## _wp09221 - Appendix Table A1) comprises “low income” or “lower middle income” according to the

## Source details

**Canonical URL:** [_wp09221 - Appendix Table A1) comprises “low income” or “lower middle income” according to the](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09221.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09221.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09221.pdf.json)

---

### Data and sample
- Sample focuses on countries classified as “low income” or “lower middle income” by the World Bank, excluding emerging markets (China, Colombia, India, Indonesia, Pakistan and Thailand).
- Largest low income country set used (for trading partner calculations and regressions with standard fundamentals) includes 59 low- and lower middle-income countries.
- Estimation samples:
  - Current account regressions: unbalanced panel of non-overlapping four-year averages over the period 1981-2005 with 6 observations for most countries.
  - RER and NFA analyses: unbalanced panel of annual data covering 1980-2006.
- Summary statistics provided in Appendix Table A2; variable definitions and construction documented in the Appendix.
- Note: many variables are computed relative to trading partners; sample size for regressions varies by data availability.

### Medium-term current account — methodology
- Regressions include country fixed effects and standard determinants: fiscal balance, old age dependency ratio, population growth, initial net foreign asset position, oil trade balance, GDP per capita relative to the US, and real per capita GDP growth.
- New medium-term determinants (aid flows, domestic financial liberalization, capital account openness, institutions, shocks, demographics) are introduced progressively across specifications; preferred parsimonious results summarized in Columns 8 and 9.
- Estimation approach: robust panel regressions on four-year averages (1981-2005), significance codes: *** p<0.01, ** p<0.05, * p<0.1.

### Key empirical findings — current account (LICs)
- Fiscal policy and demographics:
  - Fiscal balance / GDP: positive and significant (examples: coefficients range from 0.2432** to 0.3966*** across columns; Table 1 and Table 2 show coefficients such as 0.3166**, 0.3486***).
  - Population growth (relative to trading partners): negative and often significant (examples: -1.6275**, -2.6210***, -1.8436***; Table 2 examples -1.4924*, -2.7107***).
- Terms of trade:
  - Positive income effect: terms of trade improvements raise the current account (examples: 0.0228*, 0.0345**, 0.0319***; log terms of trade examples 0.0289**, 0.0352***).
  - Export prices drive the effect more than import prices.
- Official financing (aggregate foreign aid):
  - Aggregate aid is significantly and negatively correlated with the current account.
  - Estimated effect: for any 1$ of aid, the current account deteriorates by about 0.1$ during the same period.
  - Aid components (separate regressions):
    - About 80 percent of grants are fully absorbed through an increase in net imports over a 4 years period.
    - Only 30 to 50 percent of concessional loans received are absorbed through net imports over the same horizon.
  - Concessional loans / GDP: negative and significant where included (examples: -0.3122*, -0.5565***; Table 2 examples -0.4185*, -0.5318***).
  - Net grants / GDP: mixed evidence (examples: 0.2108*, 0.2168*; Table 2 examples 0.1953*, 0.2328**).
- Domestic financial liberalization:
  - Associated with higher current accounts.
  - Estimated effect: shifting from complete financial repression to the sample average level of liberalization would lift the current account by 1.5 percent of GDP.
  - Coefficients in some columns: 0.0343**, 0.0326**, 0.0317**.
- Natural disasters and capital account openness:
  - Natural disasters (exogenous negative income shocks) affect the current account conditional on capital account openness.
  - In countries with a fully open capital account, such shocks have a negative impact on the current account.
  - Natural disasters are associated with current account improvements in countries with closed capital accounts.
  - Frequency of these shocks in the sample: 0.6 over the sample period.
  - Quantitative example: in a country with a fully liberalized capital account, a natural disaster occurring at average frequency of 0.6 during a 4 year period would be associated with a -0.7 percent of GDP deterioration of the current account.
  - Interaction natural disaster * capital account (initial): negative and significant (examples: -0.0605**, -0.0654***; Table 1 and Table 2 examples -0.0664***).
- Political instability:
  - Violent political transition associated with current account improvement (consistent with capital flight); examples: violent political transition coefficient 0.0287***, Table 2 examples 0.0265**, 0.0272***.
- Persistence and robustness:
  - Limited persistence of the current account beyond a 4 year horizon (lagged current account controlled in Column 1).
  - Trade reforms (average tariffs) show no additional explanatory power in tested specifications.
  - Results robust to inclusion of emerging markets classified as lower middle income (China, Colombia, India, Indonesia, Pakistan and Thailand) and to broader capital account liberalization measures.

### RER (real effective exchange rate) — methodology
- Panel unit root tests: variables are I(1) except natural shocks; panel cointegration tests reject null of no cointegration.
- Estimation: Dynamic Ordinary Least Squares (DOLS) with fixed effects; levels coefficients interpreted as long-run cointegrating relationships; changes plus one lead and one lag of changes included.
- Two REER measures used:
  - CPI-based IMF-INS REER (log).
  - Penn World Table-derived price relative turned into REER using IMF-INS weights (log).
- Controls: net foreign assets, productivity, government consumption, terms of trade, trade restrictions, LIC-specific variables (aid flows, capital account liberalization), demographics, price controls, and natural shocks.

### Key empirical findings — RER (LICs)
- Government consumption:
  - Associated with real exchange rate appreciation (examples: government consumption to GDP # coefficients 1.1259* to 2.0271***; Table 5 and Table 6 examples 2.0186**, 3.3371***).
- Terms of trade:
  - Improvements appreciate the real exchange rate (examples: terms of trade goods (log) 0.3458*** to 0.4353*** in Table 5; Table 6 examples 0.1634**, 0.1970***).
  - Effect primarily driven by export prices in decompositions (price of exports positive, price of imports weakly negative).
- Demographics:
  - Fertility associated with RER appreciation (example coefficients 0.1239***; fertility # mean in Table A2 2.7400).
- Net foreign assets:
  - Not significant for some REER specifications (possible debt relief expectations blur intertemporal role).
- Productivity:
  - Not significant with IMF-INS REER; negative with Penn World Table REER (examples: GDP per worker PWT (log) -0.4418***, -0.4266***).
- Aid inflows:
  - Aid (Roodman (2006)) associated in the long run with a more depreciated exchange rate (examples: Aid to GDP # -2.2405***, -1.3812***; Table 6 examples -3.2405***).
  - Interpretation: may indicate positive effect on nontradable-sector productivity relative to tradables; consistent with long-run depreciation via productivity channel.
- Capital account liberalization:
  - Associated with RER appreciation (examples: capital account liberalization # 0.3152***, 0.4910***; Table 6 examples 0.1746**, 0.3048**).
- Price distortions:
  - Presence of marketing boards (“max agric. price intervention”) tends to appreciate the RER (examples in Table 9: 0.1842***).
  - Black market premium positive and significant where included (example: 0.2547*** in Table 5; 0.4344*** in Table 6); inclusion halves sample size in those regressions.
- Speed of adjustment (error-correction):
  - Lagged error correction term coefficient ~0.2 in 1–4 lag specifications (half life ~ three years).
  - Alternative REER measure yields coefficient ~0.3 (half life ~2 years).
  - These speeds align with existing literature (e.g., Rogoff (1996)).

### Net foreign assets (NFA) — methodology
- Estimation: DOLS with fixed effects on net foreign assets to trade ratio (preferred to NFA to GDP).
- Panel unit root and cointegration tests reject null of no cointegration.
- Public debt data from Jaimovich and Panizza (2006); some splicing with external debt series when needed.
- Nicaragua dropped from NFA regressions due to extreme debt and NFA values.

### Key empirical findings — NFA (LICs)
- Domestic financial liberalization:
  - Associated with higher net foreign assets (examples: domestic financial liberalization # coefficients 2.0003***, 1.8758***, 2.2360*** in Table 10).
  - Interpretation: larger positive effect on aggregate savings than on aggregate investment.
- Capital account liberalization:
  - Negative relationship with net foreign assets in low income countries (examples: -1.4137***, -1.5802*** in Table 10).
  - By contrast, positive relationship in high income countries (Table 11 example for HICs & EM: 1.2703***).
- Institutions:
  - Better institutional characteristics associated with higher net foreign assets (example: constraint on executive coefficients 0.2280***, 0.2505***).
- Public debt:
  - Long-run effect: about half of the increase in public debt is reflected in a reduction of net foreign assets (examples: public debt to trade -0.4566***, -0.4840***; effect in the order of 0.5-0.8 noted).
- Demographics and income:
  - Higher share of dependent population associated with lower savings and net foreign assets (fertility # examples -0.3135***, -0.4913***).
  - Income per capita: positive association with net foreign assets in LICs (relative productivity (log) # examples 1.5867***, 1.7841***).
- Robustness:
  - Results hold with alternative NFA measures, alternative debt indicators, and alternative control sets (columns 3-7 of Table 10).
  - After controlling for determinants, terms of trade shocks, trade restrictions, and price controls do not matter for NFA (columns 8-10).

### Cross-sample differences — LICs vs HICs
- Variables with significantly different impacts between low-income and high-income countries include:
  - Fiscal balance: positive and significant in LICs (example 0.3224***), not significant in HICs (example -0.0811); Table 3b shows p-values for equality sometimes 0.06, 0.04, 0.02.
  - Old age dependency ratio: differential effects reported across samples.
  - Initial net foreign assets: negative coefficient in advanced countries consistent with intertemporal approach.
- Capital account liberalization:
  - Opposite sign by income group: negative association with NFA in LICs (example -1.4137***), positive in HICs (example 1.2703***); equality test p-value 0.000 (Table 11).
  - Interpretation: rich countries lend while poor countries borrow; allows borrowing against disasters in LICs.
- Public debt:
  - Debt appears insignificant for the high income sample (consistent with range 0 to 0.2 found by Lane and Milesi-Ferretti (2001)); in LICs public debt has larger negative association with NFA.
- Speed of adjustment for NFA:
  - Error correction regressions deliver lagged error correction term coefficients around 0.25.
  - Interpretation: shocks to the NFA gap have a half life of about two and a half years.

### Policy-relevant conclusions and implications
- Common fundamentals coherently explain medium- to long-term patterns of:
  - Real effective exchange rate,
  - Current account,
  - Net external asset position in low income countries.
- Distinctive determinants for LICs beyond standard variables:
  - aid flows (grants and concessional loans),
  - domestic financial liberalization,
  - removal of capital account controls,
  - quality of institutions,
  - shocks (terms of trade, natural disasters),
  - demographic measures.
- Policy distortions and reforms:
  - Domestic financial reforms associated with:
    - improvement of the current account,
    - improvement of net foreign assets;
    - suggests reforms boost saving more than investment.
  - Capital account liberalization associated with:
    - lower current accounts in LICs (facilitates borrowing against disasters),
    - more appreciated long-run RER (consistent with persistent net capital inflows and higher absorption).
  - Quality of institutions positively associated with larger external wealth in the long run.
- Shocks:
  - Positive terms of trade shocks tend to improve the current account and appreciate the RER (mainly via export-price channel).
  - Natural disasters tend to improve the current account except when capital accounts are open.
  - Effect of income shocks on the current account may depend on initial NFA position (interaction evidence supports portfolio-composition predictions).
- Aid and external financing:
  - Aid increases imports progressively; part of aid flows saved in the short run.
  - In the long run, increased aid associated with RER deterioration (long-run depreciation), consistent with productivity-channel Dutch disease mechanisms (Torvik, 2001).

### Data limitations and scope for further research
- Priority: improve quality and extent of data coverage (notably black market premium, structural and financing indicators) which are missing for numerous countries.
- Need deeper understanding of nonlinearities and interactions (e.g., capital account liberalization × income; timing and lags in effects).
- Built a large database unique in indicators and country coverage to mitigate data limitations.

### Appendix: sample and descriptive statistics (highlights)
- Table A1: LICs country sample includes 59 countries in the largest low income set.
  - Note markers: * countries included in CA baseline (Table 1 column 9); # countries included in the RER baseline (Table 5 column 3); & countries included in the NFA baseline (Table 10 column 1).
- Table A2 selected descriptive statistics (mean Stdev min max):
  - Current Account to GDP — -0.05440.0785-0.47210.2478
  - Fiscal Balance to GDP # — -0.01230.0552-0.47100.3733
  - Population growth # — 0.01220.0099-0.02680.0528
  - NFA to GDP — -0.71620.7291-10.14210.4170
  - Aid to GDP — 0.08740.0984-0.03690.8246
  - Concessional loans — 0.02650.0357-0.08890.4347
  - Net grants — 0.06230.07330.00080.7781
  - Terms of trade (G&S) — 4.63580.27633.74735.9550
  - Capital account liberalization — 0.50820.27770.00001.0000
  - Log of REER (INS) — 4.65650.34282.67945.8910
  - Log of REER (PWT) — -0.79720.4644-2.26631.5944
  - Public debt to trade — 3.24643.39110.008028.7196
  - Natural disaster — 0.51950.49980.00001.0000
  - Violent political transition — 0.06800.25180.00001.0000
  - Domestic financial liberalization # — -0.29420.1710-0.69080.1458
  - Constraint on executive — 3.63521.93121.00007.0000

*Source: Excerpts from the IMF working paper content unit provided in the source PDF.*

### Appendix Table A1) comprises “low income” or “lower middle income” according to the

### _wp09221 - Appendix Table A1) comprises “low income” or “lower middle income” according to the

### Data and sample
- Sample focuses on countries classified as “low income” or “lower middle income” by the World Bank, excluding emerging markets (China, Colombia, India, Indonesia, Pakistan and Thailand) to increase homogeneity.
- The largest low income country set used (for trading partner calculations and regressions with standard fundamentals) includes 59 low- and lower middle-income countries.
- Estimation samples: unbalanced panel of non-overlapping four-year averages over the period 1981-2005 with 6 observations for most countries.
- A summary statistic of the main data is provided in Appendix Table A2. A description of all variables is provided in the Appendix.
- Note on variables: most variables are computed in relation to trading partners; number of low income countries entering regressions varies by data availability.

### B. Analysis of Medium-Term Current Accounts — benchmark regressions and methodology
- Regressions include country fixed effects and traditional variables: fiscal balance, old age dependency ratio, population growth, initial net foreign asset position, oil trade balance, GDP per capita relative to the US, and real per capita GDP growth.
- New medium-term determinants introduced progressively across Columns 1 to 7; Columns 8 and 9 summarize preferred results (standard determinants and parsimonious specification).
- Estimation approach: unbalanced panel of non-overlapping four-year averages (1981-2005).

### Key empirical findings — current account (LICs)
- Fiscal balance and population growth (computed relative to trading partners) remain strongly significant with expected signs in the LIC sample.
- Terms of trade: income effects from terms of trade fluctuations have a strong impact on the current account; a temporary improvement in the terms of trade tends to improve the current account as part of the income increase is saved.
- Official financing (aggregate foreign aid) is significantly and negatively correlated with the current account:
  - Estimated effect: for any 1$ of aid, the current account deteriorates by about 0.1$ during the same period.
  - Aid components (separate regressions in columns 2, 6, 8, and 9):
    - About 80 percent of grants are fully absorbed through an increase in net imports over a 4 years period.
    - Only 30 to 50 percent of concessional loans received are absorbed through net imports over the same horizon.
- Domestic financial liberalization is associated with higher current accounts:
  - Estimated effect: shifting from complete financial repression to the sample average level of liberalization would lift the current account by 1.5 percent of GDP.
  - Interpretation: reforms boost domestic private savings more than investment (consistent with McKinnon hypothesis).
- Natural disasters (exogenous negative income shocks) impact the current account conditional on capital account openness:
  - Such shocks have a negative impact on the current account in countries with a fully open capital account.
  - Natural disasters are associated with current account improvements in countries with closed capital accounts.
  - Frequency of these exogenous negative income shocks in the sample: 0.6 over the sample period.
  - Quantitative example: in a country with a fully liberalized capital account, a natural disaster occurring at average frequency of 0.6 during a 4 year period would be associated with a -0.7 percent of GDP deterioration of the current account.
  - Interaction with initial net foreign asset position: negative and significant coefficient on the interaction term supports portfolio-composition predictions (Kraay and Ventura).
- Political instability: violent political events tend to improve the current account, suggesting significant capital flight.

### Robustness and interactions — current account
- Limited persistence of the current account beyond a 4 year horizon (Column 1 controls for lagged current account).
- Trade reforms (measured by index of average tariffs) show no additional explanatory power (Column 2).
- Terms of trade components: export prices are strongly positively correlated with the current account; import prices are weakly and negatively correlated.
- Results robust to inclusion of emerging markets classified as lower middle income (China, Colombia, India, Indonesia, Pakistan and Thailand) and to a broader measure of capital account liberalization (updated Quinn (1997)).
- Interaction tests with capital account openness:
  - Interaction of fiscal balance and capital account openness: effect of fiscal policy on the current account not significantly affected by capital account openness in LIC sample.
  - Interaction between capital account openness and population growth: not significant; negative sign consistent with theory.
  - Interaction between capital account openness and real per capita GDP growth: positive sign but not significant (suggests more open economies may run higher current account balances when growth is temporarily above trend).

### Cross-sample differences — are LICs different from HICs?
- Standard determinants showing significantly different impacts between low-income and high-income countries include:
  - Fiscal balance
  - Old age dependency ratio
  - Initial net foreign assets
- Fiscal balance: no significant association with the current account in high income countries, consistent with previous literature and greater prevalence of Ricardian-like behavior in advanced economies.
- Initial net foreign assets: negative coefficient in advanced countries consistent with intertemporal approach (initially richer countries sustain lower medium-term current account balances).
- Capital account openness results for more developed countries are sensitive to specification and interaction terms (evidence consistent with neoclassical theory when interacting openness with GDP per capita).

### C. Empirical Analysis of the Real Exchange Rate (RER) — methodology and benchmark
- Data: unbalanced panel of annual data covering 1980-2006.
- Tests: panel unit root tests (variables are I(1) except natural shocks) and panel cointegration tests reject null of no cointegration.
- Estimation: Dynamic Ordinary Least Squares (DOLS) with fixed effects; levels coefficients interpreted as long-run cointegrating relationship; changes and one lead and one lag of changes entered due to short sample length.
- Two measures of real effective exchange rate used:
  - CPI-based IMF-INS REER.
  - Penn World Table-derived price relative turned into REER using IMF-INS weights.
- Regression controls: net foreign assets, productivity, government consumption, terms of trade, trade restrictions, LIC-specific variables (aid flows, capital account liberalization), demographics, price controls, and natural shocks (dummy).

### Key empirical findings — RER (LICs)
- Government consumption is associated with real exchange rate appreciation (consistent with higher share of government spending on non-tradables).
- Terms of trade improvement appreciates the real exchange rate (effects similar to other samples).
- Fertility is associated with real exchange rate appreciation.
- Net foreign assets not significant (expectations of debt relief may blur intertemporal role).
- Productivity: not significant with IMF-INS REER; negative with Penn World Table REER (possible measurement issues for Balassa-Samuelson effect).
- Aid inflows (Roodman (2006)) are in the long run associated with a more depreciated exchange rate (potentially indicating a positive effect on nontradable-sector productivity relative to tradables; long-run depreciation consistent with productivity channel).
- Capital account liberalization associated with real exchange rate appreciation (suggesting persistent net capital inflows).
- Price distortions: presence of marketing boards (“max agric. price intervention”) tends to appreciate the RER.
- Black market premium (included in last column of Tables 5 and 6) is positive and significant (sample size halved when included), consistent with overvalued official exchange rate scenarios; coefficient between zero and one as predicted.
- Measurement caveat: dual exchange rate systems and price level measurement issues in LICs reduce sample sizes and complicate inference.

### RER robustness and dynamics
- Splitting terms of trade into export and import price components shows effect mainly driven by export prices (Tables 8 and 9).
- Error-correction specification for speed of adjustment:
  - Lagged error correction term coefficient hovered around 0.2 in 1–4 lag specifications, suggesting a half life of about three years.
  - Using alternative REER measure yields coefficient in the order of 0.3, indicative of a half life of about 2 years.
  - These speeds of adjustment align with existing literature (e.g., Rogoff (1996)).

### D. Empirical Analysis of the Net Foreign Asset (NFA) Position — methodology and benchmark
- Data: unbalanced panel annual data covering 1980-2006.
- Panel unit root and cointegration tests performed; cointegration tests reject null of no cointegration (Table 10 column 1).
- Estimation: DOLS with fixed effects on net foreign assets to trade ratio (preferred to NFA to GDP).
- Public debt data from Jaimovich and Panizza (2006); some splicing with external debt for a few LICs due to data limitations.
- Nicaragua dropped from NFA regressions due to extreme values of debt and net foreign assets.

### Key empirical findings — NFA (LICs)
- Domestic financial liberalization is associated with higher net foreign assets (indicative of larger positive effect on aggregate savings than on aggregate investment).
- Capital account liberalization: negative relationship with net foreign assets in low income countries (Table 10) and positive in high income ones (Table 11), consistent with current account regressions and neoclassical theory.
- Better institutional characteristics are associated with higher net foreign assets (may facilitate saving).
- Standard variables with strong links to net foreign assets:
  - Public debt: in the long run, half of the increase in public debt is reflected in a reduction of net foreign assets (effect in the order of 0.5-0.8 is similar to Lane and Milesi-Ferretti (2001)).
  - Demographics: higher share of dependent population associated with lower savings and net foreign assets.
  - Income per capita: positive association with net foreign assets in LICs (consistent with poor countries borrowing in development model; contrasts with some prior findings).
- Robustness: results hold with alternative NFA measures and matching debt indicators, alternative demographics, capital account liberalization and relative income/productivity measures (columns 3-7 of Table 10).
- Terms of trade shocks, trade restrictions, and price controls in agriculture do not matter after controlling for other determinants (columns 8-10).

### Cross-sample differences — NFA (LICs vs HICs)
- Relationship of capital account liberalization with NFA differs sign by income group (negative in LICs, positive in HICs).
- Net foreign assets respond differently to public debt, demographics, and income across income groups; tests of equality of coefficients are reported in Table 11.

*Source: Excerpts from the IMF working paper content unit provided in the source PDF.*

### 3. The high income group appears quite different for about half of the indicators and not too

### 3. The high income group appears quite different for about half of the indicators and not too different for the other half.

### Differences for the high income group
- The most interesting result is the opposite coefficient for capital account liberation, consistent with the fact that rich countries lend and poor borrow.
- Debt appears insignificant for the high income sample, which is in line with the range of 0 to 0.2 found by Lane and Milesi-Ferretti (2001) for this group of countries.
- Possible interpretation: public debt is less likely to be foreign financed in higher income countries than in low income countries, possibly because of much deeper financial markets in more developed countries.
- The role of demographics, domestic financial liberalization, and institutions seem to follow the same economic patter for both set of countries, although the size maybe somewhat different.

### Speed of adjustment (error correction analysis)
- An error correction specification was implemented to gauge the speed of adjustment of the net foreign asset position.
- The error correction term was derived via the difference of the net foreign assets from the sum of the products of the fundamentals entering the baseline regression in Table 10 column 1 times the corresponding level coefficients.
- Error correction regressions deliver robust OLS coefficients of the lagged error correction term of around 0.25.
- Interpretation: shocks to the gap between the net foreign asset and its long run relationship have a half life of about two and a half years.

### Conclusions (medium- to long-term relationships for LICs)
- The paper estimates relationships between the real effective exchange rate, the current account, and the net external asset position of low income countries with respect to a set of fundamentals in the medium to long-term.
- The same broad set of economic fundamentals explain coherently the three indicators.
- Focus on low income countries motivated by limited attention in the literature and by factors that more closely characterize these countries: policies and distortions in the financial sector, access to official external financing, and higher sensitivity to exogenous shocks.
- Built a large database unique in terms of the set of indicators and of countries covered to address data limitations.

### Distinctive medium-term determinants for LICs
- In addition to standard determinants, the following factors have a significant impact on external balance indicators of low-income countries:
  - aid flows (grants and concessional loans),
  - domestic financial liberalization,
  - removal of capital account controls,
  - quality of institutions,
  - shocks (terms of trade, natural disasters),
  - demographic measures.
- Results are generally coherent across methodologies and—for standard economic indicators—are mainly in line with the existing literature.

### Policy distortions: financial reforms, capital account liberalization, institutions
- Domestic financial reforms are associated with:
  - an improvement of the current account,
  - an improvement of the net foreign assets,
  - interpretation: suggest a larger positive effect on saving than on investment.
- Capital account liberalization is associated with:
  - allowing countries to borrow against disasters (lower current account),
  - allowing lower income countries in general to borrow from higher income countries,
  - a more appreciated real exchange rate in the long run, possibly due to the effect of capital inflows on absorption.
- Quality of institutions is generally positively associated with a larger external wealth in the long run.

### Shocks: terms of trade, natural disasters, income shocks
- Positive terms of trade shock tends to:
  - improve the current account,
  - appreciate the real exchange rate,
  - caveat: this mainly occurs if it arises from a change in the export price (consistent with import prices generating an additional substitution effect working opposite the income effect common to both export and import prices).
- Natural disasters tend to be associated with an improvement of the current account, unless the country has an open capital account.
- Preliminary evidence suggests the effect of income shocks on the current account may depend on the initial net foreign asset position.

### External financing and aid
- An increase in aid arising from concessional loans or grants:
  - progressively results in higher imports,
  - evidence suggests part of aid flows are saved in the short-run.
- In the long run, an increase in aid is associated with a deterioration of the real effective exchange rate.
  - This result may be surprising relative to the standard Dutch disease argument but is consistent with more general theories of Dutch disease with learning by doing in both tradable and non-tradable sectors (Torvik, 2001).
  - Mechanism: aid may cause a short-run appreciation as expenditure on non tradables increase relative to supply, but may be associated with a long run depreciation if channeled to improve the productivity of nontradables relative to tradables.

### Scope for further research
- Priority: improve the quality and the extent of data coverage.
  - Several key indicators (not only black market premium, but also the structural and financing indicators) are generally missing for numerous countries, impairing proper economic assessment of external balance.
- Need deeper understanding of nonlinearities underlying the relationships studied.
  - Example: interactions between capital account liberalization and income, and between these factors and other determinants, lag behind the theoretical hypotheses proposed.

*Source: _wp09221 - 3. The high income group appears quite different for about half of the indicators and not too different for the other half.*

### References

### _wp09221 - References

### Key bibliographic entries
- Abdul Abiad, Enrica Detragiache, and Thierry Tressel (2008), “A New Database of Financial Reforms“, IMF Working Paper No. 08/266.
- Alfaro, Laura, Sebnem Kalemli-Ozcan and Vadym Volosovych (2007) “Capital Flows in a Globalized World: The Role of Policies and Institutions”, in Capital Controls and Capital Flows in Emerging Economies: Policies, Practices, and Consequences, edited by Sebastian Edwards, National Bureau of Economic Research. University of Chicago Press. (Also NBER Working Paper No. 11696.)
- Berg, Andrew, Shekhar Aiyar, Mumtaz Hussain, Shaun K. Roache, Tokhir N. Mirzoev, Amber Mahone (2007), The Macroeconomics of Scaling Up Aid: Lessons from Recent Experience. Occasional Paper No. 253, April 12, 2007.
- Blanchard, Olivier (1985), “Debt, deficit and finite horizons”, Journal of Political Economy 93 (April): 223-47.
- Bosworth, B., Collins, S.M. (1999), “Capital Flows to Developing Countries: Implications for Saving and Investment”, Brookings Papers on Economic Activity vol. 1, 143-180.
- Caballero, R., Farhi E., and P.O. Gourinchas (2008), “An Equilibrium Model of ‘Global Imbalances’ and Low Interest Rates”, American Economic Review 2008, 98:1, 358-393.
- Calvo, Guillermo (1987), “Balance of Payments Crises in a Cash in Advance Economy”, Journal of Money, Credit and Banking 19, Feb., 19-31.
- Cashin, Paul, Luis F. Céspedes, and Ratna Sahay (2004), “Commodity Currencies and The Real Exchange Rate,” Journal of Development Economics, Vol. 75, pp. 239–68.
- Chen, Yu-Chin and Kenneth Rogoff (2003), “Commodity Currencies,” Journal of International Economics, Vol. 60 (May), pp. 133–60.
- Chinn, M., and E. Prasad (2003), “Medium-term determinants of current accounts in industrial and developing countries: an empirical exploration”, Journal of International Economics, 59, 47-76.
- Chinn, M., and S. J. Wei (2008), “A Faith-based Initiative: Does a Flexible Exchange Rate Regime Really Facilitate Current Account Adjustment?”, NBER Working Paper 14420.
- Chinn M. and H. Ito (2007), “Current Account Balances, Financial Development and Institutions: Assaying the World ‘Saving Glut’”, Journal of International Money and Finance, 26, 546-569.
- Christiansen Lone and Stephen Tokarick (2009), “Real Exchange Rates and Fundamentals: Some Theory and Evidence“, International Monetary Fund, mimeo.
- Choudhri, E., and M. Khan (2005), “Real Exchange Rates in Developing Countries: Are Balassa-Samuelson Effects Present?”, IMF Staff Papers, Vol. 52 (December), pp.387-409.
- Chudik Alexander and Joannes Mongardini (2007) “In search of equilibrium : estimating equilibrium real exchange rates in Sub-Saharan African countries” IMF Working Paper 07/90.
- Corbo, V. and K. Schmidt-Hebbel (1991): “Public Policies and Saving in Developing Countries,” Journal of Development Economics 36 (July): 89-115.
- De Gregorio, José, Alberto Giovannini, and Holger Wolf (1994), “International Evidence on Tradables and Nontradables Inflation,” European Economic Review, Vol. 38 (June), pp. 1225–44.
- Delechat, Corinne (2008) “Assessing Real Exchange Rates in Low-Income Countries”, IMF, mimeo.
- Di Bella, Gabriel, Mark Lewis, and Aurelie Martin (2007) “Assessing Competitiveness and Real Exchange Rate Misalignment in Low-Income Countries” IMF Working Paper 07/201.
- Dimaranan, Betina V. and Robert A. McDougall, eds (2006), Global Trade, Assistance, and Production: The GTAP 6 Data Base, Center for Global Trade Analysis, Purdue University.
- Dixit, Avinash, and Victor Norman (1980), Theory of International Trade: A Dual General Equilibrium Approach, (Cambridge: Cambridge University Press).
- Easterly, W., M. Kremer, L. Pritchett, and L. Summers (1993), “Good Policy or Good Luck? Country Growth Performance and Temporary Shocks,” Journal of Monetary Economics, Vol. 32, December 1993, 459-483.
- Edwards, Sebastian (1988), “Real and Monetary Determinants of Real Exchange Rate Behavior”, Journal of Development Economics, vol. 29, pp. 311-341.
- Edwards, S. (1989), Real Exchange Rates, Devaluation, and Adjustment, Exchange Rate Policy in Developing Countries, The MIT Press, Cambridge, Mass.
- Edwards, S. (1995), “Why are Saving Rates so Different Across Countries? An International Comparative Analysis.”, NBER Working Paper 5097.
- Edwards, S., and J. Ostry (1992), “Terms of Trade Disturbances, Real Exchange Rates, and Welfare: The Role of Capital Controls and Labor Market Distortions”, Oxford Economic Papers, Vol. 44, No.1, pp.20-34.
- Edwards, Sebastian and Miguel Savastano (2000), “Exchange Rates in Emerging Economies: What Do We Know? What Do We Need to Know?” in Economic Policy Reform: The Second Stage, edited by Anne O. Krueger (Chicago: University of Chicago, Press).
- Elbadawi Ibrahim (2007) “Real Exchange Rate Misalignment in Sub-Saharan Africa: How Serious? How Dangerous?” World Bank, mimeo.
- Engel, Charles and Kenneth D. West (2005), “Exchange Rate and Fundamentals,” Journal of Political Economy, vol. 113, no. 3.
- Engel, Charles, Nelson C. Mark, and Kenneth D. West (2007), “Exchange Rate Models Are Not as Bad as You Think”, Macroeconomics Annual.
- Faria, A., Lane, P., Mauro, P., and G.M. Milesi-Ferretti (2007), “The Shifting Composition of External Liabilities”, Journal of the European Economic Association, Vol. 5, no 2-3, pp. 480-490.
- Frankel, Jeffrey, David Parsley, and Shang-Jin Wei (2005), “Slow Pass Through Around the World: A New Import For Developing Countries?”, Working Paper 11199, (Cambridge: National Bureau of Economic Research).
- Froot, Kenneth A. and Kenneth Rogoff (1995), “Perspectives on PPP and Long-Run Real Exchange Rates,” in Handbook of International Economics, Vol. 3, edited by Gene M. Grossman and Kenneth Rogoff (Amsterdam: Elsevier Science Publishers), pp. 1647–88.
- Fry, Maxwell (1995), Money, Interest, and Banking in Economic Development, The John Hopkins University Press.
- Gelb, A. (1991), “Financial Policies, Growth and Efficiency”, Washington, DC: World Bank, Country Economics Department, PPR Working Paper WPS 202, June.
- Gleditsch, Nils Petter, Peter Wallensteen, Mikael Eriksson, Margareta Sollenberg and Håvard Strand (2002), “Armed Conflict 1946–2001: A New Dataset”, Journal of Peace Research vol. 39 (5), pp. 615–637.
- Goldfajn, Ilan, and Rodrigo Valdes (1999), “The Aftermath of Appreciations,” Quarterly Journal of Economics, Vol. 114 (February), pp. 229–62.
- Greenwood, J., and B. Jovanovic (1990), “Financial Development, Growth, and the Distribution of Income”, Journal of Political Economy, XCVIII, 1076-1107.
- Guo, K., and K. Jin (2009), “Composition and Growth Effects of the Current Account: A Synthesized Portfolio View”, Mimeo Harvard University.
- Gruber, J.W., and S. B. Kamin (2007), “Explaining the Global Pattern of Current Account Imbalances”, Journal of International Money and Finance 26, 500-522.
- Hinkle Lawrence and Peter Montiel (1999), “Exchange Rate Misalignment. Concepts and Measurement for Developing Countries”, A World Bank Research Publication, Oxford University Press.
- Isard, Peter and Hamid Faruqee (1998), Exchange Rate Assessment – Extensions of the Macroeconomic Balance Approach, International Monetary Fund Occasional Paper 167.
- Jaimovich Dany and Ugo Panizza (2006) "Public Debt around the World: A New Dataset of Central Government Debt", IDB Research Department Working Paper Nr. 561.
- Jones, Ronald (1965), “The Structure of Simple General Equilibrium Models,” Journal of Political Economy, Vol. 73, (December), pp. 557–72.
- Ju, J. and S-J. Wei (2007), “Current Account Adjustment: Some New Theory and Evidence”, NBER Working Paper 13388.
- Kee, Haui Looi, Alessandro Nicita, and Marcelo Olarreaga (2008), “Import Demand Elasticities and Trade Distortions,” Review of Economics and Statistics, vol. 90, no. 4, pp. 666-682.
- Khan, Mohsin, and Morris Goldstein (1985), “Income and Price Effects in Foreign Trade,” in Handbook of International Economics, volume 2, edited by Ronald Jones and Peter Kenen, New York: North Holland, pp. 1041-1105.
- Kireyev, Alexei (2008), “Non-Econometric Methods for Exchange Rate Assessment in Low-Income Countries” IMF, mimeo.
- Kohli, Ulrich (1991), Technology, Duality, and Foreign Trade: The GNP Function Approach to Modeling Imports and Exports, (Ann Arbor: University of Michigan Press).
- Kraay, Aart, and Jaume Ventura (2000), “Current Accounts in Debtor and Creditor Countries”, Quarterly Journal of Economics, 115, 1137-1166.
- Lane, Phillip, and GianMaria Milesi-Ferretti (2002), “External Wealth, The Trade Balance, and the Real Exchange Rate”, European Economic Review, Vol. 46, pp. 1049-71.
- Lane, Phillip, and GianMaria 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.
- Lane, Phillip, and GianMaria Milesi-Ferretti (2004), “The Transfer Problem Revisited: Net Foreign Assets and Real Exchange Rates”, The Review of Economics and Statistics, November, 86(4), 841-857.
- Lane, Phillip, and GianMaria Milesi-Ferretti (2001), “Long Term Capital Movements”, NBER Macroeconomics Annual.
- Lee, J., Milesi-Ferretti, G.M., Ostry J., Prati A., and L. Ricci (2008), Exchange Rate Assessments: CGER Methodologies, International Monetary Fund Occasional Paper 261.
- Levine, Ross (1997), “Financial Development and Economic Growth: Views and Agenda.” Journal of Economic Literature.
- Loayza, N., Schmidt-Hebbel, K., and L. Servén (2000), "What drives Saving across the World?", The Review of Economics and Statistics, LXXXII (2): 165-181.
- MacDonald, Ronald and Luca Antonio Ricci (2007), “Real Exchange Rates, Imperfect Substitutability, and Imperfect Competition”, Journal of Macroeconomics, Vol. 29, No.4, pp. 639-664.
- Martin, Philippe, and Helene Rey (2006), "Globalization and Emerging Markets: With or without rash?" American Economic Review, December, vol 96, n5, 1631-51.
- Masson, P., Bayoumi, T., Samiei, H. (1998), “International Evidence on the Determinants of Private Saving”, World Bank Economic Review vol. 12, 483-501.
- Matsen, E., and R. Torvik (2004), “Optimal Dutch disease”, Journal of Development Economics, 2005, 494-515.
- McKinnon, R. (1973), Money and Capital in Economic Development, (Washington DC: Brookings Institutions).
- Meese, Richard A., and Kenneth Rogoff (1983), “Empirical Exchange Rate Models of The Seventies: Do They Fit out of Sample?” Journal of International Economics Vol. 14, pp. 3–24.
- Mendoza, E., Quadrini, V., and J-V. Rios-Rull (2008), “Financial Integration, Financial Development and Global Imbalances”, forthcoming, Journal of Political Economy.
- Modigliani, F. (1986). Life Cycle, Individual Thrift, and the Wealth of Nations. American Economic Review vol. 76 (3), pp. 297-313.
- Mongardini, J., and B. Rayner (2009) “To Transfer or Not to Transfer: The Impact of Grants and Remittances on the Equilibrium Real Exchange Rates in Sub-Saharan Countries”, IMF Working Paper 09/75.
- Montiel, P. (1999), “Determinants of the Long-Run Equilibrium Real Exchange Rate: An Analytical Model”, Chapter 6 in L. Hinkle & P. Montiel (eds.), Exchange rate misalignment: concepts and measurements for developing countries, World Bank, Washington D.C, pp. 264-290.
- Obstfeld, and Rogoff (1999), Foundations of International Macroeconomics, MIT Press.
- Ostry, Jonathan (1988), “The Balance of Trade. The Terms of Trade, and the Real Exchange Rate: An Intertemporal Optimizing Framework,” IMF Staff Papers, International Monetary Fund, Vol. 35, No4., pp.541-73.
- Ostry, Jonathan (1990), “Tariffs and the Current Account: The Role of Initial Distortions”, Canadian Journal of Economics, 23, 348-56.
- Ostry, Jonathan D., and Carmen M. Reinhart (1992), “Private Savings and Terms of Trade Shocks: Evidence from Developing Countries,” Staff Papers, Vol. 39, No. 3 (September), pp. 495–517 (Washington, D.C.: International Monetary Fund).
- Pedroni, Peter (1999), “Critical Values for Cointegration Tests in Heterogeneous Panels with Multiple Regressors”, Oxford Bulletin of Economics and Statistics, Vol. 61, pp. 653-70.
- Pedroni, Peter (2000), “Fully Modified OLS for Heterogenous Cointegrated Panels”, Advances in Econometrics, Vol. 15, pp. 93-130.
- Pedroni, Peter (2001), “Purchasing Power Parity Tests in Cointegrated Panels”, Review of Economics and Statistics, Vol. 83, pp. 727-731.
- Pedroni, Peter (2004), “Panel Cointegration; Asymptotic and Finite Sample Properties of Pooled Time Series Tests with an Application to the PPP Hypothesis”, Econometric Theory, Vol. 20, pp. 597-625.
- Pesaran, Hashem (2007), “A Simple Panel Unit Root Test in the Presence of Cross Sectional Dependence”, Journal of Applied Econometrics, Vol. 22, pp.265-312.
- Prasad, Eswar, Raghuram Rajan, and Arvind Subramanian (2007), “Foreign Capital and Economic Growth,” Brookings Papers on Economic Activity, pp. 153-230.
- Prati, A., and T. Tressel (2006), “Aid Volatility and Dutch Disease: Is There a Role for Macroeconomic Policies?”, IMF Working Paper 06/145.
- Quinn, Dennis (1997), “The Correlates of Changes in International Financial Regulation,” American Political Science Review 91 (September), 531–51.
- Razin, A., and L. Svensson (1983), “The Terms of Trade and the Current Account: The Harberger-Laursen-Metzler Effect”, Journal of Political Economy, Vol. 91, pp. 97-125.
- Reinhart, Carmen and Kenneth Rogoff (2004), “The Modern History of Exchange Rate Arrangements: A Reinterpretation”, Quarterly Journal of Economics, Vol. CXIX, No. 1, pp. 1-48, February.
- Ricci Luca Antonio, Gianmaria Milesi-Ferretti, and Jaewoo Lee (2008), “Real Exchange Rates and Fundamentals: A Cross-Country Perspective”, IMF Working Paper 08/13.
- Rogoff, Kenneth (1996), “The Purchasing Power Parity Puzzle,” Journal of Economic Literature, Vol. 34 (June), pp. 647–68.
- Roodman, David (2006), “An Index of Donor Performance”, Center for Global Development, Working Paper Number 67, November.
- Rose, A., Supaat S, and J. Braude (2008), “Fertility and the Real Exchange Rate”, mimeo Berkeley.
- Roudet, Stéphane, Magnus Saxegaard, and Charalambos G. Tsangarides (2007) “Estimation of equilibrium exchange rates in the WAEMU : a robustness approach.” IMF Working Paper 07/194.
- Sachs Jeffrey and Andrew Warner (1995), “Economic Reform and the Process of Global Integration”, Brookings Papers on Economic Activity, No. 1, pp. 1-118.
- Schmidt-Hebbel, K., Webb, S.B., Corsetti, G. (1992), “Household saving in developing countries: first cross-country evidence”, World Bank Economic Review, Vol. 6, pp. 529-547.
- Senhadji, Semlali (1997), “Time-Series of Structural Import Demand Equations—A Cross Country Analysis,” IMF Working Paper 97/132.
- Stern, Robert (1973), The Balance of Payments: Theory and Economic Policy, (New York: Aldine Publishing Company).
- Stern, Robert, Francis, Jonathan, and Bruce Schumacher (1976), Price Elasticities in International Trade—An Annotated Bibliography, London: MacMillan.
- Tornell Aaron and Phillip Lane (1998), “Are Windfalls a Curse? A Non-Representative Agent Model of the Current Account”, Journal of International Economics, Vol. 44, pp.83-112.
- Torvik, R. (2001), “Learning by Doing and the Dutch Disease”, European Economic Review, vol. 45, pp. 285-306.
- Van Wijnbergen, S. (1984), “The ‘Dutch Disease’: A Disease After All?” Economic Journal, Vol. 94, No. 373, pp. 41–55.
- Vegh, Carlos, Open Economy Macroeconomics in Developing Countries, forthcoming, MIT Press.
- Wacziarg, Romain and Karen Horn Welch (2003), “Trade Liberalization and Growth: New Evidence,” NBER Working Paper No. 10152 (December).
- Weil, Philippe (1989), “Overlapping families of infinitely lived agents”, Journal of Public Economics, Vol. 38 (March), pp.183-98.
- Woodland, Alan (1982), International Trade and Resource Allocation, North Holland.

### Empirical findings and key regression results (tables and notes)
- Table 1: Medium-term determinants of the current account (robust panel regressions; dependent variable: current account to GDP ratio; data averaged over 4 year periods covering 1981-2005)
  - Fiscal balance / GDP: coefficients range from 0.2432** to 0.3966*** across columns (Robust p-values in parentheses shown in table).
  - Population growth: negative and often significant; example coefficients include -1.6275**, -2.6210***, -1.8436***.
  - Terms of trade: positive and significant in many specifications; e.g., 0.0228* , 0.0345**, 0.0319***.
  - Concessional loans / GDP: negative and significant in specifications where included; e.g., -0.3122*, -0.5565***.
  - Net grants / GDP: mixed; examples 0.2108* , 0.2168*.
  - Domestic financial liberalization (initial): positive and significant in some columns; e.g., 0.0343**, 0.0326**.
  - Capital account liberalization (initial): positive and significant in several columns; e.g., 0.0492**, 0.0434**.
  - Natural disaster: positive and significant in several columns; e.g., 0.0349**, 0.0359***.
  - Interaction natural disaster * capital account (initial): negative and significant in several columns; e.g., -0.0605**, -0.0654***.
  - Violent political transition: positive and highly significant; e.g., 0.0287***.
  - Observations vary by column (e.g., 223, 219, 143) and R-squared values reported (e.g., 0.73, 0.84).
  - Note: Significance codes in table: *** p<0.01, ** p<0.05, * p<0.1 (Robust p-values in parentheses).

- Table 2: Robustness of medium-term current account determinants (additional controls added one at a time; broader country sample in Column (4))
  - Fiscal balance / GDP: coefficients positive and significant (e.g., 0.3166**, 0.3486***).
  - Population growth: consistently negative and significant (examples: -1.4924*, -2.7107***).
  - log (terms of trade): positive and significant in multiple columns (e.g., 0.0289**, 0.0352***).
  - Concessional loans / GDP: negative and significant across specifications (e.g., -0.4185*, -0.5318***).
  - Net grants / GDP: positive in some columns (e.g., 0.1953* , 0.2328**).
  - Violent political transition: positive and significant across columns (e.g., 0.0265**, 0.0272***).
  - Domestic financial liberalization: positive and significant in some columns (e.g., 0.0317**).
  - Capital account liberalization: positive and significant in several columns (e.g., 0.0432**).
  - Natural disaster: positive and significant (e.g., 0.0350**).
  - Natural disaster*capital account: negative and significant in several columns (e.g., -0.0664***).
  - Observations and R-squared reported; R-squared values around 0.80–0.86.

- Table 3a and 3b: Current account regressions with different slopes for LICs and HICs; F-tests of equality of coefficients across income groups
  - Differences across income groups reported for many variables; examples:
    - Fiscal balance / GDP: LICs coefficient 0.3224***; HICs coefficient -0.0811 (p-values for equality reported in Table 3b: 0.06, 0.04, 0.02 across specifications).
    - Capital account liberalization (initial): LICs positive 0.0434** in some specs; HICs negative -0.1145** in others (F-test p-values indicate significant differences).
  - Table 3b provides p-values for equality tests (Ho: coeff HIC = coeff LIC) for variables such as Fiscal balance / GDP, Old age dependency ratio, NFA / GDP, Natural disaster*Capital account, etc.

- Table 4: Panel Unit Root test statistics (Pesaran 2007 test; samples A and B correspond to minimum observations per country of 14 or 20)
  - Examples of test statistics (Sample A / Sample B):
    - Log of REER (INS): -1.59 / -1.57
    - Log of REER (PWT): -1.82 / -1.97
    - NFA to GDP: -2.02 / -2.06
    - Natural disaster: -3.66* / -3.52* (asterisk indicates reject null HO of unit root at 5% one-sided)
    - Aid to GDP: -2.19* / -2.07*
  - Notes: * means reject null HO of unit root at 5% one-sided; # indicates variable constructed relative to weighted average of trading partners; test based on Pesaran (2007).

- Table 5: Real effective exchange rate (IMF definition) regressions (Panel dynamic OLS with fixed effects; long-run coefficients)
  - Terms of trade goods (log): 0.3458*** to 0.4353*** across columns (robust p-values 0.0000).
  - Government Consumption to GDP (# relative to trading partners): coefficients range 1.1259* to 2.0271***.
  - Aid to GDP (#): large negative and significant coefficients, e.g., -2.2405***, -1.3812***.
  - Capital account liberalization (#): positive and significant, e.g., 0.3152***, 0.4910***.
  - Fertility (#): positive and highly significant (e.g., 0.1239***).
  - Black market premium (%): 0.2547*** in one specification (p=0.0015).
  - Observations reported (e.g., 522, 609, 338) and R-squared values (e.g., 0.71, 0.78).

- Table 6: Real effective exchange rate (Penn World Table) regressions (Panel dynamic OLS with fixed effects)
  - GDP per worker PWT (log): -0.4418***, -0.4266*** (significant negative).
  - Terms of trade goods (log): positive and significant (e.g., 0.1634**, 0.1970***).
  - Government Consumption to GDP (#): positive and significant (e.g., 2.0186**, 3.3371***).
  - Aid to GDP: negative and significant (e.g., -3.2405***).
  - Capital account liberalization (#): positive and significant in some columns (e.g., 0.1746**, 0.3048**).
  - Natural disaster: -0.1549*** in one specification (p=0.0057).
  - Black market premium (%) : 0.4344*** (p=0.0000).
  - Observations and R-squared values reported (e.g., Observations 522–622; R-squared 0.74–0.81).

- Table 7a and 7b: REER (IMF def.) regressions with different slopes for LICs and HICs; F-tests of equality
  - Notable heterogeneous effects:
    - NFA (w/ NPV debt) to trade: HIC coefficient 0.0276***; LIC coefficient 0.0074 (p-values for equality reported in Table 7b).
    - Relative productivity (log) #: negative for some HIC columns (e.g., -0.1346**) and positive for LICs in some columns (e.g., 0.2087***).
    - Terms of trade goods (log): LICs show larger positive responses in several columns (e.g., 0.3811***).
    - Aid to GDP #: HICs and LICs show differing signs across columns (e.g., Aid to GDP # 3.3650** for HIC in one column vs. -2.4081*** for LIC in another).
  - Table 7b reports p-values for tests Ho: coeff HIC = coeff LIC for each variable.

- Table 8: REER (IMF def.) regressions robustness (long-run coefficients)
  - Terms of trade goods (log): consistently positive and highly significant (e.g., 0.3931***, 0.4281***).
  - Government Consumption to GDP #: positive and often highly significant (e.g., 2.0271***, 2.3317***).
  - Aid to GDP #: negative and highly significant across specifications (e.g., -1.6187***, -4.6194***).
  - Capital account liberalization #: positive and significant (e.g., 0.2978***).
  - Price decomposition: Price of exports - goods 0.3698*** and Price of imports - goods -0.3754*** in one specification.
  - Demographics and other controls: Infant mortality rate (UN) #0.0052***; Old age dependency #5.2828***; GDP per worker PWT (log) -0.2360** in extended robustness tests.
  - Observations and R-squared values reported (e.g., Observations 609; R-squared 0.65–0.78).

- Table 9: REER (PWT) regressions robustness
  - Terms of trade goods (log): positive and significant in multiple specifications (e.g., 0.1970***, 0.3453***).
  - Government Consumption to GDP #: strong positive coefficients (e.g., 3.3371***, 4.0020***).
  - Aid to GDP: negative and significant in many specifications (e.g., -2.0504***, -2.6034***).
  - Capital account liberalization #: positive and significant in some columns (e.g., 0.1984**, 0.3337***).
  - Max agric. price intervention #: positive and significant in several specifications (e.g., 0.1842***).
  - Other controls: GDP per worker PWT (log) negative and significant in some columns (e.g., -0.3678***); Infant mortality and old age dependency significant in robustness checks.
  - Observations and R-squared values reported (e.g., Observations 622; R-squared ~0.69–0.75).

- Table 10: Net foreign asset regressions (Panel dynamic OLS with fixed effects; long-run coefficients)
  - Public debt to trade: large negative and highly significant across specifications (e.g., -0.4566***, -0.4840***).
  - Fertility #: negative and significant in several columns (e.g., -0.3135***, -0.4913***).
  - Relative productivity (log) #: positive and significant (e.g., 1.5867***, 1.7841***).
  - Constraints on executive: positive and highly significant (e.g., 0.2280***, 0.2505***).
  - Domestic financial liberalization #: positive and significant in several columns (e.g., 2.0003***, 1.8758***, 2.2360***).
  - Capital account liberalization #: negative and significant in some columns (e.g., -1.4137***, -1.5802***).
  - Additional controls and alternative specifications include public debt to GDP, NPV of external debt to trade, old age dependency, income per capita (relative to US), GDP per worker PWT (log), with varying significance.
  - Observations and R-squared values reported (e.g., Observations 610, R-squared up to 0.96 in one column).

- Table 11: Net foreign asset regressions with different slopes for LICs and HICs (test of equality of coefficients)
  - Public debt to trade: LICs coefficient -0.4566***; HICs & EM coefficient -0.0059 (test p-value 0.000).
  - Fertility #: LICs -0.3135***; HICs & EM -0.2623** (test p-value 0.731).
  - Relative productivity (log) #: LICs 1.5867***; HICs & EM 0.3129 (test p-value 0.030).
  - Constraints on executive: positive and significant for both groups (LICs 0.2280***; HICs & EM 0.1221***; test p-value 0.074).
  - Domestic financial liberalization #: LICs 2.0003***; HICs & EM 0.8670*** (test p-value 0.059).
  - Capital account liberalization #: LICs -1.4137***; HICs & EM 1.2703*** (test p-value 0.000).
  - Observations: 14 (presumably per group reported as "14 14" in table), R-squared 0.87 for the NFA to trade regression.

### Notes on methodology mentioned in tables
- Many regressions are robust panel regressions with country fixed effects; dependent variables often averaged over 4 year periods (except when stated otherwise) covering 1981-2005 for current account analyses and annual unbalanced panels 1980-2006 for REER and NFA analyses.
- Significance markers: *** p<0.01, ** p<0.05, * p<0.1; robust p-values are shown in parentheses in the tables.
- Several variables marked with #: constructed relative to the weighted average of the trading partners.
- Unit root testing based on Pesaran (2007) panel unit root test.

*Source: _wp09221 - References (PDF)._

### Appendix tables

### _wp09221 - Appendix tables

### Table A1: LICs country sample
- 1Albania*&31Macedonia, FYR
- 2Algeria&32Madagascar#*&
- 3Azerbaijan, Rep. of#*&33Malawi
- 4Bangladesh#*&34Mali
- 5Belarus#&35Morocco#*&
- 6Benin36Mozambique#*&
- 7Bolivia#*&37Namibia
- 8Burkina Faso#*&38Nepal&
- 9Cambodia39Nicaragua#*
- 10Cameroon#*&40Niger
- 11Chad41Nigeria#*&
- 12Congo, Dem. Rep. of42Papua New Guinea
- 13Congo, Republic of43Paraguay*&
- 14Côte d'Ivoire#*&44Peru#*&
- 15Dominican Republic#*&45Philippines#*&
- 16Ecuador#*&46Rwanda
- 17Egypt&47Senegal*&
- 18El Salvador#*&48Sierra Leone#
- 19Ethiopia#*&49Sri Lanka#*&
- 20Gambia, The#50Swaziland
- 21Georgia&51Syrian Arab Republic
- 22Ghana#&52Tanzania#*&
- 23Guatemala#*&53Togo
- 24Guinea54Tunisia#*&
- 25Honduras#55Uganda#*&
- 26Jamaica#*&56Ukraine#&
- 27Jordan#*&57Uzbekistan*
- 28Kenya#*&58Vietnam&
- 29Kyrgyz Rep.#&59Zambia#
- 30Lao People's Dem.Rep

- Note:
  - * countries included in CA baseline (Table 1 column 9)
  - # countries included in the RER baseline (Table 5 column 3)
  - & countries included in the NFA baseline (Table 10 column 1)

### Table A2: Descriptive Statistics (variable — mean Stdev min max)
- Current Account to GDP — -0.05440.0785-0.47210.2478
- Fiscal Balance to GDP # — -0.01230.0552-0.47100.3733
- Old age dependency # — -0.09460.0304-0.17900.0362
- Population growth # — 0.01220.0099-0.02680.0528
- NFA to GDP — -0.71620.7291-10.14210.4170
- Oil balance to GDP — -0.00630.0915-0.30020.6143
- Relative income percapita — 0.06820.05130.00760.2671
- GDP percapita growth # — -0.01040.0509-0.30380.2800
- Aid to GDP — 0.08740.0984-0.03690.8246
- Concessional loans — 0.02650.0357-0.08890.4347
- Net grants — 0.06230.07330.00080.7781
- Terms of trade (G&S) — 4.63580.27633.74735.9550
- Capital account liberalization (other) — 0.44140.33660.00001.0000
- Capital account liberalization — 0.50820.27770.00001.0000
- Log of REER (INS) — 4.65650.34282.67945.8910
- Log of REER (PWT) — -0.79720.4644-2.26631.5944
- NFA (w/ NPV debt) to trade — -2.59413.5943-24.58171.0167
- Government Consumption to GDP # — -0.03780.0568-0.14540.3731
- Terms of trade goods (log) — 4.66770.31253.21496.1624
- Fertility # — 2.74001.3487-0.52794.8621
- GDP per worker PWT (log) — 8.42480.83456.690210.0493
- Trade restrictions # — 0.17290.1858-0.24630.8064
- Administered prices # — 0.02230.3242-0.30870.9738
- Max price intervention # — 0.18330.4988-0.53130.9372
- Natural disaster — 0.51950.49980.00001.0000
- Violent political transition — 0.06800.25180.00001.0000
- Public debt to trade — 3.24643.39110.008028.7196
- Fertility # (duplicate entry) — 2.75121.3558-0.52794.8621
- Relative productivity (log) # — -1.91310.6679-3.56560.0631
- Domestic financial liberalization # — -0.29420.1710-0.69080.1458
- Constraint on executive — 3.63521.93121.00007.0000
- NPV of external debt to trade — 2.26572.45060.032321.4463

- Note: # indicates deviation from trading partners

### Table A3: Real deposit interest rates in selected Asian developing countries and in the US (periods: 1968-1974 | 1975-1981 | 1982-1988)
- Bangladesh — -7.88 | -3.19 | 2.33
- India — -2.46 | 0.07 | 1.13
- Indonesia — -1.21 | -5.87 | 5.7
- Korea — 1.58 | -2.58 | 5.13
- Malaysia — 2.01 | 1.79 | 5.37
- Nepal — -0.62 | 4.04 | 3.36
- Pakistan — -2.33 | -1.55 | 2.96
- Phillipines — -4.22 | -0.16 | -0.32
- Sri Lanka — -4.52 | 0.27 | 4.08
- Taiwan — 0.34 | 2.88 | 5.48
- Thailand — 1.24 | 1.52 | 7.67
- United States — 1.31 | 0.24 | 4.86

- Source: Fry (1995) and authors' calculations for the US (annual average continuously compounded rates). Data for the US are from the International Financial Statistics (line 60lc and 64)

### Appendix: the database — coverage and composition
- Database includes data since 1980 for all countries with a population larger than one million people.
- Sample includes 134 countries:
  - 31 high-income countries
  - 26 upper middle-income countries
  - 36 lower middle-income countries
  - 41 low-income countries
- LICs sample comprises the last two groups and excludes emerging markets: China, Colombia, India, Indonesia, Pakistan and Thailand.
- HICs sample includes the first two groups and the 6 countries listed above.
- Regression sample may be smaller due to dataset unbalancedness.
- Table A2 indicates variables constructed relative to trading partners using INS weights.

### Exchange rates
- Real effective exchange rate:
  - Natural logarithm of the CPI-based real effective exchange rate from the International Monetary Fund (INS).
  - Alternative measure computed from PWT variable “P” (captures price levels relative to the U.S.), natural logarithm taken and computed in deviations from trading partners using INS weights.
- Black market premium:
  - BMP = log difference between black market nominal exchange rate and official nominal exchange rate (local currency per US dollar).
  - Annual averages of monthly data from Reinhart and Rogoff (2004).

### Demographics — variable definitions and sources
- Fertility:
  - Total fertility from the UN 2006, expressed as children per woman.
- Infant mortality ratio:
  - Infant mortality rate from the UN 2006, infant deaths per 1,000 live births.
- Old-age dependency ratio:
  - Share of people older than 64 relative to working age (15-64), based on UN data annualized by World Bank; computed in deviations from trading partners.
- Population growth:
  - Computed from WB data, extended with UN projections.

### External sector — definitions and computation
- Current account:
  - Current account to GDP ratio from WEO.
- Net foreign assets (NFA):
  - NFA lagged one year, computed relative to average trade for RER regressions and relative to GDP for current account regressions.
  - Nominal NFA data provided by Gian Maria Milesi-Ferretti.
  - Trade data from WEO or IFS spliced with WEO.
  - NFA in net present value (NPV) terms: add back measure of total debt, subtract present value of debt from World Bank; subtract short-term and long-term private non-guaranteed debt for concessionality adjustment. For countries without PV data, nominal values used.
  - For a few countries, cumulative current account used in place of NFA.
- Aid:
  - Foreign aid relative to GDP, measure computed by Roodman (2006) from OECD DAC data: total net aid minus debt forgiveness plus offsetting entries for debt relief.
  - Data in millions of U.S. dollars, computed relative to WEO nominal GDP in million U.S. dollars.
  - Estimations include concessional loans and net grants (loans = foreign aid minus net grants; net grants = total grants minus debt forgiveness grants).
- Oil balance:
  - Oil balance to GDP ratio from WEO.

### Fiscal variables
- Fiscal balance:
  - General government balance relative to GDP; central government balance used where general balance unavailable. Data from WEO.
- Government consumption:
  - Government consumption relative to GDP from OECD spliced with IFS and WEO.

### Income-related variables
- GDP per capita growth:
  - Growth rate of GDP per capita from WEO.
- Relative productivity:
  - Baseline measure: real PPP GDP per capita from Penn World Tables, extended with World Bank WDI.
- GDP per worker PWT:
  - Real GDP divided by number of workers, from Penn World Tables.
- Relative income per capita:
  - PPP income per capita relative to the US, from World Bank.
- Income groups:
  - Aggregated based on World Bank income group classifications.

### Shocks
- Natural disasters:
  - Data from EM-DAT (CRED), binary dummies for drought, earthquake, extreme temperature, flood, volcano, wave surge, wildfires, windstorm; variable is the average of these indicators.
- Wars:
  - War dummies from PRIO / Uppsala Armed Conflict Dataset, Version 4-2006; four conflict types coded 0,1,2,4 corresponding to no conflict, minor, intermediate, and war (>=1000 deaths per year).
- Goods terms of trade:
  - Natural logarithm of terms of trade of goods (WEO).
- Goods and services terms of trade:
  - Natural logarithm of terms of trade of goods and services (WEO).
- Commodity-based terms of trade:
  - Robustness index: ratio of weighted average price of main commodity exports to main commodity imports.
  - Index for 6 commodity goods as in Ricci et al. (2008).
  - Index for 32 commodities: Aluminum, Bananas, Beef, Cocoa, Coconut oil, Coffee, Copper, Corn, Cotton, Crude oil, Fishmeal, Gold, Hard log, Hides, Iron, Lamb, Lead, Nickel, Palm oil, Rice, Rubber, Shrimp, Soybean meal, Soybean oil, Soybeans, Sugar, Sunflower oil, Tea, Tin, Wheat, Wool, Zinc.

### Structural reforms
- Capital account liberalization:
  - Measure from Research Department’s structural reform database, index normalized between 0 and 1, computed by Dennis Quinn.
  - Alternative measure from Abiad, Detragiache, and Tressel (2008).
- Domestic financial reforms:
  - Index coded between 0 and 1 from Research Department’s structural reform database.
- Price controls:
  - Administered prices: [0,1] dummy for strong presence of price controls in agricultural sector (agricultural product market index; code=3).
  - Max agric. price intervention: [0,1] dummy for presence of marketing boards setting prices (agricultural product market index; code=4).
  - Administered prices CGER: measure of price controls based on share of administered prices in CPI from EBRD (transition economies only).
- Trade restrictions:
  - Data from IMF Research Department’s Structural Reform Database; based on average tariffs and normalized between 0 and 1.
  - Alternative measure from Wacziarg and Welch (2003) extending Sachs and Warner (1995) liberalization years.

### Institutions and political variables
- Constraint on executive:
  - Polity IV constraint on the executive variable (xconst), censored at zero for negative values.
- Violent political transition:
  - Dummy variable equal to 1 whenever constraint on executive is negative, and 0 otherwise.

### Other controls
- Change in private credit to GDP — Source: World Bank.

### Notation (acronyms used in the Appendix)
- EBRD: European Bank for Reconstruction and Development
- IFS: International Monetary Fund’s International Financial Statistics
- INS: International Monetary Fund’s Information Notice System
- PWT: Penn World Tables
- UN: United Nations
- WB: World Bank
- WEO: International Monetary Fund’s World Economic Outlook

*Source: _wp09221 - Appendix tables (IMF PDF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09221.pdf_