Simple question re: correlation?


I am looking to assess the correlation between the average costs of services in a geographical region and the average earnings of people in those regions. I would also like to measure the relationship between the percentage change in the cost of those services and the percentage change in the average earnings for each respective region.

My question is this:

There are only 11 regions in my sample. I have used Pearson's (which I am aware is not appropriate) and Kendall's Tau and have achieved significant (p<.05) correlations for the relationship between earnings change and the change in two services. What, if any, is the appropriate measure of correlation I should be using, given that my sample is so small?

Any help is greatly appreciated.

Thank you.