Arnold Kling  

Life Expectancy Statistics

PRINT
Best Books on Health Care... Scott Sumner vs. Milton Friedm...

Tyler Cowen writes,


Matt Yglesias and Paul Krugman weigh in on interpreting life expectancy statistics across the U.S. and the Netherlands. The fact under consideration, from a few days ago, is that the U.S. has low life expectancy overall but superior life expectancy after you reach the age of 65.

One way to interpret this data (re: Yglesias and Krugman) is to think that the U.S. should spread Medicare to its entire population.

Does anyone know how, if at all, unnatural deaths affect these data? Consider four possibilities--young in the U.S., young in the Netherlands, old in the U.S., old in the Netherlands. I would bet that of these four groups, the rate of unnatural death (murder, car accidents, and so on) is highest for the young in the U.S. If true, and if the life expectancy data are not adjusted to remove unnatural deaths, then this seems to me the most likely explanation for the fact.


Comments and Sharing





COMMENTS (24 to date)
Patrick C writes:

One way to interpret the data:
- The US has a higher level of wealth and income inequality
- Poverty in the US is associated with less access to preventative health care
- Poor people then die younger
- Those people who live to be 65 are wealthier and have access to very high quality health care

Etc.

Alex J. writes:

I would look for a factor that causes certain populations in the US to have both higher infant mortality and early deaths. The high infant mortality would reduce a random American baby's chance of making it to 65. The early deaths in this population would make the surviving people be disproportionately from the healthier overall population.

Milton Recht writes:

See July 31 post on Angry Bear Blog, which references a study, by Robert L. Ohsfeldt and John E. Schneider, that adjusts OECD life expectancies for different accident and homicide rates. They find US has the highest OECD life expectancy after adjustment.

"The US has far higher fatalities from homicide and traffic accidents than other countries which impact the life expectancy statistics....
[Table omitted]
So the authors controlled for the differing non-health care related deaths to develop a life expectancy table that could more accurately reflect the relationship between health care quality and life expectancy:
[Table omitted]
The US jumps from 15th on the list with a life expectancy of 75.3 to 1st with a life expectancy of 76.9."

Read entire blog post here.

Les writes:

Patrick C offered one interpretation of the data.

I do not see much point in speculating about alternative interpretations of the data. There are an unknown number of alternative speculations.

For example, it may be that some Americans are ignorant about smoking, nutrition and exercise, and as a result die young. Others, better informed, may live to a ripe old age.

Since speculation is endless and inconclusive, I don't see any purpose to it.

Tyler writes:

Mark Perry @ Carpe Diem looked at the data almost two years on his blog and concluded that after you adjust for homicide and MVA, two problems we don't look to the healthcare system to 'fix', "the U.S. looks pretty good, and actually has the highest standardized life expectancy in the world, according to the University of Iowa study."

http://mjperry.blogspot.com/2007/11/beyond-those-health-care-numbers-us.html

Matthew writes:

I have also read that many "infant mortality" cases in the US are counted as "stillbirths" in most of the world, as those countries choose not to engage in expensive heroic efforts to save premies and other very sick infants. Not sure if this is a legitime fact or not.

Ironman writes:

Milton Recht: I'm afraid Angry Bear is, as always, late to the party. Here's Ohsfeldt's and Schneider's national life expectancy data adjusted for non-natural causes of death presented in a more dynamic format (click the column headings to sort the data) in a post from 13 September 2007.

Moving onto other points, Arnold is correct in assuming that the data giving the average number of years of remaining life expectancy at given ages by nation is not corrected for non-natural causes of death, such as vehicle accidents and homicide.

The U.S. data is also skewed by the much greater, earlier mortality for minority populations. Aside from non-natural causes of death, chronic diseases claim a disproportionate number of the lives of many U.S. minorities at younger ages. We did a series of posts on the topic, hypothesizing that a chronic vitamin deficiency may be a major contributor to that discrepancy.

Here's the conclusion to the series - links to earlier posts in the series are provided at the bottom of the post. The deficiency in question affects half of the African American population in the U.S., with the Hispanic population being the second-most affected group.

Finally, here's a more sophisticated approach to cross-national comparison of health expenditures per capita and GDP per capita, which breaks down the data for larger nations into sub-national units, which are more directly comparable to the data for much smaller nations. It's not linear.

Patrick C writes:

Les wrote:

"Patrick C offered one interpretation of the data.

I do not see much point in speculating about alternative interpretations of the data. There are an unknown number of alternative speculations.

For example, it may be that some Americans are ignorant about smoking, nutrition and exercise, and as a result die young. Others, better informed, may live to a ripe old age.

Since speculation is endless and inconclusive, I don't see any purpose to it."

Ideally someone would develop a hypothesis, and we would then test the hypothesis. I'm too lazy to do any real work on this, but it would be interesting to see any hypotheses attacked with data.

To those who are talking just about life expectancy - you are missing 1/2 of the story, specifically related to lower life expectancy pre-age 65. You are only comparing total life expectancy.

kingstu writes:

This is another example of Krugman's simplistic analytical skills when he writes about health care issues. In the U.S. we spend an extraordinary amount of money on keeping older people alive longer. We transfer huge sums of money from the working age population to the retired population to fund their health care.

If the U.S. expands Medicare to younger people, isn’t a decrease in life expectancy for those over 65 just as likely as resources are shifted from older Americans to younger Americans?

Stephen Z writes:

I have noticed that many of the countries that beat the US in overall life-expectancy charts have populations which are relatively quite homogeneous. It would be interesting to see the data broken up by ethnicity.

Troy Camplin writes:

There is no question that the high murder rate and high death rate due to accidents in the U.S. are major contributors to the high death rate among the young. No question at all. It is disingenuous for proponents of health care and insurance nationalization to disregard these facts. Also, the U.S. has an extremely high level of immigration relative to almost every other country. Immigrants tend to have the higher infant mortality rates of their countries of origin. This also has a significant downward pull. Adjust for these facts, and see what happens.

Huxley writes:

The proper comparison is white Americans to white Europeans. How does someone of northern European descent in Michigan or Wisconsin compare to someone in the Netherlands?

Tom West writes:

Looking at Angry Bear's article, I notice two sets of figures:

USA:
Including homicide + traffic: 75.3
Excluding homicide + traffic: 76.9

So far, so good. The USA loses 1.6 years of average lifespan to these two.

Canada:
Including homicide + traffic: 77.3
Excluding homicide + traffic: 76.2

So this means that some Canadians who would have died of medical causes at 76, stay alive so they can get murdered at 77? :-)

Does anyone have a less amusing, but perhaps more accurate explanation as to how Canada's (and several other nations) lifespans go *down* once you exclude homicide and traffic?

hacs writes:

The paper "Low Life Expectancy in the United States: Is the Health Care System at Fault?" from Samuel H. Preston and Jessica Y. Ho (http://repository.upenn.edu/cgi/viewcontent.cgi?article=1012&context=psc_working_papers) enlightens that point. The real challenge is the cost of health care in the USA somewhat discussed in that article also.

Dr. T writes:

Life expenctancy figures are not a good way to assess a nation's health care system for two main reasons:

#1 We count neonatal deaths differently than everyone else. What we call infant deaths, most other nations call stillbirths (which are excluded from life expectancy statistics). The US makes great efforts to save premies; every failed effort counts heavily against us because it represents a statistical loss of 80 years of life.

#2 We have high numbers of homicides, suicides, and accidental deaths. Despite having some of the best trauma care in the world, our sheer numbers of traumas produces a big death toll, especially among teens and young adults. But, these deaths are not an indictment of health care, they're an indictment of our violence-laden culture.

If we recategorize neonatal deaths to an agreed upon standard and remove trauma deaths from the life expectancy statistics, I believe you will find that the US is at or near the top in life expectancy related to health care.

Methinks writes:

It would be interesting to see the data broken up by ethnicity.

"In 2005, life expectancy at birth was 76 years for white males compared with 70 years for black males and 81 years for white females compared with 77 years for black females (data table for Figure 14). Life expectancy at birth increased more for the black than for the white population between 1990 and 2005 (Figure 14). During this period, the gap in life expectancy between white males and black males narrowed from 8 years to 6 years (data table for Figure 14). During the same period, the gap in life expectancy between white females and black females decreased from 6 years to 4 years.
The gap in life expectancy between white and black people at age 65 is narrower than at birth. Between 1990 and 2005, the difference in life expectancy at age 65 between white males and black males remained stable at 2 years. In 2005, life expectancy at age 65 was 17 years for white males and 15 years for black males. The difference in life expectancy between white and black females has also been stable in recent years; in 2005, at age 65, white females and black females could expect to live an additional 20 and 19 years, respectively."

Full report here:

http://www.cdc.gov/nchs/data/hus/hus08.pdf

I don't know why they didn't include the large Hispanic population, or the Asian population in the United States.

Methinks writes:

What Dr. T said. Unless he was pushing a political agenda, I've never heard a WHO epidemiologist claim that life expectancy differences between developed countries were due the health care system. The difference is too small not to be attributable to how the statistics are compiled and factors outside the scope of the health care system - lifestyle, for example.

The WHO uses Life expectancy as a measure of health care systems between underdeveloped countries and developed countries - unless they're pushing a political agenda. Anyone who has ever been a patient in Italy will burst into uncontrollable laughter at the WHO's #2 ranking of that country's health care.

John Fembup writes:

kingstu asks this reasonable question: "If the U.S. expands Medicare to younger people, isn’t a decrease in life expectancy for those over 65 just as likely as resources are shifted from older Americans to younger Americans?"

But it's not at all clear that this must be so. I think it likely - perhaps more likely - that even greater resources would be shifted to older Americans in Medicare-for-all scheme. In a Medicare-for-all scheme, everyone would have the same Medicare-level benefits, which would represent a reduction in benefits for younger Americans. Since the demand for medical care will remain much higher from older Americans, resources would thus be shifted away from younger Americans.

Is it true that Medicare benefits for all would represent a reduction in benefits for younger Americans? If present Medicare benefits were provided for everyone, the answer is definitely yes. Here is a comparison of my insurance (group insurance thru my employer) with my sister's (Medicare)

My insurance:

1. My deductible = 1,750 per year
2. After deductible, my plan pays 80% of allowed expenses
3. My residual max (the 20% expenses) = 2,000 per year
4. My plan pays 100% for the rest of the year, after I meet my deductible and my residual max
5. My employer funds a reimbursement account for me = $600 per year. (I have these funds automatically applied to my claims)
6. My maximum share of my own medical costs in any year, no matter what, is $3,150 (1) + (3) –(5).
7. My plan pays 100% of preventive care expenses
8. My plan covers retail Rx. All prescriptions = $15 copay then 80%. The copays are paid by the reimbursement account.
9. My plan covers the full range of dental care, subject to the above cost sharing
10. My plan has no lifetime maximum
11. My covers me anywhere in the world
12. My premium (net of my employer's contribution) is $43 per month or $516 per year.

I know I will pay $516 every year — that’s my premium. The maximum I can possibly be charged for my own expenses in any year is $3,150. So my medical care costs will vary between $516 minimum and $3,666 maximum per year. This includes full catastrophic medical care protection.

My sister’s insurance

1. Her plan has many deductibles,
a. Her plan’s inpatient expenses deductible = $1,068 per confinement
·Inpatient benefits are limited to 150 days per confinement.
·Her plan pays inpatient benefits at 100% up to 60 days per confinement
·Her plan requires a $267 per day deductible from 61-90 days
·Her plan requires a $534 per day deductible from 90-150 days.
·After 150 days – no coverage
b. Her plan’s deductible for all other types of expenses = $135 per year
2. For these other types, her plan pays 80% of allowed expenses after the deductible
3. Her plan does not limit her residual expenses (the 20% expenses)
4. Her plan would continue to reimburse 80% regardless how large her expenses may grow - and she would continue to pay her 20% no matter how large they may grow.
5. Her plan does not have a health reimbursement or health savings account.
6. Her maximum share of her own medical costs in any year cannot be known in advance.
7. Preventive care is subject to the same deductible and 80% reimbursement
8. Her plan does not cover retail Rx – no prescriptions
9. Her plan does not cover dental
10. Her plan’s benefits are limited as defined by the coverage terms
11. My sister's plan will not reimburse any expenses outside the U.S.
12. Her premium) net of the federal subsidy) is $96 per month, or $1,152 per year.

My Sister’s premium is double my premium. However, her coverage is clearly inferior to mine.

For her much higher premium, she does not have the coverage I have for routine expenses - and she has clearly worse lifetime catastrophic medical care protection. In fact, she cannot ever know her maximum possible yearly cost – because there is no such maximum in the plan.

Does this comparison represent fairly the benefit difference generally, between Medicare and employer-sponsored plans? I think it does. I'm not claiming to quantify the difference, just to gauge its direction.

My opinion is based on my observations over the years as head of benefits for an employer-sponsored plan, and as a benefits consultant. Keep in mind the large majority of people covered in private insurance plans are in employer-sponsored plans. The plan comparison above does not reflect actuarial averages for private plans, but I believe it is nevertheless a fair comparison.

FWIW, I think having the lifetime unlimited catastrophic protection in exchange for a maximum - maximum - yearly cost of around 3,000 is a fair deal. You may not think so. Wouldn't our difference of opinion at least suggest that both types of insurance should remain available to people?

PS - I will become eligible for Medicare in 2010. I can hardly contain my enthusiasm.


Methinks writes:

What Dr. T said. Unless he was pushing a political agenda, I've never heard a WHO epidemiologist claim that life expectancy differences between developed countries were due the health care system. The difference is too small not to be attributable to how the statistics are compiled and factors outside the scope of the health care system - lifestyle, for example.

The WHO uses Life expectancy as a measure of health care systems between underdeveloped countries and developed countries - unless they're pushing a political agenda. Anyone who has ever been a patient in Italy will burst into uncontrollable laughter at the WHO's #2 ranking of that country's health care.

Healthcare outcomes are repeatedly cited to claim that the expense of US healthcare is wasted, and that only governments deliver quality care. This is supported by a biased interpretation of the statistics.

USA Healthcare is First - Infant Mortality is Low


John Stossel (Why the U.S. Ranks Low on WHO's Health-Care Study) analyzes that life expectancy is a lousy measure of a health-care system. We have far more fatal transportation accidents than other countries. Our homicide rate is 10 times greater than in the U.K., eight times greater than in France, and five times greater than in Canada.

When you adjust for these "fatal injury" rates, U.S. life expectancy is actually higher than in nearly every other industrialized nation. That doesn't show a health-care problem.

The infant mortality statistics are also carefully biased. The US counts every live birth, however premature, toward its statistics, even if the infant lives only a few hours. European countries may only count infants that live at least a day or which meet other health criteria. So, they claim fewer infant deaths, which dramatically changes the average life expectancy in years.

Rdan writes:

Milton Recht,
Thanks for the link. It was written by one our more conservative readers and posted.

Ironman,
Late to the party Angry Bear comments that without checking your statement, it becomes a sad reflection on the impulse to hit the submit button.. However, nice series.

Stephen Z writes:

Methinks: Thanks for the post!

Rdan writes:

Carefully biased is a charged statement Andrew...you need to prove it...

There are other problems with statistics that cannot be answered without the data as well, so be very careful about making firm conclusions with this study...it leaves out a lot of other factors. I used it at Angry Bear as an idea post, not an accurate and precise study nor definitive in answers. It is a mistake to take it so...

Tom,
I missed that about Canada. Thanks.

David Mark writes:

The discussion is very informative. One question does extension of the dying process (ie the unnatural maintaining of body function especially ventilators, for people in comas or are brain dead) skew the data in the other direction.
I believe that most other countries are much less likely to extend the dying process as we do here.

Comments for this entry have been closed
Return to top