|
4. Population at Risk
4.1 Building the Human Exposure Data
Model
Life, health, and workers compensation insurers are
subject to claims from millions of insured people.These
people move around, changing the buildings they occupy
hour to hour, traveling to and from work, visiting shops
and cinemas, going on vacation, business trips, and staying
at home. Individual coverage depends on a person’s
location, type of injury, the particular event that occurs,
and the carrier that supplies the coverage.
In order to analyze the potential losses from catastrophe
events on this exposure, a dynamic human exposure
database was developed. This database provides an estimate
for populated locations, demographics, employment
characteristics, and insurance coverages for every
hour of the day.
The database does not model the location and activity
of every person, but tracks the activities and locations
of groups of people with similar characteristics. Time
patterns and movement characteristics are captured for
each of these statistical groups.
Figure 4.1 Location of Chicago population by day and night
4.1.1 Data Sources
The human exposure database was built from the compilation
of a large number of primary sources, including
population census data collected by the U.S. Census
Bureau, employment statistics, economic data produced
by the Bureau of Labor Statistics, and various locational
surveys including the National Household Travel Survey
and journey-to-work trends. Insurance penetration
analysis was taken from a range of insurance market data
and research from insurance bodies such as LIMRA, the
American Council of Life Insurers (ACLI), and the
Medical Expenditure Panel Survey (MEPS). RMS is also
grateful for the information and inputs provided by its
clients and sponsors of this study in the creation of the
exposure database.
Figure 4.2 Activity patterns of working age, professional females during a normal
work day---each demographic segment has its own activity patterns over time
4.1.2 Demographic Segmentation
The population is segmented into the following primary
demographic groups:
 By gender: male, female
By age: working age (16 to 64), pre-working age
(under 16), post-working age (65 & over)
By occupation: professional, manual or non-professional,
not employed
Figure 4.3 Distribution of population by age bracket for the U.S.
4.1.3 Location
The geographical basis of the mapping is a variable resolution
grid (VRG) where grid cells are sized according to
the density of population. At its finest resolution, the
data exists for a specific building, related to the mapping
of individual buildings in major city centers. Grid cells
holding human exposure data vary from 164-foot (50-
meter) cells to about a third of a mile (500-meter) in
cities and up to 6.2 mile (10 km) cells in lesser populated
rural areas. The data can be thought of as residing at
‘neighborhood’ level, or for a small part of a town, and
can be aggregated for each ZIP Code. The population
residing in each neighborhood is segmented into each of
the key demographic groups.The whereabouts of working
age people that reside in a particular neighborhood
are related to the parts of town (or other location) that they
work in.The people that work in each part of town can be
traced back to the neighborhoods in which they live.
The demographics that can be examined with this
information can be as general as patterns by sex or age
and as specific as patterns by 19 year-old part-time
working females living in New York.
4.1.4 Time Dimensions
The data is held for each hour of the day and a separate
24-hour cycle is derived for work days and non-work
days. Non-work days are comprised of weekends and
public holidays.
4.1.5 Insurance Coverage Rule-base
Some insurance lines, such as life insurance, provide continuous
coverage. Others, like workers compensation,
are conditional---if you are at work or engaged in work
activities, you are covered, but outside the workplace
you are not. Some insurance lines include compensation
for injury, others only apply if the policyholder is killed.
While life insurers may be liable for claims occurring
at any time during the day, workers compensation and
health insurers may be liable for portions of time,
depending on the location and activity of the exposed
person. An injury that is covered under workers compensation
is not also covered by health insurance---they
are mutually exclusive. Other coverages may apply in
conjunction with each other---if a policyholder is injured
at work and has both workers compensation and disability
insurance, the disability insurance will pay the portion
of salary that is outside the bounds of the workers
compensation coverage.
Life insurance policy benefits are paid irrespective of
other coverages. A group life policy, one or more individual
life policies, and an AD&D policy will all be paid
out in the event of death. If the death occurs at work,
dependants would also receive benefits from the workers
compensation insurer.
Table 4.1 Examples of the insurance coverage rule-base of the human
exposure
database
Figure 4.4 The number of people with different coverages varies by geography and
time of day
For these reasons, certain perils show varied loss distributions
to different lines of insurance, depending on
the time of day, location, and activities of the insureds.
These patterns are explained later in greater detail for
the various scenarios and perils modeled in this study. In
the case of earthquake, the scenarios consider different
times of occurrence and the benefits of corresponding
insurance coverages that would apply.
|
