So I am trying to understand the impact of the elasticity of substitution in terms of taxation. I am carrying this out for the U.S. by firstly estimating a two-input CES production function. Chirinko (2002, http://www.cesifo-group.de/DocDL/cesifo_wp707.pdf) proposed that the elasticity of substitution between capital and labor is one of the channels that determine the effectivity of tax policy. I understood that for example considering a horizontal, completely elastic demand function tax policy can be very effective to increase capital stock since firms demand is characterized by great substitutability. Now here is my first question, does the elasticity of substitution sigma (of the examined country) have to theoretically be necessarily above unity, since only then the goods are substitutes, or is it possible to have an effective tax policy even though the elasticity of substitution is smaller than unity and the goods are rather complements. My second question would be if there is a way to influence the size of the elasticity of substitution towards a more effective tax policy. It might be somehow stated in the paper I linked above, but I didn't understand it then. Is there a way and tax policy (and through what kind of taxes) can maybe rise that elasticity of substitution through f.e. price incentives so it can make it itself more effective?