Migration and Geographic Distribution

Four factors determine the proportion of elderly people in a population: past death rates, past birth rates, past in-migration rates, and past out-migration rates. Of these four factors, the least important is death rates. Although there are some variations in death rates across states, these variations are not very large and they play a very insignificant role in determining age distributions. Fertility patterns have a larger effect on the age composition of a population—high fertility rates are associated with younger age distributions. The intuitive notion that higher fertility leads to a larger proportion of children in the population, and consequently a smaller proportion of older people, is correct. This relationship between birth rates and age structure helps to explain why Utah, with the highest fertility rate of any state, ranks forty-ninth in proportion of older adults. Seven of the ten highest fertility states are among the fifteen youngest states. But there are exceptions. Arizona, for example, has the second highest fertility rate of any state, but also has an above average proportion of older age. This suggests that effects of migration on age distribution must also be considered.

In-migration and out-migration do not necessarily alter the age distribution of a county or state. If migration rates did not vary by age, then migration would affect the size of the population but not the age composition. However, migration rates almost always do vary with age. Outmigration from an area is predictably higher for young people than for older people. The tendency of young adults to migrate at higher rates than older people has been true across time and across cultures. In other words, there is almost no exception to the pattern of a declining propensity to migrate after the young adult years. Recent patterns of migration in the United States are consistent with this broad generalization. Between March of 1999 and March of 2000 it was eight times as likely for someone between twenty and twenty-four years of age to change place of residence as it was for someone age sixty-five or over (35.2 percent vs. 4.4 percent).

States (or counties) in which out-migration exceeds in-migration over a sustained period of time tend to have relatively old populations. Older people in these areas tend to age in place, while a disproportionate number of younger people move out. In recent decades, the areas that have experienced substantial net outmigration have been concentrated in the farm states of the Midwest and the declining industrial states of the Northeast. These two regions, as noted above, are the regions with the highest concentration of older people.

If a disproportionate number of migrants are young, one should anticipate that areas with substantial net in-migration would tend to have relatively young populations. This is the general pattern, and accounts for the below average proportion of older adults in the Sunbelt (which has gained population through migration in recent decades). But there is one major exception to this pattern: Florida. Florida has been such a magnet for older retirees who move out of the Northeast, that the in-migration rate for older people has exceeded that of younger people. Arizona and Oregon also have been magnets for older migrants, so despite high levels of total in-migration their populations are not especially young (Arizona ranked 22nd in percentage of population 65+ in 2000, and Oregon ranked 25th).