1: \begin{abstract}

2: For a given orthonormal basis $(f_n)$ on a probability measure space, we want to describe   all Markov operators which have the $f_n$ as eigenvectors. 

3: We introduce for that  what we call the  \emph{hypergroup property}. We study this property in three different cases.

4: 

5: On finite sets, this property appears as the dual of the GKS property linked with correlation inequalities in statistical mechanics. The representation theory of groups provides generic examples where these two properties are satisfied, although this group structure is not necessary in general.

6: 

7: The \emph{hypergroup property} also holds for Sturm--Liouville bases associated with log-concave symmetric measures on a compact interval, as stated in Achour--Trim\`eche's theorem. We give some criteria to relax   this symmetry condition in view of extensions to a more general context.

8: 

9: In the case of Jacobi polynomials with non-symmetric parameters, the  \emph{hypergroup property} is nothing else than Gasper's theorem. The  proof we present is based on a natural interpretation of these polynomials as harmonic functions and is related to analysis on spheres. The proof relies on the representation of the polynomials     as the moments of a complex variable.

10: \end{abstract}

11: