Estimating price elasticity when prices don't vary much

[George Tim]

Hi. I have sales and price data for several grocery products from 10 stores in different states and owned by the same chain. However, prices vary less frequently and are almost uniform in all stores (instrumenting in hausman style isn't possible). I am unsure if I should regress log sales on log prices without instrumenting for price. Are there conditions where IV may not be required? The retailer is known to have low prices across products and time. The only case I found when IV isn't required is when supply is not price sensitive.