Much of the variation in cost estimates was due to the following factors:
the level of international trading of emissions permits assumed;
how efficiently high carbon energy sources are replaced by less carbon intensive alternatives;
how much innovation in alternative energy sources and conservation could be expected and when substantial efforts might begin;
None of these models assumes that there will be any technological breakthrough in alternative energy sources. Only direct costs were considered here, both secondary costs and secondary benefits of an improved environment were excluded.
The final range of estimates of cost increases to the consumer, by government and nongovernment sources, were from less than $10 per month to over $100 per month per household. The average monthly cost estimated was approximately $50.
Consistent with these estimates in the poll we began by asking respondents to evaluate the amount of $50 a month per household as the cost of compliance with the treaty. This was also done to avoid possible bias introduced by starting either at a high cost or low cost. Based on their response respondents were subsequently asked about amounts as low as $10--just above the most prominent estimate of the administration--and as high as $100.
We also asked respondents to evaluate the tradeoff between paying $10 a month with flexible implementation (i.e. with credit being given for reductions being made in other countries) and $50 without. These numbers were derived as follows. Since the estimate of $70-$110 a month has been used widely by the administration as the estimate of the costs of compliance with a full regime of trad ing of emissions rights we used this as a baseline. This was rounded up to $10 a month. Conservative estimates of the ratio between the costs of full trading and no trading are approximately 5 to 1. Thus $50 was chosen as the monthly amount with flexible implementation. There are higher estimates of the absolute costs without trading, but we opted for the lower end of the spectrum so as to not make trading seem overly advantageous relative to the costs without trading. If we had used a higher baselin e with trading or a higher multiple without trading the effect in favor of trading, in the poll responses, would probably have been stronger than it was.