. 20
( 32 .)


Put differently, in the absence of re-insurance, enterprises and
households in China have to put aside more money, or the physical
resources that money represents, to guard against risks than if part
of those risks were carried by a company located in Switzerland or
the United States, which simultaneously re-insures enterprises and
homes in Argentina, Egypt, Australia, Denmark, and Fiji, as well
as in China. The amount of equipment and supplies needed to be
stockpiled to cope with the risk of floods along the Yangtze
River”tents to house people forced from their homes, canned or
preserved food to feed them, equipment to repair the damage and
rebuild”would be much less if instantaneous electronic transfers
of money from Swiss Re could purchase equipment and supplies to
be flown in from Japan in a matter of hours, if and when they are
As the London magazine The Economist put it, "China is wast-
ing capital that could be used for other things" when it puts aside
larger amounts of capital to guard against floods, fires and other
catastrophes than it would if it were re-insured by international re-
insurers. Insurance, re-insurance and other risk-sharing activities
mean that fewer idle resources are necessary around the world to
provide a given level of protection against risks. That in turn
means that the resources no longer kept idle can be used for other
things, such as raising living standards, supporting medical re-
search, or promoting greater economic development.
Lower risks mean lower costs, and market competition among
insurance and re-insurance companies means that those lower
costs must be passed on to customers in lower premiums, in order
to get their business.2 While geographic distributions of risks pro-

Tor example, Business Week magazine reported in its March 4,2002 issue: "Premiums for reinsur-
ance industrywide had already gone into a tailspin as strong profits attracted more competitors who
slashed prices to gain market share." (page 77).
Risky Business 149

vide one reason for re-insurance, that is not the only reason. For
example, more than two-thirds of American life insurance compa-
nies also have re-insurance.
As noted in the first chapter, political incentives in response to a
natural disaster like the cyclone that struck India in 1999 may be
to save face for the Indian government authorities, rather than to
rescue the victims as quickly as possible by seeking help from
other nations or from international agencies. In a competitive
marketplace, however, a private insurance company's reputation
for quick response is an enormously valuable asset. It was the
Swiss Reinsurance Company's swift response in getting money to
the victims of the devastating San Francisco earthquake and fire of
1906 which gained them international renown and promoted the
spread of their business around the world.
Although paying off the huge claims resulting from the San
Francisco disaster took half the company's annual income from
premiums, the subsequent growth in their business as a result led
eventually to Swiss Re becoming the world's largest re-insurance
company between the two World Wars.

Government Regulation of Insurance
State regulation of insurance companies in the United States
adds another dimension to the industry. The politics of insurance
regulation can be more complicated than the economics because
the very process by which insurance companies reduce risk is of-
ten under political attack. For example, some consider it "unfair"
that a safe driver with an unblemished record is charged higher
insurance premiums because he happens to live in a particular
neighborhood where others have more accidents. However, the
risks to an automobile do not depend solely on its owner or

driver, but also on others who create dangers of accidents, theft, or
Similar issues arise when women live longer than men, mak-
ing it more costly for an insurance company to offer an annuity
to a woman than to a man, or when blacks have shorter lifespans
than whites, making if more costly to offer life insurance to
someone of a given age who is black. Higher automobile acci-
dent rates in particular age brackets3 may also lead to higher
premiums for those in these brackets, even when the individual
has an unblemished safety record. State laws and regulations
may forbid insurance companies from making some or all of
these distinctions among people when charging premiums to
different policy holders. However, this sense of fairness to indi-
viduals can raise risks over all, leading to higher premiums to
policy-holders in general.
Two sets of motorists are especially affected by state regulation
of automobile insurance rates”very safe drivers and very danger-
ous drivers. In a free market, the cost of car insurance to the former
would be far less than the cost to the latter. However, under politi-
cal definitions of "fairness," safe drivers can end up subsidizing un-
safe drivers, especially when policies are set without thinking
beyond stage one.
When the price of auto insurance is set by state officials, often
they will not let the price rise to as high a level as it would reach in
a free market, when insuring drivers with a record of accidents or
serious traffic violations. Since insurance companies will not want
to insure such drivers when the premiums they pay are unlikely to
cover all the costs caused by these kinds of drivers, a common

Although fatality rates from motor vehicle deaths are highest for drivers aged 16 to 19, the de-
clining fatality rates with age end at around age 50 and then rise again, with drivers aged 80 to 84
having nearly as high a fatality rate from motor vehicle deaths as teenagers.
Risky Business 151

political "solution" is to create a "high-risk" pool of drivers unable
to get insurance through regular channels. Insurance companies
are then forced to share these high-risk drivers among themselves
and rates in general are then set at levels which will enable the in-
surers to cover the losses created by all their drivers, including
those in the high-risk pools.
What this means is that other drivers are subsidizing high-risk
drivers. Looking beyond stage one, what this also means is that
more pedestrians and motorists are likely to suffer injuries or
death because more high-risk drivers can afford to be on the
roads and highways than could do so if auto insurance rates were
allowed to rise to the very high rates required to compensate for
the damage done by reckless drivers. Among the other conse-
quences is that, as rates continue to be held down by government
regulators, even in the face of rising costs, a larger and larger
percentage of drivers will be unable to obtain insurance through
the regular channels and therefore end up in the high-risk pool.
In New Jersey, for example, only 12 percent of drivers were in
the high-risk pool in the early 1970s but, a decade later, nearly
half were.
New Jersey's experience with automobile insurance regulation
has not been unique. Nor has automobile insurance been unique in
having political criteria over-rule economic criteria in setting insur-
ance rates. The net effect, with other kinds of insurance as well, is
to have those with lower risks subsidize those with higher risks,
leading to more risky behavior whose consequences can cause both
financial losses and losses of life and limb to third parties.
This is not to say that there is never a useful role for government
to play when it comes to insurance. Two main problems affect
insurance companies in a free market: One problem already dis-
cussed earlier is that people who are insured may engage in more
risky behavior as a result. This is called "moral hazard." Another is

that, when some people choose to be insured for some things and
others choose not to be insured, those at the greatest risk are more
likely to choose to be insured, so that general statistics on the risks
to the population at large are misleading as to the risks of those
who choose to buy insurance. That is called "adverse selection."
Government regulation of risky behavior”laws against storing
flammable materials in homes or driving under the influence of al-
cohol, for example”can reduce the risks of moral hazard and laws
requiring everyone who drives to have automobile insurance can
solve the problem of adverse selection. As in other situations, how-
ever, the fact that the government can sometimes improve on the
situation that would exist in a free market does not mean that it
will in fact restrict its activities to such useful things. When the
counterproductive interventions of government in the market are
added to its beneficial interventions, it is by no means always clear
what the net balance will be.


Safety might seem to be something that you cannot get too much
of. Yet everything we do in our everyday lives belies that conclu-
sion. Often what we do makes more sense than what we say. Re-
ducing risks has costs”some of which we are willing to pay and
some of which we are not. As already noted, the very paper on
which these words are written has risks. So does the water you
drink, the food you eat, the oxygen you breathe. Sunshine causes
skin cancer. It is not safe to exercise”or not to exercise. The only
meaningful question, in all these cases, is the degree of risk, com-
pared to alternatives, and the costs of reducing those risks by given
Not all costs are money costs. For many people, the cost of re-
ducing risks would be giving up the enjoyment they get from
Risky Business 153

skiing, boating, rock climbing, skateboarding, and other risky
activities. In fact, ultimately there are only risky activities, since
nothing is 100 percent safe. Yet no one suggests that we retire into
passive inactivity”which has its own risks, and which would also
have to be away from sunlight if we were consistent, since sunlight
increases the risk of skin cancer. On the other hand, most of us are
not prepared to walk into a lion's cage or to take a short cut by
walking across a freeway during the rush hour. In other words,
there are risks we are willing to take and risks we are not willing to
take, varying somewhat from person to person, but involving a
weighing of benefits and costs in any event. We do not drive to
work in tanks, even though tanks would be safer than automo-
biles, because we are not prepared to pay the costs in either money
or commuting time.
However much we may agree with sweeping rhetoric about
safety, or even vote for those who use such rhetoric, nevertheless
when faced with choices in our own lives we weigh incremental
safety against incremental costs. We may consider it worthwhile to
avoid one chance in six of getting killed from playing Russian
roulette, but not worth it to pay a thousand dollars to avoid one
chance in six million of getting killed by some fluke occurrence.
Indeed, if the cost of avoiding one chance in six million is merely
an inconvenience, some may still refuse to pay it. In short, even
those who talk about safety in categorical terms”"if it saves just
one life, it is worth whatever it costs"”actually behave in their
own lives as if safety is an incremental decision, based on weighing
costs against benefits, not a categorical decision.
One of the costs of any given kind of safety may be an increase in
other kinds of dangers. It is always possible to make subways safer
by having the trains go more slowly, increasing the distances be-
tween trains, and having fewer cars per train, in order to reduce the
train's weight and hence reduce the distance required to bring the

train to a stop. However, all these things reduce the number of pas-
sengers who can be carried during rush hours and”since people
have to get to work somehow”forces those passengers to try other
means of transportation, most of which involve greater risks of death
than subway trains do. In short, you can always reduce the risks in
subways by policies which increase the risks elsewhere.
During the Second World War, Japanese fighter pilots usually
chose not to wear parachutes when going into aerial combat. As
one of these pilots explained after the war, although every pilot was
provided with a parachute, "the decision to fly without them was
our own" and was made because parachutes "hamstring our cockpit
movements in a battle," when split-second responses could be mat-
ters of life and death, and it "was difficult to move our arms and
legs quickly when encumbered by chute straps." In other words,
wearing a parachute increased the chances that you would get shot
down.4 Increasing one kind of safety can increase other risks, just
as with subway trains and other things.
Here we come to a crucial distinction in decision-making
processes”the distinction between individuals making decisions
for themselves and third-parties making decisions for others. Sub-
way passengers always have the option of using other means of
getting to work, but they are very unlikely to hold subways to a
standard of 100 percent safety and go elsewhere if this standard is
not met. But what of those who are making decisions for others? If
there has recently been a tragic and highly publicized subway acci-
dent with a number of fatalities, political outcries for more subway
safety may well cause the authorities to order the trains slowed
down, fewer cars to be attached to each train, and greater distances
maintained between trains.

Another reason was that Japanese pilots were usually fighting air battles over enemy territory
and bailing out would mean being captured by the enemy, which was "unthinkable" to a Japanese
Risky Business 155

Moreover, after the passage of time shows that these policies
have reduced accidents and injuries in the subways, officials who
instituted such policies are unlikely to be shy about claiming
credit. Nor are most voters likely to inquire about the total number
of injuries and deaths during rush hour on all modes of trans-
portation put together. That would require thinking beyond stage
one. In short, third-party decision-making, based on categorical
reasoning and one-stage thinking, often succeeds in the political
arena, even though individual decision-making for oneself is more
likely to involve an incremental weighing of benefits against costs.
Those skilled in rhetoric can easily ignore hidden costs which
those faced with decisions for themselves are more likely to take
into account.
There are other situations in which caution”at stage one”can
turn out to be dangerous in the long run. Nor are these situations
confined to economics. In the American Civil War, for example,
General George McClelland's cautious use of the Union army has
often been blamed for more fatalities than if he had been more ag-
gressive and not let defeated Confederate troops escape to fight


. 20
( 32 .)