On folklore and ‘Folklore’ (QJE, forthcoming): Forging dialogue, reconciling trade-offs


This page in:

Somehow one senses that authors are quite confident about the stature of their paper when they give it a one-word title (see ‘Corruption’, ‘Worms’). And upon its recent acceptance at QJE, ‘Folklore’ (by Stelios Michalopoulos and Melanie Meng Xue) seemed to vindicate that assessment, as did rapid circulation of this news among development researchers on Twitter. By scholarly life’s primary metric of stature, publication of an article in a journal like QJE deems it to be, by definition, ‘great’. It’s certainly an interesting and innovative contribution; it’s a good paper to think with (the main quality criteria I apply these days) and the authors have thoughtfully engaged with complex phenomenon. I welcome its arrival but will leave it to others to quibble about methodological details and other normal intra-disciplinary issues. Since I’m not an economist – albeit one who has spent his entire career in their close company – I offer instead some brief thoughts on it from neighboring lands.

But first, what’s the excitement about? For those who haven’t read it, the paper’s substantive contributions rest on a meta-analysis of 2,564 motifs (“a combination of images, episodes, or structural elements found in two or more texts, including sacred and profane ones”) from roughly a thousand groups around the world. Over many dutiful years, Yuri Berezkin, a Russian anthropologist, has collated this material into his Folklore and Mythology Catalogue. The authors assess the veracity of the catalogue’s constituent data by matching each case with “known geographic and social attributes”, though such a forensic analysis inevitably elicits the database’s many empirical gaps. The authors proceed to ‘fill in’ these gaps by creative deployment of machine learning techniques, thereby building an even more extensive and robust dataset. So, cutting edge empirical approaches are applied to a novel dataset on an issue long acknowledged as intriguing but presumed to be either the exclusive purview of other disciplines or to have ontological characteristics beyond assessment by economics’ standard tools. Sounds great…

What do the authors find that might be of particular interest to development researchers? Several things, including that yesterday’s folklore characteristics can manifest themselves in today’s development outcomes. In this sense, then, it adds an additional vantage point to an increasingly sophisticated literature in mainstream economic history making similar claims regarding the influence of the past on the present (including such bold conclusions as “Africa’s slave trades account for all of Africa’s poor performance relative to other developing countries” (Nunn 2020)). In ‘Folklore’, groups steeped in oral traditions celebrating the natural dominance of men, for example, are shown to have larger gender disparities today; those groups (or even specific occupations within them, such as farmers) characterized by motifs sanctioning the actions of tricksters are more likely to be prosperous in the present moment, as are those groups predominantly featuring tales in which valiant protagonists triumph over risks and setbacks. “The share of tales depicting a hero overcoming challenges”, we learn, “is twice more common in the Dutch folklore than in the Russian one.”

Which is all very interesting. But what does one do with these findings? What are the “policy implications”? What do we not learn? If it was a more familiar development issue being examined this way, one might readily conclude that since we care about gender equality, justice, and poverty reduction then, on the basis of the “rigorous evidence” provided, we now have warrant to recommend that policy steps be taken to engage with (if not actively suppress) ‘factors’ that seem to causally undermine these laudable objectives. But sacred cultural expressions aren’t tax rates. The authors, of course, make no such suggestion and would doubtless be horrified if governments or development agencies were to introduce policies seeking to modify any group’s customary understanding of its world. But we need much greater sensitivity, here and elsewhere, to how one epistemological rendering of this deeply complex phenomenon, when celebrated and privileged above others, can be used for precisely these purposes and unleash unintended harmful outcomes.

So understood, historians would quickly recognize two concerns in taking the findings of ‘Folklore’ at face value (by extension, see Bayly et al 2011). First, they would note that it was exactly this line of thinking – theory and evidence being invoked to demonstrate how material inferiority stemmed from a certain group’s ways of making sense of their world – by which colonialism was (partially) justified, and on which much of modernization theory and indeed some foundational documents of today’s multilateral institutions were based (see Escobar 1995: 3). Second, measuring complex cultural issues at scale, via necessarily standardized and standardizing instruments, is how colonialism worked and indeed how all modern public administration still works. I hasten to add that doing so is not inherently ‘bad’, since all manner of unambiguously positive outcomes flow from such management – insurance, for example, is made possible by codifying risks. And deferring unquestioningly to prevailing local mores, especially in matters pertaining to justice, has vexingly real ethical and political limits of its own.

Why does this matter? The very essence of both research and modern administration is that inherent complexity and heterogeneity must be made “legible”, the “unobservable” rendered able to be “seen” (Scott 1998). But ‘measuring’ cultural phenomenon risks triggering the social scientific equivalent of the Heisenberg Uncertainty Principle, namely that the very act of measuring something changes it. As Hostetler (2021: 290) has recently argued in her discussion of the administrative mechanisms underpinning colonial rule, new “techniques of legibility— scaled mapping, uniform map symbols, standard systems of meas­urement, and standardization of print languages— introduced new assumptions about what it meant to be, say, French, British, Russian, Thai, Chinese, or Indian.” Within countries, the introduction of the census in India, for example, as Dirks (2001) has shown, fundamentally changed how an otherwise relatively fluid social category such as ‘caste’ became a fixed and discrete (and thus mobilizable and ‘targetable’) political identity upon being reduced to a box needing to be ticked by registered citizens. The same with ‘religion’.

I can’t help but read ‘Folklore’ in a similar light. All the careful, detailed attention to assessing actual folklore ‘at scale’, in ways that are inherently a mile wide and an inch deep, yield interesting findings and will doubtless inspire a whole raft of intriguing follow-up studies. But here are some questions I’ll be looking and hoping for… By what mechanisms, and under what conditions, do orally conveyed mythologies of yore shape present material outcomes? How do we explain seemingly novel cases, in which strikingly similar tales – e.g., about the ‘Seven Sisters’, a constellation of stars – are told in contexts separated by thousands of miles, open ocean, and at least 50,000 years (see Norris and Norris 2020)? What meaning and coherence does what outsiders call “folklore” convey to insiders – to those who craft, live by, and celebrate such accounts? How can contrasting accounts be reconciled, especially when they are ontologically orthogonal – i.e., when they are grounded in qualitatively different understandings of how the world works and the very meaning of what concepts ‘are’? For some, ‘land’ and ‘water’ are essentially commercial commodities like any other, to be bought and sold in open markets by individuals; for others, they are sacred repositories, gifts from powerful gods and venerated ancestors over which contemporary custodians have temporary collective stewardship, and for whom the introduction of a worldly pricing mechanism to facilitate exchange and ‘management’ would be sacrilege. (On this dynamic, see the fabulous Spanish movie ‘Even the Rain’.)

These questions, I suggest, aren’t merely anthropological curiosities but fundamental development concerns, because so much of what is done in the name of development is itself (or is certainly perceived by others as) a form of folklore and introduces new (potentially competing) folklore. All epistemic communities, including academic disciplines, are defined by what counts as a question and what counts as an answer – see Leijonhufvud’s (1973) classic ‘tale’ on the folklore of economics (by an economist!). Requiring women to give birth in modern hospitals in the name of meeting Sustainable Development Goals on lowering maternal mortality, for example, may seem entirely uncontroversial… Unless one recognizes that, for certain groups (such as many Aboriginal communities in Australia), where exactly one is born is everything – it forever shapes one’s identity, responsibilities and obligations. Deep in the noble quest to fly a woman in labor 200 miles to a city hospital is an assumption that one’s birth location doesn’t matter... In this instance, however, the ‘tale’ of a group wearing white lab coats regarding how to safely and legitimately birth the next generation gets to override another group’s 50,000-year-old ‘tale’, in the process yielding a safe birth but a fractured community. (See Moran 2016 for similar wrenching accounts of the dynamics driving the ‘delivery’ of housing, education, and social services in northern Australia.)

Economics, of all disciplines, should be astutely attuned to trade-offs, in this instance of what is gained and what is lost when complex topics are addressed from a single standpoint, or when methodological ‘depth’ and ‘breadth’ yield rather different insights, implications, and applications. This is especially salient for issues intrinsically located at disciplinary boundaries, where monopolies in the production and distribution of ideas can be as inefficient as they are in goods and services (Rao and Woolcock 2007). The real success of ‘Folklore’ will be expanding, not constraining, the market for ideas on folklore. I welcome its publication for the great intra-disciplinary paper it is, but different kind of work will be needed to forge broader dialogue. To that end, and for that task, it may need a longer title.


Michael Woolcock

Lead Social Scientist, Development Research Group, World Bank

Join the Conversation

Alvaro S. Gonzalez
February 23, 2021

Great blog. Thanks for this!