人口予測を通して、将来のニーズを探る(英語版のみ)

  • Print
  • Connect
  • Email
  • Facebook
  • Twitter
  • LinkedIn
  • Google+
作成者 Bruce R. Mitton |  2017年6月15日

課題

ミリマンは、ネイティブ・アメリカンのコミュニティーからの依頼を受けました。同コミュニティーには、金銭的援助、埋葬給付金、ヘルスケア診療所、文化保存プログラム、教育支援など、当該市民の利益になるよう運営されている各種制度があります。これらは全て、財政支援を必要とします。コミュニティーでは、これらの制度に対する資金需要が今後どのように変化していくのかに関する情報を求めていました。コミュニティーのリーダー達は、最近の加入状況を見て、その毎年の伸び率がどうなるかという単純な推計をしましたが、その推計の検証とともに、今後20年間の伸展がどうなるかを示すより正確なモデルを求めていました。

The solution

To gain tribal membership, one must be a direct descendant of a tribal member, and while an applicant has until age 21 to sign up, the majority tend to apply early because of the benefits available. As a result, future changes to the community’s enrollment were assumed to come from births to current citizens and from deaths. As a result, Milliman gathered data on their citizens including gender, dates of birth, the dates of birth for their citizen parents, and dates of death.

Because the current community citizens are a relatively young group with over 50% of the citizens under the age of 25 and only 5% over age 65, it became clear that birth rates would likely have a much more significant impact on the future size of the community than would death rates. In fact, over the projection period, the number of births outweighed the number of deaths by a ratio of more than six to one. Normally, when looking at fertility, the focus is on the mothers. However, for this particular community, eligibility as a citizen does not require any particular blood quantum or percentage of Native American blood. As long as a person is a descendant of a registered ancestor, that person was eligible for membership. Therefore, the offspring of fathers (who may or may not have married a female citizen) are also eligible to become citizens. As a result, we looked at the fertility rates separately for male and female citizens and studied the likelihood for each gender that the other parent was also a tribal citizen. The raw rates were adjusted by gender for the percentage who married another citizen so that births would not be double counted.

Milliman’s analysis showed that female citizen fertility rates in the community were significantly higher than for the general U.S. population. We assumed that these would decline over time toward societal norms. And while there are not readily available fertility rates for males in the general U.S. population, we assumed that male fertility would also decline over time in the same proportion as was assumed for females. The age specific fertility rates developed from the data were then smoothed for use in projecting the population.

Due to the general age of the citizenry and the overall size of the group, there was insufficient data to determine death rates for this specific population. However, we had enough data to benchmark their death rates against a standard table to which we used an age adjustment to predict future deaths.

The outcome

Milliman prepared a report for the community leaders that showed that the size of the citizenry was likely to increase 50% over the next 20 years. While this is a significant increase, the previous estimate led them to believe that they would reach this size in a period of only 10 years instead of 20. We were able to provide a more realistic prediction of how their group would change over time. The citizens’ average age was expected to increase from about 26.8 years to roughly 29.6 years over the same 20-year period. Finally, we were able to provide them with count estimates in five-year age groups at five, 10, 15, and 20 years in the future, as well as the expected number of deaths. They were very happy with the refined estimate because it will enable them to match the various programs with the appropriate age groups and, using the future counts, develop budgets and make more definite plans for the future.