These pages look best viewed in Netscape!
© Copyright 1998 by the Wyoming Department of Employment, Research & Planning

The Division of Economic Analysis recently released a new report which presents estimates and projections of the resident population of Wyoming and its counties by age and gender from 1990 to 2006. It also includes estimates and projections of city and town population totals for the same period. This article will discuss how these estimates and projections were made; data excerpted is shown in the Table and the Figure.

Since actual future population trends are unknown, the projections in this report should not be considered a prediction of the future. These projections, mathematically calculated under a specific set of assumptions, indicate the population that would result given our assumptions if each population component persisted throughout the projection period. In other words, the projection procedures are only as accurate as the assumptions on which they are based. Therefore, they should be used as only one tool in the process of planning and decision making.

Methodology

The most commonly used projection technique, a cohort-component (sometimes called cohort-survival) model was applied to develop age and gender details. It involves the direct simulation of the demographic processes of fertility, mortality, and migration that produce changes in population size. This method employs the following basic demographic equation:

P₁ = Population at the end of the period,
P₀ = Population at the beginning of the period,
B = Births during the period,
D = Deaths during the period and
M = Net migration during the period.

For every projection period, the base population--disaggregated by single year age by gender, is survived to the next year period by applying the appropriate survival rates for each age and gender group. Next, net migrants by age and gender are added to the survived population, as is the population under age one. The populations under one year of age were created by applying age specific birth rates to the females of childbearing age. The entire process is then repeated for each year of the projections.

State Total Population Projections

The state total was produced from the economic forecast model which was developed by the Division of Economic Analysis and aided by the contractual services of the Wharton Econometric Forecast Associates (WEFA) Group. This econometric model uses historical data to identify relationships between economic variables. The state population tends to mirror employment in the economy, but with a slight lag. Therefore, the projections of state population follow the trends of employment forecasts to a great extent.

County Total Population Projections

A combination of extrapolative techniques and symptomatic techniques are used to produce county level projections. Trends of population variables were derived from data for the recent years since 1990, and used in a modified form to extend a population value to the future date. The arithmetic growth rates of population estimates, K-12 enrollments, employment estimates and sales tax collections in the Retail Trade sector were applied for the patterns of population change. The projected county totals obtained from the model were adjusted to achieve the consistency with state projections.

City and Town Level Projections

To determine population projections for incorporated places, a "share of growth" was utilized. It simply applies the average place/county ratios of the 1990 Census and 1996 estimates to the appropriate county projections to produce projections for municipalities.

Baseline Population

The April 1, 1990 county MARS (modified census counts by age, gender and race/Hispanic origin) are used as the starting point. A review of 1990 data by the Census Bureau revealed that people tended to report their age as of the date when they completed the census questionnaire, not as of April 1, 1990. Some of them may have also rounded their age up if they were close to their next birthday. The problem is more pronounced at age zero because there may have been more rounding from age zero to age one. Therefore, the Census Bureau modified the original 1990 data and made them more useful, particularly for the purpose of age estimates and projections.

Special Populations

Special populations (also called group quarters) display very different demographic patterns and characteristics than the population as a whole. In counties where special populations represented a significant proportion, an adjustment was made. The primary sources of special populations are prisons, colleges and military base populations. Because of the difficulty in predicting growth rates in special populations, one common procedure assumed that a fixed number of persons with a fixed set of age/gender structures will hold through the next 10 years.

Birth Projections

The number of births projected to occur in each county was derived from age-specific fertility rates for females aged 14 to 48. These rates were calculated by dividing the average of the number of births in each age cohort of mother for the years 1989, 1990 and 1991 by the 1990 appropriate female population. The three year average was used to control for year to year variation.

In the model, these rates are applied to the projected number of females of child bearing age at the beginning of the period to obtain the number of births in a future period. The gender of the births was determined by applying the usual assumption of 105 male births for every 100 female births. Again, these rates were county specific, and it was assumed to be constant throughout the projection period.

Death Projections

Because of the small death data cells for many counties, it was decided to use statewide death rates to develop the standard life table. Again, the three-year average of age- and gender-specific mortality controlled for year to year fluctuations in the number of deaths. The evaluation of the survival rates calculated from the life table revealed the similarity with the Census Bureau's 1990 life tables for Wyoming. Therefore, the statewide age- and gender-specific survival rates adopted in the model were directly derived from the Census Bureau's life table by race for 1990. For the reason of their extreme stability over time, the 1990 survival rates are assumed to remain constant during the projection period.

Migration Projections

Net migration procedures involve determining migration using residual methods which are equivalent to solving the population equation for the migration component. Thus, after the amount of total population change attributable to natural increases (births - deaths) is accounted for, the remaining difference between the total change and that due to births and deaths--the residual--is assumed to be the number of net migration.

In the model, the difference between the total population projected for a future period and the expected population obtained by surviving from the earlier period is the estimate of residual net migration. The migration pattern (distribution rates by age and gender) is statewide specific, and was derived from the Census Bureau's "1985-90 county migration file" and the "1993-94 geographical mobility file among different states." Because these patterns are quite stable over time, they were fixed to the same level over the projection period.

More Information

This report may also be obtained electronically via the Division of Economic Analysis' home page on the Internet: http://eadiv.state.wy.us/ For more information regarding Wyoming Population Projections, please contact Wenlin Liu at: Division of Economic Analysis, 327 East Emerson Building, Cheyenne, WY 82002. Telephone: (307) 777-7504 Fax: (307) 777-5852 Email: ead@missc.state.wy.us

Wenlin Liu is a Senior Economist with the Wyoming Department of Administration and Information, Division of Economic Analysis.

Table of Contents | Labor Market Information | Employment Resources | Send Us Mail |