A quarterly magazine of the IMF

June 2008, Volume 45, Number 2

A Capital Story
Thomas Dorsey

The sharp increase in private investment flows to low-income countries is an untold development success story

The dramatic growth in private capital flows to low-income countries (LICs) around the world over the past quarter century is one of the good news stories of development. Capital flows and capital-like flows, such as aid grants and remittances that can substitute for capital flows, are at historic highs, spearheaded by a sharp rise in foreign direct investment in LIC economies.

While government-to-government grants and loans to these developing countries have grown only in line with the rise in LIC GDP, private-source inflows have quadrupled relative to LIC GDP since the 1980s, and foreign direct investment (FDI) has risen even more dramatically, helping to spur development. The rise is taking place at a time when promised large increases in official aid to poor countries have so far failed to materialize, and the UN Millennium Development Goals—which include reducing poverty and improving living standards—remain stubbornly out of reach.

The LICs are a diverse group united by little other than low per capita incomes. India is the biggest, with 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. The littlest is Dominica, with about 70,000 people and a GDP of $0.02 billion. The LICs can be found around the globe: in Africa, Asia, the Caribbean, eastern Europe, Latin America, and the south Pacific.

Capital flows to LICs are a poorly understood and poorly researched topic. Most of the attention so far on international capital flows into developing economies has focused on the large emerging markets, in part because LICs account for a relatively small share (5–20 percent) of developing country capital inflows in most asset and liability categories. And the work that has been done specifically on LICs has focused on capital flight from these countries and on official flows, with a heavy emphasis on official debt and official debt forgiveness.

But new work by the IMF attempts to remedy this omission by analyzing trends in capital and capital-like flows to LICs (Dorsey and others, 2008; see box). This article looks at the story behind the numbers relating to this dramatic change in financing for the world's poorer countries and examines whether there are downsides to this trend.

Assessing the new flows
If LICs were simply poor economies that received proportionately lower inflows than other developing countries, the absence of a specific focus on LICs in the capital flow literature would not matter. However, the reality is that the composition of capital flowing into LICs is very different from that going to higher-income countries. Loans from government and foreign aid grants are much more important in LICs than in higher-income countries, whereas loans from banks and sales of stocks, bonds, and other securities are less important in LICs than in emerging markets.

For the purposes of this study, we identified LICs as the 78 countries eligible for loans from the IMF's concessional lending facility, called the Poverty Reduction and Growth Facility.

What we found is that capital and capital-like inflows to this group of countries increased from about 4 percent of LIC GDP in the 1980s to more than 10 percent of LIC GDP by 2006 (see Chart 1)—with the growth driven almost entirely by inflows from private sources. Indeed, private inflows, such as FDI and remittances, have risen from about half of total inflows in the 1980s to become the dominant source of inflows by the current decade.

Increased flows are being attracted by improved economic policies, a liberalized trade and investment environment, and more stable economic performance in many countries.

A number of poor countries, particularly in Africa, have been given a fillip by recent debt relief from official creditors, including the IMF and the World Bank. And although circumstances in individual countries vary, overall new inflows are associated with improving current account balances and are not building up new unsustainable debt.

In terms of official flows, government-to-government grants and loans have grown only at the same pace as LIC GDP, and their share of total inflows has fallen steadily. But there is a pronounced trend within official flows toward the substitution of grants for loans (see Chart 2). Consistent with announced donor policies, official current transfers (grants) tripled from roughly 0.5 percent of LIC GDP in the early 1980s to 1.3 percent of GDP in 2006.

Mirroring this, official lending to LICs (net of debt forgiveness) declined from an inflow of about 1.5 percent of LIC GDP through the 1980s and 1990s to an outflow of about 0.2 percent of LIC GDP by 2006. If debt forgiveness is not netted out from the changes in official lending (that is, if debt forgiveness is counted as a capital outflow), official lending (measured as the change in liabilities to official creditors) turned negative in 2006.

As for private flows, the study shows that they have grown more than fourfold since the 1980s (see Chart 3). The most striking feature of the shift in capital flows to LICs is the more than tenfold 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 current transfers, which include workers' remittances and other private transfers. This category 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 2006. Private inflows other than FDI and private transfers averaged only about 1/2 percent of LIC GDP in the 1980s and 1990s, but rose to just above 1 percent of GDP between 2003 and 2006.

Along with the increased inflows, LICs have also been accumulating official reserves (see Chart 4). 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 reduction in borrowing because the higher reserve levels allow LICs to forgo borrowing in the event of shocks and draw down reserves instead.

An Africa story, too
Is India or the other Asian LICs driving these results? In other words, do the results also apply to Africa? This is a fair question given the vast differences among LICs and the fact that differences in region, size, and other factors may have implications for economic analysis and performance. India accounts for roughly half of LIC GDP, and African LICs and south and east Asian LICs each account for roughly one-fifth of LIC GDP. 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 the remaining 7 percent.

As it turns out, capital flows across the three larger regions all exhibit rising private inflows and relatively steady official inflows. Private flows have surged in Africa, India, and other south and east Asian LICs in broadly similar proportions, rising from 1 percent to 3 percent of GDP in the 1980s to at least 6 percent and as much as 10 percent of GDP by 2006.

There are some differences in the composition of private inflows across regions. FDI is the most important source of inflows to the African LICs, but private transfers are more important in south and east Asian LICs, including India. However, both components are increasing strongly in all three regions.

Official flows started at very different levels in the three regions—consistently higher in African LICs than in south and east Asian LICs and very low in India. Private flows to African LICs have converged with those in Asian LICs in recent years, suggesting that the disincentives to investment in Africa in prior decades are dissipating.

The increase in private flows is broad based across other dimensions of LIC diversity as well. For example, the surge in private inflows takes place across mineral- and oil-rich countries and nonmineral countries in roughly similar proportions. Mineral- and oil-rich LICs have higher levels of FDI, whereas nonmineral LICs have higher levels of remittances and other private transfers (Dorsey and others, 2008). The increase in private inflows and broadly parallel trends in other dimensions of this story apply across LICs with different indebtedness and debt-relief histories.

Story behind the story
The reasons for the dramatic shift are diverse. Privatization or the opening of the economy to foreign acquisition of existing firms may create one-off opportunities for FDI. In addition, some LICs that had restricted even new greenfield investment by foreign firms may now be more open. These policies are hard to quantify, but there appears to have been a trend toward liberalization in recent years (Reddy, 2007).

Improvements in the general investment environment, including trade policies and policies affecting the ease and cost of setting up and operating businesses, also help (Busse and Groizard, 2006; Bénassy-Quéré, Coupet, and Mayer, 2007; and Naudé and Krugell, 2007). As with direct regulation of FDI, there has been a trend toward liberalization of trade and the business environment in LICs that coincides with the increase in FDI.

Broader policy considerations may also encourage investment in LICs. Inflation and the fiscal and external balances in LICs have improved in the past decade (Selassie and others, 2006), and this has strengthened the investment environment. Other factors such as political stability are harder to quantify but have undoubtedly promoted greater FDI where conditions have improved.

Economic developments elsewhere in the world may also have fostered increased FDI. The decline in yield on investments in advanced and emerging market economies has led to a search for higher-yielding investments, including "frontier" LIC markets. Also, the rapid increase in assets in sovereign wealth funds has added impetus to the hunt for new investment opportunities in underexplored venues such as the LICs.

Remittances are closely associated with outward migration, but policies matter, too. Lower transactions costs (Suro and others, 2002; Fajnzylber and Lopez, 2007) and the absence of exchange restrictions, black market exchange rate premiums, and unstable macroeconomic environments (IMF, 2005) have been associated with higher remittances. However, these studies are drawn mostly from the experiences of migrants from middleincome countries, and more work is needed to assess the applicability of this literature to LICs.

In any case, there is a broad consensus that remittances are less volatile than other private and official capital flows (Buch and Kuckulenz, 2004) and that remittances tend to move countercyclically with recipient country income (Chami, Fullenkamp, and Jahjah, 2005).

Policy implications
Should policymakers be concerned about the size of these flows? After all, large or increasing capital inflows often give rise to concerns that the inflows are financing unsustainable current account deficits or increasing countries' vulnerability to capital account crises. Some key considerations in evaluating capital inflows include
• whether inflows are financing a large or widening current account deficit,
• whether they are creating potentially unsustainable debt accumulations, and
• the extent to which countries are saving some of the inflows in the form of international reserves.

By each of these criteria, the capital and capital-like inflows to LICs appear benign. The aggregate current account deficit of LICs shrank from about 3 percent of LIC GDP in the mid- 1990s to near balance in 2006. Reserve accumulation moved from about zero in the late 1990s to steadily increasing inflows, reaching nearly 4 percent of GDP a year by 2006— providing some insurance against external shocks or a slowdown in inflows. Only a small part of the inflows have been debt creating because they are dominated by equity FDI and transfers. And there is little sign of offsetting nonreserve outflows (for example, capital flight). 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 data are unavailable because of long lags and substantial gaps.

The bottom line is that the trends in capital and capitallike flows to LICs do not appear to carry large risks, but they do present several new challenges for policymakers.

First, the shift from official to private financing implies a less direct role for LIC governments in determining the uses of external financing. The greater stability of these private flows and the fact that the shift to private-to-private flows may empower heretofore underdeveloped LIC private sectors suggests the trend is a positive development. With the predominance of private flows, the savings-investment and other choices for the use of inflows are being made by private parties, and government influence on these flows diminishes.

Second, the higher private inflows imply more stable and diversified financing for LICs. However, private inflows may increasingly become the main sources of external vulnerability in LICs because they could be reversed in a manner not under the control of LIC policymakers. These new inflows may require authorities to reconsider policies to address concerns about sustainability, effects on relative prices and competitiveness, and accompanying policy and institutional reforms.

However, it is difficult for LIC policymakers to calibrate their response to shifts in private flows to the extent that they cannot monitor them. Weaknesses in national balance of payments data may undermine the basis for LIC economic decision making. To the extent that timely data are unavailable to LIC policymakers, policymaking may be flying blind. Thus, it is in the interest of LIC authorities to improve balance of payments data, with priority given to these rapidly increasing sources of private inflows.

Third, regardless of data uncertainties, external sector policies in LICs need to focus on policies relevant to the new inflows. Strong debt management and good donor relations are particularly important for encouraging and managing official lending. However, maintaining a stable macroeconomic environment, a favorable business climate, and efficient mechanisms for international transfers by households are likely to be more important for private inflows.