## _wp08242

## Source details

**Canonical URL:** [_wp08242](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08242.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08242.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08242.pdf.json)

---

### I. Introduction — Purpose and main findings
- Central question: whether financial openness yields growth benefits; focus shifted from output growth to productivity (TFP) growth.
- Key empirical findings:
  - Economies with more open capital accounts generally have higher TFP growth.
  - Formal econometric analysis suggests de jure capital account openness has a causal positive effect on TFP growth even after controlling for standard growth determinants and endogeneity.
  - Overall de facto financial integration (gross stocks of foreign assets and liabilities relative to GDP) does not appear to matter for TFP growth in aggregate.
  - Disaggregating de facto integration by liability type yields:
    - FDI and portfolio equity liabilities boost TFP growth.
    - Debt liabilities are negatively correlated with TFP growth.
  - The negative relationship between external debt liabilities and TFP growth is partially attenuated in economies with better-developed financial markets and better institutional quality.

### II. Theoretical motivation and relevance
- Rationale for focusing on TFP:
  - TFP growth is emphasized as the main driver of long-term per capita income growth.
  - Direct welfare/growth gains from capital mobility may be small; benefits are likely reflected in TFP via indirect channels (financial sector development, institutional improvements, technology transfers).
- Channels highlighted:
  - FDI: technology transfers, managerial expertise, imitation, skill acquisition, competition, vertical (backward/forward) spillovers.
  - Portfolio equity: deepening of domestic financial markets, corporate governance improvements.
  - Debt: financing benefits but possible adverse effects (volatility, moral hazard, inefficient allocation) that can harm aggregate efficiency.

### III. Methodology and data
- Empirical framework:
  - Dynamic panel regression framework using non-overlapping ten-year growth rates to capture sustained effects.
  - Endogeneity and reverse causality addressed using Blundell and Bond (1998) system GMM with instrument-limiting refinements.
- Dataset and sample:
  - Annual data over the period 1966–2005 for 67 countries—21 industrial and 46 developing.
  - Primary data sources include Penn World Tables Version 6.2 and databases from the World Bank and IMF; all data in constant (2000) international prices.
- TFP and growth accounting specifics:
  - Cobb-Douglas production function: Y = A K^α (H L)^(1−α).
  - Capital stock depreciation rate: 6 percent (annual).
  - Human capital: Mincerian return parameter 0.085 per additional year of schooling; Barro and Lee (2000) data with extrapolation after 2000.
  - Baseline parameter α assumed to be one-third; alternative capital shares used in robustness checks.
- Measures of financial openness:
  - De jure: binary indicator from IMF’s AREAER (Schindler, 2007) — 1 when capital account is open, 0 otherwise.
  - De facto: ratio of gross stocks of external liabilities to GDP (Lane and Milesi-Ferretti, 2006), with splits into assets/liabilities and by type (FDI+equity vs. debt).
- Controls: trade openness, changes in terms of trade, institutional quality (sum of corruption, law and order, bureaucratic quality; range 0 to 18), financial development (private sector credit to GDP).

### IV. Stylized facts (descriptive results)
- Time periods: full sample 1966-2005 and sub-periods 1966-1985 and 1986-2005 (mid-1980s break-point).
- Grouping: More financially open (MFO) vs. less financially open (LFO) economies, split at cross-sectional median of financial openness (average over full sample).
- Descriptive patterns:
  - Contribution of TFP growth to per-worker output growth is larger in MFO economies.
  - Over 1986-2005, MFO economies enjoyed faster productivity growth; LFO economies relied more on physical and human capital accumulation.
  - Average output growth between MFO and LFO was rather similar during the globalization period, despite differences in TFP contribution shares.
  - Median TFP growth typically higher in MFO economies over 1986-2005; evidence mixed depending on de jure vs. de facto classification.

### V. Regression results — main econometric findings
- Cross-section regressions (full sample 1966-2005, OLS, Countries = 67):
  - Initial TFP (logs): coefficients ~ -0.01883*** to -0.01849*** with standard errors [0.00197]–[0.00211].
  - Trade openness coefficients near zero (examples: -0.00001, 0.00007; SEs [0.00007]–[0.00008]).
  - Terms of trade often positive and significant (example: 0.00129** [0.00063]).
  - Population growth negative and significant (examples: -0.00449*** [0.00124] to -0.00480*** [0.00127]).
  - Institutional quality positive and significant (examples: 0.00067** [0.00032] to 0.00071** [0.00033]).
  - De jure capital account openness not significant in simple cross-sections (example: -0.00396 [0.00292]).
  - Total liabilities/assets coefficients near zero and not significant (example: -0.00004 [0.00003]).
  - R-squared: 0.643–0.662; Observations: 67.
- Ten-year panel regressions (FE and System GMM; Countries = 67; Observations: 252):
  - Initial TFP (logs): strong negative in FE (e.g., -0.61287*** [0.08458]); negative and significant in some GMM columns (e.g., -0.38540** [0.16861]).
  - Trade openness positive and significant in FE (e.g., 0.00498** [0.00215]).
  - Private sector credit positive and often significant (examples: 0.00116** [0.00054]; 0.00311*** [0.00100]).
  - De jure capital account openness positive and significant in many specifications (examples: 0.07373** [0.03547]; 0.15476** [0.06056]).
  - Total liabilities/assets generally not significant (example: Total Liabilities -0.00017 [0.00037]).
  - System GMM diagnostics: Hansen Test p-values e.g., 0.211–0.464; 2nd Order Correlation p-values ~0.168–0.178; Number of instruments: 18–20.
- Quantitative GMM associations:
  - De jure openness: economies with an open capital account have, over a ten-year horizon, annual TFP growth about 0.11–0.15 percentage points higher than economies with extensive capital controls.
- Composition of liabilities (system GMM and FE results):
  - FDI & portfolio equity liabilities:
    - Positive associations; examples: 0.00201*** [0.00066], 0.00379** [0.00161], 0.00607*** [0.00220], 0.00695*** [0.00207].
    - Quantitative summary: a 10 percentage point increase in the ratio of FDI and equity liabilities to GDP is associated with about a 0.4 percentage points increase in annual TFP growth over a ten-year period.
  - Debt liabilities:
    - Negative associations; examples: -0.00178** [0.00069], -0.00247** [0.00096], -0.00383*** [0.00117], -0.00378*** [0.00087].
    - Quantitative summary: a 10 percentage point increase in the ratio of debt liabilities to GDP is associated with TFP growth lower by about 0.2 percentage points.
- Mechanisms: equity-like inflows convey technology spillovers, corporate governance benefits, and financial deepening; debt inflows are more volatile/procyclical and can harm efficiency where institutions/financial supervision are weak.

### VI. Interaction effects, thresholds, and heterogeneity
- Interaction with financial development (private sector credit to GDP):
  - Positive interaction between private sector credit and stock of debt liabilities attenuates debt’s negative impact.
  - Calculated threshold where marginal effect of external debt stock becomes positive: credit to GDP ratio nearly 150 percent (based on coefficients: debt liabilities coefficient −0.00383 and interaction coefficient 0.00261).
  - Sample mean of private sector credit to GDP in 1996-2005: 0.57 (standard deviation: 0.44).
- Interaction with institutional quality:
  - Better institutional quality reduces negative impact of debt liabilities on TFP growth.
  - Even at highest institutional quality in sample, marginal effect: FDI+equity positive; debt remains negative.
- Non-industrial countries (Countries = 46):
  - De jure openness coefficient larger and significant in some GMM specifications (example: 0.20021** [0.08287]).
  - Debt liabilities negative and strongly significant (examples: -0.00315** [0.00129], -0.00602*** [0.00177]).
  - FDI & equity effects positive but not always significant for non-industrials.
- Thresholds by level of financial integration (MFO vs LFO; Countries ~33–34):
  - MFO economies: positive coefficients on de jure openness and on FDI & equity liabilities; negative on debt liabilities.
  - LFO economies: coefficients much smaller and not statistically significant.
  - Contextual shares (2000-04): debt ~52 percent and FDI ~37 percent of gross external liabilities for emerging markets; in 1980-84 debt ~85 percent and FDI ~14 percent.
  - Commodity-exporter dummy and interactions made little difference to key results.

### VII. Horizons, persistence, alternative specifications, and robustness
- Different horizons (FE estimates reported across 3-, 5-, 7-, 10-, 15-year non-overlapping averages):
  - Coefficients on de jure openness decline almost monotonically from 0.008 to 0.004 when moving from 3-year to 15-year growth rates.
  - Absolute values of coefficients on FDI+equity and debt liabilities are larger at shorter horizons and decline with horizon length.
  - Coefficients less statistically significant at longer horizons (fewer observations), but effects remain economically/statistically significant up to ten years.
- Alternative specifications (difference-in-difference variants and other de jure indices):
  - De jure openness associated with higher TFP growth for two of the three de jure measures in DiD-like specifications.
  - Chinn-Ito index and Edwards measures yield broadly similar results; Bekaert-Harvey indicator positive but not always statistically significant.
- Alternative TFP measures and capital shares:
  - Using alternative capital share parameters (Gollin; Bernanke-Gurkaynak) preserves main findings: de jure openness significance and positive FDI+equity, negative debt coefficients.
- Outlier and additional control checks:
  - Exchange rate regime (Reinhart and Rogoff “fine” 15 categories) and terms-of-trade volatility add little to key coefficients.
  - Dropping observations >2 standard deviations or using Hadi (1994) multivariate outlier detection typically drops <2 percent of panel observations and preserves main results.
  - Results robust across sample splits (industrial vs. non-industrial), alternative measures, and TFP constructions.

### VIII. Policy-relevant implications and concluding remarks
- Main empirical conclusions:
  - Strong evidence that de jure capital account openness is associated with higher medium-term TFP growth.
  - De facto financial integration (stock of external liabilities to GDP) is not correlated with TFP growth overall.
  - Disaggregated liabilities show FDI and equity inflows contribute to TFP growth; debt inflows have the opposite effect.
  - Negative effect of external debt liabilities on TFP is partially attenuated by stronger financial markets and institutional quality.
- Policy implications:
  - Capital account liberalization (de jure openness) may contribute to productivity gains, particularly when combined with complementary domestic reforms.
  - Policies that attract FDI and portfolio equity are more likely to generate positive productivity spillovers than reliance on external debt accumulation.
  - Domestic thresholds matter: financial deepening (private credit development) and institutional quality mitigate adverse debt effects; countries may need to attain certain levels to reap full integration benefits.
  - Caution regarding debt-driven integration: large external debt stocks in countries with weak financial systems and institutions pose risks to TFP and macroeconomic stability.
- Suggested research agenda:
  - Delineate channels (technological spillovers, competition, corporate governance) through which financial openness boosts productivity.
  - Resolve why de jure openness delivers TFP benefits while de facto openness does not, noting de facto openness may relate more to GDP growth benefits.
  - Further examine domestic financial and institutional thresholds that raise marginal benefits of integration.

*Content based on the provided excerpt from _wp08242 (IMF Working Paper).*

### 1. Sample Statistics: Median Values by Country Group .........................................................21

### 1. Sample Statistics: Median Values by Country Group .........................................................21

### Major sections (chapter/section headings and page numbers)
- 1. Sample Statistics: Median Values by Country Group .........................................................21
- 2. Financial Openness and TFP Growth: Cross-Section Regressions .....................................22
- 3. Financial Openness and TFP Growth: Panel Regressions ...................................................23
- 4. Does the Composition of Exernal Liabilities Matter?  ........................................................24
- 5. Alternative Measures of TFP ...............................................................................................25
- 6. Alternative Measures of Capital Account Openness ...........................................................26
- 7. Financial Openness and TFP Growth in Non-Industrial Countries .....................................27
- 8. Is There a Threshold Level of Financial Integration? .........................................................28
- 9. Effects on TFP Growth at Different Horizons .....................................................................29
- 10. Financial Openness and TFP Growth: Difference-in-Differences.....................................30

### Figures
- Figure 1. Growth Accounting for More and Less Financially Open Economies................................32
- Figure 2. Growth Accounting for More and Less Financially Open Economies................................33
- Figure 3. Growth Accounting for More and Less Financially Open Economies................................34

### Tables
- Table 1a. Summary Statistics.............................................................................................................31

*Source: _wp08242 - 1. Sample Statistics: Median Values by Country Group .........................................................21 (PDF).*

### References..............................................................................................................

### _wp08242 - References..............................................................................................................

### I. Introduction — Purpose and main findings
- Central question: whether financial openness yields growth benefits; focus shifted from output growth to productivity (TFP) growth.
- The paper analyzes the relationship between financial openness and TFP growth using multiple measures of openness and an extensive cross-country dataset.
- Key empirical findings:
  - Economies with more open capital accounts generally have higher TFP growth.
  - Formal econometric analysis suggests de jure capital account openness has a causal positive effect on TFP growth even after controlling for standard growth determinants and endogeneity.
  - Overall de facto financial integration (gross stocks of foreign assets and liabilities relative to GDP) does not appear to matter for TFP growth in aggregate.
  - Disaggregating de facto integration by liability type yields:
    - FDI and portfolio equity liabilities boost TFP growth.
    - Debt liabilities are negatively correlated with TFP growth.
  - The negative relationship between external debt liabilities and TFP growth is partially attenuated in economies with better-developed financial markets and better institutional quality.

### II. Theoretical motivation and relevance
- Rationale for focusing on TFP:
  - TFP growth is emphasized as the main driver of long-term per capita income growth in recent literature.
  - Direct welfare/growth gains from capital mobility may be small; benefits are likely reflected in TFP via indirect channels (financial sector development, institutional improvements, technology transfers).
- Channels highlighted:
  - FDI: technology transfers, managerial expertise, imitation, skill acquisition, competition, vertical (backward/forward) spillovers.
  - Portfolio equity: deepening of domestic financial markets, corporate governance improvements.
  - Debt: financing benefits but possible adverse effects (volatility, moral hazard, inefficient allocation) that can harm aggregate efficiency.

### III. Methodology and data
- Empirical framework:
  - Dynamic panel regression framework using non-overlapping ten-year growth rates to capture sustained effects.
  - Address endogeneity and reverse causality concerns using Blundell and Bond (1998) system GMM with instrument-limiting refinements.
- Dataset and sample:
  - Annual data over the period 1966–2005 for 67 countries—21 industrial and 46 developing.
  - Primary data sources include Penn World Tables Version 6.2 and databases from the World Bank and IMF; all data in constant (2000) international prices.
- TFP measurement and growth accounting:
  - Cobb-Douglas production function: Y = A K^α (H L)^(1−α) (presentation preserved as in source).
  - Capital stock constructed using standard capital formation equation with an annual depreciation rate of 6 percent.
  - Human capital estimated via a Mincerian function with a Mincerian return parameter of 0.085 for each additional year of schooling using Barro and Lee (2000) and extrapolation after 2000.
  - Growth decomposition of output per worker into contributions from TFP growth, capital deepening, and human capital accumulation.
  - Baseline parameter α (capital share) assumed to be one-third; alternative capital share measures are considered in robustness checks.
- Measures of financial openness:
  - De jure benchmark: binary indicator from IMF’s AREAER (Schindler, 2007) — 1 when capital account is open, 0 otherwise.
  - De facto benchmark: ratio of gross stocks of external liabilities to GDP (from Lane and Milesi-Ferretti, 2006) and alternative splits into assets/liabilities and by type (FDI+equity vs. debt).
- Control variables:
  - Trade openness, changes in terms of trade, institutional quality (sum of corruption, law and order, and bureaucratic quality indices from ICRG; range 0 to 18), and financial sector development (private sector credit to GDP).

### IV. Stylized facts (descriptive results)
- Time periods analyzed: full sample 1966-2005 and sub-periods 1966-1985 and 1986-2005 (mid-1980s as break-point).
- Grouping for descriptive comparisons:
  - More financially open (MFO) vs. less financially open (LFO) economies, split at the cross-sectional median of financial openness (average over full sample).
- Descriptive patterns:
  - Cross-sectional medians: contribution of TFP growth to per-worker output growth is larger in MFO economies.
  - Over 1986-2005 (globalization period), MFO economies enjoyed faster productivity growth; LFO economies relied more on physical and human capital accumulation.
  - Average output growth between MFO and LFO was rather similar during the globalization period, despite differences in TFP contribution shares.
  - Median TFP growth typically higher in MFO economies compared to LFO economies over 1986-2005; mixed evidence depending on de jure vs. de facto classification.

### V. Regression results — main econometric findings
- Cross-section regressions:
  - Little evidence that de jure or de facto measures of financial integration matter for TFP growth in simple cross sections.
  - Only three variables consistently matter for TFP growth in the basic cross-section: convergence term (initial TFP), population growth, and institutional quality; trade openness and financial depth initially do not.
- Dynamic panel (ten-year averages) and fixed effects:
  - De jure capital account openness is significantly positively associated with TFP growth in fixed effects specifications.
  - De facto integration measures (total assets, total liabilities, sum) generally do not matter in FE cross-specifications.
- System GMM (addressing endogeneity):
  - De jure capital account openness continues to be positively related to TFP growth after controlling for endogeneity.
  - Economies with an open capital account have, over a ten-year horizon, annual TFP growth about 0.11-0.15 percentage points higher than economies with extensive capital controls.
- Composition of liabilities (de facto disaggregation):
  - Splitting gross external liabilities into (i) FDI+portfolio equity liabilities and (ii) debt liabilities yields:
    - Strong evidence FDI and equity liabilities boost TFP growth.
    - Debt liabilities reduce TFP growth.
  - Quantitative associations (system GMM):
    - A 10 percentage point increase in the ratio of FDI and equity liabilities to GDP is associated with about a 0.4 percentage points increase in annual TFP growth over a ten-year period.
    - A 10 percentage point increase in the ratio of debt liabilities to GDP is associated with TFP growth lower by about 0.2 percentage points.
- Mechanisms and interpretation:
  - Equity-like inflows (FDI and portfolio equity) are more likely to convey indirect benefits: technology spillovers, corporate governance improvements, and financial deepening.
  - Debt inflows are more volatile/procyclical and can exacerbate inefficient allocation or moral hazard, particularly where institutions and financial supervision are weak.

### VI. Interaction effects, thresholds, and heterogeneity
- Interaction with financial development (private sector credit to GDP):
  - Positive interaction between private sector credit and stock of debt liabilities implies well-developed financial markets attenuate the negative impact of debt inflows on TFP growth.
  - Calculated threshold: level of financial development beyond which the marginal effect of increases in external debt stock on TFP growth becomes positive corresponds to a credit to GDP ratio of nearly 150 percent (based on coefficients: debt liabilities coefficient −0.00383 and interaction coefficient 0.00261).
  - Sample mean of private sector credit to GDP in 1996-2005: 0.57 (standard deviation: 0.44).
- Interaction with institutional quality:
  - Better institutional quality reduces the negative impact of debt liabilities on TFP growth.
  - Improvements in institutional quality also reduce the estimated effects of FDI and portfolio equity liabilities on TFP growth; however, even at the highest institutional quality in the sample, marginal effect of FDI+equity remains positive and debt remains negative.
- Non-industrial country subsample:
  - When restricting to non-industrial countries, the coefficient on de jure openness is larger (and significant in GMM), FDI+equity coefficient positive but not always significant, and debt liabilities remain negative and strongly significant.
  - For many non-industrial countries, debt historically dominated inflows; negative debt effects are pronounced in this group.

### VII. Extensions and robustness checks
- Alternative TFP measures:
  - Recalculated TFP using alternative capital share parameters from Gollin and Bernanke-Gurkaynak; main results preserved:
    - De jure openness significant with total liabilities/GDP in some specifications.
    - FDI+equity positive association with TFP growth; debt coefficient negative (significant in FE specifications).
- Alternative de jure openness measures:
  - Chinn-Ito index (principal components of four restriction categories) yields broadly similar results: de jure openness significant when combined with total liabilities/GDP (GMM coefficient significant at the 11 percent level).
  - Other measures (e.g., Edwards) give essentially the same outcomes.
  - Equity market liberalization (binary indicator for stock market opening) is not significantly correlated with TFP growth once total liabilities are controlled for.
- Sensitivity to sample splits, time horizons, other controls and outliers:
  - Results robust across various checks including alternative country samples (industrial vs. non-industrial), alternative measures of openness, and different TFP constructions.
  - Key persistent regularities: de jure openness positively linked to TFP growth; composition of inflows matters—FDI/equity beneficial, debt potentially harmful absent strong institutions/financial depth.

### VIII. Policy-relevant implications distilled from results
- De jure capital account openness is associated with higher TFP growth; policy reforms that liberalize the capital account may contribute to productivity gains, particularly when paired with complementary reforms.
- Composition of inflows matters:
  - Policies that attract FDI and portfolio equity are more likely to generate positive productivity spillovers than policies that rely on external debt accumulation.
- Importance of domestic thresholds:
  - Financial deepening (private credit development) and institutional quality mitigate the adverse effects of external debt on TFP; countries may need to attain certain levels of financial and institutional development to reap full benefits of integration.
- Caution on debt-driven integration:
  - Large stocks of external debt accumulated by countries with under-developed financial systems and weak institutions pose risks to TFP and macroeconomic stability.

*Italicized source attribution: Content based on the provided excerpt from _wp08242 - References (IMF working paper content).*

### Section V, we discussed the possibility of other thresholds based on levels of financial and

### _wp08242 - Section V, we discussed the possibility of other thresholds based on levels of financial and

### Thresholds based on level of financial integration
- Hypothesis: benefits of financial integration may not be proportional to integration level and may appear only after a certain level of integration is achieved.
- Empirical approach: sample divided into above-median and below-median levels of financial integration (based on gross stocks of external liabilities to GDP); regressions run separately for these two groups (results reported in Table 8).
- Key findings:
  - For More Financially Open (MFO) economies:
    - Positive coefficient on de jure capital account openness (first two columns).
    - Positive coefficient on the stock of FDI and equity liabilities.
    - Negative coefficient on debt liabilities.
  - For Less Financially Open (LFO) economies:
    - These coefficients are much smaller and not statistically significant.
  - Robustness: very similar results when using the de jure capital account openness variable to distinguish MFO and LFO economies.
- Contextual shares of external liabilities:
  - In 2000-04, debt accounted for about 52 percent of gross external liabilities of emerging markets, while FDI accounted for 37 percent. Portfolio equity liabilities accounted for most of the remainder.
  - In 1980-84, the corresponding shares for debt and FDI were 85 percent and 14 percent, respectively.
- Commodity-exporter control:
  - Inclusion of a dummy for commodity-exporting countries and interactions with stocks of external liabilities made little difference to key results.

### Different time horizons
- Baseline used non-overlapping ten-year growth rates to avoid short-term/business cycle effects; results suggest persistent effects of financial openness on TFP growth.
- Re-estimation over multiple horizons (fixed effects specification from first column of Table 4); results reported for non-overlapping 3-year, 5-year, 7-year, 10-year and 15-year periods (Table 9).
- Coefficient patterns:
  - Coefficients on de jure capital account openness decline almost monotonically from 0.008 to 0.004 when moving from 3-year growth rates to 15-year growth rates.
  - Absolute values of coefficients on FDI plus equity liabilities and on debt liabilities are larger at shorter horizons and decline as horizon lengthens.
  - Coefficients become less statistically significant at longer horizons (fewer observations).
- Conclusion: effects of financial openness on TFP growth tend to wear off over time but remain economically and statistically significant at horizons of up to ten years.

### Alternative specification
- Comparison with Bonfiglioli (2008) and Henry (2007): they use a variant interpreted as a difference-in-difference specification (country and time fixed effects remove country-specific and time-specific variation).
- Table 10 estimates this specification using three measures of de jure capital account openness and three measures of de facto financial openness.
- Main result preserved:
  - An open capital account is associated with higher TFP growth for two of the three de jure measures.
  - The Bekaert-Harvey indicator: positive coefficients but not quite statistically significant.
  - Edwards (2007) measure produced results in line with the IMF and Chinn-Ito measures.
- Interpretation: results broadly supportive of the core conclusion that de jure openness is associated with higher TFP growth.

### Other controls and outliers
- Exchange rate regime:
  - Controlled using Reinhart and Rogoff (2004) “fine” classification with 15 categories.
  - Once controlling for financial openness, exchange rate regime has little additional influence on TFP growth and only a marginal effect on key coefficients.
  - Terms of trade volatility control also did not matter.
- Outlier sensitivity checks:
  - Eliminated observations with financial openness values more than two standard deviations from full sample means in two ways:
    - Dropping specific country-period observations only.
    - Dropping a country altogether if any of its observations had to be dropped.
  - Number of observations dropped was typically less than 2 percent of the full sample of panel data.
  - Re-estimation with these smaller samples preserved the main baseline results.
  - Hadi (1994) multivariate outlier detection applied; eliminating such outliers made little difference to key results.

### Concluding remarks
- Main empirical conclusions:
  - Strong evidence that financial openness, measured by de jure capital account openness, is associated with higher medium-term TFP growth.
  - Results robust to attempts to address endogeneity and reverse causality; authors view the relationship as potentially causal, though subtle.
  - De facto financial integration (stock of external liabilities to GDP) is not correlated with TFP growth overall.
  - Disaggregation of external liabilities reveals:
    - FDI and equity inflows (cumulated over decade-long periods) contribute to TFP growth.
    - Debt inflows have the opposite effect.
    - Negative effect of external debt liabilities on TFP is partially attenuated in economies with better-developed financial markets and better institutional quality.
- Possible mechanisms explaining stronger TFP effects than output effects:
  - Timing differences in adjustment of TFP and output; new general-purpose technologies may slow short-run output growth while raising TFP.
  - Reallocation effects: financial openness alters entry/exit, favoring more productive firms and raising aggregate TFP without significant employment gains.
  - Adjustment costs that delay TFP gains translating into output growth; longer horizons may reveal growth effects as capital stock adjusts.
- Research agenda and policy implications:
  - Need to delineate channels (technological spillovers, increased competition, improved corporate governance) through which financial openness boosts productivity.
  - Need to resolve why de jure openness delivers TFP benefits while de facto openness does not, noting that de facto openness seems more tied to GDP growth benefits.
  - Domestic financial and institutional development matter for productivity gains from openness; higher levels of integration appear to raise marginal benefits of additional integration.
  - Overall implication: capital account liberalization, especially when associated with FDI and portfolio equity flows, warrants reconsideration as it can lead to significant gains in efficiency and TFP growth.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08242.pdf*

### 2.  A de facto measure (the ratio of the stock of external liabilities to GDP) is used to define MFO and LFO economies f

### _wp08242 - 2.  A de facto measure (the ratio of the stock of external liabilities to GDP) is used to define MFO and LFO economies f

### Methodology and definitions
- MFO and LFO economies are defined using:
  - A de facto measure: the ratio of the stock of external liabilities to GDP for each sub-period (1966-1985, 1986-2005).
  - A de jure measure (Schindler, 2007) for the full sample (1966-2005) and for each sub-period (1966-1985, 1986-2005).
- Dependent variable in regressions: growth rate of TFP (total factor productivity) averaged over each period (full sample 1966-2005 or ten-year panels).
- External position components:
  - Total liabilities and assets refer to gross external liabilities and assets.
  - FDI and equity liabilities are the sum of gross FDI and gross portfolio equity liabilities.
  - Debt liabilities are gross external debt liabilities, including sovereign and portfolio debt.
- Estimation methods reported: OLS, FE, System GMM, Difference-in-Differences. Robust standard errors reported in brackets. Significance indicators: *, **, *** at the 10%, 5% and 1% levels.

### Key cross-section findings (Table 2: Dependent variable -- TFP growth; full sample 1966-2005)
- Initial TFP (in logs): coefficients range from -0.01883*** to -0.01849*** with standard errors between [0.00197] and [0.00211].
- Trade Openness (% GDP): coefficients near zero (e.g., -0.00001, 0.00007) with standard errors [0.00007]–[0.00008].
- Terms of Trade (% Change): positive and significant in many specifications, e.g., 0.00129** [0.00063], 0.00132** [0.00061].
- Population Growth: negative and significant across specifications, e.g., -0.00449*** [0.00124] to -0.00480*** [0.00127].
- Private Sector Credit (% GDP): small positive coefficients, sometimes significant at 10% (e.g., 0.00007* [0.00004]).
- Institutional Quality: positive and significant in many specifications, e.g., 0.00067** [0.00032] to 0.00071** [0.00033].
- Capital Account Openness (de jure): not significant in Table 2 (e.g., -0.00396 [0.00292]).
- Total Liabilities (% GDP), Total Assets (% GDP), Total Liabilities + Assets (% GDP): coefficients near zero and not significant (e.g., -0.00004 [0.00003]).
- R-squared: range 0.643 to 0.662. Observations: 67. OLS estimator.

### Panel regressions and dynamic evidence (Table 3: ten-year panel)
- Initial TFP (in logs): strongly negative and significant in FE/System GMM; examples:
  - FE: -0.61287*** [0.08458] to -0.61490*** [0.08493].
  - System GMM (alternative columns): -0.38540** [0.16861] to -0.23962* [0.12245].
- Trade Openness (% GDP): positive and significant in FE (e.g., 0.00498** [0.00215]) but smaller and not always significant in some GMM columns (e.g., 0.00175 [0.00209]).
- Private Sector Credit (% GDP): positive and often significant, e.g., 0.00116** [0.00054]; stronger in some GMM columns (e.g., 0.00251** [0.00102] to 0.00311*** [0.00100]).
- Capital Account Openness (de jure): positive and significant in many specifications, e.g., 0.07373** [0.03547]; larger in some GMM columns, e.g., 0.15476** [0.06056].
- Total Liabilities (% GDP), Total Assets (% GDP), Total Liabilities + Assets (% GDP): coefficients close to zero and not robustly significant (e.g., Total Liabilities -0.00017 [0.00037]).
- R squared: around 0.674–0.676. Countries: 67. Observations: 252. Specification tests: Hansen Test p-values reported (e.g., 0.211 to 0.464); 2nd Order Correlation p-values ~0.168–0.178. Number of instruments: 18–20.

### Composition of external liabilities (Table 4 and related)
- FDI & Equity Liabilities (% GDP): generally positive and significant:
  - Examples: 0.00201*** [0.00066], 0.00379** [0.00161], 0.00607*** [0.00220], 0.00695*** [0.00207].
- Debt Liabilities (% GDP): generally negative and significant:
  - Examples: -0.00178** [0.00069], -0.00247** [0.00096], -0.00383*** [0.00117], -0.00378*** [0.00087].
- Interaction terms reported (Institutional Quality * FDI & Equity Liabilities; Institutional Quality * Debt Liabilities; Private Sector Credit * FDI & Equity Liabilities; Private Sector Credit * Debt Liabilities) with varying coefficients and statistical significance (see table for exact values).
- R squared: reported values around 0.702–0.715. Countries: 67. Observations: 248 (in many columns). Specification tests: Hansen Test p-values (e.g., 0.470, 0.849, 0.295); 2nd Order Correlation (e.g., 0.126, 0.253, 0.248); Number of instruments: 23–26.

### Robustness: alternative TFP measures and capital account openness metrics (Tables 5 and 6)
- Alternative capital share parameters (Gollin 2002; Bernanke and Gurkaynak 2002) used to compute TFP produce broadly similar results.
- Examples from Table 5:
  - FDI & Equity Liabilities (% GDP): positive and significant in several specifications, e.g., 0.00198*** [0.00067], 0.00492** [0.00206].
  - Debt Liabilities (% GDP): negative and significant in several specifications, e.g., -0.00175** [0.00071].
- Alternative de jure measures of capital account openness (Chinn-Ito; Bekaert-Harvey) produce:
  - Capital Account Openness (de jure): coefficients vary; examples: 0.02895** [0.01308], 0.03059 [0.01861], 0.04532 [0.03849], 0.09231 [0.08015].
- FDI & Equity Liabilities remain positive and significant across alternative measures (e.g., 0.00195*** [0.00068], 0.00446*** [0.00134]).
- Debt Liabilities remain negative and significant in many specifications (e.g., -0.00172** [0.00070], -0.00230** [0.00091]).

### Non-industrial countries (Table 7)
- Sample: Non-industrial countries (Countries = 46).
- Total Liabilities (% GDP) and Debt Liabilities (% GDP): negative and significant in several specifications:
  - Total Liabilities: -0.00312** [0.00133], -0.00566*** [0.00198].
  - Debt Liabilities: -0.00315** [0.00129], -0.00602*** [0.00177].
- FDI & Equity Liabilities (% GDP): small or insignificant in some columns (e.g., 0.00001 [0.00271], 0.00419 [0.00560]).
- Capital Account Openness (de jure): mixed significance; example 0.20021** [0.08287] in a System GMM column.

### Thresholds and heterogeneity (Table 8)
- Threshold analysis comparing MFO vs LFO subsamples (MFO, LFO; Countries ~33–34) reports:
  - Capital Account Openness (de jure): sometimes strongly positive (e.g., 0.12139** [0.04834], 0.24200*** [0.08750]).
  - FDI & Equity Liabilities (% GDP): positive and significant in some columns (e.g., 0.00233*** [0.00078], 0.00515** [0.00226]); negative or insignificant in others.
  - Debt Liabilities (% GDP): negative and sometimes significant (e.g., -0.00158** [0.00072], -0.00360* [0.00179]).
- Samples and observations vary: Countries 33–34; Observations range 121–129. R squared examples: 0.735, 0.780, 0.660, 0.656.

### Horizons and persistence (Table 9)
- Dependent variable: average annual TFP growth over horizons of 3, 5, 7, 10, 15 years.
- Capital Account Openness (de jure): coefficients decline with horizon and become less significant:
  - 3 years: 0.00775* [0.00397]
  - 5 years: 0.00716* [0.00381]
  - 7 years: 0.00677* [0.00347]
  - 10 years: 0.00525 [0.00385]
  - 15 years: 0.00414 [0.00425]
- FDI & Equity Liabilities (% GDP): positive and significant at all horizons reported:
  - 3 years: 0.00027*** [0.00007]
  - 5 years: 0.00023*** [0.00006]
  - 7 years: 0.00017*** [0.00005]
  - 10 years: 0.00020*** [0.00007]
  - 15 years: 0.00015** [0.00006]
- Debt Liabilities (% GDP): negative and significant at all horizons, magnitude declining with horizon:
  - 3 years: -0.00024*** [0.00007]
  - 5 years: -0.00021*** [0.00006]
  - 7 years: -0.00018*** [0.00006]
  - 10 years: -0.00018** [0.00007]
  - 15 years: -0.00013* [0.00006]
- R-squared rises with horizon: 0.359 (3 years), 0.526 (5 years), 0.533 (7 years), 0.702 (10 years), 0.763 (15 years).
- Observations: 742 (3 years), 487 (5 years), 362 (7 years), 248 (10 years), 185 (15 years).

### Difference-in-differences and levels (Table 10)
- Dependent variable: TFP (in logs), ten-year panel; specifications include IMF, Chinn-Ito, Bekaert-Harvey measures.
- Trade Openness (% GDP): consistently positive and significant across specifications, e.g., 0.00629** [0.00241], 0.00688** [0.00261].
- Private Sector Credit (% GDP): coefficients positive, some significant at 10%/5% in certain columns, e.g., 0.00187* [0.00097], 0.00230** [0.00105].
- Capital Account Openness (de jure): positive and significant in many DiD specifications with larger magnitudes in some de jure measures:
  - Examples: 0.15967** [0.04915], 0.15778** [0.04765], 0.04553** [0.01729], 0.04691** [0.01712].
- Total Liabilities (% GDP), Total Assets (% GDP), Total Liabilities + Assets (% GDP): coefficients negative but not consistently significant (e.g., Total Liabilities -0.00094 [0.00091]).
- R squared: high, e.g., 0.971–0.972 in many columns. Countries ~66–67. Observations ~183–188.

### Growth accounting and visual evidence (Figures 1–3)
- Figures present growth accounting decompositions for More Financially Open (MFO) and Less Financially Open (LFO) economies under:
  - De facto measure of financial integration (ratio of stock of external liabilities to GDP).
  - De jure measure of capital account openness (Schindler, 2007).
  - Constant sample comparisons for pre-globalization (1966-1985) and globalization (1986-2005).
- Labels in figures: Real GDP per worker, TFP contribution, K/Y Contribution, H Contribution.
- Notes emphasize:
  - De facto measure used to define MFO and LFO economies in Figures 1a, 2.
  - De jure measure (Schindler, 2007) used in Figures 1b, 3.
  - Pre-globalization: 1966-1985; Globalization: 1986-2005.

### Summary statistics (Table A1; exact values highlighted)
- TFP (% change): Mean 0.39; Std dev 1.85 (Full Sample, 1966-2005). Other reported means and std devs across subsamples are present in table.
- Trade Openness (% GDP): Mean 28.74; Std dev 14.23 (Full Sample, 1966-2005).
- Population Growth: Mean 1.73; Std dev 1.04 (Full Sample).
- Terms of Trade (% change): Mean -0.36; Std dev 3.43 (Full Sample).
- Volatility of Terms of Trade: Mean 9.74; Std dev 7.80 (Full Sample).
- Private Sector Credit (% GDP): Mean 42.62; Std dev 34.32 (Full Sample).
- Institutional Quality: Mean 9.37; Std dev 4.07 (Full Sample).
- De Jure IMF *: Mean 0.32; Std dev 0.42 (Full Sample).
- Chinn-Ito: Mean 0.13; Std dev 1.42 (Full Sample).
- Edwards: Mean 58.27; Std dev 24.16 (Full Sample).
- De Facto: Total Assets & Liabilities (% GDP): Mean 117.38; Std dev 120.54 (Full Sample).
- Assets (% GDP): Mean 41.76; Std dev 64.59 (Full Sample).
- FDI & Equity (% GDP): Mean 9.65; Std dev 22.68 (Full Sample).
- Debt (% GDP): Mean 24.14; Std dev 44.27 (Full Sample).
- Liabilities: Mean 75.62; Std dev 62.10 (Full Sample).
- Liabilities FDI & Equity (% GDP): Mean 21.19; Std dev 28.34 (Full Sample).
- Liabilities Debt (% GDP): Mean 53.50; Std dev 39.98 (Full Sample).

*Italic source attribution: IMF Working Paper ("_wp08242 - 2.  A de facto measure (the ratio of the stock of external liabilities to GDP) is used to define MFO and LFO economies f")*

### References

### References

### Financial globalization, capital account liberalization, and integration
- Aizenman, Joshua, and Ilan Noy, 2008, “Endogenous Financial and Trade Openness,” Review of International Economics, forthcoming.
- Aizenman, Joshua, Brian Pinto and Arthur Radziwill, 2008, “Sources for Financing Domestic Capital--Is Foreign Saving a Viable Option for Developing Countries?” Journal of International Money and Finance, forthcoming.
- Aoki, Kosuke, Gianluca Benigno and Nobuhiro Kiyotaki, 2005, “Adjusting to Capital Account Liberalization,” Working Paper, London School of Economics.
- Bekaert, Geert, Campbell R. Harvey and Christian Lundblad, 2005, “Does Financial Liberalization Spur Economic Growth,” Journal of Financial Economics, Vol. 77, pp. 3-55.
- Edwards, Sebastian, 2007, “Capital Controls, Sudden Stops and Current Account Reversals,” in International Capital Flows, ed. Sebastian Edwards (Chicago, IL: University of Chicago Press).
- Fischer, Stanley, 1998, “Capital Account Liberalization and the Role of the IMF,” in Should the IMF Pursue Capital-Account Convertibility? Essays in International Finance 207, (Princeton, NJ: Princeton University), pp. 1–10.
- Forbes, Kristin J., 2007, “The Microeconomic Evidence on Capital Controls: No Free Lunch,” in International Capital Flows, ed. Sebastian Edwards (Chicago, IL: University of Chicago Press).
- Henry, Peter B., 2007, “Capital Account Liberalization: Theory, Evidence and Speculation” Journal of Economic Literature, Vol. 45, pp. 887-935.
- Henry, Peter B., and Diego Sasson, 2008, “Capital Account Liberalization, Real Wages, and Productivity,” NBER Working Paper No. 13880.
- Kose, M. Ayhan, Eswar Prasad, Kenneth Rogoff and Shang-Jin Wei, 2008, “Financial Globalization: A Reappraisal,” IMF Staff Papers, forthcoming.
- Kose, M. Ayhan, Eswar Prasad, and Ashley Taylor, 2008, “Thresholds in the Process of Financial Integration,” manuscript in preparation.
- Kose, M. Ayhan, Eswar Prasad, and Marco Terrones, 2008, “Does Openness to International Financial Flows Contribute to Productivity Growth?” IZA Discussion Paper No. 3634.
- Lane, Philip R., and Gian Maria Milesi-Ferretti, 2006, “The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004,” IMF Working Paper 06/69.
- Obstfeld, Maurice, 2008, “International Finance and Growth in Developing Countries: What Have We Learned?” Manuscript, University of California, Berkeley.
- Prasad, Eswar, and Raghuram Rajan, 2008, “A Pragmatic Approach to Capital Account Liberalization,” Journal of Economic Perspectives, forthcoming.
- Rodrik, Dani, 1998, “Who Needs Capital-Account Convertibility?” Essays in International Finance, No. 207 (Princeton, NJ: Princeton University Press).
- Schindler, Martin, 2007, “Measuring Financial Integration: A Disaggregated Dataset of Capital Account Restrictions,” IMF Working Paper, forthcoming.
- Stiglitz, Joseph, 2004, “Capital-Market Liberalization, Globalization, and the IMF,” Oxford Review of Economic Policy, Vol. 20, No. 1, pp. 57–71.
- Wei, Shang-Jin, and Zhiwei Zhang, 2007, “Collateral Damage: Exchange Controls and International Trade,” IMF Working Paper 07/08 (Washington: International Monetary Fund).

### Foreign direct investment, multinational enterprises, and spillovers
- Borensztein, Eduardo, Jose De Gregorio, and Jong Wha-Lee, 1998, “How Does Foreign Direct Investment Affect Economic Growth,” Journal of International Economics, Vol. 45, No. 1, pp. 115-135.
- Choe, Jong Il, 2003, “Do Foreign Direct Investment and Gross Domestic Investment Promote Economic Growth?” Review of Development Economics, Vol. 7, No. 1, pp. 44–57.
- Görg, Holger, and David Greenaway, 2004, “Much Ado about Nothing? Do Domestic Firms Really Benefit from Foreign Direct Investment?” World Bank Research Observer, Vol. 19, No. 2, pp. 171–97.
- Haskell, Jonathan E., Sonia C. Pereira, and Matthew J. Slaughter, 2007, “Does Inward Foreign Direct Investment Boost the Productivity of Domestic Firms?” Review of Economics and Statistics, Vol. 89, No. 3, pp. 482-496.
- Javorcik, Beata S., 2004, “Does Foreign Direct Investment Increase the Productivity of Domestic Firms? In Search of Spillovers through Backward Linkages,” American Economic Review, Vol. 94, No. 3, pp. 605–27.
- Lipsey, Robert E., and Fredrik Sjöholm, 2005, “The Impact of Inward FDI on Host Countries: Why Such Different Answers?” in Does Foreign Investment Promote Development? eds. Theodore H. Moran, Edward M. Graham, and Magnus Blomström (Washington, DC: Institute for International Economics Center for Global Development).
- Razin, Assaf, Efraim Sadka, and Hui Tong, 2005, “Bilateral FDI Flows: Threshold Barriers and Productivity Shocks,” NBER Working Paper No. 11639.
- Xu, Bin, 2000, “Multinational Enterprises, Technology Diffusion, and Host Country Productivity Growth,” Journal of Development Economics, Vol. 62, No. 2, pp. 477–93.

### Financial development, markets, and firm-level effects
- Bekaert, Geert, and Campbell R. Harvey, 2000, “Foreign Speculators and Emerging Equity Markets,” Journal of Finance, Vol. 55, No. 2, pp. 565–613.
- Benhabib, Jess, and Mark M. Spiegel, 2000, “The Role of Financial Development in Growth and Investment,” Journal of Economic Growth, Vol. 5, pp. 341-360.
- Greenwood, Jeremy, and Boyan Jovanovic, 1990, “Financial Development, Growth and the Distribution of Income,” Journal of Political Economy, Vol. 98, pp. 1076-1107.
- Henry, Peter B., 2000, “Stock Market Liberalization, Economic Reform, and Emerging Market Equity Prices,” Journal of Finance, Vol. 55, No. 2, pp. 529–64.
- Mitton, Todd, 2006, “Stock Market Liberalization and Operating Performances at the Firm Level,” Journal of Financial Economics, Vol. 81, pp. 625-647.
- Mishkin, Frederic S., 2006, The Next Great Globalization: How Disadvantaged Nations Can Harness Their Financial Systems to Get Rich (Princeton, N.J.: Princeton University Press).
- Gourinchas, Pierre-Olivier, and Olivier Jeanne, 2006, “The Elusive Gains from International Financial Integration,” Review of Economic Studies, Vol. 73, No. 3, pp. 715-741.
- Gourinchas, Pierre-Olivier, and Olivier Jeanne, 2007, “Capital Flows to Developing Countries: The Allocation Puzzle,” NBER Working Paper No. 13602.
- Levchenko, Andrei A., Romain Ranciere, and Matthias Thoenig, 2008, “Growth and Risk at the Industry Level: The Real Effects of Financial Liberalization,” Journal of Development Economics, forthcoming.
- Bonfiglioli, Alessandra, 2007, “Financial Integration, Productivity and Capital Accumulation,” Manuscript (Barcelona, Spain: Universitat Pompeu Fabra).

### Economic growth, productivity, and development theory
- Aghion Philippe, and Peter Howitt, 1998, “Capital Accumulation and Innovation as Complementary Factors in Long-Run Growth,” Journal of Economic Growth, Vol. 3, pp. 111-130.
- Alcala, Francisco, and Antonio Ciccone, 2004, “Trade and Productivity,” Quarterly Journal of Economics, Vol. 119, No. 2, pp. 613-646.
- Barro, Robert and Jong-Wha Lee, 2000, “International Data on Education Attainment: Updates and Implications,” NBER Working Paper No. 7911.
- Caselli, Francesco, 2005, “Accounting for Cross-Country Income Differences.” in Handbook of Economic Growth, eds. Philippe Aghion and Steven Durlauf (The Netherlands: Elsevier Science), pp. 679-741.
- Caselli, Francesco, and James Feyrer, 2007, “The Marginal Product of Capital,” Quarterly Journal of Economics, Vol. 122, No. 2, pp. 535-568.
- Easterly, William, and Ross Levine, 2001, “It’s Not Factor Accumulation: Stylized Facts and Growth Models,” World Bank Economic Review, Vol. 15, pp. 177–219.
- Gollin, Douglas, 2002, “Getting Income Shares Right,” Journal of Political Economy, Vol. 110, No. 2, pp. 458-474.
- Hall, Robert E., and Charles I. Jones, 1999, “Why Do Some Countries Produce So Much More Output Per Worker Than Others?” Quarterly Journal of Economics, Vol. 114, No. 1, pp. 83-116.
- Klenow, Peter, and Andres Rodriguez-Clare, 2005, “Externalities and Growth,” in the Handbook of Economic Growth, eds. Philippe Aghion and Steven Durlauf (The Netherlands: Elsevier Science), pp. 817-861.
- Parente, Stephen and Edward Prescott, 2005, “A Unified Theory of the Evolution of International Income Levels,” in the Handbook of Economic Growth, eds. Philippe Aghion and Steven Durlauf (The Netherlands: Elsevier Science), pp. 1371-1416.
- Prasad, Eswar, Raghuram Rajan and Arvind Subramanian, 2007, “Foreign Capital and Economic Growth,” Brookings Papers on Economic Activity, Vol. 2007:1, pp. 153-209.
- Jones, Benjamin F., and Benjamin A. Olken, 2008, “The Anatomy of Start-Stop Growth,” Review of Economics and Statistics, forthcoming.

### Empirical methods, data sources, and robustness
- Arellano, Manuel, and Olympia Bover, 1995, “Another Look at the Instrumental Variable Estimation of Error-Correction Models,” Journal of Econometrics, Vol. 68, pp. 29-52.
- Blundell, Richard, and Stephen Bond, 1998, “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models,” Journal of Econometrics, Vol. 87, pp. 115-143.
- Bond, Stephen R., Anke Hoeffler and Jonathan Temple, 2001, “GMM Estimation of Empirical Growth Models,” CEPR Working Paper No. 3048 (Centre for Economic Policy Research: London, UK).
- Hadi, Ali S., 1994, “A Modification of a Method for the Detection of Outliers in Multivariate Samples,” Journal of the Royal Statistical Society, Series (B), Vol. 56, pp. 393-396.
- Heston, Alan, Robert Summers and Bettina Aten, Penn World Tables Version 6.2, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, September 2006.
- Roodman, David, 2007, “A Short Note on the Theme of Too Many Instruments,” Working Paper No. 125, Center for Global Development, Washington, DC.

### Trade, exchange rates, and related macro topics
- Bhagwati, Jagdish, 1998, “The Capital Myth: The Difference between Trade in Widgets and Dollars,” Foreign Affairs, Vol. 7, No. 3, pp. 7-12.
- Reinhart, Carmen M., and Kenneth S. Rogoff, 2004, “The Modern History of Exchange Rate Arrangements: A Reinterpretation,” Quarterly Journal of Economics, Vol. 119, No. 1, pp. 1-48.
- Summers, Lawrence H, 2000, “International Financial Crises: Causes, Prevention, and Cures.” American Economic Review, Vol. 90, No. 2, pp. 1–16.
- Collins, Susan M., 2007, “Comments on “Financial Globalization, Growth, and Volatility in Developing Countries,” by Eswar Prasad, Kenneth Rogoff, Shang-Jin Wei, and M. Ayhan Kose,” in Globalization and Poverty, ed. Ann Harrison (Chicago, IL: University of Chicago Press).
- Johnson, Simon, and Todd Mitton, 2003, “Cronyism and Capital Controls: Evidence from Malaysia,” Journal of Financial Economics, Vol. 67, No. 2 (February), pp. 351–82.
- Forbes, Kristin J., 2007, “The Microeconomic Evidence on Capital Controls: No Free Lunch,” in International Capital Flows, ed. Sebastian Edwards (Chicago, IL: University of Chicago Press).

*Source: _wp08242 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08242.pdf_