Journal Issue
Share
Article

The Landscape of Capital Flows to Low-Income Countries

Author(s):
Thomas Dorsey, Zuzana Brixiova, Sukhwinder Singh, and Helaway Tadesse
Published Date:
February 2008
Share
  • ShareShare
Show Summary Details

I. Introduction

Capital flows to low-income countries are a poorly-understood and poorly-researched topic.1 Most of the attention to international capital flows into developing economies has been focused on emerging markets. Even work on capital flows to developing countries as a whole (i.e., middle- and low-income countries) tends to focus mostly on the large emerging markets, in part because LICs account for a relatively small share (e.g., 5–20 percent) of developing country capital inflows in most asset or liability categories.2 The work that has been done specifically on low-income countries has focused on capital flight from low-income countries and on official flows, with a heavy emphasis on official debt and official debt forgiveness.

Differences in the composition of external flows reinforce the case for taking a separate look at inflows to LICs. For example, debt forgiveness, net transfers (e.g., remittances) and the stock of official debt all account for much higher shares of LIC GDP than they do in other developing countries, but portfolio investment liabilities and the stock of commercial banks debt both account for a much small share of LIC GDP

This paper describes trends in capital and related flows and considers factors behind the observed trends.3Section II describes aggregate capital and related flows to LIC economies as a whole relative to aggregate LIC GDP (i.e., GDP-weighted averages).4Section III considers the applicability of the general cross-country observations to regional and other subgroups of countries (e.g., African LICs, HIPCs, or hydrocarbon exporters).5 Possible causes and policy implications of the trends in capital and related flows are briefly considered in Section IV. The annexes to this paper discuss data definitions, sources, and consistency across sources.

This paper does not attempt to reassess the literature on the causes of private capital inflows nor does it seek to provide detailed policy prescriptions based on the existing literature. Most of the empirical work on capital flows is focused on middle-income countries or a mix of middle-income countries and LICs. Thus, it may not be fully applicable to the different types of capital flows in LICs. Moreover, the quality of data in LICs, particularly on remittances and other private transfers, makes any empirical assessments tentative at best. While the paper provides some basic policy conclusions, it also suggests areas in which future work would be needed to underpin more specific and authoritative conclusions.

II. A Global View of Capital and Related Flows to LICs

Capital and capital-like inflows to LICs have increased strongly over the last quarter century. Total inflows have increased from roughly 4 percent to more than 10 percent of LIC GDP between the 1980s and 2006 and the upward trend in inflows has accelerated in recent years (Figure 1, Tables 1 and 2).6

Figure 1.Official and Private Inflows

percent of GDP
Table 1.All LICs: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy1.31.31.61.51.61.62.02.62.42.52.33.1
Other investment liabilities0.11.10.01.60.70.30.10.70.50.80.1-0.3
Net external borrowing from commercial banks0.0-0.50.00.00.10.00.00.10.10.30.20.3
Liabilities to offical creditors1.00.50.21.21.30.60.70.80.20.30.0-1.0
Other liabilities, incl. trade credits, other loans, & deposits-0.71.20.40.3-0.6-0.4-0.6-0.30.20.2-0.20.4
Portfolio investment liabilities0.50.60.40.00.30.30.40.00.70.80.90.6
Financial account liabilities, total1.93.02.03.12.72.12.53.43.74.13.43.4
Capital-like inflows
Capital account0.50.61.10.60.70.50.60.50.50.50.62.1
debt forgiveness0.21.01.50.10.10.10.20.30.10.10.21.2
other capital account transfers0.4-0.4-0.40.50.50.40.30.20.40.40.40.8
Current transfers3.53.84.03.73.74.04.34.55.04.64.94.9
Private transfers3.23.82.82.52.62.93.03.23.73.43.63.6
Official transfers0.30.01.21.21.01.11.31.31.31.21.31.3
“Capital-like” flows, total4.04.45.14.34.44.64.95.05.65.15.57.0
Total capital and capital-like inflows6.07.57.17.47.06.77.38.49.29.28.910.4
Asset items
Financial account assets, total (non-reserve)-0.3-0.20.0-0.1-0.2-0.5-0.5-0.9-0.5-0.8-0.8-1.4
International reserves0.0-1.1-0.8-0.1-0.4-1.1-1.2-2.3-3.1-3.2-2.2-3.7
Errors and omissions0.90.70.30.1-0.4-0.2-0.50.0-0.30.0-0.4-0.3
Asset items, total0.6-0.5-0.5-0.2-0.9-1.8-2.2-3.2-4.0-4.0-3.4-5.4
Memorandum items
Official inflows, total1.41.52.82.52.51.92.22.41.71.61.51.6
Private inflows, total4.26.34.64.34.04.44.75.87.17.27.08.0
Unallocated capital account0.4-0.4-0.40.50.50.40.30.20.40.40.40.8
Other invest. liab. to official creditors, net of debt forgiveness1.11.51.61.41.50.81.01.10.40.40.20.2
Current account balance-3.1-3.2-2.6-3.6-2.4-0.8-0.8-0.7-0.2-0.6-0.6-0.1
Current account balance, less transfers-6.3-6.9-5.4-6.1-5.1-3.8-3.8-3.9-3.9-4.0-4.2-3.7
Table 2.All LICs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.20.30.91.52.43.1
Other investment liabilities2.21.91.40.70.5-0.3
Net external borrowing from commercial banksn.a.n.a.n.a.-0.10.10.3
Liabilities to offical creditors1.51.41.30.80.4-1.0
Other liabilities, incl. trade credits, other loans, & depositsn.a.n.a.n.a.0.2-0.10.4
Portfolio investment liabilities0.00.00.40.30.60.6
Financial account liabilities, total2.42.22.72.63.43.4
Capital-like inflows
Capital account0.10.40.60.70.52.1
debt forgiveness0.00.00.00.60.21.2
other capital account transfers0.10.30.60.10.30.8
Current transfers1.51.73.23.84.74.9
Private transfers1.11.22.52.93.43.6
Official transfers0.50.50.70.91.31.3
“Capital-like” flows, total1.72.03.94.55.27.0
Total capital and capital-like inflows4.14.26.67.18.610.4
Asset items
Financial account assets, total (non-reserve)0.00.00.0-0.2-0.7-1.4
International reserves0.20.1-0.4-0.7-2.4-3.7
Errors and omissions0.40.5-0.10.1-0.2-0.3
Asset items, total0.60.6-0.6-0.8-3.4-5.4
Memorandum items
Official inflows, total2.01.92.12.31.91.6
Private inflows, total1.92.03.94.86.48.0
Unallocated capital account0.10.30.60.10.30.8
Other invest. liab. to official creditors, net of debt forgiveness1.51.41.41.30.60.2
Current account balance-3.1-3.2-2.8-2.5-0.6-0.1
Current account balance, less transfers-4.2-4.3-5.3-5.5-4.0-3.7

Box 1.Data and Terms Used in this Note

Capital inflows: This note retains the conventional use of the term “capital inflows” to refer to net changes in external liabilities (e.g., FDI, portfolio flows, and loans).

Capital-like inflows: Other balance of payments flows may be substitutes for financial account flows to varying degrees. For example, official transfers (i.e., grants) and debt forgiveness can substitute for official loans, the use of reserves can substitute for new borrowing, and private transfers (e.g., remittances) are substitutable for ODA in some circumstances. This note uses the umbrella term “capital-like flows” to cover: current transfers (official and private) in the current account, and debt forgiveness and other capital transfers in the capital account. To the extent that underlying data permit, the note systematically distinguishes between private and official components.

Asset outflows: The assessment of the impact of capital inflows is influenced by the counterpart items in the balance of payments. For this paper, these asset transactions are divided into changes in international reserves and the accumulation of other financial account assets. Net errors and omissions are sometimes considered as the reflection of disguised or unrecorded capital flows and are also included under the heading of asset outflows in this paper, although many factors other than assets outflows may be contributing to net errors and omissions.

The relationship between these terms and the terminology in the Balance of Payments Manual are reviewed in more detail in Annex I.

The increase in inflows has not been uniform; a rapid increase in inflows from private sources has coincided with comparative stability in official-source inflows. Private inflows (e.g., FDI and remittances) have accelerated rapidly, rising fourfold from about 2 percent of LIC GDP in the 1980s to more than 8 percent by 2006. In contrast, net official-source inflows (loans, grants, and debt forgiveness from bilateral governments and international financial institutions) have averaged slightly more than 2 percent of LIC GDP per year since 1981, peaking at about 3.5 percent of LIC GDP in 2000 and declining back to about 1.6 percent of GDP by 2006.

The net result of these trends has been a steady shift toward private sources of inflows for LICs. Private and official inflows were roughly equal through the 1980s. Private inflows had risen to roughly double the scale of official inflows by the late 1990s, and to roughly four times the size of official inflows by 2006.

A. Official Financing for LICs

There is a pronounced trend within official flows toward the substitution of grants for loans (Figure 2, Tables 1 and 2). Consistent with announced donor policies, official current transfers (i.e., grants) have tripled from roughly 0.5 percent of LIC GDP in the early 1980s to 1.5 percent of GDP in 2006. Mirroring this, official lending to LICs (net of debt forgiveness) has declined from an inflow of about 1.5 percent of LIC GDP through the 1980s and 1990s to an outflow of about 0.5 percent of LIC GDP by 2006. To the extent that some of the recorded debt forgiveness reflects liabilities to private creditors, the decline in official other investment flows would be even larger (see Box 2).

Figure 2.Official Inflows to LICs

percent of GDP

Box 2:Debt Forgiveness

Balance of payments data record debt forgiveness as equivalent to amortization in the financial account. For example, the change in “other liabilities to official creditors” (i.e., loans) is -5.9 percent of GDP for African LICs in 2006. Debt forgiveness for African LICs in 2006 is 4.9 percent of GDP. When the debt forgiveness is “added back” to the change in other liabilities to official creditors, the result is the item “other liabilities to official creditors, net of debt forgiveness” equal to -1.0 percent of GDP (Tables 1 and 2). This “net” item is included as a proxy for net lending by official creditors in Figures 2, 8, 12, and 16. Both the net data and the components of the net figures are shown in each of the tables.

However, official lending net of debt forgiveness is an imperfect proxy for net lending. WEO data have a single entry for debt forgiveness. This could apply to debt forgiveness on loans from official sources, loans from commercial banks, other loans, or debt securities. Thus, netting the entire stock of debt forgiveness against loans from official sources may overstate the net lending from official sources and understate net lending from other sources to the extent that debt forgiveness is also reflected in a reduction in commercial bank or portfolio investment debt.

B. Private Financing for LICs

Private-source inflows to LICs have grown more than fourfold since the 1980s (Figure 3, Table 1 and 2). While there are many sources of private sector inflows, the inflows to LICs have been overwhelmingly non-debt-creating and have been dominated by FDI and private transfers (e.g., remittances).7

Figure 3.Private Inflows

percent of GDP

The most striking aspect of the shift in capital flows to LICs is the more than ten-fold increase in FDI as a share of LIC GDP between the 1980s and 2006. FDI inflows averaged only 0.2 percent of LIC GDP in the early 1980s, but rose steadily to more than 3 percent of GDP by 2006.

Only slightly less striking is the rise in private transfers. Private current transfers more than tripled as a share of LIC GDP since the 1980s, rising from 1.1 percent of GDP in the early 1980s to 3.6 percent of GDP in 2005. In principle, private transfers can be decomposed into workers’ remittances and other (likely smaller) types of private transfers. However, data problems preclude such a decomposition for LICs. Also, other categories that might contain remittances fall outside private transfers, but these are also difficult to disentangle from the data. In the absence of better alternatives, private transfers is used as a proxy for remittances in this paper; because the biases associated with data problems go in opposite directions, it is difficult to determine whether using private transfers understates or overstates total remittances (see Box 3).

Private inflows other than FDI and private transfers averaged only about 0.5 percent of LIC GDP in the 1980s and 1990s, but rose just above 1 percent of GDP in the 2003–06 period. Flows in this group have been volatile, but there appears to be a shift from unsecuritized debt (e.g., commercial bank loans and trade credit) to securitized portfolio investment within this category. Commercial bank loans and other private, debt-creating liabilities of LICs have risen over the last quarter century, but they have done so from a very low base, and they remain a minor component of aggregate private inflows to LICs.

Given the large scale and rapid rise in combined FDI and private transfer flows, some consideration of whether these data reflect changes in actual inflows is in order. While there are reasons to question the reliability of the FDI data (see Bhinda et al, 1999) and especially the private transfers/remittances data (see Box 3), there is less reason to suspect a systematic bias. Consider first the possibility that previously unrecorded transactions have gradually come “on the books.” Other things equal, this should produce an offsetting shift in net errors and omissions of similar scale but opposite sign (i.e., in excess of 5 percent of LIC GDP). In fact, net errors and omissions have shifted only modestly over the period and have rarely exceeded 0.5 percent of LIC GDP in either direction. Another possibility is that some other aggregate has been gradually reclassified as either FDI or private transfers. In the case of private transfers, the most likely candidate would be a decline in recorded income receipts as it includes compensation of employees (the main remittance item not included in private transfers). However, income receipts for LICs has moved in a narrow range between 0.01 and 0.80 percent of LIC GDP between 1981 and 2006 and reached its peak value in 2006, the opposite direction from what would be expected if data on compensation of employees were being redefined as private transfers over time. There are many other possibilities for error in admittedly weak data, but the comparisons cited above and similar cross-checks do not reveal any clear candidates for unrecorded or misclassified data.

Assuming that the FDI and private transfers data can be taken at something close to face value, the causes of such sharp shifts merit examination. The factors behind the increases in FDI and private transfers may overlap, but are likely to differ in significant degree.

Policies directly affecting investment include privatization and explicit restrictions on FDI. Privatization in LICs or the opening of the economy to foreign acquisition of existing firms may create one-off opportunities for FDI, and success by foreign acquiring firms may encourage further investment in such enterprises. Alternatively, some LICs restrict investment by foreign firms even in new enterprises and may chose to open up FDI opportunities by sectors or on the basis of other criteria. Such policies are hard to quantify, but there appears to have been trend toward liberalization in recent years. India provides an example of a country making explicit choices in terms of these policies in a liberalizing direction (Reddy, 2007).

The increase in FDI inflows may also be due to improvements in the investment environment more generally. Many classes of policies have a strong effect on the environment for foreign investors, most notably trade policies, policies affecting the ease and cost of setting up and continuing business operation, and other policies cataloged in the many indices of investment environments. As with policies on direct regulation of FDI, there has been a trend toward liberalization of trade and other aspects of the business environment in LICs that coincides with the increase in FDI. This suggests that the recent emphasis on improving the business environment is appropriate and timely (Busse and Groizard, 2006; Bénassy-Quéré et al., 2007; and Naude and Krugell, 2007).

Broader policy considerations may also play a role in encouraging investment in LICs. Studies of factors behind FDI often emphasize the importance of a stable macroeconomic and political environment. The improvement in the macroeconomic policies of LICs in the last decade in terms of fiscal policies, inflation, and external balance (see Selassie et al., 2006) has presumably been an encouragement to FDI. Non-economic factors such as political stability and lessened risk of civil conflict or expropriation are harder to quantify, but have undoubtedly contributed to greater FDI to the extent that these factors have improved in individual countries.

Economic developments elsewhere in the world economy may also have contributed to increased FDI. The decline in yield on investments in advanced and emerging market economies has led to a search for new and higher yielding opportunities outside of traditional investment markets. The LICs include many of the “frontier” markets that some investors have looked toward in the search for yield. Apart from the search for yield, the rapid increase in assets in sovereign wealth funds of resource exporters and other emerging markets has added additional impetus to the hunt for new investment opportunities in under-explored venues such as the LICs.

Assessing the causes of increased private transfers/remittances is more complicated. Aside from the manifest data problems, the factors underpinning remittances are less studied than those on the factors underpinning FDI. There is a considerable recent literature on financial institution-level transaction costs and impediments to remittance transfers that suggests measures to facilitate remittances, although mostly drawn on the experience of (im)migrants from middle-income countries, in particular immigrants from Latin America to the United States (Suro et al, 2002; Fajnzilber et al. 2007). Reflecting the data difficulties in this area, much of the research is drawn from private survey data of migrants, money transfer firms, and others involved in remittances from the United States rather than official balance of payments data. More work would be needed to assess the applicability of this literature to LICs and to countries in other parts of the world.

The relationship between remittances and the policy environment is more complicated than for FDI. LICs with better educational systems, the absence of confiscatory foreign exchange practices for household-level transactions, and perhaps more efficient and less financially-repressed banking systems presumably all contribute to having higher earning migrants more willing to remit income to their home countries and thus higher levels of remittances. In particular, remittances appear to be deterred by exchange restrictions and black market premia and—to some extent—unstable macroeconomic environments (International Monetary Fund, 2005a). On the other hand, the source of remittances is a population of migrants often induced to leave their home countries by the absence of employment opportunities and economic stagnation more generally, civil conflict, or the breakdown in law and order. In this context, the effects of an improvement or deterioration in the policy environment are ambiguous and may differ in sign between the short run and the longer term.

The evidence on the impact of remittances on recipient economies is mixed. There is a broad consensus that remittances are less volatile than other private and official capital flows (Buch and Kuckulenz, 2004; International Monetary Fund, 2005a) and that remittances tend to move counter-cyclically with recipient country income (Chami et al., 2005; International Monetary Fund, 2005a). However, studies differ on the macroeconomic effects of remittances. Some have found a positive impact on domestic investment (Fajnzylber and Lopez, 2007), while others have found no impact on investment (Chami et al. 2005). In some studies, the impact on growth is assessed to be positive (Fajnzylber and Lopez, 2007), while others find no clear relationship between remittances and growth (Buch and Kuckulenz, 2004). However, all of these studies are based on a sample of mostly middle-income countries or a mix of middle-income countries and LICs, and thus the results may not be fully applicable to LICs.

Box 3.Remittances in BOP Data

Remittances, roughly speaking, are household-levels payments across international borders, generally between family members. Remittances are poorly defined in balance of payments statistics as they are spread over multiple categories of Balance of Payments Manual, Fifth edition (BPM5) transactions and are generally only part of any one category. There are also serious measurement issues related to data availability and the practicality of making some of the distinctions needed for the BPM5 classification.

Remittances are distributed across three main items in BOP data, none of which consists exclusively of remittances.

  • Private transfers is the most important of these BOP items, and workers’ remittances account for three quarters of private transfers in the available BOPSY data for LICs. Workers remittances data should include all worker income transfers derived from paid employment from migrants to one economy to residents of another economy. However, similar private transfers such as remittances from self-employment income are not included in workers remittances and are instead combined with transfers such as private charitable donations. Because no usable disaggregation is available, this entire category is treated as a proxy for worker remittances in this paper.

  • Compensation of employees is a component of the income account and included paid employment income transfers from workers living temporarily in an economy (i.e., for less than one year). However, in WEO data, this item is subsumed in income credits together with investment income, and net income is not disaggregated into income credits and income debits for some countries. Because of these data limitations, discussion of remittances data in this paper does not include the contribution of compensation of employees included in income credits or net income.

  • Migrant transfers are a component of the capital account, encompassing the assets that migrants bring with them when moving from one country of residence to another. However, the WEO data disaggregate capital transfers into debt forgiveness and all other capital transfers. As with compensation of employees, no attempt is made to estimate the fraction of the latter category that should be assigned to remittances.

Reinke (2007) concisely summarizes many of these problems as well as proposals for more analytically useful and practical classification schemes in future editions of the balance of payments manual. Apart from the data limitations discussed above, he notes that practitioners report that the distinctions between temporary and permanent residents and between remittances derived from wage and self-employment income are neither practical nor analytically useful. On measurement, among many problems, he notes with regard to remittances transferred through ATM cards that “it would seem very unlikely that such transactions are accurately recorded.” As a result of these data problems, studies of remittances frequently use survey data rather than BOP data and estimates (e.g., Suro et al., 2002).

C. Capital Going the Other Way—LIC Holdings of Foreign Assets

Much of the relatively thin literature on capital flows to low-income countries focuses on other international flows that might represent unrecorded, misclassified, or disguised capital transactions. In this context, various capital or financial account items have been explicitly or implicitly defined as capital flight together with errors and omissions to create a residual estimate of capital flight to and from low-income countries. Given the extent of the data problems with low-income country capital and financial account information, the disparate definitions of capital flight (Box 4), and the often complicated edifices of assumptions and data splices underpinning these estimates, this note does not attempt to define capital flight. Instead, the tables and charts report net errors and omissions, reserve accumulation, and other financial account asset items, without attempting to assess whether these should be taken at face value or treated as disguised capital flight.

Asset outflows were modest and variable in sign during the 1980s and 1990s, but they shifted toward a persistent and accelerating accumulation of official reserves from 2000 (Figures 4, Table 1 and 2). In every year from 2002 through 2005, reserve accumulation as a share of LIC GDP exceeded that of any of the preceding 12 years (Figure 4). Also noteworthy is the fact that reserve accumulation is roughly equal to total official inflows in 2002 and substantially in excess of total official inflows in 2003–05. The stock of reserves in LICs rose from the equivalent of 3.1 months of imports of goods and services in 1995 to 6.5 months by 2006. This self-insurance may be a contributing factor to the diminishing other investment inflows as the higher reserve levels allow LICs to forego borrowing in the event of shocks.

Figure 4.All LICs, Asset Outflows

percent of GDP

Other asset outflows do not provide much basis for concern about large scale capital flight (Tables 1 and 2). Recorded asset outflows other than the accumulation of international reserves are generally small. While net errors and omissions are sometimes cited in the literature as a reflection of unrecorded or disguised capital flows, these do not appear to be a substantial net contributor to “apparent” capital outflows. As noted above, they have never exceeded 0.5 percent of LIC GDP in either direction since 1996 (Table 1). However, it is possible that trends in unrecorded capital movements in these flows may be masked by offsetting trends in other balance of payments items.

Box 4.Capital Flight In Low-Income Countries

Much of what literature is available on capital flows to low-income countries focuses on capital flight. Particularly among those studies focusing on Africa, estimates of capital flight are often very large, to the point where some estimates indicate that Africa is a net creditor region due to capital flight in excess of external lending (Boyce and Ndikumana, 2001).

There is not an agreed definition of capital flight in the literature; indeed much of the literature distinguishes itself from earlier studies by devising new measures of the concept. What most measures of capital flight have in common is an approach of taking some elements of the balance of payments at face value, and explicitly or implicitly defines the rest as flight capital. For example, one common approach defines capital flight as the change in debt stock, plus FDI, less the current account deficit and the change in official reserves (World Bank, 1985). Thus by implication, portfolio equity liabilities, non-reserve financial account assets, the entire capital account as defined by BPM5, and errors and omissions are all part of capital flight. Other estimates use c.i.f./f.o.b. differentials from trade data to impute an additional element of estimated capital flight, adjust changes in debt stocks for exchange rate changes, compound imputed cumulative flows by using a proxy rate of return on the estimated stock of flight capital (Boyce and Ndikumana, 2001 and 2002), or deduct foreign assets of domestic commercial banks (Hermes, 2002)

Most of these approaches share two problems: (i) they use a residual approach to measuring capital flight in a data environment in which there are many other sources of errors and mis-measurement, such as smuggling or poor data compilation; and (ii) they tend to mix data from different sources (e.g., IMF data for flows, World Bank data for debt stocks, and LIC and partner country trade data for c.i.f./f.o.b. differentials), adding an additional element of noise to calculations (see Annex for details on the inconsistency between some of these data sources). Approaches that take an internally consistent set of data (e.g., Bosworth, et al., 1999) are the exceptions to the rule.

D. Policy Concerns: Current Accounts, Reserves, Sustainability, and Vulnerability

Large and increasing capital inflows often give rise to concerns that the inflows are financing unsustainable current account deficits or contributing to vulnerability to capital account crises. An approach to assessing such risks in the context of emerging market economies has been to compare current account deficits to capital inflows, particularly debt-creating inflows, and consider whether current account deficits are creating a demand for capital account financing that might not be sustainable (Ghosh et al., 2007). For the longer-term, the effect of the compounding of interest payments on the debt-creating components of capital inflows may also be important. A key related indicator of vulnerability is the trend in official reserve accumulation or loss and how these reserve flows relate to the current account balance.

LICs seem to have modest and declining risks based on comparisons of the current account to capital inflows or reserves (Figure 5). Prior to 2000, LICs as a group ran modest current account deficits, financed by larger and rising capital and capital-like inflows. Net reserve accumulation was near zero until 1995. However, the aggregate current account deficit of LICs shrank almost disappeared from 2000 onward and dropped substantially even when transfers are excluded from the current account, and reserve accumulation moved to steadily increasing inflows (negative values in Figure 5).8

Figure 5.Current Account Balance and Reserve Flows

percent of GDP

The different nature of the inflows to LICs relative to those to emerging markets also suggests somewhat benign conclusions on vulnerability and sustainability for LICs. The key difference is that very little of the current account financing for LICs has so far been debt-creating, as it is dominated by FDI and transfers. The fact that outflows consist primarily of official reserve accumulation also suggests that there is less basis for concern about the controllability of outflows than for the authorities than in emerging markets.

III. The Diversity of LICs and Their Capital Flows

LICs are united by little other than their low per-capita incomes. They vary widely in size, and they can be found in Africa, Asia, the Caribbean, Europe, Latin America, and the South Pacific. At one extreme, India has a more than a billion people and a 2006 GDP of $772 billion; it is classed with Brazil, Russia, and China as one of the “BRIC” major emerging markets. Toward the other end of the spectrum, Dominica has about 70,000 people and a GDP of $0.02 billion. LICs also differ markedly in per capita income, with 2006 levels ranging from $100 to more than $2,000 (and as much as $5,000 for certain small island economies). The production side of LIC economies is similarly diverse and includes oil exporters (e.g., Nigeria), remittance-recipient economies (e.g., Albania), emerging manufacturing powers (e.g., Vietnam), and tourist-based economies (e.g., the Maldives). There are also differences induced by debt forgiveness policies: LICs include pre-completion point HIPCs, post-completion point HIPCs, and LICs that are not HIPCs.

The heterogeneity of LICs implicitly raises the question as to whether conclusions derived from aggregate results apply across the full range of LICs. Groups of LICs differ because of history and other factors in ways that have implications for economic analysis and performance. The persistent robustness of Africa and other regional dummy variables in explaining economic growth differences is one example of this (Sala-i-Martin, 1997; Barro and Sala-i-Martin, 1995). The international community has made policy decisions to alter capital flows and the incentives for capital flows that differ across country groups in the context of the HIPC initiative. Given that the intention of the HIPC initiative is in part to change capital flows, it would be surprising if HIPC status did not imply some difference in capital flows across LICs. Mineral endowments may also be important. The high costs and technological sophistication of oil, gas, and other mineral extraction and price swings in commodities presumably have implications for FDI inflows, and the “resource curse” literature suggests effects on other inflows and outflows.

The next three sub-sections examine whether the conclusions outlined above for all LICs are applicable to subgroups of LICs differentiated by region, HIPC status, and mineral endowment. The coverage is necessarily selective in light of the many dimensions across which LICs differ from each other (Annex B present some additional sensitivity analysis). However, the persistence of the main patterns across different country groups suggests that the trends in inflows to LICs is broad-based and is not an artifact of some other trends affecting some subset of LICs.

A. Regional Patterns: Is India Driving These Results? Does this Apply to Africa?

LIC GDP is unevenly divided across regions of varying size. India accounts for roughly half of LIC GDP, and African LICs and South and East Asian LICs (other than India) each account for roughly one fifth of LIC GDP (Figure 5).9 LICs in all other regions (the Caribbean, the Caucasus region, southeastern Europe, Central Asia, Latin America, the Middle East, and the Pacific islands) account for only seven percent of LIC GDP.

Figure 6.2006 LIC GDP by Region

Capital flows across the three larger regions all exhibit rising private inflows and relatively steady official inflows (Figure 7).10 Private flows have surged in Africa, India, and other South and East Asian LICs in broadly similar proportions. Official flows started at very different levels in the three regions and have grown roughly in line with GDP.

Figure 7.Total, Official, and Private Inflows by Region

percent of GDP

Inflows from official sources are consistently higher in African LICs than in South and East Asian LICs, and official capital inflows to India are much lower than either of the other groups. (Figure 7, Tables 3-8). Inflows from official sources have been volatile, but they have averaged about 4 percent of GDP in African LICs, and about 2 percent of GDP in South and East Asian LICs other than India, and have never exceeded 1.0 percent of GDP in India, except at the time of India’s 1991 financial crisis.

Table 3.India: 1995-2006percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy0.60.70.90.60.50.81.21.10.80.90.91.9
Other investment liabilities-0.31.20.71.31.21.10.21.41.01.40.72.7
Net external borrowing from commercial banks0.10.00.00.00.10.2-0.10.20.30.40.20.7
Liabilities to offical creditors-0.10.10.10.10.20.10.10.1-0.80.00.30.2
Other liabilities, incl. trade credits, other loans, & deposits-0.31.10.61.20.90.80.21.11.51.10.21.8
Portfolio investment liabilities0.51.10.6-0.10.50.50.60.21.41.41.61.0
Financial account liabilities, total0.83.02.21.82.22.32.02.83.23.73.15.5
Capital-like inflows
Capital account0.00.00.00.00.00.00.00.00.00.00.00.0
debt forgiveness0.00.00.00.00.00.00.00.00.00.00.00.0
other capital account transfers0.00.00.00.00.00.00.00.00.00.00.00.0
Current transfers2.43.13.42.52.72.93.13.23.83.03.03.0
Private transfers2.33.03.32.42.62.83.03.13.72.93.03.0
Official transfers0.10.10.10.10.10.10.10.10.10.10.00.0
“Capital-like” flows, total2.43.13.42.52.72.93.13.23.83.03.03.0
Total capital and capital-like inflows3.26.15.64.24.95.25.16.07.06.66.18.5
Asset items
Financial account assets, total (non-reserve)0.0-0.10.00.00.0-0.1-0.3-0.3-0.3-0.3-0.3-1.0
International reserves0.6-0.7-1.1-0.6-1.4-1.3-1.8-3.8-4.5-3.5-1.9-3.5
Errors and omissions0.2-0.5-0.30.5-0.10.1-0.20.00.10.10.10.1
Asset items, total0.8-1.3-1.5-0.1-1.5-1.3-2.3-4.2-4.7-3.8-2.1-4.4
Memorandum items
Official inflows, total0.00.20.20.10.20.10.20.2-0.70.00.30.2
Private inflows, total3.25.95.44.14.75.14.95.87.76.65.88.3
Unallocated capital account0.00.00.00.00.00.00.00.00.00.00.00.0
Other invest. liab. to official creditors, net of debt forgiveness-0.10.10.10.10.20.10.10.1-0.80.00.30.2
Current account balance-1.6-1.7-0.7-1.7-0.7-1.00.31.41.50.1-1.0-1.1
Current account balance, less transfers-3.9-4.8-4.1-4.1-3.4-3.9-2.8-1.8-2.3-2.9-4.0-4.1
Table 4.India: 1981-2006percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.00.10.20.71.01.9
Other investment liabilities1.52.01.31.11.02.7
Net external borrowing from commercial banks0.40.80.20.10.20.7
Liabilities to offical creditors0.60.70.80.1-0.10.2
Other liabilities, incl. trade credits, other loans, & deposits0.60.40.20.90.81.8
Portfolio investment liabilities0.00.00.60.51.01.0
Financial account liabilities, total1.52.12.12.32.95.5
Capital-like inflows
Capital account0.00.00.00.00.00.0
debt forgiveness0.00.00.00.00.00.0
other capital account transfers0.00.00.00.00.00.0
Current transfers1.41.01.92.93.23.0
Private transfers1.20.91.82.83.23.0
Official transfers0.20.20.10.10.10.0
“Capital-like” flows, total1.41.01.92.93.23.0
Total capital and capital-like inflows2.93.14.05.26.28.5
Asset items
Financial account assets, total (non-reserve)0.00.00.50.0-0.3-1.0
International reserves0.00.3-1.1-1.0-3.1-3.5
Errors and omissions0.1-0.1-0.4-0.10.00.1
Asset items, total0.10.2-1.0-1.1-3.4-4.4
Memorandum items
Official inflows, total0.80.90.90.20.00.2
Private inflows, total2.12.23.15.06.28.3
Unallocated capital account0.00.00.00.00.00.0
Other invest. liab. to official creditors, net of debt forgiveness0.60.70.80.1-0.10.2
Current account balance-1.6-2.3-1.1-1.20.5-1.1
Current account balance, less transfers-3.0-3.3-3.0-4.1-2.8-4.1
Table 5.African LICs 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy1.71.62.52.73.73.04.05.65.45.65.35.7
Other investment liabilities1.6-2.1-0.42.71.5-0.4-0.4-0.1-0.7-0.9-2.3-8.3
Net external borrowing from commercial banks-0.3-2.30.0-0.5-0.1-0.3-0.3-0.1-0.10.50.3-0.2
Liabilities to offical creditors1.9-0.4-0.62.92.31.70.82.21.10.6-1.6-5.9
Other liabilities, incl. trade credits, other loans, & deposits0.00.60.10.2-0.6-1.8-0.9-2.1-1.7-1.9-1.0-2.3
Portfolio investment liabilities0.70.40.40.50.40.20.3-0.10.10.10.30.3
Financial account liabilities, total4.0-0.12.45.85.72.93.95.44.84.83.4-2.3
Capital-like inflows
Capital account1.23.34.01.91.91.52.01.61.41.71.77.7
debt forgiveness0.22.63.40.40.50.41.01.00.70.30.94.9
other capital account transfers1.00.60.61.51.31.11.00.60.71.40.82.8
Current transfers4.43.73.94.03.74.14.24.04.95.35.24.6
Private transfers1.31.32.02.01.92.42.12.22.73.33.43.0
Official transfers3.02.42.02.11.81.82.11.82.11.91.91.6
“Capital-like” flows, total5.67.07.96.05.55.76.25.66.27.06.912.3
Total capital and capital-like inflows9.76.910.411.811.28.610.111.011.111.810.310.0
Asset items
Financial account assets, total (non-reserve)-0.2-0.5-0.1-0.1-0.2-0.5-0.1-1.7-0.7-1.4-1.6-3.2
International reserves-1.0-2.6-0.91.01.0-2.7-0.70.3-0.9-4.9-4.7-5.9
Errors and omissions1.12.3-1.8-1.0-0.8-0.9-1.50.3-0.61.30.51.2
Asset items, total-0.1-0.8-2.7-0.10.0-4.2-2.4-1.1-2.2-5.0-5.7-8.0
Memorandum items
Official inflows, total5.14.64.85.44.63.93.95.03.92.81.20.6
Private inflows, total3.51.65.04.95.33.65.25.46.57.68.36.5
Unallocated capital account1.00.60.61.51.31.11.00.60.71.40.82.8
Other invest. liab. to official creditors, net of debt forgiveness2.02.22.93.32.82.11.83.21.70.9-0.7-1.0
Current account balance-5.2-2.4-3.7-7.7-7.5-0.2-3.5-5.9-4.0-1.50.42.4
Current account balance, less transfers-9.5-6.1-7.7-11.7-11.2-4.4-7.7-9.9-8.9-6.8-4.8-2.2
2/ Disaggregated data on transfers unavailable for Comoros (1981-83) and Sudan (1981-91).
Table 6.African LICs: 1981-2006percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.40.91.32.75.25.7
Other investment liabilities3.32.41.60.3-0.9-8.3
Net external borrowing from commercial banks-0.23.4-0.5-0.60.1-0.2
Liabilities to offical creditors2.42.61.51.20.6-5.9
Other liabilities, incl. trade credits, other loans, & deposits1.2-3.60.6-0.3-1.5-2.3
Portfolio investment liabilities0.20.10.40.40.10.3
Financial account liabilities, total3.93.43.33.44.5-2.3
Capital-like inflows
Capital account0.51.11.62.51.77.7
debt forgiveness0.00.20.41.50.84.9
other capital account transfers0.40.91.11.00.92.8
Current transfers1.22.74.23.94.74.6
Private transfers-0.30.51.11.92.73.0
Official transfers1.52.23.22.02.01.6
“Capital-like” flows, total1.63.85.86.46.412.3
Total capital and capital-like inflows5.57.29.19.810.810.0
Asset items
Financial account assets, total (non-reserve)0.00.40.0-0.3-1.1-3.2
International reserves0.90.20.2-0.8-2.2-5.9
Errors and omissions-0.40.00.7-0.40.01.2
Asset items, total0.60.60.9-1.6-3.3-8.0
Memorandum items
Official inflows, total3.84.95.14.63.40.6
Private inflows, total1.31.42.94.16.66.5
Unallocated capital account0.40.91.11.00.92.8
Other invest. liab. to official creditors, net of debt forgiveness2.42.81.92.71.4-1.0
Current account balance-4.9-5.2-5.8-4.3-2.92.4
Current account balance, less transfers-6.1-7.9-10.1-8.2-7.6-2.2
1/ Includes all LICs on the African continent plus Cape Verde, the Comoros, Madagascar, and Sao Tome and Principe (excluded from the countries on the African continent are the non-LICs Algeria, Botswana, Egypt, Equatorial Guinea, Gabon, Libya, Morocco, Namibia, South Africa, and Swaziland, and Tunisia). Data are unavailable for Eritrea prior to 1992 and Liberia and Somalia for all years due to lack of data; excludes Mozambique for all years due to large reclassifications of assets between private and official sources.2/ Disaggregated data on transfers unavailable for Comoros (1981-83) and Sudan (1981-91).
Table 7.South and East Asia, excl. India: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy1.72.01.71.51.41.21.31.81.71.72.12.6
Other investment liabilities-0.21.81.20.9-1.4-1.40.2-0.30.0-0.10.50.9
Net external borrowing from commercial banks0.00.00.00.00.00.00.00.00.00.00.00.0
Liabilities to offical creditors0.91.11.22.22.31.51.81.51.00.31.31.1
Other liabilities, incl. trade credits, other loans, & deposits-1.10.70.0-1.3-3.7-2.9-1.6-1.8-1.0-0.5-0.8-0.2
Portfolio investment liabilities0.70.10.20.00.1-0.3-0.1-0.2-0.10.10.20.3
Financial account liabilities, total2.23.93.12.50.1-0.51.31.31.61.62.83.8
Capital-like inflows
Capital account1.20.60.60.60.50.60.60.51.10.40.80.8
debt forgiveness0.00.00.00.20.00.00.00.00.00.00.00.0
other capital account transfers1.20.60.60.40.50.60.60.51.10.40.80.8
Current transfers4.64.74.95.04.95.66.07.07.06.77.58.0
Private transfers3.53.84.04.34.04.55.15.86.26.17.07.4
Official transfers1.10.90.90.70.91.10.91.30.90.60.50.6
“Capital-like” flows, total5.85.35.55.65.46.16.67.58.17.18.38.8
Total capital and capital-like inflows8.09.28.68.15.55.67.98.89.78.711.212.6
Asset items
Financial account assets, total (non-reserve)-0.10.20.0-0.3-0.2-0.2-0.8-1.10.0-0.20.00.7
International reserves0.00.30.70.10.91.5-0.4-1.7-3.1-1.0-0.7-2.0
Errors and omissions1.00.5-0.5-0.2-0.2-0.6-0.70.1-0.4-0.6-1.9-1.4
Asset items, total0.91.00.1-0.50.60.7-1.9-2.8-3.5-1.8-2.6-2.7
Memorandum items
Official inflows, total2.02.12.03.13.22.52.72.81.90.91.71.7
Private inflows, total4.86.55.94.61.72.54.75.56.87.48.610.1
Unallocated capital account1.20.60.60.40.50.60.60.51.10.40.80.8
Other invest. liab. to official creditors, net of debt forgiveness0.91.11.22.42.31.51.81.51.00.31.31.1
Current account balance-4.3-5.5-3.9-2.6-1.2-0.80.01.00.8-0.3-1.1-1.9
Current account balance, less transfers-8.9-10.2-8.7-7.6-6.0-6.4-6.0-6.0-6.3-7.0-8.6-9.9
Table 8.South and East Asia, excluding India: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.10.21.11.61.72.6
Other investment liabilities3.32.91.60.20.10.9
Net external borrowing from commercial banks0.10.10.10.00.00.0
Liabilities to offical creditors1.70.81.01.71.21.1
Other liabilities, incl. trade credits, other loans, & deposits1.42.10.5-1.4-1.1-0.2
Portfolio investment liabilities0.00.10.30.00.00.3
Financial account liabilities, total3.43.23.01.81.73.8
Capital-like inflows
Capital account0.20.61.40.60.70.8
debt forgiveness0.00.00.10.00.00.0
other capital account transfers0.20.61.30.60.70.8
Current transfers1.92.24.65.06.98.0
Private transfers1.51.73.54.16.07.4
Official transfers0.40.51.10.90.80.6
“Capital-like” flows, total2.12.86.05.67.58.8
Total capital and capital-like inflows5.56.08.97.49.312.6
Asset items
Financial account assets, total (non-reserve)-0.1-0.1-0.3-0.1-0.40.7
International reserves-0.10.0-0.40.7-1.4-2.0
Errors and omissions1.40.8-0.1-0.2-0.7-1.4
Asset items, total1.30.7-0.80.4-2.5-2.7
Memorandum items
Official inflows, total2.11.22.22.62.01.7
Private inflows, total3.24.15.54.36.610.1
Unallocated capital account0.20.61.30.60.70.8
Other invest. liab. to official creditors, net of debt forgiveness1.70.81.11.71.21.1
Current account balance-3.4-4.5-3.6-2.80.1-1.9
Current account balance, less transfers-6.8-6.7-8.2-7.8-6.8-9.9

The composition of official flows has also varied across regions and over time. Inflows to African LICs have been dominated by grants rather than loans from the early 1990s, even after netting out the impact of debt forgiveness on official loans (Table 5 and 6). In South and East Asia, official loans have exceeded grants from the early 1980s to the present, and debt forgiveness has been unimportant. In India, the modest amount of official-source inflows has consistently taken the form of loans.

Private source inflows into LICs reveal much more similarity across regions, with a pronounced and accelerating trend in all three. Private inflows as a share of GDP are similar across regions over time and in terms of their rates of increase. However, the composition differs markedly across regions. FDI is the most important source of inflows to the African LICs, running at roughly twice the share of GDP as private transfers and at roughly two thirds of total private inflows. The pattern is reversed in South and East Asian including India with private transfers running at roughly three times the level of FDI since the early 1990s. Notwithstanding the differences in the relative importance of private transfers and FDI, both components are increasing strongly in all three regions.

Asset outflows present a more mixed picture. All three regions had an acceleration in reserve accumulation with broadly similar levels and trends across the three regions (albeit with greater volatility in African LICs). Other asset flows present a more mixed, but still largely benign picture. Non-reserve financial account assets have been trending toward net outflows in India and even more so in African LICs in the last few years (Tables 34 and 56 respectively) but relatively neutral in South and East Asian LICs other than India. On the other hand, errors and omissions have been trending toward net inflows in African LICs and net outflows in South and East Asian LICs over the same period, with Indian errors and omissions consistently near zero. Consolidating net outflows with errors and omissions yields broadly similar patterns of non-reserve outflows across the three regions.

Regional patterns of reserve accumulation relative to current account balances give little reason for concern (Figure 8). Reserve movements have shifted from a relatively neutral stance to substantial accumulation in all three regions. Current account balance trends also appear benign. African LICs have shifted sharply from current account deficits to surpluses in recent years. India and other South and East Asian LICs have shown only modest movements in current account deficits, with both surpluses and deficits that are small a shares of regional GDP.

Figure 8.Private Transfers, FDI, and Official Lending

percent of GDP

The comparability of India and other LICs in attracting private inflows is worth noting. Much of the literature and investment industry attention directed at emerging markets includes India, but few other LICs, among their number. India is exceptional among LICs mainly in the low level of official flows. On private inflows as a whole and their major components, African LICs and other South and East Asian LICs have done as well or better than India in attracting inflows, relative to the size of their economies. This suggests both that India’s experience might have lessons for other LICs and that the experience of other LICs might provide examples of how India could manage greater inflows, should it choose to further liberalize FDI inflows (Reddy, 2007)

It is also worth noting that flows in African LICs have converged with those in Asian LICs in recent years. While the African LIC line is readily distinguishable from the other two in most of the charts in Figures 79 during the 1980s and 1990s, there is a convergence in most items since 2000, and FDI inflows to Africa are larger than for other regions. This suggests that the disincentives to investment in Africa in prior decades are dissipating in the current decade.11

Figure 9.Reserves and the Current Account Balance by Region

percent of GDP

B. HIPC Debt Relief Status: Free-Riding by Private Creditors?

Any consideration of capital flows to LICs must take account of the HIPC Initiative. The HIPC Initiative has been a decade-long effort by the international community to change the capital and financial account structures of the HIPCs, and it would be surprising if there were no difference between HIPCs and other LICs in capital and capital-like inflows. The HIPC Initiative also seeks to create a sharp series break at the time of the completion point (CP). The period before the completion point is characterized by unsustainable debt levels and the prospect of debt forgiveness; after the CP, creditor relations should be normal and debt levels sustainable and in many cases quite low. This series break also makes it important to distinguish between pre- and post-CP HIPCs in consideration of capital flows.

There are policy questions relating to new indebtedness by HIPCs, with concerns about a resumption of unsustainable borrowing trends after the CP or “free-riding” by new creditors taking advantage of the debt service created by the HIPC Initiative and the Multilateral Debt Relief Initiative (MDRI). While substantial re-accumulation of debt might take several years, some HIPCs had reached completion point as early as 2000 and almost all of the post-CP point HIPCs received MDRI relief at the beginning of 2006.

Capital and capital-like inflows for both pre-CP and post-CP HIPCs are higher than those for non-HIPCs (Figures 11, Tables 9-14).12 Almost all of the difference between HIPCs and non-HIPC LICs is attributable to official inflows, as aggregate inflows from private sources were similar in both level and trends among pre-CP, post-CP and non-HIPC LICs. Among the HIPCs, official inflows were consistently higher in post-CP relative to pre-CP HIPCs.

Figure 10.2006 LIC GDP by Debt Relief Status

Figure 11.Inflows by Debt Relief Status

Table 9.Post-Completion Point HIPCs: 1995-2006percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy0.91.42.52.33.03.02.63.12.22.32.11.9
Other investment liabilities3.25.5-2.91.51.30.70.81.91.60.80.5-19.1
Net external borrowing from commercial banks0.00.10.00.20.30.20.21.10.30.30.40.5
Liabilities to offical creditors5.22.0-2.61.92.42.51.42.72.51.30.6-15.1
Other liabilities, incl. trade credits, other loans, & deposits-2.03.4-0.3-0.6-1.5-2.0-0.8-1.9-1.2-0.8-0.5-4.5
Portfolio investment liabilities0.90.50.50.40.40.60.70.30.50.71.21.0
Financial account liabilities, total5.07.40.14.34.74.34.15.24.33.83.8-16.2
Capital-like inflows
Capital account1.6-0.97.13.34.12.73.62.22.33.54.021.5
debt forgiveness2.15.06.90.71.20.71.91.11.10.82.514.2
other capital account transfers-0.5-5.90.12.62.91.91.61.11.22.81.57.3
Current transfers6.75.74.95.25.46.27.17.07.88.59.49.0
Private transfers2.22.02.02.12.43.13.43.64.14.95.65.8
Official transfers4.53.72.93.13.03.13.73.33.73.63.83.2
“Capital-like” flows, total8.34.811.98.59.58.810.79.210.112.013.330.4
Total capital and capital-like inflows13.312.212.012.814.213.114.814.414.415.817.214.2
Asset items
Financial account assets, total (non-reserve)-0.4-0.1-0.1-0.30.10.20.90.60.10.50.3-0.2
International reserves-1.3-1.8-0.9-0.4-1.0-0.6-0.9-1.5-2.6-1.9-1.6-2.7
Errors and omissions-0.20.20.4-0.20.30.4-1.5-0.30.7-0.4-0.32.7
Asset items, total-1.9-1.7-0.5-1.0-0.60.0-1.5-1.3-1.8-1.8-1.6-0.2
Memorandum items
Official inflows, total11.810.77.35.76.66.37.07.17.35.66.82.2
Private inflows, total2.07.44.64.54.74.86.26.25.97.48.94.7
Unallocated capital account-0.5-5.90.12.62.91.91.61.11.22.81.57.3
Other invest. liab. to official creditors, net of debt forgiveness7.37.04.42.63.63.23.33.83.62.03.0-0.9
Current account balance-4.7-4.8-6.7-6.5-8.2-6.8-6.2-6.2-4.8-5.5-6.2-5.0
Current account balance, less transfers-11.4-10.5-11.5-11.7-13.6-13.0-13.3-13.1-12.6-14.0-15.6-14.0
1/ Benin, Bolivia, Burkina Faso, Cameroon, Ethiopia, Ghana, Guyana, Honduras, Madagascar, Malawi, Mali, Mauritania, Nicaragua, Niger, Rwanda, Senegal, Tanzania, Uganda, Zambia.
Table 10.Post-Completion Point HIPCs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.00.10.52.42.51.9
Other investment liabilities2.52.32.21.21.1-19.1
Net external borrowing from commercial banks-0.10.00.00.10.50.5
Liabilities to offical creditors2.32.73.91.31.7-15.1
Other liabilities, incl. trade credits, other loans, & deposits0.2-0.4-1.8-0.2-1.0-4.5
Portfolio investment liabilities0.40.30.60.50.71.0
Financial account liabilities, total2.82.73.34.14.2-16.2
Capital-like inflows
Capital account0.30.71.53.23.121.5
debt forgiveness0.10.31.12.91.514.2
other capital account transfers0.20.40.40.31.67.3
Current transfers2.04.46.75.57.99.0
Private transfers0.11.31.92.34.35.8
Official transfers1.93.14.83.13.63.2
“Capital-like” flows, total2.35.18.18.711.130.4
Total capital and capital-like inflows5.17.811.412.815.314.2
Asset items
Financial account assets, total (non-reserve)0.40.40.30.00.5-0.2
International reserves-0.1-0.4-0.7-1.0-1.7-2.7
Errors and omissions1.20.81.30.2-0.42.7
Asset items, total1.40.80.8-0.8-1.6-0.2
Memorandum items
Official inflows, total4.36.19.87.36.82.2
Private inflows, total0.61.31.25.26.94.7
Unallocated capital account0.20.40.40.31.67.3
Other invest. liab. to official creditors, net of debt forgiveness2.43.05.04.23.1-0.9
Current account balance-4.5-4.2-5.5-6.6-5.8-5.0
Current account balance, less transfers-6.5-8.6-12.2-12.1-13.7-14.0
Table 11Pre-Completion Point HIPCs: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy1.50.61.32.02.01.62.53.74.34.34.25.1
Other investment liabilities0.74.93.24.62.2-1.60.11.9-0.9-0.5-0.3-1.4
Net external borrowing from commercial banks-0.30.00.10.00.00.00.1-0.40.10.0-0.20.0
Liabilities to offical creditors0.43.52.54.13.01.62.22.80.70.20.1-0.7
Other liabilities, incl. trade credits, other loans, & deposits0.71.40.60.4-0.8-3.2-2.2-0.4-1.8-0.7-0.1-0.6
Portfolio investment liabilities1.30.60.70.90.80.00.5-0.7-0.4-0.3-0.50.0
Financial account liabilities, total3.56.15.17.55.0-0.13.14.93.03.53.53.7
Capital-like inflows
Capital account3.32.83.73.02.52.62.12.91.40.90.90.9
debt forgiveness0.20.71.00.40.10.40.32.00.30.00.10.0
other capital account transfers3.12.12.72.72.42.21.80.91.10.90.80.9
Current transfers3.03.23.23.64.35.15.76.06.06.57.16.4
Private transfers-0.20.51.51.72.23.13.63.84.05.05.14.3
Official transfers3.22.61.82.02.02.02.12.22.01.62.02.1
“Capital-like” flows, total6.36.06.96.76.87.77.88.97.47.47.97.4
Total capital and capital-like inflows9.812.012.014.111.87.610.913.810.410.911.411.1
Asset items
Financial account assets, total (non-reserve)-0.70.20.3-0.2-0.30.2-0.3-0.5-0.8-0.90.10.6
International reserves-2.50.81.51.60.8-0.30.0-1.51.1-2.7-1.4-1.2
Errors and omissions3.7-3.1-6.2-5.2-3.31.40.6-0.4-0.71.80.20.8
Asset items, total0.4-2.1-4.4-3.8-2.81.40.4-2.4-0.5-1.8-1.20.2
Memorandum items
Official inflows, total3.76.85.36.55.24.04.57.03.01.72.11.4
Private inflows, total3.03.14.05.04.31.44.55.96.38.38.58.8
Unallocated capital account3.12.12.72.72.42.21.80.91.10.90.80.9
Other invest. liab. to official creditors, net of debt forgiveness0.54.23.54.53.12.02.44.81.00.20.1-0.7
Current account balance-7.3-6.8-4.4-6.7-4.7-3.7-5.5-5.2-4.0-2.7-4.1-5.6
Current account balance, less transfers-10.3-10.0-7.6-10.3-9.0-8.8-11.2-11.2-10.0-9.2-11.2-12.0
Table 12Pre-Completion Point HIPCs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.31.11.01.53.85.1
Other investment liabilities5.13.01.12.70.1-1.4
Net external borrowing from commercial banks-1.81.00.00.0-0.10.0
Liabilities to offical creditors3.05.40.52.91.2-0.7
Other liabilities, incl. trade credits, other loans, & deposits3.9-3.40.7-0.3-1.0-0.6
Portfolio investment liabilities-0.10.00.50.6-0.30.0
Financial account liabilities, total5.44.12.64.73.63.7
Capital-like inflows
Capital account1.42.94.32.91.60.9
debt forgiveness0.00.10.20.50.50.0
other capital account transfers1.42.84.12.41.10.9
Current transfers1.51.21.93.96.36.4
Private transfers-0.8-0.9-0.51.84.34.3
Official transfers2.32.22.42.12.02.1
“Capital-like” flows, total2.94.16.36.87.97.4
Total capital and capital-like inflows8.28.38.911.511.511.1
Asset items
Financial account assets, total (non-reserve)-0.51.30.30.0-0.50.6
International reserves-0.11.80.20.9-0.9-1.2
Errors and omissions-2.4-0.82.0-3.30.30.8
Asset items, total-3.02.32.5-2.3-1.10.2
Memorandum items
Official inflows, total5.37.73.15.53.71.4
Private inflows, total1.5-2.21.63.66.78.8
Unallocated capital account1.42.84.12.41.10.9
Other invest. liab. to official creditors, net of debt forgiveness3.05.50.73.51.7-0.7
Current account balance-3.7-9.4-9.5-5.3-4.3-5.6
Current account balance, less transfers-5.2-10.6-11.4-9.2-10.5-12.0
Table 13.Non-HIPC LICs: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liabilities
Direct investment in reporting ecomomy1.31.41.51.41.41.41.92.52.32.42.23.0
Other investment liabilities-0.20.3-0.51.40.40.30.00.40.50.90.01.7
Net external borrowing from commercial banks0.0-0.60.1-0.10.00.0-0.10.00.10.40.20.3
Other investment liabilities to official creditors0.40.0-0.31.01.00.40.60.5-0.10.2-0.10.4
Other liabilities, incl. trade credits, other loans, & deposits-0.70.9-0.20.5-0.6-0.1-0.5-0.10.50.3-0.11.0
Portfolio investment liabilities0.40.60.4-0.10.30.20.30.10.80.81.00.6
Financial account liabilities, total1.52.31.52.72.12.02.22.93.74.13.25.4
Capital-like inflows
Capital account0.20.70.20.20.20.20.10.30.30.20.30.4
debt forgiveness0.00.61.40.10.20.10.10.10.10.10.00.0
other capital account transfers0.20.1-1.20.10.10.10.10.20.30.10.20.3
Current transfers3.23.63.93.53.53.83.94.14.64.14.34.4
Private transfers3.64.23.12.72.83.13.13.33.73.43.53.4
Official transfers-0.4-0.60.80.80.80.70.80.90.90.80.80.8
“Capital-like” flows, total3.44.34.13.73.74.04.14.45.04.34.54.8
Total capital and capital-like inflows4.96.65.66.35.85.96.37.38.78.47.710.1
Asset items
Financial account assets, total (non-reserve)-0.2-0.20.0-0.1-0.2-0.6-0.6-1.0-0.6-0.9-0.8-1.8
International reserves0.4-1.1-1.0-0.2-0.4-1.2-1.3-2.4-3.5-3.4-2.3-4.0
Errors and omissions0.81.11.30.50.0-0.2-0.40.1-0.4-0.1-0.4-0.6
Asset items, total1.0-0.30.40.2-0.6-2.1-2.4-3.3-4.5-4.3-3.5-6.4
Memorandum items
Official inflows, total0.00.01.91.91.91.11.51.50.91.10.71.2
Private inflows, total4.76.64.94.43.94.74.85.87.57.46.88.4
Unallocated capital account0.20.1-1.20.10.10.10.10.20.30.10.20.3
Other invest. liab. to official creditors, net of debt forgiveness0.40.51.11.11.10.50.70.60.00.2-0.10.4
Current account balance-2.7-2.7-2.0-3.1-1.7-0.10.00.10.50.00.10.7
Current account balance, less transfers-5.9-6.3-6.0-6.5-5.2-3.9-3.9-4.0-4.2-4.1-4.2-3.7
Table 14.Non-HIPC LICs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.20.30.91.42.33.0
Other investment liabilities2.01.81.40.40.41.7
Net external borrowing from commercial banksn.a.n.a.0.1-0.10.10.3
Liabilities to offical creditors1.30.90.90.40.20.4
Other liabilities, incl. trade credits, other loans, & depositsn.a.n.a.-0.50.10.01.0
Portfolio investment liabilities0.00.00.40.30.60.6
Financial account liabilities, total2.22.12.72.13.25.4
Capital-like inflows
Capital account0.00.10.20.30.20.4
debt forgiveness0.00.00.00.50.10.0
other capital account transfers0.00.10.2-0.20.20.3
Current transfers1.51.42.93.64.24.4
Private transfers1.61.72.93.23.43.4
Official transfers0.30.00.00.50.80.8
“Capital-like” flows, total1.51.53.14.04.44.8
Total capital and capital-like inflows3.73.55.86.17.710.1
Asset items
Financial account assets, total (non-reserve)0.0-0.1-0.1-0.2-0.8-1.8
International reserves0.20.0-0.5-0.8-2.6-4.0
Errors and omissions0.50.6-0.40.5-0.2-0.6
Asset items, total0.70.4-1.0-0.5-3.6-6.4
Memorandum items
Official inflows, total1.60.90.91.41.11.2
Private inflows, total2.52.84.64.96.48.4
Unallocated capital account0.00.10.2-0.20.20.3
Other invest. liab. to official creditors, net of debt forgiveness1.30.90.90.90.30.4
Current account balance-2.9-2.6-1.9-1.90.10.7
Current account balance, less transfers-4.4-4.0-4.8-5.6-4.1-3.7

Capital and capital-like inflows have been on an increasing trend in all three groups of countries, due largely to inflows from private sources. The increase in private-source inflows is most pronounced among the HIPCs, where inflows rose from near zero levels in the 1980s to the high single digit percentages of GDP from about 2002 onward. Non-HIPC LICs started from higher initial GDP shares in the 1980s, but largely converged with the HIPCs by the late 1990s.

The main components of private and official flows are consistent with the broader aggregates for all groups (Figure 11). Both private transfers and FDI increase for all three groups, with a more pronounced increase from a lower level for the pre-CP HIPCs. There is convergence in the patterns of official lending (net of debt forgiveness), which is consistently low for non-HIPCs LICs and falls sharply from a mid-1990s peak for both groups of HIPCs.

Private debt-creating inflows to HIPCs have been modest, but increasing (Tables 9-12). There has been a modest net inflow of portfolio liabilities to LICs, including HIPCs and non-HIPCs. However, the net portfolio inflows to post-CP HIPCs have been on an increasing trend, and all of the portfolio liabilities in post-CP HIPCs are in the form of debt securities. By contrast, portfolio liabilities of non-HIPC LICs have increased by a smaller share of GDP and consist almost entirely of equity securities. Commercial bank lending has been on a upward trajectory in post-CP HIPCs and non-HIPC LICs, but starting from very low levels in both cases. As with portfolio liabilities, commercial bank lending to non-HIPC LICs has shown less of an upward trend and is at a lower share of GDP in non-HIPCs. One puzzling item is “other” other liabilities, a category that includes currency and deposits, trade credits, and loans from sources other than commercial banks and official creditors. In non-HIPCs, this aggregate tends to fluctuate between small negative and small positive values (Table 14). In both pre-CP and post-CP HIPCs, this items is a consistent and substantial outflow. This could imply consistent net repayment of trade credit or other loans or the draw down of bank deposits, but many of the measures of capital flight would include these items as a component of capital flight.

Financial account asset outflows have been increasing in all three groups of countries, but these are dominated by the accumulation of official reserves. Reserve asset accumulation has been particularly pronounced in non-HIPCs. Reserve assets have also been accumulating at accelerating, but lower, rates in pre-CP and post-CP HIPCs (Figure 13). Non-HIPC LICs have seen an increase in non-reserve financial account assets to 1.8 percent of GDP by 2006, although roughly nine-tenths of this outflow is in the form of FDI. By contrast, non-reserve assets flows have been more modest in size and without a clear trend in HIPCs.

Figure 12.Private Transfers, FDI, and Net Lending by Debt-Relief Status

percent of GDP

Figure 13.Reserve and Current Account Balance by Debt-Relief Status

Current account balances have diverged between HIPCs and non-HIPC LICs in recent years. Current account deficits in non-HIPC LICs have been on a declining trend reaching rough balance from 2000 onwards. However, current account deficits in both pre-CP and post-CP HIPCs have stabilized in the range of 4-6 percent of GDP in the last decade.

The greater levels of current account deficits in HIPCs could be taken as a sign of vulnerability, but there are factors that suggest relatively benign interpretations. As noted above, debt-creating inflows are low and declining, at least in the case of debt to official creditors. Also, the fact that these countries are accumulating reserves indicates that these deficits are over-financed, and the reserve stock serves to ameliorate the risks associated with a slowdown or reversal of the inflows. Finally, given that HIPCs are viewed as particularly needy in terms of domestic investment in human and physical capital, there is a case to be made that higher current account deficits and lower rate of reserve accumulation are needed to accommodate the resource transfer associated with the absorption of foreign aid and other inflows (Berg et al., 2005).

Overall, the patterns in capital and capital-like flows across LICs of differing debt-relief status are converging. As with regions, differences present in earlier decades were more of degree than overall direction. Among the differences that are evident in recent years, non-HIPC LICs have higher reserve accumulations and stronger current accounts, and commercial bank lending to post-CP LICs is higher than to other LICs, although only about half a percent of GDP by 2006.

The relationship between HIPC status and the pattern of inflows over time does not lend itself to easy explanations as to the effect of the HIPC Initiative itself. The higher levels of official inflows to HIPCs have been present throughout the past quarter century. Moreover, official inflows peaked in the mid-1990s for post-CP HIPCs, before the creation of the original HIPC Initiative in 1996, and official flows to both groups of HIPCs fell sharply by 2006.13 The data could be interpreted to conclude that the higher levels of official inflows to HIPCs in the period through the mid-1990s contributed to the buildup of official debt that motivated the HIPC Initiative. Also, the onset of HIPC Initiative and MDRI debt-relief is associated with a decline in official inflows, perhaps as other forms of official support were redirected into financing the budgetary costs of debt forgiveness. (The provision of debt forgiveness itself has no impact on measured official flows as a reduction in financial account liabilities is offset by the debt forgiveness item which finances it in the capital account.)

C. Resources Endowments: Is This Investment Related to the Commodity Boom?

Mineral endowments appear to be an important determinant of private investment, particularly in the recent period of sharply rising prices for oil and other commodities. Private investment in LICs, in particular FDI, has often been associated with extractive industries. While WEO data are not available by sector, it is possible to group countries by their mineral resource endowments. This section divides LICs into “hydrocarbon-rich,” “mineral-rich,” and “non-mineral” countries based on the classification of countries in the IMF’s Guide to Resource Revenue Transparency.14

Figure 14.2006 LIC GDP by Export Type

Private inflows to hydrocarbon-rich economies have been higher than to other LICs, although not overwhelmingly so and with signs that non-mineral LICs may be catching up (Figure 15). The difference is largely explained by FDI inflows to hydrocarbon-rich LICs. In most other respects, there is little difference between hydrocarbon-rich, mineral-rich, and non-mineral LICs in terms of trends and levels, although inflows to hydrocarbon-rich and mineral-rich are more volatile than those to non-mineral LICs (Figure 16).

Figure 15.Inflows by Resource Endowment Type

percent of GDP

Figure 16.Private Transfers, FDI, and Net Official Lending by Resource Endowment

percent of GDP

Figure 17.Reserves and Current Account Balance by Resource Endowment

percent of GDP

Asset outflows and errors and omissions in hydrocarbon-rich and mineral-rich LICs show both reserve accumulation and some patterns that are often attributed to capital flight (Tables 1518; and Figure 18). In both groups of countries, non-reserve financial assets and net errors and omissions have been consistent outflows. These groups have also been accumulating reserve assets, but non-reserve outflows have exceeded reserve accumulation in most years since the mid-1990s. There have also been outflows of “other” other financial account liabilities (e.g., net reduction in liabilities such as trade credits or bank deposits). These non-reserve outflows do not necessarily represent capital flight, but they include some of the core components of most capital flight estimates in the literature and may merit further investigation. Alternatively, this might reflect different investment strategies (in the case of non-reserve assets) or other factors.

Figure 18.Non-Reserve Asset Outflows

percent of GDP
Table 15.Hydrocarbon Rich LICs: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy3.73.44.64.65.64.46.210.09.38.47.06.9
Other investment liabilities0.9-3.6-2.13.52.40.7-0.20.40.00.7-1.5-3.0
Net external borrowing from commercial banks-0.1-3.20.0-0.40.0-0.4-0.3-0.5-0.20.40.3-0.3
Liabilities to offical creditors2.0-1.1-2.93.33.21.61.62.51.61.5-1.2-1.8
Other liabilities, incl. trade credits, other loans, & deposits-1.10.70.70.6-0.8-0.5-1.5-1.6-1.4-1.2-0.6-0.8
Portfolio investment liabilities0.10.10.10.10.00.10.2-0.10.10.20.30.2
Financial account liabilities, total4.7-0.12.68.28.05.36.110.29.49.35.94.1
Capital-like inflows
Capital account-0.22.50.40.00.00.10.10.00.10.20.22.2
debt forgiveness-0.42.84.80.30.00.30.10.00.00.00.02.0
other capital account transfers0.2-0.4-4.4-0.30.0-0.20.00.00.10.10.10.3
Current transfers2.83.23.94.23.93.73.44.14.34.24.34.0
Private transfers2.73.15.45.64.84.23.53.83.93.93.52.6
Official transfers0.60.60.30.30.20.00.10.20.20.10.20.1
“Capital-like” flows, total2.65.74.34.23.93.83.54.14.44.34.46.2
Total capital and capital-like inflows7.35.76.912.411.99.19.614.313.813.610.310.3
Asset items
Financial account assets, total (non-reserve)-0.3-0.9-0.4-0.7-1.2-2.4-1.9-4.8-1.9-3.0-3.1-4.2
International reserves0.2-3.4-2.41.31.0-4.0-0.61.2-1.0-5.3-5.7-6.8
Errors and omissions2.55.73.1-0.8-3.3-3.1-2.60.1-2.2-1.0-1.9-2.9
Asset items, total2.31.50.3-0.2-3.5-9.5-5.1-3.5-5.2-9.2-10.6-13.9
Memorandum items
Official inflows, total2.22.32.23.93.41.91.82.81.81.6-1.00.2
Private inflows, total5.44.110.810.59.67.98.011.511.711.810.58.5
Unallocated capital account0.2-0.4-4.4-0.30.0-0.20.00.00.10.10.10.3
Other invest. liab. to official creditors, net of debt forgiveness1.61.71.93.63.21.91.72.51.61.5-1.20.1
Current account balance-6.8-3.9-3.3-8.0-4.64.2-1.1-6.8-4.4-0.34.57.6
Current account balance, less transfers-9.6-7.1-7.2-12.2-8.40.5-4.5-10.8-8.6-4.50.33.6
Table 16.Hydrocarbon Rich LICs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.81.13.44.58.26.9
Other investment liabilities3.12.71.50.2-0.1-3.0
Net external borrowing from commercial banks-0.44.8-0.8-0.8-0.1-0.3
Liabilities to offical creditors3.82.61.20.81.2-1.8
Other liabilities, incl. trade credits, other loans, & deposits-0.3-4.71.00.2-1.3-0.8
Portfolio investment liabilities0.30.20.10.10.10.2
Financial account liabilities, total4.24.15.04.88.24.1
Capital-like inflows
Capital account0.10.2-0.40.60.12.2
debt forgiveness0.00.0-0.51.60.02.0
other capital account transfers0.10.20.2-1.10.10.3
Current transfers1.92.32.73.84.04.0
Private transfers-0.61.22.64.63.72.6
Official transfers3.42.40.60.30.20.1
“Capital-like” flows, total1.92.52.34.44.16.2
Total capital and capital-like inflows6.16.67.39.212.310.3
Asset items
Financial account assets, total (non-reserve)-0.2-0.6-1.1-1.1-2.9-4.2
International reserves1.90.31.9-1.5-2.3-6.8
Errors and omissions0.20.70.30.3-1.5-2.9
Asset items, total2.00.41.1-2.3-6.7-13.9
Memorandum items
Official inflows, total7.25.11.22.71.40.2
Private inflows, total-0.22.76.48.610.78.5
Unallocated capital account0.10.20.2-1.10.10.3
Other invest. liab. to official creditors, net of debt forgiveness3.82.60.62.51.20.1
Current account balance-6.2-4.7-5.7-3.1-1.67.6
Current account balance, less transfers-8.1-7.0-8.3-6.9-5.63.6
Table 17.Mineral Rich LICs: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy2.61.51.52.22.72.52.22.23.64.83.53.2
Other investment liabilities-3.31.42.01.20.00.21.34.21.10.9-3.6-17.4
Net external borrowing from commercial banks0.1-0.20.00.30.80.1-0.3-0.1-0.3-0.3-0.3-0.2
Liabilities to offical creditors6.32.42.42.61.72.52.25.22.61.7-2.8-10.4
Other liabilities, incl. trade credits, other loans, & deposits-9.7-0.8-0.4-1.7-2.4-2.5-0.5-0.9-1.2-0.5-0.5-6.7
Portfolio investment liabilities0.6-0.2-0.1-0.3-0.2-0.20.20.00.2-0.21.41.0
Financial account liabilities, total-0.12.73.53.22.52.43.76.44.85.51.3-13.2
Capital-like inflows
Capital account1.23.24.34.56.15.56.64.64.73.18.515.8
debt forgiveness0.20.10.40.51.91.22.42.72.31.26.613.3
other capital account transfers1.03.13.94.04.14.34.22.02.41.91.82.4
Current transfers4.84.13.64.54.35.97.07.98.09.210.39.0
Private transfers-0.30.10.71.11.32.02.43.43.35.55.95.4
Official transfers5.03.92.93.42.93.84.64.54.73.84.43.6
“Capital-like” flows, total5.97.37.99.010.311.413.612.612.712.318.724.7
Total capital and capital-like inflows5.810.011.312.212.913.817.219.017.617.820.011.5
Asset items
Financial account assets, total (non-reserve)-5.7-2.3-1.7-2.6-2.3-1.90.5-1.8-4.1-3.4-3.0-3.5
International reserves0.0-0.80.81.40.6-1.3-0.9-1.9-2.5-2.8-1.0-4.5
Errors and omissions4.50.3-0.11.20.51.2-4.5-3.2-0.81.2-1.38.0
Asset items, total-1.1-2.8-1.00.0-1.1-2.0-4.9-6.9-7.5-5.0-5.30.0
Memorandum items
Official inflows, total11.66.45.66.56.57.59.212.49.56.68.26.5
Private inflows, total-6.80.51.81.72.21.93.94.65.69.310.02.6
Unallocated capital account1.03.13.94.04.14.34.22.02.41.91.82.4
Other invest. liab. to official creditors, net of debt forgiveness6.52.52.83.13.63.74.67.94.92.83.82.9
Current account balance0.1-3.2-6.8-7.7-7.5-5.7-5.3-4.0-2.1-3.6-5.7-4.1
Current account balance, less transfers-4.7-7.3-10.3-12.2-11.8-11.5-12.2-11.9-10.1-12.8-16.0-13.0
Table 18.Mineral-Rich LICs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy-0.6-0.30.52.13.23.2
Other investment liabilities3.94.8-0.81.00.8-17.4
Net external borrowing from commercial banks0.1-0.30.00.2-0.3-0.2
Liabilities to offical creditors-0.10.61.62.31.8-10.4
Other liabilities, incl. trade credits, other loans, & deposits2.84.4-2.4-1.6-0.7-6.7
Portfolio investment liabilities-0.10.00.3-0.20.31.0
Financial account liabilities, total3.24.50.02.94.4-13.2
Capital-like inflows
Capital account1.03.84.64.75.515.8
debt forgiveness0.00.20.70.83.013.3
other capital account transfers1.03.63.93.92.52.4
Current transfers0.81.93.94.58.59.0
Private transfers-0.4-0.4-0.21.14.15.4
Official transfers1.22.34.13.44.43.6
“Capital-like” flows, total1.85.78.59.214.024.7
Total capital and capital-like inflows4.910.38.412.018.311.5
Asset items
Financial account assets, total (non-reserve)0.51.6-0.7-2.1-2.4-3.5
International reserves0.5-0.3-0.20.1-1.8-4.5
Errors and omissions0.0-1.2-1.90.6-1.78.0
Asset items, total1.10.1-2.8-1.4-5.90.0
Memorandum items
Official inflows, total1.13.16.36.59.26.5
Private inflows, total2.83.5-1.81.66.72.6
Unallocated capital account1.03.63.93.92.52.4
Other invest. liab. to official creditors, net of debt forgiveness-0.10.82.23.14.82.9
Current account balance-5.2-8.5-1.7-6.2-4.1-4.1
Current account balance, less transfers-6.0-10.4-5.6-10.6-12.6-13.0
Table 19.Non-Mineral or Hydrocarbon-Rich LICs: 1995-2006 1/percentages of GDP
199519961997199819992000200120022003200420052006
Financial account liablities
Direct investment in reporting ecomomy0.70.91.00.90.81.01.11.21.01.11.12.0
Other investment liabilities0.22.1-0.31.30.20.10.10.50.60.80.51.1
Net external borrowing from commercial banks0.00.10.10.00.00.10.00.20.20.40.20.5
Liabilities to offical creditors0.40.60.00.90.90.40.50.3-0.10.00.4-0.4
Other liabilities, incl. trade credits, other loans, & deposits-0.31.4-0.40.4-0.6-0.4-0.40.00.60.50.01.0
Portfolio investment liabilities0.60.80.50.00.40.30.40.10.90.91.10.7
Financial account liabilities, total1.53.81.32.21.51.41.71.82.52.92.73.8
Capital-like inflows
Capital account0.70.21.10.60.70.50.50.60.50.50.51.7
debt forgiveness0.30.71.50.10.20.10.20.30.10.10.10.6
other capital account transfers0.3-0.5-0.40.50.40.30.20.30.40.40.41.1
Current transfers3.63.94.03.53.64.14.44.55.14.64.95.0
Private transfers3.64.22.82.32.52.93.13.13.73.33.53.5
Official transfers0.0-0.31.21.21.11.21.31.41.41.31.41.5
“Capital-like” flows, total4.34.05.14.24.34.64.95.05.65.15.36.7
Total capital and capital-like inflows5.87.86.46.45.86.06.66.88.17.98.010.6
Asset items
Financial account assets, total (non-reserve)0.00.10.10.10.1-0.1-0.20.0-0.2-0.20.0-0.6
International reserves0.0-0.6-0.6-0.4-0.6-0.5-1.4-3.0-3.6-2.8-1.3-2.9
Errors and omissions0.4-0.30.40.20.40.50.10.10.10.20.00.2
Asset items, total0.4-0.80.0-0.2-0.2-0.1-1.5-2.9-3.7-2.8-1.3-3.4
Memorandum items
Official inflows, total0.81.02.72.22.21.72.11.91.41.31.81.7
Private inflows, total4.77.34.13.73.23.94.24.66.36.25.87.8
Unallocated capital account0.3-0.5-0.40.50.40.30.20.30.40.40.41.1
Other invest. liab. to official creditors, net of debt forgiveness0.71.31.51.01.10.50.80.60.00.10.40.2
Current account balance-2.6-3.1-2.4-2.7-2.0-1.8-0.70.50.7-0.6-1.8-2.1
Current account balance, less transfers-6.2-7.0-6.4-6.2-5.6-5.9-5.1-4.0-4.4-5.2-6.7-7.2
Table 20.Non-Mineral or Hydrocarbon-Rich LICs: 1981-2006 1/percentages of GDP
1981-851986-901991-951995-20002001-052006
Financial account liablities
Direct investment in reporting ecomomy0.10.20.40.91.12.0
Other investment liabilities1.91.61.50.70.51.1
Net external borrowing from commercial banksn.a.n.a.0.00.00.20.5
Liabilities to offical creditors1.21.21.20.50.2-0.4
Other liabilities, incl. trade credits, other loans, & depositsn.a.n.a.0.20.10.11.0
Portfolio investment liabilities0.00.00.50.40.70.7
Financial account liabilities, total2.01.82.42.02.33.8
Capital-like inflows
Capital account0.10.20.60.60.51.7
debt forgiveness0.00.00.20.50.20.6
other capital account transfers0.10.20.40.10.31.1
Current transfers1.51.63.43.84.75.0
Private transfers1.51.32.82.93.33.5
Official transfers0.00.30.60.91.41.5
“Capital-like” flows, total1.71.83.94.55.26.7
Total capital and capital-like inflows3.73.66.36.57.510.6
Asset items
Financial account assets, total (non-reserve)0.00.10.20.1-0.1-0.6
International reserves-0.30.0-0.9-0.5-2.4-2.9
Errors and omissions0.50.60.00.20.10.2
Asset items, total0.20.7-0.7-0.3-2.4-3.4
Memorandum items
Official inflows, total1.21.52.02.01.71.7
Private inflows, total2.41.94.04.45.47.8
Unallocated capital account0.10.20.40.10.31.1
Other invest. liab. to official creditors, net of debt forgiveness1.21.21.41.10.40.2
Current account balance-2.3-2.7-2.3-2.4-0.4-2.1
Current account balance, less transfers-3.8-4.3-5.7-6.2-5.1-7.2

Non-mineral LICs have also had consistent asset outflows in recent years, but these are dominated by reserve accumulation. Non-reserve financial assets accumulation and net errors and omissions have been small and frequently switch signs. Non-mineral LICs also have very small net flows of “other” other financial account liabilities, with a positive sign (i.e., an inflow) more often than not.

IV. Conclusions and Policy Implications

Before moving to consider policy implications and next steps, it is worth recapitulating the main trends and their likely causes

  • Over the last quarter century there has been a strong surge in capital and capital-like flows to LICs that has accelerated in recent years.

  • All of the increase in capital and capital-like flows to LICs (as a share of their GDP) has come from private sources. Private capital and capital-like flows are now roughly four times their level in the 1980s at around 8 percent of LIC GDP, while official inflows have remained unchanged at roughly 2 percent of GDP. Whatever the balance between these factors, the sharp increase in private inflows presumably reflects the improvement in macroeconomic stability and other aspects of the investment environment.

  • Private transfers (e.g., remittances) and FDI have dominated the inflows from private sources. Both FDI and private transfers have been increasing strongly, but FDI has increased tenfold as a share of LIC GDP; private transfers have risen threefold as a share of LIC GDP, but remain slightly larger due to a higher initial level. Increases in FDI are likely a response to a mix of corporate and financial globalization, wide-ranging capital account liberalization, increased commodity prices, and improved macroeconomic policies in LICs. The growth of private transfers (e.g., remittances) likely reflects on-going migration patterns, rising incomes in “host” countries, and falling costs of financial intermediation and transfers.

  • Other, debt-creating inflows from private sources, such as loans from commercial banks and portfolio investment, had been minor and declining in relative importance until the last few years. The relative absence of such flows may represent a hard-learned risk aversion by lenders, LICs, or both. However, it may also reflect a lack of market development (e.g., the absence of tradable securities).

  • Although flows from official sources are little changed as a share of LIC GDP, their composition has evolved as donors and lenders have shifted resources from loans to grants and debt forgiveness.

  • LICs have saved much of these increased inflows in the form of international reserves with the result that the stock of reserves relative to imports has more than doubled since 1995.

These pattern of a strong rise in private inflows is consistent across the major regions and groups of LICs, with only modest differences of degree.

  • Private flows have surged in Africa, India, and other South and East Asian LICs in broadly similar proportions.

  • Private capital and capital-like inflows have increased irrespective of countries’ debt-relief (i.e., HIPC Initiative) status.

  • Inflows are also similar across countries with differing natural resource endowments (e.g., hydrocarbon-rich countries, other mineral-rich countries, and LICs without substantial mineral endowments).

  • There are differences in the levels of official flows to these different regions and groups, however, official flows as a share of GDP in the different country groups have generally remained relatively stable for individual country groups.

The policy implications of these trends are mostly benign.

  • New inflows appear to be rising steadily as a share of LIC GDP with no sign of a slowdown.

  • Private inflows are concentrated in non-debt-creating FDI and private transfers rather than loans; official inflows have shifted from loans to grants and debt forgiveness.

  • The inflows are associated with improving current account balances rather than financing widening and potentially difficult-to-sustain current account deficits.

  • The higher international reserve levels in LICs provide some insurance against external shocks or a slowdown in inflows.

  • There is little sign of offsetting non-reserve outflows (e.g., capital flight) except possibly in hydrocarbon-rich and mineral-rich LICs.

However, the trends and their policy implications should be interpreted cautiously in light of the severe problems with LIC data.

  • IMF staff estimates have had to fill in where national BOP data are unavailable due to long lags and substantial gaps in the official data. Thus, significant revisions to much of the individual country data are certain, and material revisions to even some of the aggregate trends cannot be ruled out.

  • In particular, information on lending from nontraditional creditors (official and private) is often poor and the staff estimates for more recent years may need to be revised significantly, possibly resulting in a much greater pickup in loans than is currently reported.

  • The relationship between reported data and the underlying economic concepts for private transfers/remittances is likely to be loose and subject to very substantial measurement error.

The new inflows present new challenges to economic policy making in LICs.

  • The shift from official to private financing implies a different and less direct role for LIC governments in determining the uses of external financing. With the predominance of private flows, the intermediation of these flows will take place through commercial banks. While the proportion of inflows ultimately sold to the central bank may be similar, they will not be intermediated through the government and the savings/investment and other choices for the use of inflows will be made by private parties.

  • The rise of private inflows makes it likely that these private flows will increasingly become the main sources of external vulnerability in LICs. Concerns about sustainability, effects on relative prices and competitiveness, and accompanying policy and institutional reforms need to be considered in the context of these new inflows to LICs.

  • External sector policies in LICs need to focus on policies relevant to the new inflows. The policies to manage the scale and economic impact of official flows do not translate well into policies for encouraging and managing private, non-debt-creating flows. Strong debt management and good donor relations are particularly important for encouraging and managing official lending. However, maintaining a stable macroeconomic environment and favorable business climate, and efficient mechanisms for international transfers by households are likely to be more important for private inflows.

  • Weaknesses in national data on the balance of payments data may undermine the basis for LIC economic decision-making. To the extent that data are unavailable to LIC policy makers in a timely manner, policy-making may be flying blind. It is in the interest of LIC authorities to improve the recording and compilation of balance of payments, with a particular focus on these rapidly increasing sources of private inflows.

References

    Asiedu, Elizabeth, 2002, “On the Determinants of Foreign Direct Investment to Developing Countries: Is Africa Different?World Development, pp. 107-119

    Barro, Robert J. and XavierSala-i-Martin, 1995, Economic Growth, (New York: McGraw-Hill).

    Bénassy-Quéré, Agnes, MaylisCoupet, and ThierryMayer, 2007, “Institutional Determinants of Foreign Direct Investment,” World Economy, 2007, pp. 764-82.

    Berg, Andrew, ShekharAiyar, MumtazHussain, ShaunRoche, TokhirMirzoev, and AmberMahone, 2007, The Macroeconomic of Scaling Up Aid: Lessons from International Experience, IMF Occasional Paper 253 (Washington: International Monetary Fund)

    Bhinda, Nils, JonathanLeape, MatthewMartin, StephanyGriffith-Jones, 1999, Private Capital Flows to Africa: Perception and Reality, (The Hague, FONDAD).

    Bosworth, Barry, Susan M.Collins; and Carmen M.Reinhart, 1999, Capital Flows to Developing Countries: Implications for Savings and Investment, Brookings Papers on Economic Activity, Vol. 1999, No. 1. pp. 143180.

    Boyce, James K. and LéonceNdikumana, 2002, Public Debts and Private Assets, Explaining Capital Flight from Sub-Saharan African Countries, World Development, Vol. 31, No. 1, pp. 107130.

    Boyce, James K. and LéonceNdikumana, 2001, “Is Africa a Net Creditor? New Estimates of Capital Flight from Severely Indebted Sub-Saharan Africa Countries, 1970–96.” Journal of Development Studies, December2001, pp. 2756.

    Buch, Claudia and AnjaKuckulenz, 2004, Worker Remittances and Capital Flows to Developing Countries, Discussion Paper 04-31 (Mannheim: Center for European Economic Research).

    Busse, Matthias, and JoséLuis Groizard, 2006, Foreign Direct Investment, Regulations, and Growth, World Bank Policy Research Working Paper 3882 (Washington: World Bank).

    Chami, Ralph, ConnelFullenkamp, and SamirJahjah(2005) “Are Immigrant Remittance Flows a Source of Capital for Development?IMF Staff Papers, Vol. 52, No. 1 (Washington: International Monetary Fund).

    Dudine, Paolo, JamesJohn, MarkLewis, LuzmariaMonasi, HelawayTadesse, and JoergZeuner, 2006, Weathering the Storm So Far: The Impact of the 2003–05 Oil Shock on Low-Income Countries, IMF Working Paper 06/171 (Washington: International Monetary Fund).

    Evrensel, Ayşe, 2004, “Lending to developing countries revisited: changing nature of lenders and balance of payments problems.” Economic Systems, 28, pages 235256.

    Fajnzylber, Pablo and J HumbertoLopez, eds.2007Closer to Home (Washington: World Bank).

    Ghosh, Atish, ManuelaGoretti, BikasJoshi, AlunThomas, and JuanZalduendo, 2007, “Capital Flows, the Balance of Payments, and Policy Responses,” (Washington, International Monetary Fund). [forthcoming]

    Hermes, Niels, RobertLensink, and VictorMurinde, Flight Capital and Its Reversal for Developmental Financing,” WIDER Discussion Paper 2002/99 (Helsinki: World Institute for Development Economics Research).

    International Monetary Fund, 1993, Balance of Payments Manual, Fifth Edition (Washington: International Monetary Fund)

    International Monetary Fund, 2005a, Guide to Resource Revenue Transparency (Washington: International Monetary Fund).

    International Monetary Fund, 2005b, World Economic Outlook, April2005, (Washington: International Monetary Fund).

    International Monetary Fund, 2007, World Economic Outlook, Fall 2007 (Washington: International Monetary Fund).

    Naudé, W.A., and W.F.Krugell, 2007, “Investigating Geography and Institutions as Determinants of Foreign Direct Investment in Africa Using Panel Data,” Applied Economics, 39, pp. 1223-1233.

    Reddy, Yaga Venugopal(2007) Point of View: Converting a Tiger, Finance and Development, March2007, Vol. 44, No 1. (IMF, Washington)

    Reinke, Jens, Remittances in the Balance of Payments Framework: Current Problems and Forthcoming Improvements, presentation to Semin ar on Remittance Statistics at the Center of Excellence in Finance, Ljubljana, Slovenia, February 26–March 2, 2007. imf.org/external/np/sta/bop/pdf/rem.pdf

    Sala-i-Martin, Xavier X., 1997, “I Just Ran 2 Million Regressions,” American Economic Review, Papers and Proceedings, May1997, pages 17883.

    Selassie, Abebe Aemro, BenedictClements, ShamsuddinTareq, Jan KeesMartijn, and Gabriel diBella(2006), Designing Monetary and Fiscal Policy in Low-Income Countries, IMF Occasional Paper 250 (Washington: International Monetary Fund).

    Suro, Roberto, SergioBendixen, B.Lindsay Lowell, and Dulce C.Benavides(2002) Billions in Motion: Latino Immigrants, Remittances and Banking, (Washington: Pew Hispanic Center and Multilateral Investment Fund).

    Williamson, John and Donald R.Lessard, 1987, Capital Flight: The Problem and Policy Responses, (Washington: Institute for International Economics).

    World Bank, 1985, World Development Report, 1985, (Washington: World Bank).

    World Bank, 2006a, Global Development Finance, 2006 (Washington: World Bank).

    World Bank, 2006b, Global Economic Perspectives, 2006 (Washington: World Bank).

ANNEX – A

Data Issues

A. Defining Capital and Capital-Like Flows

1. Capital and capital-like flows are distributed across all of the major components of the balance of payments (Box A1). The “core” capital flows are those in the financial account: foreign direct investment (FDI), portfolio investment, and the residual category “other investment,” that includes loans, trade credits and deposits. However, some components of the current account, the capital account, and net errors and omissions are considered as capital flows or close substitutes. These include debt forgiveness, official transfers (e.g., grants), and private transfers (e.g., remittances). These capital and capital-like flows can be combined and rearranged to get a more complete picture of the relative importance of private and official sources of capital. Finally, examination of financial account assets of LICs can shed some light on the role of these assets as a use of inflows and as a buffer against fluctuations in inflows.

2. The specific categories used in this note are defined as follows; more precise definitions of the individual components are presented in Box AI. 1 below.

  • Capital inflows: This note retains the conventional use of the term “capital inflows” to refer to net changes in external liabilities (e.g., FDI, portfolio flows, and loans) even though these transactions are classified as financial account transactions for the purposes of the Balance of Payments Manual, 5th Edition (BPM5).

  • Capital-like inflows: Other balance of payments flows may be substitutes for financial account flows to varying degrees. For example, official grants and debt forgiveness are seen as substitutes for official loans, the use of reserves can substitute for new borrowing, and private transfers (e.g., remittances) are substitutable for ODA in some circumstances. This note uses the umbrella term “capital-like flows” to cover: current transfers (official and private) in the current account, and debt forgiveness and other capital transfers in the capital account.

  • Asset outflows: The assessment of the impact of capital inflows is influenced by the counterpart items in the balance of payments. For this paper, these asset transactions are divided into changes in international reserves and the accumulation of other financial account assets. Net errors and omissions are sometimes considered as the reflection of disguised or unrecorded capital flows and are also included under the heading of asset outflows in this paper, although many factors other than assets outflows may be contributing to net errors and omissions.15

Box AI.1.Balance of Payments Classification of Assets, Liabilities, and Flows

The IMF’s World Economic Outlook (WEO), the World Bank’s Global Development Finance (GDF) database, and the IMF’s International Financial Statistics and Balance of Payments Statistics Yearbook (IFS/BOPSY) follow the classification scheme set out in the 1993IMF’s Balance of Payments Manual, 5thedition (BPM5). The key concepts used in this note are defined in the BPM5 classification scheme as follows:

The BPM5 Financial Account is defined to cover “all transactions associated with changes in ownership in the foreign financial assets and liabilities of an economy,” including the creation or liquidation of such claims. This account includes most of the flows considered to be in the capital account in early editions of the BPM. These comprise:

  • Foreign Direct Investment (FDI), which includes loans as well as equity from the foreign direct investor;

  • Portfolio Investment which includes both debt and equity securities (apart from those included in FDI); and

  • Other Investment, a residual category that includes cross-border loans, trade credits, deposits, and all other financial account asset flows not included in FDI or portfolio investment.

The financial account items measure net changes in stocks, and does not distinguish between changes due to new lending, amortization, and debt forgiveness (an offset for the last of these is recorded separately in the capital account

  • International Reserves are an asset item within the financial account.

The BPM5 retains a Capital Account but this is limited to:

  • Debt Forgiveness (a positive entry to balance the reduction in stocks shown in the financial account) and

  • Other capital transfers, which may be official or private transfers, and other items.

The WEO data separate out debt forgiveness, but the other items are identifiable only in the aggregate defined by the gap between total capital account and debt forgiveness.

Flows and transactions outside the capital and financial accounts that are used in this note include:

  • Current Transfers, which includes foreign aid grants (within official transfers), remittances from non-residents (in private transfers), and private aid donations to the government or local NGOs (also in private transfers).

  • Net Errors and Omissions: Much of the literature of capital flight and net flows to low-income countries emphasizes the potential of net errors to capture disguised capital flows, particularly when legal restrictions deter open reporting of financial account outflows.

B. Data Sources

3. WEO data are used as the primary source for this note because they are the most comprehensive in terms of coverage of the balance of payments components and also in terms of countries and time periods. Two alternative sources, the World Bank’s Global Development Finance (GDF data) and data from the Balance of Payments Statistics Yearbook (BOPSY)/International Financial Statistics (IFS) database are discussed below.

4. WEO data have a comprehensive coverage of current, capital, and financial account asset and liability items. WEO data together with a systematic division of transactions into private and official sources. WEO data have nearly complete country coverage (74 out of 78 LICs have usable data for at least the most recent years), and relatively few omitted items. Offsetting this is the fact that WEO data are often IMF staff estimates, based on preliminary or incomplete information.

5. The BOPSY/IFS data have the same comprehensiveness as WEO data in terms of current, capital, and financial account coverage and potentially greater disaggregation into more detailed components, but are severely plagued by missing observations. For example, the October 2007 WEO database (compiled in August 2007) has data through 2006 for 74 of 78 LICs (albeit Fund staff estimates rather than official data in many cases), the August 2007 published IFS (approximately the same vintage) has data through 2006 for only 22 countries accounting for much less than half of LIC GDP. For the other countries, there is no usable IFS financial or capital account data at all for 15 of these countries, no usable data more recent than 2004 for another 14, and data only through 2005 for further 27 LICs. Even where data are available for some periods, there are often omissions for particular data categories. While greater disaggregation of BOP items than in the WEO data is available in principle, the additional detail is very often unavailable for LICs, and there are many omissions in the data for even major categories.

6. The GDF data resemble the WEO data in terms of country coverage and the use of staff estimates in lieu of official data, but they are less comprehensive in coverage of the full range of current, capital, and financial account items discussed in this paper. While GDF data are classified along the same BPM5 lines as the WEO data, the coverage of BOP data items is more limited and more focused on liabilities to official creditors, rather than LIC assets, or non-financial account flows.

7. The table below summarizes some features of the WEO, GDF, and BOPSY/IFS data as well as the OECD’s Official Development Finance statistics; the United Nations data on foreign direct investment; and BIS data series on (private) external loans.

C. Comparing WEO, IFS, BOPSY and GDF data

8. The level of agreement among various data sources is strongest for FDI data, and relatively weaker for data on portfolio and other investment flows of the financial account. As the following tables show (see Annex Tables 2 and 3), in recent years FDI data reported by the WEO, GDF and UN sources are relatively similar; the markedly different FDI data seen in IFS and BOPSY sources reflect gaps in country coverage of these 2 cases. Data for “other investment” show large differences across data sources, which is not unexpected since—besides gaps in country coverage—there are known differences in definitional coverage. More specifically, the GDF, OECD, and BIS offer only partial coverage of “other investment” flows focusing on loans, official loans, and private loans respectively.

ANNEX TABLE 1Low-Income Country Capital Flows: Overview of Data Sources
IMF’s

World Economic

Outlook
IMF’s

International

Financial Statistics
IMF’s

Balance of

Payments Statistical

Yearbook
World Bank’s

Global Development

Finance
OECD’s

Official

Development

Finance
United Nations

Investment

flows data
BIS

Quarterly Review

of External Loans

and Deposits
Capital flows covered 1/All “capital flows” (see footnote 1)All “capital flows” (see footnote 1)All “capital flows” (see footnote 1)All “capital flows”, except income credits, official transfers, reserve assetsOnly grants, concessional loans, debt reliefOnly FDIOnly private bank loans
General featuresMost comprehensive source; staff estimates are used, as needed.Reliant on official reporting; extensive gaps in data seriesReliant on official reporting; extensive gaps in data seriesComprehensive series on loan flows and debt stocks; uses official statistics & estimatesData from creditor and donor countriesData compiled by UN, but sourced from IMF and WB.Data from private commercial banks
Data availabilityVery few missing data pointsMissing data points for about one-third of PRGFsMissing data points for about one-third of PRGFsVery few missing data pointsVery few missing data pointsVery few missing data pointsVery few missing data points
Private vs. Official flowsAvailableAvailableAvailableAvailableOfficial flows onlyPrivate FDI flowsPrivate flows only
ANNEX TABLE 2Comparisons of Capital Flows Data Among Alternative Data Sources
IMF’s

World Economic

Outlook
IMF’s

International

Financial

Statistics
IMF’s

Balance of

Payments

Statistical

Yearbook
World Bank’s

Global

Development

Finance
OECD’s

Official

Development

Finance
United Nations

Investment

flows data
BIS

Quarterly Review

of External Loans

and Deposits
Availability of data on...
Foreign Direct InvestmentYESYESYESYESYES
Portfolio FlowsYESYESYESYES
Other Investment flowsYESYESYESPARTIALPARTIALPARTIAL
Sample Data for PRGF Group: 2000-2004
IMF’s

World Economic

Outlook
IMF’s

International

Financial

Statistics
IMF’s

Balance of

Payments

Statistical

Yearbook
World Bank’s

Global

Development

Finance
OECD’s

Official

Development

Finance
United Nations

Investment

flows data
BIS

Quarterly Review

of External Loans

and Deposits
Reported data on...
Foreign Direct Investment
Year 200011.212.312.613.513.3
Year 200114.314.214.316.816.7
Year 200218.416.516.719.819.4
Year 200322.419.419.623.322.9
Year 200425.515.315.524.223.9
Portfolio Flows
Year 20002.22.32.95.7
Year 20013.13.43.50.2
Year 20020.20.70.61.7
Year 20037.88.38.13.0
Year 20049.20.30.113.6
Other Investment flows
Year 20001.9-11.5-3.75.76.3-3.4
Year 20013.2-8.7-3.50.27.4-6.7
Year 20028.6-12.8-1.91.78.31.2
Year 20037.6-13.8-9.23.06.011.4
Year 20049.5-20.9-18.313.65.110.3
Notes: Italicized entries indicate that definitional coverage used by given source does not match WEO definitions.Shaded entries indicate that there are gaps in country coverage that render source data incomparable with WEO aggregates.Entries marked with “…” indicate data is not available from that particular data source.
Annex Table 3:Correlations in Financial Account Aggregates Across Sources All LICs
WEO vs. BOPSY

1990–2005
WEO vs. GDF

1990–2004
GDF vs. BOPSY

1990–2004
All liabilities-0.070.730.05
FDI0.800.960.98
Portfolio liabilities0.620.700.45
Other investment0.750.770.62
WEO vs. BOPSY

1990–2000
WEO vs. GDF

1990–2000
GDF vs. BOPSY

1990–2000
All liabilities0.710.630.67
FDI0.970.971.00
Portfolio liabilities0.890.670.67
Other investment0.750.710.67

Figure A1.All Financial Account Liabilities in LICs WEO versus GDF data

Figure A2.FDI Liabilities

Figure A3.Portfolio Liabilities in LICs WEO versus GDF data

Figure A4.Other Liabilities (e.g., loans) in LICs WEO versus GDF data

ANNEX – B

How Robust Are These Patterns?

A. Sensitivity Analysis on Weights and Country Groupings

1. This paper outlines trends across LICs as a group and major subgroups of LICs, and finds a broad-based and strong increase in private inflows, dominated by private transfers and FDI. Trends are considered for regional subgroups (African LICs, India, and other South and East Asian LICs), groups of LICs differentiated by commodity export status (hydrocarbon-rich LICs, mineral-rich LICs, and other LICs) and HIPC Initiative status. However, with more than 70 LICs varying widely across regions, widely differing in size, and differentiated across many other economic and non-economic characteristics, there is an almost unlimited array of country cases and regional and other sub-groupings to consider. This annex describes some of the robustness checks underpinning the conclusion that these trends are broad-based, even beyond the specific groups that are the focus of the main text.

Alternative country weights

2. The main text reports results in terms of flows as shares of a country group’s collective GDP (i.e., total flows for the group divided by total GDP for the group). This is equivalent to GDP-weighted averages. Two alternatives suggested in comments on earlier drafts of the paper are population weights and an unweighted average across countries.

3. Weights based on population rather than GDP would make only a marginal difference in the reported results given the relatively similar levels of per-capita GDP across groups. For the three regional groups (India, African LICs, and other South and East Asian LICs), the per capita income of all three groups is within 20 percent of the average per capita income for all LICs in 2006, 1996, and 1986. Moreover, each of the three country groups has the highest per capita income in one of these three years. There is somewhat more dispersion among the groups classified by resource endowment and debt-relief status, but all observations for these six groups in 2006, 1996, and 1986 are within 50 percent of the average for all LICs and all but three of the 18 observations are within 20 percent of the average per capita income for all LICs in the relevant year.

Figure B1.Private Transfers and FDI GDP Shares Weighted (excluding India) and Unweighted Averages

4. Unweighted averages across countries represent a more radical departure from the GDP weighted averages presented in the main text, but the difference in results is modest. As noted in the main text, the half of LICs at the lower end of the GDP size range account for less than 4 percent of aggregate LIC GDP, while India accounts for more than half of LIC GDP. An unweighted average across LICs reduces India’s weight (and that of other large LICs) to just over one percent each, while 14 small island states collectively accounting for less than 0.4 percent of LIC GDP account for 19 percent of the total weight. However, apart from the effect of the near-elimination of India’s weight, the unweighted averages are tell essentially the same story as the GDP-weighted averages (Figure B.1). The unweighted averages show very similar strong increases in FDI to the GDP-weighted LIC averages (excluding India). The private transfers trends show similar trends, but somewhat higher absolute levels in the unweighted average, which is consistent with small PRGF-eligible islands and other small states having some of the highest remittance-to-GDP ratios in the world (IMF, 2005).

Weights Within Country Sub-groups

5. A different but related concern with GDP weights is the possibility that a country in a sub-group of LICs with a large share of the group’s GDP could swamp the influence of the other countries. Just as India is considered separately from the other regional groupings, there is the possibility that trends in the largest countries in some groups may overwhelm the influence of the smaller members with very different patterns of capital and capital-like flows. While no member of any subgroup discussed in the main text accounts for a majority of the group’s GDP, Nigeria accounts for 32 and 28 percent of the 2006 GDP of hydrocarbon-rich and African LICs, respectively, Pakistan accounts for 37 percent of the 2006 GDP of the South and East Asian LICs other than India, and Sudan accounts for 42 percent of the GDP of the pre-CP HIPCs.

6. Comparison of trends in these “dominant” countries within the subgroups with those of the other members reveals little difference between them in most cases.Figures B.2 and B.3 show the evolution of private transfers and FDI between Nigeria and other African LICs, Nigeria and other hydrocarbon-rich LICs, Pakistan and other South and East Asian LICs, and Sudan and other pre-CP HIPCs. In every case, the broad trends identified for the subgroup (and LICS as a whole) seem to be applicable both to the dominant member of the group and the GDP-weighted average of the other members. For private transfers, the trends are similar in all cases although the “dominant” country tends to exhibit more volatility than the average of the other countries. For FDI, there is also a great deal of similarity, except in the case of Sudan relative to other pre-CP HIPCs. FDI surges in Sudan after 2000 (presumably due to hydrocarbon-related FDI) while the increase in FDI to the other countries is much more modest.

Figure B2.Private Transfers: Percent of GDP

Figure B3.Foreign Direct Investment: Percent of GDP

Notable exceptions are Bhinda et al. (1999) which examines several aspects of private capital flows to Africa in the early and mid-1990s and Dudine et al. (2006) which examines the financing response of LIC oil importers to the recent oil shock.

See for example, the most recent World Economic Outlook (IMF, 2007) and the World Bank’s Global Development Finance, 2006 (World Bank 2006).

Low-income countries (LICs) are defined throughout this paper as the current (2007) set of PRGF-eligible countries. Countries that have been removed from the PRGF-eligible group in earlier years (e.g., China and Egypt) are excluded from the data for all years. Countries that have been added to the list (e.g., Papua New Guinea) are included in the data set for the entire 1980–2005 period. Usable data are unavailable for Afghanistan, Liberia, and Somalia; data for Mozambique are excluded because large reclassifications between private and official sources cannot be disentangled from actual transactions.

The discussion in this paper is discusses inflows and outflows largely as shares of LIC country GDP because the large number and disparate sizes of the LICs makes comprehensive reporting of individual country data unwieldy and potentially misleading (e.g., half of these LICs collectively account for less than 4 percent of LIC GDP).

All data cited in the text and charts in this note are taken from the Fall 2007 WEO database unless indicated otherwise.

The WEO database subdivides most capital and financial account items into private and official components, for most but not all items, leaving a small unallocated residual. In particular, the WEO database separates debt forgiveness from the total capital account, but does not further subdivide capital transfers (and other minor items) further into private and official components.

Foreign direct investment potentially includes debt for foreign direct investors as well as equity. However, the share of debt in FDI that is reported in the WEO data for LICs is low and on a downward trend; it has not exceeded three percent of total FDI in LICs since 1997.

The data used in Figure 5 show the current account balance less current transfers to avoid double counting, as transfers are included in capital and capital-like flows. Data on the current account with and without transfers are included in the Tables 1 and 2.

African LICs are defined as all PRGF-eligible countries on the African continent (including Djibouti and Mauritania), plus Cape Verde, the Comoros, Madagascar, and Sao Tome and Principe. South and East Asian LICs are defined as the PRGF-eligible countries on the Asian continent from Pakistan eastward, plus Sri Lanka (i.e., Bangladesh, Bhutan, Cambodia, India, Lao P.D.R., Mongolia, Myanmar, Nepal, Pakistan, Sri Lanka, and Vietnam).

Discussion of regional patterns is confined to the three major regions identified in Figure 6. The small size and diverse composition of the other LICs suggest that consideration of this residual as a group would not be meaningful while consideration of more cohesive subgroups would be unwieldy.

Among the studies that conclude that Africa is particularly disadvantaged in attracting FDI inflows are Asiedu (2002) and UNCTAD (2005). Studies that find an adverse effect to African location in more general economic context include Barro and Sala-i-Martin (1995) and Sala-i-Martin (1997).

Countries are classified as pre- or post-completion point HIPCs based upon their status at end-2006.

Official-source inflows to pre-CP HIPCs are more volatile with multiple peaks and troughs. This is true even setting aside a very high peak in 1987 that is attributable to a debt restructuring in Cote d’Ivoire which transformed a large amount of liabilities to private creditors into liabilities to official creditors.

The Guide to Resource Revenue Transparency (IMF 2005) defines 2 groups of countries as hydrocarbon-rich and mineral-rich. The two groups have some overlap and this paper assigns those in both categories to “hydrocarbon-rich” to avoid double counting. Those LICs in neither group are categorized as “non-mineral” LICs in this paper.

Among these other explanations for net errors and omissions are incomplete or overlapping coverage, inconsistent times and basis for recording, unavailability of data, non-uniform prices, and inconsistent currency conversions.

Other Resources Citing This Publication