Does Uncertainty Affect Non-response to the European Central Bank’s Survey of Professional Forecasters?

Abstract

This paper explores how changes in macroeconomic uncertainty have affected the decision to reply to the European Central Bank's Survey of Professional Forecasters (ECB's SPF). The results suggest that higher (lower) aggregate uncertainty increases (reduces) non-response to the survey. This effect is statistically and economically significant. Therefore, the assumption that individual ECB's SPF data are missing at random may not be appropriate. Moreover, the forecasters that perceive more individual uncertainty seem to have a lower likelihood of replying to the survey. Consequently, measures of uncertainty computed from individual ECB's SPF data could be biased downwards.

Quasi ex-ante inflation forecast uncertainty

“…Indeed it is often noted that the experts participating in such pools can be inattentive and fail to update their forecasts, may disagree when updating, and do not always learn from past experience (Andrade and Le Bihan, 2013). Furthermore, forecasts in a panel may be highly correlated (Makarova, 2014), and the constitution of panels of forecasters can frequently change, depending on the phase of the business cycle (López-Pérez, 2016). Additionally, psychological bias, overconfidence and underconfidence may play a role when probabilistic and interval forecasts are being formulated (Soll and Klayman, 2004;Hansson, Juslin and Winman, 2008;Clements, 2014).…”

Effects of Monetary Policy Decisions on Professional Forecasters’ Expectations and Expectations Uncertainty

“…The European Survey of Professional Forecasters (to be abbreviated as ECB SPF) has been analyzed quite extensively on both the aggregated level (e.g. Tsenova 2012; Andrade and Le Bihan 2013) and on the micro level (Dovern and Kenny 2017, Lopez-Peres 2016a, 2016bAbel et al 2016;Oinonen and Paloviita 2017). Individual responses of other surveys have also been examined by some authors.…”