military nor the Federal civil service will necessarily relate to the effects of reduced fertility.) Thus, these pension cost increases will constitute a claim on the same resources many look to for relief from rising social security taxes.

State and local pension systems likewise will be affected by improvements in mortality and will be somewhat affected by reduced fertility as well, since State/local sector employment has borne a closer relationship to population trends, and some of the primary services provided by this work force are delivered to children and their families.

For fully funded State/local plans, the impact of aging is analogous to that for private plans. That is, assets will build up rapidly in anticipation of higher costs (to $0.5 trillion by 2000 for the sector as a whole as projected by Munnell and Connolly), with possibly added unplanned liabilities to adjust benefits for inflation. According to Tilove (1976), over half of State/local employees are covered by automatic benefit adjustments, but few are assured complete protection against price increases.

In the case of State/local plans, such unplanned costs will have to be made up by the governments involved, either through tax increases, reduced wage increases, or displacement of other claims on State/local treasuries. And, of course, the underfunded systems may have to look to greater employee contributions and/or tax hikes to meet rising benefit costs, much as social security tax increases have been found unavoidable at the Federal level. But during the 1970's, a number of State and local governments have moved toward greater prefunding, in some cases with less promise of postretirement benefit adjustments or adoption of measures for prefunding them.

D. FINANCIAL MARKETS

As already pointed out above, the aging of the population has prompted a rapid buildup in the assets of private and State/local pension plans. By the end of 1975, these plans' assets plus the social security reserve amounted to one-fifth of the total equity and debt of the U.S. nonfinancial corporate business sector. Private and State/ local pension assets increased by a net amount of $24 billion in 1975, which was more than a quarter of total personal savings and 65 percent of the nonfinancial business sector's requirements for external funds (Friedman, 1976). It has been predicted that pension plans will control a majority of all U.S. corporate equities by 1985 (Drucker, 1976).

The most significant issue posed by this dominance in the capital markets is the heightened potential for institutional control over the allocation of capital. For example, it may be attractive for State governments to influence the shape of capital formation through encouragement of mortgage investments or a capital market break for in-State businesses. As Stone (1976) indicates, there is the following economic argument that can be made for government intervention in the investment of pension assets.

Certain economists have contended that there is an inherent market bias which causes funds to be overcommitted to private purposes and undercommitted to public purposes. It is certainly arguable that our society is not allocating a proper share of its investment capital to such public goods as education, scientific research, transportation, or

housing. If the returns to those investments are difficult for an investor to measure or capture, there is likely to be a distortion in our pattern of capital formation.

A second bias results from the fact that large investment institutions are prone to confine their asset purchases to the largest issues of the largest issuers. A large private or public investor may well think it efficient to study only a small number of situations. Only the most substantial investment opportunities attract large investors' attention. Moreover, if they wish to remain fairly liquid, their opportunities are narrowed still further to large investment opportunities which are small fractions of even larger investment opportunities.

This mostly theoretical debate over the social efficiency of private investment will become a heated practical debate as the accumulation of government pension assets grows. Until now, government involvement in capital formation could only have come through mandatory controls, a notion extremely difficult politically to bring about. But the political balance could change as funds for government pension liabilities grow. When governments hold the funds, governments must make the investment decisions themselves.

Pension plan assets have been concentrated in corporate securities. Greenough and King (1976) reported that, in 1973, 79 percent of these funds were invested in corporate stocks and bonds. Within the corporate sector, investments have concentrated on the offerings of major corporations (over 90 percent of stocks held are listed on the New York Stock Exchange). Portfolio turnover has been at a fairly slow rate. Thus, it is feared that pension funds have underutilized opportunities for long-range investments in smaller companies or in new enterprises and have helped create a shortage of risk capital. Continued growth in pension fund assets, coupled with a trend toward homogeneity in portfolios, suggests this pattern of influence on the capital markets may become more pronounced.

In the long run, another concern may arise with respect to the influence of pension funds on the capital markets. That is, as more and more private and State/local plans become fully funded, and as they accumulate funds in anticipation of the retirement of the "baby boom" generation, a rapid líquidation of funds will occur in the next century. As early as 1996 to 2001 for State/local plans, and 2001 to 2006 for private plans, substantial numbers of people born during the 1941-60 period of high fertility will claim pensions. Sometime during the first half of the 21st century, there may well be a net withdrawal of pension funds from the capital markets as liquidations exceed accruals for future liabilities (Hsiao, 1976). Such a withdrawal could result in downward pressures on stock prices and rising interest rates. Unless long-term investment planning for the overall economy properly anticipates this period of pension asset liquidation and appropriate investment strategies are followed, an acute capital shortage could occur during that period.

V. THE IMPLICATIONS OF MIGRATION AND REGIONAL DEVELOPMENT PATTERNS

Movements of elderly individuals among States and regions follow different patterns than those for migrating persons of working age and their families. Should such differences continue, the various economic effects of population aging will have differential impacts on

regions and States, particularly with respect to labor force size and structure, health and welfare costs, and demand for goods and services. This section describes selected studies on the migration of the elderly. In the decade between 1960 and 1970, the elderly population in the United States increased by 21 percent. However, growth rates varied dramatically across regions and States. The regional growth rates ranged from 12.3 percent for the north-central region to 31.2 percent in the South. Over the period 1970-77, the South absorbed another 1.1 million elderly individuals, an increase of an additional 18.8 percent. Within regions, the South Atlantic was the largest gainer with a 39.2 percent increase over the 1960-70 period and a 23 percent increase over the 1970-75 period.34

Even within census divisions, the variation in growth rates is dramatic. Within the South Atlantic census division, the 1970-75 increase in the elderly population ranged from 1.6 percent in the District of Columbia to 36.8 percent in Florida. Similar patterns emerge for the 1960-70 time period. The elderly population in the District of Columbia grew by only 1.9 percent while the increase in Florida amounted to 78.2 percent. In 1960, 553,000 elderly persons resided in Florida; by 1970 their presence had grown to 985,000, and by 1975 over 1.3 million elderly persons resided in that State.

However, rates of increase can be deceiving in that States with small increases in the number of elderly residents may actually rank among the highest in terms of absolute numbers of elderly individuals. Illinois experienced only a 5.9 percent increase in its elderly population during the 1970-75 period, yet by 1975 over 1.1 million older individuals lived in that State. In fact, 45 percent of the total U.S. elderly population reside in the seven States of New York, California, Pennsylvania, Florida, Illinois, Texas, and Ohio, with each containing over a million elderly. By 1977, 17.3 percent of Florida's total population was aged 65 and over.35 However, of the seven States with the largest absolute numbers of elderly, four-New York, Pennsylvania, Ohio, and Illinois-experienced net outmigration of elderly persons during the 1960-70 period. In some States, such as Nebraska and South Dakota, the proportion of elderly in the total population grew due to outmigration of young persons.

Net migration rates for the 1960-70 time period show a movement of elderly persons out of the Mid-Atlantic and East North Central States and into the South Atlantic, West South Central, and Pacific States. If such movements continue, the increased proportion of elderly individuals in the total population could exert a significant impact on regional development. An examination of the determinants of elderly migration may help to establish the likelihood of a continuation of past trends.

The literature dealing with migration patterns is enormous in size; however, the subset which deals with migration behavior of the elderly is quite limited. Barsby and Cox (1975) have conducted the most comprehensive review of the migration patterns of the elderly. The factors which they identify as determinants of elderly migration behavior include: (1) Family and social attachment; (2) health and climate; (3) homeownership; (4) location at time of retirement; and (5) educational level.

24 The data cited in this section are taken from Current Population Reports, No. 59, p. 23, May 1976. 35 University of Florida, College of Business Administration, Florida Statistical Abstract 1978, The Uni

Family and social attachments are negatively correlated with migration when the family and friends of the elderly person are located within the community, and positively related when they are located outside the community. Harold Geist (1968) found that about 80 percent of sampled retirees who chose not to move after retirement did so as a result of social ties. Langford (1962) found that social contacts were a more significant factor in determining migration behavior than were the availability of community or health facilities or services.

Serow (1976) found that while older persons are less mobile than the population observed as a whole, once the elderly decide to move, they are more likely to return to their State of birth. The tendency to return to one's State of birth may reflect the influence of social ties-i.e., familiarity with the environment, community structure and past social relationships. Serow states that, while the migration of the elderly will most likely continue to be concentrated in areas of favorable climate (the "frostbelt" to "sunbelt" movement), increasing numbers of elderly persons will be returning to their States of birth. Serow states that planning for such a trend is crucial if the needs of the elderly are to be met in the future.

Health and climate are two variables which may work in conjunction or in opposition. While poor health may inhibit mobility, it may actually increase the probability of movement to an area of improved climate. Mobility may also be affected by climate independent of health, reflecting personal preferences.

Homeownership may act as a deterrent or a spur to migration. Since a home is a relatively nonliquid asset, homeownership may reduce mobility. A study of Prasad and Johnson (1964) found this deterrent effect. However, Lansing and Mueller (1967) found that the sale of a home provides a stimulus for out-migration.

Barsby and Cox state that a tentative relationship between residency at time of retirement and migration does exist. However, the studies completed to date have not conclusively defined this relationship. It does tentatively appear that residents of suburbs or SMSA's are more likely to migrate than residents of other locations. Job opportunity is another factor for which the relationship with migration behavior is tentative at best. It would appear reasonable, however, that the inducement to migrate because of job opportunities declines with age. Postretirement income appears to be positively related with the propensity to migrate. This is not surprising since migration involves economic costs.

Given the extended durations of unemployment experienced by older persons, it is not surprising to find that elderly persons in the labor force are less mobile than labor force nonparticipants. Bowen and Finnegan (1969) reported a negative relationship between labor force participation and net migration rates of older persons. Other studies of Miller (1965) and Goldstein (1968) confirm these findings. The study by Lansing and Mueller and another study by Bultena and Wood (1969) reveal that higher education and skill levels are related to higher migration rates in the elderly population. Elderly migrants were more likely to have held jobs in the professional or managerial occupations prior to retirement than were elderly nonmigrants.

Thus, based on the existing literature, Barsby and Cox (1975) conclude that "high mobility among older persons is concentrated

among those who have moderate income levels, live in suburban areas, are well-educated, and held managerial or professional positions prior to retirement." However, the authors caution that much of the evidence supporting these conclusions is of questionable validity.

In terms of patterns of elderly migration, several observations are made of Barsby and Cox. First, older persons are less mobile than younger persons. Gross migration rates of young persons were over twice as high as the rates for older persons during the 1955-60 and 1965-70 time periods. Second, the elderly tend to move from larger States to smaller States. Third, elderly persons show a greater tendency to move into a few specific States than do the nonelderly. During the 1965-70 period, the net inmigration rates of those aged 65 and over exceeded that of the under-65 population in eight States: Arizona, Florida, Nevada, North Carolina, Oklahoma, Oregon, South Carolina, and Tennessee.

While the determinants of elderly migration may remain only tentatively identified, the observed trends indicate that the migration patterns of the elderly do differ from those of the nonelderly population. The economic consequences of such differences in migratory behavior will certainly increase as the population ages, both at the regional level and within specific States. Some economic consequences of population aging observed at the national level will be felt more keenly at the regional and State level. Migratory patterns of the population, coupled with prolonged low birth rates, may well serve to magnify at the subnational level the economic impacts of population aging observed in the United States as a whole.

REFERENCES

Abbott, Julian, "Socioeconomic Characteristics of the Elderly: Some Black-White Differences," Social Security Bulletin, vol. 40, No. 7, Washington, D.C., U.S. GPO, July 1977.

Anderson, Kathryn; Clark, Robert; and Johnson, Thomas; "Retirement in Dual Career Families," unpublished paper. Department of Economics and Business, North Carolina State University, October 18, 1978.

Barfield, Richard E., "The Automobile Worker and Retirement: A Second Look." Survey Research Center, Institute for Social Research, University of Michigan, 1970.

Barfield, Richard E.; and Morgan, James N.; "Early Retirement," Survey Research Center, Institute for Social Research, University of Michigan, 1970.

Barro, Robert J., "Social Security and Private Saving: Evidence from the U.S. Time Series," American Enterprise Institute, 1978. Barsby, Steve L.; and Cox, Dennis R.; "Interstate Migration of the Elderly: An Economic Analysis." Heath, Lexington Books. Massachusetts, 1975.

Becker, Gary S., "Human Capital: A Theoretical and Empirical Analysis," National Bureau of Economic Research, Columbia University Press, New York 1964.

Bixby, Lenore, "Retirement Patterns: Research and Policy Interaction," Social Security Bulletin, vol. 39, No. 8. Washington, D.C.,