Observations of an expert

For various reasons, statistics, sometimes, are not enough, complete or available to legitimate policy decision-making. Then recourse has to be taken to other kinds of observations. These are the observations made by human beings, so they are not the registrations of a measuring device, nor mechanical observations, nor 'objective' observations. The standard view (in science) is that these non-statistical observations run the risk of being more biased than statistics, therefore often denoted as 'subjective' observations. Some people, however, are assumed to be less biased than others, because of training, because of experience, or because they have better intuitions. These people are called experts.
This paper discusses the role of expert's observations in different practices of decision making. In these practices it is never the case that the observations of one sole expert is being used, so discussing the role of expert’s observations implies a discussion of how these observations are combined. The practices that will be discussed are cases from economics and econometrics, with the exception of a case in risk analysis, which provides a rather lucid example of a practice in between statistics and the intuitive expert. These practices will be discussed in their historical context, starting with a notorious debate in econometrics on the role of statistics.