The EKS index avoids base-group bias in multilateral comparisons as it is the geometric mean of bilateral Fisher indices computed using all possible bases. Is there an easy way to compute it?
Also, with panel data, to make cross-sectional EKS indices comparable over time, Robert Hill (AER, 2004) suggests a method that uses a time-series of price indices to link them over time. Is there a Shazam procedure to do that?