The tax was unpopular - no real surprise. But when people were asked if they supported making the carbon offset mandatory - which is of course exactly equivalent - the response was highly favourable (around 2 points more positive on a scale from -3 to +3).
Not only was the "mandatory offset" more popular, but it was regarded identically by Democrats and Republicans. The tax, on the other hand, was strongly disliked by Republicans while Democrats made no distinction between taxes and mandatory offsets. Thus the entire effect appears to be due to Republicans' attitudes to tax.
This is an example of a well-known cognitive bias called 'framing'. The paper's purpose is to study the cognitive processes that generate the framing effect, and the authors find evidence that it arises partly from the order in which advantages and disadvantages are evaluated by a subject.
The original paper is here and contains lots of other interesting details including:
- Only 46% of Republicans were willing and able to follow the instructions of one study which asked them to list the advantages (if any) of carbon taxes before the disadvantages (the reverse of the "natural" order). 100% followed the instructions when asked to list the disadvantages first. Democrats and Independents generally followed the instructions in both cases.
- When this natural order was reversed - even for those Republicans who did not follow instructions - the framing effect was eliminated!