By Peter Murray, Content Markerting Manager, Carto
In 1970, the number of Malaysians aged 65 and over made up 3.3% of the population. According to the Department of Statistics Malaysia, this figure has risen steadily over the years along with the country’s population size. Last year, the figure stood at 6.2%. In fact, it is projected that this figure will go up to 14.5% by the year 2040, and this is equivalent to about 6 million people!
Countries worldwide are experiencing similar predicaments.
In the US, the unprecedented growth of people aged 65 and over is driving demand in the senior housing sub-sector. The Census Bureau expects the US senior population to increase by 60% between 2016 and 2040. Preparing for this prodigious population growth presents challenges for deciding when, where, and what type of medical facilities to open.
In the past, site analysts have relied on census data for finding areas where the senior population had increased significantly, and then selected sites in those areas. Correlating market growth to population growth can work in certain industries. However, there are more factors involved in healthcare site planning.
Today, for instance, more seniors are moving to cities and metropolitan areas, so they can retire closer to their adult children. These recent behavioural trends are hard to notice with traditional site planning workflows, like the one above, because sources like the US Census update demographic data only periodically.
Location Intelligence provides a modern solution for site planning. It applies different types of spatial analysis for identifying origin and destination patterns among retirees. This then helps determine what constraints exist in these areas, and then decides the optimal site for new facilities.
Site Planning with Location Intelligence in Washington, D.C.
If seniors are more likely to retire near their adult children, then identifying areas with a high population density of “adult children” can help determine growth market corridors where demand for senior care services is likely to grow.
Erickson Living has had success with this strategy, and as senior vice president Adam E. Kane recently explained, even if “there’s not really a plethora of ageing demographics in a local area, its a growing market where you have a lot of adult children moving to and living there.”
One growth corridor market is Washington, D.C., which has experienced a drastic population increase. But, as D.C. Policy Center reports, from 2010 to 2016, population growth outpaced the growth of new housing with the number of residents increasing 13.2% while housing stock increased only 5%. This situation has led to increases in property values as the map below illustrates:
In the map below, the 19 current nursing homes within the District of Columbia are represented with proportional circles whose size corresponds to the residential capacity for each site. In addition, each census tract is styled based on the total population count.
As expected, most nursing homes are located in densely populated areas, but to identify growth market corridors, let’s segment total population by age, and style each tract by total population to men and women aged 35 to 49.
Taking the higher end of the distribution of adults ages 30-49 gives a better sense of possible growth market corridors for new nursing homes. Next, set some criteria for each site to help locate optimal sites among these areas.
Ideally, potential sites should be:
- easily accessible by public transportation,
- have low crime rates, and
- have high population density of adult children at lowest possible cost.
Now let’s add enriched data from our Data Observatory and run some basic types of spatial analysis to refine site selection based on established criteria:
- Import transportation data to find tracts with accessible public transit based on bus ridership rates
- Run an intersect and aggregate analysis to add data on crime incidents
- Run another intersect and aggregate analysis to add average real estate listing price for each tract
These steps added a lot of criteria-related data to the map below, but there’s no meaningful information just yet for determining new nursing home locations.
Making sense of this data requires running a few more types of analysis for finding areas where our desired criteria overlapped.
- First, conduct the ‘create centroid of geometries’ analysis to find the centre point of areas where our criteria would overlap.
- Next, create clusters so that points are grouped around the centroids created in the last step, and set the cluster count to three.
- Finally, run the ‘create centroid of geometry’ analysis again and categorise by the number of clusters column created in the last step.
These steps produce three sets of coordinates that identify which census tracts, and where in those census tracts are the most suitable area for new nursing homes. In this case, our three neighborhoods of interest are Mount Pleasant, Arboretum, and Congress Heights.
In the map below, these coordinates are displayed and the map has been styled by average real estate listing price.
The most expensive neighborhood by far is Mount Pleasant with listing prices between $1 and $1.1 million dollars. In the other two neighborhoods, however, prices were substantially less ranging between $370,000-$450,000 in Arboretum and $370,000-$400,000 in Congress Hill.
Finally, let’s check to see which, if any, of these sites are located in growth market corridors by styling this map by population of adult children in the map below.
All three neighborhoods are on the higher end of target segment distribution, with the highest concentration of adult children located in Mount Pleasant and an equal share in Arboretum and Congress Hill.
The final decisions on where to open a nursing home will depend on budgetary constraints and available real estate, but Location Intelligence has certainly provided a modern approach that eliminates guesswork from site planning.