Quality of the buildings is often considered just as a mix of financial variables and objective characteristics. The main problem of this approach is that price is supposed to well reflect the quality of the property. However, real estate markets are not transparent, goods are complex and the information on the characteristics is difficult to obtain. The informational content of price might also reveal others aspects such as rarity, or just willingness to pay in an imperfect world. It suggests that, for market participants, price is less a mirror of the quality, than the origin of the idea of quality itself. For instance, transaction prices are often based on similar properties prices, more than on in-depth studies of the quality of the buildings. This situation can lead to disconnections with the reference value.

To address this issue, we develop a methodology directly based on the dimensions of the quality. Certifications or labels already provide indications of that kind, but these approaches are often normative, with arbitrary weights explicitly decided by experts committees. Specific approaches also exist (Breeam, HQE, Wired…) but they are partial and oriented toward a goal (environmental, connectivity). Here, we propose a new approach where the levels of quality for each characteristics are judged by market participants themeselves. More precisely, this approach consists in making explicit market quality judgments.

In the first phase, an experimental experience is realized. It is asked to a group of market participants to formulate global quality judgements on buildings of the Paris office market, described by their characteristics. Then an econometric procedure is applied to identify what are the most valued characteristics. This approach allows different buildings to be positioned relatively to a benchmark and provide an alternative way to consider quality, i.e. the reference value of the building. It also makes possible to track what are the evolutions in the quality judgements in the market across time.