In 1997, two major survey of the literature on financial institutions’ efficiency were published. The first one presented an overview of the results in an international perspective and proposed new research themes. (Berger & Humphrey, 1997) The second survey focused more on the various approaches used by the researchers and showed the sensitivity of the results on the methodological hypotheses. (Berger & Mester, 1997) Recently, a third paper emphasized on the results of the studies analyzing the effect of mergers on financial institutions’ efficiency. (Amel, Barnes, Panetta, & Salleo, 2004) The specific situation of the financial services co-operatives was analyzed by some authors. Let us mention, for example, the works of Worthington and Ralston (1999 & 2001) on Australian credit unions, of Sifakis-Kapetanakis (2001) on French co-operative banks, Fried, Lovell & Eeckaut (1993) on American credit unions and of Fortin & Leclerc (2003, 2007a & 2007b) and Leclerc & Fortin (2004, 2008 et 2009) on Acadian and Quebec caisses populaires.

This paper study the link between the local market conditions in which financial services co-operatives operate and their level of efficiency. To establish an organization’s performance, one focus especially on the quantity of resources used to reach a certain level of production. Some of these variables are effectively under the control of the managers. However, other variables connected to local market conditions can have an impact on the co-operative’s performance. For example, a credit union operating in a market where the proportion of old persons is very high will have to offer more services more expensive counter and cannot benefit completely from the substitution rate between two types of physical capital, buildings and equipments. Heterogeneity of markets places the co-operative in front of a series of variables which affect her performance but on which it has not enough control. If we do not recognize the contribution of this environment, we risk to base our evaluation on biased results. Using a data bank grouping together Acadian and Quebec caisses populaires, this study try to measure the importance of this bias.