This article presents information regarding social data and response models related to crime rates. According to the dictionary there is not much to distinguish "prediction" from "forecasting" in meaning, they can and probably mostly are used as close synonyms. The former may be a shade less technical and more general in application. In what follows attention will be centered on mass phenomena and the future behaviour of indicators of response in the aggregate, and the conjectures that may be formed of that behaviour in advance. Here "forecasting" is certainly a legitimate term and will be preferred, for instance the forecasting of crime trends. There is first the classical case of a strictly periodic system, for instance the planets revolving around the sun. On the whole this model, which no doubt implies predictability, has been a source of error and confusion more than anything else with its assumptions of cyclical schemes which have little foundation in behavioural realities. More interesting is "experimental predictability" by which is meant prediction of response when something is added, or withdrawn from a set of conditions.