## Annex I. Technical Results

## Source details

**Canonical URL:** [Annex I. Technical Results](https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023252-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2023/english/wpiea2023252-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2023/english/wpiea2023252-print-pdf.pdf.json)

---

### Contents and Structure
- Annex I appears on page "17".
- The document also includes:
  - "Annex II. Regression Results" on page "19".
  - "Annex III. Robustness Results" on page "25".
  - "References" on page "28".
- Figures Referenced:
  - Figure 1. Legal Betas, Female LFP Rate
  - Figure 2. Legal Betas Versus Weights, Female LFP Rate
  - Figure 3. Legal Betas, Male LFP Rate
  - Figure 4. Legal Betas, Gender Gap in LFP Rate
  - Figure 5. Proportion of Countries with Selected Laws
  - Figure 6. Geographical Distribution of Selected Laws, 2022
- The document includes a section labeled "TABLES" (no specific table titles provided in the supplied content).

### Regression Results for the WBL Index — Overview and Key Aggregate Findings
- Motivation and objective:
  - Indices (such as the overall WBL index) obscure which specific laws matter, their magnitudes, and assume homogeneous effects across components; the paper analyzes each of the 35 WBL law indicators individually to assess associations with female and male prime-age (25–54) labor force participation (LFP) rates.
  - Primary objective: identify laws that increase female LFP while not adversely affecting male LFP, thereby narrowing the gender gap and increasing overall LFP.
- Data and methods:
  - WBL database: 35 law indicators, 190 countries, calendar years 1970–2021; data range 1971–2022.
  - ILO LFP rates for ages 25 to 54 (Modelled Estimates from 1990 onward; national estimates prior to 1990).
  - World Bank WDI real GDP per capita.
  - Time horizon: uses data up to 2019 to avoid COVID-19 anomalies.
  - Empirical specification: fixed-effects panel regressions estimated by OLS; controls include country fixed effects, year fixed effects, and three-year lag of log real GDP per capita; standard errors clustered at the country level.
- Index definition and interpretation:
  - Index: I_{i,t} = 1/35 ∑ P_{k,i,t} (equation (1)).
  - The index estimate equals a weighted sum of the individual law estimates: β̂ = ∑ w_k β̂_k (35 k=1) (equation (4)) where β̂_k are “legal betas” and w_k β̂_k are “weighted betas.”

- Table 1 (index-level regression results; all regressions include country and year fixed effects and three-year lag of log real GDP per capita; robust standard errors clustered at the country level):
  - Women: WBL index 0.110*** (standard error (0.042))
  - Men: WBL index 0.018 (standard error (0.017))
  - Gender Gap (male LFP rate minus female LFP rate): WBL index -0.092** (standard error (0.041))
  - N 5,512 for all regressions.
  - R2: Women 0.536; Men 0.156; Gender Gap 0.609.
  - Significance notation: *** p<0.01, ** p<0.05, * p<0.1.

- Average-marginal-effect interpretation:
  - Dividing the index coefficients by 35 yields the average marginal effect of enacting one additional law (implicitly assuming homogeneous effects across laws).
  - Enacting one additional law (index/35) increases female LFP by 0.31 percentage points.
  - Enacting one additional law decreases the gender gap by 0.26 percentage points.
  - These average effects are statistically significant but small in magnitude, motivating the disaggregated analysis.

### Indicator-by-Indicator Analysis — Heterogeneity and Notable Estimates
- Heterogeneity across the 35 laws:
  - Point estimates range from -3.5 to 6.3 percentage points for effects on female LFP, with many estimates imprecise.
  - Of the 35 laws, 25 are not statistically significantly related to female LFP rates.
  - Ten laws are statistically significant for female LFP; of these ten, 9 are positively related to female LFP.
- Selected specific-law magnitudes reported exactly:
  - Allowing women to obtain a judgment of divorce in the same way as men: associated with an average increase in female LFP of 6.3 percentage points (the largest positive point estimate).
  - Legislation explicitly addressing domestic violence: associated with an average increase in female LFP of 1.4 percentage points (the least positively correlated among significant positive laws).
  - Allowing women to travel outside the country in the same way as men: negatively correlated with female LFP (the only significant negative estimate reported; specific point estimate falls within the -3.5 to 6.3 percentage point range).

- Weights and index composition:
  - The index estimate is a weighted sum of law-specific estimates; weights w_k do not necessarily sum to one and do not systematically align with higher or more positive legal betas.
  - Example: the divorce law has the highest point estimate but a very small weight in the index; a law prohibiting employment discrimination by gender has a near-zero point estimate but is heavily weighted in the index.
  - Conclusion: the aggregate index can mask the most relevant laws for female LFP due to uneven weighting.

- Male LFP and gender gap:
  - Most law-specific estimates for male LFP are close to zero and not statistically significant.
  - Only three of the 35 laws have a significant relationship with male LFP rates.
  - In general, improving women’s legal rights has no adverse effect on men’s labor force participation.
  - Any law associated with an increase (decrease) in female LFP is also associated with a decrease (increase) in the gender gap (male minus female LFP).

- Robustness checks:
  - Additional controls tested: squared term of GDP per capita, fertility rate (total births per woman), and age dependency ratio (percentage of working-age population).
  - Sample size remains nearly identical after these controls (only 7 observations lost).
  - Results are consistent with the baseline approach (robustness results provided in Annex III; education controls not included due to severe sample reduction concerns).

### Discussion: Mechanisms, Temporal Patterns, and Geographic Distribution
- Laws with strong positive associations (mechanisms described):
  - Equal rights to divorce and to remarry: increase autonomy, financial independence, incentives to invest in education and skills, and reduce stigma, raising labor market participation.
  - Equal rights in industrial jobs: expand access to male-dominated fields, training, and occupational diversification.
  - Equal rights to a passport, travel, and choice of residence: allow mobility for work and education.
  - Paid maternity leave: provides financial support during childbirth and early childcare; caveat that maternity leave can induce faster skill loss relative to men if paternity leave is absent or short, but findings suggest positive effects on women’s LFP outweigh negatives on average.
  - Prohibition of gender-based credit discrimination: expands access to credit, enabling entrepreneurship and greater labor force attachment.
  - Legislation addressing domestic violence: provides legal protection and support, enabling safer labor market participation.

- Negative association example and interpretation:
  - Allowing women to travel outside the country in the same way as men is negatively correlated with female LFP in the country of origin; a possible mechanism is that women access job opportunities abroad, reducing domestic LFP.

- Evolution over time and prevalence (exact figures preserved):
  - Restrictions on women’s right to travel outside the home were rare in 1970 and remained largely unchanged up to 2019.
  - Share of countries requiring at least 14 days of paid paternity leave rose from 20 percent in 1970 to around 60 percent in 2019.
  - Equal legal access to credit: around 40 percent of countries now guarantee it.
  - Domestic violence laws: approximately 80 percent of countries have domestic violence laws in 2019, up from about 3 percent in 1990.
  - Between-country variation remains substantial; some regions (e.g., Middle East and Northern Africa) retain many gender-unequal laws relevant for female LFP.

- Geographic distribution:
  - Maps (Figure 6) identify countries/regions where women lack equal legal rights in laws relevant to female LFP as of 2022.

### Two-stage Estimation Procedure, Residuals, and Weight Construction (Technical)
- Two-stage estimation:
  - First stage (purge controls from main independent variables):
    - I_{i,t-3} = δ + φ X_{i,t-3} + λ_i + θ_i + u_{i,t}, (5)
    - P_{k,i,t-3} = δ_k + φ_k X_{i,t-3} + λ_{k,i} + θ_{k,i} + u_{k,i,t}, (6)
    - u_{i,t} and u_{k,i,t} are error terms.
  - Second stage (regress residuals on dependent variables):
    - y_{i,t} = ρ + β R_{i,t} + v_{i,t}, (7)
    - y_{i,t} = ρ_k + β_k R_{k,i,t} + v_{k,i,t}, (8)
    - Estimated β̂ and β̂_k from (7) and (8) are identical to those from regressions (2) and (3).
- Relationship between index residuals and specific-law residuals:
  - R_{i,t} = I_{i,t-3} − Î_{i,t-3} = (1/35) ∑_{k=1}^{35} (P_{k,i,t-3} − P̂_{k,i,t-3}) = (1/35) ∑_{k=1}^{35} R_{k,i,t}. (9)
  - Interpretation: the index residual R_{i,t} equals the simple average of the 35 specific-law residuals R_{k,i,t}.
- Weight construction:
  - w_k ≡ (1/35) × var(R_{k,i,t}) / var(R_{i,t}). (14)
- Second-stage OLS formulas (sample covariance/variance representation):
  - β̂ = cov(y_{i,t}, R_{i,t}) / var(R_{i,t}). (10)
  - β̂_k = cov(y_{i,t}, R_{k,i,t}) / var(R_{k,i,t}). (11)
  - cov(y_{i,t}, R_{i,t}) = (1/35) ∑_{k=1}^{35} cov(y_{i,t}, R_{k,i,t}). (12)
  - Combining results yields: β̂ = ∑_{k=1}^{35} w_k β̂_k. (13)

### Annex II — Regression Results: Structure and Selected Exact Estimates
- Tables 2–4 report, for each policy variable, the following fields in regressions for Female LFP Rate (Table 2), Male LFP Rate (Table 3), Gender Gap in LFP Rate (Table 4):
  - β̂_k, S.E., p-value, R^2, w_k, w_k β̂_k, N.
- All table rows use N = 5,512.

- Selected exact entries from Table 2 (Female LFP Rate):
  - "Can a woman apply for a passport in the same way as a man?": β̂_k = 0.037; S.E. = 0.015; p-value = 0.015; R^2 = 0.531; w_k = 0.073; w_k β̂_k = 0.003; N = 5,512.
  - "Can a woman travel outside the country in the same way as a man?": β̂_k = −0.032; S.E. = 0.008; p-value = 0.000; R^2 = 0.529; w_k = 0.017; w_k β̂_k = −0.001; N = 5,512.
  - "Can a woman travel outside her home in the same way as a man?": β̂_k = 0.032; S.E. = 0.006; p-value = 0.000; R^2 = 0.529; w_k = 0.013; w_k β̂_k = 0.000; N = 5,512.

- Selected exact entries from Table 3 (Male LFP Rate):
  - "Can a woman apply for a passport in the same way as a man?": β̂_k = −0.001; S.E. = 0.005; p-value = 0.827; R^2 = 0.154; w_k = 0.073; w_k β̂_k = 0.000; N = 5,512.
  - "Can a woman travel outside the country in the same way as a man?": β̂_k = 0.043; S.E. = 0.002; p-value = 0.000; R^2 = 0.160; w_k = 0.017; w_k β̂_k = 0.001; N = 5,512.
  - "Can a woman travel outside her home in the same way as a man?": β̂_k = −0.028; S.E. = 0.002; p-value = 0.000; R^2 = 0.156; w_k = 0.013; w_k β̂_k = 0.000; N = 5,512.

- Selected exact entries from Table 4 (Gender Gap in LFP Rate):
  - "Can a woman apply for a passport in the same way as a man?": β̂_k = −0.038; S.E. = 0.016; p-value = 0.016; R^2 = 0.606; w_k = 0.073; w_k β̂_k = −0.003; N = 5,512.
  - "Can a woman travel outside the country in the same way as a man?": β̂_k = 0.076; S.E. = 0.008; p-value = 0.000; R^2 = 0.606; w_k = 0.017; w_k β̂_k = 0.001; N = 5,512.
  - "Can a woman travel outside her home in the same way as a man?": β̂_k = −0.060; S.E. = 0.006; p-value = 0.000; R^2 = 0.605; w_k = 0.013; w_k β̂_k = −0.001; N = 5,512.

- Further entries in Tables 2–4 follow the same format and report β̂_k, S.E., p-value, R^2, w_k, w_k β̂_k, with N = 5,512 for each policy row.

### Conclusion and Policy Implications (from the annex)
- Detailed, law-by-law analysis provides actionable insights that aggregate indices obscure.
- Laws concerning intra-household bargaining power (women’s equal rights to divorce and remarry) appear particularly associated with larger increases in female LFP.
- Improving specific legal rights can reduce the gender gap in LFP primarily via increased female participation without lowering male participation.
- Policy guidance:
  - Governments should prioritize legal reforms with the largest estimated associations with female LFP, taking into account country-specific norms and implementation costs.
  - Legal reforms should be complemented by efforts to ensure de-facto implementation and to consider welfare, human rights, and non-labor-market objectives (for example, protection from workplace harassment and domestic violence).
- Next steps for research and data:
  - Investigate de-facto implementation of laws and the time profile from legal change to observable labor-market outcomes.
  - The World Bank’s ongoing collection of data on implementation of gender equality laws will enhance understanding.

*Source: wpiea2023252-print-pdf - Annex I. Technical Results (https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023252-print-pdf.pdf)*

### Annex I. Technical Results .............................................................................................

### Annex I. Technical Results

### Contents and Structure
- Annex I appears on page "17".
- The document also includes:
  - "Annex II. Regression Results" on page "19".
  - "Annex III. Robustness Results" on page "25".
  - "References" on page "28".

### Figures Referenced in the Document
- Figure 1. Legal Betas, Female LFP Rate
- Figure 2. Legal Betas Versus Weights, Female LFP Rate
- Figure 3. Legal Betas, Male LFP Rate
- Figure 4. Legal Betas, Gender Gap in LFP Rate
- Figure 5. Proportion of Countries with Selected Laws
- Figure 6. Geographical Distribution of Selected Laws, 2022

### Tables
- The document includes a section labeled "TABLES" (no specific table titles provided in the supplied content).

*Source: wpiea2023252-print-pdf - Annex I. Technical Results (https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023252-print-pdf.pdf)*

### 1. Regression Results for the WBL Index ................................................................................

### 1. Regression Results for the WBL Index

### Introduction
- Around 2.4 billion working-age women live in economies that do not grant them the same rights as men.
- Women globally enjoy only 77 percent of the legal rights that men do.
- At the current pace of reform, it would take at least 50 years to reach legal gender equality everywhere.
- Motivation: indices (such as the overall WBL index) obscure which specific laws matter, their magnitudes, and assume homogeneous effects across components; this paper analyzes each of the 35 WBL law indicators individually to assess associations with female and male prime-age (25–54) labor force participation (LFP) rates.
- Primary objective: identify laws that increase female LFP while not adversely affecting male LFP, thereby narrowing the gender gap and increasing overall LFP.

### Data and Methods
- Data sources: Women, Business and the Law (WBL) database (35 law indicators, 190 countries, calendar years 1970–2021; data range 1971–2022), ILO LFP rates for ages 25 to 54 (Modelled Estimates from 1990 onward; national estimates prior to 1990), and World Bank WDI real GDP per capita.
- Time horizon for analysis: uses data up to 2019 to avoid COVID-19 anomalies.
- Empirical specification:
  - Fixed-effects panel regressions estimated by OLS.
  - Main independent variables: overall WBL index (mean of 35 law indicators) and each of the 35 law indicators (푃푘,푖,푡).
  - Controls: country fixed effects, year fixed effects, and three-year lag of log real GDP per capita (variables measured with a three-year lag).
  - Standard errors clustered at the country level.
- Interpretation:
  - The index is defined as 퐼푖,푡 = 1/35 ∑푃푘,푖,푡 (equation (1)).
  - Regression forms shown in equations (2) and (3); the index estimate equals a weighted sum of the individual law estimates (equation (4)): 훽̂ = ∑푤푘 훽̂푘 (35 k=1), where 훽̂푘 are “legal betas” and 푤푘훽̂푘 are “weighted betas.”

### Index-Level Analysis (Aggregate WBL Index)
- Table 1: Regression results for the WBL index (all regressions include country and year fixed effects and three-year lag of log real GDP per capita; robust standard errors clustered at the country level).
  - Women: WBL index 0.110*** (standard error (0.042))
  - Men: WBL index 0.018 (standard error (0.017))
  - Gender Gap (male LFP rate minus female LFP rate): WBL index -0.092** (standard error (0.041))
  - N 5,512 for all regressions.
  - R2: Women 0.536; Men 0.156; Gender Gap 0.609.
  - Significance notation: *** p<0.01, ** p<0.05, * p<0.1.
- Average-marginal-effect interpretation:
  - Dividing the index coefficients by 35 yields the average marginal effect of enacting one additional law (implicitly assuming homogeneous effects across laws).
  - Enacting one additional law (index/35) increases female LFP by 0.31 percentage points.
  - Enacting one additional law decreases the gender gap by 0.26 percentage points.
  - These average effects are statistically significant but small in magnitude, motivating the disaggregated analysis.

### Analysis of Specific Laws (Indicator-by-Indicator)
- Heterogeneity across the 35 laws:
  - Point estimates range from -3.5 to 6.3 percentage points for effects on female LFP, with many estimates imprecise.
  - Of the 35 laws, 25 are not statistically significantly related to female LFP rates.
  - Ten laws are statistically significant for female LFP; of these ten, 9 are positively related to female LFP.
- Examples of specific law estimates (indicator-level magnitudes reported exactly as in the text):
  - Allowing women to obtain a judgment of divorce in the same way as men: associated with an average increase in female LFP of 6.3 percentage points (the largest positive point estimate).
  - Legislation explicitly addressing domestic violence: associated with an average increase in female LFP of 1.4 percentage points (the least positively correlated among significant positive laws).
  - Allowing women to travel outside the country in the same way as men: negatively correlated with female LFP (the only significant negative estimate reported; specific point estimate falls within the -3.5 to 6.3 percentage point range).
- Weights and index composition:
  - The index estimate is a weighted sum of law-specific estimates; weights 푤푘 do not necessarily sum to one and do not systematically align with higher or more positive legal betas.
  - Example: the divorce law has the highest point estimate but a very small weight in the index; a law prohibiting employment discrimination by gender has a near-zero point estimate but is heavily weighted in the index.
  - Conclusion: the aggregate index can mask the most relevant laws for female LFP due to uneven weighting.
- Male LFP and gender gap:
  - Most law-specific estimates for male LFP are close to zero and not statistically significant.
  - Only three of the 35 laws have a significant relationship with male LFP rates.
  - In general, improving women’s legal rights has no adverse effect on men’s labor force participation.
  - Any law associated with an increase (decrease) in female LFP is also associated with a decrease (increase) in the gender gap (male minus female LFP).
- Robustness:
  - Additional controls tested: squared term of GDP per capita, fertility rate (total births per woman), and age dependency ratio (percentage of working-age population).
  - Sample size remains nearly identical after these controls (only 7 observations lost).
  - Results are consistent with the baseline approach (robustness results provided in Annex III; education controls not included due to severe sample reduction concerns).

### Discussion: Mechanisms and Patterns
- Laws related to intra-household bargaining and women’s agency show strong positive associations with female LFP:
  - Equal rights to divorce and to remarry: can increase women’s autonomy, financial independence, incentives to invest in education and skills, and reduce stigma—all channels that can increase labor market participation.
  - Equal rights in industrial jobs: expand access to traditionally male-dominated fields, training, and occupational diversification, reducing occupational segregation.
  - Equal rights to a passport, travel, and choice of residence: allow mobility for work and education, broadening employment opportunities and access to skill development.
  - Paid maternity leave: provides financial support during childbirth and early childcare, easing the economic burden and facilitating remaining in or returning to the workforce (paper notes a caveat that maternity leave can induce faster skill loss relative to men if paternity leave is absent or short; nevertheless, findings suggest positive effects of maternity leave on women’s LFP outweigh negative effects on average).
  - Prohibition of gender-based credit discrimination: expands women’s access to credit, enabling entrepreneurship and greater labor force attachment.
  - Legislation addressing domestic violence: provides legal protection and support, enabling safer labor market participation and increased economic independence.
- Negative association example:
  - Allowing women to travel outside the country in the same way as men is negatively correlated with female LFP in the country of origin; possible mechanism: women may access job opportunities abroad, reducing domestic LFP.
- Evolution over time and prevalence:
  - Restrictions on women’s right to travel outside the home were rare in 1970 and remained largely unchanged up to 2019.
  - Share of countries requiring at least 14 days of paid paternity leave rose from 20 percent in 1970 to around 60 percent in 2019.
  - Equal legal access to credit: around 40 percent of countries now guarantee it.
  - Domestic violence laws: approximately 80 percent of countries have domestic violence laws in 2019, up from about 3 percent in 1990.
  - Between-country variation remains substantial; some regions (e.g., Middle East and Northern Africa) retain many gender-unequal laws relevant for female LFP.
- Geographic distribution:
  - Maps (Figure 6) identify countries/regions where women lack equal legal rights in laws relevant to female LFP as of 2022.

### Conclusion and Policy Implications
- Detailed, law-by-law analysis provides actionable insights that aggregate indices obscure.
- Laws concerning intra-household bargaining power (women’s equal rights to divorce and remarry) appear particularly associated with larger increases in female LFP.
- Improving specific legal rights can reduce the gender gap in LFP primarily via increased female participation without lowering male participation.
- Policy guidance:
  - Governments should prioritize legal reforms with the largest estimated associations with female LFP, taking into account country-specific norms and implementation costs.
  - Legal reforms should be complemented by efforts to ensure de-facto implementation and to consider welfare, human rights, and non-labor-market objectives (for example, protection from workplace harassment and domestic violence).
- Next steps for research and data:
  - Investigate de-facto implementation of laws and the time profile from legal change to observable labor-market outcomes.
  - The World Bank’s ongoing collection of data on implementation of gender equality laws will enhance understanding.

*Source: wpiea2023252-print-pdf — 1. Regression Results for the WBL Index.*

### Annex I. Technical Results

### Annex I. Technical Results

### Two-stage estimation procedure
- Regressions (2) and (3) can be estimated using a two-stage process:
  - First stage: purge controls from main independent variables by estimating
    - I_{i,t-3} = δ + φ X_{i,t-3} + λ_i + θ_i + u_{i,t}, (5)
    - P_{k,i,t-3} = δ_k + φ_k X_{i,t-3} + λ_{k,i} + θ_{k,i} + u_{k,i,t}, (6)
    - where u_{i,t} and u_{k,i,t} are the error terms.
  - Second stage: regress the residuals from (5) and (6), denoted R_{i,t} and R_{k,i,t}, on the dependent variables:
    - y_{i,t} = ρ + β R_{i,t} + v_{i,t}, (7)
    - y_{i,t} = ρ_k + β_k R_{k,i,t} + v_{k,i,t}, (8)
    - where v_{i,t} and v_{k,i,t} are the error terms.
- The estimated coefficients β̂ and β̂_k from (7) and (8) are identical to those from regressions (2) and (3).

### Relationship between index residuals and specific-law residuals; weight construction
- Let Î_{i,t-3} and P̂_{k,i,t-3} be the fitted values from (5) and (6). By definition of residuals:
  - R_{i,t} = I_{i,t-3} − Î_{i,t-3} = (1/35) ∑_{k=1}^{35} (P_{k,i,t-3} − P̂_{k,i,t-3}) = (1/35) ∑_{k=1}^{35} R_{k,i,t}. (9)
  - Interpretation: the index residual R_{i,t} equals the simple average of the 35 specific-law residuals R_{k,i,t}.
- The weights w_k used to express the index beta as a weighted average of specific-law betas are determined by residual variances:
  - w_k ≡ (1/35) × var(R_{k,i,t}) / var(R_{i,t}). (14)

### Expressions for estimated coefficients (sample covariance/variance representation)
- Second-stage OLS formulas:
  - β̂ = cov(y_{i,t}, R_{i,t}) / var(R_{i,t}). (10)
  - β̂_k = cov(y_{i,t}, R_{k,i,t}) / var(R_{k,i,t}). (11)
- Using (9) and properties of covariance:
  - cov(y_{i,t}, R_{i,t}) = (1/35) ∑_{k=1}^{35} cov(y_{i,t}, R_{k,i,t}). (12)
- Combining results yields that the index coefficient equals a weighted average of specific-law coefficients:
  - β̂ = ∑_{k=1}^{35} w_k β̂_k, (13)
  - where w_k are given by (14) above.

### Annex II — Regression results: structure and selected exact estimates
- Tables 2–4 report, for each policy variable, the following fields in the regressions for:
  - Female LFP Rate (Table 2), Male LFP Rate (Table 3), Gender Gap in LFP Rate (Table 4):
    - β̂_k, S.E., p-value, R^2, w_k, w_k β̂_k, N.
- All table rows use N = 5,512.
- Selected exact entries from Table 2 (Female LFP Rate):
  - "Can a woman apply for a passport in the same way as a man?": β̂_k = 0.037; S.E. = 0.015; p-value = 0.015; R^2 = 0.531; w_k = 0.073; w_k β̂_k = 0.003; N = 5,512.
  - "Can a woman travel outside the country in the same way as a man?": β̂_k = −0.032; S.E. = 0.008; p-value = 0.000; R^2 = 0.529; w_k = 0.017; w_k β̂_k = −0.001; N = 5,512.
  - "Can a woman travel outside her home in the same way as a man?": β̂_k = 0.032; S.E. = 0.006; p-value = 0.000; R^2 = 0.529; w_k = 0.013; w_k β̂_k = 0.000; N = 5,512.
- Selected exact entries from Table 3 (Male LFP Rate):
  - "Can a woman apply for a passport in the same way as a man?": β̂_k = −0.001; S.E. = 0.005; p-value = 0.827; R^2 = 0.154; w_k = 0.073; w_k β̂_k = 0.000; N = 5,512.
  - "Can a woman travel outside the country in the same way as a man?": β̂_k = 0.043; S.E. = 0.002; p-value = 0.000; R^2 = 0.160; w_k = 0.017; w_k β̂_k = 0.001; N = 5,512.
  - "Can a woman travel outside her home in the same way as a man?": β̂_k = −0.028; S.E. = 0.002; p-value = 0.000; R^2 = 0.156; w_k = 0.013; w_k β̂_k = 0.000; N = 5,512.
- Selected exact entries from Table 4 (Gender Gap in LFP Rate):
  - "Can a woman apply for a passport in the same way as a man?": β̂_k = −0.038; S.E. = 0.016; p-value = 0.016; R^2 = 0.606; w_k = 0.073; w_k β̂_k = −0.003; N = 5,512.
  - "Can a woman travel outside the country in the same way as a man?": β̂_k = 0.076; S.E. = 0.008; p-value = 0.000; R^2 = 0.606; w_k = 0.017; w_k β̂_k = 0.001; N = 5,512.
  - "Can a woman travel outside her home in the same way as a man?": β̂_k = −0.060; S.E. = 0.006; p-value = 0.000; R^2 = 0.605; w_k = 0.013; w_k β̂_k = −0.001; N = 5,512.
- Further entries in Tables 2–4 follow the same format and report β̂_k, S.E., p-value, R^2, w_k, w_k β̂_k, with N = 5,512 for each policy row.

*Annex I. Technical Results — IMF working paper (selected equations, derivations, and regression-summary excerpts).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023252-print-pdf.pdf_