The previous chapter ended with a description of the (econometric) properties of optimal expert adjustment, given that an expert has received a statistical model forecast, which he or she could incorporate in the expert-adjusted forecast. In the present chapter some of the currently available and recent empirical evidence of expert adjustment is reviewed. This chapter consists of three sections. First, a review of empirical results is given for such diverse companies and institutions as KLM (Royal Dutch Airlines), Bayer and Organon (both pharmaceutical companies) and the CPB (the Netherlands Bureau of Economic Policy Analysis). The results of various other available studies will also be included. Next, there is a brief discussion on what experts themselves say that they apparently do, based on anecdotal evidence and also based on interviews with the experts from Organon. Finally, this chapter deals with possible explanatory variables for the observed behaviour of the experts like personal traits (age and experience) and job characteristics (how often the experts make the forecasts).

The key lesson to be learned from the evidence summarized in this chapter is that there are substantial differences between the optimal behaviour as put forward in Chapter 2 and the observable behaviour of experts. Moreover, there seem to be various explanatory variables to predict this behaviour, and these are predominantly associated with past behaviour, past realizations and past forecasts. Personal traits do not seem to matter much.