Recent scholarship finds that communicating descriptive norms, such as the fact that 97% of climate scientists have concluded that human-caused climate change is happening, is an effective gateway to changing individuals' beliefs and attitudes about climate change and support for climate policies. Other scholars, however, have offered an alternative explanation: priming people with any "97%" figure serves as an anchoring heuristic that leads people to adjust their estimates of the scientific consensus. This study investigates this proposed anchoring effect. Results from three parallel experiments indicate that participants consistently update their estimates of the scientific consensus when they receive a relevant statement about the scientific consensus on climate change, but not when receiving a semantically equivalent but otherwise irrelevant consensus cue that might be used as an anchor. Further, we find that perceived consensus mediates the effect of the consensus treatment on individuals' attitudes. We also find that conservatives updated their consensus estimates significantly more than liberals. We replicate the results using three large samples from different sources. The findings indicate that when people update their estimates of the scientific consensus, it reflects a genuine update in their beliefs and is unlikely to be a result of the anchoring and adjustment heuristic. This work contributes to a growing literature on the value of communicating expert consensus about contested scientific issues.