A good story costs a lot of money – Newspaper Kommersant No. 161 (7362) of 09/02/2022

Explanations of how certain numbers, indicators and measurements are related to each other have their own power in the eyes of people – as economists Chad Kendall and Constantine Charles demonstrate in an experimental study, the human mind prefers to create “stories” from numbers and believe them, moreover formally refuting them statistical, factual and logical arguments are ineffective. Economists undertake the study of such effects, among other things, to study how they work in the expectation economy. For them to work, as it turns out, they do not need to be justified or even rational at all.

In the preprint paper “Causal Narratives” (NBER Publication Series), Chad Kendall of the University of Southern California and Constantine Charles show the results of their own experiments on how specific types of narratives can work, tied into a single narrative story. In this case, “causal narratives” were studied – a linear narrative similar to those used in modern technologies of “storytelling” and linking heterogeneous data sets one after another in a chain. Kendall and Charles were actually experimentally testing the assumptions of a 2020 article by Kfir Eliaz and Rani Shpiler, who studied different types of narratives, including “smoothing,” which assumes a logical alignment of contradictory facts, and “annoying,” which pits facts against each other. Its authors assumed that, on average, people would prefer the first type of narrative construction.

It is no coincidence that these studies are undertaken by economists, and not by psychologists, linguists or sociologists, who in recent years have also been extremely interested in the study of narratives and the source of their persuasiveness for listeners – it sometimes surpasses the power of logic and evidence. For behavioral economics, narratives are one of the foundations for the formation of expectations that underlie the explanatory models of many modern economic schools. In the narrow sense, expectations management has been the basis of the neo-Keynesian practice of monetary policy of most central banks since the end of the 20th century. In other words, any Central Bank would like to know how economic explanations for the relationship of certain figures (for example, government spending and inflation) are accepted by the markets, how different types of explanations compete with each other, and why sometimes the wildest narratives outperform scientifically based reports.

The main findings of the experiments of Kendall and Charles, which tested the ability of people to form, borrow and reject narratives, are as follows. “Causal narratives” (at a minimum, placing additional intermediate variables between two sets of data, “A depends on B through C”) do have a special power – if the story is created in this way, then it works better than narratives of any other type, and the demonstration of refuting a constructed narrative of statistical or factual data can only reduce the spread of the “causal narrative” by 20-30%, but de facto does not refute it. People do prefer “flattening” narratives to “annoying” ones, but to a greater extent they love the narratives that they create themselves and are ready to spread them more vigorously than any others. But it is the “intermediate data series” that is of primary importance: even if the initial and final data series are not connected in any way, putting correlating data “explaining everything” between them has almost magical power. The stories concocted in this way about how, for example, information consumption by teenagers of social media and depression can be connected, there is little that can resist, including objective data about the absence of such a connection, and a formally equal (and no less logical) message that teen depression leads to social media use. It is not hard to imagine that the study of narratives by economists will remain at the forefront of research: the potential results are very enticing for all players from marketing to public administration.

Dmitry Butrin