The venerable Mr. Green is correct in assessing the environmental benefits of the decision -- which obviously, after the first 50,000 miles, favor the new Prius.  But he doesn't look at the cost of achieving those benefits.  Factoring in costs, buying a new Prius loses a bit of its luster.

Let's assume, for the sake of argument, that Heath drives about 10,000 miles per year, that his Accord gets 25 mpg in actual driving, and that a Prius would get 50 mpg.  (The gas mileage figures are the ones that Mr. Green seems to be using, so I'll grant him that.)  After 5 years, the Accord has burned up 2,000 gallons of gas; the Prius 1,000, plus an additional 1,000 gallons equivalent consumed during the car's construction. 

In other words, from a carbon standpoint at least, 5 years is the environmental break-even point.  Before that, and the Accord emits less carbon overall.  After that, the environmental benefits of driving a Prius start to add up.

Now, let's look at the costs.  Right now, a fairly bare-bones Accord is worth somewhere in the vicinity of $1,200 to $2,500, according to the Kelley Blue Book.  Let's say that after 5 more years of driving, its value declines to $0 -- which is a pretty good assumption, given that a 20 year old Accord with 200,000 miles on it has a resale value of $0 (see here).  And let's say that to keep the older car running for 5 more years will require an extra $1,500 in repairs and maintenance, above and beyond the cost of maintaining the Prius.  And finally, let's assume that the average cost of gas over the next 5 years will be higher than today--say, $3 per gallon, or a total of $6,000 to fuel the Accord for 50,000 miles of driving. 

Heath would save about $3,000 in gas costs over 5 years if he goes with the Prius.  But the Prius is expected to depreciate much more than the Accord.  By one estimate, a new 2005 Prius will lose about $9,200 in value over the next five years.  (This is from a site that rates the Prius "The Best Overall Value of the Year" for a midsized car.)  These depreciation costs will likely overwhelm the savings on gas and repairs.  I'm no expert, but looking through the numbers, my best guess is that after 5 years, the total cost of buying, owning, and driving the new Prius is about $4,000, over and above the costs of holding on to the Accord.

That is, after 5 years Heath has spent $4,000, and achieved no net benefits for global warming. 

So one strategy might be for Heath to hang on to the car, and commit to spending the $4,000 on something else -- weatherizing his house, buying carbon credits or green electricity, or (tee-hee) donating it to the Sierra Club.  Then, in 5 years, he can buy a used Prius.  From the standpoint of Heath's own personal ecological footprint, he's coming out $4,000 ahead -- or, really, whatever $4,000 can buy.  (And, by the way, if used wisely $4,000 can buy a lot.  On the European carbon market, a ton of CO2 costs about $19, at today's prices and exchange rates.  So $4,000 would offset a little over 200 tons of CO2 -- about what a Prius emits if driven 10,000 miles a year for 100 years.)

Now, obviously, there are tons of caveats here.  Heath could drive much more than 10,000 miles a year, which could make the Prius a much better buy.  And we can't know what fuel prices are going to be like over the next five years; if prices rise quickly, the Prius' cost advantage on gasoline might be greater, and the car might not depreciate so fast.  In fact, there've been some reports that used Priuses have sometimes sold for more than new ones, since there's no waiting list for a used Prius, but until recently a pretty long waiting time for a new one.  Also, an old car -- even one in good condition -- can have unexpected and costly repairs.  So there's definitely some risk to hanging onto the Accord.

In addition, there may be some additional benefits to buying a Prius now rather than later.  Buying now stimulates the market for hybrids, which is a good thing.  And hanging onto the Accord may make Heath a bit of a free rider: after all, if everyone followed this reasoning, then nobody would buy a Prius, which would defeat the "buy used" part of this strategy.

But my real point is that you shouldn't assess an environmental choice simply by looking at the benefits -- you have to look at what opportunities you're giving up in the process, and that means assessing the costs, too.  Once you understand both costs and benefits, you're in a position to make the best and greenest buys first -- which is a strategy that even Mr. Green should get behind.