Life Expectancy Isn’t Always Fair · 2018-04-11 · 2 Life Expectancy, Gains Have Varied by Gender But life expectancy is a potentially misleading statistic 0 5 10 15 20 25 30 1940

Life Expectancy Isn’t Always Fair

Enrolled Actuaries Meeting General Session 3

Steve GossOffice of the Chief Actuary

Social Security Administration

April 11, 2018

2

Life Expectancy, Gains Have Varied by GenderBut life expectancy is a potentially misleading statistic

2

0

5

10

15

20

25

30

1940 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090

Period Life Expectancy at 652017 Trustees Report Intermediate

Male Female

Males falling behind

Males catching

up

Survival Curves Tell a More Complete StoryWe are living longer, but there are limits, and progress is slowing

33

0.00.10.20.30.40.50.60.70.80.91.0

0 10 20 30 40 50 60 70 80 90 100 110

Survival Curve U.S. Female: Period Data

19001925

1950

1975

2000

2013

0.00.10.20.30.40.50.60.70.80.91.0

0 10 20 30 40 50 60 70 80 90 100 110

Survival Curve U.S. Male: Period Data

1900

1925

1950

1975

2000

2013

Variation by Education and Earnings Level

Distribution has been changing for each– So best to look at Percentiles

Death rates, and change in death rates– Allow for comparison of actual experience

Education and earnings level correlated– Both reflect relative advantages– Nature and nurture

4

Female Death Rates by Education Level at 65-85John Bound et al.(2014)

Female Non-Hispanic-White Population Annualized Death Rates from Age 65 to 85

by Educational Attainment

0%

1%

2%

3%

4%

5%

All < HighSchool

Low 25%Education

High 75%Education

19902010

5

Male Death Rates by Education Level at 65-85John Bound et al. (2014)

Male Non-Hispanic-White Population Annualized Death Rates from Age 65 to 85

by Educational Attainment

0%1%2%3%4%5%6%7%8%

All < HighSchool

Low 25%Education

High 75%Education

19902010

6

Death Rates Vary by Career Earnings Ranking Difference has increased for females (2014)

Female 65-69 Retired-Worker Relative Death Rates by AIME Quartile

00.20.40.60.8

11.21.41.6

1 2 3 4

19902010

7

Death Rates Vary by Career Earnings Ranking Difference has increased for males too (2014)

Male 65-69 Retired-Worker Relative Death Rates by AIME Quartile

00.20.40.60.8

11.21.41.6

1 2 3 4

19902010

8

Relative Death Rates for Age 65-9 Social Security RetireesDifferential has been expanding slowly for females (Bosley 2018)

9

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2000 2005 2010 2015

Relat

ive M

orta

lity

Ratio

Year

Females

Lowest AIMEQuintile

2nd AIMEQuintile

3rd AIMEQuintile

4th AIMEQuintile

Highest AIMEQuintile

Relative Death Rates for Age 65-9 Social Security RetireesExpansion may be reversing for males?? (Bosley 2018)

10

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2000 2005 2010 2015

Rela

tive

Mor

talit

y Ra

tio

Year

Males

Lowest AIMEQuintile

2nd AIMEQuintile

3rd AIMEQuintile

4th AIMEQuintile

Highest AIMEQuintile

Mortality Decline Varies Over TimeConditions: Antibiotics/economy 1936-54; Medicare/Medicaid 1968-82

Female Historical and Projected (2014 Trustees Report) Annual Percent Reduction in U.S. Mortality Rates

0

1

2

3

4

5

Under Age 15 Ages 15 - 49 Ages 50 - 64 Ages 65 - 84 Ages 85 andolder

Total

1900 to 19361936 to 19541954 to 19681968 to 19821982 to 2010

Male Historical and Projected (2014 Trustees Report) Annual Percent Reduction in U.S. Mortality Rates

-10

123

45

Under Age 15 Ages 15 - 49 Ages 50 - 64 Ages 65 - 84 Ages 85 andolder

Total

1900 to 19361936 to 19541954 to 19681968 to 19821982 to 2010

11

Mortality Decline by Cause of Death: Rate of change from 1979 to 2013

-4

-3

-2

-1

0

1

2

3

4

Under 15 15-49 50-64 65-84 85+

CardiovascularCancerViolenceRespiratoryOther

FEMALE

-3

-2

-1

0

1

2

3

4

Under 15 15-49 50-64 65-84 85+

MALE

12

Mortality Experience: Ages 65 and OlderReductions since 2009 continue to fall short of expectations

131313

Factors Affecting Future ChangeSmoking - still some effect for females who

started and stopped later than menObesity - major factor yet to be fully

exploredHealth spending and access - will this be

even, or differ greatly by earnings level?

14

Trends in Obesity: US 1971-2006 Sam Preston 2010—must consider cumulative effects

Increasing duration of obesity for aged in future

1515

Health Spending Cannot Continue to Rise at Historical RatesNote Trustees’ deceleration

16

-1%

0%

1%

2%

3%

4%

5%

1970-1980 1980-2010 2010-2040 2040-2070 2070-2085

Annual Percent Change in Medicare Cost per Beneficiary Relative to GDP per Worker: 2015 TR

1616

17

For More Information…http://www.ssa.gov/oact/

Documentation of Trustees Report data & assumptions https://www.ssa.gov/oact/TR/2017/2017_Long-Range_Demographic_Assumptions.pdf

Historical and projected mortality rateshttps://www.ssa.gov/oact/HistEst/DeathHome.html

Actuarial Studies and Noteshttps://www.ssa.gov/oact/NOTES/actstud.html

17