Taking everything into account . . .

. . . you can't.

New research is saying that local foods don't necessarily have a lower carbon footprint. When you consider technical economies of scale (i.e. more efficient transport -- containers on trains instead of boxes in pickups), mass-produced fruits and vegetables shipped across the country use less gas than local ones from small farms. This is based on, for example, the gas per strawberry. Many strawberries makes light footprint.

There are other things the article didn't go into (machinery efficiency, farm supply shipping, etc.) that large monoculture farms have going for them as far as carbon footprint goes. With all of these things, I have to admit that buying all your groceries at Wal-Mart, especially with only one trip a week, uses the least carbon.

If that's your goal.

Me, I eat locally for other reasons. I want to support local farmers and local industry. I want the flavor of the heritage varieties (wonder why they taste better than the shipped ones? flavor and storage value are inversely related for most fruits and veggies). I want to know where my food came from.

Yes, I feel guilty when I drive my car all alone to the farmer's market, only to go to the grocery store once or twice a week on separate trips. I used to bike it, but that was when I lived in Kansas (ah, flatness; the prospect of hauling my shopping up two blocks at 40% grade or more is rather discouraging). I used to do all my shopping on Saturday morning, going without rather than doing quick runs for one ingredient. I need to get back into at least some of those good habits.

Still, I can't help thinking that basing a drastic shift in behaviors (an all-local diet, for example) on one good-hearted idea is dooming it to failure. As soon as you start bringing in other calculations, you'll run into complications, realize how complicated life is, and give up. Yes, we need to consider confounders. We still need to act, though, and it's up to you what confounders you use in your analysis.