## _wp1287 — Appendix Tables (IMF working paper excerpt)

## Source details

**Canonical URL:** [_wp1287 — Appendix Tables (IMF working paper excerpt)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1287.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1287.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1287.pdf.json)

---

### Conceptual framework and goals
- Framework: Operationalizes North, Wallis and Weingast (NWW, 2009) transition framework distinguishing two social orders: limited access (natural state) and open access.
- Three doorstep conditions (NWW):
  - Doorstep 1 (D1): rule of law for the elites.
  - Doorstep 2 (D2): existence of perpetually lived organizations (including the state).
  - Doorstep 3 (D3): consolidated political control of all military power.
- Aims:
  - Construct indices representing the three doorsteps by combining existing institutional variables.
  - Test relationships between countries’ transition paths and economic and political performance using graphical analysis, OLS and nonlinear regressions.
  - Use Identification through Heteroskedasticity (IH) to empirically test interactions between doorsteps and economic/political variables.

### Data sources, sample, and index composition
- Sample:
  - Database consists of 108 low-, middle-, and high-income countries (Appendix Table A–5 lists full country coverage).
  - IH econometric sample: 93 countries with data for 2009; colonized = 57 countries; non-colonized = 36 countries.
- Data sources and databases used (selected):
  - Institutional Profiles Database (IPD) (2006, 2009), Polity IV (2008/2009), Economic Freedom of the World (EFW) (2008/2010), Database of Political Institutions (DPI) (2009).
- Indicator components:
  - Total of 31 selected criteria across the three doorsteps.
  - 17 of 31 criteria from Institutional Profiles database; remaining 14 from Polity IV, IPD, and Economic Freedom of the World databases.
- Doorstep-specific composition (categories and example variables):
  - D1 — three categories: Existence of rules for the elite (XREG, XCONST, PARREG, ICC 2009), Independence and impartiality of the judicial system (A3050, 2B, 2E), Elites’ respect for courts (A604, A600, A601).
  - D2 — two categories: Stability and permanence of political system (A105, A500, Interregnum, Age of parties, Finittrm, Constitutional stability), Ease of creating organizations (A602, A504, B300, B700, D100, D101).
  - D3 — three categories: Political control of legal armed forces (A201, MILITARY, Parliamentary oversight), Military interference in political life (A1002, 2D, DEFMIN), Level of armed violence (A2000, A2001, A2002).

### Aggregation methodology
- Variable standardization:
  - Individual variables standardized on a 0 (worst) to 1 (best) scale.
- Aggregation rules:
  - Each category = simple average of standardized variables in that category.
  - Each doorstep sub-index = simple average of its categories, then rescaled by the sample average.
    - Example: D1 = average of its three categories divided by 0.62 (sample average for D1 categories).
  - Overall doorsteps index (D_Overall) = simple average of the three sub-indices (described range between 0 and 1 in methods).

### Country scores and descriptive statistics
- Database: 108 countries (Table A–5 provides D1, D2, D3, D_OVERALL for each country).
- Selected country entries (D1, D2, D3, D_OVERALL):
  - Australia — 1.55, 1.36, 1.31, 1.41
  - Denmark, Finland, Netherlands, New Zealand: listed among top performers (consistently top ten).
  - Zimbabwe — 0.40, 0.72, 0.56, 0.56 (among worst ten); Mauritania also among worst ten.
  - United States (USA) — 1.32, 1.32, 1.18, 1.27
  - United Kingdom (UK) — 1.55, 1.31, 1.27, 1.38
  - India — 0.91, 0.99, 1.06, 0.99
  - China — 0.70, 0.89, 0.80, 0.80
- Average D_Overall by income group:
  - High income countries (HIC): 1.27
  - Middle income countries (MIC): 0.91
  - Low income countries (LIC): 0.80
- Variance: greatest for MIC (because low- and upper-MIC are grouped together); top MIC performers are upper-MIC; worst MIC are lower-MIC.
- Robustness: PCA-weighted aggregation did not materially affect relative country ranks.
- Relation to external index: D1 scores are very similar to the Bertelsman Transformation Index (political transformation) category.

### Correlations within and between doorsteps (Spearman & categories)
- Sub-index correlations and internal drivers:
  - D_Overall vs D1: 0.96 (p-value (0.00))
  - D_Overall vs D2: 0.86 (p-value (0.00))
  - D_Overall vs D3: 0.92 (p-value (0.00))
  - D1 vs D2: 0.82; D1 vs D3: 0.82; D2 vs D3: 0.67 (all p-value (0.00))
  - None of the three sub-indices is driven by a single category; the three sub-indices are strongly and significantly correlated with each other.
  - Note: high correlations may partly reflect a “consistency bias” from using different variables drawn from the same databases.
- Individual categories most correlated with D_Overall (decreasing magnitude):
  - Military interference in political life; Elites’ respect for courts’ decisions; Independence and impartiality of judicial system.
  - Weakest correlation with D_Overall: D2-Category 1 (stability and permanence of the political system).

### Graphical correlations: doorsteps and political/economic “symptoms” (selected ρ values; significance *** 1%)
- Doorsteps and Democracy (EIU Overall Score):
  - D1 Rescaled vs EIU Overall Score: ρ = 0.8567***
  - D2 Rescaled vs EIU Overall Score: ρ = 0.8593***
  - D3 Rescaled vs EIU Overall Score: ρ = 0.6820***
  - D Overall vs EIU Overall Score: ρ = 0.8715***
  - Key patterns: Only countries more than three quarters up D1 are fully democratic in charts; kernel density shows three clusters (high doorsteps/full democracy; below-average doorsteps/very low democracy; below-average to average doorsteps/medium democracy).
- Doorsteps and Equal Access (IPD A9040–A9043):
  - D1 Rescaled vs Access: ρ = 0.7124***
  - D2 Rescaled vs Access: ρ = 0.7029***
  - D3 Rescaled vs Access: ρ = 0.5909***
  - D Overall vs Access: ρ = 0.7303***
  - Note: D3 correlation is quadratic; D_Overall correlation strongest, implying all three doorsteps needed to provide equal access.
- Doorsteps and Corruption (CPI):
  - D1 Rescaled vs CPI: ρ = 0.8838***
  - D2 Rescaled vs CPI: ρ = 0.7763***
  - D3 Rescaled vs CPI: ρ = 0.6597***
  - D Overall vs CPI: ρ = 0.8534***
  - Relationship quadratic: increasing returns in reducing corruption as doorstep conditions improve; about three-quarters of a doorstep condition before corruption reaches 50 percent on CPI scale.
- Doorsteps and Government Expenditure / Education:
  - D1 Rescaled vs Expenditure: ρ = 0.6863***
  - D2 Rescaled vs Expenditure: ρ = 0.5465***
  - D3 Rescaled vs Expenditure: ρ = 0.6575***
  - D Overall vs Expenditure: ρ = 0.7010***
  - D1 Rescaled vs Education/GDP: ρ = 0.4863***
  - D2 Rescaled vs Education/GDP: ρ = 0.5023***
  - D3 Rescaled vs Education/GDP: ρ = 0.4809***
  - D Overall vs Education/GDP: ρ = 0.5308***
  - Education expenditure correlation strongest for D2 and is quadratic (increasing returns).
- Doorsteps and Territorial Coverage of Public Services (A905):
  - D1 Rescaled vs A905: ρ = 0.7392***
  - D2 Rescaled vs A905: ρ = 0.6640***
  - D3 Rescaled vs A905: ρ = 0.6368***
  - D Overall vs A905: ρ = 0.7490***
  - Territorial-coverage indicator is quadratically positive and significantly correlated with all three doorsteps.
- Doorsteps and Income Inequality (Adjusted Gini):
  - D1 Rescaled vs Adjusted Gini: ρ = -0.4410***
  - D2 Rescaled vs Adjusted Gini: ρ = -0.3876***
  - D3 Rescaled vs Adjusted Gini: ρ = -0.3821***
  - D Overall vs Adjusted Gini: ρ = -0.4408***
  - Form: negative quadratic (Kuznets-like): inequality increases in initial transition, then decreases at accelerated rate; correlation strongest for D3.
- Doorsteps and Inflation:
  - D1 Rescaled vs CPI Inflation: ρ = -0.3973***
  - D2 Rescaled vs CPI Inflation: ρ = -0.3567***
  - D3 Rescaled vs CPI Inflation: ρ = -0.3070***
  - D Overall vs CPI Inflation: ρ = -0.3944***
  - Shape: negative quadratic; inflationary risk may rise during initial transition then fall after threshold.
- Doorsteps and GDP per Capita:
  - D1 Rescaled vs GDPpc: ρ = 0.7924***
  - D2 Rescaled vs GDPpc: ρ = 0.7058***
  - D3 Rescaled vs GDPpc: ρ = 0.6253***
  - D Overall vs GDPpc: ρ = 0.7773***
  - Note: strong positive quadratic correlation; correlation coefficient for squared value of D1 and GDP per capita is over 0.8.
- Doorsteps and Private Sector Credit:
  - D1 Rescaled vs Credit: ρ = 0.6876***
  - D2 Rescaled vs Credit: ρ = 0.5972***
  - D3 Rescaled vs Credit: ρ = 0.5221***
  - D Overall vs Credit: ρ = 0.6723***
  - Pattern: private sector credit to GDP increases strongly and nonlinearly with doorstep advancement, though good doorway performance does not guarantee high private credit.

### Econometric strategy: Identification through Heteroskedasticity (IH)
- Identification approach: IH (Rigobon and Rodrik (2005)) exploiting different variances of structural shocks across subsamples rather than external instruments.
- Sample split used primarily: colonized vs non-colonized countries (s1: 36; s2: 57).
- Variables analyzed:
  - First model: D1, D2, D3, democracy, GDP per capita.
  - Second model: adds income inequality (Gini).
- Means and variances across colony split (selected):
  - mean s1: democracy -0.55; Gini 0.36; GDPpc -0.55; D1 -0.63; D2 -0.59; D3 -0.12; nat'l. 1.94; res 0.02
  - variance s1: democracy 0.18; Gini 0.18; GDPpc 0.16; D1 0.08; D2 0.14; D3 0.39; natl. 0.34; res 1.42
  - mean s2: democracy 0.33; Gini -0.21; GDPpc 0.32; D1 0.35; D2 0.34; D3 0.08; nat'l 7.75; res -0.02
  - variance s2: democracy 0.18; Gini 1.11; GDPpc 0.16; D1 0.15; D2 0.20; D3 0.50; nat'l. 1.33; res 0.72
  - F-test p-values: democracy 0.46; Gini 4.59E-10*; GDPpc 0.47; D1 0.03*; D2 0.13; D3 0.21; natl 2.29E-06*; res 0.01*
  - Interpretation: in four out of eight cases variances differ significantly across subsamples, sufficient for heteroskedasticity-based identification.
- GEO split alternative (East-West vs North-South) broadly confirms main results; 5x5 model equality test F-test p-value 0.98 (cannot reject equality). 6x6 GEO model introduces some uncertainty on doorsteps’ impacts on Gini.

### Key IH econometric findings (selected, from 6x6 IH model with colony split, Table 2)
- Doorsteps → democracy and GDP per capita:
  - D1 and D2 have positive and significant impacts on democracy and GDP per capita.
  - D3 impact on democracy and GDP per capita is insignificant.
  - Interpretation: D1 (rule of law among elites) and D2 (perpetual organizations) are more decisive for transition to democracy and higher GDP per capita than D3.
- Feedbacks (democracy and GDPpc → doorsteps):
  - Strong positive feedback from democracy on all three doorsteps.
  - GDP per capita has strong impact on D1 and D2, but not on D3.
  - Implication: D1 and D2 can initiate transitions; democratization and higher GDP per capita reinforce doorsteps thereafter.
- Inter-doorstep dynamics:
  - D1 and D2 significantly affect other doorsteps; D3 impact on others is insignificant.
  - Strong feedback between D1 and D2 supports NWW thesis of mutual reinforcement.
- Interactions with income inequality (selected IH coefficients and t-values shown in source):
  - Democracy equation: democracy on itself 0.424* (2.177); Gini 0.211* (3.665); GDPpc 0.332* (10.89); D1 0.284* (3.669); D2 0.054 (0.562); D3 -0.178 (0.886); res 0.175* (2.638).
  - Gini equation: Gini on itself -0.138* (6.159); GDPpc -0.041* (2.439); D1 -0.036* (3.412); D2 0.107* (4.433); D3 -0.147* (5.219); area -0.493* (10.01).
  - D1 equation: democracy 0.502* (4.268); GDPpc 0.507* (5.258); D1 0.355* (2.456); area -0.024* (2.373).
  - D3 equation: democracy 0.514* (7.994); Gini 0.849* (5.078); D1 0.072* (2.394); D2 0.37* (5.15).
- Top-left quadrant (democracy, GDPpc, inequality):
  - Greater income inequality leads to more democracy (significant and positive) and to lower GDP per capita (significant at the 10 percent level).
  - Democracy reduces income inequality; higher GDP per capita reduces inequality.
  - Interpretation: inequality can act as a “push factor” for social transitions.

### Interactions of income inequality with doorsteps
- Progress along D1 and D3 leads to reduction in income inequality.
- Progress in D3 can reduce income inequality even when it has no impact on democracy and GDP per capita (e.g., transition from military dictatorship to civil rule; disarmament/unification of armed factions).
- Progress toward more perpetual organization (D2) appears to lead to more income inequality (noted as a “mystery” deserving future analysis).
- Income inequality has a statistically significant impact only on D3 in the model (positive sign: more inequality pushes for improvements in D3).

### Joint dynamics and policy-relevant implications
- Public services and expenditure:
  - Proportional increase in provision of public services is far bigger than proportional increase in government expenditure as countries advance the doorsteps.
  - Moving toward open access increases both the level and territorial coverage of public goods; economies-of-scale may explain nonlinear increase in provision relative to linear increase in expenditure.
  - Finding: an increase in institutional advancement is associated with an increase in the ratio of government expenditure on education to GNI by 40 percent (text statement).
- GDP per capita, inequality, and democracy thresholds:
  - Income inequality decreases only well beyond average overall index levels.
  - GDP per capita accelerates after threshold approximately D_Overall = 0.8.
  - First moves toward democracy occur before reductions in income inequality; reductions in inequality coincide with countries having gone about half way through doorstep conditions and entering “flawed democracies” (0.6 on the democracy scale); genuine reductions occur in final stages and around full democracy (around 0.8 marker).
- Four synthesized empirical findings:
  - (i) Progress in D1 and D2 contributes to improvements in political environment and GDP per capita.
  - (ii) Once doorstep progress starts, strong feedback mechanisms from democracy and GDP per capita exist.
  - (iii) Improvements in D1 and D2 reinforce each other and lead to progress in D3; D3 has negligible effects on other doorsteps.
  - (iv) Higher income inequality pushes directly for more democracy and induces upward changes in D3.
- Policy-relevant implications:
  - D1 and D2 are main drivers for transitions to democracy and higher GDP per capita; D3 mainly reduces income inequality and tends to follow other changes.
  - Rising income inequality can act as a pressure for democratic reforms and institutional change.
  - Contemporary transitions differ from historical NWW examples due to foresight, urbanization, transparency, information technology, and globalization.

### Appendix tables and methodological references (selected)
- Table inventory includes: Table A–1 Description of Databases; Table A–2 Components of the First Doorstep; Table A–3 Components of the Second Doorstep; Table A–4 Components of the Third Doorstep; Table A–5 Country Doorsteps Scores; Tables A–6 to A–8 Spearman Rank Correlations.
- Spearman correlations within doorway categories reported (selected):
  - D1 vs D1-Category 1: 0.84; D1 vs D1-Category 2: 0.87; D1 vs D1-Category 3: 0.90 (all p-value 0.00).
  - D2 vs D2-Category 1: 0.73; D2 vs D2-Category 2: 0.89 (p-value 0.00).
  - D3 vs D3-Category 1: 0.81; D3 vs D3-Category 2: 0.91; D3 vs D3-Category 3: 0.89 (p-value (0.00)).
- Methodological literature cited includes Rigobon (2003) and Rigobon and Rodrik (2005) on IH, and literature on institutions, democracy, inequality and growth (selected references listed in source).

*Source: _wp1287 - Appendix Tables*

### Appendix Tables

### Appendix Tables

### Overview and table inventory
- Table A–1. Description of Databases (page 52)
- Table A–2. Components of the First Doorstep (page 52)
- Table A–3. Components of the Second Doorstep (page 53)
- Table A–4. Components of the Third Doorstep (page 53)
- Table A–5. Country Doorsteps Scores (page 54)
- Table A–6. Spearman Rank Correlations (page 56)
- Table A–7. Spearman Rank Correlations: D_Overall, D1, D2, and D3 (page 57)
- Table A–8. Spearman Rank Correlations: D_Overall and all Categories (page 57)
- References section appears (page 37 referenced in listing)

### Conceptual framework and goals
- The work operationalizes North, Wallis and Weingast (NWW, 2009) transition framework distinguishing two social orders: limited access (natural state) and open access.
- NWW’s three doorstep conditions are central:
  - Doorstep 1: rule of law for the elites.
  - Doorstep 2: existence of perpetually lived organizations (including the state).
  - Doorstep 3: consolidated political control of all military power.
- The paper’s aims include:
  - Constructing indices representing the three doorsteps by combining existing institutional variables.
  - Testing relationships between countries’ transition paths and economic and political performance using graphical analysis, OLS and nonlinear regressions.
  - Using Identification through Heteroskedasticity (IH) to empirically test interactions between doorsteps and economic/political variables.

### Data sources and composition of indices
- Sample selection: low-, middle-, and high-income countries for which consistent data from the same databases are available across selected criteria.
- Indicator components:
  - Total of 31 selected criteria across the three doorsteps.
  - 17 out of 31 criteria drawn from the Institutional Profiles database (de Crombrugghe et al. (2009)).
  - The remaining 14 criteria taken from Polity IV, the World Bank’s IPD, and Economic Freedom of the World databases.
- Doorstep-specific components (described in Appendix Tables A2–A4):
  - Doorstep 1 (D1) components: existence of rules for elites; independence and impartiality of the judicial system; elites’ respect for and compliance with court decisions.
  - Doorstep 2 (D2) components: (i) stability and permanence of the political system as a proxy for state perpetuity; (ii) ease of creating independent organizations (legal role and societal attitude toward organizations).
  - Doorstep 3 (D3) components: (i) political control of the legal armed forces; (ii) military interference in political life; (iii) level of armed violence in society as a de facto measure of political control.

### Aggregation methodology
- Variable standardization:
  - Individual variables within each category are standardized on a 0 (worst) to 1 (best) scale.
- Category and sub-index aggregation:
  - Each category = simple average of standardized variables contained in that category.
  - Each doorstep sub-index = simple average of its categories, then rescaled by the sample average.
  - Example: D1 (rule of law for the elites) = average of its three categories divided by 0.62 (the sample average for the constructed D1 categories).
- Overall index:
  - The overall doorsteps index = simple average of the three sub-indices (range stated as between 0 and 1 in the method description).

### Key empirical and contextual statements (as presented)
- Historical context highlighted by NWW:
  - Limited access (natural state) has existed for the past 10,000 years.
  - Open access societies have emerged over the last 300 years.
- The indices are intended to capture the specific NWW doorstep criteria (rule of law for elites; perpetually lived organizations; political control of the military) rather than more general measures of rule of law or governance.
- Interdependencies:
  - The three doorsteps interact and reinforce one another; meeting doorstep 1 typically precedes the others but is not a strict prerequisite.
- Prior literature connections and motivations:
  - The appendix situates the index construction within related empirical literature on institutions, growth, democracy, and inequality and references prior datasets used (e.g., de Crombrugghe et al. (2009), Gollwitzer and Quintyn (2010)).

*Source: _wp1287 - Appendix Tables*

### 1. Our database consists of 108 low-, middle-, and high-income countries (Appendix Table

### 1. Our database consists of 108 low-, middle-, and high-income countries (Appendix Table

### B. Country Doorsteps Scores
- Database: 108 countries.
- Overall index (D_Overall) patterns:
  - Five countries consistently among top ten performers: Denmark, Finland, Australia, the Netherlands, and New Zealand.
  - Two countries consistently among worst ten across all four indicators: Zimbabwe and Mauritania.
- Average D_Overall by income group:
  - High income countries (HIC): 1.27
  - Middle income countries (MIC): 0.91
  - Low income countries (LIC): 0.80
- Variance and MIC:
  - Variance is greatest for MIC because low- and upper-MIC are grouped together.
  - All top ten MIC performers are upper-MIC; the five worst MIC scores are lower-MIC.
- Robustness check:
  - Aggregating the eight categories into an overall indicator using weights from Principal Component Analysis did not affect the relative ranks of countries in any significant way.
- Relation to external index:
  - Scores for D1 are very similar to the Bertelsman Transformation Index (political transformation) category; BTI explores the state of development achieved by 128 countries on their way to democracy and market economy safeguards.

### C. Correlations within and between the Doorsteps
- Internal drivers:
  - None of the three sub-indices (D1, D2, D3) is driven by an individual category.
  - The first doorstep (D1) is most strongly correlated with the overall index (see Appendix Table A7), yet no single sub-index dominates.
  - The three sub-indices are strongly and significantly correlated with each other.
- Consistency bias note:
  - High correlations may partly reflect a “consistency bias” from using different variables drawn from the same databases.
- Individual categories most correlated with D_Overall (in decreasing magnitude):
  - Military interference in political life.
  - Elites’ respect for the courts’ decisions.
  - Independence and impartiality of the judicial system.
- Weakest correlation with D_Overall:
  - First category of the second doorstep: stability and permanence of the political system.

### IV. DOORSTEPS AND PERFORMANCE – A GRAPHICAL ANALYSIS
- Objective:
  - Jointly consider country performance under doorstep indicators and selected political and economic “symptom” variables from the NWW transition framework.
- Method:
  - Run correlations between D1, D2, D3 and D_Overall and symptom variables; depict correlations in scatterplots with correlation coefficient ρ and significance stars (*** 1%, ** 5%, * 10%).
- Symptoms analyzed:
  - Political: democracy, equal access, corruption.
  - Economic: size of government, social welfare, income inequality, macroeconomic stability, bank credit to the private sector.
- Variable scaling:
  - Whenever appropriate, variables are standardized to range between 0 (worst) and 1 (best).

### A. Political Symptoms
- Democracy (EIU Overall Score of Democracy Index):
  - EIU overall score composed of five categories: the electoral process, functioning of the government, political participation, political culture, and civil liberties.
  - EIU regime categories and score ranges:
    - Authoritarian regimes: 0 to 3.99
    - Hybrid regimes: 4–5.99
    - Flawed democracies: 6–7.99
    - Full democracies: 8–10
  - Correlations:
    - Strong, positive, and significant correlation between EIU democracy score and all doorstep indicators.
    - Correlation weakest for D3 (political control of the military).
  - Key graphical inferences:
    - Only countries more than three quarters up the ladder of doorstep one are fully democratic (rating above .8 in the study’s charts).
    - D2 concentration: strongest concentration of countries in the 0.6-0.8 democracy range.
    - D3: high concentration of full democracies at the extreme end of doorstep achievement; average D3 associated with flawed and hybrid democracies.
  - Kernel density findings (D_Overall vs democracy):
    - Three clusters: (1) high doorsteps & full democracy; (2) below-average doorsteps & very low democracy; (3) below-average to average doorsteps & medium democracy (hybrid/flawed democracies).
    - Flat region in upper-left indicates it is very unlikely for a country with low D_Overall to be a perfect democracy.

- Equal Access:
  - Constructed as average of four IPD variables: equal access to schooling, equal access to public health care, equal treatment in administrative procedures, equal access to public employment (A9040–A9043).
  - Correlations:
    - Significant and positive correlation between doorsteps and equal access.
    - For D3, the correlation is quadratic.
    - Strongest correlation for the average of the three doorsteps (D_Overall), suggesting all three doorstep conditions are needed to provide equal access to institutions.
  - Note on scaling:
    - Original variables have four possible scores (1 to 4); averages are not normally distributed.

- Corruption (Transparency International Corruption Perception Index, CPI):
  - Relationship with doorsteps is quadratic; correlations reported also with squared doorstep scores (ρ_{D^2}).
  - Correlation strongest for D1.
  - Finding on returns:
    - Increasing returns in reducing corruption as doorstep conditions improve.
    - It takes about three-quarters of the doorstep condition before corruption levels reach the 50 percent mark on the CPI scale.
  - Kernel density (D_Overall vs CPI):
    - Two clusters: (1) below-average doorsteps with high corruption (low CPI); (2) higher D_Overall with low corruption (higher CPI).
  - Corruption versus Democracy:
    - As countries develop, initial progress tends to be greater in democracy than in reducing corruption.
    - Countries well-advanced from fragile to mature natural state show medium democracy but very poor corruption scores.
    - Only once countries pass the average score of D_Overall does corruption wane faster while democratic reforms continue.

### B. Economic Symptoms
- Size of Government:
  - Hypothesis: transition from limited to open access orders is associated with growth of government and increased public provision.
  - Proxies used: total government expenditure as ratio to GDP, education expenditure as percentage of GNI (WEO variables).
  - Findings:
    - Positive and significant correlation between doorsteps and government expenditure to GDP ratio.
    - Kernel density confirms governments with lower D_Overall tend to spend less than those with higher D_Overall.
    - Significant quadratic correlation between doorstep indicators and education expenditure to GNI; correlation strongest for D2, indicating increasing returns in education provision as doorstep conditions improve.
  - Illustrative scenario (text cut off in source):
    - Source begins: "Assume a country starts off with an average performance (1.00) under the first doorstep and achieves an improvement of" — the sentence is incomplete in the supplied content.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1287.pdf*

### 0.2 points. This will lead to an increase in the ratio of government expenditure on education

### _wp1287 - 0.2 points. This will lead to an increase in the ratio of government expenditure on education

### Social welfare and territorial coverage of public services
- Finding: An increase in institutional advancement is associated with an increase in the ratio of government expenditure on education to GNI by 40 percent.
- Interpretation: Governments of institutionally advanced countries tend to spend more on public goods which enhance equality across society.
- Measurement: Social welfare is captured by the territorial coverage of public services (Institutional Profiles database), comprising six categories: public schooling, basic healthcare, potable water and sewage purification, electricity, road network, and solid waste.
- Relationship with doorsteps:
  - The territorial-coverage indicator is quadratically positively and significantly correlated with all three individual doorsteps.
  - As countries progress along the doorsteps, provision of schooling, healthcare, clean water, electricity, road networks, and waste management becomes ever increasing and more equally spread.
  - In limited access societies, patron-client relations limit supply of these services to elite bases; advancement replaces this with impersonal, nationwide provision.

### Income inequality
- Measurement: Income distribution measured with an adjusted Gini coefficient (Gini coefficients adjusted for type of income, survey coverage and continent dummies).
- Finding: The correlation between doorstep development and income inequality is nonlinear (quadratic), negative and significant.
- Specifics:
  - Correlation strongest for the third doorstep: better political control of the military is associated with reduction in income inequality.
  - The correlation with the first and second doorstep has the same negative quadratic form but is less pronounced.
  - The curve displays a Kuznets-curve: inequality tends to increase in initial transition stages, then decreases at an accelerated rate.
- Distributional pattern: Kernel density plot (Figure 19) distinguishes two groups: a large group with low to average doorstep scores and relatively high income inequality, and a smaller group with near-perfect doorstep scores and low income inequality.
- Conclusion: Findings confirm the claim that transition from limited to open access order is associated with reduction in income inequality.

### Macroeconomic stability (inflation and GDP per capita)
- Measurement: Macroeconomic stability measured by CPI inflation and GDP per capita.
- Inflation:
  - Correlation between inflation and all doorstep indicators is negative and quadratic.
  - Shape suggests inflationary risk may increase during initial transition from fragile/basic to mature natural states, then decrease progressively after a threshold of institutional development is reached.
  - Correlation with the second doorstep is weakest.
- GDP per capita:
  - Figure 21 shows a strong positive quadratic correlation of GDP per capita with the doorsteps.
  - Correlation coefficient for squared value of the first doorstep and GDP per capita is over 0.8.
  - Correlations with the other two doorsteps are slightly weaker but still highly significant and above 0.7.
  - Interpretation: As countries develop rules for elites, average GDP per capita grows at an increasing rate.

### Financial development: bank credit to the private sector
- Expectation: Open access orders facilitate private sector access to credit due to impersonal exchange and enforceable impartial rules.
- Finding: Ratio of private sector credit to GDP exhibits a strongly positive and significant quadratic correlation with the doorstep conditions, especially with the average doorstep indicator.
- Pattern: As a country advances along doorstep criteria, private sector access to commercial bank credit increases exponentially.
- Caveat: Good doorway performance is not a guarantee of high private sector credit—some well-performing-doorstep countries still have limited private credit provision.

### Joint dynamics: public services, government expenditure, GDP per capita, income inequality, and democracy
- Government expenditure vs territorial coverage:
  - Proportional increase in provision of public services is far bigger than proportional increase in government expenditure as a country advances along the doorsteps.
  - Implication: Moving toward open access increases not only the level but also the targeting/restructuring of government expenditure toward social welfare (public services).
  - Economies-of-scale may partially explain nonlinear increase in provision relative to linear increase in expenditure.
- GDP per capita and income inequality:
  - Income inequality only starts to decrease well beyond average overall index levels.
  - GDP per capita accelerates after threshold approximately D_Overall = 0.8.
  - Interpretation: (i) wealth must be sufficiently large before elites/governments are willing to share it (Kuznets-effect); (ii) this coincides with transition from mature natural state into open access society.
- Income inequality and democracy:
  - Reduction in income inequalities is preceded by economic development and political reforms.
  - First moves toward democratic institutions occur long before income inequalities are reduced.
  - First signs of reductions in income inequality coincide with countries having gone about half way through doorstep conditions and entering “flawed democracies” (0.6 on the democracy scale).
  - Genuine moves toward reduction in income inequalities occur only in final stages of transition and when countries transit into full democracy (around the 0.8 marker).

### Econometric strategy and sample
- Endogeneity challenge: Doorstep conditions both affect and are affected by politico-economic variables.
- Identification approach: Identification through Heteroskedasticity (IH) following Rigobon and Rodrik (2005).
  - IH exploits plausible differences in variances of structural shocks across sub-samples rather than requiring an external instrument.
- Sample split: Countries split into those colonized by European powers and others.
  - Final sample includes 93 countries with data for 2009: colonized = 57 countries; non-colonized = 36 countries.
- Variables analyzed in first model: three doorsteps (D1, D2, D3), democracy, GDP per capita.
- Variables added in second model: income inequality.

### Key econometric findings (IH estimation, Table 1)
- Top-right quadrant (doorsteps → democracy and GDP per capita):
  - Doorsteps 1 and 2 have a positive and significant impact on both democracy and GDP per capita.
  - Doorstep 3 impact on democracy and GDP per capita is insignificant.
  - Interpretation: Establishing rule of law among elites and establishing perpetual organizations (D1 and D2) are more decisive for transition to democracy and higher GDP per capita than bringing military under political control (D3).
- Bottom-left quadrant (feedback from democracy and GDP per capita to doorsteps):
  - Strong positive feedback from democracy on all three doorsteps.
  - Strong impact of GDP per capita on doorsteps 1 and 2, but not on doorsteps 3.
  - Interpretation: Doorsteps 1 and 2 can initiate transition toward democracy; democratization further improves D1, D2, and D3.
  - Political unification of military (D3) appears more an outcome of democratization than a driver of it.
- Bottom-right quadrant (interactions among doorsteps):
  - D1 and D2 have significant impacts on other doorsteps; D3 impact on others is insignificant.
  - Strong feedback between D1 and D2 supports NWW thesis that elites accepting rules among themselves promotes creation of perpetual political organizations and broader obedience of rule of law.
- Economic–political interactions (top-left quadrant):
  - GDP per capita and democracy have positive and significant impacts on each other.
  - Consistent with literature (Lipset (1959); Barro (1999)) that higher GDP per capita increases propensity for democracy; also consistent with studies finding democracy affects income levels indirectly.

### Robustness and interpretation notes from authors
- The exogenous variables included in the IH models are insignificant in most cases.
- The authors acknowledge alternative channels where higher GDP per capita preceded doorstep improvements in some countries (Meisel and Ould Aoudia, 2008), implying government coordination can offset low D1-type rule of law in certain development trajectories.
- The authors maintain that initial improvements in D1 are necessary for better economic performance despite such counterexamples.

*Italic: Source — Excerpt from the IMF working paper content provided in the supplied PDF chapter/section.*

### part in this model concerns the interactions with income inequality. The results are somewhat

### _wp1287 - part in this model concerns the interactions with income inequality. The results are somewhat

### Interactions of income inequality with doorsteps, democracy, and GDP per capita
- Progress along D1 and D3 leads to a reduction in income inequality.
- Progress in D3 can reduce income inequality even when it has no impact on democracy and GDP per capita. Progress along D3 may reflect transitions such as:
  - transition from military dictatorship to civil rule; and/or
  - disarmament or unification of armed factions associated with elites.
- Progress toward more perpetual organization (D2) appears to lead to more income inequality — an unexpected result noted as a “mystery” that deserves future analysis.
- Feedback mechanisms from democracy and GDP per capita to the doorsteps are confirmed. Income inequality has a statistically significant impact only on D3, with the sign positive (i.e., more inequality pushes for improvements in D3).

### Empirical estimates and statistical details (selected)
- Reported coefficients and significance (t-values in brackets; asterisks indicate significance level of 95%) from the 6x6 IH model with colony split (Table 2) include:
  - democracy equation: democracy coefficient on itself 0.424* (2.177); Gini 0.211* (3.665); GDPpc 0.332* (10.89); D1 0.284* (3.669); D2 0.054 (0.562); D3 -0.178 (0.886); nat'l -0.039 (0.275); res 0.175* (2.638).
  - Gini equation: Gini on itself -0.138* (6.159); GDPpc -0.041* (2.439); D1 -0.036* (3.412); D2 0.107* (4.433); D3 -0.147* (5.219); nat'l -0.108* (1.877); res 0.029 (0.767); area -0.493* (10.01).
  - GDPpc equation: democracy 0.188* (2.255); Gini -0.336* (1.629); D1 0.327* (9.023); D2 0.403* (4.895); D3 -0.052 (0.476); nat'l 0.359 (1.569); res 0.116 (0.787); area 0.285* (3.876).
  - D1 equation: democracy 0.502* (4.268); Gini -0.002 (0.006); GDPpc 0.507* (5.258); D1 0.355* (2.456); D2 0.198 (1.12); D3 0.047 (0.129); nat'l -0.091 (0.367); area -0.024* (2.373).
  - D2 equation: democracy 0.239* (4.015); Gini -0.025 (0.151); GDPpc 0.275* (6.135); D1 0.208* (7.335); D2 -0.119 (1.43); D3 -0.296* (1.785); nat'l -0.002 (0.025); area -0.839* (13.717).
  - D3 equation: democracy 0.514* (7.994); Gini 0.849* (5.078); GDPpc 0.009 (0.174); D1 0.072* (2.394); D2 0.37* (5.15); D3 0.017 (0.101); nat'l 0.081 (0.709); area -0.049 (0.534).

### Top-left quadrant results: democracy, GDP per capita, and inequality
- Greater income inequality leads to:
  - more democracy (significant and positive effect); and
  - lower GDP per capita (significant at the 10 percent level).
- Democracy has a negative and significant impact on income inequality (i.e., more democracy leads to less income inequality).
- Higher levels of GDP per capita also lead significantly to lower inequality.
- The positive effect of inequality on democracy is interpreted as inequality acting as a “push factor” for social transitions (consistent with Acemoglu and Robinson (2000) and Zak and Feng (2003), as discussed in the source).

### Robustness checks: colony split and GEO split
- Identification using the colony split requires different relative variances of structural shocks across subsamples (s1: 36 countries; s2: 57 countries). Table 3 reported means and variances:
  - mean s1: democracy -0.55; Gini 0.36; GDPpc -0.55; D1 -0.63; D2 -0.59; D3 -0.12; natl. 1.94; res 0.02
  - variance s1: democracy 0.18; Gini 0.18; GDPpc 0.16; D1 0.08; D2 0.14; D3 0.39; natl. 0.34; res 1.42
  - mean s2: democracy 0.33; Gini -0.21; GDPpc 0.32; D1 0.35; D2 0.34; D3 0.08; natl. 7.75; res -0.02
  - variance s2: democracy 0.18; Gini 1.11; GDPpc 0.16; D1 0.15; D2 0.20; D3 0.50; natl. 1.33; res 0.72
  - F-test p-values: democracy 0.46; Gini 4.59E-10*; GDPpc 0.47; D1 0.03*; D2 0.13; D3 0.21; natl. 2.29E-06*; res 0.01*
  - Interpretation: in four out of eight cases the differences in variance are significantly different from zero across the subsamples, sufficient for heteroskedasticity-based identification.
- Alternative split (GEO split: East-West versus North-South continents) broadly confirms prior results:
  - 5x5 model under GEO: signs of coefficients remain the same; size of most coefficients remains in the same range. F-test p-value 0.98 for equality of coefficient estimates across the two 5x5 models (cannot reject equality).
  - 6x6 model under GEO (Table 5) mostly confirms prior findings; one notable change: impact of D1 on Gini becomes positive and significant while impact of D2 on Gini becomes insignificant, introducing uncertainty about true impacts of the doorsteps on income inequality.

### Synthesis of empirical section: four key findings (verbatim structure)
- (i) progress with respect to doorstep conditions one and two (rule of law for the elites, establishment of perpetual institutions) contributes to improvements in the political environment and in economic performance, the latter measured here by GDP per capita;
- (ii) once progress in meeting the doorstep conditions is set in motion, we experience strong feedback mechanisms from both democracy and GDP per capita;
- (iii) we find strong indications that, in line with the NWW theory, improvements in the rule of law for elites and the development of perpetual organizations strongly reinforce each other, while both also lead to progress in D3. The latter seems to have negligible effects on both other doorsteps; and finally
- (iv) higher income inequality pushes directly for more democracy and induces upward changes in D3.

### Conclusions and policy-relevant implications
- The NWW transition framework is generally confirmed: the three doorstep conditions (D1: acceptance of the rule of law for elites; D2: move toward perpetual organizations; D3: consolidated political control of the military) are critical in politico-economic development.
- D1 and D2 are the main driving forces behind shifts toward more democracy and higher GDP per capita; D3 has almost negligible direct impact on polity and economic performance but plays a role in reducing income inequality.
- Strong feedback loops exist from democracy and GDP per capita back to the doorsteps, creating potential virtuous cycles once institutional change begins.
- D1 and D2 strongly interact and reinforce each other; they also promote progress in D3, but there is limited evidence of D3 feeding back to D1 and D2 — suggesting that political unification of armed forces tends to follow, rather than lead, other institutional and economic changes.
- Income inequality plays a crucial transitional role: greater inequality appears to drive moves toward more democracy (mostly directly), implying that growing inequality can serve as a force that pressures elites toward transition.
- Contemporary transitions may differ from historical NWW examples because societies today have better foresight about desired institutional outcomes and external forces (urbanization, transparency, information technology, globalization) play larger roles; rising income inequality is one such external/internal force that can precipitate democratic reforms.

*Source: IMF working paper content provided in the supplied PDF excerpt.*

### REFERENCES

### _wp1287 - REFERENCES

### Institutions and Economic Growth (selected references)
- Acemoglu, Daron and Simon Johnson, 2005, “Unbundling Institutions,” Journal of Political Economy, Vol. 113, No. 5.
- Acemoglu, Daron and James Robinson, 2000, “Why did the West Extend the Franchise? Democracy, Inequality and Growth in Historical Perspective,” Quarterly Journal of Economics, Vol. 115, No. 4 (November), pp. 1167–1199.
- Acemoglu, Daron, S. Johnson, and J. Robinson, 2004, “Institutions as the Fundamental Cause of Long-Run Growth,” NBER Working Paper No. 10481, May.
- Acemoglu, Daron, S. Johnson, J. Robinson, and P. Yared, 2007, “Reevaluating the Modernization Hypothesis,” NBER Working Papers No. 13334, August.
- Clague, Christopher, ed. Institutions and Economic Development: Growth and Governance in Less-Developed and Post-Socialist Countries (Baltimore and London: The Johns Hopkins University Press).
- Clague, Christopher, P. Keefer, S. Knack, and M. Olson, 1996a, “Institutions and Economic Performance: Property Rights and Contract Enforcement,” pp. 67–90 in Institutions and Economic Development: Growth and Governance in Less-Developed and Post-Socialist Countries, ed. by Clague, C. (Baltimore and London: The Johns Hopkins University Press).
- Clague, Christopher, P. Keefer, S. Knack, and M. Olson, 1996b, “Democracy, Autocracy, and the Institutions Supportive of Economic Growth,” pp. 91–120, in Institutions and Economic Development: Growth and Governance, Less-Developed and Post-Socialist Countries, ed. by Clague, C. (Baltimore and London: The Johns Hopkins University Press).
- Knack, Stephen and Philip Keefer, 1995, “Institutions and Economic Performance: Cross Country Tests using Alternative Measures,” Economics and Politics, Vol. 7, pp. 207–227.
- Rodrik, Dan, A. Subramanian, and F. Trebbi, 2004, “Institutions Rule,” Journal of Economic Growth, Vol. 9, pp. 131–165.
- Glaesser, E., R. La Porta, F. Lopez-de-Silanes, and A. Schleifer, 2004, “Do Institutions Cause Growth?” NBER Working Papers No. 10568.

### Democracy, Political Regimes, and Redistribution (selected references)
- Alesina, Alberto and Dani Rodrik, 1994, “Distributive Politics and Economic Growth,” Quarterly Journal of Economics, Vol. 109, pp. 465–490.
- Alesina, Alberto and Roberto Perotti, 1996, “Income Distribution, Political Instability and Investment,” European Economic Review, Vol. 40, pp. 1203–1228.
- Alesina, Alberto, Sule Ozler, Nouriel Roubinio, and Philip Swagel, 1996, “Political Instability and Economic Growth,” Journal of Economic Growth, Vol. 1, No. 2, pp. 189–211.
- Barro, Robert, 1996, “Democracy and Growth,” Journal of Economic Growth, Vol. 1, No. 1, March.
- Barro, Robert, 1999, “Determinants of Democracy,” Journal of Political Economy, Vol. 107, No. 6, pp. 158–183.
- Boix, Carles and Susan Stokes, 2003, “Endogenous Democratization,” World Politics, Vol. 55, No. 4 (July), pp. 517–549.
- Przeworski, Adam, 2005, “Why do Political Parties Obey Results of Elections?” pp. 114–145 in, 2005, Democracy and the Rule of Law, ed. by Maravall, Jose Maria and Adam Przeworski (Cambridge University Press).
- Przeworski, Adam, M. Alvarez, J. Cheibub, and F. Limongi, 2000, Democracy and Development: Political Institutions and Well-Being in the World, 1950-1990 (Cambridge University Press), 321 pp.
- Przeworski, Adam and Fernando Limongi, 1993, “Political Regimes and Economic Growth,” Journal of Economic Perspectives, Vol. 7, No. 3 (summer), pp. 51–69.
- Lipset, Symour, 1959, “Some Social Requisites of Democracy: Economic Development and Political Legitimacy,” American Political Science Review, Vol. 53 (March), pp. 69–105.

### Inequality, Redistribution, and Social Outcomes (selected references)
- Deininger, Klaus and Lyn Squire, 1997, “Economic Growth and Income Inequality: Reexamining the Links,” Finance & Development, Vol. 34, No. 1 (March), pp. 38–41.
- Persson, Thorsten and Guido Tabellini, 1994, “Is Inequality Harmful for Growth? Theory and Evidence,” American Economic Review, Vol. 84, pp. 600–621.
- Perotti, Roberto, 1996, “Growth, Income Distribution, and Democracy: What the Data Say,” Journal of Economic Growth, Vol. 1, pp. 149–187.
- Ravaillion, Martin, 2009, “The Developing World’s Bulging (but Vulnerable) Middle Class,” World Development, Vol. 38, No. 4, pp. 445–454.
- Okun, Arthur, 1975, Equality and Efficiency: The Big Trade-Off (Washington: Brookings Institution Press).
- Muller, Edward, 1995, “Income Inequality and Democratization: Reply to Bollen and Jackman,” American Sociological Review, Vol. 60, No. 6 (December), pp. 990–996.
- De Crombrugghe, Denis, Kristine Farla, Nicholas Meisel, Chris de Neubourg, Jacques Ould Aoudia and Adam Szirmai, 2009, “Institutional Profiles Database III. Presentation of the Institutional Profiles Database 2009. http://www.cepii.fr/anglaisgraph/bdd/form_instit/documents.asp

### Governance, Corruption, and Public Finance (selected references)
- Berg, Andy, J. Ostry, and J. Zettelmeyer, 2008, “What Makes Growth Sustained?” IMF Working Paper 08/59 (Washington: International Monetary Fund).
- Berg, Andy and Jonathan Ostry, 2011, “Equality and Efficiency”, Finance & Development, Vol. 48, No. 3 (September), pp. 12–15.
- Gollwitzer, Sophia and Marc Quintyn, 2010, “The Effectiveness of Macroeconomic Commitment in Weak(er) Institutional Environments,” IMF Working Paper 10/193 (Washington: International Monetary Fund).
- Quintyn, Marc and Genevieve Verdier, 2010, “Mother, Can I Trust the Government? Sustained Financial Deepening—A Political Institutions View,” IMF Working Paper 10/210 (Washington: International Monetary Fund).
- Kaufman, Dany and A. Kraay, 2002, “Growth without Governance,” Economia, Vol. 3, No. 1. pp. 169–215.
- Haber, Stephen, 2008, “Political Institutions and Financial Development: Evidence from the Political Economy of Bank Regulation in Mexico and the United States,” pp. 10–59 in Haber, Stephen, Douglass North, and Barry Weingast, 2008, Political Institutions and Financial Development, 300 pp (Stanford, California: Stanford University Press).
- Zak, Paul, 2001, “Institutions, Property Rights, and Growth,” the Gruter Institute Working Paper on Law, Economics and Evolutionary Biology, Vol. 2, Issue 1.

### Methodology and Empirical Techniques (selected references)
- Rigobon, Roberto, 2003, “Identification through Heteroskedasticity,” The Review of Economic and Statistics, Vol. 85, No. 4 (November), pp. 777–792.
- Rigobon, Roberto and Dani Rodrik, 2005, “Rule of Law, Democracy, Openness and Income,” Economics in Transition, Vol. 13, No. 3, pp. 533–564.
- Mulligan, Casey, R. Gil, and X. Sala-i-Martin, 2004, “Do Democracies Have Different Public Policies than Nondemocracies?” Journal of Economic Perspectives, Vol. 18, No. 1 (winter), pp. 51–74.
- Rodrik, Dan, A. Subramanian, and F. Trebbi, 2004, “Institutions Rule,” Journal of Economic Growth, Vol. 9, pp. 131–165.

### Doorsteps (empirical correlations reported in figures)
- Doorsteps and Democracy
  - D1 Rescaled vs EIU Overall Score: ρ = 0.8567***
  - D2 Rescaled vs EIU Overall Score: ρ = 0.8593***
  - D3 Rescaled vs EIU Overall Score: ρ = 0.6820***
  - D Overall vs EIU Overall Score: ρ = 0.8715***
- Doorsteps and Equal Access
  - D1 Rescaled vs Access: ρ = 0.7124***
  - D2 Rescaled vs Access: ρ = 0.7029***
  - D3 Rescaled vs Access: ρ = 0.5909***
  - D Overall vs Access: ρ = 0.7303***
- Doorsteps and Corruption
  - D1 Rescaled vs CPI: ρ = 0.8838***
  - D2 Rescaled vs CPI: ρ = 0.7763***
  - D3 Rescaled vs CPI: ρ = 0.6597***
  - D Overall vs CPI: ρ = 0.8534***
- Doorsteps and Government Expenditure
  - D1 Rescaled vs Expenditure: ρ = 0.6863***
  - D2 Rescaled vs Expenditure: ρ = 0.5465***
  - D3 Rescaled vs Expenditure: ρ = 0.6575***
  - D Overall vs Expenditure: ρ = 0.7010***
- Doorsteps and Education Expenditure
  - D1 Rescaled vs Education/GDP: ρ = 0.4863***
  - D2 Rescaled vs Education/GDP: ρ = 0.5023***
  - D3 Rescaled vs Education/GDP: ρ = 0.4809***
  - D Overall vs Education/GDP: ρ = 0.5308***
- Doorsteps and Territorial Coverage of Public Services
  - D1 Rescaled vs A905: ρ = 0.7392***
  - D2 Rescaled vs A905: ρ = 0.6640***
  - D3 Rescaled vs A905: ρ = 0.6368***
  - D Overall vs A905: ρ = 0.7490***
- Doorsteps and Income Inequality
  - D1 Rescaled vs Adjusted Gini: ρ = -0.4410***
  - D2 Rescaled vs Adjusted Gini: ρ = -0.3876***
  - D3 Rescaled vs Adjusted Gini: ρ = -0.3821***
  - D Overall vs Adjusted Gini: ρ = -0.4408***
- Doorsteps and Inflation
  - D1 Rescaled vs CPI Inflation: ρ = -0.3973***
  - D2 Rescaled vs CPI Inflation: ρ = -0.3567***
  - D3 Rescaled vs CPI Inflation: ρ = -0.3070***
  - D Overall vs CPI Inflation: ρ = -0.3944***
- Doorsteps and GDP per Capita
  - D1 Rescaled vs GDPpc: ρ = 0.7924***
  - D2 Rescaled vs GDPpc: ρ = 0.7058***
  - D3 Rescaled vs GDPpc: ρ = 0.6253***
  - D Overall vs GDPpc: ρ = 0.7773***
- Doorsteps and Private Sector Credit
  - D1 Rescaled vs Credit: ρ = 0.6876***
  - D2 Rescaled vs Credit: ρ = 0.5972***
  - D3 Rescaled vs Credit: ρ = 0.5221***
  - D Overall vs Credit: ρ = 0.6723***

### Appendix — Table A–1: Description of Databases (excerpt)
- NAME OF DATABASE / DESCRIPTION OF DATABASE
  - Institutional Profiles Database (IPD): A database on the institutional characteristics of developed and developing countries compiled by Meisel, Ould Aoudia et al. We use both the 2006 and the 2009 version of this database. The 2006 version contains information on 85 countries, whereas the 2009 version contains information on 123 countries. Note that a few of the variables of the 2006 database are not included in the 2009 database. We always used the most recent available information.
  - Polity IV: A dataset compiled by Marshall and Jaggers. It contains information on political regime characteristics and transitions covering “all major, independent states in the global system” over the period between 1800 and 2008 Unless otherwise indicated, we use data for 2008.
  - Economic Freedom of the World (EFW): A database developed for the Economic Freedom Network and compiled by Gwartney, Lawson et al. The current (2010) version contains information on economic freedom in 141 nations between 1979 and

*Source: _wp1287 - REFERENCES (IMF PDF)._*

### 2008. Unless otherwise indicated, we use data for 2008.

### _wp1287 - 2008. Unless otherwise indicated, we use data for 2008.

### Data sources and coverage
- Database of Political Institutions (DPI): published by the World Bank; compiled by Keefer, Clarke, Walsh, and Groff. Most recent version (2010) contains data on political institutions in 178 countries between 1975 and 2009. Unless otherwise indicated, DPI data used from 2009 (DPI (2006) entries noted where applicable).
- Polity IV: variables from Polity IV cited with (2009) where indicated.
- International Property Database (IPD): variables cited with (2006) or (2009) as indicated.
- Economic Freedom of the World (EFW): variables cited with (2008) where indicated.
- ICC 2009: Dummy indicator: 1 if ICC membership is enacted (used as measure of the immunity of members of the government).

### Composition of Doorstep 1 (DOORSTEP 1)
- Category 1 — Existence of rules for the elite
  - XREG, Polity IV (2009): Regulation of chief executive recruitment.
  - XCONST, Polity IV (2009): Constraints on the executive (institutionalized limitations).
  - PARREG, Polity IV (2009): Existence and range of binding rules on when, whether and how political preferences are expressed.
  - ICC 2009: Dummy: 1 if ICC membership is enacted (measure of immunity of members of the government).
- Category 2 — Independence and impartiality of the judicial system
  - A3050, IPD (2009): Independence of the justice system from the government.
  - 2B, EFW (2008): Impartiality of the courts.
  - 2E, EFW (2008): Integrity of the legal system.
- Category 3 — Elite’s respect for the courts’ decisions
  - A604, IPD (2009): Government’s respect for contracts.
  - A600, IPD (2009): Security of property rights.
  - A601, IPD (2009): Security of contracts between private agents.

### Composition of Doorstep 2 (DOORSTEP 2)
- Category 1 — Stability and permanence of the political system
  - A105, IPD (2009): Stability of the political system (measured by change in and compliance with rules concerning the head of state’s accession to power).
  - A500, IPD (2009): Dialogue structures headed by the political authority to find common interest among stakeholders.
  - Interregnum, Polity IV: Number of Interregnum periods in Polity IV between 1995 and 2008 divided by 14.
  - Age of parties, DPI (2006): Average age of political parties in 2009.
  - Finittrm, DPI (2006): Dummy: 1 if there is a constitutional limit on the number of years an executive can serve.
  - Constitutional stability: Number of changes in the constitution between 1995 and 2009 divided by 15.
- Category 2 — Ease of creating organizations
  - A602, IPD (2006): Form of contracts between private agents (oral or written form; with or without mediation).
  - A504, IPD (2006): Degree to which organizations are representative of all stakeholders.
  - B300, IPD (2009): Administrative steps required for the creation of an organization.
  - B700, IPD (2009): Ease of market entry for new firms in production sector.
  - D100, IPD (2009): Freedom of association.
  - D101, IPD (2009): Pluralism and autonomy of trade unions.

### Composition of Doorstep 3 (DOORSTEP 3)
- Category 1 — Political control over the legal armed forces
  - A201, IPD (2009): Political authority’s control over the legal armed forces.
  - MILITARY, DPI (2009): Dummy: 1 if the Chief Executive is a military officer (0-1 reversed for consistency).
  - Parliamentary oversight (2009): Dummy: 1 if parliament oversees military- and defense-related operations.
- Category 2 — Military interference in political life
  - A1002, IPD (2009): De jure and de facto participation of the armed forces in political life.
  - 2D, EFW (2008): Measure of military interference in the rule of law and in the political process.
  - DEFMIN, DPI (2009): Dummy: 1 if the Defense Minister is a military officer (0-1 reversed for consistency).
- Category 3 — Level of armed violence
  - A2000, IPD (2009): Security of persons and goods.
  - A2001, IPD (2009): Conflicts of ethnic, religious and regional nature.
  - A2002, IPD (2009): Violent actions by underground political organizations.

### Country Doorsteps Scores (Table A–5) — selected country entries (D1, D2, D3, D_OVERALL)
- Algeria — D1: 0.61, D2: 0.85, D3: 0.74, D_OVERALL: 0.74
- Angola — D1: 0.58, D2: 0.74, D3: 0.72, D_OVERALL: 0.68
- Argentina — D1: 0.89, D2: 0.93, D3: 1.18, D_OVERALL: 1.00
- Australia — D1: 1.55, D2: 1.36, D3: 1.31, D_OVERALL: 1.41
- Austria — D1: 1.57, D2: 1.27, D3: 1.27, D_OVERALL: 1.37
- Bangladesh — D1: 0.74, D2: 0.94, D3: 0.87, D_OVERALL: 0.85
- Belgium — D1: 1.34, D2: 1.25, D3: 1.27, D_OVERALL: 1.29
- Brazil — D1: 1.20, D2: 1.03, D3: 1.16, D_OVERALL: 1.13
- Canada — D1: 1.47, D2: 1.25, D3: 1.31, D_OVERALL: 1.34
- China — D1: 0.70, D2: 0.89, D3: 0.80, D_OVERALL: 0.80
- France — D1: 1.44, D2: 1.10, D3: 1.30, D_OVERALL: 1.28
- Germany — D1: 1.54, D2: 1.21, D3: 1.31, D_OVERALL: 1.35
- India — D1: 0.91, D2: 0.99, D3: 1.06, D_OVERALL: 0.99
- Japan — D1: 1.37, D2: 1.04, D3: 1.27, D_OVERALL: 1.23
- Mexico — D1: 0.97, D2: 1.05, D3: 0.89, D_OVERALL: 0.97
- Netherlands — D1: 1.57, D2: 1.33, D3: 1.31, D_OVERALL: 1.40
- Norway — D1: 1.55, D2: 1.33, D3: 1.27, D_OVERALL: 1.38
- Russia — D1: 0.75, D2: 0.81, D3: 1.11, D_OVERALL: 0.89
- South Africa — D1: 1.15, D2: 1.11, D3: 1.01, D_OVERALL: 1.09
- Sweden — D1: 1.59, D2: 1.30, D3: 1.30, D_OVERALL: 1.40
- Switzerland — D1: 1.54, D2: 1.24, D3: 1.31, D_OVERALL: 1.36
- Turkey — D1: 1.01, D2: 1.04, D3: 0.75, D_OVERALL: 0.93
- United Kingdom (UK) — D1: 1.55, D2: 1.31, D3: 1.27, D_OVERALL: 1.38
- United States (USA) — D1: 1.32, D2: 1.32, D3: 1.18, D_OVERALL: 1.27
- Venezuela — D1: 0.54, D2: 0.88, D3: 0.48, D_OVERALL: 0.63
- Zimbabwe — D1: 0.40, D2: 0.72, D3: 0.56, D_OVERALL: 0.56
- (Table A–5 contains full country coverage with D1, D2, D3, and D_OVERALL for all listed countries.)

### Spearman rank correlations — key statistics (Tables A–6 to A–8)
- Table A–6–a. D1 and its Categories
  - D1 vs D1-Category 1: 0.84 (p-value 0.00)
  - D1 vs D1-Category 2: 0.87 (p-value 0.00)
  - D1 vs D1-Category 3: 0.90 (p-value 0.00)
  - D1-Category 1 vs D1-Category 2: 0.60 (p-value 0.00)
  - D1-Category 1 vs D1-Category 3: 0.62 (p-value 0.00)
  - D1-Category 2 vs D1-Category 3: 0.75 (p-value 0.00)
- Table A–6–b. D2 and its Categories
  - D2 vs D2-Category 1: 0.73 (p-value 0.00)
  - D2 vs D2-Category 2: 0.89 (p-value 0.00)
  - D2-Category 1 vs D2-Category 2: 0.36 (p-value 0.00)
- Table A–6–c. D3 and its Categories
  - D3 vs D3-Category 1: 0.81 (p-value (0.00))
  - D3 vs D3-Category 2: 0.91 (p-value (0.00))
  - D3 vs D3-Category 3: 0.89 (p-value (0.00))
  - D3-Category 1 vs D3-Category 2: 0.63 (p-value (0.00))
  - D3-Category 1 vs D3-Category 3: 0.64 (p-value (0.00))
  - D3-Category 2 vs D3-Category 3: 0.72 (p-value (0.00))
- Table A–7. D_Overall, D1, D2, and D3
  - D_Overall vs D1: 0.96 (p-value (0.00))
  - D_Overall vs D2: 0.86 (p-value (0.00))
  - D_Overall vs D3: 0.92 (p-value (0.00))
  - D1 vs D2: 0.82 (p-value (0.00))
  - D1 vs D3: 0.82 (p-value (0.00))
  - D2 vs D3: 0.67 (p-value (0.00))
- Table A–8. D_Overall and all Categories (selected)
  - D_Overall vs D1-Category 1: 0.81 (p-value (0.00))
  - D_Overall vs D1-Category 2: 0.83 (p-value (0.00))
  - D_Overall vs D1-Category 3: 0.86 (p-value (0.00))
  - D_Overall vs D2-Category 1: 0.57 (p-value (0.00))
  - D_Overall vs D2-Category 2: 0.80 (p-value (0.00))
  - D_Overall vs D3-Category 1: 0.81 (p-value (0.00))
  - D_Overall vs D3-Category 2: 0.87 (p-value (0.00))
  - D_Overall vs D3-Category 3: 0.77 (p-value (0.00))

*Content based on tables and variable descriptions in the source PDF.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1287.pdf_