W O R L D    B A N K    P O L I C Y    A N D    R E S E A R C H

Why do savings rates vary across countries?

A recent World Bank research project sheds light on the many factors affecting saving. The work has important policy implications for developing countries, particularly in Latin America and Sub-Saharan Africa, where savinfs rates have stagnated or devlined over the past 30 years.

Over the past three decades differences in savings rates across countries have widened significantly. The gap between industrial and developing countries has grown, and so have differences across regions (figure 1). Savings rates nearly doubled in a handful of countries in East Asia, where they averaged almost 30 percent of disposable income between 1984 and 1993. But in most of the rest of the world saving fell. In OECD countries saving declined from more than 25 percent of gross national disposable income in 1965 to less than 20 percent in 1995. In transition economies the rate of saving has declined since the onset of reform in the late 1980s. Latin America and the Caribbean's already modest savings rates fell to about 15 percent of disposable income, and the low rates in Sub-Saharan Africa to between 10 and 15 percent.

Savings rates and long-term income growth move together, suggesting a virtuous cycle in which high rates of saving lead to greater prosperity, which in turn leads to even higher rates of saving. Countries with low saving may be caught in a poverty trap. In these countries a low rate of saving leads to stagnation, which further reduces the rate of saving, dooming the country to poverty.

A large body of literature has shed light on different aspects of consumption and savings behavior, but it has left a long list of policy-related questions on saving largely unanswered. A recently completed World Bank research project, Saving across the World, addressed many of these questions, examining three broad issues:

• Why do savings rates differ so markedly across (and within) regions and periods?
• What is behind the savings-growth relationship? Does a higher growth rate lead to a higher savings rate, or a higher savings rate to a higher growth rate?
• Which policy measures are most effective in raising national saving? And which are unlikely to work?

The World Saving Database

Data on saving are affected by a host of measurement problems that make analysis of cross-country saving difficult. The Bank's Saving across the World project corrected some of these problems, creating the largest and most systematic annual time series on country saving and savings-related variables available.

The World Saving Database covers 112 developing and 22 industrial countries for up to 35 years (1960-95). It includes a broader set of countries than earlier databases, corrects inconsistencies in national accounts data and savings determinants, standardizes definitions relating to the composition of the public sector, and includes a set of related savings and consumption determinants, such as income growth rates, interest rates, monetary aggregates, and demographic indicators. The database includes both raw data and data that have been adjusted to account for capital gains and losses from inflation and real exchange rate devaluation. (The data set and underlying documentation are available on the Internet; for information go to www.worldbank.org/research/projects/savings/data.htm.)

What drives private savings rates?

The research project identified eight factors that drive saving: income, growth, fiscal policy, pension reform, financial liberalization, external borrowing and foreign aid, demographics, and uncertainty. It yielded a rich set of empirical results on the effect of each of these factors on private (and national) savings rates during the past three decades.

Income

Real per capita income is positively correlated with private savings rates, suggesting that policies that spur development are an effective way of raising private saving. The influence of per capita income is larger in developing economies than in industrial countries, suggesting that the effect of the level of development on saving tapers off at medium or high levels of income (figure 2).

Growth

Growth and saving are positively correlated. Much of the causation appears to run from growth to saving, however, rather than from saving to growth. A sustained increase in the level of saving is typically followed by an acceleration of growth that persists for several years but eventually disappears. By contrast, sustained increases in growth are associated with permanent increases in the rate of saving. The effect of real income growth on saving appears substantial, with a 1 percentage point increase in the rate of growth raising the private savings rate by 1 percentage point.

Fiscal policy

Often advocated to increase national savings rates is changing the rate of public saving or introducing tax incentives. A permanent increase in public saving of 4 percent of gross national disposable income will raise national saving by 3 percent of disposable income within a year and 1.5 percent over the long term. The private savings response to changes in public saving varies widely across countries, however. In India private saving falls by 30 percent of the increase in public saving; in Mexico it declines by almost 80 percent.

Tax incentives have only small effects on national saving, particularly when the negative effects on public saving are taken into account. But tax incentives may change the composition of saving without changing the overall savings rate.

Pension reform

Many developing countries, particularly in Latin America, are replacing their pay-as-you-go pension system with a fully funded scheme-a reform often advocated because of the belief that pay-as-you-go pension systems reduce saving.

Countries with pay-as-you-go systems do tend to have lower savings rates than those with fully funded systems, other things equal. But the direct, short-term effects of pension reform on saving will depend on how the government finances the transition deficit. Theoretical models suggest that debt financing has no effect on savings rates in the short run, while the effects of tax financing are ambiguous. The welfare of low-income, borrowing-constrained earners may be reduced, however, if pension reform forces them to save more than they would have chosen to save.

In the long run the indirect effects of pension reform may predominate. If reform is effective in reducing labor market distortions and spurring capital market development, these indirect effects may increase efficiency and growth, raising the rate of saving.

Financial liberalization

Financial liberalization has long been thought to encourage aggregate saving. But analysis of eight financial liberalization episodes (in Chile, Ghana, Indonesia, the Republic of Korea, Malaysia, Mexico, Turkey, and Zimbabwe) failed to find a systematic direct effect on saving. (The effect was negative in Korea and Mexico, positive in Ghana and Turkey, and negligible in Chile, Indonesia, Malaysia, and Zimbabwe.)

Financial liberalization that includes interest rates usually results in a rise in real interest rates on deposits. This means that each unit of resources saved yields more interest income, reducing the need to save. But the "price" of current consumption rises with interest rates: households might save more because a unit of income saved now would yield greater interest income in the future. Liberalizing interest rates will increase saving only if this intertemporal substitution effect overcomes the income effect.

There is no consistent evidence supporting a positive net effect of liberalization on saving. If liberalization expands the supply of credit to agents that had been credit-constrained, this can reduce saving, because easier access to credit reduces the need to set aside resources in anticipation of adverse income changes. Cross-country evidence reveals that a 1 percentage point increase in the private credit flow to income ratio reduces the long-term private savings rate by 0.74 percentage point.

External borrowing and foreign aid

Higher levels of external borrowing tend to replace private (and thus national) saving. But the effect of foreign aid on saving is unclear. Cross-country time-series data from Africa suggest that higher foreign aid tends to reduce national saving (that is, a large fraction of aid is consumed). But this evidence may reflect the fact that aid to poor countries increases at times of adverse income shocks, when saving is lowest.

Scrutiny of the countries that have moved from low to high savings rates reveals that increases in foreign aid are positively associated with takeoffs of both private and national saving. Aid may reduce saving in some countries and increase it in others. If a country is unable to borrow all it wants, foreign aid would be expected to replace domestic saving. But if the country is undergoing economic reform that invites more aid and induces higher investment and growth, aid and savings rates would tend to move together. This is probably what we see in the takeoff economies.

Demographics

The data confirm that an increase in the share of either young or elderly dependents in the population tends to reduce private saving. An increase in the young age dependency ratio of 3.5 percentage points leads to about a 1 percentage point decline in private saving; an increase in the old age dependency ratio has a negative savings effect more than twice as large. Developing countries undergoing a demographic transition-in which the working-age population is a large and growing share of the population-may witness a transitory increase in their savings rates. This may persist until the next demographic stage, in which elderly dependents become a large share of the population.

Uncertainty

Increases in uncertainty should increase saving, as risk-averse consumers set resources aside to protect themselves against possible adverse changes in income, taxes, and other factors. Evidence shows that inflation (conventionally taken as a measure of macroeconomic volatility) is positively correlated with private saving (holding income, growth, real interest rates, and other factors constant).

Research conclusions and policy lessons

The Saving across the World project has increased our knowledge of saving and yielded several important policy lessons:

• Disparities in savings rates across countries are large and have increased over the past three decades. Cross-country disparities can be explained largely by differences in income levels, growth rates, economic policies, and demographics.
• Income growth and savings rates move together. The positive correlation between growth and saving is partly explained by the fact that growth contributes to saving. But increases in the rate of saving are followed by permanent increases in income levels, suggesting that higher saving accelerates the move to higher income.
• Increasing public saving can increase the national savings rate. But offering tax incentives is not likely to increase the private savings rate.
• Contrary to the traditional view, financial liberalization can reduce private saving in the short run. But in the long run indirect effects of financial liberalization are likely to increase saving.
• Replacing a pay-as-you-go pension system with a fully funded scheme may not increase saving in the short run. But again, in the long run indirect effects of reform are likely to increasing saving.
• Better access to foreign financing tends to reduce national saving, with aid flows in particular largely financing consumption. Some economies have achieved lasting increases in their savings rates following increases in foreign aid, however.
• Macroeconomic stabilization can both increase and decrease saving. On the one hand, macroeconomic stability reduces the precautionary motive for saving, thereby decreasing saving. On the other hand, fiscal adjustment through higher public saving raises national saving. The net effect on saving is likely to be positive because growth will increase as a result of stabilization.

For further reading

Loayza, Norman, Klaus Schmidt-Hebbel, and Luis Servén. 1999.

"What Drives Saving across the World?"
www.worldbank.org/research/projects/savings/whatdriv.htm.

Rodrik, Dani. 1998.

"A Study of Transitions from Low to High Saving Rates."
www.worldbank.org/research/projects/savings/trnsition.htm.

Schmidt-Hebbel, Klaus, and Luis Servén, eds. 1999.

The Economics of Saving. Cambridge: Cambridge University Press.

_____. 1997.

Saving across the World: Puzzles and Policies.
World Bank Discussion Paper 3354. Washington, DC.

Other papers and data from the savings project are available at www.worldbank.org/research/projects/savings/policies.htm.