Conventional wisdom about the economic future. The role of “Consensus” in the world of macroeconomic forecasting

In the chapter 12 of his General Theory, Keynes (1936) famously states that “Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally” – leading to the resemblance between stock markets and beauty contests. In a sense, the Keynesian stress on conventions is an early expression of social dynamics and social embeddedness on financial markets – a path which may lead to the late discovery of Durkheim by economists (see for example, Orléan, 2014). The purpose of this presentation is to highlight the role of devices in shaping practices by implementing such conventions, in the case of macroeconomic forecasting. Exposing some preliminary results of an on-going research, it focuses on forecasts production rather than on the way economic actors use them.

As a professional activity, forecasting aims at reducing uncertainty and provides a basis upon which economic actors may form expectations. It helps them to imagine futures (Beckert, 2013). Being a notoriously difficult, if not impossible, task, it seems rational for economic actors to turn to collective rather than individual assessments of the values macroeconomic variables should take. Such devices, among which the studied firm is one of the most prominent and regarded as one of the most reliable forecasting tool (Batchelor, 2001; Ager, Kappler & Osterloh, 2009), give access to the “average opinion” of forecasters, and embody a Keynesian “convention”. In practice, this firm aggregates country economic forecasts and topical analyses covering various countries, mostly from private sector forecasting institutions. They are published by a Europe-based organization established in the late 1980s which claims to be “the world’s leading macroeconomic survey firm”.

The presentation will first tackle the historic origins of this device since its creation in the late 1980s. It will also explore its meaning for professional forecasters. All those who have been interviewed so far, without exception, acknowledged its key role as a resource for forecast-production: it is one of the “primary source” (Evans, 2007) to produce forecasts. Consequently, one of its most important feature lies in its self-referential nature: aggregating the outputs of a large number of forecasting institutions (even though their respective weights remain partly related to symbolic hierarchies). In this perspective, it reminds of Keynesian mechanisms. However, some its effects challenge Keynes’ statement, especially insofar as the difference between forecasters from private and public sectors is taken into account. Forecasters’ reputation lies in their ability to produce accurate forecasts, or more accurate than the average forecast. Their relation to the “consensus” is therefore paradoxical: on the one hand, forecasters who stand close to the “consensus” take the risk of “failing conventionally”, on the other, those who don’t exhibit distinctive values upon which their (and their employers’) reputation may be built or jeopardized. As a result, devices like “Consensus Forecasts” give rise to market dynamics which display a mix of conformity and originality – “beating the Consensus” being regarded as necessity, even though hard to achieve.