<p>Estimates of residential water consumption responses to price changes often conflate the structural price effect with behavioral responses resulting from public discourse preceding the enactment of new water rates. This paper examines this identification challenge using data from Norman, Oklahoma, where a municipal charter mandates voter approval for any water rate increase. Norman experienced two recent elections. One in August 2010, which failed to raise rates, and another in January 2015, which increased rates. A difference-in-differences model is estimated to compare behavioral responses to the successful January 2015 rate election with those to the failed August 2010 election, thereby attempting to disentangle the effects of the information campaign from those of the price increase. Since both campaigns exposed residents to similar conservation-oriented public discourse, contrasting post-2015 behavior with post-2010 behavior yields an estimate of the net behavioral response associated with the enactment of higher water prices following a successful election. Heterogeneous net responses are measured across four consumption groups defined by pre-election usage. Net reductions in use ranged from 7% for low-volume users to 32% for medium-high volume users. Finally, the evidence suggests that both price increases and information campaigns lead to conservation, but the largest effects were observed around Norman’s 2015 rate change, where both non-price and price factors were present. These results are consistent with existing results on nonlinear water pricing. These estimates should be interpreted as context-specific net effects rather than fully identified structural elasticities, since pre-trends are not perfectly parallel across groups.</p>