When we think about food, how often do we think about what it's going to do for us (in terms of nutrition, taste, satiety), and how often do we focus on what was required to get it to our tables?

Back when I was a wee chemistry student learning how to solve problems in thermodynamics, my teachers described the importance for any given problem of identifying the system and the surroundings. The system was the piece of the world that was the focus of the the problem to be solved -- on the page or the chalkboard (I'm old), it was everything inside the dotted line you drew to enclose it. The surroundings were outside that dotted line -- everything else.

Those dotted lines we drew were very effective in separating the components that would get our attention from everything else -- exactly what we needed to do in order to get our homework problems done on a deadline. But it strikes me that sometimes we can forget that what we've relegated to surroundings still exists out there in the world, and indeed might be really important for other questions that matter, too.

In recent years, there seems to be growing public awareness of food as something that doesn't magically pop into existence at the supermarket or the restaurant kitchen. People now seem to recall that there are agricultural processes that produce food -- and to have some understanding that these processes have impacts on other pieces of the world. The environmental impacts, especially, are on our minds. However, figuring out just what the impacts are is challenging, and this makes it hard for us to evaluate our choices with comparisons that are really apples-to-apples.

A couple years ago at the New York Times Room for Debate Blog, a bunch of commentators were asked to weigh in with easy-to-make changes Americans might adopt to reduce their environmental impact. One of those commentators, Juliet Schor, recommended eating less meat:

Rosamond Naylor, a researcher at Stanford, estimates that U.S. meat production is especially grain intensive, requiring 10 times the grain required to produce an equivalent amount of calories than grain, Livestock production, which now covers 30 percent of the world's non-ice surface area, is also highly damaging to soil and water resources.

Compared to producing vegetables or rice, beef uses 16 times as much energy and produces 25 times the CO2. A study on U.S. consumption from the University of Chicago estimates that if the average American were to reduce meat consumption by just 20 percent, that would be the equivalent of switching from driving a Camry to a Prius.

Americans currently rank second in world in meat consumption, weighing in at 271 pounds a year, up from 196 pounds 40 years ago. And that doesn't include dairy. We get an estimated 75 grams of protein a day from animals, and 110 grams total; the government recommends only 50 grams a day.

Back in 2007, an article in the New York Times made a similar claim, that "switching to a plant-based diet does more to curb global warming than switching from an S.U.V. to a Camry." This particular article (which appeared in the "Media & Advertising" section) was quite short on details as far as numbers go. As I'm not an agronomist, I don't have all the relevant numbers at my finger tips, but I made a preliminary attempt to set up some equations into which reliable numbers can be plugged once they are located.

First, there's some unclarity in the sentence quoted above as far as what exactly is being compared when we're weighing the relative importance of choices about driving and choices about eating. Is it a comparison of greenhouse gas (GHG) emissions? Petroleum consumption? The whole bundle of factors that have an effect on global warming?

Also, which SUV and which Camry are we talking about? What animal sources of protein and what plant sources of protein? There are lots of specific details that could make a significant difference.

Your driving decision's impact seems like it could be calculated as:

((Average miles/day)/(Average miles/gallon)) x (kg GHG emitted/gallon of gas consumed)

Of course, if you're just tracking how much gas you're burning, don't multiply by the last term.

I don't know whether there's a uniform rate of GHG emitted per gallon of gas a car consumes -- either for a single vehicle or vehicle type or across many vehicle types. In the absence of expert opinion on this, I'd probably assume it's a constant. (If you have an expert opinion on this, please chime in!)

I drive neither an SUV nor a Camry, so my gas efficiency numbers here depend on the accuracy of what others have told me. A pretty fuel efficient SUV gets maybe 20 miles per gallon. A Camry is probably not too far from a Corolla in average fuel efficiency, so we could estimate that at around 30 miles per gallon. For someone with a daily commute about the same distance as mine (a 50 mile round-trip), this hypothetical commuter would burn 2.5 gallons of gas with the SUV and 1.7 gallons of gas with the Camry. However much crud is being emitted from the burning of the gas (assuming a constant rate of crud per gallon of gas consumed), the SUV commute puts out 47% more of it than the Camry commute.

That was the easy calculation. Now for the food.

To simplify, I'm going to deal with only one macronutrient, protein, since that's the one people seem to be most interested in getting from their meat. (The fat is like the prize at the bottom of the cereal box.) If we're sticking to GHG as the thing we're comparing, we'll be looking at:

(kg protein consumed/day) x ((GHG emissions/kg to produce the protein source) + (GHG emissions/kg to transport the protein sources))

The USDA says that my daily protein intake should be around 50 g (or 0.050 kg), but it doesn't tell me what particular source of protein to favor, or how to apportion those 50 g among different potential sources. And depending on the source, the balance sheet may come out quite differently.

An obvious source of GHG is transportation (on trucks or planes or cargo ships), either within the process of raising the food or of the "finished" food product to market. For example, the GHG emissions involved in producing 1 kg of beef, pork, chicken, or (farmed) fish may all include GHG emissions related to shipping corn to feedlots. However, cows, pigs, chickens, and fish have different rates of conversion of grain input to flesh on their bones. Also, ruminant animals like cows also put out their own GHG emissions via farting; chickens don't fart as much.

Note that you can't even easily pin down GHG emissions in the production of 1 kg of a particular kind of meat. Corn fed beef will have GHG emissions from transport of the corn to the feedlot, while grass fed beef will presumably be grazing on the grass where they are, eliminating some of the GHG emissions. (Do they ship grass -- or truck cattle that are fed grass considerable distances to the grass?)

For a plant source of protein, there's less farting (by the plants, anyway), but there are still GHG emissions associated with shipping. In addition, to the extent that your plant based protein might be processed (like tofu), it's quite conceivable that some part of the processing might involve GHG emissions.

Really, without specifics about what you're eating, how it was fed or produced, and how much fuel was consumed getting it to you, it's impossible to hold up a kilogram of protein from one source and a kilogram of protein from another and say anything very sensible about which contributes more to global warming. Locally grown whole soybeans probably contribute less GHG emissions than beef from gassy cows fed corn trucked in from 1000 miles away, and then itself put on a truck for hundreds of miles to get to the store. But it's not obvious that soybeans that are shipped 1000 miles to make tofu and then shipped hundreds of miles to get to the store would lead to lower GHG emissions than an equivalent amount of beef protein from locally raised grass fed cows.

Of course, there are other inputs to meat and plant based proteins that are environmentally relevant -- like how much water is required to produce 1 kg of protein of each sort, how much soil is degraded in the process, and so forth. But if this is included, how can we ignore the other environmentally relevant inputs (and outputs) involved in the production of the Camry and the SUV? In a fair fight between changing what you drive and changing what you eat, you'd have to take this all into account.

Indeed, even in a vegetarian household, where none of your protein is coming from animal flesh, there are plenty of perplexing comparisons about the environmental impacts of our food choices. The labels in the supermarket don't list how much water, land, or petroleum-based fertilizer went into producing your pound of potatoes or peas. Nor do they reflect the amount of energy used to process your tofu, nor the natural resources to can your chickpeas, nor the fuel to ship your bananas.

All of which makes it very hard to know how to make better choices about the foods we eat.

My family eats more legumes than we used to. Nutritionally, they're a slam-dunk, high in fiber and protein. And, since legumes tend to enrich the soil they're grown in, I'm hopeful that we're expending less fertilizer than we would be otherwise. So far, so good. But now, in our quest to be environmentally responsible, should we be buying beans in cans or bags of dried beans?

The dried beans need to be soaked in water, then boiled in water (and it's generally recommended that you don't boil them in the water you soaked them in). So, here you're consuming water (which we're very short on in California) plus whatever source of power you use for cooking -- just to get your legumes ready for your recipe.

Canned beans, on the other hand, don't need to be soaked or boiled ... at least not at home. I'm unclear on what state the beans are in when they enter the processing plant -- maybe they're freshly picked, but maybe they're dried beans that need to be soaked and boiled at the factory before being canned. (The latter option strikes me as a silly one, but as a consumer, I can't rule it out.) The beans are packed in steel cans -- it takes both metal and energy to make those cans and to put the beans in them. And while steel cans are recyclable, recycling requires energy inputs, too.

Indeed, there's packaging for the dried beans, too -- plastic bags -- which requires petroleum and energy to make. And unless the legumes are sun-dried, there's likely an energy input required to dry the beans.

So, dry or canned? Which choice is more responsible? How on earth is a consumer supposed to evaluate this choice?

The cooking method introduces another decision point where I don't have as much information as I'd like. Lately, I've been rocking the slow-cooker. It tends to use less water than cooking stuff in a pot on the electric range, but over a much longer cooking time. Am I saving water only to use more electricity? Is there a good way to evaluate this?

None of this is to say that we shouldn't bother thinking about how our choices impact our climate or environment. To me, the take-home message is really how alienated we are from where things come from, what's involved in making them and getting them to us, and what the hidden tolls might be from the choices we make on autopilot.

Milton Friedman said the only thing we need to know about a product is its price. Maybe that's true if we're just interested in buying and selling stuff. But, seeing as how we live here on the planet we're using to make and move around the products, we might be better served by the reminder (from a character in the movie Shallow Grave) that knowing how much money you paid for a product is not the same as knowing what it cost you.

Even if the complicated path our protein sources (and cars) take to come into being make straight comparisons hard, it probably wouldn't hurt us to notice some of the costs that are incurred before we make our purchase. But given that consumers have a growing interest in these costs, it would also be good if the producers of our food (and are cars, and other stuff) were more transparent about them.