Excerpt from Free to Choose:
"Almost everything we do has some third-party effects, how-
ever small and however remote. In consequence, Adam Smith's
third duty may at first blush appear to justify almost any proposed
government measure. But there is a fallacy. Government measures
also have third-party effects. "Government failure" no less than
"market failure" arises from "external" or "neighborhood" effects.
And if such effects are important for a market transaction, they
are likely also to be important for government measures intended
to correct the "market failure." The primary source of significant
third-party effects of private actions is the difficulty of identifying
the external costs or benefits. When it is easy to identify who is
hurt or who is benefited, and by how much, it is fairly straight-
forward to replace involuntary by voluntary exchange, or at least
to require individual compensation. If your car hits someone
else's because of your negligence, you can be made to pay him
for damages even though the exchange was involuntary. If it
were easy to know whose collars were going to be dirtied, it would
be possible for you to compensate the people affected, or alterna-
tively, for them to pay you to pour out less smoke.
If it is difficult for private parties to identify who imposes costs
or benefits on whom, it is difficult for government to do so. As a
result a government attempt to rectify the situation may very
well end up making matters worse rather than better—imposing
costs on innocent third parties or conferring benefits on lucky
bystanders. To finance its activities it must collect taxes, which
themselves affect what the taxpayers do—still another third-
party effect. In addition, every accretion of government power for
whatever purpose increases the danger that government, instead
of serving the great majority of its citizens, will become a means
whereby some of its citizens can take advantage of others. Every
government measure bears, as it were, a smokestack on its back.
Voluntary arrangements can allow for third-party effects to a
much greater extent than may at first appear. To take a trivial
example, tipping at restaurants is a social custom that leads you
to assure better service for people you may not know or ever
meet and, in return, be assured better service by the actions of
still another group of anonymous third parties. Nonetheless, third-
party effects of private actions do occur that are sufficiently im-
portant to justify government action. The lesson to be drawn from
the misuse of Smith's third duty is not that government interven-
tion is never justified, but rather that the burden of proof should
be on its proponents. We should develop the practice of examining
both the benefits and the costs of proposed government interven-
tions and require a very clear balance of benefits over costs be-
fore adopting them. This course of action is recommended not
only by the difficulty of assessing the hidden costs of government
intervention but also by another consideration. Experience shows
that once government undertakes an activity, it is seldom ter-
minated. The activity may not live up to expectation but that is
more likely to lead to its expansion, to its being granted a larger
budget, than to its curtailment or abolition."
"The impact of global warming on temperature-induced human mortality has long been a concern, where it has been hypothesized that rising temperatures will lead to an increase in the number of deaths due to an increase in the frequency and intensity of heat waves. Others claim that rising temperatures will also reduce the number of deaths at the cold end of the temperature spectrum (fewer and less severe cold spells), resulting in possibly no net change or even fewer total temperature-related deaths in the future.
The largest study—by far—on temperature-related mortality was published by Gasparrini et al. in the journal Lancet in 2015. They examined over 74 million (!) deaths worldwide from 1985 to 2012 and found that the ratio of cold-related to heat-related deaths was a whopping 17 to 1. Moreover, the temperature percentile for minimum mortality was around the 60th in the tropics and “80–90th” in the temperate zones. Based upon real-world data, it is obvious that global warming is going to directly prevent a large number of deaths.
One of us (Michaels) co-authored a peer-reviewed literature article showing that as heat waves become more frequent, heat-related deaths decrease because of adaptation. Given that our cities are heating up on their own—without needing a push from greenhouse gases—under our hypothesis, heat-related mortality should be dropping, which it is.
But what about morbidity (sickness), as opposed to mortality? For that, we should be looking at emergency room visits, where people go because they are really feeling crummy or have a physical injury. Turns out everyone has been looking at death, but few at debilitation.
Now comes a new paper from Zhao et al. (2017). They examined the association between daily mean ambient temperature and emergency department visits in twelve Chinese cities over the period 2011–2014. Two were in the cool north, six from the central region and four in the hot and humid south.
As represented by the pooled national data as shown in the figure below, the relative risk of emergency department visits increases as temperatures become both warm and cold. However, the risk is far greater for cold temperatures, where the cumulative relative risk is 1.80 (nearly twice as likely compared to average temperatures) versus a much smaller 1.15 (a 15% increase in prevalence) that was associated with hot temperatures. Additionally, Zhao et al. determined that the effects of cold spells on emergency department visits were much more persistent, lasting a full 30 days compared to the more acute, but short lived, effects of warm spells that lasted a mere three days, or one-tenth of the time.
Figure 1. Pooled national level cumulative association between temperature and emergency department visits over a lag of 0–32 days during 2011–2014. Adapted from Zhao et al. (2017).
Other important findings included the observation that the temperature percentile associated with the least amount of emergency department visits was 64. Given that the average climate in China varies from tropical to pretty darned cold, it’s probably somewhat more tropical than temperate. Their optimum temperature is very consistent with what was found in the Gasparrini et al. study. At 14 percentage points higher than 50, this fact (along with Gasparrini et al.) suggests that humans are much better adapted to warmer temperatures than cold. Zhao et al. also found that the temperature effect on emergency department visits varied by latitude; the effect of extreme cold was higher in the southern cities and declined northward, whereas the effect of extreme heat was higher in the northern cities and declined southward, which suggests a form of regional adaptation to temperature, similar to what we (Michaels) found for U.S. cities in our work on urban heat-related mortality.
In stratifying their analysis by gender and age, the thirteen researchers report that the temperature/emergency department visit relationship was unaffected by gender but was attenuated with increasing age, which contradicts other, more speculative work on Chinese urban heat-related mortality. At the national level, the relative risk of emergency department visits due to cold declined from 2.27 for the youngest age group (0–14 years) to 2.17 for ages 15–34, 1.60 for ages 35–64 and 1.41 for the elderly aged 65 and older. Similarly, the risk of emergency department visits due to hot temperatures also declined from 1.51 for 0–14 years to 1.19 from ages 15–34, 1.14 for ages 35–64 and 1.08 for those over age 65. Children of ages 0–14 were the most vulnerable to cold spells and heat waves over the period of study.
In considering all of the above findings, plus those reported in numerous other studies of the subject, it is clear that the impact of cold weather on human health is much more severe and longer lasting than that caused by heat waves. The truth be told, as shown by real-world numbers, humanity has much more to gain in terms of physical heath from rising, as opposed to falling, temperatures."
Gasparrini, A., et al. (2015) “Mortality risk attributable to high and low ambient temperature: a multicountry observational study.” The Lancet 386 (9991): 369–375.
Davis, R.E., Knappenberger, P.C., Michaels, P.J. and W. Novicoff (2003) “Changing heat-related mortality in the United States.” Environmental Health Perspectives 111 (14): 1712–1718.
Zhao, Q., Zhang, Y., Zhang, W., Li, S., Chen, G., Wu, Y., Qui, C., Ying, K., Tang, H., Huang, J., Williams, G., Huxley, R. and Guo, Y. (2017) “Ambient temperature and emergency department visits: Time-series analysis in 12 Chinese cities.” Environmental Pollution 224: 310–316.