In support of Keith

[David Friedman]

I had dinner tonight with one of my temporary (i.e. GMU) colleagues who,
among other things, teaches a course on health economics. According to
him, it is very difficult to find evidence linking life expectancy to
any of the obvious inputs. In particular, there is apparently a database
of some millions of people, coming out of (I think) medicare, which can
be analyzed to see how health care expenditure and mortality results
vary from one place to another. Expenditure per capita varies over about
a factor of two, with no detectable effect on outcome.

I was reminded of Keith's claim that the health care provided by
insurance has only a small effect on life expectancy. It sounds as
though such evidence as is available is consistent with that claim.

--
http://www.daviddfriedman.com/
http://daviddfriedman.blogspot.com/
_Salamander_: http://tinyurl.com/6957y7e
_How to Milk an Almond,..._ http://tinyurl.com/63xg8gx

[Cryptoengineer]

On Nov 10, 9:49 pm, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:

> I had dinner tonight with one of my temporary (i.e. GMU) colleagues who,
> among other things, teaches a course on health economics. According to
> him, it is very difficult to find evidence linking life expectancy to
> any of the obvious inputs. In particular, there is apparently a database
> of some millions of people, coming out of (I think) medicare, which can
> be analyzed to see how health care expenditure and mortality results
> vary from one place to another. Expenditure per capita varies over about
> a factor of two, with no detectable effect on outcome.
>
> I was reminded of Keith's claim that the health care provided by
> insurance has only a small effect on life expectancy. It sounds as
> though such evidence as is available is consistent with that claim.

I think its important to separate the issues of

1: Does having health insurance increase life expectancy in the US?
2: Does getting medical care increase life expectancy?

The thing is, very few people are willing to refuse vitally needed
care, either to receive or to give, to people with no insurance but a
life-threatening condition. Between charity and emergency rooms, a lot
of lives are saved.

pt

[Mike Benveniste]

On 11/10/2011 9:49 PM, David Friedman wrote:

> I was reminded of Keith's claim that the health care provided by
> insurance has only a small effect on life expectancy. It sounds as
> though such evidence as is available is consistent with that claim.

Y'all may find this report worth reading:
https://download.nap.edu/catalog.php?record_id=12511

As a quick preview -- 4 studies examined found that having health
insurance had a statistically significant impact on mortality at the
95% level. 2 studies did not find statistical significance.

--
Mike Benveniste -- m...@murkyether.com (Clarification Required)
You don't have to sort of enhance reality. There is nothing
stranger than truth. -- Annie Leibovitz

[Keith F. Lynch]

Mike Benveniste <m...@murkyether.com> wrote:
> Y'all may find this report worth reading:
> https://download.nap.edu/catalog.php?record_id=12511

> As a quick preview -- 4 studies examined found that having health
> insurance had a statistically significant impact on mortality at
> the 95% level. 2 studies did not find statistical significance.

It's surprising that two studies did not find statistical
significance. I've never denied that there is statistical
significance. To me, the important question isn't whether medical
insurance buys you any additional life expectancy, but *how much*
additional life expectancy it does buy you. If it buys you one month,
but it costs a decade's income, then it's not unreasonable for someone
to choose to do without it. Especially if the extra life it buys you
is of very low quality, i.e. you're alive, but far too sick to enjoy
anything.
--
Keith F. Lynch - http://keithlynch.net/
Please see http://keithlynch.net/email.html before emailing me.

[David Friedman]

In article <9i3jtr...@mid.individual.net>,

Mike Benveniste <m...@murkyether.com> wrote:

> On 11/10/2011 9:49 PM, David Friedman wrote:
>
> > I was reminded of Keith's claim that the health care provided by
> > insurance has only a small effect on life expectancy. It sounds as
> > though such evidence as is available is consistent with that claim.
>
> Y'all may find this report worth reading:
> https://download.nap.edu/catalog.php?record_id=12511
>
> As a quick preview -- 4 studies examined found that having health
> insurance had a statistically significant impact on mortality at the
> 95% level. 2 studies did not find statistical significance.

Which pages are those studies on? There seem to be a lot of studies
listed, with rather brief descriptions of what was done and what the
results were.

[David Friedman]

In article
<08a446ae-cfe8-49f3...@o9g2000vbc.googlegroups.com>,

Cryptoengineer <pete...@gmail.com> wrote:

> I think its important to separate the issues of
>
> 1: Does having health insurance increase life expectancy in the US?
> 2: Does getting medical care increase life expectancy?
>

I agree. The question is how one gets data on the latter question?

If you can't, then the former might be the best proxy, on the theory
that having health insurance results, on average, in getting more and
better medical care. The size of the effect would be reduced by the fact
that the uninsured get some care, but the sign shouldn't be.

The colleague I mentioned said that the Rand study, which I think is the
one Keith referred to in the past, was the best such information out
there, despite its problems.

[garabik-ne...@kassiopeia.juls.savba.sk]

Keith F. Lynch <k...@keithlynch.net> wrote:
> Mike Benveniste <m...@murkyether.com> wrote:
>> Y'all may find this report worth reading:
>> https://download.nap.edu/catalog.php?record_id=12511
>
>> As a quick preview -- 4 studies examined found that having health
>> insurance had a statistically significant impact on mortality at
>> the 95% level. 2 studies did not find statistical significance.
>
> It's surprising that two studies did not find statistical
> significance. I've never denied that there is statistical
> significance. To me, the important question isn't whether medical
> insurance buys you any additional life expectancy, but *how much*
> additional life expectancy it does buy you. If it buys you one month,
> but it costs a decade's income, then it's not unreasonable for someone
> to choose to do without it.

OTOH, if you earn and save enough that a decade's income is expendable
without major burden on your lifestyle, it might be worth it.
Especially if your average savings amount to a decade's income - the
difference between additional month and "taking a decade's income worth
into your grave month earlier" is quite in favour of the former.
OT3H, a decade's income might well pay for the additional month.
OT4H, you might have relatives who you care about and prefer to leave
some heritage for them.

Without good statistical data and a game theory analysis, you cannot say
one way or another.

A real life example: we have compulsory insurance here. I knew a guy who
was self employed and was NOT paying it (financial difficulties). Then
he got sick (cancer) and was almost denied treatment (beyond the
emergency life threatening situations). He managed to talk his way in
and continued making debts at an alarming rate, since the treatments
were directly billed to him. After some months he died, leaving the
hospital with a sore hole in their budget.
Net gain for him.

--
-----------------------------------------------------------
| Radovan Garabík http://kassiopeia.juls.savba.sk/~garabik/ |
| __..--^^^--..__ garabik @ kassiopeia.juls.savba.sk |
-----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!

[Doug Wickström]

On Thu, 10 Nov 2011 21:49:11 -0500, David Friedman
<dd...@daviddfriedman.nopsam.com> wrote:

>I had dinner tonight with one of my temporary (i.e. GMU) colleagues who,
>among other things, teaches a course on health economics. According to
>him, it is very difficult to find evidence linking life expectancy to
>any of the obvious inputs. In particular, there is apparently a database
>of some millions of people, coming out of (I think) medicare, which can
>be analyzed to see how health care expenditure and mortality results
>vary from one place to another. Expenditure per capita varies over about
>a factor of two, with no detectable effect on outcome.

I'm not surprised. The mortality rate, over time, is 100%.

>I was reminded of Keith's claim that the health care provided by
>insurance has only a small effect on life expectancy. It sounds as
>though such evidence as is available is consistent with that claim.

Keith's claim is true, but arguably for different reasons than he
claims.

_Necessary_ health care, particularly preventives such as
immunizations, extends the life span of the general population by
decades. So does good nutrition, and healthy living conditions
(malaria was endemic in much of Scandinavia, until houses began
to have windows that closed tightly. Screens would help even
more, but for some reason screens aren't as prevalent as in North
America.

_Unnecessary_ health care often kills. It carries all of the
risks associated with necessary health care, and very little
benefit. I had a co-worker whose father died due to an infection
contracted during bypass surgery. The surgery was to improve the
man's quality of life; there was no expectation that it would
significantly extend it. Instead, his life was shortened by
several years.

In the United States, simple medical mistakes kill approximately
100,000 people a year.

Much of the high cost of medical care is driven by costly new
medicines, that may or may not work as well as older ones, and
may or may not have fewer or less negative side effects. I know
that I'd rather live with my psoriasis than deal with the side
effects (including death) of the medications that can actually
treat it into remission.

If you have three feet of steel rebar rammed through your
shoulder, you probably need surgery. If you have an enlarged
prostate, you probably do not, and surgery would be a greater
risk than prostate cancer.

The hard cases are those in the middle. No one has ever died of
a migraine. But migraines are debilitating enough that not
treating them is a poor option.

--
Doug Wickström

[Mike Benveniste]

On 11/10/2011 11:30 PM, David Friedman wrote:

> Which pages are those studies on? There seem to be a lot of studies
> listed, with rather brief descriptions of what was done and what the
> results were.

The summary table is on page 70, the review of the various
studies begins on page 74, the discussion of higher overall
mortality rates begins on page 80, and the reference table
is, as expected at the end.

I'm not sure all of the referenced studies are available online
for free, but I was able to find this one fairly quickly:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1702516/?tool=pubmed

[David Friedman]

In article <9i4en8...@mid.individual.net>,

Mike Benveniste <m...@murkyether.com> wrote:

> On 11/10/2011 11:30 PM, David Friedman wrote:
>
> > Which pages are those studies on? There seem to be a lot of studies
> > listed, with rather brief descriptions of what was done and what the
> > results were.
>
> The summary table is on page 70, the review of the various
> studies begins on page 74, the discussion of higher overall
> mortality rates begins on page 80, and the reference table
> is, as expected at the end.
>
> I'm not sure all of the referenced studies are available online
> for free, but I was able to find this one fairly quickly:
>
> http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1702516/?tool=pubmed

Thanks.

[David Friedman]

In article <cropb75j3f93un8bn...@4ax.com>,

Doug Wickström <nims...@comcast.net> wrote:

> _Necessary_ health care, particularly preventives such as
> immunizations, extends the life span of the general population by
> decades. So does good nutrition, and healthy living conditions

According to the colleague, actually finding results along those
lines--I think he specifically mentioned sanitation--is surprisingly
hard.

[cryptoengineer]

On Fri, 11 Nov 2011 06:38:57 -0500, David Friedman
<dd...@daviddfriedman.nopsam.com> wrote:
> In article <cropb75j3f93un8bn...@4ax.com>,
> Doug Wickström <nims...@comcast.net> wrote:


> > _Necessary_ health care, particularly preventives such as
> > immunizations, extends the life span of the general population by
> > decades. So does good nutrition, and healthy living conditions


> According to the colleague, actually finding results along those
> lines--I think he specifically mentioned sanitation--is
surprisingly
> hard.

> http://daviddfriedman.blogspot.com/

I'd be curious as to how he explains the doubling in Western life
expectancy over the last century or so.

Pt

[David Friedman]

In article <almarsoft.7601...@news.giganews.com>,

He doesn't. That's the puzzle.

[cryptoguy]

On Nov 11, 6:37 pm, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article <almarsoft.7601897304706330...@news.giganews.com>,

>
>
>
>  cryptoengineer <petert...@Gmail.com> wrote:
> > On Fri, 11 Nov 2011 06:38:57 -0500, David Friedman

> > <d...@daviddfriedman.nopsam.com> wrote:
> > > In article <cropb75j3f93un8bnn2bjf80gb3fbj0...@4ax.com>,

> > >  Doug Wickström <nimshu...@comcast.net> wrote:
>
> > > > _Necessary_ health care, particularly preventives such as
> > > > immunizations, extends the life span of the general population by
> > > > decades.  So does good nutrition, and healthy living conditions
>
> > > According to the colleague, actually finding results along those
> > > lines--I think he specifically mentioned sanitation--is
> > surprisingly
> > > hard.
>
> > >http://daviddfriedman.blogspot.com/
>
> > I'd be curious as to how he explains the doubling in Western life
> > expectancy over the last century or so.
>
> He doesn't. That's the puzzle.

Barring more data, my conclusion is that his analysis is crocked.

pt

[Doug Wickström]

On Fri, 11 Nov 2011 06:38:57 -0500, David Friedman
<dd...@daviddfriedman.nopsam.com> wrote:

>In article <cropb75j3f93un8bn...@4ax.com>,
> Doug Wickström <nims...@comcast.net> wrote:
>
>> _Necessary_ health care, particularly preventives such as
>> immunizations, extends the life span of the general population by
>> decades. So does good nutrition, and healthy living conditions
>
>According to the colleague, actually finding results along those
>lines--I think he specifically mentioned sanitation--is surprisingly
>hard.

One of the counter-intuitives that contributes to this is the
unexpected _increase_ in deaths due to polio because of modern
sanitation. The cleaner we got, the more we tried to prevent
polio through public sanitation, the more virulent the disease
became.

--
Doug Wickström

[David Friedman]

In article
<01d24eb9-ed5d-4cbf...@cu3g2000vbb.googlegroups.com>,

I don't think it's his analysis--he teaches the course, but health
economics isn't his specialty. It's his report on what surprised him
about the literature in the field.

[Keith F. Lynch]

Doug Wickström <nims...@comcast.net> wrote:

> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
>> Doug Wickström <nims...@comcast.net> wrote:
>>> _Necessary_ health care, particularly preventives such as
>>> immunizations, extends the life span of the general population by
>>> decades. So does good nutrition, and healthy living conditions

Decades for each, or decades in total?

>> According to the colleague, actually finding results along
>> those lines--I think he specifically mentioned sanitation--is
>> surprisingly hard.

There's no doubt that life expectancies have increased in the US. I'm
pretty sure the main causes are knowledge of sanitation, nutrition,
and safety, and the increased per capita wealth necessary for the
majority of the population to be able to afford to act on that
knowledge. (Now that per capita wealth is dropping in the US, I
expect life expectancies to decrease. Knowledge is of little use
to those who can't afford to apply it.)

Medical science's main contribution was primarily to learn what
habits are and aren't healthy. And secondly to provide inexpensive
treatments such as vaccinations. What's really relevant to the
question of whether medical insurance is worthwhile is how much
benefit the average person gets from medical treatments that aren't
affordable out of pocket. My contention is that that benefit *can*
be large -- hence lots of anecdotes of people saved by million-dollar
treatments -- but on average it's small to nonexistant.

And even with those anecodtes, you can seldom be sure that the person
wouldn't have survived just fine with less treatment or none at all.

> One of the counter-intuitives that contributes to this is the
> unexpected _increase_ in deaths due to polio because of modern
> sanitation. The cleaner we got, the more we tried to prevent polio
> through public sanitation, the more virulent the disease became.

Nitpick: I think it's more a matter of that disease, like mumps,
being far worse for adults than for children. With less sanitation,
most people got it as children.

[Keith F. Lynch]

Doug Wickström <nims...@comcast.net> wrote:
> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

>> Expenditure per capita varies over about a factor of two, with no
>> detectable effect on outcome.

> I'm not surprised. The mortality rate, over time, is 100%.

It's logically possible for the life expectancy of people with medical
insurance to be 90, and for people without medical insurance to be 40.
That wouldn't preclude the mortality rate, over time, being 100%. And
it would be a good reason to want everyone to have medical insurance.

It still wouldn't mean that it was *possible* for everyone to have
medical insurance. If the average American can't afford to pay for
medical care for the average American, then every American pooling
their money won't help, not even if insurance companies had no
overhead and paid their workers no salary.

But as it is, the gain in life expectancy is very small, a matter of
at most a few weeks. So I see no point at all in the debate. Let
each person decide for himself whether medical insurance is worth
buying. It *certainly* has far less effect on life expectancy than
smoking does, and yet people are allowed to smoke. So why not allow
people to be uninsured?

> _Unnecessary_ health care often kills. It carries all of the risks
> associated with necessary health care, and very little benefit. I
> had a co-worker whose father died due to an infection contracted
> during bypass surgery. The surgery was to improve the man's quality
> of life; there was no expectation that it would significantly extend
> it. Instead, his life was shortened by several years.

ObFandom: That's also what killed Isaac Asimov.

> In the United States, simple medical mistakes kill approximately
> 100,000 people a year.

I'm not sure what that means. Anyone who dies while under medical
care was arguable killed by a medical mistake, in the sense that he
probably would have lived a little longer had his dose of medication
been slightly different, or had his surgery been performed on a
different day or by a different surgeon. There's a smooth continuum
between that and the kind of gross medical errors that are shocking to
the conscience, such as doing a major operation on the wrong patient.

> Much of the high cost of medical care is driven by costly new
> medicines, that may or may not work as well as older ones, and
> may or may not have fewer or less negative side effects.

Yes. As far as I'm concerned, medications that are still under patent
are exclusively for the wealthy. Let them pay for the privilege of
being guinea pigs. Patents only last 20 years. 1991 wasn't exactly
the Black Death, so I for one would be perfectly satisfied with
insurance that paid for the best medications of that year and earlier.

> I know that I'd rather live with my psoriasis than deal with the
> side effects (including death) of the medications that can actually
> treat it into remission.

Some people will choose otherwise, depending on how bad the psoriasis
is, how vain they are, and what proportion of its users the medication
kills. And, of course, how much the medication costs.

> If you have three feet of steel rebar rammed through your shoulder,
> you probably need surgery. If you have an enlarged prostate, you
> probably do not, and surgery would be a greater risk than prostate
> cancer.

I might reverse that. The rebar, I could, with some risk, pull out
myself. Or, also with some risk, I could cut level with the skin and
leave it there indefinitely.

I don't think there's any link between enlarged prostate and prostate
cancer. The former is a problem mainly in that it makes urination
difficult or impossible. I'd rather risk losing an arm than be unable
to urinate. Infection or blood loss from the rebar *might* kill me,
but being unable to urinate would *certainly* kill me.

> The hard cases are those in the middle. No one has ever died of a
> migraine. But migraines are debilitating enough that not treating
> them is a poor option.

Are migraine treatments so expensive that the median-income American
can't afford them without medical insurance?

[Keith F. Lynch]

<garabik-ne...@kassiopeia.juls.savba.sk> wrote:
> Keith F. Lynch <k...@keithlynch.net> wrote:
>> It's surprising that two studies did not find statistical
>> significance. I've never denied that there is statistical
>> significance. To me, the important question isn't whether medical
>> insurance buys you any additional life expectancy, but *how much*
>> additional life expectancy it does buy you. If it buys you one
>> month, but it costs a decade's income, then it's not unreasonable
>> for someone to choose to do without it.

> OTOH, if you earn and save enough that a decade's income is
> expendable without major burden on your lifestyle, it might be
> worth it.

Certainly. But it should be your choice. I don't object to medical
insurance. I object to it being made mandatory. (Or forbidden.) I'd
feel the same even if there was an obviously right choice to make, as
there is with smoking, as I value freedom above all else.

> Especially if your average savings amount to a decade's income - the
> difference between additional month and "taking a decade's income
> worth into your grave month earlier" is quite in favour of the former.

Of course it's a month *on average*. You may get an additional 30
years of healthy life, or you may get nothing. Or the medical care
may even shorten your life.

Nobody would buy lottery tickets for a dollar if they always returned
exactly 30 cents. But plenty of people buy lottery tickets for a
dollar that return 30 cents *on average*.

> OT3H, a decade's income might well pay for the additional month.
> OT4H, you might have relatives who you care about and prefer to
> leave some heritage for them.

> Without good statistical data and a game theory analysis, you cannot
> say one way or another.

Indeed. So I object to the high-handed way Obama presumes to order
every American to buy medical insurance.

> A real life example: we have compulsory insurance here. I knew a guy
> who was self employed and was NOT paying it (financial difficulties).

"Compulsory" meaning someone who doesn't buy it can be sent to jail?
Was that guy sent to jail?

> Then he got sick (cancer) and was almost denied treatment (beyond
> the emergency life threatening situations). He managed to talk his
> way in and continued making debts at an alarming rate, since the
> treatments were directly billed to him. After some months he died,
> leaving the hospital with a sore hole in their budget. Net gain
> for him.

Net loss for the hospital, maybe, but it doesn't sound like he gained
very much. The value of the treatment was obviously far less than
its cost.

Reducing costs should be the number one priority for medical research.
If treatments only affordable by the wealthy should be researched at
all, that research should be paid for only by the wealthy.

[Jette Goldie]

perhaps they would like to think that they will be the one who gets an
extra 30 years of healthy life.


--
Jette Goldie jette....@gmail.com
http://www.flickr.com/photos/wolfette/ http://wolfette.livejournal.com/
("reply to" is spamblocked - use the email addy in sig)

[David Friedman]

In article <j9mh1f$iq$1...@reader1.panix.com>,

"Keith F. Lynch" <k...@KeithLynch.net> wrote:

> But as it is, the gain in life expectancy is very small, a matter of
> at most a few weeks.

If I understand the situation correctly, that figure is based on an
experiment done some decades back. It was an experiment with individuals
randomly assigned to insurance/no insurance (or something
equivalent--I'm going on a second hand description), and may be the best
data we have, but it still had a variety of problems.

Given how often you cite the result, have you read the actual
description of the study?

[cryptoguy]

On Nov 12, 11:29 am, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article
> <01d24eb9-ed5d-4cbf-a5a0-824491616...@cu3g2000vbb.googlegroups.com>,

>
>
>
>  cryptoguy <treifam...@gmail.com> wrote:
> > On Nov 11, 6:37 pm, David Friedman <d...@daviddfriedman.nopsam.com>
> > wrote:
> > > In article <almarsoft.7601897304706330...@news.giganews.com>,
>
> > >  cryptoengineer <petert...@Gmail.com> wrote:
> > > > On Fri, 11 Nov 2011 06:38:57 -0500, David Friedman
> > > > <d...@daviddfriedman.nopsam.com> wrote:
> > > > > In article <cropb75j3f93un8bnn2bjf80gb3fbj0...@4ax.com>,
> > > > >  Doug Wickström <nimshu...@comcast.net> wrote:
>
> > > > > > _Necessary_ health care, particularly preventives such as
> > > > > > immunizations, extends the life span of the general population by
> > > > > > decades.  So does good nutrition, and healthy living conditions
>
> > > > > According to the colleague, actually finding results along those
> > > > > lines--I think he specifically mentioned sanitation--is
> > > > surprisingly
> > > > > hard.
>
> > > > >http://daviddfriedman.blogspot.com/
>
> > > > I'd be curious as to how he explains the doubling in Western life
> > > > expectancy over the last century or so.
>
> > > He doesn't. That's the puzzle.
>
> > Barring more data, my conclusion is that his analysis is crocked.
>
> I don't think it's his analysis--he teaches the course, but health
> economics isn't his specialty. It's his report on what surprised him
> about the literature in the field.

Health care - including such things as clean water, and malaria
elimination, have been clearly linked to prevention of disease. If
your colleague is unaware of such famous examples as the Broad Street
pump handle incident [1], his qualification to give such a course is
in question.

[1] https://secure.wikimedia.org/wikipedia/en/wiki/1854_Broad_Street_cholera_outbreak

IMO, he's making an extraordinary claim.

pt

[David Friedman]

In article
<eeafe2bc-49e9-4b22...@m7g2000vbc.googlegroups.com>,

cryptoguy <treif...@gmail.com> wrote:

> > I don't think it's his analysis--he teaches the course, but health
> > economics isn't his specialty. It's his report on what surprised him
> > about the literature in the field.
>
> Health care - including such things as clean water, and malaria
> elimination, have been clearly linked to prevention of disease. If
> your colleague is unaware of such famous examples as the Broad Street
> pump handle incident [1], his qualification to give such a course is
> in question.

If you are unfamiliar with the use of statistics to establish causal
relations, your qualifications to judge his are in question.

[karl.j...@gmail.com]

On Nov 12, 11:22 am, "Keith F. Lynch" <k...@KeithLynch.net> wrote:
> Doug Wickström <nimshu...@comcast.net> wrote:

> > David Friedman <d...@daviddfriedman.nopsam.com> wrote:
> >> Expenditure per capita varies over about a factor of two, with no
> >> detectable effect on outcome.
> > I'm not surprised.  The mortality rate, over time, is 100%.
>
> It's logically possible for the life expectancy of people with medical
> insurance to be 90, and for people without medical insurance to be 40.
> That wouldn't preclude the mortality rate, over time, being 100%.  And
> it would be a good reason to want everyone to have medical insurance.
>
> It still wouldn't mean that it was *possible* for everyone to have
> medical insurance.  If the average American can't afford to pay for
> medical care for the average American, then every American pooling
> their money won't help, not even if insurance companies had no
> overhead and paid their workers no salary.

Lots of things cost less if bought in bulk.

> > _Unnecessary_ health care often kills.  It carries all of the risks
> > associated with necessary health care, and very little benefit.  I
> > had a co-worker whose father died due to an infection contracted
> > during bypass surgery.  The surgery was to improve the man's quality
> > of life; there was no expectation that it would significantly extend
> > it.  Instead, his life was shortened by several years.
>
> ObFandom:  That's also what killed Isaac Asimov.
>
> > In the United States, simple medical mistakes kill approximately
> > 100,000 people a year.
>
> I'm not sure what that means.  Anyone who dies while under medical
> care was arguable killed by a medical mistake, in the sense that he
> probably would have lived a little longer had his dose of medication
> been slightly different, or had his surgery been performed on a
> different day or by a different surgeon.  There's a smooth continuum
> between that and the kind of gross medical errors that are shocking to
> the conscience, such as doing a major operation on the wrong patient.

Interesting and well said.

Karl Johanson

[Keith F. Lynch]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
> "Keith F. Lynch" <k...@KeithLynch.net> wrote:
>> But as it is, the gain in life expectancy is very small, a matter
>> of at most a few weeks.

> If I understand the situation correctly, that figure is based on
> an experiment done some decades back. It was an experiment with
> individuals randomly assigned to insurance/no insurance (or
> something equivalent--I'm going on a second hand description), and
> may be the best data we have, but it still had a variety of problems.

> Given how often you cite the result, have you read the actual
> description of the study?

It's my understanding that this information comes from a variety of
sources. And that the only sources that disagree with it say that
the benefit of medical insurance is statistically indistinguishable
from *zero*.

I first learned of it from my father, who was an actuary. I
independently learned of it from Michael Vassar, the head of the
Singularity Institute, who had also been trained as an actuary. I
didn't think it was at all controversial. *Do* you know of any
studies that conclude that medical insurance makes a large positive
difference to life expectancies? Thanks.

[Keith F. Lynch]

karl.j...@gmail.com <karl.j...@gmail.com> wrote:
> Lots of things cost less if bought in bulk.

Medical insurance is already bought in bulk. Medical insurance is
expensive because medical care is expensive. Not all medical care.
Not most medical care. But enough medical care to push the average
way up for everyone. If there's a one percent chance you'll need
a million-dollar operation, insuring against it will cost at least
$10,000 -- even if that million-dollar operation only has a small
chance of saving your life.

What's needed are ways to bring the cost of medical care down.
Robotic surgery, perhaps. (The term "robotic surgery" has
unfortunately already been appropriated to mean something that *does*
involve the usual expensive team of surgeons and support staff, so I
suppose we'll have to call true robotic surgery something else.)

[Keith F. Lynch]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

> cryptoguy <treif...@gmail.com> wrote:
>> Health care - including such things as clean water, and malaria
>> elimination, have been clearly linked to prevention of disease.
>> If your colleague is unaware of such famous examples as the Broad
>> Street pump handle incident [1], his qualification to give such a
>> course is in question.

> If you are unfamiliar with the use of statistics to establish causal
> relations, your qualifications to judge his are in question.

By "medical care" most of us mean one-on-one care by a doctor, nurse,
or equivalent, and medical devices and medications, i.e. things that
medical insurance pays for.

[David Friedman]

In article <j9mu64$67f$1...@reader1.panix.com>,

"Keith F. Lynch" <k...@KeithLynch.net> wrote:

> > Given how often you cite the result, have you read the actual
> > description of the study?
>
> It's my understanding that this information comes from a variety of
> sources. And that the only sources that disagree with it say that
> the benefit of medical insurance is statistically indistinguishable
> from *zero*.
>
> I first learned of it from my father, who was an actuary. I
> independently learned of it from Michael Vassar, the head of the
> Singularity Institute, who had also been trained as an actuary. I
> didn't think it was at all controversial. *Do* you know of any
> studies that conclude that medical insurance makes a large positive
> difference to life expectancies? Thanks.

Someone posted links to references to some studies recently, I think in
this group. One was:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1702516/?tool=pubmed

[cryptoguy]

On Nov 12, 6:16 pm, "Keith F. Lynch" <k...@KeithLynch.net> wrote:
> David Friedman <d...@daviddfriedman.nopsam.com> wrote:

> > cryptoguy <treifam...@gmail.com> wrote:
> >> Health care - including such things as clean water, and malaria
> >> elimination, have been clearly linked to prevention of disease.
> >> If your colleague is unaware of such famous examples as the Broad
> >> Street pump handle incident [1], his qualification to give such a
> >> course is in question.
> > If you are unfamiliar with the use of statistics to establish causal
> > relations, your qualifications to judge his are in question.
>
> By "medical care" most of us mean one-on-one care by a doctor, nurse,
> or equivalent, and medical devices and medications, i.e. things that
> medical insurance pays for.

If Keith had followed the thread, he'd know that David had mentioned
that 'sanitation' was under discussion.

pt

[David Friedman]

In article
<34c81832-3aae-4be8...@i6g2000vbe.googlegroups.com>,

Yes.

The basic point my friend was making, as I understood it, was that we
knew there had been large increases in life expectancy, that there were
some obvious conjectures as to the causes, but that actual attempts to
test those conjectures were surprisingly unsuccessful.

[cryptoguy]

On Nov 12, 5:54 pm, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article
> <eeafe2bc-49e9-4b22-84af-b7caf545d...@m7g2000vbc.googlegroups.com>,

>
>  cryptoguy <treifam...@gmail.com> wrote:
> > > I don't think it's his analysis--he teaches the course, but health
> > > economics isn't his specialty. It's his report on what surprised him
> > > about the literature in the field.
>
> > Health care - including such things as clean water, and malaria
> > elimination, have been clearly linked to prevention of disease. If
> > your colleague is unaware of such famous examples as the Broad Street
> > pump handle incident [1], his qualification to give such a course is
> > in question.
>
> If you are unfamiliar with the use of statistics to establish causal
> relations, your qualifications to judge his are in question.

Read up on the case I cited, and which you clipped out. It's
considered the founding event of statistical epidemiology.

pt

[Doug Wickström]

On Sat, 12 Nov 2011 18:35:34 +0000 (UTC), "Keith F. Lynch"
<k...@KeithLynch.net> wrote:

>Doug Wickstr?nims...@comcast.net> wrote:
>> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

Well, duh.

BTW, your newsreader is still broken.

--
Doug Wickström

[Doug Wickström]

On Sat, 12 Nov 2011 19:22:23 +0000 (UTC), "Keith F. Lynch"
<k...@KeithLynch.net> wrote:

Absolute medical error. Wrong patient, wrong medicine, wrong
dosage, wrong treatment, dropping the patient in the operating
room, getting the wrong blood type for the transfusion, giving
NSAIDS to patients known to be allergic, etc., etc., etc. ... .

And then there are the errors that don't actually kill, such as
amputating the wrong leg.

--
Doug Wickström

[garabik-ne...@kassiopeia.juls.savba.sk]

Keith F. Lynch <k...@keithlynch.net> wrote:

> <garabik-ne...@kassiopeia.juls.savba.sk> wrote:

>> I don't object to medical
> insurance. I object to it being made mandatory. (Or forbidden.) I'd
> feel the same even if there was an obviously right choice to make, as
> there is with smoking, as I value freedom above all else.

The problem is that most people would not pay it, and would not save
enough, and the time would come when they'd need serious medical
attention later in their life (unless they drop dead on the street).
And then they (or their relatives) will rebel and demand the authorities
to do something. I could see the heart breaking news stories - "the
hospital let my dad die because he could not pay".

What about emergency, life threatening situations? It would be most
disturbing to see the emergency medical crew fix broken hand of one car
accident victim, while the second victim is slowly bleeding to death in
front of their eyes, because he chose not to pay the insurance.

>
> Nobody would buy lottery tickets for a dollar if they always returned
> exactly 30 cents. But plenty of people buy lottery tickets for a
> dollar that return 30 cents *on average*.

If you naïvely apply game theory to lottery, you of course do not buy
the ticket. However, you have to realize that you should measure the
costs and payoffs in utilons, not in dollars. $1 is almost
infinitesimally small (in utilons), while $1M payoff is huge - it is
perfectly rational to buy the ticket.

Value of additional month of life is relatively small, if it is 30 years
in future, but immensely huge when it is the following month.

>
>> A real life example: we have compulsory insurance here. I knew a guy
>> who was self employed and was NOT paying it (financial difficulties).
>
> "Compulsory" meaning someone who doesn't buy it can be sent to jail?
> Was that guy sent to jail?

You do not go to jail for each breach of law.

Not paying the insurance is considered more of a breach of contractual
obligation - for some months it goes unnoticed (unless you need medical
attention), then you start getting mails, the insurance company starts
charging interests, then they send a collection agency. Usually up this
time, if you pay it off, they will cancel the accumulated interests.
If you do not pay, the collection agency will sell your possession, and
if it does not cover your debts, you can apply for personal bankruptcy.
Or live on with ever increasing debts (which is actually rather easy,
once you get past the psychological issues and have some friends you can
trust).

Interestingly, from wikipedia:
"More than a third of all U.S. states allow borrowers who can't or won't
pay to be jailed."

>
>> treatments were directly billed to him. After some months he died,
>> leaving the hospital with a sore hole in their budget. Net gain
>> for him.
>
> Net loss for the hospital, maybe, but it doesn't sound like he gained
> very much. The value of the treatment was obviously far less than
> its cost.

Yes - in his case, it bought him maybe a months... plus some palliative
care.

>
> Reducing costs should be the number one priority for medical research.

No, number one should be finding treatments for so far untreatable
frequent diseases. _Then_ you can work on reducing the costs.

> If treatments only affordable by the wealthy should be researched at
> all, that research should be paid for only by the wealthy.

Do not forget that this diffuses down - first only the wealthy can
afford it, but in 20 years everyone with modest income, in 40 years it
is a common, cheap treatment. Usually.
--
-----------------------------------------------------------
| Radovan Garabík http://kassiopeia.juls.savba.sk/~garabik/ |
| __..--^^^--..__ garabik @ kassiopeia.juls.savba.sk |
-----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!

[David Friedman]

In article <j9o6m2$4kf$1...@speranza.aioe.org>,

garabik-ne...@kassiopeia.juls.savba.sk wrote:

> If you naïvely apply game theory to lottery, you of course do not buy
> the ticket. However, you have to realize that you should measure the
> costs and payoffs in utilons, not in dollars. $1 is almost
> infinitesimally small (in utilons), while $1M payoff is huge - it is
> perfectly rational to buy the ticket.

Provided, of course, that you have increasing rather than decreasing
marginal utility of income, which there is no reason to expect you to
have. That is why the purchase of lottery tickets--especially the
purchase of lottery tickets by people who also buy insurance, which is
evidence of declining marginal utility of income--is a puzzle.

[David Friedman]

In article
<a9e8456c-ed68-492d...@p5g2000vbm.googlegroups.com>,

It provided evidence for how cholera was transmitted. It does not tell
us whether cleaner water reduced overall mortality and so helps to
explain the increase in life expectancy, which was the question at issue.

If the distinction isn't clear, consider current discussions of the
apparent increase in allergies, some of them life threatening. One
explanation that has been offered is that it is a result of our living
in a cleaner environment and so being exposed to fewer pathogens in
childhood. I don't know if that explanation is correct, but it provides
one example of how a change could prevent particular deaths without
increasing life expectancy.

[garabik-ne...@kassiopeia.juls.savba.sk]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

> Provided, of course, that you have increasing rather than decreasing
> marginal utility of income, which there is no reason to expect you to
> have. That is why the purchase of lottery tickets--especially the
> purchase of lottery tickets by people who also buy insurance, which is
> evidence of declining marginal utility of income--is a puzzle.
>

There are several factors - first is positive utility gained form the
thrill of checking the lottery results (cf. people who buy tickets for a
football match) + the expected outcome, weighted against time spent
buying the ticket + $1.
(that is why I do not buy lottery tickets - the first value oscillates
around zero for me).

Second issue is that small amount of money cease to be[1] real (decimal)
numbers and start to be infinitesimals (I often do not collect 1¢ of change
money).

The third issue is the "phase transition" if your income grows over
a certain sum. Winning $10k is nice, but does not change your life.
Winning $100M completely changes everything. For some (many) people the
expected outcome of winning $100M _vastly_ exceeds the raw mathematical
$1 * 1e8.

[1] psychologically, of course

[David Friedman]

In article <j9omuh$h86$1...@speranza.aioe.org>,

garabik-ne...@kassiopeia.juls.savba.sk wrote:

> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
>
> > Provided, of course, that you have increasing rather than decreasing
> > marginal utility of income, which there is no reason to expect you to
> > have. That is why the purchase of lottery tickets--especially the
> > purchase of lottery tickets by people who also buy insurance, which is
> > evidence of declining marginal utility of income--is a puzzle.
> >
>
> There are several factors - first is positive utility gained form the
> thrill of checking the lottery results (cf. people who buy tickets for a
> football match) + the expected outcome, weighted against time spent
> buying the ticket + $1.

...

I would add to this the pleasure of daydreaming about being rich, which
is easier to do if there is some possible causal sequence you can
imagine that would make you rich.

> Second issue is that small amount of money cease to be[1] real (decimal)
> numbers and start to be infinitesimals (I often do not collect 1¢ of change
> money).

One cent is a very small amount of money, and it may not be worth the
trouble of dealing with it. But the price of a lottery ticket is a sum
that people routinely do care about; it will buy them things--a drink, a
dinner--that they value. Do you also reject ten dollars in change?

> The third issue is the "phase transition" if your income grows over
> a certain sum. Winning $10k is nice, but does not change your life.
> Winning $100M completely changes everything. For some (many) people the
> expected outcome of winning $100M _vastly_ exceeds the raw mathematical
> $1 * 1e8.

Why would the benefit from a hundred million be more than ten thousand
times that from ten thousand? If anything, the standard arguments for
declining MUI would seem to apply. The ten thousand can be spent on
things important to you. Most of the hundred million goes for a third
mansion you rarely visit, a car five times as expensive and only a
little better than the alternative car, ... .

One can, of course, draw a utility function to explain the behavior,
convex up for the low range and concave up for the high, as per the
Friedman and Savage article on the puzzle, but how do you provide a
plausible justification for that shape?

[cryptoguy]

On Nov 13, 9:51 am, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article
> <a9e8456c-ed68-492d-9f7b-b07ed985c...@p5g2000vbm.googlegroups.com>,

>
>
>
>  cryptoguy <treifam...@gmail.com> wrote:
> > On Nov 12, 5:54 pm, David Friedman <d...@daviddfriedman.nopsam.com>
> > wrote:
> > > In article
> > > <eeafe2bc-49e9-4b22-84af-b7caf545d...@m7g2000vbc.googlegroups.com>,
>
> > >  cryptoguy <treifam...@gmail.com> wrote:
> > > > > I don't think it's his analysis--he teaches the course, but health
> > > > > economics isn't his specialty. It's his report on what surprised him
> > > > > about the literature in the field.
>
> > > > Health care - including such things as clean water, and malaria
> > > > elimination, have been clearly linked to prevention of disease. If
> > > > your colleague is unaware of such famous examples as the Broad Street
> > > > pump handle incident [1], his qualification to give such a course is
> > > > in question.
>
> > > If you are unfamiliar with the use of statistics to establish causal
> > > relations, your qualifications to judge his are in question.
>
> > Read up on the case I cited, and which you clipped out. It's
> > considered the founding event of statistical epidemiology.
>
> It provided evidence for how cholera was transmitted. It does not tell
> us whether cleaner water reduced overall mortality and so helps to
> explain the increase in life expectancy, which was the question at issue.

That's a *really* strained take on the evidence.

pt

[David Friedman]

In article
<370995c7-6b68-4817...@y12g2000vba.googlegroups.com>,

cryptoguy <treif...@gmail.com> wrote:

> > It provided evidence for how cholera was transmitted. It does not tell
> > us whether cleaner water reduced overall mortality and so helps to
> > explain the increase in life expectancy, which was the question at issue.
>
> That's a *really* strained take on the evidence.

On the contrary. I am describing my colleague's comment. You are
straining to turn it into something you can attack.

[rksh...@rosettacondot.com]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
> In article <j9omuh$h86$1...@speranza.aioe.org>,
> garabik-ne...@kassiopeia.juls.savba.sk wrote:
>
>> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
>>
>> > Provided, of course, that you have increasing rather than decreasing
>> > marginal utility of income, which there is no reason to expect you to
>> > have. That is why the purchase of lottery tickets--especially the
>> > purchase of lottery tickets by people who also buy insurance, which is
>> > evidence of declining marginal utility of income--is a puzzle.
>> >
>>
>> There are several factors - first is positive utility gained form the
>> thrill of checking the lottery results (cf. people who buy tickets for a
>> football match) + the expected outcome, weighted against time spent
>> buying the ticket + $1.
>
> ...
>
> I would add to this the pleasure of daydreaming about being rich, which
> is easier to do if there is some possible causal sequence you can
> imagine that would make you rich.
>
>> Second issue is that small amount of money cease to be[1] real (decimal)

>> numbers and start to be infinitesimals (I often do not collect 1¢ of change

>> money).
>
> One cent is a very small amount of money, and it may not be worth the
> trouble of dealing with it. But the price of a lottery ticket is a sum
> that people routinely do care about; it will buy them things--a drink, a
> dinner--that they value. Do you also reject ten dollars in change?

Ten dollars, probably not. A dollar? Sure...I did that last night. I
gave the waiter $10 to cover a $6.50 tab and told him to keep the
change.

>> The third issue is the "phase transition" if your income grows over
>> a certain sum. Winning $10k is nice, but does not change your life.
>> Winning $100M completely changes everything. For some (many) people the
>> expected outcome of winning $100M _vastly_ exceeds the raw mathematical
>> $1 * 1e8.
>
> Why would the benefit from a hundred million be more than ten thousand
> times that from ten thousand? If anything, the standard arguments for
> declining MUI would seem to apply. The ten thousand can be spent on
> things important to you. Most of the hundred million goes for a third
> mansion you rarely visit, a car five times as expensive and only a
> little better than the alternative car, ... .
>
> One can, of course, draw a utility function to explain the behavior,
> convex up for the low range and concave up for the high, as per the
> Friedman and Savage article on the puzzle, but how do you provide a
> plausible justification for that shape?

I don't find this difficult to envision at all...$10K is enough to
finish paying off a car loan, make a (probably negligible) dent in a
mortgage, pay down a credit card, take the family on a vacation overseas
(once), etc., but it's not going to make a permanent change in someone's
lifestyle. They're not going to quit their job, buy a house, or buy a bigger
house if they already have one, send their kids to a good college, etc.
The money's gone and back to work on Monday.
$100M is life changing. Virtually anyone could use that sum to wipe out
all of their debt, buy a new house outright, retire and at the same time
vastly improve their lifestyle.
For me, personally, there's a threshold effect...a somewhat fuzzy point
where the amount transitions from "nice" to "oh, wow"...the point at
which it would make a fundamental difference to my life rather than just
an increase in the balance of my savings account. That point, right now,
is well above $10K and well below $100M.
You said it yourself above...$10 will buy things that $1 won't, and $1
will buy things that a penny won't. In that sense I get more than 100
times the utility out of a dollar than I get out of a penny...the penny
is just a weight in my pocket until it accumulates enough friends, the
dollar will buy me a soda.

Robert
--
Robert K. Shull Email: rkshull at rosettacon dot com

[Keith F. Lynch]

<rksh...@rosettacondot.com> wrote:
> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
>> I would add to this the pleasure of daydreaming about being rich,
>> which is easier to do if there is some possible causal sequence you
>> can imagine that would make you rich.

In my case, I get that by envisioning my proving the Riemann
Hypothesis, for which there's a million-dollar reward. This would
bring not just wealth, but also fame and excellent job offers.

And I think I have a better chance of proving it than I have of
winning a million-dollar lottery. And I learn interesting math, and
get good mental exercise, in the process.

> $10K is enough to ... make a (probably negligible) dent in a
> mortgage, ...

ObFandom: Heinlein sold his first story for $50 so as to make a
substantial dent in his mortgage. Damned inflation.

> They're not going to ... send their kids to a good college, ...

With $10,000? What does tuition at a good college cost these days?

> In that sense I get more than 100 times the utility out of a dollar
> than I get out of a penny...the penny is just a weight in my pocket
> until it accumulates enough friends, the dollar will buy me a soda.

That doesn't make sense to me, except in the sense that the time it
takes to deal with a single penny may not be worth it. If I see a
penny on the ground, it takes me about two seconds to pick it up, and
about one second to put it in a roll. (I don't count the time it
takes me to take the roll to the bank, since I only do so when I'm
going to the bank anyway.) So that's about $12 per hour or $24,000
per year, plus an exercise benefit. As such, it's worth it to me, but
I'm not surprised that many people don't bother to pick up pennies. I
*am* surprised that so many people don't bother to pick up nickels, as
that's equivalent to a hefty $120,000 income, and few Americans other
than government employees earn that much.

[Keith F. Lynch]

<garabik-ne...@kassiopeia.juls.savba.sk> wrote:
> There are several factors - first is positive utility gained form
> the thrill of checking the lottery results (cf. people who buy
> tickets for a football match) + the expected outcome, weighted
> against time spent buying the ticket + $1.

> (that is why I do not buy lottery tickets - the first value
> oscillates around zero for me).

The overriding reason why I don't buy lottery tickets is that the
money goes straight to the government. As you know, I believe
governments do far more harm than good. Similarly, I'm willing to
spend $2 to avoid $1 in taxes, since I regard each dollar I pay in
taxes as contributing to evil. If I were wealthier I'd be willing
to spend a lot more to avoid each dollar in taxes.

> Second issue is that small amount of money cease to be[1] real
> (decimal) numbers and start to be infinitesimals (I often do not

> collect 1\202¢ of change money).

As I explained in my previous message, one cent is not infinitesimal.
For an median-income American, it's worth two or three seconds of
income. That's not very much, but it does mean that if you're handed
a penny in change, it's worth keeping it, rolling it with other
pennies once you get enough of them, and taking them to the bank
the next time you're going there anyway.

There are also machines in stores that will roll coins for you -- you
just pour the coins in, and the rolls come out. Or perhaps dollar
bills do, I'm not sure. I never use them since they keep several
percent as a convenience fee.

Of course the smallest unit of currency in your country might be worth
far less than a US penny, I don't know.

[garabik-ne...@kassiopeia.juls.savba.sk]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

> One cent is a very small amount of money, and it may not be worth the
> trouble of dealing with it. But the price of a lottery ticket is a sum
> that people routinely do care about; it will buy them things--a drink, a
> dinner--that they value. Do you also reject ten dollars in change?

Ten dollars (better: euro), no. One euro, under some
circumstances. 10¢ often. The transition from infinitesimals to real
money depends on the situation, amount of sum being paid and mood.

>
> Why would the benefit from a hundred million be more than ten thousand
> times that from ten thousand?

Because if I start adding 10000€ ten thousand times, I'd need many
lifetimes to arrive to 1e8€. Whereas if I win 1e8€, the money would
be there ready to use.

> If anything, the standard arguments for
> declining MUI would seem to apply. The ten thousand can be spent on
> things important to you. Most of the hundred million goes for a third
> mansion you rarely visit, a car five times as expensive and only a
> little better than the alternative car, ... .

People often do not bear well sudden immense wealth, and they often end
badly.
I would probably put 1M€ to nice personal use, with 10M€ I could try to
change some minor aspects of the world, but anything beyond that would
be past my ability to deal reasonably with the money.

>
> One can, of course, draw a utility function to explain the behavior,
> convex up for the low range and concave up for the high, as per the
> Friedman and Savage article on the puzzle, but how do you provide a
> plausible justification for that shape?
>

Psychology, of course. Plus transitions in a stratified society.
IMHO the personal phase transitions go roughly in this order:
1. dirt poverty (homeless etc.)
2. struggle to meet the bills
3. very well paid professional
4. rich bastard
5. fairy tale wealth

Lottery offers you a way to go from 1 or 2 to 5. This in itself adds
a lot to the expected utility.

[Keith F. Lynch]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
> It provided evidence for how cholera was transmitted. It does not
> tell us whether cleaner water reduced overall mortality and so helps
> to explain the increase in life expectancy, which was the question
> at issue.

I'd think the past causes of death would give us plenty of information
on the causes of the increase in life expectancy. For instance to the
extent that a common 19th-century cause of death was being trampled by
horses, we could credit today's smaller number of horses for today's
greater longevity.

> If the distinction isn't clear, consider current discussions of
> the apparent increase in allergies, some of them life threatening.
> One explanation that has been offered is that it is a result of
> our living in a cleaner environment and so being exposed to fewer
> pathogens in childhood. I don't know if that explanation is
> correct, but it provides one example of how a change could prevent
> particular deaths without increasing life expectancy.

My own family provides some evidence for that. When I was an infant,
my parents were, let's say, relaxed housekeepers. I'm told I used to
crawl behind the couch and eat dead flies, play in the cat's dirty
litter-box, etc. So I've been healthy all my adult life. I never get
so much as a cold, and I have no allergies. I've even gotten away
with handling poison ivy and eating whole tomato plants.

When working on medical software, I traveled to dozens of hospitals,
including ones in Iceland and Guantanamo Bay. I was warned that I
would get sick. I never did. A few months ago when my mother was in
a nursing home, she was in isolation due to an MRSA infection, and I
was asked to wear protective garments when visiting her. As soon as
I learned they were for my protection, not hers, I didn't bother.

ObFandom: Neither do I get sick eating con suite food that's been
handled by hundreds of other fen from all over the world, or even
that's fallen to the floor.

When visiting Guantanamo Bay, I sipped some water directly from the
bay, adding it my collection of bodies of water I've tasted (Atlantic,
Pacific, Caribbean, Lake Michigan, etc.) I probably would have
refrained had I known that both the Cubans and the Americans dump
raw untreated sewage into the bay. Neverthless, as always, I didn't
get sick.

By the time my brother was born, a decade after I was, my parents had
learned "better." He spent most of his infancy confined in a playpen
or high chair, and if any toy fell to the floor, it could not be
returned to him until it was sterilized. Today, he has lots of
allergies, and also gets colds.

[cryptoguy]

On Nov 13, 12:04 pm, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article
> <370995c7-6b68-4817-a554-6f78fe81e...@y12g2000vba.googlegroups.com>,

>
>  cryptoguy <treifam...@gmail.com> wrote:
> > > It provided evidence for how cholera was transmitted. It does not tell
> > > us whether cleaner water reduced overall mortality and so helps to
> > > explain the increase in life expectancy, which was the question at issue.
>
> > That's a *really* strained take on the evidence.
>
> On the contrary. I am describing my colleague's comment. You are
> straining to turn it into something you can attack.

People died when the pump's supply became contaminated by cholera-
infected sewage. They hadn't been dying before, and people in the same
area who used different water sources did not suffer from the
cholera.

The pump's well was dug only 3 feet from an old cesspit. It's probable
that sewage had been leaking into the water for some time, but until
the cholera was flushed into the cesspit (from a known source - a sick
baby's diapers), there was no *noticeable* extra disease.

One of the fundamental principles of not getting sick is 'don't shit
where you drink'. This principle was breached in this case, and 616
people died. Cholera did not spread beyond the neighborhood (with the
exception of an unfortunate woman who had the water shipped to her,
because she preferred the taste). Even when sufferers moved out of
the neighborhood (which became partially deserted as people, sick and
well, fled), they did not start new epidemics, since most wells didn't
have the leakage problem. It's a pretty open and shut case showing
that clean water prevented a highly lethal epidemic from spreading.

I strongly suspect that if you asked your colleague for studies
showing that sanitation led to overall *higher* rates of mortality,
he'd come up empty as well. There aren't many research grants given to
investigate whether or not 'In most places, most of the time, the Sun
will rise in the East' either.

pt

[David Friedman]

In article
<238a1e24-1dac-4a65...@y12g2000vba.googlegroups.com>,

cryptoguy <treif...@gmail.com> wrote:

> I strongly suspect that if you asked your colleague for studies
> showing that sanitation led to overall *higher* rates of mortality,
> he'd come up empty as well. There aren't many research grants given to
> investigate whether or not 'In most places, most of the time, the Sun
> will rise in the East' either.

His point was that attempts to find statistical links between the
obvious candidates for causes and reductions in mortality had been
surprisingly unsuccessful, not that they hadn't been made.

[Keith F. Lynch]

<garabik-ne...@kassiopeia.juls.savba.sk> wrote:
> The problem is that most people would not pay it, ...

Most Americans *do* pay for medical insurance, even though it's not
yet mandatory (except in Massachusetts, which has about 2% of the
US population). The proportion of Americans who get it is rapidly
dropping, however, as it becomes increasingly unaffordable.

> and would not save enough, and the time would come when they'd need
> serious medical attention later in their life (unless they drop
> dead on the street). And then they (or their relatives) will rebel
> and demand the authorities to do something. I could see the heart
> breaking news stories - "the hospital let my dad die because he
> could not pay".

Why doesn't this already happen? Insurance won't pay for everything.
For instance my mother's life could be saved by a heart-lung
transplant. But her insurance (Medicare) won't pay for it. And of
course we can't pay out of pocket. So it's not happening, meaning
that her quality of life is poor and she will probably die within a
year or two. And yet nobody is protesting.

> What about emergency, life threatening situations? It would be most
> disturbing to see the emergency medical crew fix broken hand of one
> car accident victim, while the second victim is slowly bleeding to
> death in front of their eyes, because he chose not to pay the
> insurance.

For any reasonable expense, the patient can be billed and can be
expected to pay. Brief emergency treatment isn't -- or shouldn't
be -- fantastically expensive.

> $1 is almost infinitesimally small (in utilons), while $1M payoff is
> huge - it is perfectly rational to buy the ticket.

I think that under most circumstances, the utility of money decreases,
not increases, with how much money you have. Your first dollar is
worth a lot more to you than your millionth dollar.

> Value of additional month of life is relatively small, if it is 30
> years in future, but immensely huge when it is the following month.

It's never that stark: "With this treatment you will certainly live
exactly 30 days in perfect health, then die. Without it you will die
today." It's all a matter of odds. Each day you flip a fair coin N
times. If it comes up heads fewer than M of those N times, you will
die that day. Insurance changes the values of N and M slightly.

> Not paying the insurance is considered more of a breach of

> contractual obligation ...

I don't agree that anyone has any contractual obligation they didn't
explicitly agree to.

> Interestingly, from wikipedia:
> "More than a third of all U.S. states allow borrowers who can't or
> won't pay to be jailed."

That's disturbing. I wish it applied to the biggest debtor of all,
the US government. I wouldn't mind seeing Obama in prison until he's
paid off the $15 trillion federal debt. Federal prisoners are paid 30
cents per hour. So it should take him about 25 billion years to pay
off the principal. As for 25 billion years of compound interest on
$15 trillion ....

But at least those debtors voluntarily incurred their debt. People
signed up for insurance against their will did not.

>> Reducing costs should be the number one priority for medical research.

> No, number one should be finding treatments for so far untreatable
> frequent diseases. _Then_ you can work on reducing the costs.

As long as the general public is being forced to pay, the research
should go into what will benefit them the most. A treatment which
is unaffordable to 99% of them is of no value to 99% of them.
And no, insurance won't help, since it just makes everything else
proportionately less affordable to them. If millionaires want to
pay for research into treatments affordable by millionaires, fine.
But nobody should be taxed to discover treatments which only people
wealthier than that taxpayer can afford.

> Do not forget that this diffuses down - first only the wealthy can
> afford it, but in 20 years everyone with modest income, in 40 years
> it is a common, cheap treatment.

True of medications, perhaps, as they become generic. But research on
any medications which are under patent are -- or ought to be -- paid
for entirely by the patent holder. The cost of such research is what
the patent holder uses to justify the high prices they charge, after
all. I certainly hope no taxpayer money is being used to enrich drug
companies.

As for MRI scans, major surgery, etc., prices have been going up with
time, not down.

[rksh...@rosettacondot.com]

Keith F. Lynch <k...@keithlynch.net> wrote:

> <rksh...@rosettacondot.com> wrote:
>> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
>>> I would add to this the pleasure of daydreaming about being rich,
>>> which is easier to do if there is some possible causal sequence you
>>> can imagine that would make you rich.
>
> In my case, I get that by envisioning my proving the Riemann
> Hypothesis, for which there's a million-dollar reward. This would
> bring not just wealth, but also fame and excellent job offers.
>
> And I think I have a better chance of proving it than I have of
> winning a million-dollar lottery. And I learn interesting math, and
> get good mental exercise, in the process.
>
>> $10K is enough to ... make a (probably negligible) dent in a
>> mortgage, ...
>
> ObFandom: Heinlein sold his first story for $50 so as to make a
> substantial dent in his mortgage. Damned inflation.

Yes...although inflation doesn't matter so much as long as incomes keep
up with it. My dad's family rented a house for $10/month when he was
growing up, but they also thought a good salary was $0.50 per 10+
hour day.

>> They're not going to ... send their kids to a good college, ...
>
> With $10,000? What does tuition at a good college cost these days?

$10,000 is about midpoint for yearly tuition at a state college. Private
colleges can be three or more times that much. Add in books, which have
had the same sort of inflation as tuition. If they're not lucky enough to
live near the college, add in room and board. For the full four years, figure
at least $60,000 for a state college, double or triple that for a private
college.

>> In that sense I get more than 100 times the utility out of a dollar
>> than I get out of a penny...the penny is just a weight in my pocket
>> until it accumulates enough friends, the dollar will buy me a soda.
>
> That doesn't make sense to me, except in the sense that the time it
> takes to deal with a single penny may not be worth it. If I see a
> penny on the ground, it takes me about two seconds to pick it up, and
> about one second to put it in a roll. (I don't count the time it
> takes me to take the roll to the bank, since I only do so when I'm
> going to the bank anyway.) So that's about $12 per hour or $24,000
> per year, plus an exercise benefit. As such, it's worth it to me, but
> I'm not surprised that many people don't bother to pick up pennies. I
> *am* surprised that so many people don't bother to pick up nickels, as
> that's equivalent to a hefty $120,000 income, and few Americans other
> than government employees earn that much.

Sure...but I don't find one penny every two seconds on the ground. I find
one per week or so, which means that in two years I could buy a soda. In
the same way, getting $10K every two seconds would be an entirely different
circumstance. That's $432 million per day, as opposed to once, conceivably
twice, in a lifetime, which is what we're talking about with lotteries.
I can't quite stretch my imagination far enough to get excited about
finding a nickel on the basis that doing so every few seconds would give
me a million dollars within the next decade. Worth doing, sure. Worth
getting excited about, nope.

[Keith F. Lynch]

That's different from a disease become more virulent. Diseases
generally evolve to become *less* virulent, so that they can be
passed on before killing their carrier.

> BTW, your newsreader is still broken.

"Wickstr?nimshubur" appeared in *your* message, in reply to a message
in which I quoted your name and address correctly, exactly as I'm
doing with the first instance of your name and address in this reply.
There was a different recent message, in a different thread, from
Carson Chittom, in which your last name appeared as "Wickstr\303\266m,"
and I quoted it as such in my reply to him.

[garabik-ne...@kassiopeia.juls.savba.sk]

Keith F. Lynch <k...@keithlynch.net> wrote:

> That doesn't make sense to me, except in the sense that the time it
> takes to deal with a single penny may not be worth it. If I see a
> penny on the ground, it takes me about two seconds to pick it up, and
> about one second to put it in a roll. (I don't count the time it

But it also diverts your thought process for those 3 seconds, and you
might also need some additional time to restart your previous stream of
reasoning. Depending on the situation, my thoughts could be worth more.

> takes me to take the roll to the bank, since I only do so when I'm
> going to the bank anyway.) So that's about $12 per hour or $24,000
> per year, plus an exercise benefit. As such, it's worth it to me, but
> I'm not surprised that many people don't bother to pick up pennies. I

I certainly wouldn't want a job where I earn $12/h by continuous bending
and straighting up, unless I'd be in difficulties and couldn't find
any other employment.

[garabik-ne...@kassiopeia.juls.savba.sk]

Keith F. Lynch <k...@keithlynch.net> wrote:

> The overriding reason why I don't buy lottery tickets is that the
> money goes straight to the government. As you know, I believe
> governments do far more harm than good. Similarly, I'm willing to
> spend $2 to avoid $1 in taxes, since I regard each dollar I pay in
> taxes as contributing to evil. If I were wealthier I'd be willing
> to spend a lot more to avoid each dollar in taxes.

Yes, that puts some weighting factors to your expected utility and your
not buying lottery tickets is perfectly rational.

>
>> Second issue is that small amount of money cease to be[1] real
>> (decimal) numbers and start to be infinitesimals (I often do not
>> collect 1\202¢ of change money).
>
> As I explained in my previous message, one cent is not infinitesimal.
> For an median-income American, it's worth two or three seconds of
> income. That's not very much, but it does mean that if you're handed
> a penny in change, it's worth keeping it, rolling it with other
> pennies once you get enough of them, and taking them to the bank
> the next time you're going there anyway.

Since I don't do cash deposits, I'd put it into my wallet together with
other coins. The increased weight and tear+wear of the wallet would
probably offset the value of the penny.

>
> Of course the smallest unit of currency in your country might be worth
> far less than a US penny, I don't know.

1 eurocent, it is actually a little more. Before euro, the smallest unit
in use was 50 hellers, which was even more (~1.67 eurocents). I would
have picked up 1 crown coin, but 50 hellers... I am not sure.

[Keith F. Lynch]

<rksh...@rosettacondot.com> wrote:
> Keith F. Lynch <k...@keithlynch.net> wrote:
>> <rksh...@rosettacondot.com> wrote:

> Yes...although inflation doesn't matter so much as long as incomes
> keep up with it.

But they haven't been, at least not for the past 30 to 40 years.

> $10,000 is about midpoint for yearly tuition at a state college.
> Private colleges can be three or more times that much. Add in
> books, which have had the same sort of inflation as tuition. If
> they're not lucky enough to live near the college, add in room and
> board. For the full four years, figure at least $60,000 for a state
> college, double or triple that for a private college.

Thanks. And one other thing I had momentarily forgotten: Many
colleges require all students to have medical insurance. I think this
is misguided, yet another example of the best being the enemy of the
good. It ensures that lots of people who could have been temporarily
uninsured graduates will instead be permanently uninsured non-graduates.

Perhaps it doesn't matter, since college degrees are positional goods.
If college degrees were more affordable, more people would have them,
and they'd be worth less.

As for textbooks, it's noteworthy that the same books that cost
hundreds of dollars new are available for a few pennies as soon as
they're supplanted with something slightly newer. So I've bought
far more college textbooks than most college graduates have, and
have probably paid less for all of them put together than students
typically pay for just one. And I've read and understood them. As
I've said, education is cheap, it's *credentials* that are expensive.

> Sure...but I don't find one penny every two seconds on the ground.
> I find one per week or so, which means that in two years I could buy
> a soda.

That's relevant to whether you should quit your day job to become a
full-time penny-picker-upper. It's not relevant to the comparison of
the value of the time it takes to pick up one penny with the value of
the penny, which is the same whether you pick up one of them each year
or ten million of them each year. (Except that if you encountered
millions of them every year, it would be worth investing in a tool
that makes picking them up faster and easier.)

> I can't quite stretch my imagination far enough to get excited about
> finding a nickel on the basis that doing so every few seconds would
> give me a million dollars within the next decade.

Nobody suggested getting excited.

[Keith F. Lynch]

<garabik-ne...@kassiopeia.juls.savba.sk> wrote:
> I certainly wouldn't want a job where I earn $12/h by continuous
> bending and straighting up, unless I'd be in difficulties and
> couldn't find any other employment.

Good point. Soon after I agreed to pick up sticks in my mother's
yard, I invested in a tool that lets me do so without bending over.
If pennies on the sidewalk were plentiful enough, I'd carry a tool
for that, too.

[rksh...@rosettacondot.com]

Keith F. Lynch <k...@keithlynch.net> wrote:
> <rksh...@rosettacondot.com> wrote:
>> Keith F. Lynch <k...@keithlynch.net> wrote:
>>> <rksh...@rosettacondot.com> wrote:
>
>> Yes...although inflation doesn't matter so much as long as incomes
>> keep up with it.
>
> But they haven't been, at least not for the past 30 to 40 years.

We could ban women from working outside the home, or institute a massive
penalty for multiple-earner households. That should reverse the trend.

>> $10,000 is about midpoint for yearly tuition at a state college.
>> Private colleges can be three or more times that much. Add in
>> books, which have had the same sort of inflation as tuition. If
>> they're not lucky enough to live near the college, add in room and
>> board. For the full four years, figure at least $60,000 for a state
>> college, double or triple that for a private college.
>
> Thanks. And one other thing I had momentarily forgotten: Many
> colleges require all students to have medical insurance. I think this
> is misguided, yet another example of the best being the enemy of the
> good. It ensures that lots of people who could have been temporarily
> uninsured graduates will instead be permanently uninsured non-graduates.
>
> Perhaps it doesn't matter, since college degrees are positional goods.
> If college degrees were more affordable, more people would have them,
> and they'd be worth less.
>
> As for textbooks, it's noteworthy that the same books that cost
> hundreds of dollars new are available for a few pennies as soon as
> they're supplanted with something slightly newer. So I've bought
> far more college textbooks than most college graduates have, and
> have probably paid less for all of them put together than students
> typically pay for just one. And I've read and understood them. As
> I've said, education is cheap, it's *credentials* that are expensive.

Education is cheap for people that can educate themselves. That's likely
a majority of our personal circles of friends. It's a vanishingly small
fraction of the students that my wife encounters.

>> Sure...but I don't find one penny every two seconds on the ground.
>> I find one per week or so, which means that in two years I could buy
>> a soda.
>
> That's relevant to whether you should quit your day job to become a
> full-time penny-picker-upper. It's not relevant to the comparison of
> the value of the time it takes to pick up one penny with the value of
> the penny, which is the same whether you pick up one of them each year
> or ten million of them each year. (Except that if you encountered
> millions of them every year, it would be worth investing in a tool
> that makes picking them up faster and easier.)

Except that what I was replying to originally was a discussion of the
marginal utility of money, and whether it was realistic to have it increase
relative to the amount of money involed. The time value of income wasn't
there.

>> I can't quite stretch my imagination far enough to get excited about
>> finding a nickel on the basis that doing so every few seconds would
>> give me a million dollars within the next decade.
>
> Nobody suggested getting excited.

Again, the original discussion was on the fundamental difference between
a small chance of winning a large sum ($100M) and a larger chance of
winning a smaller sum ($10K), and whether (or why) the benefit of the
former exceeded the benefit of the latter by greater than the simple
ratio of the amounts. Essentially, why having a tiny chance of winning a
huge payoff is more exciting than a much larger chance of a smaller
payoff.

[Keith F. Lynch]

<rksh...@rosettacondot.com> wrote:
> Keith F. Lynch <k...@keithlynch.net> wrote:
>> <rksh...@rosettacondot.com> wrote:
>>> Yes...although inflation doesn't matter so much as long as incomes
>>> keep up with it.

>> But they haven't been, at least not for the past 30 to 40 years.

> We could ban women from working outside the home, or institute a
> massive penalty for multiple-earner households. That should reverse
> the trend.

That definitely wouldn't reverse the trend, either for individual
income or for household income. Quite the opposite. Not unless you
don't count people without jobs when calculating median income. But
that would be bogus, since people without jobs still need to eat.

> Education is cheap for people that can educate themselves. That's
> likely a majority of our personal circles of friends. It's a
> vanishingly small fraction of the students that my wife encounters.

What exactly constitutes being educated by others? I believe it's
often possible to audit classes, or to watch videos of classroom
lectures. Does that count? How about lots of chats with educated
friends or relatives? How about reading a textbook and posting
questions to the net whenever there's something you don't understand?

> ... Essentially, why having a tiny chance of winning a huge payoff

> is more exciting than a much larger chance of a smaller payoff.

I don't agree that it is. Actually *winning* is exciting. Getting
yet another piece of junk mail from Publishers Clearinghouse is not,
no matter how big the numbers Ed McMahon claims you've definitely,
no fooling, actually won.

[David Harmon]

On Sun, 13 Nov 2011 18:48:57 +0000 (UTC) in rec.arts.sf.fandom,
"Keith F. Lynch" <k...@KeithLynch.net> wrote,

>ObFandom: Heinlein sold his first story for $50 so as to make a
>substantial dent in his mortgage. Damned inflation.

According to _Grumbles From the Grave_, Heinlein wrote the story in
response to a $50 contest in a pulp mag, but then decided he could
do better and sold it to Campbell for $70 instead.

[rksh...@rosettacondot.com]

Keith F. Lynch <k...@keithlynch.net> wrote:
> <rksh...@rosettacondot.com> wrote:
>> Keith F. Lynch <k...@keithlynch.net> wrote:
>>> <rksh...@rosettacondot.com> wrote:
>>>> Yes...although inflation doesn't matter so much as long as incomes
>>>> keep up with it.
>
>>> But they haven't been, at least not for the past 30 to 40 years.
>
>> We could ban women from working outside the home, or institute a
>> massive penalty for multiple-earner households. That should reverse
>> the trend.
>
> That definitely wouldn't reverse the trend, either for individual
> income or for household income. Quite the opposite. Not unless you
> don't count people without jobs when calculating median income. But
> that would be bogus, since people without jobs still need to eat.

I am, of course, being facetious, but the idea is to reduce the labor
pool and increase the value of the remaining workers. BTW, it sounds
like you're now arguing for household incomes as a measure. Median
individual income doesn't include people who aren't part of the labor
force. Children, retirees, people on disability, students, people who
choose not to work for whatever reason. Or do you want per capita
income? That's entirely different, although fairly meaningless without
knowing something about the distribution among households or families.

>> Education is cheap for people that can educate themselves. That's
>> likely a majority of our personal circles of friends. It's a
>> vanishingly small fraction of the students that my wife encounters.
>
> What exactly constitutes being educated by others? I believe it's
> often possible to audit classes, or to watch videos of classroom
> lectures. Does that count? How about lots of chats with educated
> friends or relatives? How about reading a textbook and posting
> questions to the net whenever there's something you don't understand?

I'm not sure if colleges allow unpaid audit anymore. Videos...maybe, if
they could get something from them. Most of them need a lot of question
and answer time. As far as educated friends and relatives...most of them
don't have any. Many of them are the first of their family to ever
attend college, some are among the few in their family to graduate from
high school. As far as posting questions...that might work, if someone
took the time to train them to use the net.

>> ... Essentially, why having a tiny chance of winning a huge payoff
>> is more exciting than a much larger chance of a smaller payoff.
>
> I don't agree that it is. Actually *winning* is exciting. Getting
> yet another piece of junk mail from Publishers Clearinghouse is not,
> no matter how big the numbers Ed McMahon claims you've definitely,
> no fooling, actually won.

I won a pound of bologna once, and that wasn't especially exciting. I
also won two dollars on a one dollar lottery ticket. That didn't do much
for me either. I *was* excited when I won a toy fire engine, but I was
also five years old at the time. It's the big payoff that's exciting in
the lottery, not the consolation prize. My friends who gamble don't talk
about the times they won ten dollars, they talk about the time they won
a thousand at the penny slots. The small prizes may give people the
illusion that they're not losing because they're "getting something
back", but there's not much thrill. I used to buy lottery tickets for my
dad and his friends...they didn't even want to know about the small
prizes, only if they'd hit the jackpot.

[Keith F. Lynch]

David Harmon <b...@example.invalid> wrote:
> "Keith F. Lynch" <k...@KeithLynch.net> wrote,
>> ObFandom: Heinlein sold his first story for $50 so as to make a
>> substantial dent in his mortgage. Damned inflation.

> According to _Grumbles From the Grave_, Heinlein wrote the story in
> response to a $50 contest in a pulp mag, but then decided he could
> do better and sold it to Campbell for $70 instead.

Thanks for the correction. It's been a while since I read that. And
I'm pretty sure I read it long before GFtG came out.

[Keith F. Lynch]

<rksh...@rosettacondot.com> wrote:
> Keith F. Lynch <k...@keithlynch.net> wrote:
>> That definitely wouldn't reverse the trend, either for individual
>> income or for household income. Quite the opposite. Not unless
>> you don't count people without jobs when calculating median income.
>> But that would be bogus, since people without jobs still need to
>> eat.

> I am, of course, being facetious, but the idea is to reduce the
> labor pool and increase the value of the remaining workers.

Well, that is the union approach. And it works *if* you're already a
union member -- until your employer goes out of business because they
were outcompeted by a firm overseas. In retrospect all those folks in
Detroit would have been better off making a good wage for a lifetime
than making a *great* wage for a few years, then making nothing.

> BTW, it sounds like you're now arguing for household incomes as
> a measure.

No, I'm pointing out that even if you use that perverse measure,
people would still be worse off on average if some people were
forbidden from working. Unless you arbitrarily choose not to count
those people.

It's happening now, thanks to E-Verify, a error-prone whitelist of
people allowed to work in the US, which more and more firms are being
pressured into adopting.

> Median individual income doesn't include people who aren't part of
> the labor force.

It should count everyone who needs to eat.

> Children, retirees, people on disability, students, people who
> choose not to work for whatever reason.

Someone is as financially well off supporting three children and two
retired parents as is someone else with the same income supporting
only himself? I don't think so.

> Or do you want per capita income?

That's not a good measure. Otherwise I could become much better off
by moving to Bill Gates's neighborhood, even if I personally never
earned another cent.

> As far as posting questions...that might work, if someone took the
> time to train them to use the net.

I think these days even most high school dropouts know how to use the
net. Isn't the proportion of the US population that uses the Internet
higher than the proportion who have graduated high school?

> I also won two dollars on a one dollar lottery ticket.

So did I, as I mentioned here at the time. (No, I didn't buy it.
It was given away in a promotion.)

> It's the big payoff that's exciting in the lottery, not the
> consolation prize.

Certainly. But only if you win. Having a lottery ticket that has a
one in ten million chance of winning a million dollars is about as
exciting as having a dime, i.e. not exciting at all.

It is of course very exciting if you win. But that's not very likely.

[David Friedman]

In article <j9p996$s5n$1...@reader1.panix.com>,

"Keith F. Lynch" <k...@KeithLynch.net> wrote:

> > $1 is almost infinitesimally small (in utilons), while $1M payoff is
> > huge - it is perfectly rational to buy the ticket.
>
> I think that under most circumstances, the utility of money decreases,
> not increases, with how much money you have. Your first dollar is
> worth a lot more to you than your millionth dollar.

I think that in this discussion, people tend to mix two different issues:

1. Lottery as entertainment. It's probably true that if the reason you
buy a lottery ticket is for the excitement, or the aid in daydreaming
about being wealthy, a ticket with one chance in a million of paying you
a million dollars is a better buy than a ticket with once chance in ten
of paying you ten dollars.

2. Lottery as a gamble, judged by the payoffs. This is where declining
marginal utility of income comes in, and, considered alone, makes a
lottery a bad deal for most people.

Of course, point 1 might be important enough for many people so that
combining the two points it was still worth their buying the low
probability/high payoff ticket.

[Doug Wickström]

On Sun, 13 Nov 2011 20:39:18 +0000 (UTC), "Keith F. Lynch"

<k...@KeithLynch.net> wrote:

>Doug Wickstr?nims...@comcast.net> wrote:
>> "Keith F. Lynch" <k...@KeithLynch.net> wrote:
>>> Doug Wickstr?nims...@comcast.net> wrote:
>>>> One of the counter-intuitives that contributes to this is the
>>>> unexpected _increase_ in deaths due to polio because of modern
>>>> sanitation. The cleaner we got, the more we tried to prevent
>>>> polio through public sanitation, the more virulent the disease
>>>> became.
>
>>> Nitpick: I think it's more a matter of that disease, like mumps,
>>> being far worse for adults than for children. With less sanitation,
>>> most people got it as children.
>
>> Well, duh.
>
>That's different from a disease become more virulent. Diseases
>generally evolve to become *less* virulent, so that they can be
>passed on before killing their carrier.

I was unclear. It was the _epidemics_ that were becoming more
virulent.

--
Doug Wickström

[garabik-ne...@kassiopeia.juls.savba.sk]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
> I think that in this discussion, people tend to mix two different issues:
>

Of course - deliberately.

> 1. Lottery as entertainment.
...

> 2. Lottery as a gamble, judged by the payoffs.

...

> Of course, point 1 might be important enough for many people so that
> combining the two points it was still worth their buying the low
> probability/high payoff ticket.

You have to combine those two. Very few people (if any) have their
financial affairs decoupled from their life. There is the positive
outcome of the lottery-as-entertainment, there is the expected payoff in
non monetary value if you win (which I imagine can have increasing
marginal utility) and the expected payoff of money. You take these
two payoffs, multiply by their weights and calculate with that (of
course, it is impossible to quantify the weights)

[rksh...@rosettacondot.com]

They need to eat, but they don't all need individual housing, transportation,
utilities, etc. IIRC housing is a big portion of your expenses. I can
assure you that five people living together don't spend five times as
much on housing.

>> Children, retirees, people on disability, students, people who
>> choose not to work for whatever reason.
>
> Someone is as financially well off supporting three children and two
> retired parents as is someone else with the same income supporting
> only himself? I don't think so.

Consider the opposite...three people with income who've been living apart
and decide to move in together. Do you really think their standard of
living doesn't change? I've done it...I can testify that it goes way,
way up. Let me add a few more "ways" in there. That's why you need to
know something about the distribution. Six people living on one income
is very different from one earner living alone and five with no income
at all.
We're back to the point we were before...the information you want
doesn't exist because the government, intrusive as it is, doesn't have
that kind of detailed insight into everyone's life, financial or
otherwise.

>> Or do you want per capita income?
>
> That's not a good measure. Otherwise I could become much better off
> by moving to Bill Gates's neighborhood, even if I personally never
> earned another cent.
>
>> As far as posting questions...that might work, if someone took the
>> time to train them to use the net.
>
> I think these days even most high school dropouts know how to use the
> net. Isn't the proportion of the US population that uses the Internet
> higher than the proportion who have graduated high school?

There are various degrees of "using the net". I doubt they have any
trouble finding Facebook but many of them have immense difficulty
finding information. They also have problems with calculators, meter
sticks, scales and a few other things that you and I probably take for
granted.

>> I also won two dollars on a one dollar lottery ticket.
>
> So did I, as I mentioned here at the time. (No, I didn't buy it.
> It was given away in a promotion.)
>
>> It's the big payoff that's exciting in the lottery, not the
>> consolation prize.
>
> Certainly. But only if you win. Having a lottery ticket that has a
> one in ten million chance of winning a million dollars is about as
> exciting as having a dime, i.e. not exciting at all.
>
> It is of course very exciting if you win. But that's not very likely.

I'm not a fan of the lottery, but the people I know that buy tickets seem
to find it very exciting...the anticipation, the discussion of what they'd
do with the money, waiting for the drawing, watching the numbers
appear. They get even more excited if the initial numbers match one of
their tickets. Of course that's followed by disappointment when they
fail to win, but they're certainly excited up until that point. And
there's always next week...

[cryptoguy]

On Nov 13, 3:28 pm, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article
> <238a1e24-1dac-4a65-981e-d72f4658d...@y12g2000vba.googlegroups.com>,

>
>  cryptoguy <treifam...@gmail.com> wrote:
> > I strongly suspect that if you asked your colleague for studies
> > showing that sanitation led to overall *higher* rates of mortality,
> > he'd come up empty as well. There aren't many research grants given to
> > investigate whether or not 'In most places, most of the time, the Sun
> > will rise in the East'  either.
>
> His point was that attempts to find statistical links between the
> obvious candidates for causes and reductions in mortality had been
> surprisingly unsuccessful, not that they hadn't been made.

If such studies have been made, I suspect that there are so many
contributing factors that, except in special cases like the Broad
Street event [1], its difficult to tease them apart.

Lets not confuse absence of evidence with evidence of absence.

[1] https://en.wikipedia.org/wiki/1854_Broad_Street_cholera_outbreak

pt

[Lowell Gilbert]

cryptoguy <treif...@gmail.com> writes:

> On Nov 13, 3:28 pm, David Friedman <d...@daviddfriedman.nopsam.com>
> wrote:
>> In article
>> <238a1e24-1dac-4a65-981e-d72f4658d...@y12g2000vba.googlegroups.com>,
>>
>>  cryptoguy <treifam...@gmail.com> wrote:
>> > I strongly suspect that if you asked your colleague for studies
>> > showing that sanitation led to overall *higher* rates of mortality,
>> > he'd come up empty as well. There aren't many research grants given to
>> > investigate whether or not 'In most places, most of the time, the Sun
>> > will rise in the East'  either.
>>
>> His point was that attempts to find statistical links between the
>> obvious candidates for causes and reductions in mortality had been
>> surprisingly unsuccessful, not that they hadn't been made.
>
> If such studies have been made, I suspect that there are so many
> contributing factors that, except in special cases like the Broad
> Street event [1], its difficult to tease them apart.

Not to mention the fact that conducting such a study with meaningful
controls would be unethical, so investigators tend to make do with
indirect measures.

> Lets not confuse absence of evidence with evidence of absence.

No one has claimed an absence of evidence, including the lecturer that
David Friendman quoted. His point was that it was *surprising* to him
that there wasn't any evidence of absence. David started a new thread to
relate the anecdote, so there is no context from which to extract
implied meanings.

On the other hand, I'd be interested in knowing more precisely what
statistics were surprising to miss. I doubt very much that there's a
lack of statistical reason to believe that building a sewer system in
Port au Prince would fail to reduce deaths from cholera, so my first
assumption would be that the claim was more specific than that --
particularly in light of the specific reference to Medicare data. I'm
perfectly willing to believe that access to sanitation is good enough
for nearly *all* Medicare recipients that it is not a major health
factor for them.

In terms of the one actual claim that David Friedman made in the initial
post, though, I'm sure he's right: a study of outcomes in a study of
Medicare patients provides absolutely no evidence that having medical
insurance improves life expectancy.

[David Friedman]

In article
<5ee2f9ea-ffd8-4e97...@r9g2000vbw.googlegroups.com>,

cryptoguy <treif...@gmail.com> wrote:

> On Nov 13, 3:28 pm, David Friedman <d...@daviddfriedman.nopsam.com>
> wrote:
> > In article
> > <238a1e24-1dac-4a65-981e-d72f4658d...@y12g2000vba.googlegroups.com>,
> >
> >  cryptoguy <treifam...@gmail.com> wrote:
> > > I strongly suspect that if you asked your colleague for studies
> > > showing that sanitation led to overall *higher* rates of mortality,
> > > he'd come up empty as well. There aren't many research grants given to
> > > investigate whether or not 'In most places, most of the time, the Sun
> > > will rise in the East'  either.
> >
> > His point was that attempts to find statistical links between the
> > obvious candidates for causes and reductions in mortality had been
> > surprisingly unsuccessful, not that they hadn't been made.
>
> If such studies have been made, I suspect that there are so many
> contributing factors that, except in special cases like the Broad
> Street event [1], its difficult to tease them apart.

That's what statistics, in particular multivariate statistics, is
for--teasing apart multiple causes.

>
> Lets not confuse absence of evidence with evidence of absence.

I said nothing at all about evidence of absence.

[David Friedman]

In article <448vnib...@be-well.ilk.org>,

Lowell Gilbert <lgus...@be-well.ilk.org> wrote:

> I doubt very much that there's a
> lack of statistical reason to believe that building a sewer system in
> Port au Prince would fail to reduce deaths from cholera, so my first
> assumption would be that the claim was more specific than that --
> particularly in light of the specific reference to Medicare data. I'm
> perfectly willing to believe that access to sanitation is good enough
> for nearly *all* Medicare recipients that it is not a major health
> factor for them.
>
> In terms of the one actual claim that David Friedman made in the initial
> post, though, I'm sure he's right: a study of outcomes in a study of
> Medicare patients provides absolutely no evidence that having medical
> insurance improves life expectancy.

The particular point with regard to medicare was that it provided
evidence that substantial variations in expenditure from one area to
another did not correspond to substantial variations in outcomes.

But that was only one example, and I don't think it was the basis for
the comment on the difficulty of relating sanitation to life expectancy.
My guess is that that would be based on considerably older data, or on
data from somewhere other than the US.

[David Dyer-Bennet]

"Keith F. Lynch" <k...@KeithLynch.net> writes:

> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
>> It provided evidence for how cholera was transmitted. It does not
>> tell us whether cleaner water reduced overall mortality and so helps
>> to explain the increase in life expectancy, which was the question
>> at issue.
>
> I'd think the past causes of death would give us plenty of information
> on the causes of the increase in life expectancy. For instance to the
> extent that a common 19th-century cause of death was being trampled by
> horses, we could credit today's smaller number of horses for today's
> greater longevity.

Though we should compare that number to the deaths in auto accidents.

[David Friedman]

In article <j9qir8$7or$1...@speranza.aioe.org>,

garabik-ne...@kassiopeia.juls.savba.sk wrote:

> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
> > I think that in this discussion, people tend to mix two different issues:
> >
>
> Of course - deliberately.
>
> > 1. Lottery as entertainment.
> ...
> > 2. Lottery as a gamble, judged by the payoffs.
> ...
>
> > Of course, point 1 might be important enough for many people so that
> > combining the two points it was still worth their buying the low
> > probability/high payoff ticket.
>
> You have to combine those two. Very few people (if any) have their
> financial affairs decoupled from their life. There is the positive
> outcome of the lottery-as-entertainment, there is the expected payoff in
> non monetary value if you win (which I imagine can have increasing
> marginal utility) and the expected payoff of money. You take these
> two payoffs, multiply by their weights and calculate with that (of
> course, it is impossible to quantify the weights)

I interpreted the original comment, which I was questioning, as a claim
about 2 making the purchase rational. You wrote:

"If you naïvely apply game theory to lottery, you of course do not buy
the ticket. However, you have to realize that you should measure the
costs and payoffs in utilons, not in dollars. $1 is almost

infinitesimally small (in utilons), while $1M payoff is huge - it is
perfectly rational to buy the ticket."

The point about utiles (the usual term in American economicese) rather
than dollars appeared to be a reference to a rising marginal utility of
income argument, not to the fact that, in addition to the expected
utility from the dollars there was also an additional utility from
daydreaming, excitement, or whatever.

[David Dyer-Bennet]

"Keith F. Lynch" <k...@KeithLynch.net> writes:

> Doug Wickström <nims...@comcast.net> wrote:
>> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

>>> Doug Wickström <nims...@comcast.net> wrote:
>>>> _Necessary_ health care, particularly preventives such as
>>>> immunizations, extends the life span of the general population by
>>>> decades. So does good nutrition, and healthy living conditions
>
> Decades for each, or decades in total?
>
>>> According to the colleague, actually finding results along
>>> those lines--I think he specifically mentioned sanitation--is
>>> surprisingly hard.
>
> There's no doubt that life expectancies have increased in the US. I'm
> pretty sure the main causes are knowledge of sanitation, nutrition,
> and safety, and the increased per capita wealth necessary for the
> majority of the population to be able to afford to act on that
> knowledge. (Now that per capita wealth is dropping in the US, I
> expect life expectancies to decrease. Knowledge is of little use
> to those who can't afford to apply it.)
>
> Medical science's main contribution was primarily to learn what
> habits are and aren't healthy. And secondly to provide inexpensive
> treatments such as vaccinations. What's really relevant to the
> question of whether medical insurance is worthwhile is how much
> benefit the average person gets from medical treatments that aren't
> affordable out of pocket. My contention is that that benefit *can*
> be large -- hence lots of anecdotes of people saved by million-dollar
> treatments -- but on average it's small to nonexistant.

Emergency medicine does pretty well. Look at the 18th century (say)
statistics for mortality from something as minor (today) as a broken
femur. And gunshot wounds.

Compare appendicitis mortality before and after anaethesia and sterile
surgical technique. Diabetes before insulin.

Really an awful lot of people I know have had one or more of those in
their lives to date.

I know multiple people who have gained years of life (and much-improved
quality of life) from a kidney transplant, and one with a heart
transplant (I didn't know her when she got it, but over two decades ago
I believe). And even more with bypass surgery or angioplasty. It looks
to me like modern medicine has made a BIG difference among people I know.

> And even with those anecodtes, you can seldom be sure that the person
> wouldn't have survived just fine with less treatment or none at all.

Generally you can't for individual cases, no.

>> One of the counter-intuitives that contributes to this is the
>> unexpected _increase_ in deaths due to polio because of modern
>> sanitation. The cleaner we got, the more we tried to prevent polio
>> through public sanitation, the more virulent the disease became.
>
> Nitpick: I think it's more a matter of that disease, like mumps,
> being far worse for adults than for children. With less sanitation,
> most people got it as children.

There's an awful lot we don't know about polio. I think you're right
for mumps and chicken pox, anyway.

[David Dyer-Bennet]

David Friedman <dd...@daviddfriedman.nopsam.com> writes:

> In article <j9o6m2$4kf$1...@speranza.aioe.org>,

> garabik-ne...@kassiopeia.juls.savba.sk wrote:
>
>> If you naïvely apply game theory to lottery, you of course do not buy
>> the ticket. However, you have to realize that you should measure the
>> costs and payoffs in utilons, not in dollars. $1 is almost
>> infinitesimally small (in utilons), while $1M payoff is huge - it is
>> perfectly rational to buy the ticket.
>

> Provided, of course, that you have increasing rather than decreasing
> marginal utility of income, which there is no reason to expect you to
> have. That is why the purchase of lottery tickets--especially the
> purchase of lottery tickets by people who also buy insurance, which is
> evidence of declining marginal utility of income--is a puzzle.

Taking a dollar a week out of my income stream is invisible, down in the
noise. (I actually spend considerably less than that on lottery
tickets).

Winning two hundred million dollars would completely and totally change
my entire existence, in lots and lots of different ways, most of which I
consider desirable (and I don' actually see a lot of papparazzi photos
of lottery winners, so they probably don't have to put up with many of
the worst aspects of celebrity).

So it feels to me like gambling "nothing" against "a wonderful
improvement".

(Yes, winning "$200,000,000" from Powerball means a little over half
that before taxes really, if you select the immediate payout option.
That's still enough to make that "wonderful improvement". Twenty
million would make a MUCH smaller difference; I could certainly finance
an improved lifestyle for the rest of my life without working, but I
couldn't do much at all of the "very rich plaything" stuff. And in fact
I don't buy many lottery tickets until the prize gets up there a bit.)

[David Dyer-Bennet]

David Friedman <dd...@daviddfriedman.nopsam.com> writes:

> In article <j9omuh$h86$1...@speranza.aioe.org>,

> garabik-ne...@kassiopeia.juls.savba.sk wrote:
>> Second issue is that small amount of money cease to be[1] real (decimal)

>> numbers and start to be infinitesimals (I often do not collect 1¢ of change
>> money).
>
> One cent is a very small amount of money, and it may not be worth the
> trouble of dealing with it. But the price of a lottery ticket is a sum
> that people routinely do care about; it will buy them things--a drink, a
> dinner--that they value. Do you also reject ten dollars in change?

A dollar will not buy me a drink, or a dinner, anywhere I've been in the
last month anyway.

>> The third issue is the "phase transition" if your income grows over
>> a certain sum. Winning $10k is nice, but does not change your life.
>> Winning $100M completely changes everything. For some (many) people the
>> expected outcome of winning $100M _vastly_ exceeds the raw mathematical
>> $1 * 1e8.
>
> Why would the benefit from a hundred million be more than ten thousand
> times that from ten thousand? If anything, the standard arguments for
> declining MUI would seem to apply. The ten thousand can be spent on
> things important to you. Most of the hundred million goes for a third
> mansion you rarely visit, a car five times as expensive and only a
> little better than the alternative car, ... .

The first big break-point is being able to stop working. That comes
fairly high -- if you need to fund the rest of your life including
health insurance and including protection from inflation, it takes quite
a bit.

Putting a price tag on that is hard, and different people put it in
different places (e.g. retire early and accept a lesser lifestyle, by
choice).

What kind of return can you reasonably expect, inflation-corrected, for
investments in the $5M to $100M range? 5%? 2%? (What you can get at
any given point in time isn't inherently the answer, you need a bunch of
data points for both return and inflation, AND a definition of inflation
that you like well enough.) So maybe you need $5 million invested just
to maintain a $100,000 income (more to include health insurance; and
most people with $100,000 incomes have health insurance). (Maybe
buying a basket of annuites works better? Also depends on whether you
have children you want to fund for life when you die, or what other
dependents and what level of support you want to give them).

Now, if you're making $40,000 now, going up to $100,000 is a big jump.
But lots of people make in the $1xx,000 these days, very low six
figures.

> One can, of course, draw a utility function to explain the behavior,
> convex up for the low range and concave up for the high, as per the
> Friedman and Savage article on the puzzle, but how do you provide a
> plausible justification for that shape?

Well, observing how people talk and behave around buying lottery tickets
sort of suggests they think it's worth it. Especially people who have
little expectation of actually winning (i.e. their expectation is in
line with reality).

[Cryptoengineer]

On Nov 14, 11:30 am, David Friedman <d...@daviddfriedman.nopsam.com>
wrote:
> In article
> <5ee2f9ea-ffd8-4e97-b9fd-0f9b1f500...@r9g2000vbw.googlegroups.com>,

>
>
>
>
>
>
>
>
>
>  cryptoguy <treifam...@gmail.com> wrote:
> > On Nov 13, 3:28 pm, David Friedman <d...@daviddfriedman.nopsam.com>
> > wrote:
> > > In article
> > > <238a1e24-1dac-4a65-981e-d72f4658d...@y12g2000vba.googlegroups.com>,
>
> > >  cryptoguy <treifam...@gmail.com> wrote:
> > > > I strongly suspect that if you asked your colleague for studies
> > > > showing that sanitation led to overall *higher* rates of mortality,
> > > > he'd come up empty as well. There aren't many research grants given to
> > > > investigate whether or not 'In most places, most of the time, the Sun
> > > > will rise in the East'  either.
>
> > > His point was that attempts to find statistical links between the
> > > obvious candidates for causes and reductions in mortality had been
> > > surprisingly unsuccessful, not that they hadn't been made.
>
> > If such studies have been made, I suspect that there are so many
> > contributing factors that, except in special cases like the Broad
> > Street event [1], its difficult to tease them apart.
>
> That's what statistics, in particular multivariate statistics, is
> for--teasing apart multiple causes.
>
>
>
> > Lets not confuse absence of evidence with evidence of absence.
>
> I said nothing at all about evidence of absence.

That wasn't clear - at least to me.

One thing I notice is that your colleague is using medicare records.
Medicare started only in 1965, by which time most Americans had
reasonably good sanitation. Even in the rural South, the work of the
Rockefeller Sanitary Commission for the Eradication of Hookworm had
provided enough sanitation to largely eliminate that parasite decades
earlier.

I'd suggest that most of the benefits of improved sanitation were
realized in the century prior to Medicare. Life expectancy has
increased by about 7 years since it was introduced, but rose by 28
years from 1850 to1950. What improvements to sanitation is your
colleague looking at? Does the data date back to the point where major
sanitation improvements where actually being put into effect? That
would be pre-Medicare data.

pt

[David Dyer-Bennet]

"Keith F. Lynch" <k...@KeithLynch.net> writes:

> <rksh...@rosettacondot.com> wrote:
>> David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

>>> I would add to this the pleasure of daydreaming about being rich,
>>> which is easier to do if there is some possible causal sequence you
>>> can imagine that would make you rich.
>
> In my case, I get that by envisioning my proving the Riemann
> Hypothesis, for which there's a million-dollar reward. This would
> bring not just wealth, but also fame and excellent job offers.
>
> And I think I have a better chance of proving it than I have of
> winning a million-dollar lottery. And I learn interesting math, and
> get good mental exercise, in the process.

Could well be; the odds of the lottery are small. On the other hand, I
don't play lotteries for less than 10 times your million dollars. A
million dollars wouldn't even let me quit my job.

>> $10K is enough to ... make a (probably negligible) dent in a
>> mortgage, ...

>
> ObFandom: Heinlein sold his first story for $50 so as to make a
> substantial dent in his mortgage. Damned inflation.

Yeah. I get that about camera and lens prices just over the 40 years
I've been buying the things.

>> They're not going to ... send their kids to a good college, ...
>
> With $10,000? What does tuition at a good college cost these days?

Oh, $30,000 a year to $60,000 a year, depending how "good" and such.
You can top $30k at a state school, and the school I went to is around
$58k this year if I'm remembering correctly.

>> In that sense I get more than 100 times the utility out of a dollar
>> than I get out of a penny...the penny is just a weight in my pocket
>> until it accumulates enough friends, the dollar will buy me a soda.

>
> That doesn't make sense to me, except in the sense that the time it
> takes to deal with a single penny may not be worth it. If I see a
> penny on the ground, it takes me about two seconds to pick it up, and
> about one second to put it in a roll. (I don't count the time it

> takes me to take the roll to the bank, since I only do so when I'm
> going to the bank anyway.) So that's about $12 per hour or $24,000
> per year, plus an exercise benefit. As such, it's worth it to me, but
> I'm not surprised that many people don't bother to pick up pennies. I

> *am* surprised that so many people don't bother to pick up nickels, as
> that's equivalent to a hefty $120,000 income, and few Americans other
> than government employees earn that much.

More evidence that people aren't economically rational actors.

I wonder if encountering a quarter super-glued to the ground once
changes people's likelihood of stooping for other coins later?

[David Dyer-Bennet]

"Keith F. Lynch" <k...@KeithLynch.net> writes:

> <rksh...@rosettacondot.com> wrote:

>> Education is cheap for people that can educate themselves. That's
>> likely a majority of our personal circles of friends. It's a
>> vanishingly small fraction of the students that my wife encounters.
>
> What exactly constitutes being educated by others? I believe it's
> often possible to audit classes, or to watch videos of classroom
> lectures. Does that count? How about lots of chats with educated
> friends or relatives? How about reading a textbook and posting
> questions to the net whenever there's something you don't understand?

I think the important distinction is getting to interact with the
teacher, and having them take your response into account in presenting
the material. So learning "by myself" is from fixed materials (whether
static, streaming, paper, or whatever).

Friends, colleagues, and netizens who will answer questions can help a
lot.

There is, of course, the additional issue of credentials involved in
most people's need for education.

[Lowell Gilbert]

David Friedman <dd...@daviddfriedman.nopsam.com> writes:

> In article <448vnib...@be-well.ilk.org>,
> Lowell Gilbert <lgus...@be-well.ilk.org> wrote:
>
>> I doubt very much that there's a
>> lack of statistical reason to believe that building a sewer system in
>> Port au Prince would fail to reduce deaths from cholera, so my first
>> assumption would be that the claim was more specific than that --
>> particularly in light of the specific reference to Medicare data. I'm
>> perfectly willing to believe that access to sanitation is good enough
>> for nearly *all* Medicare recipients that it is not a major health
>> factor for them.
>>
>> In terms of the one actual claim that David Friedman made in the initial
>> post, though, I'm sure he's right: a study of outcomes in a study of
>> Medicare patients provides absolutely no evidence that having medical
>> insurance improves life expectancy.
>
> The particular point with regard to medicare was that it provided
> evidence that substantial variations in expenditure from one area to
> another did not correspond to substantial variations in outcomes.
>
> But that was only one example, and I don't think it was the basis for
> the comment on the difficulty of relating sanitation to life expectancy.
> My guess is that that would be based on considerably older data, or on
> data from somewhere other than the US.

You're wandering very close to the silliness that Peter was
(prematurely, I thought at the time) accusing you of. The fact that
there is such a relationship is easy to establish. To an economist,
maybe it's hard to quantify in economic terms, but in scientific terms
it is quite clear that there is a strong positive correlation between
sanitation and life expectancy.

That is, unless you think this person you quote was denying one or both
of: (1) the germ theory of disease, or (2) the idea that many diseases
reduce life expectancy. But I tend to doubt that; I rather expect he was
trying to say something much more nuanced than your presentation of his
comments carried here from your discussion with him.

[David V. Loewe, Jr]

On Sat, 12 Nov 2011 19:22:23 +0000 (UTC), "Keith F. Lynch"
<k...@KeithLynch.net> wrote:

>Anyone who dies while under medical
>care was arguable killed by a medical mistake, in the sense that he
>probably would have lived a little longer had his dose of medication
>been slightly different, or had his surgery been performed on a
>different day or by a different surgeon.

This is so obviously wrong that I cannot fathom why no one has called
Keith on this.
--
"It's raining soup and we haven't built any soup bowls."
Dr. Jerry Pournelle

[David Friedman]

In article <ylfkhb26...@dd-b.net>,

David Dyer-Bennet <dd...@dd-b.net> wrote:

> > With $10,000? What does tuition at a good college cost these days?
>
> Oh, $30,000 a year to $60,000 a year, depending how "good" and such.
> You can top $30k at a state school, and the school I went to is around
> $58k this year if I'm remembering correctly.

It's worth noting that schools engage in extensive price discrimination
via scholarships and such. So while you are probably correct about the
"list price" at a good school, many, perhaps most, students would pay
less. My memory from when we were looking at schools for the kids is
that a number of the elite schools claimed that if they admit you they
will find a way of making it possible for you to pay to attend, however
low your parents' income.

Sometimes, of course, that involves jobs and loans--but I'm pretty sure
at least one and perhaps more claimed they would make it possible
without loans.

[David Friedman]

In article <ylfkobwe...@dd-b.net>,

I think the usual estimate for the real interest rate over the long term
is something like 2%. Of course, you can plan on using up the capital
sum over time or, perhaps more prudently, using it to buy an annuity.

...

[David Friedman]

In article <ddfr-2BB1B1.1...@news.giganews.com>,

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:

> In article <448vnib...@be-well.ilk.org>,
> Lowell Gilbert <lgus...@be-well.ilk.org> wrote:
>
> > I doubt very much that there's a
> > lack of statistical reason to believe that building a sewer system in
> > Port au Prince would fail to reduce deaths from cholera, so my first
> > assumption would be that the claim was more specific than that --
> > particularly in light of the specific reference to Medicare data. I'm
> > perfectly willing to believe that access to sanitation is good enough
> > for nearly *all* Medicare recipients that it is not a major health
> > factor for them.
> >
> > In terms of the one actual claim that David Friedman made in the initial
> > post, though, I'm sure he's right: a study of outcomes in a study of
> > Medicare patients provides absolutely no evidence that having medical
> > insurance improves life expectancy.
>
> The particular point with regard to medicare was that it provided
> evidence that substantial variations in expenditure from one area to
> another did not correspond to substantial variations in outcomes.
>
> But that was only one example, and I don't think it was the basis for
> the comment on the difficulty of relating sanitation to life expectancy.
> My guess is that that would be based on considerably older data, or on
> data from somewhere other than the US.

I asked my friend today. It sounds as though at least some of the
attempt to measure the effect of sanitation involved people in countries
such as the Philippines, with individual level data.

[Mike Benveniste]

On 11/14/2011 3:26 PM, David V. Loewe, Jr wrote:

>> Anyone who dies while under medical
>> care was arguable killed by a medical mistake, in the sense that he
>> probably would have lived a little longer had his dose of medication
>> been slightly different, or had his surgery been performed on a
>> different day or by a different surgeon.
>
> This is so obviously wrong that I cannot fathom why no one has called
> Keith on this.

Keith said "arguable", and he's making this argument. So his statement
is true by tautology. Anything, no matter how absurd, is arguable.

--
Mike Benveniste -- m...@murkyether.com (Clarification Required)
You don't have to sort of enhance reality. There is nothing
stranger than truth. -- Annie Leibovitz

[David Friedman]

In article <4439dqb...@be-well.ilk.org>,

What he was saying was that actual statistical studies, looking at
people in countries where the level of sanitation is much lower than
here and varies from place to place, with individual level data, didn't
show the correlation between sanitation and life expectancy that one
would expect to observe.

And, I gather, that analogous studies on other plausible explanations
for the increase in life expectancy over time, had also had more
difficulty than one would expect in supporting the connection.

[David Friedman]

In article
<d5e90dc3-53b7-4c79...@e15g2000vba.googlegroups.com>,

I wrote:

"According to him, it is very difficult to find evidence linking life
expectancy to any of the obvious inputs."

That describes the absence of evidence. It says nothing at all about
evidence of absence.

Perhaps you could point at what I wrote that you interpreted as making
the latter claim?

> One thing I notice is that your colleague is using medicare records.
> Medicare started only in 1965, by which time most Americans had
> reasonably good sanitation. Even in the rural South, the work of the
> Rockefeller Sanitary Commission for the Eradication of Hookworm had
> provided enough sanitation to largely eliminate that parasite decades
> earlier.

The medicare data was relevant to a particular absence of evidence--the
absence of evidence of a connection between expenditure and outcome.
Which is why I wrote:

"In particular, there is apparently a database
of some millions of people, coming out of (I think) medicare, which can
be analyzed to see how health care expenditure and mortality results
vary from one place to another. Expenditure per capita varies over about
a factor of two, with no detectable effect on outcome."

I said nothing about that database being used for evidence about the
effect of sanitation. That came up in a later post, which contained no
mention of the medicare database.

...

> What improvements to sanitation is your
> colleague looking at? Does the data date back to the point where major
> sanitation improvements where actually being put into effect? That
> would be pre-Medicare data.

It would be in the U.S. As I mentioned in a more recent post--almost
certainly after you made the one I'm responding to--I asked him about
the sanitation data today. At least some of it involved data from the
Philippines and (I think) Bangladesh, where sanitation is a current
issue. If I correctly understood his description, researchers had
individual level data on thousands of people, including information
about sanitation conditions where they were, and did not find the
expected link between sanitation and mortality.

I don't know if that was the only such attempt or merely the one he
described to me.

[Mike Benveniste]

On 11/13/2011 1:48 PM, Keith F. Lynch wrote:

> I surprised that so many people don't bother to pick up nickels, as

> that's equivalent to a hefty $120,000 income, and few Americans other
> than government employees earn that much.

For values of "few" well over 1 million, according to the tax policy
center. And that's just counting single people _not_ filing as head
of household.

[David Loewe, Jr.]

On Mon, 14 Nov 2011 17:28:50 -0500, Mike Benveniste <m...@murkyether.com>
wrote:

>On 11/14/2011 3:26 PM, David V. Loewe, Jr wrote:
>
>>> Anyone who dies while under medical
>>> care was arguable killed by a medical mistake, in the sense that he
>>> probably would have lived a little longer had his dose of medication
>>> been slightly different, or had his surgery been performed on a
>>> different day or by a different surgeon.
>>
>> This is so obviously wrong that I cannot fathom why no one has called
>> Keith on this.
>
>Keith said "arguable", and he's making this argument. So his statement
>is true by tautology. Anything, no matter how absurd, is arguable.

So argue it.

If someone comes in to the ER after a horrific car accident and dies
there instead of at the scene, they are "under medical care" and there
is nothing that could have been done. The idea that people, especially
trauma victims, who die "under medical care" are killed by "medical
mistakes" is absurdist theater.
--
"There is a time and a place for tact (and there are times when
tact is entirely misplaced)."
-Laurence VanCott Niven

[Cryptoengineer]

On Nov 14, 5:39 pm, Mike Benveniste <m...@murkyether.com> wrote:
> On 11/13/2011 1:48 PM, Keith F. Lynch wrote:
>
> > I surprised that so many people don't bother to pick up nickels, as
> > that's equivalent to a hefty $120,000 income, and few Americans other
> > than government employees earn that much.
>
> For values of "few" well over 1 million, according to the tax policy
> center.  And that's just counting single people _not_ filing as head
> of household.

In fact, a wide variety of professionals make that much, and more,
including experienced software engineers. But don't expect it as a
starting salary straight out of school.

pt

[Jay E. Morris]

And government employees too after about twenty-five years of service,
if they rise high enough. At least, in the government I work for.

[Lowell Gilbert]

"David V. Loewe, Jr" <dave...@charter.net> writes:

> On Sat, 12 Nov 2011 19:22:23 +0000 (UTC), "Keith F. Lynch"
> <k...@KeithLynch.net> wrote:
>
>>Anyone who dies while under medical
>>care was arguable killed by a medical mistake, in the sense that he
>>probably would have lived a little longer had his dose of medication
>>been slightly different, or had his surgery been performed on a
>>different day or by a different surgeon.
>
> This is so obviously wrong that I cannot fathom why no one has called
> Keith on this.

Because there are a lot of possible definitions of death by medical
mistake, and more than a couple of them are useful. I personally think
Keith's is silly, but that's trivial to insisting that there's one
correct definition. [I don't think most nations even agree on a solid
definition.]

[karl.j...@gmail.com]

On Nov 12, 3:13 pm, "Keith F. Lynch" <k...@KeithLynch.net> wrote:
> karl.johan...@gmail.com <karl.johan...@gmail.com> wrote:
> > Lots of things cost less if bought in bulk.
>
> Medical insurance is already bought in bulk.  Medical insurance is
> expensive because medical care is expensive.  Not all medical care.
> Not most medical care.  But enough medical care to push the average
> way up for everyone.  If there's a one percent chance you'll need
> a million-dollar operation, insuring against it will cost at least
> $10,000 -- even if that million-dollar operation only has a small
> chance of saving your life.
>
> What's needed are ways to bring the cost of medical care down.

Preventative care is one way. Laws protecting people from dangerous
stuff is one form of preventative care. When the government has a
stake in medical costs, there is an incentive for the government to
pass protective laws, besides ethics. Laws protecting non-smokers will
likely have an effect on life expectancies and on medical costs.

> Robotic surgery, perhaps.  (The term "robotic surgery" has
> unfortunately already been appropriated to mean something that *does*
> involve the usual expensive team of surgeons and support staff, so I
> suppose we'll have to call true robotic surgery something else.)

Another way is to get better information on what medical procedures
are actually useful. Not removing tonsils as readily as they did in
the past, for example.

Karl Johanson

[Keith F. Lynch]

Mike Benveniste <m...@murkyether.com> wrote:
> Keith F. Lynch wrote:
>> I surprised that so many people don't bother to pick up nickels, as
>> that's equivalent to a hefty $120,000 income, and few Americans other
>> than government employees earn that much.

> For values of "few" well over 1 million, according to the tax policy
> center. And that's just counting single people _not_ filing as head
> of household.

One million is one third of one percent of the US population. And how
many of them are *not* government employees?
--
Keith F. Lynch - http://keithlynch.net/
Please see http://keithlynch.net/email.html before emailing me.

[Lowell Gilbert]

David Friedman <dd...@daviddfriedman.nopsam.com> writes:

> In article <4439dqb...@be-well.ilk.org>,
> Lowell Gilbert <lgus...@be-well.ilk.org> wrote:
>
>> You're wandering very close to the silliness that Peter was
>> (prematurely, I thought at the time) accusing you of. The fact that
>> there is such a relationship is easy to establish. To an economist,
>> maybe it's hard to quantify in economic terms, but in scientific terms
>> it is quite clear that there is a strong positive correlation between
>> sanitation and life expectancy.
>>
>> That is, unless you think this person you quote was denying one or both
>> of: (1) the germ theory of disease, or (2) the idea that many diseases
>> reduce life expectancy. But I tend to doubt that; I rather expect he was
>> trying to say something much more nuanced than your presentation of his
>> comments carried here from your discussion with him.
>
> What he was saying was that actual statistical studies, looking at
> people in countries where the level of sanitation is much lower than
> here and varies from place to place, with individual level data, didn't
> show the correlation between sanitation and life expectancy that one
> would expect to observe.

"expect to observe?" That's hard to measure in the absence of other
factors, so for it to be a meaningful statement, I assume you mean it
trivial or less.

> And, I gather, that analogous studies on other plausible explanations
> for the increase in life expectancy over time, had also had more
> difficulty than one would expect in supporting the connection.

In an academic sense, that's interesting. But to me, it's just a
question of what's wrong with how the economists are measuring things.
Because, as I said earlier, the only other possibilities are that the
germ theory of disease is wrong, or that disease doesn't shorten life
expectancy.

But to get back to what I keep emphasizing, I think that you're probably
not presenting us with what your colleague intended to get at (given
that, as I recall, you were relating a dinner conversation, this doesn't
imply any failure on your part). There are probably multiple masters'
theses (if not a PhD colloquium) on why specific cross-cultural
statistical studies don't show life expectancy increases with
improvements in sanitation. But epidemiology is a fairly well
established experimental field, and it indicates rather more strongly
that sanitation does, in fact, decrease multiple classes of disease. If
population studies fail to back that up, then either there are
correlating factors that are confounding those studies (which would be
useful to explore, to avoid being fooled by the same correlations in the
future), or the studies are just plain wrong.

[Keith F. Lynch]

Mike Benveniste <m...@murkyether.com> wrote:
> David V. Loewe, Jr wrote:

>> "Keith F. Lynch" <k...@KeithLynch.net> wrote:
>>> Anyone who dies while under medical care was arguable killed by a
>>> medical mistake, in the sense that he probably would have lived a
>>> little longer had his dose of medication been slightly different,
>>> or had his surgery been performed on a different day or by a
>>> different surgeon.

>> This is so obviously wrong that I cannot fathom why no one has
>> called Keith on this.

> Keith said "arguable", and he's making this argument. So his
> statement is true by tautology. Anything, no matter how absurd,
> is arguable.

I'm trying to figure out what's meant by death by medical error.

Presumably you'd agree that if someone died shortly after receiving 40
milligrams of a drug intended to save him, that he probbaly either:

* Would have lived slightly longer had he been given 41 mg, and
slightly shorter had he been given 39 mg.

or

* Would have lived slightly longer had he been given 39 mg, and
slightly shorter had he been given 41 mg.

A person lives longest when he's given a precisely optimal treatment.
But since doctors aren't omniscient, treatments are never precisely
optimal.

As I said, it seems to be there's a continuum between medical "errors"
such as this, and medical errors that are so egregious as to be
shocking to the conscience. There doesn't seem to be any sharp
dividing line. So I have no idea what 100,000 deaths due to medical
errors is intended to mean.

[Keith F. Lynch]

David Friedman <dd...@daviddfriedman.nopsam.com> wrote:
> What he was saying was that actual statistical studies, looking at
> people in countries where the level of sanitation is much lower than
> here and varies from place to place, with individual level data,
> didn't show the correlation between sanitation and life expectancy
> that one would expect to observe.

I suspect it only makes a significant difference when there's an
actual epidemic. For instance in the "pump handle" case, had there
been no initial cholera case, drinking the sewage-contaminated well
water probably would have done little harm.

[Keith F. Lynch]

David Dyer-Bennet <dd...@dd-b.net> wrote:
> There is, of course, the additional issue of credentials involved in
> most people's need for education.

Yes. That was the point I was making: That education and
credentialing are two very different things which have little to do
with each other. Education, unlike credentialing, is inexpensive.
Unfortunately, it's also of little value in today's America.

[Keith F. Lynch]

David Dyer-Bennet <dd...@dd-b.net> wrote:
> A million dollars wouldn't even let me quit my job.

You must either be very young or have an expensive lifestyle. I could
comfortably retire today on half a million dollars. If nothing else,
it would let me take a very long vacation and be very choosy about
what jobs I'll take.

That does assume that inflation doesn't exceed the after-tax interest
I can get on that money. And that I won't have any major emergencies.

However, as I mentioned, one advantage of getting money by solving a
hard math problem rather than by winning a lottery is that it would
get me good job offers.

> I wonder if encountering a quarter super-glued to the ground once
> changes people's likelihood of stooping for other coins later?

I can't see why it would unless they have reason to think that a large
proportion of coins on the ground are glued down.

(The next time I planned to be passing near the glued quarter, I'd
be sure to bring a screwdriver to pry it up. But I wouldn't make a
special trip for the purpose.)

[Keith F. Lynch]

David Dyer-Bennet <dd...@dd-b.net> wrote:
> A dollar will not buy me a drink, or a dinner, anywhere I've been in
> the last month anyway.

What sort of drink? Water is effectively free. And last I checked,
canned soft drinks were still under a dollar. As for dinner, it
depends on how hungry you are and how choosy. Vegetables are fairly
cheap, as are some junk foods. It's an interesting question how many
calories you can get for a dollar.

> What kind of return can you reasonably expect, inflation-corrected,
> for investments in the $5M to $100M range? 5%? 2%?

If we get hyperinflation, it's likely to be very negative. Your rent
might be $100 million per month a decade from now. I'd put a lot of
my money into gold just to be safe.

> Maybe buying a basket of annuites works better?

Unless the company that owns them goes bankrupt.

> Also depends on whether you have children you want to fund for life

> when you die, ...

Conventional wisdom is that giving your children enough money that
they never need to work is a bad idea.

[Keith F. Lynch]

David Dyer-Bennet <dd...@dd-b.net> wrote:
> Winning two hundred million dollars would completely and totally
> change my entire existence, in lots and lots of different ways, ...

Would you continue to post here?

[karl.j...@gmail.com]

On Nov 13, 12:28 pm, "Keith F. Lynch" <k...@KeithLynch.net> wrote:
> Federal prisoners are paid 30
> cents per hour.

I like the idea of paying prisoners minimum wage, with a significant
portion of it donated to charity. When the prisoner gets out, they get
tax receipts for the donated money, that can be written off of
legitimate income. I'd spread the receipts over a few years as well.
Provides an additional incentive for finding a legitimate job,
provides cash to charities and lets the prisoners know that their work
is valued more than $.30 an hour.

Not a perfect idea by any means, but I think it could be an
improvement.

The sign factory I used to work at made use of prison labour, for
cutting sign blanks and making sign stands. I only talked to a couple
of the prisoners. One mentioned they appreciated the work. The other
tried over and over (and failed) to talk me into smuggling pot into
the prison for him.

Karl Johanson

[Jay E. Morris]

On 11/14/2011 8:25 PM, Keith F. Lynch wrote:
> Mike Benveniste<m...@murkyether.com> wrote:
>> Keith F. Lynch wrote:
>>> I surprised that so many people don't bother to pick up nickels, as
>>> that's equivalent to a hefty $120,000 income, and few Americans other
>>> than government employees earn that much.
>
>> For values of "few" well over 1 million, according to the tax policy
>> center. And that's just counting single people _not_ filing as head
>> of household.
>
> One million is one third of one percent of the US population. And how
> many of them are *not* government employees?

You know, you're really an ass about this government employee thing. As
I've said before, my organization is made up probably 2/3 engineers,
scientists, and the like but if 1/3 of the organization makes six
figures I'd be greatly surprised. I don't have access to how much they
make, but I do have access to what level they're at (mostly GS-11, 12, 13).

The organization next to us is 2/3 under GS-11, mostly 7s and 9s so
they're making $40,000-$60,000.

A GS employee would have to be GS-14 (second from the top) with 8-10
years in grade to break $120,000. Even with DC locality pay one would
still have to be GS-14. Of the 800 people in our org, there's around 25
GS-14s and a dozen GS-15s. One SES.

You, in your very specific, very limited, case may run into a greater
number of higher ranking employees, or possibly those not GS but some
higher paying scale. Perhaps there's an over abundance of Senior
Executive Service employees there. I'm quite sure there a lot more
Executive Service (150,000-199,000) employees there than any place else.
But at the most only 20% of all government employees are in the DC area.

And if you continue to use the all inclusive term "government employees"
then I'm going to start adding in the wage grade, law enforcement, park
service, etc schedules and really start pulling the average down.

Anybody interested in the pay scales they are here:
http://www.opm.gov/oca/11tables/index.asp