You report: "According to the California Department of Insurance, the Environmental Defense Fund estimates that if 30 percent of the state's drivers opt for the voluntary pay-as-you-drive coverage, the state could avoid 55 million tons of carbon dioxide between 2009 and 2020 -- the equivalent of taking 10 million cars off the road..."

It would be interesting to see their research. California has 37 million residents. One may assume, therefore, that "30% of the state's drivers" would not exceed 10 million people. So how would this work, whereby if 30% of the states drivers "opt for the voluntary pay-as-you-drive coverage" the state would avoid CO2 emissions equivalent of "taking 10 million cars off the road"? To accomplish such a result implies, using elementary calculations, that "pay-as-you-drive" coverage equates to not driving at all. It is absurd.

Proclamations from government agencies and nonprofits should be scrutinized with at least the rigor reserved for corporate announcements. Unless pay-as-you-drive insurance were considerably more expensive per mile (on average) than flat rate insurance, why should it result in any significant reduction in driving mileage?

Imagine that you paid a flat rate for gas - if you rarely drove, mostly you would be paying to cover the cost of someone else to drive more. That is the insurance system that we have now. Pay as You Drive lets people who drive less pay less - and everyone benefits from less pollution, congestion, and accidents from people who would choose to reduce their driving so that they can save money.

Not to belittle the observation that proclamations from government agencies should be scrutinized, but I think the figures actually do make sense. The figure for CO2 emission reductions applies to the running total for all the years between 2009 and 2020, and the 10-million-car figure refers to the equivalent of the annual emissions of 10 million cars. So, really, they're saying the CO2 savings are roughly equivalent to the total annual emissions of a *million* cars each year.