## _wp12242

## Source details

**Canonical URL:** [_wp12242](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12242.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12242.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12242.pdf.json)

---

### Abstract and headline findings
- Investigates whether better information about the macroeconomic environment increases capital inflows (portfolio and foreign direct investment (FDI)).
- Information asymmetries proxied by compliance with the IMF's Special Data Dissemination Standard (SDDS).
- SDDS subscription increased FDI inflows by about 60 percent (statistically significant and robust).
- No significant aggregate effect of SDDS subscription on portfolio flows in the reported models.
- Evidence of aversion against political and macroeconomic risk as determinants of portfolio and FDI flows.
- Uses a non-parametric test for spatial correlation in residuals; finds no significant spatial dependence.

### Empirical framework, identification and data
- Estimation sample: generally covers N=55 countries between 1989 and 2008; unbalanced panel (actual observations < 55×20).
- Main specification: log-linear static panel with two-way fixed effects:
  - y^j_it = Ψ_it θ^j + SDDS_it λ^j_SDDS + η^j_t + α^j_i + ε^j_it
  - y^j_it = logarithm of capital flow of type j (FDI or portfolio) to country i in year t.
  - SDDS_it = 1 if country i meets SDDS specification in year t; 0 otherwise.
  - Year dummies η_t control for global/source-country effects; country fixed effects α_i absorb time-invariant country characteristics.
- Data sources:
  - Capital flows: IMF International Financial Statistics (IFS) balance of payments; where IFS not available before 1993, use IMF WEO data and include WEO data dummy.
  - Controls: WEO (GDP, GDP growth, investment rate, PPP), ICRG (political risk), IFS (money market rate, exchange rates), Chinn and Ito index (capital account openness), Barro and Lee (years of schooling), World Bank WDI (high-tech exports), OECD (patents).
- Measurement choices:
  - FDI: host enterprise threshold 10 percent ownership rule.
  - Flows expressed in constant USD and natural logarithm taken.
  - Use HAC (“cluster-robust”) standard errors.

### Risk measures, exchange rate volatility and SDDS timing
- Political risk: ICRG political risk ratings, annual averages, range 0 (high risk) to 100 (least risk); lowest de facto rating in sample = 56.
- Exchange rate volatility (national currency per SDR):
  - Exrtvol_it = sum_{m=t(6/12)}^{m=t−1(7/12)} (e_m − e_{m−1})^2 / e_{m−1}
  - Mean exchange rate volatility in sample = 0.0026 (Std. Dev. 0.024).
- SDDS historical details in data:
  - SDDS launched March, 1996.
  - 1999 is the first year with observed 1-values for some countries in the sample.
  - SDDS dummy mean = 0.40 (Std. Dev. 0.49).

### Main empirical results — FDI determinants (preferred specifications)
- SDDS effect on FDI (Table 1 reported coefficients; SDDS coefficients and standard errors):
  - model(2a) SDDS = 0.4462*** (0.1522)
  - model(3a) SDDS = 0.4760*** (0.1355)
  - model(4a) SDDS = 0.4590*** (0.1310)
- Authors' marginal-effect interpretation: SDDS increases FDI inflows by 56.2 to 61 percent.
- Selected control coefficients from preferred models (coefficient (standard error)):
  - ln(GDP): 0.7724*** (0.0995) model(1a); 0.7732*** (0.0952) model(2a); 0.7116*** (0.0795) model(3a); 0.8287*** (0.0550) model(4a).
  - GDP growth: 4.1104*** (0.9197) model(1a); 4.4241*** (0.8887) model(2a); 2.9280*** (0.9967) model(3a); 3.8167*** (1.2328) model(4a).
  - investment rate (-1): 3.1872*** (0.9126) model(1a); 3.4468*** (0.8909) model(2a); 2.1197* (1.1974) model(3a); 2.0197** (1.0252) model(4a).
  - capital account open: 0.1445** (0.0646) model(1a); 0.1299** (0.0625) model(2a); 0.1399** (0.0677) model(3a); 0.1377** (0.0554) model(4a).
  - political risk: 0.1143** (0.0458) model(1a); 0.1140** (0.0473) model(2a); 0.0207 (0.0380) model(3a); 0.0021 (0.0240) model(4a).
  - exchange rate volatility: -2.9418*** (0.7488) model(1a); -2.8234*** (0.7428) model(2a); -3.3692*** (0.8635) model(3a); -3.4416*** (1.0731) model(4a).
  - exchange rate volatility (-1): -0.7471 (0.7864) model(1a); -0.7105 (0.7611) model(2a); -1.4820** (0.5685) model(3a); -1.1573 (0.7046) model(4a).
  - interest rate (-1): 0.0000 (0.0000) model(1a); 0.0000 (0.0000) model(2a); -0.0002*** (0.0001) model(3a); -0.0003*** (0.0001) model(4a).
  - real exchange rate (-1): -0.0001 (0.0001) model(1a); -0.0001 (0.0001) model(2a); -0.0003* (0.0001) model(3a); -0.0003*** (0.0001) model(4a).
  - trade share: 0.1873 (0.2047) model(1a); 0.0990 (0.1899) model(2a); -0.1835 (0.2890) model(3a); 0.4382*** (0.1227) model(4a).
  - yrs. of schooling: -0.1155 (0.1383) model(2a); -0.0926 (0.1322) model(3a).
  - export unit value (-1): -0.0098** (0.0042) model(2a); -0.0043 (0.0036) model(3a).
  - WEO data dummy: 0.8398* (0.4916) model(1a); 0.8663* (0.5097) model(2a); -0.2562 (0.4287) model(3a); -0.6221 (0.4193) model(4a).
- Tests and interpretations:
  - Joint insignificance of both lags of exchange rate volatility in FDI models rejected at 1 percent in model(2a) and (3a).
  - Real exchange rate devaluation has negative impact on FDI in these specifications.
  - De jure capital account openness effect robustly positive and significant.
  - Human capital (aggregate years of schooling) not significant; sectoral technology measures mixed.

### Main empirical results — Portfolio determinants
- SDDS effect on portfolio flows (Table 2 reported coefficients):
  - model(1b) SDDS = -0.2079 (0.2185)
  - model(2b) SDDS = -0.0545 (0.2425)
  - model(3b) SDDS = 0.0713 (0.2022)
  - No statistically significant impact of SDDS on portfolio inflows.
- Selected control coefficients from portfolio models (coefficient (standard error)):
  - ln(GDP) (-1): 1.0519*** (0.1024) model(1b); 1.0516*** (0.1064) model(2b); 1.2999*** (0.2473) model(3b); 1.1935*** (0.0924) model(4b).
  - GDP growth: 3.2752 (2.3892) model(1b); 3.2010 (2.4053) model(2b); 6.1644*** (2.1640) model(3b); 4.4890* (2.3619) model(4b).
  - investment rate (-1): 2.9249** (1.3292) model(1b); 2.7624** (1.3441) model(2b); 2.3474 (2.2172) model(3b); -0.0970 (1.4085) model(4b).
  - political risk: 0.1107*** (0.0413) model(1b); 0.1221*** (0.0398) model(2b); 0.1121** (0.0512) model(3b); -0.1451*** (0.0381) model(4b).
  - exchange rate volatility: -0.6470 (1.6459) model(1b); -0.6689 (1.6538) model(2b); -1.3807 (1.4529) model(3b); -2.5401** (1.0597) model(4b).
  - exchange rate volatility (-1): -5.4158 (5.5546) model(1b); -5.2558 (5.6120) model(2b); -22.1873** (10.7241) model(3b); -22.5561** (8.9123) model(4b).
  - interest rate (-1): 0.0000 (0.0000) model(1b); 0.0000 (0.0000) model(2b); 0.0007** (0.0003) model(3b); 0.0007** (0.0003) model(4b).
  - # of patents (-1): -0.0000 (0.0000) model(1b); -0.0001** (0.0000) model(2b).
  - WEO data dummy: 0.4811* (0.2723) model(1b); 0.4806* (0.2759) model(2b); 0.4611 (0.5525) model(3b); 1.0629 (0.7039) model(4b).
- Tests and interpretations:
  - Can reject null that both lags of exchange rate volatility have no impact on portfolio flows at 10 percent in model(3b) but not in model(2a).
  - Portfolio investment responds positively to interest rate spreads in model(3b).
  - No consistent statistically significant effect of de jure financial liberalization on portfolio inflows across fixed-effects specifications.
  - Portfolio investment reacts more elastically to changes in current market size than FDI.

### Robustness, timing, heterogeneity and diagnostics
- Fixed effects preferred; Hausman test p-value = 0.000 (rejects random effects consistency).
- Residual diagnostics (preferred model (3a)):
  - QQ and PP plots and kernel density show no strong deviation from normality.
- Spatial interdependencies:
  - Non-parametric Conley-type approach with smoothing spline; estimated spatial correlation spline is essentially horizontal at zero.
  - Cannot reject null of no spatial correlation in residuals; no correction for spatial dependence needed.
- Timing of SDDS effect:
  - Subscriber-specific year dummies: capital inflows increase considerably in first four years after subscription; effect less robust later but overall suggests increased inflows concentrated early.
  - Period-specific dummies: most increased inflows occur in first years after subscription; effect decays afterwards but suggests slightly higher FDI inflows at later periods.
- Selection and early-subscriber exclusion:
  - Excluding early subscribers yields parameter estimate = 0.50 (sample of 19 countries) and remains statistically significant.
  - Excluding 1998 subscribers yields parameter estimate = 0.45 and highly statistically significant.
- Parameter heterogeneity by income groups (World Bank classification 1987):
  - High income countries: SDDS parameter = 0.69 (reported as 0.69 / 0.6905 in table 4).
  - Upper medium: 0.20.
  - Lower medium: 0.22.
  - Extended model LR test statistic = 13.24 with 2 degrees of freedom; model with separate high-income parameter outperforms alternatives.
  - Interpretation: effect driven by high-income countries (industrialized countries experience an even larger impact of 100 percent as reported in table 4); other countries without strong fundamentals show comparable impact somewhat above 20 percent.
- Recursive regressions (sample truncations 2000–2005): estimated impact smaller than approximately 0.45 but never falls short of 0.16.

### Policy implications and interpretation
- More accurate and timely public dissemination of macroeconomic data (SDDS compliance) is associated with substantially higher FDI inflows (about 60 percent in conditional estimates).
- Effect concentrated in the first years after compliance; suggests information triggers an adjustment process raising capital stock via initial inflows.
- Public information provision likely attracts more long-term oriented investment and can help develop productive resources in capital-scarce countries by attracting more stable FDI.
- Informational frictions may partly explain weak or non-robust growth benefits of financial openness; improving transparency is a potential tool to manage capital inflows and shift flows toward longer-term and less volatile forms.
- Distinction between FDI and portfolio flows is important for policy: FDI responds strongly to macro-data transparency; portfolio flows show mixed responses and may depend more on firm-level information and investor composition.

### Key descriptive statistics (selected values from Appendix B, Table 7)
- ln(FDI): Mean 22.04 (Std. Dev. 1.70) — Source IFS (WEO).
- ln(portfolio): Mean 22.10 (Std. Dev. 2.40) — Source IFS (WEO).
- SDDS dummy: Mean 0.40 (Std. Dev. 0.49) — Source IMF.
- ln(GDP): Mean 25.95 (Std. Dev. 1.79) — Source WEO.
- GDP growth: Mean 0.031 (Std. Dev. 0.036) — calculated from WEO.
- investment rate: Mean 0.233 (Std. Dev. 0.058) — calculated from WEO.
- capital account open: Mean 1.12 (Std. Dev. 1.40) — Source Chinn/Ito (2011).
- political risk (ICRG): Mean 65.89 (Std. Dev. 3.49).
- exchange rate volatility: Mean 0.0026 (Std. Dev. 0.024).
- interest rate (MMR−LIBOR spread): Mean 19.08 (Std. Dev. 231.37).
- real exchange rate (implied PPP, national currency per USD): Mean 85.90 (Std. Dev. 371.63).
- trade share: Mean 0.696 (Std. Dev. 0.601).
- yrs. of schooling: Mean 9.04 (Std. Dev. 11.99).
- high-tech exports: Mean 1.84e+10 (Std. Dev. 3.47e+10).
- export unit value: Mean 87.31 (Std. Dev. 20.59).
- # of patents: Mean 841.73 (Std. Dev. 2,592.6).
- WEO data dummy: Mean 0.121 (Std. Dev. 0.327).

- Sample of countries used in model (3a) includes: Algeria, Argentina, Australia, Austria, Brazil, Bulgaria, Canada, Chile, Colombia, Croatia, Czech Republic, Estonia, Finland, France, Germany, Greece, Hong Kong (China), Iceland, Indonesia, Ireland, Italy, Jamaica, Japan, Jordan, Korea, Rep., Kuwait, Latvia, Lithuania, Malaysia, Mexico, Moldova, Morocco, Netherlands, New Zealand, Norway, Pakistan, Peru, Philippines, Poland, Romania, Russian Federation, Singapore, Slovak Republic, Slovenia, South Africa, Spain, Sri Lanka, Switzerland, Thailand, Turkey, Ukraine, United Kingdom, United States, Uruguay, Venezuela.

*Source: IMF Working Paper WP/12/242 — “The Role of Risk and Information for International Capital Flows: New Evidence from the SDDS” (sections 1–9 as provided).*

### Section 1

### The Role of Risk and Information for International Capital Flows: New Evidence from the SDDS

### Abstract
- Investigates whether better information about the macroeconomic environment increases capital inflows (portfolio and foreign direct investment (FDI)).
- Information asymmetries are quantified by compliance with the IMF's Special Data Dissemination Standard (SDDS).
- For FDI: SDDS subscription increased inflows by about 60 percent (statistically significant and robust).
- Finds evidence of aversion against political and macroeconomic risk as determinants of portfolio and FDI flows.
- Uses a non-parametric test for spatial correlation in the residual of capital flows.
- JEL Classification Numbers: C33, F21, G14.

### Introduction — key empirical observations and scope
- SDDS established in 1996 with the aim of enhancing member countries’ access to the international capital market through improved macroeconomic data provision.
- Descriptive evidence: average levels of FDI inflows (relative to GDP) were higher for almost all subscriber countries after SDDS subscription than before; portfolio flows show a similar but less definite pattern.
  - Note: Unweighted average over the 5 years prior to and after subscription. Outliers were deleted to provide the graph on a meaningful scale.
- Main empirical finding highlighted: better data dissemination through SDDS increases FDI inflows by about 60 percent, but has no significant aggregate effect on portfolio flows.
- Also finds:
  - Evidence for macroeconomic risk-aversion for portfolio and FDI flows.
  - More robust evidence of political-risk aversion for portfolio flows.
- The paper contrasts FDI and portfolio flows and proposes new measures for productivity relevant to FDI literature.
- Considers spatial interdependencies; finds no evidence for significant spatial patterns in empirical models (in line with results implied by Baltagi et al. (2008)).

### Econometric approach, model and data (overview from Section III)
- Empirical framework: panel data econometric models linking capital inflows to information provision (SDDS) and risk measures.
- Variables and summary statistics are provided in appendix B (not reproduced here).
- Spatial interdependencies are examined using a non-parametric approach to avoid stringent assumptions about the underlying spatial process.

### Literature review — salient points and gaps addressed
- Prior work documents the role of information frictions in limiting cross-border diversification and shaping capital flows (French and Poterba (1991); Tesar and Werner (1995); Mody and Taylor (2003); Brennan and Cao (1997); Gelos and Wei (2005)).
- Measures used previously to proxy information (distance, telephone traffic, newspaper trade, immigrant stocks) may conflate information with other transaction-cost or proximity effects; these proxies risk omitted variable bias.
- Daude and Fratzscher (2008) and related studies find higher elasticity of FDI to information proxies than for other capital forms; however, their information proxies cover many types of transaction costs.
- SDDS offers a direct measure of macroeconomic data provision that allows more explicit identification of the information channel than many prior proxies.
- FDI literature: agglomeration and information-sharing channels identified (Wheeler and Mody (1992); Head et al. (1995); Blonigen et al. (2005); Kinoshita and Mody (2001); Bobonis and Shatz (2007); Davies et al. (2009); Harding and Javorcik (2011)). Some studies attribute treaty or institutional effects to reductions in uncertainty/information frictions.

### Main empirical claims and contributions
- Direct estimation of capital flows on SDDS-based information measurement provides stronger identification of the information channel than many previous studies.
- Core empirical results:
  - SDDS subscription → increase in FDI inflows of about 60 percent (statistically significant and robust).
  - No significant aggregate effect of SDDS subscription on portfolio flows in the models reported.
  - Political-risk aversion strongly influences portfolio flows; macroeconomic risk-aversion affects both portfolio and FDI flows.
  - Spatial correlation tests on residuals do not indicate significant spatial dependence in the estimated models.

### Structure and further material (as presented)
- The paper contains:
  - Section II: Literature review (summarized above).
  - Section III: Empirical model and data (variable list and summary statistics in appendix B).
  - Section IV: Main empirical results.
  - Section V: Robustness and further results.
  - Section VI: Discussion and conclusion.
  - Appendices: A (Information on SDDS), B (Sample, variables and descriptive statistics), References, Tables and Figures.

*Source: IMF Working Paper WP/12/242 — “The Role of Risk and Information for International Capital Flows: New Evidence from the SDDS” (Section 1).*

### Section 2

### III. INVESTMENT: MODEL AND DATA

### Empirical strategy and identification
- Focus: investors’ motives (supply side) toward host country effects; source-country fundamentals taken as externally given.
- Use of flow data (instead of stock data) to reduce autocorrelation and spurious regression risk and to limit reversed causality; follows Harding and Javorcik (2011).
- Data sample: generally covers N=55 countries between 1989 and 2008; unbalanced panel (actual observations < 55×20).
- Identification: two-way fixed effects (country and year) exploit within-country variation over time; requires no omitted variables that influence both capital inflows and SDDS compliance and no reverse causality from capital inflows to SDDS compliance.
- SDDS compliance variable: dummy that equals 1 in country i in year t if the country met the SDDS specifications by then; interpretation of the SDDS parameter as a difference-in-difference effect under model correctness.
- Controls for global (source-country) effects via year dummies; SDDS subscriptions are multilateral and most countries joined at a single point in time (1996), while meeting SDDS specifications usually occurs three to four years after subscription (predetermined main explanatory variable).

### Portfolio and foreign direct investment (FDI) definitions and data handling
- FDI classification rule: host-country enterprise in which a foreign investor owns 10 percent or more of the voting power is classified as a direct investment enterprise.
- Portfolio investment: residual category of cross-border transactions involving debt or equity securities (covers, but is not limited to, securities traded on financial markets).
- Economic implication: portfolio and FDI capture different investment concepts and may respond differently to explanatory variables (e.g., public provisioning of macroeconomic data may affect MNCs and portfolio investors differently).
- Capital flows data sources:
  - Primary: IMF International Financial Statistics (IFS) balance of payments data.
  - Where IFS data do not start before 1993, use IMF World Economic Outlook (WEO) data.
  - Correction: include a dummy variable equal to 1 if WEO data are used and 0 if IFS data are used.
  - Flows expressed in constant USD and natural logarithm is taken.

### Econometric model specification
- Log-linear static model with real capital flows (log) as dependent variable; use heteroscedasticity and autocorrelation consistent (HAC) variance estimator (Huber (1967), White (1980)) — referred to as “cluster-robust” standard errors.
- Model equation (as presented):
  - y^j_it = Ψ_it θ^j + SDDS_it λ^j_SDDS + η^j_t + α^j_i + ε^j_it
  - where y^j_it is the logarithm of capital flow of type j = 1,2 (either FDI or portfolio) to country i = 1,...,N in year t = 1,...,T; Ψ is a matrix of (up to) K−N−T−1 control variables; SDDS_it is the SDDS compliance dummy; η_t and α_i are time and country fixed effects respectively.
- Interpretation:
  - α_i = country-fixed effect interpreted as average inflow to country i over time.
  - η_t = time-fixed effect controls for overall global cross-border capital volume in year t (source-country effects, oil price, general trend).
- Logs are taken for GDP so it can be interpreted as an elasticity.

### Determinants of international investment flows (controls and measurement)
- Market size and potential:
  - Current market size: GDP data from WEO, natural logarithm of real value in USD.
  - Short-run growth (future market potential proxy): percentage change of real GDP per capita in national currency from WEO.
  - Long-run growth proxy: investment rate measured as gross capital formation at current national prices to GDP at current national prices (both from WEO).
- Competitiveness and technology measures:
  - Education: Barro and Lee (2010) dataset (average years of schooling).
  - Supplementary measures: high-tech exports (World Bank Development Indicators), total number of patents (OECD), export unit values (IMF WEO).
- Trade openness:
  - Trade share = (imports (including c.i.f.) + exports) in current USD (IFS) relative to GDP in current USD (WEO); used to proxy openness and potential FDI–trade interdependencies.
- Exchange rate effects:
  - Use implied PPP exchange rate (national currency per USD) from WEO (lagged by one year to mitigate reverse causality).
- Interest rate / financing costs:
  - Include the spread of the money market rate (MMR) over LIBOR in percent p.a. (both from IFS) to proxy for the interest rate; MMR defined as the rate at which banks lend to each other for short term.
  - Note distinction: low local interest rates favor MNC affiliates (access to local credit), while capital flows in Mundell–Fleming respond positively to domestic–foreign interest rate spread.
- Financial openness:
  - Control using Chinn and Ito (2006, 2008, 2011) index of capital account openness (available up to 2009); index is de jure, based on IMF Annual Report on Exchange Arrangements and Exchange Restrictions codified dummies; higher index = higher openness to cross-border financial transactions.
  - Acknowledge index includes current account restrictions; import restrictions can have effects similar to capital inflow controls (Jeanne, 2011).
- Omitted variables considerations:
  - Wages and taxes not included due to persistence and absorption by fixed effects; complex MNC responses to taxation and wage structures noted in literature.

### Risk: political and macroeconomic considerations
- Two broad risk categories considered: political risk (instability) and macroeconomic risk.
- Political risk:
  - Direct and indirect costs to investors (e.g., ownership shares in MNCs decline as political risk increases per Kesternich and Schnitzer (2010)).
  - Empirical studies (Wei (2000a,b); Papaioannou (2009); Daude and Fratzscher (2008)) find negative impacts of political risk on FDI inflows.
  - Daude and Fratzscher (2008): portfolio investment is more sensitive to institutional indicators and market openness than FDI; investor protection has a large effect on portfolio investment but not on FDI.

### Data construction and estimation choices
- Use of lags (e.g., one-year lag for PPP exchange rate) to avoid reversed causality since capital inflows affect exchange rates if they float.
- Note on FDI classification sensitivity: raising classification criterion (from 10 percent) to 20 or 50 percent has minor impact on measurement of US firms under foreign control (Graham and Krugman, 1989: 10).
- Use of first differences vs. flows: earlier studies sometimes use first difference of BEA stock data; this study uses flows to better capture responsiveness to policy changes in informational asymmetries.
- Robustness checks planned/tested include tests for omitted time-dependent variable problems.

*Source: _wp12242 - Section 2*

### Section 3

### _wp12242 - Section 3

### Risk measures and exchange rate volatility
- Political risk: ICRG political risk ratings used (monthly data averaged over one year). Ratings range from a high of 100 (least risk) to a low of 0 (highest risk); lowest de facto rating in the sample is 56.
- Exchange rate volatility measure (national currency per SDR): calculated using monthly IFS data as the sum of squared deviations from the expected exchange rate divided by last month’s exchange rate over the first 6 months of yeart and the last 6 months of yeart−1. Formal expression:
  - Exrtvol_it = sum_{m=t(6/12)}^{m=t−1(7/12)} (e_m − e_{m−1})^2 / e_{m−1} .
- Rationale:
  - Uses SDR denomination to avoid single reference currency effects.
  - Relies on Engel and West (2005) justification that short-run exchange rate fluctuations are nearly unpredictable, so today’s rate is a reasonable predictor for tomorrow’s.
  - Intuition: investors base decisions on past exchange rate volatility as an estimate of future volatility.

### Information disclosure: SDDS
- SDDS dummy definition:
  - SDDS_it = { 1, if country i meets SDDS specification in yeart; 0, else }.
- Historical note: SDDS launched March, 1996; Canada and the United States first met specifications on February 19, 1999. As of 2012, SDDS has 71 member countries listed in appendix A.
- Implementation detail: 1999 is the first year where 1-values are observed for at least some countries in the sample.
- Robustness: authors perform robustness checks with lagged SDDS subscription and dynamics (see section V in original).

### Spatial interdependencies and econometric treatment
- Concern: Distance may proxy many aspects beyond informational frictions; spatial interdependencies can bias covariance matrix if untreated.
- Methodology:
  - Apply Conley (1999) and Conley and Ligon (2002) approach adapted to cross-border capital flows.
  - Non-parametrically estimate correlation of Pearson-transformed error terms with respect to spatial distance.
  - Bootstrap a (pointwise) 90 percent confidence region for the null hypothesis of no spatial correlation.
  - Use a smoothing spline rather than a moving-average estimator for spatial correlation.
- Result: The analysis indicates there is no need to correct inference for spatial interdependencies.

### Main empirical results — FDI determinants (summary of Table 1 and accompanying text)
- Estimation setup:
  - Equation (1) estimated for FDI and portfolio investment.
  - Columns: baseline (no SDDS), baseline+SDDS, preferred specification with sectoral skill/technology measures, and random effects.
  - Fixed effects preferred; Hausman test p-value = 0.000 (rejects random effects consistency).
- SDDS effect on FDI:
  - SDDS coefficients in Table 1: model(1a) (no SDDS), model(2a) SDDS = 0.4462*** (0.1522), model(3a) SDDS = 0.4760*** (0.1355), model(4a) SDDS = 0.4590*** (0.1310).
  - Authors calculate conditional marginal effect in log-linear models: providing high-quality information under SDDS increases FDI inflows by 56.2 to 61 percent.
- Key control variable coefficients and notes (Table 1 excerpts):
  - ln(GDP): 0.7724*** (0.0995) in model(1a); 0.7732*** (0.0952) in model(2a); 0.7116*** (0.0795) in model(3a); 0.8287*** (0.0550) in model(4a).
  - GDP growth: 4.1104*** (0.9197) in model(1a); 4.4241*** (0.8887) in model(2a); 2.9280*** (0.9967) in model(3a); 3.8167*** (1.2328) in model(4a).
  - investment rate (-1): 3.1872*** (0.9126) in model(1a); 3.4468*** (0.8909) in model(2a); 2.1197* (1.1974) in model(3a); 2.0197** (1.0252) in model(4a).
  - capital account open: 0.1445** (0.0646) in model(1a); 0.1299** (0.0625) in model(2a); 0.1399** (0.0677) in model(3a); 0.1377** (0.0554) in model(4a).
  - political risk: 0.1143** (0.0458) model(1a); 0.1140** (0.0473) model(2a); 0.0207 (0.0380) model(3a); 0.0021 (0.0240) model(4a).
  - exchange rate volatility: -2.9418*** (0.7488) model(1a); -2.8234*** (0.7428) model(2a); -3.3692*** (0.8635) model(3a); -3.4416*** (1.0731) model(4a).
  - exchange rate volatility (-1): -0.7471 (0.7864) model(1a); -0.7105 (0.7611) model(2a); -1.4820** (0.5685) model(3a); -1.1573 (0.7046) model(4a).
  - interest rate (-1): 0.0000 (0.0000) model(1a); 0.0000 (0.0000) model(2a); -0.0002*** (0.0001) model(3a); -0.0003*** (0.0001) model(4a).
  - real exchange rate (-1): -0.0001 (0.0001) model(1a); -0.0001 (0.0001) model(2a); -0.0003* (0.0001) model(3a); -0.0003*** (0.0001) model(4a).
  - trade share: 0.1873 (0.2047) model(1a); 0.0990 (0.1899) model(2a); -0.1835 (0.2890) model(3a); 0.4382*** (0.1227) model(4a).
  - yrs. of schooling: -0.1155 (0.1383) in model(2a); -0.0926 (0.1322) in model(3a).
  - export unit value (-1): -0.0098** (0.0042) in model(2a); -0.0043 (0.0036) in model(3a).
  - # of patents (-1): 0.0000 (0.0000) in model(2a); 0.0000 (0.0000) in model(3a).
  - WEO data dummy: 0.8398* (0.4916) model(1a); 0.8663* (0.5097) model(2a); -0.2562 (0.4287) model(3a); -0.6221 (0.4193) model(4a).
- Statistical tests and interpretations:
  - Joint insignificance of both lags of exchange rate volatility in FDI models is rejected at the 1 percent level in model (2a) and (3a) (F-test).
  - The authors find a negative impact of a real exchange rate devaluation on FDI inflows (contrary to Blonigen (1997)), and discuss alternative channels (e.g., Dixit-Stiglitz preferences).
  - Human capital measures: years of schooling not significant and sign contrary to expectation; number of patents borderline weak significance (t-statistic 1.46) with expected sign.
  - De jure capital account openness effect robustly positive and significant; omitting it biases SDDS impact only modestly (without capital account openness, SDDS impact ~65 percent).

### Main empirical results — Portfolio determinants (summary of Table 2 and accompanying text)
- SDDS effect on portfolio flows:
  - SDDS coefficients in Table 2: model(1b) SDDS = -0.2079 (0.2185); model(2b) SDDS = -0.0545 (0.2425); model(3b) SDDS = 0.0713 (0.2022). No statistically significant impact of SDDS on portfolio inflows is found.
- Key control variable coefficients and notes (Table 2 excerpts):
  - ln(GDP) (-1): 1.0519*** (0.1024) model(1b); 1.0516*** (0.1064) model(2b); 1.2999*** (0.2473) model(3b); 1.1935*** (0.0924) model(4b).
  - GDP growth: 3.2752 (2.3892) model(1b); 3.2010 (2.4053) model(2b); 6.1644*** (2.1640) model(3b); 4.4890* (2.3619) model(4b).
  - investment rate (-1): 2.9249** (1.3292) model(1b); 2.7624** (1.3441) model(2b); 2.3474 (2.2172) model(3b); -0.0970 (1.4085) model(4b).
  - capital account open: 0.0198 (0.0799) model(1b); 0.0298 (0.0809) model(2b); -0.0552 (0.0853) model(3b); 0.1462*** (0.0559) model(4b).
  - political risk: 0.1107*** (0.0413) model(1b); 0.1221*** (0.0398) model(2b); 0.1121** (0.0512) model(3b); -0.1451*** (0.0381) model(4b).
  - exchange rate volatility: -0.6470 (1.6459) model(1b); -0.6689 (1.6538) model(2b); -1.3807 (1.4529) model(3b); -2.5401** (1.0597) model(4b).
  - exchange rate volatility (-1): -5.4158 (5.5546) model(1b); -5.2558 (5.6120) model(2b); -22.1873** (10.7241) model(3b); -22.5561** (8.9123) model(4b).
  - interest rate (-1): 0.0000 (0.0000) model(1b); 0.0000 (0.0000) model(2b); 0.0007** (0.0003) model(3b); 0.0007** (0.0003) model(4b).
  - # of patents (-1): -0.0000 (0.0000) model(1b); -0.0001** (0.0000) model(2b).
  - WEO data dummy: 0.4811* (0.2723) model(1b); 0.4806* (0.2759) model(2b); 0.4611 (0.5525) model(3b); 1.0629 (0.7039) model(4b).
- Statistical tests and interpretations:
  - Can reject null that both lags of exchange rate volatility have no impact on portfolio flows at the 10 percent level in model (3b) but not in model (2a).
  - Portfolio investment responds positive to interest rate spreads in model (3b).
  - No statistically significant impact of de jure financial liberalization on portfolio inflows found.
  - Portfolio investment shies away from political risk in fixed effects models; random effects model (4b) shows positive correlation, possibly due to omitted variable bias.
  - Portfolio investment reacts more elastically to changes in current market size than FDI.

### Comparative interpretations and implications
- Information (SDDS) matters strongly for FDI but not for portfolio flows in these specifications: SDDS associated with a 56.2 to 61 percent increase in FDI inflows conditional on controls, but no significant SDDS effect on portfolio inflows.
- Macroeconomic risk (exchange rate volatility) is detrimental to FDI (statistically significant negative coefficients) and has mixed but notable effects on portfolio flows (lags jointly relevant in some models).
- De jure capital account openness is a robust positive correlate of FDI; effect on portfolio flows is not consistently significant across fixed-effects specifications.
- Aggregate years of schooling may be a poor predictor of multinational investment; sectoral measures (patents, high-tech exports, export unit values) show mixed evidence, with patents borderline significant for FDI.
- Econometric robustness: fixed effects preferred; spatial correlation testing indicates no correction necessary; cluster-robust standard errors used.

*Source: _wp12242 - Section 3 (extracted content).*

### Section 4

### _wp12242 - Section 4

### Main estimate and residual diagnostics
- Estimated highly significant effect of SDDS subscription on FDI inflows with a parameter estimate around 0.45.
- QQ and PP plots and a kernel density estimate for residuals of preferred model (3a) show no specific pattern, lending support to the model specification and the assumption that the main test statistic follows a t-distribution.
- Kernel density and QQ/PP diagnostics presented (figures 2 and 3) indicate no strong deviation from normality in residuals.

### Omitted variables and time-trend tests
- To investigate potential omitted variable bias from country-specific time-varying factors, the authors estimate model (5):
  y_it = α_i + Ψ_it θ + δ_SDDS_t + γ_non−SDDS_t + ε_it
  and perform a Wald test of H0: δ_SDDS = γ_non−SDDS for FDI flows up until certain points in time.
- Table 3 results (different time trends between SDDS subscribers and non-subscribers?):
  - before year... 1996 | 1997 | 1998 | 1999
  - SDDS trend (ˆδ): -0.0764 | -0.0898 | -0.1173 | 0.0534
  - non-SDDS trend (ˆγ): -0.0468 | -0.0890 | -0.1295 | -0.0046
  - difference significant (p-val)?: 0.563 | 0.989 | 0.822 | 0.377
  - observations: 150 | 190 | 234 | 278
- Conclusion: Results do not provide evidence of an underlying time-dependent omitted process influencing both joining SDDS and FDI (i.e., no evidence for an omitted variable bias from these tests).
- Note: When control variables are excluded, p-values of the test become 0.0135 (1996), 0.0076 (1997), 0.0057 (1998), and 0.0052 (1999), indicating some power in finding an omitted variable bias in that specification.

### Selection, exclusion of early subscribers, and timing of effect
- Excluding early subscribers (bulk of 1996 subscribers) yields a parameter estimate equal to 0.50 (first column of table 4) and remains statistically significant; only 19 countries remain in this sample.
- When 1998 subscribers are excluded (in 1997, Portugal was the only subscribing country), the parameter estimate becomes 0.45 and is highly statistically significant.
- Investigating the timing: introduce SDDS-subscriber-specific year dummies (model (6)) to measure effect at each year before and after subscription; subscription year taken as reference year 0.
  - Findings from year-specific coefficients (left panel of figure 4): in the first four years after subscription, capital inflows considerably increase; impact is not as robust thereafter with negative estimates for years 5 and 7, but an overall picture suggesting increased inflows.
- Period-specific dummies (right panel of figure 4):
  - Period definitions: period 0 = three years prior to subscription; period 1 = year of subscription and two years thereafter; period 2 = 3 to 5 years after; period 3 = 6 to 8 years after; period 4 = 8 to 10 years after.
  - Results: Less volatile than year-specific estimates but same general pattern — most increased inflows occur in the first years after subscription, effect decays afterwards but suggests slightly higher FDI inflows at later periods.
- Theoretical interpretation: Increased information results in a permanently higher capital stock; the adjustment process operates through increased inflows initially, with replacement-induced flows remaining slightly higher after reaching new equilibrium.

### Parameter heterogeneity by income groups
- Parameter estimates by World Bank classification 1987:
  - High income countries: 0.69 (reported as 0.69 / 0.6905 in table 4)
  - Upper medium: 0.20
  - Lower medium: 0.22
- Model fit and tests:
  - Extended model with three income categories: LR test statistic 13.24 with 2 degrees of freedom (better fit by AIC/BIC and LR).
  - Comparing a model with one parameter for high income and another for all other countries versus the three-way model: LR statistic 0.01 with 1 degree of freedom (latter outperforms extended model along all three lines).
- Reported coefficients in Table 4 (parameter robustness):
  - SDDS (general): 0.5016** (0.1907) ; 0.4935** (0.2362) ; 0.2056 (0.1550)
  - SDDS (high inc): 0.6905*** (0.2340)
  - observations: 160 | 634 | 587
- Interpretation: Impact of SDDS on FDI inflows is driven by high-income countries; estimated parameter for other countries is positive and economically relevant (+23 percent) with a t-statistic 1.33 — indicating potential benefit for lower-income countries but highlighting that data provision alone is not sufficient without sound macroeconomic fundamentals.

### Stability checks and recursive regression
- Recursive regression results (section 5.3 of Hashimoto and Wacker, 2012): for samples truncated in the years 2000 - 2005 the impact is somewhat smaller than approximately 0.45, but it never falls short of 0.16.
- Implication: Estimated effect is stable over alternative sample truncations and remains positive.

### Spatial correlation in residuals
- Estimated spatial correlation pattern (smoothing spline, figure 5):
  - The estimated spline function is almost identically a horizontal line of zero-correlation.
  - Estimated spatial correlation does not reach a level beyond randomness; cannot reject the null hypothesis of no spatial correlation in the error terms.
- Conclusion: Inference in the models is not plagued by spatial correlation in the residuals.

*Source: _wp12242 - Section 4*

### Section 5

### _wp12242 - Section 5

### Main findings
- No substantive textual content or findings are present in the supplied content unit.

### Key statistics
- No numeric values or statistics are present in the supplied content unit.

### Policy recommendations
- No policy recommendations are present in the supplied content unit.

* _wp12242 - Section 5*

### Section 6

### _wp12242 - Section 6

### Verbatim bullets from source
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●

*_wp12242 - Section 6_*

### Section 7

### VI. DISCUSSION AND CONCLUSION (Section 7)

### A. Main findings
- Compliance with the IMF’s Special Data Dissemination Standard (SDDS) increases FDI inflows by about 60 percent.
- The most important impact occurs in the first years after subscription to (or compliance with) the SDDS.
- Industrialized countries experience an even larger impact of 100 percent (as reported in table 4).
- For other countries without strong fundamentals the comparable impact may be somewhat above 20 percent.
- In the longer run these countries can also catch up to the overall parameter of 60 percent, likely because financial markets’ monitoring of macroeconomic and financial data strengthens incentives to improve fundamentals.
- Portfolio and FDI flows both shy away from political and macroeconomic risk:
  - Political risk does not seem to matter much for FDI in the most advanced countries.
  - The aversion against macroeconomic risk in the form of exchange rate volatility is less robust for portfolio investment.

### B. Relation to other findings in the literature
- The main finding aligns with Daude and Fratzscher (2008): FDI is more responsive to information than other forms of capital flows.
- Information about the macro environment is probably more important for FDI; portfolio investment is likely more connected with firm-level information in the host economy.
- Portfolio investors are supposed to acquire information; large institutional investors may have incentives to acquire private information, reducing short-run arbitrage gains once information becomes public.
- Smaller investors with a longer-term perspective who cannot acquire information on their own may rely on public information.
- These two opposing public information effects on portfolio decisions (large investors acquiring private information vs. small investors relying on public information) might offset each other, resulting in an overall small and statistically not significant public-information impact on portfolio flows.
- This does not necessarily contradict Gelos and Wei (2005), who find emerging market portfolio funds respond positively to information:
  - Their sample captures only one third of the total portfolio flows to the relevant countries (p. 2989).
  - Their SDDS-comparable measure for macrodata opacity loses statistical significance when including fixed region effects (p. 3002).

### C. Further results
- The study empirically supports the distinction of portfolio flows from FDI flows in the balance of payments.
- The elasticity of portfolio investment toward current market size is significantly higher than for foreign direct investment.
- Portfolio and FDI flows show differing responses to the interest rate, highlighting that:
  - FDI is not simply a capital flow.
  - MNCs’ finance strategy for investments should play a more important role in microeconomic attempts to understand multinational firm behavior.

*Source: _wp12242 - Section 7*

### Section 8

### Section 8

### Main findings on information, FDI, and portfolio investment
- Empirical modeling should focus more specifically on know-how in sectors with comparative advantages when assessing (firm demand for) education and its link to investment.
- Spatial correlations do not play a significant role when FDI flows are estimated at an aggregate level using fixed effects.
- The study finds no significant overall effect of information on portfolio investment at the aggregate level, but this does not imply that the structure of portfolio investment is unaffected.
- Informational asymmetries may attract market makers and hence more short-term oriented portfolio flows, which can generally result in more volatility in capital markets (cf. Diamond and Verrecchia, 1991; Du and Wei, 2004) and can cause adverse externalities (cf. Bianchi, 2011).
- Public dissemination of more accurate and timely economic and financial data is a potentially important macroeconomic tool to manage capital inflows because it may attract more long-term oriented portfolio flows.
- Evidence strongly suggests that more FDI will be attracted by stronger macro fundamentals and by macroeconomic information; public dissemination of accurate economic data will initiate more stable and advantageous forms of capital inflows.

### Policy implications and expected macroeconomic effects
- Greater public dissemination of accurate economic data:
  - Should result in a more sustainable and healthy international investment position.
  - May help to lessen the degree of international balance of payments imbalances.
  - Would develop the productive resources of capital-scarce countries by attracting more stable FDI.
- Informational frictions may partly explain why growth effects of financial openness have been somewhat disappointing and generally far from being robust (cf. Jeanne et al., 2012, ch. 3 and 4); this channel requires further investigation.
- Managing capital inflows via improved information transparency is presented as a tool to shift flows toward longer-term and less volatile investment.

### Research perspectives and gaps
- The results highlight the need for further, more disaggregated research about the role of information for portfolio (and potentially other forms of) investment.
- The channel through which informational frictions affect the growth benefits of financial openness still needs detailed investigation.
- The study did not find a significant spatial pattern in the residuals of the random effects specification.

### Appendix A: SDDS information (summary)
- Table 5: List of SDDS subscribers (subscription year “subsc”, metadata posting year “meta”, first year meeting specification “spec”) is provided in the appendix.
- Table 6: SDDS Data Coverage — illustrative categories and components include:
  - Real Sector: GDP by categories; industrial / commodity production; (un)employment, wages; CPI, producer price index.
  - Fiscal Sector: General Government Operations (revenue, expenditure, financing); Central Government Operations; Central Government Debt (domestic / foreign by currency).
  - Financial Sector: Analytical Banking Accounts (money, credit); Analytical Central Bank Accounts (reserve money, domestic claims, external position); Interest Rates (government security rates); Stock Market (share price index).
  - External Sector: Balance of Payments (goods and services); Reserves; Merchandise Trade (exports and imports); International Investment Position; Exchange Rates (spot rates, 3- and 6-month forward markets); External Debt (debt of different sectors).
  - Population.
- Note: The appendix states this is an illustrative list and refers to the SDDS website for comprehensive information.

### Appendix B: Sample, variables and descriptive statistics (selected key statistics from Table 7)
- ln(FDI): Logarithm of real foreign direct investment inflows in USD — Mean 22.04 (Std. Dev. 1.70) — Source IFS (WEO).
- ln(portfolio): Logarithm of real foreign portfolio investment inflows in USD — Mean 22.10 (Std. Dev. 2.40) — Source IFS (WEO).
- SDDS: dummy variable equal 1 if countryimet ‘Special Data Dissemination Standard’ in yeartt — Mean 0.40 (Std. Dev. 0.49) — Source IMF.
- ln(GDP): Logarithm of real GDP in USD — Mean 25.95 (Std. Dev. 1.79) — Source WEO.
- GDP growth: percentage change of real GDP per capita — Mean 0.031 (Std. Dev. 0.036) — calculated from WEO.
- investment rate: gross capital formation / GDP — Mean 0.233 (Std. Dev. 0.058) — calculated from WEO.
- capital account open: Chinn and Ito (2006) index for capital account openness — Mean 1.12 (Std. Dev. 1.40) — Source Chinn/Ito (2011).
- political risk: political risk rating of International Country Risk Guide (ICRG, yearly average), 0 (high risk) - 100 — Mean 65.89 (Std. Dev. 3.49) — Source ICRG.
- exchange rate volatility: see equation(3) on page16 — Mean 0.0026 (Std. Dev. 0.024) — calculated from IFS.
- interest rate: spread of money market rate over LIBOR (both in percent p.a.) — Mean 19.08 (Std. Dev. 231.37) — calculated from IFS.
- real exchange rate: implied purchasing power parity exchange rate measured in national currency per USD — Mean 85.90 (Std. Dev. 371.63) — Source WEO.
- trade share: sum of imports and exports to/from the world in current USD / GDP in current USD — Mean 0.696 (Std. Dev. 0.601) — calculated from IFS, WEO.
- yrs. of schooling: Years of Schooling from Barro and Lee (2010) using a smoother — Mean 9.04 (Std. Dev. 11.99) — Source Barro/Lee (2010).
- high-tech exports: country’s share of high-tech exports in global high-tech exports at year t — Mean 1.84e+10 (Std. Dev. 3.47e+10) — calculated from WDI.
- export unit value: export unit value index — Mean 87.31 (Std. Dev. 20.59) — Source WEO.
- # of patents: total number of patents — Mean 841.73 (Std. Dev. 2,592.6) — Source OECD.
- WEO data dummy: dummy variable equal 1 if WEO data was used for dependent variable (instead of IFS data) — Mean 0.121 (Std. Dev. 0.327) — calculated.

- The appendix lists the sample of countries used in model (3a), including Algeria, Argentina, Australia, Austria, Brazil, Bulgaria, Canada, Chile, Colombia, Croatia, Czech Republic, Estonia, Finland, France, Germany, Greece, Hong Kong (China), Iceland, Indonesia, Ireland, Italy, Jamaica, Japan, Jordan, Korea, Rep., Kuwait, Latvia, Lithuania, Malaysia, Mexico, Moldova, Morocco, Netherlands, New Zealand, Norway, Pakistan, Peru, Philippines, Poland, Romania, Russian Federation, Singapore, Slovak Republic, Slovenia, South Africa, Spain, Sri Lanka, Switzerland, Thailand, Turkey, Ukraine, United Kingdom, United States, Uruguay, Venezuela.

*Source: _wp12242 - Section 8*

### Section 9

### _wp12242 - Section 9

### References on corruption, political risk, and taxation
- Javorcik, Beata S. and Shang-Jin Wei (2009), “Corruption and cross-border investment in emerging markets: Firm-level evidence.” Journal of International Money and Finance, 28, 605–624.
- Wei, Shang-Jin (2000a), “How taxing is corruption on international investors?” The Review of Economics and Statistics, 82, 1–11.
- Wei, Shang-Jin (2000b), “Local corruption and global capital flows.” Brookings Papers on Economic Activity, 31, 303–354.
- Kesternich, Iris and Monika Schnitzer (2010), “Who is afraid of political risk? Multinational firms and their choice of capital structure.” Journal of International Economics, 82, 208 – 218.
- Scholes, Myron and Mark Wolfson (1990), “The effects of changes in tax law on corporate reorganization activity.” Journal of Business, 63, S141–S164.
- Razin, Assaf, Efraim Sadka, and Chi-Wa Yuen (1998), “A pecking order of capital inflows and international tax principles.” Journal of International Economics, 44, 45–68.

### References on information, transparency, and informational asymmetries
- Mody, Ashoka, Efraim Sadka, and Assaf Razin (2003), “The role of information in driving FDI flows: Host-country transparency and source-country specialization.” IMF Working Papers 03/148, International Monetary Fund.
- Mody, Ashoka and Mark P. Taylor (2003), “International capital crunches: The time-varying role of informational asymmetries.” Discussion Papers 3757, C.E.P.R. Discussion Papers.
- Kinoshita, Yuko and Ashoka Mody (2001), “Private information for foreign investment in emerging economies.” Canadian Journal of Economics/Revue canadienne d’économique, 34, 448–464.
- Portes, Richard, Helene Rey, and Yonghyup Oh (2001), “Information and capital flows: The determinants of transactions in financial assets.” European Economic Review, 45, 783–796.
- van Nieuwerburgh, Stijn and Laura Veldkamp (2009), “Information immobility and the home bias puzzle.” Journal of Finance, 64(3), 1187–1215.
- Tesar, Linda L. and Ingrid M. Werner (1995), “Home bias and high turnover.” Journal of International Money and Finance, 14, 467 – 492.
- Warnock, Francis E. (2002), “Home bias and high turnover reconsidered.” Journal of International Money and Finance, 21, 795–805.
- Pagano, Marco, Ailsa A. Röell, and Josef Zechner (2002), “The geography of equity listing: Why do companies list abroad?” The Journal of Finance, 57, 2651–2694.

### References on exchange rates, volatility, and FDI
- Kiyota, Kozo and Shujiro Urata (2004), “Exchange rate, exchange rate volatility and foreign direct investment.” The World Economy, 27, 1501–1536.
- Pain, Nigel and Desirée van Welsum (2003), “Untying the Gordian Knot: The multiple links between exchange rates and foreign direct investment.” Journal of Common Market Studies, 41, 823–846.
- Schmidt, Christian and Udo Broll (2009), “Real exchange-rate uncertainty and us foreign direct investment: an empirical analysis.” Review of World Economics, 145, 513–530.
- Jeanne, Olivier (2011), “Capital account policies and the real exchange rate.” John Hopkins University, mimeo.
- Jeanne, Olivier, Arvind Subramanian, and John Williamson (2012), Who Needs to Open the Capital Account? Peterson Institute for International Econometrics, Washingon, DC.

### References on multinational firms, measurement, and financial markets
- Lipsey, Robert E. (2002), “Home and host country effects of FDI.” Working Paper No. 9293, National Bureau of Economics Research, Cambridge, MA.
- Lehmann, Alexander and Hyoung G. Kang (2004), “Multinational affiliates and local financial markets.” Working Paper 04/107, International Monetary Fund, Washington, DC, URL http://www.imf.org/.
- Wacker, Konstantin M. (2012), “Measurement and Determinants of Activities of Multinational Firms and Foreign Direct Investments.” mimeo.
- Milesi-Ferretti, Gian Maria and Philip R. Lane (2004), “International investment patterns.” IMF Working Papers 04/134, International Monetary Fund, URL http://ideas.repec.org/p/imf/imfwpa/04-134.html.
- Stehrer, Robert and Julia Woerz (2009), “‘Attract FDI!’ - A universal golden rule? Empirical evidence for OECD and selected non-OECD countries.” European Journal of Development Research, 21, 95–111.
- Papaioannou, Elias (2009), “What drives international financial flows? Politics, institutions and other determinants.” Journal of Development Economics, 88, 269 – 281.
- Portes, Richard and Hélène Rey (2005), “The determinants of cross-border equity flows.” Journal of International Economics, 65, 269 – 296.

### Methodology, econometrics, and theoretical foundations
- Maddala, Gangadharrao Soundalyarao (1971), “The use of variance components models in pooling cross section and time series data.” Econometrica, 39, 341–358.
- Stiglitz, Joseph E and Andrew Weiss (1981), “Credit rationing in markets with imperfect information.” American Economic Review, 71, 393–410.
- Vaizey, John (1964), “The residual factor and economic growth.” Study Group in the Economics of Education, Paris: OECD.
- Scholes, Myron and Mark Wolfson (1990), “The effects of changes in tax law on corporate reorganization activity.” Journal of Business, 63, S141–S164.
- White, Halbert (1980), “A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity.” Econometrica, 48, 817â
 ̆
A ̧S830.

*Source: _wp12242 - Section 9*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12242.pdf_