ages and years indicated. Note that the
simulated deaths are based on an artificial mortality table and improvement
scale. One, therefore, should look not at
the absolute values but at the range of
variation from year to year and age to age.
To my eye, the simulated deaths show at
least as much variation as the actual. It’s
important that the simulated deaths be
“lifelike.” I believe that they are.
1998
85
71
38
21
11
10
236
1999
87
60
36
26
19
7
234
2000
101
60
39
20
16
3
238
Total
511
387
238
132
93
49
1410
Developing Average Mortality
The real work began after I had the simulated deaths. I applied my method to
each set of deaths, inferring exposures
from the simulated deaths and then
graduating the raw mortality rates to
develop an average mortality table for
each scenario.
The amount of additional work to do
wasn’t significant. I already had a program
to graduate a mortality table from a set of
deaths. All I had to do was obtain the deaths
from an array rather than from a worksheet
and put a loop around the calculation.
Figure 2 compares the “actual” mortality rates with the rates calculated for
the first three scenarios. The “actual”
rates, of course, aren’t really actual.
They are the assumed set of Gompertz
mortality rates, modified by a cohort
improvement scale. The matrix of mortality rates for each year and age isn’t
wildly different from what may have
fIgURE 1: Actual deaths and Simulated deaths
Actual Deaths for Canadian Males
Age 1995 1996 1997
100 70 76 93
101 72 66 58
102 41 33 52
103 21 20 24
104 18 16 13
105 10 12 7
Total 232 223 247
Simulated Deaths, Scenario 1
Age 1995 1996
100 102 91
101 72 89
102 43 57
103 32 26
104 22 19
105 13 16
Total 284 298
Simulated Deaths, Scenario 2
Age 1995 1996
100 89 110
101 84 93
102 44 48
103 35 33
104 22 18
105 13 12
Total 287 314
Simulated Deaths, Scenario 3
Age 1995 1996
100 116 96
101 66 85
102 55 39
103 30 37
104 20 18
105 22 16
Total 309 291
1997
92
74
38
33
20
13
270
1998
120
73
65
36
29
12
335
1999
97
80
50
24
32
9
292
2000
138
97
52
29
27
14
357
Total
640
485
305
180
149
77
1836
1997
93
63
47
30
19
9
261
1998
89
76
50
32
19
24
290
1999
108
87
47
29
20
15
306
2000
109
78
47
38
22
19
313
Total
598
481
283
197
120
92
1771
1997
116
64
59
26
20
10
295
1998
116
67
49
38
24
8
302
1999
101
73
46
38
19
15
292
2000
103
80
59
34
27
8
311
Total
648
435
307
203
128
79
1800
