Trade expansion channel
  - Joining GATT/WTO increases trade by 140% (Larch et. al, 2024).
  - Survey evidence: economic growth is roughly 1.0–1.5 percentage points higher than a benchmark after trade reforms by developing economies, cumulating to 10–20 percent higher income after a decade (Irwin, 2024).
  - Feyrer (2019): a 10 percent increase in trade leads to a 5 percent increase in national income; trade can explain 17 percent of the variation in growth rates across economies between 1960 and 1995.
- Institutional quality and governance channel
  - Prospective members undertake legally binding commitments across subsidies, agriculture, textile and clothing, trade-related investment, services, and intellectual property.
  - Commitments can include limits on export duties, elimination of fees and charges other than tariffs on imports, strict limits on industrial subsidies geared towards exporters, and limits on price controls.
  - Institutional upgrading can lower cost of doing business and promote investment by securing property rights.
  - The paper finds the pace of reform commitments during accession is positively correlated with some institutional quality indicators.
- Predictability and uncertainty channel
  - WTO commitments can underpin a more predictable trading environment, reducing uncertainty that deters investment and slows capital accumulation.
  - WTO and PTA membership can decrease volatility of trade flows through enhanced rule of law, enforcement of market-access commitments, deterrence of new protectionist barriers, transparency, and policy convergence.
  - WTO commitments can reduce trade policy uncertainty directly.

### Data, Pre-Accession Commitment Index, and sample
- Novel contribution: first study to capture timing and pace of trade reforms during the accession process by constructing a pre-accession index of commitments and reforms.
- Sample focused on the 36 economies that acceded to the WTO under Article XII of the Marrakesh Agreement between 1995 and 2023 (study sample broader for some estimations).
- Pre-Accession Commitment Index
  - Definition: share of questions answered by an applicant in a given year out of the total number of questions answered at conclusion of the accession process; normalized between 0 and 1.
  - Construction: manually reviewed Working Party (WP) Q&A documents to count questions answered; index proxies progress of reforms and alignment with WTO framework.
  - Operational rules:
    - Index is a step function: number of answers can only increase in consecutive years.
    - Index can only change in WP meeting years (answers released in non-meeting years are counted in the next meeting year).
    - Only answers released prior to meetings are considered.
    - In the years after accession the pre-accession index ends and takes the value 0.
  - Empirical patterns:
    - On average, an applicant answers 80% of all questions 5 years after MFTR release.
    - Large heterogeneity in pace of Q&A completion (examples: Vanuatu 16 years, Seychelles 18 years, Kazakhstan 19 years; Ecuador 1 year, Armenia/Georgia/Kyrgyz Republic 2 years).
    - Correlations: questions and commitments 0.69; questions and legislation enacted 0.77.
    - Article XII members (sample) aggregate stats: 1,454 accession-specific commitments total (40 on average), 4,905 legislations enacted total (136 on average), and over 36,000 questions answered total (1,019 on average).

### Empirical specification and main estimation results
- Main regression specification:
  - ∆log Y_{i,t} = β0 + β1 log(Y_{i,t−1}) + β2 GATT/WTO_{i,t} + β3 WTO Accession_{i,t} + β4 WTO Pre-Accession Index_{i,t} + β5 X_{i,t} + α_i + γ_t + ε_{i,t}
  - Dependent variable: first difference in log GDP per capita.
  - Controls X_{i,t}: log(Gross Fixed Capital Formation/GDP) and log(Trade Openness/GDP). Robustness checks include Human Capital Index and incidence of armed conflicts.
  - Sample: yearly data for 150 economies, 1951-2019; Observations: 8,147; Country F.E.: Yes; Year F.E.: Yes; R-squared 9.6%.
- Benchmark findings (Table 3, Column (3) preferred specification):
  - WTO Pre-Accession Index coefficient: 0.027* (standard error 0.015).
  - At mean value of the index (0.56) the average effect of WTO entry negotiations is reported as 1.5 percentage points.
  - WTO Accession coefficient: 0.036*** (standard error 0.012).
  - GATT/WTO_{i,t} coefficient: -0.002 (insignificant).
  - log(GDP pc)_{t-1} coefficient: -0.032*** (0.006).
  - log(Investment/GDP) coefficient: 0.017*** (0.005).

### Heterogeneous and subgroup results
- Descriptive patterns (Application and Accession 5-year windows, Article XII sample):
  - Application, Article XII (1995-2019): Average 5 years before -2.04%; Average 5 years after 3.11%; Difference 5.15p.p.; faster than before (share) 65.6%.
  - Accession, Article XII (1995-2019): Average 5 years before 4.03%; Average 5 years after 4.77%; Difference 0.74p.p.; faster than before (share) 62.5%.
  - For All economies (1948-2019): Application difference 1.76p.p.; Accession difference 0.37p.p.
- Subgroup interaction findings (Table 5):
  - Interaction with negotiation duration (median 9 years):
    - WTO Pre-Accession Index 0.039*** (0.014); interaction (Index * 1 if shortest duration) -0.030** (0.012).
    - Implied examples: 2.1 percentage points for above-median negotiation duration and 0.5 percentage points for below-median.
  - Interaction with total questions (median 875):
    - Index 0.037** (0.018); interaction -0.022 (0.015) (economies with more questions benefit more).
  - Interaction with recency of accession (median year 2003):
    - Index 0.043** (0.018); interaction -0.033*** (0.012) (more recent accessions drive larger effects).
    - Example: for 16 most recent accessions estimated pre-accession effect 2.2 percentage points.
  - Robust to treating Article XXVI accessions differently by changing accession date to metropole accession date.
  - GATT pre-accession indicator: coefficient 0.013* (0.007), indicating smaller but significant GATT pre-accession effect.

### Endogeneity, instrumental variables, and falsification
- Political economy and selection evidence:
  - Long lag from application to accession: average length from application to accession 9.3 years (range examples: Seychelles 19.9 years, Kyrgyz Republic 2.8 years).
  - Only 12 of 34 Article XII members had the same chief executive at application and accession.
- Instrumental variables (Table 6):
  - Instruments: Pre-Accession Index of the closest acceding member identified by geographic distance or by accession date; averages weighted by distance or time; overidentified combinations.
  - First-stage statistics: Kleibergen-Paap Wald rk F-stat reported as large values (examples: 788.7, 508.7, 1148.6, 1712.8, 1481.1, 1358.2) indicating instrument strength.
  - Second-stage IV results (selected):
    - Closest Distance: Index 0.022* (0.011); WTO Accession 0.036*** (0.010).
    - Closest Time: Index 0.026** (0.011); WTO Accession 0.036*** (0.010).
    - Index coefficients remain positive across instrument specifications (range 0.022* to 0.026**); WTO Accession consistently 0.036***.
- Falsification tests (Table 7):
  - Randomized by member: Index -0.001 (0.007) (insignificant).
  - Randomized by year: Index -0.020* (0.010) (weakly negative).
  - Significant positive effects disappear under randomization, supporting the identification strategy.

### Robustness and sensitivity checks
- Results robust to:
  - Inclusion of Human Capital Index (drops 38 economies) — Human Capital coefficient insignificant; main coefficients unchanged in significance.
  - Inclusion of Conflicts — Conflicts coefficient -0.007 (insignificant); main coefficients marginally reduced.
  - Excluding China, excluding EU enlargement countries, excluding six growth outliers — Pre-Accession Index coefficients remain positive and (weakly) significant. Example exclusion result: Pre-Accession Index 0.030** (0.013); WTO Accession 0.039*** (0.011).
  - Instrumental variable strategies using geographic and temporal proximity instruments.
  - Falsification tests where randomization removes significant effects.

### Key quantitative magnitudes
- Negotiating WTO accession associated with a 1.5 percentage points higher growth than would have occurred otherwise (authors’ calculation at mean index 0.56).
- Benchmark accession (general GATT/WTO dummy) post-accession growth increase: 0.8 percentage points (Table 3 Column (1)).
- WTO (Article XII) accession post-accession coefficient: 0.036*** in preferred specifications.
- Pre-Accession Index coefficient in preferred specifications: around 0.025–0.030 (Table 3 Column (3): 0.027* (0.015); IV range 0.022* to 0.026**).
- Article XII sample averages:
  - Total questions answered: 1,019.
  - Total commitments per Article XII member: 40.
  - Total legislations enacted per Article XII member: 136.

### Conclusions and policy implications
- Reforms required to join the WTO have become more demanding in several policy areas compared to accessions under the GATT.
- The accession negotiation process functions as a mechanism prompting domestic reforms and institutional alignment with WTO standards, which can materialize prior to formal accession.
- The WTO accession process is associated with:
  - faster growth before accession, and
  - continued faster growth after accession.
- Robustness checks, including instrumental variable estimations and falsification tests, support the interpretation that pre-accession negotiations and commitments contributed to growth improvements.
- Research agenda / Open questions:
  - Further research is needed to identify the specific types of reforms during WTO accession that have the greatest impact on growth.
  - Proposed next step: map specific reforms to commitments undertaken during negotiations to compare the effectiveness of different types of reforms for growth.

### Annex highlights
- Annex 1: Replication of Tang and Wei (2009) — time-profile regression coefficients for Article XII, Article XXVI, and Article XXXIII accessions (detailed coefficients and standard errors reported in source).
- Annex 2: Details of the Pre-Accession Index illustrated by Viet Nam example (Total answers 3511; WP meetings attended 14 between 1998 and 2006; index accumulation described).
- Annex 3: Examples of Q&A leading to reforms
  - Viet Nam: Decree No. 61/2003/ND-CP (6 June 2003); Decree No. 27/2003/ND-CP (19 March 2003); Decree No. 38/2003/ND-CP (15 April 2003); unified Law on Investment and Law on Enterprise drafted and scheduled for submission in 2006.
  - Comoros: timeline of Q&A leading to end of State monopoly on rice by Article 25 of Decree No. 23/60/PR of 30 June 2023; VAT implementation scheduled for 2027 with IMF technical assistance requested.
- Annex 4: Economies in the estimation sample
  - Full list provided in source. Footnote classifications:
    - ¹ Article XII
    - ² Article XXVI 5(c)

*Source: IMF Working Paper “Committing to Grow: The Full Impact of WTO Accessions” — Introduction, Sections 5–6, Conclusions, and Annexes (excerpt).*

### Introduction ...........................................................................................................

### Introduction

### Accession process and nature of commitments
- Accession to the WTO is the outcome of negotiations (Williams, 2008) and entails a series of reforms that can fundamentally transform the economy of the acceding member and serve as catalysts for institutional improvements.
- Commitments are reflected in Protocols of Accession and are enforceable under the Dispute Settlement Mechanism.
- Implementation of reforms often starts years ahead of the accession date; accession is not set until a candidate has met all necessary conditions.
- Negotiations can take several years and differ across applicants in the number of meetings, length, and set of issues discussed.
- The study exploits heterogeneity in accession proceedings (timing of meetings, number of issues raised and answered) to explore the relationship between reforms and economic growth.

### Channels through which WTO accession may affect growth
- Trade expansion channel
  - The WTO reduces trade barriers and boosts trade; joining GATT/WTO increases trade by 140% (Larch et. al, 2024).
  - Trade reforms that reduce import tariffs boost economic growth; lower tariffs on intermediate goods increased productivity of domestic final good producers (Estevadeordal and Taylor, 2013; Amiti et al., 2020).
  - A survey by Irwin (2024) found economic growth is roughly 1.0–1.5 percentage points higher than a benchmark after trade reforms implemented by developing economies, cumulating to 10–20 percent higher income after a decade.
  - Feyrer (2019): a 10 percent increase in trade leads to a 5 percent increase in national income, implying trade can explain 17 percent of the variation in growth rates across economies between 1960 and 1995.
- Institutional quality and governance channel
  - Prospective members undertake extensive domestic reforms and make legally binding commitments across areas including subsidies, agriculture, textile and clothing, trade-related investment, services, and intellectual property.
  - Commitments can include limits on export duties, elimination of fees and charges other than tariffs on imports, strict limits on industrial subsidies geared towards exporters, and limits on price controls.
  - Trade-related investment rules can limit local content requirements or trade-balancing; some commitments include privatization plans (examples: Montenegro, Russia, Samoa, Tajikistan).
  - Institutional upgrading can lower cost of doing business and promote investment by securing property rights (Acemoglu and Robinson, 2008; Glaeser et al., 2004).
  - The paper finds the pace of reform commitments during accession is positively correlated with some institutional quality indicators.
- Predictability and uncertainty channel
  - WTO commitments can underpin a more predictable trading environment, reducing uncertainty that deters investment and slows capital accumulation (Leahy and Whited, 1996; Guiso and Parigi, 1999).
  - Mansfield and Reinhardt (2008) find WTO and PTA membership can decrease volatility of trade flows, attributing this to enhanced rule of law, enforcement of market-access commitments, deterrence of new protectionist barriers, transparency, and policy convergence.
  - Jakubik and Piermartini (2023) show WTO commitments can reduce trade policy uncertainty directly.

### Existing empirical literature and gaps
- Evidence on the impact of WTO accession on trade is substantial; evidence on growth effects is limited.
- Tang and Wei (2009): GATT/WTO membership promotes growth, but only for members that undertake commitments to accede; members that were not required to make commitments (mostly former colonies or territories under GATT Article XXVI 5(c)) did not benefit. Their estimate: the economy of an acceded member is on average permanently 20% larger if it undertook commitments.
- Brotto, Jakubik, and Piermartini (2021): extending analysis to members that acceded under Article XII up to 2017, find that on average joining the WTO resulted in a 30% permanent increase in GDP.
- Grassi (2024): using synthetic control, joining the EU in 2004 led to an increase in GDP per capita of 32% for new members by 2019.
- Literature cautions that merely observing membership status is insufficient; the quality and circumstances of accession matter (Subramanian and Wei, 2007; Tomz et al., 2007; Eicher and Henn, 2011; Dutt, 2020).

### This study: data, index, sample, and main findings
- Novel contribution: first study to capture the timing and pace of trade reforms undertaken during the accession process to estimate WTO impact on growth by constructing a pre-accession index of commitments and reforms.
- Focus sample: the 36 economies that acceded to the WTO under Article XII of the Marrakesh Agreement between 1995 and 2023.
- Index: tracks the pattern of commitments undertaken over time during the pre-accession period using detailed data on negotiation progress (timing of meetings, number of issues raised and answered).
- Main quantitative findings:
  - Negotiating WTO accession is associated with a 1.5 percentage points higher growth than would have occurred otherwise.
  - This 1.5 percentage points is nearly 30% of the overall growth that the new acceding economies experienced during the negotiations, equal to 5.2 percentage points.
  - These new WTO members also grew faster than others after accession.
- Methods: baseline estimates complemented with instrumental variables and falsification tests; qualitative evidence and robustness checks are presented to support consistency of estimates.

### Organization of the paper (overview)
- Section 2: describes patterns of growth around date of accession and application.
- Section 3: outlines important features of the accession process and presents two commitment indices.
- Section 4: describes empirical specification and presents baseline results.
- Subsequent sections and annexes provide data, robustness checks, IV results, falsification tests, replication of prior results, index details, examples of accession Q&A leading to reforms, list of economies in sample, first-stage IV results, and additional documentation.

*Source: IMF Working Paper “Committing to Grow: The Full Impact of WTO Accessions” — Introduction (excerpt).*

### Section 5 addresses identification issues and presents robustness checks. Section 6 concludes.

### wpiea2024207-print-pdf - Section 5 addresses identification issues and presents robustness checks. Section 6 concludes.

### Heterogeneous effects of GATT/WTO accessions
- Sample: 150 developing and transition economies over the period 1951–2019; advanced economies at time of accession excluded.
- Key descriptive facts:
  - Clear difference between members acceding to the WTO since 1995 (Article XII) and those acceding earlier under GATT (non-Article XII) in GDP per capita growth trajectories around application and accession dates.
  - Article XII members show more significant positive growth coefficients already prior to accession and remain the only group with positive coefficients several years after accession.
- Accession modality groups described:
  - Article XII Members: 38 members (as of 2024) acceded since 1995; more rigorous accession process; heterogeneous group with a quarter LDCs; examples include China (2001) and Russia (2012).
  - Article XXVI Members: more than 60 ex-colonies/overseas territories/provisional members admitted by notification, often without required reforms.
  - Article XXXIII Members and Founding Members: earlier GATT accession modalities with less comprehensive reform obligations.
- Table 1 (trends in economic growth, Application and Accession 5-year windows):
  - Application, Article XII (1995-2019): Average 5 years before -2.04%; Average 5 years after 3.11%; Difference 5.15p.p.; faster than before (share) 65.6%.
  - Accession, Article XII (1995-2019): Average 5 years before 4.03%; Average 5 years after 4.77%; Difference 0.74p.p.; faster than before (share) 62.5%.
  - For All economies (1948-2019): Application difference 1.76p.p.; Accession difference 0.37p.p.
- Working hypothesis: heterogeneity arises from timing and depth of commitments undertaken during accession negotiations.

### WTO accession process, commitments, and Pre-Accession Commitment Index
- Typical timeline:
  - Average length from application to accession: 9.3 years; range example: Seychelles 19.9 years, Kyrgyz Republic 2.8 years.
  - Main steps: application → MFTR (Memorandum on the Foreign Trade Regime; released on average 2 years after application) → Q&A rounds → Working Party (WP) meetings → bilateral market access negotiations → Draft WP Report adopted ad referendum → accession.
- Commitments during accession:
  - Two types in accession packages: (1) market access (tariffs, quantitative restrictions, services commitments); (2) non-market-access regulatory and institutional reforms (price controls, intellectual property rights, trading rights, privatization plans, etc.).
  - Commitments are expected to be implemented or on track before accession, motivating potential pre-accession growth effects.
- Pre-Accession Commitment Index:
  - Definition: share of questions answered by an applicant in a given year out of the total number of questions answered at conclusion of the accession process; normalized between 0 and 1.
  - Construction: manually reviewed WP Q&A documents to count questions answered; index proxies progress of reforms and alignment with WTO framework.
  - Empirical patterns:
    - On average, an applicant answers 80% of all questions 5 years after MFTR release.
    - Large heterogeneity across members in pace of Q&A completion (e.g., Vanuatu 16 years, Seychelles 18 years, Kazakhstan 19 years; Ecuador 1 year, Armenia/Georgia/Kyrgyz Republic 2 years).
  - Correlations:
    - Positive correlation between questions and commitments (0.69) and questions and legislation enacted (0.77).
    - Article XII members (sample) aggregate stats: 1,454 accession-specific commitments total (40 on average), 4,905 legislations enacted total (136 on average), and over 36,000 questions answered total (1,019 on average).
  - Examples linking Q&A to reforms:
    - Viet Nam: Law on Investment passed 2005, effective July 2006 (prior to 2007 accession), addressing WP concerns on foreign investment restrictions; Decree 108/2006/ND-OP introduced details in September 2006.
    - Comoros: ended state rice import monopoly (ONICOR) through Presidential decree No. 23-060/PR in June 2023 after repeated WP questioning; committed to replace internal taxes with a VAT system and reaffirmed commitment in Q&A of 6th December 2023.

### Empirical specification and data
- Sample: yearly data for 150 economies, 1951-2019; 88 members acceded through Article XII or Article XXVI; 37 additional members and 28 non-members complete sample.
- Main regression:
  - ∆log Y_{i,t} = β0 + β1 log(Y_{i,t−1}) + β2 GATT/WTO_{i,t} + β3 WTO Accession_{i,t} + β4 WTO Pre-Accession Index_{i,t} + β5 X_{i,t} + α_i + γ_t + ε_{i,t}
  - Dependent variable: first difference in log GDP per capita.
  - GATT/WTO_{i,t}: dummy denoting membership.
  - WTO Accession_{i,t}: dummy for Article XII members in post-accession periods.
  - Pre-Accession Index_{i,t}: between 0 and 1, positive only for acceding Article XII members during negotiations.
  - Controls X_{i,t}: log(Gross Fixed Capital Formation/GDP) and log(Trade Openness/GDP). Robustness checks include Human Capital Index and incidence of armed conflicts.
- Data sources:
  - GDP and macro variables: Penn World Tables (expenditure-side GDP in chained PPP, USD 2011, millions).
  - Accession dates, MFTR, Q&A, tariff bindings: WTO.
  - Conflicts: UCDP Onset Dataset.
- Econometric issues addressed:
  - First-order autocorrelation: include lagged log(GDP) to address Wooldridge test.
  - Cross-sectional dependence and heteroskedasticity: use Driscoll and Kraay (1998) spatial correlation consistent standard errors.

### Benchmark empirical results (Table 3 highlights)
- Column (1): General accession dummy (GATT/WTO_{i,t}) positive and significant: accession increases growth rates by 0.8 percentage points on average.
- Column (2): Adding WTO Accession dummy shows most variation explained by WTO (Article XII) accessions; WTO Accession coefficient 0.033***.
- Column (3) (benchmark): include WTO Pre-Accession Index. Estimated effect:
  - WTO Pre-Accession Index coefficient: 0.027* (standard error 0.015) in specification where index included.
  - At mean value of the index (0.56) the average effect of WTO entry negotiations is reported as 1.5 percentage points (note: calculated using index times 2.7 percentage points as described in the source).
  - WTO Accession coefficient: 0.036*** (standard error 0.012).
  - GATT/WTO_{i,t} coefficient: -0.002 (insignificant).
  - log(GDP pc)_{t-1} coefficient: -0.032*** (0.006).
  - log(Investment/GDP) coefficient: 0.017*** (0.005).
  - Observations: 8,147; Country F.E. Yes; Year F.E. Yes; R-squared 9.6%; Sample size 150.
- Robustness columns:
  - Column (4): adding Human Capital Index (drops 38 economies) — Human Capital coefficient insignificant; main coefficients unchanged in significance.
  - Column (5): adding Conflicts — Conflicts coefficient -0.007 (insignificant); main coefficients marginally reduced.

### Robustness to excluding outliers and subgroup heterogeneity (Tables 4–5)
- Table 4 (excluding outliers):
  - Exclusions tested: -China; -EU enlargements; -Outliers (six growth outliers).
  - WTO Pre-Accession Index coefficients remain positive and significant/weakly significant across exclusions:
    - example Column (4) (excluding outliers): Pre-Accession Index 0.030** (0.013); WTO Accession 0.039*** (0.011).
  - Results quantitatively and qualitatively robust.
- Table 5 (subgroup interactions exploring drivers):
  - Interaction with negotiation duration (median 9 years):
    - WTO Pre-Accession Index coefficient 0.039*** (0.014); interaction term (Index * 1 if shortest duration) -0.030** (0.012).
    - Interpretation: pre-accession effects larger for accessions longer than median duration; for shortest-duration group effect smaller.
    - Reported example calculations in source: Column (1) yields 2.1 percentage points for above-median negotiation duration and 0.5 percentage points for below-median.
  - Interaction with total questions (median 875):
    - Index 0.037** (0.018); interaction -0.022 (0.015) (implying economies with more questions benefit more).
  - Interaction with recency of accession (median year 2003):
    - Index 0.043** (0.018); interaction -0.033*** (0.012) (more recent accessions drive larger effects).
    - Example: for 16 most recent accessions estimated pre-accession effect 2.2 percentage points.
  - Column (4): treating Article XXVI accessions differently by changing accession date to metropole accession date — results robust.
  - Column (5): GATT pre-accession indicator (dummy for years between application and accession for GATT economies): coefficient 0.013* (0.007), indicating smaller but significant GATT pre-accession effect.

### Endogeneity concerns, instrumental variables, and falsification
- Political economy and selection concerns:
  - Long lag between application and accession reduces selection on timing; evidence that only 12 of 34 Article XII members had the same chief executive at application and accession.
  - No significant relationship found between length of negotiations and governing party vote share; most pre-accession effects concentrated in longer negotiations, contrary to hypothesis that politically powerful governments accede quickly and drive results.
- Instrumental variables (Table 6):
  - Instruments: Pre-Accession Index of the closest acceding member identified by geographic distance or by accession date; additional instruments: averages weighted by distance or proximity in accession time; overidentified model combining instruments tested.
  - First-stage: instruments strongly related to index; weak identification tests (Kleibergen-Paap Wald rk F-stat) reported as large values indicating instrument strength (examples: 788.7, 508.7, 1148.6, 1712.8, 1481.1, 1358.2).
  - Second-stage IV results (selected columns):
    - Column (1) (Closest Distance): WTO Pre-Accession Index 0.022* (0.011); WTO Accession 0.036*** (0.010).
    - Column (2) (Closest Time): Index 0.026** (0.011); WTO Accession 0.036*** (0.010).
    - Columns (3)–(6): Index coefficients remain positive and significant across instrument specifications (range 0.022* to 0.026**); WTO Accession consistently 0.036***.
  - Observations: 8,147; Country and Year fixed effects included.
- Falsification tests (Table 7):
  - Randomization exercises: randomize Pre-Accession Index by member and by year.
  - Results:
    - Randomized by member: Index -0.001 (0.007) (insignificant).
    - Randomized by year: Index -0.020* (0.010) (weakly negative).
  - Conclusion: significant positive effects disappear in falsification samples, supporting validity of instruments and identification.

### Key quantitative findings and magnitudes
- Benchmark estimates imply:
  - General accession (all members) post-accession growth increase: 0.8 percentage points (Column (1) general dummy).
  - WTO (Article XII) accession post-accession increase: 0.033 to 0.036 (3.3 to 3.6 percentage points?) — note: coefficients reported are in decimal form representing percentage point changes in growth rates (e.g., 0.036***).
  - Pre-Accession Index coefficient around 0.025–0.030 in preferred specifications (Table 3 Column (3): 0.027* (0.015); IV Table 6 columns 1–6: 0.022* to 0.026**).
  - At mean index value 0.56, authors report average effect of WTO entry negotiations of 1.5 percentage points (as calculated in source).
- Article XII members summary (sample):
  - Total questions answered (sample average): 1,019.
  - Total commitments per Article XII member (sample average): 40.
  - Total legislations enacted per Article XII member (sample average): 136.
  - Sample size Table 1: Article XII sample size 32; All economies sample size 117 for the table’s years.

### Robustness summary and interpretation
- Results robust to:
  - Inclusion of Human Capital Index and Conflicts (coefficients largely unchanged; some loss of observations when adding Human Capital).
  - Excluding China, excluding EU enlargement countries, excluding growth outliers.
  - Subsample heterogeneity checks showing stronger pre-accession effects for longer negotiations, more questions, and more recent accessions.
  - Instrumental variable strategies using geographic and temporal proximity instruments.
  - Falsification tests where randomization removes significant effects.
- Interpretation:
  - Pre-accession negotiations (captured by the Pre-Accession Commitment Index) are associated with measurable increases in GDP per capita growth for Article XII acceding members, consistent with reforms and commitments undertaken during negotiations driving part of the growth effect.
  - The accession negotiation process functions not only to secure improved market access but also as a mechanism prompting domestic reforms and institutional alignment with WTO standards, which can materialize prior to formal accession.

*Source: IMF Working Paper (excerpt provided from wpiea2024207-print-pdf).*

### Conclusions

### wpiea2024207-print-pdf - Conclusions

### Main conclusions
- Reforms required to join the WTO have become more demanding in several policy areas compared to accessions under the GATT, including agriculture, structural reforms, tariff bindings, and many others.
- The study concludes that these efforts pay off and help acceding members develop their economies.
- Using a novel index that proxies for the incremental nature of accession-related reforms and a difference-in-difference methodology, the WTO accession process is associated with:
  - faster growth before accession, and
  - continued faster growth after accession.
- Several robustness tests were run, including instrumental variable estimations, to address potential concerns of endogeneity.

### Methodology and robustness
- Novel pre-accession index: proxies for the incremental nature of accession-related reforms.
- Empirical approach: difference-in-difference methodology.
- Robustness: several tests including instrumental variable estimations to address endogeneity concerns.

### Research agenda / Open questions
- The results call for further research to identify the specific types of reforms undertaken during WTO accession that have the greatest impact on growth.
- Proposed next step: map the specific reforms to commitments undertaken during negotiations to compare the effectiveness of different types of reforms for growth.

### Annex 1 — Replication of Tang and Wei (2009): Figure 2 regression results
- Dependent Variable: [log(Y)t - log(Y)t-1]
- Sample details:
  - Observations: 8,022
  - Number of groups: 150
  - Country F.E.: Yes
  - Year F.E.: Yes
  - R-squared: 9.8% (Column 1), 9.6% (Column 2), 9.3% (Column 3)
- Time profile coefficients (with standard errors in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1):
  - Column (1) Article XII:
    - -5: 0.035** (0.016)
    - -4: 0.029* (0.016)
    - -3: 0.001 (0.016)
    - -2: 0.033** (0.016)
    - -1: 0.039** (0.016)
    - 0: 0.026* (0.016)
    - 1: 0.019 (0.016)
    - 2: 0.032** (0.016)
    - 3: 0.023 (0.016)
    - 4: 0.047*** (0.016)
    - 5: 0.036** (0.017)
    - Beyond: 0.047*** (0.007)
  - Column (2) Article XXVI:
    - -5: -0.001 (0.015)
    - -4: 0.007 (0.015)
    - -3: 0.015 (0.014)
    - -2: 0.003 (0.013)
    - -1: 0.037*** (0.013)
    - 0: 0.005 (0.013)
    - 1: 0.018 (0.013)
    - 2: 0.033*** (0.013)
    - 3: 0.013 (0.013)
    - 4: -0.014 (0.013)
    - 5: 0.014 (0.013)
    - Beyond: -0.010* (0.005)
  - Column (3) Article XXXIII:
    - -5: 0.017 (0.021)
    - -4: -0.004 (0.021)
    - -3: 0.013 (0.021)
    - -2: 0.023 (0.020)
    - -1: 0.008 (0.020)
    - 0: 0.016 (0.019)
    - 1: 0.019 (0.019)
    - 2: 0.025 (0.019)
    - 3: -0.004 (0.017)
    - 4: 0.011 (0.017)
    - 5: -0.005 (0.017)
    - Beyond: 0.003 (0.006)
- Additional regressors:
  - log(Y)t-1:
    - Column (1): -0.034*** (0.003)
    - Column (2): -0.032*** (0.003)
    - Column (3): -0.031*** (0.003)
  - log(Investment/GDP)it (Gross Fixed Capital Formation):
    - Column (1): 0.017*** (0.002)
    - Column (2): 0.017*** (0.002)
    - Column (3): 0.016*** (0.002)
  - log(Trade Openness/GDP)it (Imports+Exports/GDP):
    - Column (1): 0.002 (0.001)
    - Column (2): 0.004*** (0.001)
    - Column (3): 0.004*** (0.001)
  - Constant:
    - Column (1): 0.341*** (0.026)
    - Column (2): 0.332*** (0.026)
    - Column (3): 0.325*** (0.026)

### Annex 2 — Details of the Pre-Accession Index (Viet Nam example)
- The index is a step function: the number of answers can only increase in every consecutive year.
- Viet Nam case:
  - Total answers: 3511 questions answered.
  - WP meetings attended: 14 meetings between 1998 and 2006.
  - Before first meeting (July 1998): 655 questions answered (document released in March 1998).
  - By third meeting (July 1999): additional Q&A documents released April (463 questions) and July 1999 (68 questions); index in 1999 sums to 1216 replies (655+463+68).
- Three operational rules of the index:
  - The index can only change in WP meeting years (answers released in non-meeting years are counted in the next meeting year).
  - Only answers released prior to meetings are considered.
  - In the years after accession the pre-accession index ends and takes the value 0 (implicit interaction with a dummy for the pre-accession period).

### Annex 3 — Examples of questions and answers leading to reforms
- Example 1: Viet Nam — excerpts concerning treatment of domestic and foreign investment
  - Timeline and legislative reforms referenced:
    - Decree No. 61/2003/ND-CP (6 June 2003): Department of Foreign Investment set up to administer foreign investment activities.
    - Decree No. 27/2003/ND-CP (19 March 2003): amendment extending list of projects subject to registration, narrowing conditional sectors, clarifying criteria for issuing or rejecting licences.
    - Decree No. 38/2003/ND-CP (15 April 2003): allowing transformation of FDI enterprises into stock-holding companies.
    - Unified Law on Investment and Law on Enterprise drafted and scheduled for submission to the National Assembly for approval in 2006.
    - Government requested National Assembly to pass these two laws at its 8th Session in October 2005; drafts to be submitted to the Working Party when ready.
  - Specific accession-related issues and replies:
    - Request to introduce a “one-stop” service and simplify procedures; reply describes administrative reforms, decentralization plans, coordination among agencies, and a review to abolish out-dated/overlapping provisions.
    - Question on allowing stock-holding companies; reply cites Decree No. 38/2003/ND-CP and consideration of a uniformly applied enterprise law for domestic and foreign investors.
    - Questions on eliminating discrimination between foreign and domestic investors; reply indicates gradual elimination and drafting of unified laws to be submitted in 2006.
    - Conditional investment sectors: draft Investment Law will be supported by a Government Decree detailing lists and procedures; maximum foreign shareholding is subject to Viet Nam's commitments in services (capped equity capital of 49 percent referenced as not determined by the draft law text).

- Example 2: Comoros — excerpts concerning liberalization of the state monopoly on rice imports and VAT implementation
  - Rice import monopoly (ONICOR):
    - 23 June 2020 (WT/ACC/COM/27): Question expressed concern that only one company (ONICOR) imports ordinary rice; reply: Government maintained monopoly to ensure continuous supply, guarantee quality, and stabilize market prices.
    - 21 June 2021 (WT/ACC/COM/30): Request to eliminate exclusive right prior to accession; reply: Comoros is considering opening the rice market to competition and outlined steps including establishment of a Board of Directors, a socioeconomic impact study (2022), and review of the 1982 law.
    - 5 May 2022 (WT/ACC/COM/34): Update request; reply: World Bank conducted an impact assessment; two short/medium-term solutions considered for ONICOR: 1. role as a central procurement agency (transition); 2. management of strategic rice stocks.
    - 14 October 2022 (WT/ACC/COM/38): Reiteration to eliminate exclusivity and discontinue government validation of administered prices; reply: review under consideration and impact assessment to be shared.
    - 6 September 2023 (WT/ACC/COM/42): Timeline request; reply: Comoros decided to end the State monopoly on rice pursuant to Article 25 of Decree No. 23/60/PR of 30 June 2023 — importation and sale of ordinary rice liberalized and practice of government validating administered prices discontinued.
  - VAT implementation:
    - 7 December 2023 (WT/ACC/COM/49): Question noted VAT implementation was foreseen in 2026 and asked for state of play; reply: Government scheduled VAT implementation for 2027 and will shortly request technical assistance from the IMF; a detailed action plan for VAT implementation will be submitted to WTO Members as soon as possible (see document WT/ACC/COM/47).

*IMF WORKING PAPERS Committing to Grow: The Full Impact of WTO Accessions — Conclusions and Annexes*

### Annex 4: Economies in the estimation sample

### Annex 4: Economies in the estimation sample

### Economies (as listed in the source)
- Albania¹
- Algeria
- Angola²
- Anguilla
- Antigua and Barbuda²
- Argentina
- Armenia¹
- Aruba
- Azerbaijan
- Bahrain²
- Bangladesh
- Barbados²
- Belarus
- Belize²
- Benin²
- Bermuda
- Bhutan
- Bolivia
- Bosnia and Herzegovina
- Botswana²
- Brazil
- British Virgin Islands
- Brunei Darussalam²
- Bulgaria¹
- Burkina Faso²
- Burundi²
- Cabo Verde¹
- Cambodia¹
- Cameroon²
- Cayman Islands
- Central African Republic²
- Chad²
- Chile
- China¹
- Colombia
- Comoros
- Costa Rica
- Croatia¹
- Côte d'Ivoire²
- Curacao
- Czech Republic
- D.R. Congo
- Djibouti²
- Dominica²
- Dominican Republic
- Ecuador¹
- Egypt
- El Salvador
- Equatorial Guinea
- Eritrea (not listed) 
- Estonia¹
- Eswatini²
- Ethiopia
- Fiji²
- Gabon²
- Gambia, The (The Gambia²)
- Georgia¹
- Ghana²
- Greece (not listed)
- Grenada²
- Guatemala
- Guinea²
- Guinea-Bissau²
- Haiti
- Honduras
- Hungary
- India
- Indonesia²
- Iraq
- Iran
- Ireland (not listed)
- Isle of Man (not listed)
- Jamaica²
- Japan (not listed)
- Jordan¹
- Kazakhstan¹
- Kenya²
- Korea, Rep. (not listed)
- Kosovo (not listed)
- Kurdistan (not listed)
- Kuwait²
- Kyrgyz Republic
- Lao P.D.R.¹
- Latvia¹
- Lebanon
- Lesotho²
- Liberia¹
- Libya (not listed)
- Lithuania¹
- Luxembourg (not listed)
- Madagascar²
- Malawi²
- Malaysia²
- Maldives²
- Mali²
- Malta²
- Mauritania²
- Mauritius²
- Mexico
- Moldova¹
- Mongolia¹
- Montenegro¹
- Morocco
- Mozambique²
- Myanmar
- Namibia²
- Nepal¹
- Netherlands (not listed)
- New Zealand (not listed)
- Nicaragua
- Niger²
- Nigeria²
- North Macedonia
- Norway (not listed)
- Oman¹
- Pakistan
- Panama¹
- Paraguay
- Peru
- Philippines
- Poland
- Portugal (not listed)
- Qatar²
- Romania
- Russia¹
- Rwanda²
- St. Kitts and Nevis²
- St. Lucia²
- St. Vincent and the Grenadines²
- Saint Martin / Sint Maarten (Sint Marteen)
- Samoa (not listed)
- Sao Tomé and Príncipe
- Saudi Arabia¹
- Senegal²
- Serbia
- Seychelles¹
- Sierra Leone²
- Singapore (not listed)
- Sint Marteen
- Slovak Republic
- Slovenia
- Solomon Islands (not listed)
- Somalia (not listed)
- South Africa
- South Sudan (not listed)
- Spain (not listed)
- Sri Lanka
- State of Palestine
- Sudan
- Suriname²
- Sweden (not listed)
- Switzerland (not listed)
- Syria
- Tajikistan¹
- Tanzania²
- Thailand
- Togo²
- Tonga (not listed)
- Trinidad and Tobago²
- Tunisia
- Turkey (Türkiye)
- Turkmenistan
- Turks and Caicos Islands
- Uganda²
- Ukraine¹
- United Arab Emirates²
- United Kingdom (not listed)
- United States (not listed)
- Uruguay
- Uzbekistan
- Venezuela
- Viet Nam¹
- Yemen¹
- Zambia²
- Zimbabwe

### Footnotes and classifications
- ¹ Article XII
- ² Article XXVI 5(c)

*Annex 4: Economies in the estimation sample — IMF Working Paper*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024207-print-pdf.pdf_