This Selected Issues paper on the United States explains the behavior of inflation and unemployment during 1997–98. The paper highlights that a simple Philips curve equation relating inflation to the unemployment gap has overpredicted inflation since 1993. The mean forecast error for 1994–97 is greater than zero by an amount equivalent to two-thirds of the standard deviation of the forecast error. The paper also examines the developments in the U.S. stock prices. Alternative approaches to social security reform are also discussed.

Abstract

This Selected Issues paper on the United States explains the behavior of inflation and unemployment during 1997–98. The paper highlights that a simple Philips curve equation relating inflation to the unemployment gap has overpredicted inflation since 1993. The mean forecast error for 1994–97 is greater than zero by an amount equivalent to two-thirds of the standard deviation of the forecast error. The paper also examines the developments in the U.S. stock prices. Alternative approaches to social security reform are also discussed.

VI. LONG-TERM PROSPECTS FOR THE U.S. CURRENT ACCOUNT BALANCE1

1. The persistence of external current account deficits in the 1980s led to a shift in the international investment position of the United States from a substantial net asset to a net liability position, and with continuing current account deficits being registered in the 1990s, the net liability position has grown rapidly. The appreciation of the U.S. dollar since mid-1995, together with the effects of the financial crisis in Asia, is contributing to a substantial widening in the current account deficit in 1998.

2. This situation has given rise to concerns that if deficits of the magnitude expected in 1998 persist, at some point in the next few years a disruptive adjustment in the current account might occur. Such an adjustment could be prompted by a sharp decline in the value of the dollar, as foreign investors become less willing to continue to accumulate substantial amounts of dollar-denominated assets. Also, given the current “low” level of private saving and the aging of the U.S. population, concerns have been expressed that, over the longer term, national saving in the United States may not be sufficient to meet domestic investment needs without continuing heavy reliance on foreign saving (i.e., large external current account deficits). In such circumstances, the exchange rate of the dollar would be expected to remain under downward pressure in real terms and real interest rates would be high, holding down investment and growth.

3. The concerns about the longer-term sustainability of the U.S. external current account position are often illustrated by simulations derived from partial-equilibrium models of U.S. international transactions. Such simulations, based on an assumption of a constant real effective value of the U.S. dollar at its current level, show growing imbalances in the U.S. external position over the long term and the need for a sizable real depreciation of the dollar to bring the current account back to a sustainable position.

4. A broader assessment of the long-term prospects for the U.S. current account derived using a general equilibrium model of the world economy, such as the IMF’s multicountry model (MULTIMOD), provides a somewhat different picture. In particular, scenarios using MULTIMOD were developed to look at the implications for the U.S. current account of population aging, alternative fiscal policy rules, and the behavior of private saving and investment. These scenarios suggest that, under a long-term fiscal policy rule that provides for a solution to the financing needs of Social Security and Medicare while maintaining balance in the rest of the budget, deficits in the U.S. current account would decline gradually, and the current account would move to a small surplus in the longer term. World demographic trends would also contribute to an improvement in the U.S. external position, because the U.S. population is expected to age less rapidly than the population in other industrial countries. Under these circumstances, the U.S. dollar would depreciate in real terms on the order of 10–15 percent over the medium term, before appreciating slowly over the long term. The scenarios suggest that U.S. private saving would stabilize as a ratio to GDP, following a decline through the early years of the next century, and then gradually rise, exceeding private investment over the long term. The general conclusion from this analysis is that, provided that appropriate macroeconomic policies continue to be pursued, the longer-term current account position of the United States is expected to strengthen significantly, while the expected real depreciation of the U.S. dollar and the reduction in the external imbalance in the medium term are likely to be orderly.

A. Simulations from a Partial-Equilibrium Model

5. A simulation based on a simple econometric model of the current account illustrates the concern that the external position of the United States is not sustainable over the long term (Table 1). Assuming that the real effective value of the dollar remains unchanged at its current level, the U.S. current account deficit is expected to rise from slightly less than 2 percent of GDP in 1997 to around 2¾ percent in 1998 and remain at that level through 2003. Subsequently, the deficit would rise sharply, reaching roughly 7 percent of GDP after 2020.2 The steady rise in the deficit over this period is primarily driven by increasing debt service on a growing net international debt position; by the middle of the next century, net foreign debt would rise to more than 150 percent of GDP, compared with 15 percent at present.

Table 1.

United States: Projections on U.S. Current Account

(Percent of GDP)

article image
Source: Fund staff estimates.

Defined as the ratio of gross national income to gross domestic product

6. It is easy to draw the conclusion from this scenario that a significant improvement in U.S. competitiveness would be needed over the medium term in order to achieve a sustainable external position. The model suggests that a 40–45 percent real depreciation of the U.S. dollar between 1998 and 2003 would be required to balance the current account by 2003. However, with a real depreciation of this magnitude, the current account would shift into growing surpluses after 2003, and the U.S. dollar would need to appreciate in real terms to maintain the external position in rough balance. The cumulative real depreciation that would keep the external current account in balance by 2020 would be around 20–25 percent.

7. The gloomy picture for the current account deficit that emerges from the simulation of the simple current account model reflects a significant difference in the long-run income elasticity of U.S. exports relative to U.S. imports. In the model used for the simulation, an income elasticity of 1.3 was estimated for exports, while one of 2.2 was estimated for imports.3 The effect of this difference in income elasticities is offset to some extent in the scenario because GDP growth in U.S. trading partners is assumed to exceed that in the United States by roughly 1¼ percentage points, which is roughly in line with the historical average for this differential. The faster rate of GDP growth in U.S. trading partners primarily reflects the growth of developing countries. Over the long run, it would be expected that the income elasticities of U.S. exports and imports would converge, as well as the growth rates of the United States and its trading partners. Such a convergence of income elasticities over time would significantly improve the outlook for the U.S. external position.

8. The large size of the real depreciation of the dollar needed to bring the current account into balance reflects an estimated inelastic response of import volume to the change in relative prices in the model. While the price elasticity of export volume is slightly greater than one, the import price elasticity is somewhat less than one half.4 Also, the size of the potential move in the real exchange rate is overstated by a partial-equilibrium model like the one used here. The real exchange rate in this model bears all of the burden of adjustment, whereas, in the context of a more complete model of the world economy, interest rates would adjust as well to help equilibrate the external position. Moreover, the role of other key factors, such as economic policy developments (particularly fiscal policy) and demographics, in determining the path of the current account over the long term is not captured in a partial-equilibrium model

B. A Multicountry Model

9. To examine more comprehensively the interaction between developments in the United States and those in the rest of the world, long-term prospects for the U.S. external position were assessed using a version of MULTIMOD. Illustrative scenarios were generated using the model over the period to 2070.

10. In the version of MULTIMOD used for this analysis, the world is divided into four areas: the United States; other industrial countries; developing countries that are net creditors; and developing countries that are net debtors. The model is also modified to capture the expected effects of population aging in the United States and other industrial countries. The macroeconomic implications of population aging are likely to be manifested in several ways. On the supply side, the resultant slowdown in growth of the labor force could have implications for the growth and level of aggregate production and gross capital formation. On the demand side, to the extent that consumption patterns vary over individuals’ lifetimes, a changing age structure of the population will affect the saving propensity in the overall economy. Based on World Bank projections, the total dependency ratio5 is expected to increase by around 17 percentage points in the United States and 24 percentage points in other industrial countries over the next half century.6 This trend mainly reflects the relative increase in the number of elderly, stemming from increases in life expectancy and declines in fertility rates.

11. To illustrate the potential effects of changing demographics on key macroeconomic variables, Table 2 presents panel estimates for 21 industrial countries relating saving, investment, and the current account to the dependency ratio. The pooled estimates suggest that a rise in the dependency ratio (DEM) relative to that in other countries tends to lower a particular country’s saving-investment balance as a share of GDP, reflecting a fall in domestic saving that would exceed a decline in investment. Based on the projected increase in dependency ratios, the long-run distributional and aggregate implications of population aging on saving-investment balances for the United States and other industrial countries are shown in Table 3. In spite of the rise in its dependency ratio in the period after 2010, the U.S. saving-investment balance (and its counterpart, the current account balance) is expected to improve in the long run, given the faster rise in the dependency ratio in other industrial countries.7 The results from these panel estimates were incorporated in the simulation version of MULTIMOD.

Table 2.

Panel Estimates: Saving, Investment, and Current Account Equations

article image
Source: World Economic Outlook data base, except as indicated in table notes.a * (**) indicates significance at the 5 (1) percent level

General government balance as percent of GDP minus GDP-weighted average ratio of general government balance to GDP for all countries in the panel.

Ratio of populations age 65 and older and 19 and younger relative to age 20–64, minus GDP-weighted average ratio. Data source: United Nations, World Population Prospects.

Per capita GDP relative to the United States, PPP-adjusted. Data source: Organization for Economic Cooperation and Development, Main Economic Indicators.

Actual output minus potential output (logarithmic difference).

GDP-weighted average ratio of general government balance to GDP for all countries.

GDP-weighted average ratio of dependency ratio for all countries.

Table 3.

Long-Run Effects of Demographic Trend in Industrial Countries

(Deviations from baseline, in percent of GDP)

article image

12. In the main scenarios developed in MULTIMOD, U.S. fiscal policy was assumed to aim at ensuring the long-term actuarial balance of Medicare and Social Security, while maintaining balance in the remainder of the budget (referred to here as the “central” fiscal policy rule).8 For simplicity, it was assumed that this policy rule would be achieved by increasing payroll taxes for Social Security and Medicare by the amounts required to restore actuarial balance in these programs (2¼ and 2 percentage points, respectively) according to the latest reports of the programs’ trustees.9 As a result, sufficient assets would be set aside to meet the future spending requirements of these programs.

13. Long-term projections for the rest of the budget are taken from estimates in the U.S. Administration’s budget for FY 1999, adjusted to reflect differences in the economic assumptions (mainly interest rates and economic growth) underlying these estimates and the results for major economic variables derived in the MULTIMOD simulations. To maintain balance in this part of the budget, federal taxes are assumed to be cut through most of the scenario period, although they do rise slightly toward the end of the projection period when the assets in the Social Security and Medicare trust funds begin to decline. State and local governments are expected to maintain their budgets roughly in balance over the entire scenario period.

14. The central fiscal policy rule would result in a significant rise in the general government budget surplus from about ½ percent of GDP in 1998 to an average of about 3½ percent of GDP in 2004–20 (Table 4). As the baby-boom generation retires, the surplus would decline gradually, to an average of 2½ percent of GDP in 2021–30 and would shift to an average deficit of about ½ percent of GDP in 2061–70.

Table 4.

United States: Long-Term Macroeconomic Projections 1/

article image
Source: Fund staff estimates.

Based on the central fiscal policy rule and the effects of demographic changes.

On a NIP A basis.

C. Long-Term External Prospects in a Multicountry Context

15. On the basis of the central fiscal policy rule and abstracting from demographic effects, the U.S. current account deficit would decline sharply over the medium term, and would be roughly in balance over the longer term (Chart 1). When the effects of worldwide demographic trends are also incorporated in the simulations, the current account moves into a small surplus in the long term, because the population in other industrial countries is expected to age faster than the U.S. population (see Chart 1 and Table 4).10

CHART 1
CHART 1

United States Current Account Balances

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 105; 10.5089/9781451839500.002.A006

Source: IMF staff estimates.

16. The budget surpluses that would arise under the central fiscal policy rule would boost national saving and improve the current account.11 The movement of the external balance to surplus would be facilitated by a depreciation of the real effective exchange rate in the medium term, reflecting the fiscal contraction and lower domestic interest rates. Over the long run, the improvement in the external position (and in the net international investment position as a share of GDP) would induce a long-run real appreciation of the dollar. The private savings-investment balance also would shift from a deficit to a surplus over the long run. While savings would rise slightly in relation to GDP, private investment—after declining through the first two decades of the next century—would be largely unchanged over the longer term, reflecting the effects of a declining labor force growth on the marginal productivity of capital and on the returns to capital investment (Chart 2).

CHART 2
CHART 2

United States Private Saving-Investment Balance

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 105; 10.5089/9781451839500.002.A006

Source: IMF staff estimates.

17. To illustrate the importance of the long-term fiscal policy rule that is pursued (and the rough range of results for different fiscal policy rules), an alternative scenario was developed assuming that the long-term fiscal objective would be to maintain balance in the unified U.S. federal budget (Chart 3).12 Under this scenario, the U.S. external position would improve gradually over the medium run and stabilize at modest deficit levels over the long run. Private saving would rise slightly in the medium term but would remain relatively stable in relation to GDP in the long term (Chart 4). Private investment would not differ significantly from the baseline scenario.

CHART 3
CHART 3

United States Current Account Balances under Alternative Fiscal Policy Rules

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 105; 10.5089/9781451839500.002.A006

Source: IMF staff estimates.
CHART 4
CHART 4

United States Private Saving-Investment Balance under Balanced Budget Rule

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 105; 10.5089/9781451839500.002.A006

Source: IMF staff estimates.

18. The greater fiscal effort required in the scenario incorporating the central fiscal policy rule results in a higher level of real national income over the scenario period than in the scenario based on a balanced budget fiscal rule (Chart 5). Over most of the period through 2070, the level of national income is about 2 percent higher, as fiscal surpluses reduce interest rates and crowd in private investment.

CHART 5
CHART 5

United States Ratio of Real Gross National Product under Central Fiscal Policy Rule Relative to Balanced Budget Rule

Citation: IMF Staff Country Reports 1998, 105; 10.5089/9781451839500.002.A006

Source: IMF staff estimates.

List of References

  • Cline, W., 1995, “Predicting External Imbalances for the United States and Japan,” Institute for International Economics.

  • Diamond, Peter, 1977, “A Framework for Social Security Analysis,” Journal of Public Economics, Vol.8, pp. 27598.

  • Feldstein, M., 1996, “The Missing Piece in Policy Analysis: Social Security Reform,” National Bureau of Economic Research, Working Paper No. 5413, January.

    • Search Google Scholar
    • Export Citation
  • Feldstein, M., and Pellechio, a., 1979, “Social Security and Household Wealth Accumulation: New Microeconomic Evidence,” Review of Economics and Statistics, Vol. 61, August.

    • Search Google Scholar
    • Export Citation
  • Frenkel, J. and Goldstein, M., 1991, “Essays in Honor of Jacques Polak,” IMF.

  • Hakkio, C., 1995, “The U.S. Current Account: The Other Deficit,” in Economic Review, Federal Reserve of Kansas City.

  • Hooper, P., Johnson, K., and Marquez, J. 1998, “Trade Elasticities for G-7 Countries,” International Finance Discussion Papers, No.609, April.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 1998, “Multimod Mark III: The Core Dynamic and Steady-State Model,” IMF Occasional Paper, No. 164.

  • Leimer, D. and Lesnoy, S., 1982, “Social Security and Private Saving: New Time-Series Evidence,” Journal of Political Economy, Vol. 90(3).

    • Search Google Scholar
    • Export Citation
  • OECD, 1996 “Future Global Capital Shortages: Real Threat or Pure Fiction?.”

  • Williamson, J. and Mahar, M., 1996, “Current Account Targets,” Institute for International Economics, unpublished manuscript.

1

Prepared by Martin Cerisola, Hamid Faruqee, and Yutong Li.

2

In the period after 2050, the current account deficit improves somewhat in relation to GDP. Import growth slows, as the growth of GNP declines because an increasing share of domestic output (GDP) in the United States is being transferred abroad to service external debt, leaving less income available for domestic consumption.

3

These elasticity estimates are based on annual data for the period 1968–97 in the, case of exports and 1961–97 in the case of imports. Hooper (1998) reports long-run income elasticities for U.S. exports and imports of 0.8 and 1.8, respectively.

4

Hooper (1998) reports long-run price elasticities for U.S. exports and imports of 1.5 and 0.3, respectively.

5

The dependency ratio is defined as the number of persons aged 65 and older or 19 and younger as a share of the working population aged 20 to 64 years. The World Bank projections are used because of their extensive country coverage.

6

The time profile for the long-run increase in the dependency ratio differs somewhat between the United States and other industrial countries. The dependency ratio declines in the United States until 2010, and then rises substantially above its 1998 level; for the other industrial countries, the average dependency ratio rises steadily from a level that is below that in the United States initially to one that exceeds the U.S. ratio in the long run.

7

In the panel estimates, changes in the total saving-investment balance in response to fiscal or demographic shocks are identically zero by construction (i.e., adding-up restriction). In the MULTIMOD simulations, this adding-up restriction for the United States and other industrial countries is not imposed as developing country groups are also included. Broad adding-up between the United States and other industrial countries may still obtain, however, given their relative economic size.

8

Monetary policy in the United States is assumed to aim at holding the rate of increase in the GDP price deflator to around 2 percent throughout the scenario period. Projections for the rest of the world in this scenario are based on the WEO projections to 2003. After that time, the scenario reflects an assumption that fiscal policy in the rest of the world seeks to stabilize government spending and debt in relation to GDP.

9

The impact on national savings, while not identical, would be similar if reductions in benefits (or a combination of benefits reductions and contribution increases) were used instead to ensure actuarial balance in Social Security and Medicare. Benefit reductions would stimulate private saving in a similar magnitude as increases in contribution rates. This is based on the substitution hypothesis advanced by Diamond (1977), in which rational decision makers would substitute expected future social security benefits for private wealth accumulation. Feldstein (1996) notes that several empirical studies have found evidence of the substitution hypothesis. However, Feldstein and Pellechio (1979) have also argued that a change in future social security benefits may have no impact on private saving behavior, if for example, households discount heavily future benefits. Leimer and Lesnoy (1982) also present empirical evidence that Social Security may have no impact on private saving.

10

In MULTIMOD, the long-run income elasticities for exports and imports are restricted to being 1. To test the effect of this restriction on the long-run current account projections, an alternative scenario was developed based on an income elasticity of 1 for exports and 2 for imports. The results show that the U.S. current account would still improve considerably over the next 25 years, achieving surpluses averaging 0.1 percent of GDP during the period 2006–24. As fiscal surpluses narrow, the U.S. current account would return gradually to deficits averaging about 0.9 percent of GDP in 2025–70. The real effective value of the U.S. dollar would depreciate by around 30 percent by 2025, before subsequently appreciating by about 5–10 percent. To an extent, this alternative scenario might be seen as a worse case. Over the 72-year horizon in the scenarios presented here, it would be expected that the income elasticities of exports and imports would converge.

11

The panel estimates in Table 2 show qualitatively similar effects of the fiscal budget on the saving-investment balance. Overall, the long-run fiscal effects on the current account are somewhat larger in the panel estimates than in the dynamic simulations.

12

In the balanced budget scenario, it is assumed that revenues would be adjusted to maintain budget balance. Accordingly, the ratio of revenues to GDP would initially decline from 31.8 percent in 1997 to an average of 30.3 percent in 2011–20. Subsequently, the revenue-to-GDP ratio would rise, reaching an average of almost 35 percent in 2061–70, about 7 percentage points of GDP higher than the ratio in the scenarios assuming the central fiscal policy rule. The rise in the revenue-to-GDP ratio in the balanced-budget scenario reflects the needed adjustments to revenues in the latter part of the simulation period to keep the budget in balance largely owing to rising expenditures for Social Security and Medicare.

United States: Selected Issues
Author: International Monetary Fund
  • View in gallery

    United States Current Account Balances

    (In percent of GDP)

  • View in gallery

    United States Private Saving-Investment Balance

    (In percent of GDP)

  • View in gallery

    United States Current Account Balances under Alternative Fiscal Policy Rules

    (In percent of GDP)

  • View in gallery

    United States Private Saving-Investment Balance under Balanced Budget Rule

    (In percent of GDP)

  • View in gallery

    United States Ratio of Real Gross National Product under Central Fiscal Policy Rule Relative to Balanced Budget Rule