Proposition 4: The Economic Impact of Narratives May Change
Through Time
An economic narrative’s impact on behavior depends on details of the narrative’s
current mutation and other related narratives. When we rely on digitized data on
words or phrases that are flags for narratives, we must resist the temptation to
assume that all the narratives with these flags have the same meaning through
time. We have to read the narratives in terms of their implication for action, in
the context in which they were spoken, at least. In the future, some information-
processing innovation might make this undertaking less dependent on human
judgment.
Let’s look again at the October 19, 1987, stock market crash, the biggest one-
day crash in percentage terms in history. The topic still comes up regularly, often
on major anniversaries of that event. We might believe that memories of that
crash make stock markets vulnerable to another crash, because fear of a crash
may cause people to react to the apparent beginnings of a drop in stock prices.
But the narrative of the 1987 crash need not have any such effect if people do
not think current circumstances are similar. In 1987, there was much discussion
of a new computerized trading program called portfolio insurance. Along with
other factors, narratives about portfolio insurance led to a predisposition to
consider selling that was peculiar to that time.5
Other disturbing stock market events were surrounded by narratives that had
nothing to do with portfolio insurance. After Austria-Hungary declared war on
Serbia on July 28, 1914, touching off World War I, stock prices began to fall
precipitously. Reacting to the panic, the New York Stock Exchange and all the
major European stock exchanges closed their doors. Even though the United
States was not involved in the war, the New York Stock Exchange did not reopen
until December 12. In his 2014 book about this closing, When Washington Shut
Down Wall Street, William Silber details a number of stories and rumors that
contributed to the market’s severe reaction. Notably, panicky European investors
scrambled to get their investments out of the United States while they could.
During this “European gold rush,” massive amounts of gold were shipped from
the United States to Europe despite increasing danger to transatlantic shipping.
There was much talk about the Panic of 1907 as proof that US markets were
unstable, along with fears that another panic might occur. In addition, there was

a baseless rumor that the assassination of Archduke Franz Ferdinand, which
triggered World War I, was part of a conspiracy involving the Russians, who
were hoarding gold in preparation for a great war.
In contrast, the beginning of World War II in 1939 did not close the US stock
market. After the United Kingdom declared war on Germany on September 3,
1939, marking the beginning of World War II, the Standard & Poor’s Composite
Index gained 9.6% in one trading day. Newspapers expressed general surprise at
such a positive market reaction and were mostly at a loss to explain why the
market did not repeat its 1914 experience. Apparently the very different response
had something to do with a narrative that World War I had, ultimately, proven
very profitable for some investors who’d held on to their stock market
investments and profited from selling armaments or supplies to Europe.6 The
human stories of World War I and World War II might be very similar, but there
was a huge difference in the narratives describing successful investors around
the start of each war.
We must pay attention to the names that people attach to their narratives.
Seemingly minor changes in the name of a narrative can matter a lot, especially
if the new name attaches to a different constellation of narratives. In linguistics,
synonyms never have exactly the same meaning. If pressed, people can state
complex thoughts about the slightly different connotations of synonyms. In
neurolinguistics, synonyms have different connections in the neural network.
Some of those connections can matter a lot in terms of the economic ideas they
support.