## 1. Introduction

## Source details

**Canonical URL:** [1. Introduction](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025243-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025243-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025243-source-pdf.pdf.json)

---

### Scope, data, and empirical framework
- Sample: 38 emerging market economies across all regions; quarterly panel spanning 1990 to 2020.
- Primary dependent variables: gross private capital inflows and outflows (USD), standardized by annual trend GDP.
- Main political uncertainty proxy: election dummies constructed from DPI2020 and NELDA (quarter = 1 if an election occurred during the quarter).
- Number of election episodes in sample period: 261 elections.
  - Incumbent lost in: 152 elections.
  - Average duration between consecutive elections: close to 4 years.
- Empirical model: fixed-effects panel regressions with country fixed effects, country-specific time trends, controls for push (global) and pull (domestic) factors, and Driscoll–Kraay corrected standard errors to account for serial correlation and cross-sectional dependence.
- Key control variables:
  - Push: log(VIX), composite global money supply growth (US, euro area, Japan, UK).
  - Push–pull: real interest rate differentials (real domestic rate – real risk-free rate using shadow federal funds rate), growth rate differentials (country vs. advanced-world average).
  - Pull: ICRG political stability index (0–100), Chinn-Ito capital account openness index, IMF Financial Market Development Index.
  - Alternative uncertainty measures used in robustness: World Uncertainty Index (WUI) and Economic Policy Uncertainty (EPU).

### Stylized facts on capital flows and election cycles
- Aggregate dynamics:
  - Peak gross private inflows pre-global financial crisis: over 10% of trend quarterly GDP.
  - FDI share: stable at 2 percent of aggregate GDP over last 25 years.
  - Banking/other (cross-border lending) flows surged before 2007 and collapsed after the GFC.
  - Outflows increased from below 1% of trend quarterly GDP in the 1990s to over 3% in the most recent decade.
- Election-cycle pattern:
  - Average private inflows decline sharply in the quarter immediately after the election; do not immediately return to pre-election averages.
  - No consistent, statistically significant pre-election decline in the pooled sample, but declines are observed in the immediate aftermath (up to two post-election quarters) for many specifications.

### Main empirical findings — timing, magnitude, and heterogeneity
- Timing and persistence:
  - Gross private capital inflows subside during election cycles with the most pronounced effects in the election quarter and extending through the two post-election quarters.
  - Persistent adverse effects (beyond immediate quarter) arise when political uncertainty lingers after elections (e.g., violence, incumbent defeat, non-predetermined elections).
- Magnitude for least politically stable countries (Table 6):
  - Countries in the lowest ICRG political stability quartile: change in gross inflows during the election quarter = -1.25 percent of trend GDP.
  - This corresponds to a -28.3% relative change in gross inflows compared with the average flows during the 6 months before the election.
  - For non-predetermined elections: gross inflows change = -2.15 percent of trend GDP; relative change = -48.7%.
  - When incumbent lost: gross inflows change = -2.53 percent of trend GDP; relative change = -57.3%.
  - When incumbent won: gross inflows change = 0.33 percent of trend GDP; relative change = 7.6%.
- Flow-type heterogeneity:
  - Debt vs. non-debt: election-quarter hit is larger and more statistically significant for debt flows than for non-debt flows.
  - Instrument types:
    - FDI: negative and statistically significant effect in election quarter; more persistent declines post-election in some specifications.
    - Portfolio flows: do not show a statistically significant decline around elections in pooled regressions.
    - Other investment flows (banking/cross-border lending): decline in election quarters and are sensitive to global liquidity and risk aversion.
  - Net flows: pattern aligns with gross inflows given inflows dominate net dynamics in EMs; political variables weaker for net flows due to dilution by outflows.
- Role of political stability and institutions:
  - ICRG political stability index: higher scores associated with larger capital inflows; interaction with election dummy positive and significant — political stability mitigates negative election effects.
  - Countries with ICRG score above the sample average: no statistically significant decline in capital inflows around elections in baseline specifications.
  - Decomposed ICRG pillars: socioeconomic conditions, investment profile, and tensions/conflicts matter as standalone determinants; interaction terms with elections are positive and significant for most pillars (Table 7).
  - World Bank governance indicators (annual): overall governance and specific dimensions — control of corruption, government effectiveness, rule of law, regulatory quality — are significant drivers of inflows and mitigate election-related declines when interacted with election dummies (Table 8).
- Election characteristics influencing uncertainty and persistence:
  - Violence (post- or peri-election): both peaceful and violent elections see lower inflows in the election quarter, but violent elections show persistent negative effects lasting one to two quarters post-election (Table 5).
  - Non-predetermined (off-cycle/snap) elections: negative effect endures up to four quarters post-election (Table 5).
  - Incumbent turnover: when incumbent loses, negative effects persist up to two quarters post-election; if incumbent wins, negative effect is smaller/short-lived (Table 5).

### Robustness and endogeneity checks
- Endogeneity addressed by:
  - Instrumental variable approach using time since last election; Durbin–Wu–Hausman test does not reject exogeneity of election dummy.
  - Granger causality tests: elections Granger-cause capital flows; capital flows do not Granger-cause elections.
  - Subsample of predetermined (constitutionally fixed) elections: main conclusions confirmed.
- Alternative uncertainty measures:
  - WUI and EPU included in various specifications; coefficients on WUI/EPU not statistically significant in most specifications, while election-related variables remain robust.
- Alternative model/specification checks:
  - Normalizing flows by quarterly GDP instead of trend GDP: baseline inference largely unchanged (some sensitivity with pre-election quarters).
  - Fixed effects variations (exclude country trends; add overall time trend; add annual time fixed effects): findings robust.
  - Excluding 2020 and including GFC dummy: main conclusions unaffected.
  - Controlling for U.S. election timing: no significant relationship with EM private inflows; domestic election dummies remain significant.
  - Controlling for common factors: include quarterly cross-sectional averages of flows and year-region fixed effects — main results robust.
- Limitations:
  - Dataset end point: 2020 (more recent years excluded due to lags in political variable databases).
  - Low R2 in flow regressions is typical for capital flows research; large unexplained variation remains.

### Descriptive statistics — key variables (Appendix A / Appendix B)
- Private inflows (% of trend GDP): Mean: 5.2; Std. dev.: 6.5; Min: -23.6; Max: 44.9.
- Private outflows (% of trend GDP): Mean: 1.7; Std. dev.: 4.2; Min: -28.1; Max: 34.4.
- Net private flows (% of trend GDP): Mean: 3.6; Std. dev.: 6.3; Min: -21.8; Max: 45.4.
- FDI inflows (% of trend GDP): Mean: 2.5; Std. dev.: 3.1; Min: -5.0; Max: 28.1.
- Portfolio inflows (% of trend GDP): Mean: 1.1; Std. dev.: 3.3; Min: -17.3; Max: 37.1.
- Private other inflows (% of trend GDP): Mean: 0.6; Std. dev.: 3.4; Min: -36.2; Max: 19.3.
- Debt inflows (% of trend GDP): Mean: 2.6; Std. dev.: 5.5; Min: -24.4; Max: 39.3.
- Non-debt inflows (% of trend GDP): Mean: 2.8; Std. dev.: 3.4; Min: -11.2; Max: 29.3.
- VIX index: Mean: 18.9; Std. dev.: 5.2; Min: 12.0; Max: 29.9.
- Global liquidity growth (% qoq): Mean: 1.2; Std. dev.: 3.0; Min: -4.4; Max: 8.8.
- Real interest rate differential (%): Mean: -0.7; Std. dev.: 24.5; Min: -238.5; Max: 276.8.
- Growth differential (%): Mean: 1.1; Std. dev.: 5.1; Min: -20.4; Max: 21.3.
- Financial market development index: Mean: 0.2; Std. dev.: 0.2; Min: 0.0; Max: 0.6.
- Chinn-Ito index: Mean: -0.2; Std. dev.: 1.4; Min: -1.9; Max: 2.3.
- Stability (share): Mean: 0.2; Std. dev.: 0.3; Min: 0.0; Max: 1.0.
- Polarization (unit): Mean: 0.6; Std. dev.: 0.9; Min: 0.0; Max: 2.0.
- Victory margin (% point): Mean: 21.6; Std. dev.: 21.1; Min: 0.3; Max: 80.0.
- WUI (index): Mean: 0.1; Std. dev.: 0.2; Min: 0.0; Max: 1.3.
- EPU (index): Mean: 76.0; Std. dev.: 13.8; Min: 59.3; Max: 112.9.
- ICRG_Political (0,100): Mean: 64.2; Std. dev.: 10.2; Min: 22.0; Max: 87.0.
- ICRG_Institutions (0,16): Mean: 8.8; Std. dev.: 2.3; Min: 2.0; Max: 14.0.
- ICRG_Tensions (0,36): Mean: 28.4; Std. dev.: 5.3; Min: 7.0; Max: 36.0.
- ICRG_Political_Pillar (0,24): Mean: 15.3; Std. dev.: 3.4; Min: 4.0; Max: 23.0.
- ICRG_Investment_Risk (0,12): Mean: 6.6; Std. dev.: 1.9; Min: 1.0; Max: 11.0.
- ICRG_Soc_Cond (0,12): Mean: 5.5; Std. dev.: 1.7; Min: 0.0; Max: 10.0.
- Polity (-10,10): Mean: 5.2; Std. dev.: 4.9; Min: -9; Max: 10.
- WB control of corruption (percentile): Mean: 48.3; Std. dev.: 20.9; Min: 1; Max: 91.
- WB government effectiveness (percentile): Mean: 52.6; Std. dev.: 17.9; Min: 11; Max: 87.
- WB political stability and absence of violence (percentile): Mean: 43.7; Std. dev.: 23.0; Min: 3; Max: 93.
- WB rule of law (percentile): Mean: 47.8; Std. dev.: 18.2; Min: 16; Max: 87.
- WB regulatory quality (percentile): Mean: 56.6; Std. dev.: 17.6; Min: 22; Max: 94.
- WB voice and accountability (percentile): Mean: 53.9; Std. dev.: 17.7; Min: 10; Max: 88.

### Policy-relevant implications
- Institutional strengthening as a buffer:
  - Enhancing control of corruption, rule of law, government effectiveness, and regulatory quality reduces vulnerability of capital inflows to election-related uncertainty and can stabilize investment during electoral cycles.
- Monitoring high-uncertainty election episodes:
  - Elections with violence, non-predetermined timing, or incumbent defeat warrant closer monitoring for capital flow volatility and potential financial stability risks given the documented persistence and magnitude of outflows in such episodes.
- Consideration for macrofinancial planning:
  - Countries in the lowest ICRG political stability quartile face particularly large election-quarter declines in gross inflows (e.g., -1.25 percent of trend GDP); macroeconomic policymakers should account for temporary financing gaps and possible exchange rate and credit market pressures around election cycles.
- Differentiated approach by flow type:
  - Policies to manage short-term liquidity and borrowing risks should prioritize debt flows and other investment flows, which are more sensitive to global risk aversion and liquidity shifts; long-term FDI is also affected by election uncertainty and may require measures to preserve investor confidence.

*Source: IMF Working Paper — Elections Matter: Capital Flows and Election Cycles (Appendix B: Descriptive statistics; Appendix A: Countries and data sources).*

### 1. Introduction ........................................................................................................

### 1. Introduction (Content Unit: wpiea2025243-source-pdf - 1. Introduction)

### Major sections (chapter/section inventory)
- 1. Introduction ..................................................................................................................................................... 3
- 2. Literature Review ............................................................................................................................................ 6
  - 2.1. Drivers of Capital Flows .............................................................................................................................. 6
  - 2.2. Political Uncertainty and Capital Flows ...................................................................................................... 7
- 3. Data Sources ................................................................................................................................................... 8
  - 3.1. Capital Flows .............................................................................................................................................. 9
  - 3.2. Elections and Related Variables ............................................................................................................... 10
  - 3.3. Other Control Variables ............................................................................................................................ 12
- 4. Stylized Facts ................................................................................................................................................ 13
  - 4.1. Evolution of Capital Flows ........................................................................................................................ 13
  - 4.2. Election Cycles ......................................................................................................................................... 15
- 5. Empirical Analysis and Main Findings ........................................................................................................ 16
  - 5.1. Empirical Framework ................................................................................................................................ 16
  - 5.2. Empirical Findings .................................................................................................................................... 18
    - 5.2.1. Push–Pull Drivers and Election-Induced Political Uncertainty ........................................................... 18
    - 5.2.2. Pre- and Post-Election Political Landscape and Gross Capital Inflows ............................................. 24
    - 5.2.3. Quantifying the Impact of Political Uncertainty .................................................................................. 28
    - 5.2.4. The Role of Institutional Capacity ...................................................................................................... 29
  - 5.3. Robustness Checks .................................................................................................................................. 33
- 6. Conclusions ................................................................................................................................................... 38
- Appendix ............................................................................................................................................................ 40
- References ......................................................................................................................................................... 42

### Figures listed in the content unit
- Figure 1: Dynamics of private capital inflows and outflows ......................................................................... 14
- Figure 2: Structure of private capital inflows ................................................................................................. 15
- Figure 3: Private capital inflows (% of trend GDP) and election cycles ...................................................... 16

### Tables listed in the content unit
- Table 1: Election characteristics ..................................................................................................................... 15
- Table 2: Gross capital inflows and elections ................................................................................................. 19
- Table 3: Various types of capital flows and elections ................................................................................... 22
- Table 4: Gross capital inflows vs. pre-election political landscape ............................................................. 25
- Table 5: Gross capital inflows: the persistence of high-uncertainty election effects ................................ 27
- Table 6: Change in capital flows during the election quarter: countries in the lowest ICRG political 
  stability quartile ................................................................................................................................................. 29
- Table 7: Gross capital inflows and institutional characteristics (ICRG) ...................................................... 30
- Table 8: Gross capital inflows and institutional characteristics (WB Governance) ................................... 32
- Table 9: Gross capital inflows and alternative uncertainty measures ......................................................... 34
- Table 10: Gross capital inflows and indicators of democracy ..................................................................... 35
- Table 11: Gross capital inflows: Alternative model specifications .............................................................. 37

### Appendix
- Appendix A: Countries and data sources ....................................................................................................... 40

*Source: wpiea2025243-source-pdf - 1. Introduction (https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025243-source-pdf.pdf)*

### Appendix B: Descriptive statistics .....................................................................................

### Appendix B: Descriptive statistics

### Scope, data, and empirical framework
- Sample: 38 emerging market economies across all regions; quarterly panel spanning 1990 to 2020.
- Primary dependent variables: gross private capital inflows and outflows (USD), standardized by annual trend GDP.
- Main political uncertainty proxy: election dummies constructed from DPI2020 and NELDA (quarter = 1 if an election occurred during the quarter).
- Number of election episodes in sample period: 261 elections.
  - Incumbent lost in: 152 elections.
  - Average duration between consecutive elections: close to 4 years.
- Empirical model: fixed-effects panel regressions with country fixed effects, country-specific time trends, controls for push (global) and pull (domestic) factors, and Driscoll–Kraay corrected standard errors to account for serial correlation and cross-sectional dependence.
- Key control variables:
  - Push: log(VIX), composite global money supply growth (US, euro area, Japan, UK).
  - Push–pull: real interest rate differentials (real domestic rate – real risk-free rate using shadow federal funds rate), growth rate differentials (country vs. advanced-world average).
  - Pull: ICRG political stability index (0–100), Chinn-Ito capital account openness index, IMF Financial Market Development Index.
  - Alternative uncertainty measures used in robustness: World Uncertainty Index (WUI) and Economic Policy Uncertainty (EPU).

### Stylized facts on capital flows and election cycles
- Aggregate dynamics:
  - Peak gross private inflows pre-global financial crisis: over 10% of trend quarterly GDP.
  - FDI share: stable at 2 percent of aggregate GDP over last 25 years.
  - Banking/other (cross-border lending) flows surged before 2007 and collapsed after the GFC.
  - Outflows increased from below 1% of trend quarterly GDP in the 1990s to over 3% in the most recent decade.
- Election-cycle pattern:
  - Average private inflows decline sharply in the quarter immediately after the election; do not immediately return to pre-election averages.
  - No consistent, statistically significant pre-election decline in the pooled sample, but declines are observed in the immediate aftermath (up to two post-election quarters) for many specifications.

### Main empirical findings — magnitude, timing, and heterogeneity
- Timing and persistence:
  - Gross private capital inflows subside during election cycles with the most pronounced effects in the election quarter and extending through the two post-election quarters.
  - Persistent adverse effects (beyond immediate quarter) arise when political uncertainty lingers after elections (e.g., violence, incumbent defeat, non-predetermined elections).
- Magnitude for least politically stable countries:
  - Countries in the lowest ICRG political stability quartile: change in gross inflows during the election quarter = -1.25 percent of trend GDP (Table 6).
  - This corresponds to a -28.3% relative change in gross inflows compared with the average flows during the 6 months before the election (Table 6).
  - For non-predetermined elections: gross inflows change = -2.15 percent of trend GDP; relative change = -48.7% (Table 6).
  - When incumbent lost: gross inflows change = -2.53 percent of trend GDP; relative change = -57.3% (Table 6).
  - When incumbent won: gross inflows change = 0.33 percent of trend GDP; relative change = 7.6% (Table 6).
- Flow-type heterogeneity:
  - Debt vs. non-debt: election-quarter hit is larger and more statistically significant for debt flows than for non-debt flows.
  - Instrument types:
    - FDI: negative and statistically significant effect in election quarter; more persistent declines post-election in some specifications.
    - Portfolio flows: do not show a statistically significant decline around elections in pooled regressions.
    - Other investment flows (banking/cross-border lending): decline in election quarters and are sensitive to global liquidity and risk aversion.
  - Net flows: pattern aligns with gross inflows given inflows dominate net dynamics in EMs; political variables weaker for net flows due to dilution by outflows.
- Role of political stability and institutions:
  - ICRG political stability index: higher scores associated with larger capital inflows; interaction with election dummy positive and significant — political stability mitigates negative election effects.
  - Countries with ICRG score above the sample average: no statistically significant decline in capital inflows around elections in baseline specifications.
  - Decomposed ICRG pillars: socioeconomic conditions, investment profile, and tensions/conflicts matter as standalone determinants; interaction terms with elections are positive and significant for most pillars (Table 7).
  - World Bank governance indicators (annual): overall governance and specific dimensions — control of corruption, government effectiveness, rule of law, regulatory quality — are significant drivers of inflows and mitigate election-related declines when interacted with election dummies (Table 8).
- Election characteristics influencing uncertainty and persistence:
  - Violence (post- or peri-election): both peaceful and violent elections see lower inflows in the election quarter, but violent elections show persistent negative effects lasting one to two quarters post-election (Table 5).
  - Non-predetermined (off-cycle/snap) elections: negative effect endures up to four quarters post-election (Table 5).
  - Incumbent turnover: when incumbent loses, negative effects persist up to two quarters post-election; if incumbent wins, negative effect is smaller/short-lived (Table 5).

### Robustness and endogeneity checks
- Endogeneity addressed by:
  - Instrumental variable approach using time since last election; Durbin–Wu–Hausman test does not reject exogeneity of election dummy.
  - Granger causality tests: elections Granger-cause capital flows; capital flows do not Granger-cause elections.
  - Subsample of predetermined (constitutionally fixed) elections: main conclusions confirmed.
- Alternative uncertainty measures:
  - WUI and EPU included in various specifications; coefficients on WUI/EPU not statistically significant in most specifications, while election-related variables remain robust.
- Alternative model/specification checks:
  - Normalizing flows by quarterly GDP instead of trend GDP: baseline inference largely unchanged (some sensitivity with pre-election quarters).
  - Fixed effects variations (exclude country trends; add overall time trend; add annual time fixed effects): findings robust.
  - Excluding 2020 and including GFC dummy: main conclusions unaffected.
  - Controlling for U.S. election timing: no significant relationship with EM private inflows; domestic election dummies remain significant.
  - Controlling for common factors: include quarterly cross-sectional averages of flows and year-region fixed effects — main results robust.
- Limitations noted:
  - Dataset end point: 2020 (more recent years excluded due to lags in political variable databases).
  - Low R2 in flow regressions is typical for capital flows research; large unexplained variation remains.

### Policy-relevant implications (from empirical results)
- Institutional strengthening as a buffer:
  - Enhancing control of corruption, rule of law, government effectiveness, and regulatory quality reduces vulnerability of capital inflows to election-related uncertainty and can stabilize investment during electoral cycles.
- Monitoring high-uncertainty election episodes:
  - Elections with violence, non-predetermined timing, or incumbent defeat warrant closer monitoring for capital flow volatility and potential financial stability risks given the documented persistence and magnitude of outflows in such episodes.
- Consideration for macrofinancial planning:
  - Countries in the lowest ICRG political stability quartile face particularly large election-quarter declines in gross inflows (e.g., -1.25 percent of trend GDP); macroeconomic policymakers should account for temporary financing gaps and possible exchange rate and credit market pressures around election cycles.
- Differentiated approach by flow type:
  - Policies to manage short-term liquidity and borrowing risks should prioritize debt flows and other investment flows, which are more sensitive to global risk aversion and liquidity shifts; long-term FDI is also affected by election uncertainty and may require measures to preserve investor confidence.

*Source: IMF Working Paper — Elections Matter: Capital Flows and Election Cycles (Appendix B: Descriptive statistics), authors’ estimations using DPI, NELDA, IMF BOPS, WEO, ICRG, WUI, EPU, Haver, IFS, and other datasets as described in the text.*

### Appendix A: Countries and data sources

### Appendix A: Countries and data sources

### Descriptive statistics (Appendix B)
Sources: See Table A

- Variable: Private inflows  
  - Measure: % of trend GDP  
  - Mean: 5.2  
  - Std. dev.: 6.5  
  - Min: -23.6  
  - Max: 44.9

- Variable: Private outflows  
  - Measure: % of trend GDP  
  - Mean: 1.7  
  - Std. dev.: 4.2  
  - Min: -28.1  
  - Max: 34.4

- Variable: Net private flows  
  - Measure: % of trend GDP  
  - Mean: 3.6  
  - Std. dev.: 6.3  
  - Min: -21.8  
  - Max: 45.4

- Variable: FDI inflows  
  - Measure: % of trend GDP  
  - Mean: 2.5  
  - Std. dev.: 3.1  
  - Min: -5.0  
  - Max: 28.1

- Variable: Portfolio inflows  
  - Measure: % of trend GDP  
  - Mean: 1.1  
  - Std. dev.: 3.3  
  - Min: -17.3  
  - Max: 37.1

- Variable: Private other inflows  
  - Measure: % of trend GDP  
  - Mean: 0.6  
  - Std. dev.: 3.4  
  - Min: -36.2  
  - Max: 19.3

- Variable: Debt inflows  
  - Measure: % of trend GDP  
  - Mean: 2.6  
  - Std. dev.: 5.5  
  - Min: -24.4  
  - Max: 39.3

- Variable: Non-debt inflows  
  - Measure: % of trend GDP  
  - Mean: 2.8  
  - Std. dev.: 3.4  
  - Min: -11.2  
  - Max: 29.3

- Variable: VIX index  
  - Measure: index  
  - Mean: 18.9  
  - Std. dev.: 5.2  
  - Min: 12.0  
  - Max: 29.9

- Variable: Global liquidity growth  
  - Measure: % qoq  
  - Mean: 1.2  
  - Std. dev.: 3.0  
  - Min: -4.4  
  - Max: 8.8

- Variable: Real interest rate differential  
  - Measure: %  
  - Mean: -0.7  
  - Std. dev.: 24.5  
  - Min: -238.5  
  - Max: 276.8

- Variable: Growth differential  
  - Measure: %  
  - Mean: 1.1  
  - Std. dev.: 5.1  
  - Min: -20.4  
  - Max: 21.3

- Variable: Financial market development index  
  - Measure: index  
  - Mean: 0.2  
  - Std. dev.: 0.2  
  - Min: 0.0  
  - Max: 0.6

- Variable: Chinn-Ito  
  - Measure: index  
  - Mean: -0.2  
  - Std. dev.: 1.4  
  - Min: -1.9  
  - Max: 2.3

- Variable: Stability  
  - Measure: share  
  - Mean: 0.2  
  - Std. dev.: 0.3  
  - Min: 0.0  
  - Max: 1.0

- Variable: Polarization  
  - Measure: unit  
  - Mean: 0.6  
  - Std. dev.: 0.9  
  - Min: 0.0  
  - Max: 2.0

- Variable: Victory margin  
  - Measure: % point  
  - Mean: 21.6  
  - Std. dev.: 21.1  
  - Min: 0.3  
  - Max: 80.0

- Variable: WUI (World Uncertainty Index)  
  - Measure: index  
  - Mean: 0.1  
  - Std. dev.: 0.2  
  - Min: 0.0  
  - Max: 1.3

- Variable: EPU (Economic Policy Uncertainty)  
  - Measure: index  
  - Mean: 76.0  
  - Std. dev.: 13.8  
  - Min: 59.3  
  - Max: 112.9

- Variable: ICRG_Political  
  - Measure: index (0,100)  
  - Mean: 64.2  
  - Std. dev.: 10.2  
  - Min: 22.0  
  - Max: 87.0

- Variable: ICRG_Institutions  
  - Measure: index (0,16)  
  - Mean: 8.8  
  - Std. dev.: 2.3  
  - Min: 2.0  
  - Max: 14.0

- Variable: ICRG_Tensions  
  - Measure: index (0,36)  
  - Mean: 28.4  
  - Std. dev.: 5.3  
  - Min: 7.0  
  - Max: 36.0

- Variable: ICRG_Political_Pillar  
  - Measure: index (0,24)  
  - Mean: 15.3  
  - Std. dev.: 3.4  
  - Min: 4.0  
  - Max: 23.0

- Variable: ICRG_Investment_Risk  
  - Measure: index (0,12)  
  - Mean: 6.6  
  - Std. dev.: 1.9  
  - Min: 1.0  
  - Max: 11.0

- Variable: ICRG_Soc_Cond  
  - Measure: index (0,12)  
  - Mean: 5.5  
  - Std. dev.: 1.7  
  - Min: 0.0  
  - Max: 10.0

- Variable: Polity  
  - Measure: index (-10,10)  
  - Mean: 5.2  
  - Std. dev.: 4.9  
  - Min: -9  
  - Max: 10

- Variable: WB control of corruption  
  - Measure: percentile  
  - Mean: 48.3  
  - Std. dev.: 20.9  
  - Min: 1  
  - Max: 91

- Variable: WB government effectiveness  
  - Measure: percentile  
  - Mean: 52.6  
  - Std. dev.: 17.9  
  - Min: 11  
  - Max: 87

- Variable: WB political stability and absence of violence  
  - Measure: percentile  
  - Mean: 43.7  
  - Std. dev.: 23.0  
  - Min: 3  
  - Max: 93

- Variable: WB rule of law  
  - Measure: percentile  
  - Mean: 47.8  
  - Std. dev.: 18.2  
  - Min: 16  
  - Max: 87

- Variable: WB regulatory quality  
  - Measure: percentile  
  - Mean: 56.6  
  - Std. dev.: 17.6  
  - Min: 22  
  - Max: 94

- Variable: WB voice and accountability  
  - Measure: percentile  
  - Mean: 53.9  
  - Std. dev.: 17.7  
  - Min: 10  
  - Max: 88

*Appendix A: Countries and data sources — IMF Working Paper Elections Matter: Capital Flows and Election Cycles (Working Paper No. WP/2025/243)*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025243-source-pdf.pdf_