Conferences

The integration of objective and subjective indicators represents a crucial approach in order to come up with a solid scientific result and understanding of relevant social phenomena. The subjective and objective dimensions can provide new perspectives and allow the units of interest to be evaluated also in social policy.

This is confirmed also by many international initiatives and events revealing the increasing attention on individual perception of living conditions and quality of life and the necessity of a correct scientific approach to their measurement and analysis. In this perspective, we can include, for example, the theoretical debate around the relationship between economics and happiness.

The need to integrate subjective and objective information comes from different sources (statistical offices and survey) and is causing a growing demand in the study of well-being and happiness of societies (see OECD and its agenda of the “Measuring and Fostering the Progress of Societies” World Fora on Statistics, Knowledge and Policy).

In a policy perspective, the need for subjective indicators arises during (i) the assessment of policy results and (ii) in the selection of policy objectives (Veenhoven, 2002). The first item concerns the need to assess if a policy has been successfully implemented (e.g. “is there perception of more security in the streets after increasing police staff?”). The second item refers to what people desire (e.g. less air pollution, more cultural events in the city). The need arises also in consideration of the limits of objective indicators (Veenhoven, 2002), as reality cannot be fully reduced only to objective facts. Moreover, objective facts are measured referring to a design or a model that is “subjective” in its definition.

For this reason the definition of an integrating model is needed. This model requires firstly the definition of a conceptual framework from which it is possible to identify the proper analytical approach (causal analysis, multilevel analysis, life-course analysis, or explorative analyses). Secondly, it requires an organizational context in which the integration can be accomplished by relying on structured and systematic data, observed in long-term longitudinal perspective (e.g. systems of indicators) and in which particular technical issues (i.e. aggregation issues) can be managed.

The paper will discuss these aspects by referring in particular to the feasibility of the different statistical approaches taking into account their specific assumptions. The goal is to describe a procedure able to yield results, not only statistically valid and consistent with reference to the defined conceptual framework, but also easy to read and interpret at policy level.