involving other contagious ideas, but they tend to lose their punch and identity in
the process. Their ability to direct thought and action becomes much diminished.
A key proposition of this book is that economic fluctuations are substantially
driven by contagion of oversimplified and easily transmitted variants of
economic narratives. These ideas color people’s loose thinking and actions. As
with disease epidemics, not everyone becomes infected. In the case of narrative
epidemics, the people who miss the epidemic may tell you that there was no
such important popular narrative. But in a historic epidemic, for most people the
narrative will be fundamental to their reasons for doing, or not doing, things that
affect the economy. Just like the economic theories in Figure 3.3, popular
theories among the general public grow on an upward epidemic path, but only
for a while. They then recede unless they get renewed.
FIGURE 3.3. Frequency of Appearance of Four Economic Theories, 1940–2008
The figure shows four important models: the IS-LM model (Hicks, 1937), the multiplier-accelerator model
(Samuelson, 1939), the overlapping generations model (Samuelson, 1958), and the real business cycle
model (Kydland and Prescott, 1982). All four show hump-shaped patterns through time. Source: Google
Ngrams, no smoothing.
It is noteworthy that Keynes’s book The General Theory of Employment,
Interest, and Money (1936) put forth the idea of a perfectly mechanical
contagion without using that phrase. According to Keynesian theory, an

economic boom starts when some initial stimulus, such as government deficit
spending, causes an initial increase in some people’s income. These people then
spend much of their additional income, which in turn generates income for other
people who sell to them or work for companies that sell to them. They in turn
spend much of this extra income, thus generating another round of income
increases for yet other people, and so on in multiple rounds of expenditure. The
Keynesian theory can be tweaked to add some investment dynamics, as Paul
Samuelson showed in 1939 with his multiplier-accelerator model, thus creating
hump-shaped responses in national income as a result of an economic stimulus.
These hump-shaped responses resemble the epidemic curves we have seen. We
can view the Keynes-Samuelson model as an epidemic model of sorts, where the
contagious element is income. However, it is not enough to think solely in terms
of mechanical, multiple rounds of expenditure. We must think of multiple rounds
of expansion of economic narratives, and of the ideas and feelings embodied in
them.

Constellations and Confluences of Narratives
Just as the world experiences co-epidemics of diseases, where two or more
diseases interact positively with each other, we also see co-epidemics of
narratives in which the narratives are perceived as sharing a common theme,
such as case studies that illuminate a political argument, creating a picture in the
mind that is hard to see if one focuses on just one of the narratives. In other
words, large-scale economic narratives are often composed of a constellation of
many smaller narratives. Each smaller narrative may suggest a part of a larger
story, but we need to see the full constellation to discern the full theme.
The 
analogy 
to 
constellations 
should 
be 
clarified. 
Astronomical
constellations, such as Cygnus the Swan, are chance alignments of stars, but
humans interpret them in a way that seems natural to the human mind—in this
case, as a swan. Sometimes humans co-opt constellations for certain purposes.
For example, Christians have renamed Cygnus as the Northern Cross to put one
of their symbols in the sky. They also paired it with another constellation, the
Southern Cross, for people living in the Southern Hemisphere. Other groups and
cultures have different narratives with other motivations.
Narratives appear in constellations partly because their credibility relies on a
set of other narratives that are currently extant. That is, they sound plausible and
interesting in the context of the other narratives. The storyteller does not need to
refute the other narratives to set the stage for the current one. Also, the narrative
may be based on certain assumed facts that the teller and the listener do not
know how to test. Some narratives are contagious because they seem to offer a
confirming fact. We can say with some accuracy that most people put on a show
of their own knowledgeability and try to conceal their ignorance of millions of
facts. Hence narratives that seem contrary to prevailing thought may have lower
contagion rates that do not result in epidemics.
Some narrative constellations may at their peak infect only a small fraction of
the population, but if that fraction of the population curtails its spending
substantially, the narrative may matter a lot. For example, if the narrative has
reached only 20% of a country’s population, but that fraction decides to
postpone purchasing a new car or fixing up their house, the impact of its
decreased spending may be big enough to tip the country into a recession.
In addition to a constellation of narratives, there is a confluence of narratives