Monetary asset user costs are functions of spreads between a benchmark rate of return and the own rates of return on the monetary assets. We analyze the impact of the benchmark rate on a Euro area Divisia M2 aggregate, on estimated elasticities of substitution, and on estimated impulse response functions. Substitution in response to changes in the user cost of M1 is generally elastic, but we find evidence of inelastic substitution along other dimensions. When a loan rate is used as the benchmark, substitution in response to changes in the user costs of the two components of M2-M1 is inelastic throughout the sample and the corresponding elasticity estimates are near their lowest levels during the pandemic. This is strong evidence that Divisia monetary aggregates are preferable to conventional monetary aggregates. Annual growth rates of simple sum and Divisia M2 monetary aggregates differ significantly in some periods, but not during the pandemic. Estimated impulse response functions using both Divisia and simple sum money measures indicate that money shocks have positive and statistically significant effects on real output. The response of the price level to a money shock tends to be more persistent when the models are estimated using Divisia aggregates.