A town, a county or a village are all “places” in the eyes of the U.S. Census Bureau, but a park is not. The Adirondack Park is spread over 12 counties of northern New York (Figure 5), accounting for more than one-half of the total area of the region and only one-eighth of the population. A majority of the Park boundary (the “Blue Line”) does not coincide with town boundaries or census tracts. This has resulted in a series of imprecise census “estimates” over the years and little in the way of accurate demographic study.
Estimates and Projections
Demographers speak in terms of estimates and projections, reflecting the use of mathematics and statistical methods to further our knowledge of a particular data set. A projection can give us a glance into the future. An estimate provides a peek into the past or a more thorough understanding of the present. After the 2000 census, the Adirondack Park Agency (APA) produced the first credible estimate for the population inside the Park. A known ratio, in this case the occupancy rate of all structures, was used to estimate the population in a census tract divided by the Blue Line. Count the buildings and you can estimate the population on one side or the other. Add the partial counts to the counts of whole census tracts in the Park and you get a total population. By 2010, an updated APA estimate demonstrated, for the first time, a decline in Park population.
The Program for Applied Demographics at Cornell University* has also had an interest in Park demographics. At the request of APRA 2014, Cornell developed a data set for the Adirondack Park in September of 2013. Their analysis employs the Hamilton-Perry method, a standard statistical technique, to project median age, population and age distribution for each census from 2000 to 2030. Over time, a commonality of results has emerged from independent studies by APA, APRA 2014 and Cornell. Similar results from differing methods offer credibility to each. The Park Agency’s initial estimate of total Park population has evolved into a variety of demographics for the areas within and beyond the Blue Line in these 12 upstate counties.
The Cornell population projections will be used in this report for the sake of consistency. APRA 2014 is solely responsible for the interpretations and charting of the Cornell data used in this report. The Program for Applied Demographics simply provided the raw data.
The Adirondack Park population declined by more than 2,100 people from 2000 to 2010. The rate of population decline is projected to increase in the current decade and increase still more from 2020-30. By the end of 2014, the estimated population of the Park will be 128,000 residents. Figure 6 shows the steeper population downturns resulting from lower birth rates, continued out-migrations of young adults and their children, as well as the ongoing demise of the baby-boom generation. The current population density of the Park is 14 people per square mile. By comparison, the population density of New York City is approximately 27,000 people per square mile.
APRA’s projection, based on Cornell data, shows a rate of population decline averaging 525 people per year throughout the current decade. The number of young adults and children (<30 Age Group) is projected to decline at more than 680 individuals per year during this period. The difference between these two numbers is the reliable in-migration of older people that will continue at an average pace of more than 155 people annually.
Between 2020 and 2030, the Park is projected to lose approximately 940 people each year. By 2030, the Park population is projected to drop below 116,000, with a median age greater than 50 years. The last time the Adirondack Park had a population at that level was 45 years ago, however, the median age at that time was 20 years younger.
The projections provided in a Hamilton-Perry analysis are limited to 15 years going forward, so any projection beyond 2030 would be in the speculative realm. Keep in mind that more than one-third of the Park’s residents will be in the aging boomers (60+) category in 2030. The mortality of this age group is likely to have a significant effect on total Park population for two more decades or until the middle of the century.
Cornell’s Projection of Park Population
While there are links between the numbers shown at the right, it is noteworthy that each is about 17,000. The over-all population decline may seem alarming, but it is clearly subordinate to the losses in the age group that includes young adults and children. The changes in this age group will have long-term consequences for the Park.
In the projection shown in Figure 7, the decline of the 0-29 Age Group (blue columns) averages 14% each decade. The K-12 Public School Enrollment studies undertaken by APRA 2014 (see next chapter) are supportive of Cornell’s projections up to the present. Over the past ten years, the number of these students in the Park has gone down by more than 21%. Cornell projects an average loss of 604 residents per year from 2000-2020 in the 0-29 Age Group, while APRA 2014 projects an average loss of 422 students per year in grades K-12 from 2004-2013. That constitutes a commonality of results.
Population and median age of a community are affected by changes in birth rate, migrations and mortality. In equilibrium, the median age of a population should remain relatively stable. The baby-boomers changed the nation’s population significantly after World War II and they lowered the median age from the late 1940s through the late 1960s. At that time, the median age began to rise, as shown in Figure 8. In 1990, a divergence occurs between the median age of the Park and that of the rest of the nation. The slopes of the state and national lines represent the aging baby-boomers. So what explains the steeper slope in Park median age? Why were Adirondackers aging at more than twice the pace of the state or the nation? Why only since 1990?
In 2014, the people of the Park had a median age nearly nine years older than the rest of America, eight years older than the rest of New York State and five years older than their neighbors outside the Blue Line in the same 12-county region. As stated earlier, in equilibrium, the median age of a population should remain relatively stable over time. Estimates made from 2010 census data confirmed that the median age of the residents of the Park increased 10 years during the prior two decades.
Looking forward, Cornell’s Program for Applied Demographics has also projected median ages using the Hamilton-Perry analysis. In the current decade, the Park median age is expected to rise four years. From 2020 to 2030, the median age increase should slow to just two years, resulting in a median age of nearly 51 years old. By then, the age distribution and median age for the Adirondack Park should resemble Hamilton County as it exists today.
Approximately one in every 10 counties in America is currently experiencing this kind of aging.** Most of these counties are small and scattered about the rural Midwest. Nineteen of these counties can be found in the northeastern United States, including Warren, Essex and Hamilton in the Adirondacks.
An age distribution chart can show the composition of the Park population over time. Figure 9 uses the same data set as the Declining Population of the Adirondack Park By Age Group chart in Figure 7, but the groupings are reduced to five years. Each of the lines projects a different decade and the crests in each of those lines represent the aging of the baby-boom generation. The ever-rising nature of these crests demonstrates the in-migration of more baby-boomers until the peak in 2020.
The movement of these crests in Figure 9 could be interpreted as a wave about to curl. In 2030, that wave represents more than a third of the population of the Park. The demise of the baby-boomers will clearly affect the total population over the next 30 years, as will the decline in the number of children and young adults.
The Cornell projections demonstrate significant declines in the number of young adults residing in the Park after high school graduation. The decline is quite steep for the 2000 Age Group Distribution in Figure 9 and it flattens somewhat in each successive decade. The Adirondack Park appears to be transitioning to a lower level of population and these young adults are leading the way. The loss of young adults leads to a loss of children, which can be seen in the decennial projections of Figure 9. Beyond 2030, there will be a rapid loss of baby-boomers. Eventually these curves will stabilize or “flatten out,” as they might in equilibrium, but the total population may be 15-30% less.
It is important to note that the projections in Figure 9 include a substantial prison population. Inmates are counted where they are housed during a U.S. Census survey. The Park has a very large prison population relative to its over-all population. The 2010 age distribution line (dark blue) from Figure 9 is duplicated in Figure 10 as a dashed line from age 15 to age 60. Once the prison population for the Park is removed from the total Park population, the resulting age distribution line for the Park is substantially lower, particularly between the ages of 20 and 40. This would be true for each of the decennial projections in Figure 9. The revised line is then used in the Age Distribution Less Prison Population chart in Figure 11 to accurately compare the Park with the 12-county region and the rest of the State.
The Park had five prisons in 2010 (Lyon Mt. is now closed). There were 4,104 inmates “residing” in North Elba, Moriah and Dannemora, which represented more than 3% of the total Park population. In 2010, the portions of the 12 counties outside the Park had an inmate/total population ratio of 1.65% (14,735 / 890,098). The New York State inmate/total population ratio was 0.30%. The concentration of inmates in the Adirondack Park was 10 times greater than the statewide ratio.
Since the 2010 census, the inmate count for the Park has increased and the Park population has decreased. By the Fall of 2013, the ratio of inmates at State and Federal facilities increased to nearly 3.5% of the total Park population. In this same period, the inmate/total population ratio decreased for the State and for the portions of the 12 counties that are outside the Park. So the spreads shown in Figures 10 and 11 are now noticeably greater. In short, the inmate count increases the 20-40 age groups by about 15%. The median age for inmates is currently a dozen years younger than that of other Park residents.
In the age distribution chart in Figure 11, APRA 2014 removed the inmate count from each of the respective populations shown. For many years the 12-county region has been substituted for the Park in comparative analyses. That assumption may no longer be valid. The age group distribution for the “North Country” of New York is quite different from the State as a whole. However, the age composition for the residents of the Park is now significantly different from that of their neighbors across the Blue Line. The gap is projected to widen for several decades.
The entire 12-county region has more elderly residents and fewer adults from the ages of twenty to forty-five years, when compared to the State as a whole. The differences are more profound when comparisons are made to “In Park” population alone. The “In Park” projections show far fewer children and an even greater loss of younger adults. The number of adults beyond 50 years of age is significantly greater.
The demographic characteristics of the people of the 12 counties that are residents of the Park are now known to be significantly different, than their neighbors living outside the Blue Line.
* The data for the Hamilton-Perry analysis of the Park was developed by Jan Vink, a Research Support Specialist, Program for Applied Demographics, Cornell University. The data set can be found here. The general website for Program for Applied Demographics is here.
** Woods & Poole Economics, Inc. - 2013 Projections and Database.