The Huge Link Between Food and Carbon Footprint

Over the last couple of weeks, Paul and I have been trying to hash out an estimate as to what portion of carbon footprint is related to food, both directly and indirectly.

I knew the number was going to be bigger than most people would expect, but other than that I wasn't really sure. After spending a couple of hours doing some humming and hah'ing with some charts in front of me, I came up with a rough estimate of 22%.

I thought 22% was pretty significant and then Paul took a look and said something like: "no way, it's gotta be closer to 35%."

And I thought he was crazy. So we agreed that I would set out to prove him wrong. After 4 hours of hard research and number crunching, I confirmed my opinion that I thought Paul's estimate was crazy. I was willing to raise my estimate, but no higher than 28%. Anything over that and I would have been very uncomfortable.

So I talked to Paul again and said "okay fine, more than 22%, but not 35%. I'm okay with saying 28%" and then Paul said something like "I was kinda thinking 40% but said less to be generous." And I thought this was crazier still. I thought there was little chance that I would crack 28% and no chance I would go above 30%.

So we spent an hour together going through all of the research that I did, debating the finer points of what it means to be related to food, and compiling spreadsheet calculations. As I was putting the last numbers into the spreadsheet, I felt pretty darn good about myself. I had made a few concessions but felt like I had somehow pulled more concessions out of Paul. Thus, I thought the spreadsheet was about to proclaim me victorious! Then I hit enter.

40%.

Nooooooooo! I thought I was gonna win!

---

Some takeaways from this experience:

1. It is not practically possible to perform a totally accurate calculation in this space. Part of that is a lack of detailed information and another part is the abundance of information that is misleading, whether intentionally or not. So this calculation includes a lot of opinionated guesstimating because otherwise we couldn't do it.

2. Although finding the exact number may not be possible, the information we were able to find, coupled with appropriate guesstimating, suggests that food is a HUGE part of carbon footprint. The calculator popped out 40%, but we're gonna go with 35% just to compensate for potential over-guesstimating. Given the amount of uncertainty in these calculations, it felt like the final number could have ended up anywhere between 20% and 50%, so it seems fitting to go with 35% in that sense too - right in the middle.

3. Arguing with an engineer is like wrestling with a pig in mud. After a while you realize the pig likes it. Paul and I both have engineering backgrounds... so this was kinda fun. Until I was wrong. :)

---

Paul and I are writing a book on building a better world in your backyard instead of being angry at bad guys. And one of the big things we talk about is growing your own food and the MASSIVE impact that food has on global footprints. We won't have room in the book (or in most readers' attention spans anyways) to include the in-depth reasoning behind how we came up with the numbers we did. So for anyone who reads the book and really wants to know how we got the numbers, we're going to send them to this thread.

---

Now I'm going to attempt to explain how we arrived at these numbers. Please keep in mind how squishy all of these numbers are, both the ones we come up with and (most likely) the ones that others came up with. We're shooting for a ballpark number, not 4 decimal places. Please please please don't look at these numbers and complain about them because you live your life differently. It's okay that you live your life differently than the numbers (I sure do), but we're just trying to capture the "average person."

With that now said, let's begin with this chart:



Let's start off with transportation, which according to this chart accounts for 27.2% of US greenhouse gas emissions. Here is a chart with data from the US Department of Transportation on transportation greenhouse emissions by source:



If you consider that some light-duty trucks and some passenger cars are used for business purposes, we figured that the emissions would be pretty well split 50/50 between "commercial" and "personal" use.

---

Starting with personal use, I created this chart from 2009 data on the average number of miles travelled for different purposes from the NHTS:



I was surprised by these numbers. I thought that commuting to work would be more than 50%. Apparently people drive a lot more than I thought. So in terms of food, it isn't nicely laid out for us. Let's go into each category:

Shopping: We agreed that for most people, 85% (or more) of the shopping they drive out to do is going to be for food. For me personally this might approach 98% right now - I've never really liked shopping.

Family & errands: Visiting grandma and paying the bills. Not much food involved... except of course the best time to visit grandma is meal time. But it's hard to know what "personal errands" all includes when shopping is separate. I'm having trouble coming up with ideas besides paying the bills (now done online by many people) and dry cleaning (which a lot of people don't do). I asked my wife and we figured that somewhere between 30% and 50% of this category would include food. So let's go with 40%.

To/from work: At first blush I was going to say 0. And Paul and I even agreed on it. But then as I am writing this I remember that most people go to work so that they can pay for things... including their food. On average Americans spend roughly 10% of their budget directly on food, so if we're going to talk about indirect food footprint, 10% of the work mileage should count towards food too.

Social and recreational: Going to the bar, going out for a meal, going to the theatre (which has "food"), going to a sporting event (which has "food"), etc. Maybe a road trip which will also involve some food. Obviously not all of some of these should be counted in the food category, but it is important to note the relation to food. We debated this one quite a bit but in the end I think we agreed upon 70%.

Other: I'm really unsure as to where else people are driving that don't fit into the other categories... but we figured probably somewhere around half of this category must be related to food somehow.

Running the math, the percentage of personal transportation related to food comes out to be just over 50%.

---

Moving on to non-personal transportation. I was unable to find any helpful information that broke down trucking emissions by cargo type. So this devolves into more of a guesstimate.

I am going to make a claim here that might not be quite true but I think has a lot of truth to it: If you follow the countless links and connect the dots, all industrial and commercial products and services eventually connect back somehow to the personal consumer. So if we look at what the consumer is consuming, we can get a feel for how everything on the back end is related.

We figured that if people are doing 85% of their shopping for food, it would be reasonable to assume that 85% of the stuff that they bring home was food. But a bag of apples is smaller than, say, a couch so the weight distribution might not be as high as 85%. But people generally buy less couches than apples, so we figure that at least half of the weight of products brought into the home will be food.

So if half of the weight of products being brought home is food, it might be that half of commercial transportation will in some way, shape, or form be related to food. That might be in shipping food to processing plants, shipping food to stores, shipping boxes to put food in, shipping farm equipment to dealerships, shipping steel with which to build the store, shipping office supplies to the lawyer who works for the food corporation... the list goes on for a long time. Yes this is a huge generalization. But I think it's a realistic one.

---

So we think that half of personal and half of commercial transportation is directly or indirectly related to food. What follows then is that half of all transportation is related to food. Going back to the chart at the top, half of 27.2% is 13.6%. So just looking at transportation, food is already 13.6% of our overall footprint.

---

Next up on the big chart from the top is electricity and heat for residential buildings.

A refrigerator and freezer usually take about 5% of home energy use. Then there's usually lighting in the kitchen, an oven, a stove, and hot water for dishes. At least. Combined, we figure this might account for 10% of residential building emissions, which are 15.3% of overall emissions. So not much here, but there's another 1.5% of overall emissions related to food, bringing our running total to 14.1%.

---

Electricity and heat from commercial buildings account for 12% of overall emissions. At first I thought this had nothing to do with food. And then I thought of grocery stores. So there's a little bit. And then I thought about restaurants, so there's some more. And then we got to talking about indirect relationships with food. And then I realized the percentage was likely much higher than I thought. So I found this chart:



This chart shows some interesting things. I think that when we're talking about emissions footprints in commercial buildings, it's a pretty fair rule of thumb to say that they are proportional to floor space. Paul and I went behind the scenes to the data that produces this nice image in order to calculate what percentage might be related to food. What follows are some of the thoughts and ideas that went behind our estimates. I think it is possible that some of these numbers are high while some of them are low. I think it'll balance out for the most part.

Office - 50% - When we first started this discussion I thought it was completely absurd to say 50% of office space was related to food. Then we started talking about all of the lawyers, accountants, managers, insurance people, investors, media, engineers, software developers, web designers, etc. that are directly involved in the food industry. And once we started down the rabbit hole of discussion about the indirect associations to food, it was no longer so difficult to believe that 50% of office space might be related to food.

Warehouse and storage - 50% - As discussed earlier, we think that 50% of commercial transportation will somehow be related to food. It seems a reasonable assumption to suggest that 50% of warehouse and storage space will thus also be related to food.

Service - 20% - This category was difficult to break down because we had a hard time defining what's all included. Things like laundromats, hair salons, and such places came to mind, but there might also be some service-based businesses that are somehow related to food. So we gave it 20%.

Mercantile - 40% - This will include places to buy clothing, books, furniture, etc. But it will also include places to buy cookware, tools (used to build things like offices and restaurants), and other items that are directly or indirectly related to food. On the surface it seems like it might be 10% related to food but I suspect that once you dig further it is closer to 40%.

Religious worship - 10% - Many places of religious worship have a kitchen that is active in serving the community somehow.

Education - 30% - Many places of education also have places where one can buy food. But it is generally a small portion of the floor space. Want to get into the indirect? It gets uncomfortable. For example, what if someone is in school to get an MBA and goes on to manage a grocery store?  As much as I would rather not accept it, I think this means that at least a part of the classroom being kept warm for this MBA student is related to food. I think if you really go down the indirect rabbit hole this could be more than half, but I'm going to go more conservatively with 30%. Good luck ever measuring this one...

Food service - 100% - Finally, an easy one.

Public Assembly - 10% - I'm thinking of places like conference centres, arenas, and movie theatres. They aren't entirely about food, but food is often a part of the experience. And then there's sure to be some areas of public assembly that don't have food.

Vacant - 50% - It seems like a lot of vacant buildings we see are restaurants. And if occupied offices are 50% related to food, it stands to suggest that 50% of vacant offices should also be related to food.

Food sales - 100% - Hurray, another easy one.

Lodging - 5% - Most hotels, motels, etc. have some sort of bar and/or restaurant involved. The portion of restaurant for a big hotel might be quite small, but for a small motel might be larger.

Health care - 5% - I've never received a food at the doctor's office. Or at the dentist. Or even when visiting the naturopath. But I have been to hospitals which nearly always have a place where visitors can grab a bite to eat. And many of the patients staying in the hospital are served "food."

Other - 20% - I'm always curious as to what fits in this category. I'm going to be say 20% even though it might be a bit less but could be quite a bit more.

Public order and safety - 5% - Yes the police are probably going to have a coffee maker... but that's nitpicking. I was just about to give this a 0 and then I thought about prisons. And people there are served "food." So I'm going to go with 5%.


Running all the numbers, this comes out to suggest that 30% of commercial building emissions are related to food. 30% of 12% of overall emissions means another 3.6% of our footprint is related to food. The current total: 17.7%.

---

Moving on, let's skip ahead to the portion of the chart explicitly labeled "agriculture." I was going to say it's safe to say this is 100% related to food. And then someone kindly reminded me that this category also includes things like cotton, flax, and silk that are used to make textiles. But some textiles are used for things like cushions in restaurants, so part of that category is still related to food. However, in recognition of things like clothes and other potential "agriculture" emissions that are not related to food, I'm going to go down to 95%. 95% of 6.2% is 5.9%, bringing the running total to 23.6%

---

Let's tackle the last one on the chart and then go back to the ones I skipped. Waste. Primarily methane from landfills. This time I was able to find some info online that seemed like they had done some homework. It seems like 75% of emissions in this category are pretty much directly related to food - the food that is wasted and put in a big pile of junk instead of returning it to the soil where it belongs. There will be some indirect waste too, but let's just keep the number at 75% to be generous. That's another 2% for the overall footprint, bringing the running total to 25.6%.

---

Okay. Almost done. Kinda.

The remaining categories can all be lumped together under the category of "industry." Let's talk briefly about each piece and then I will suggest a number to cover these category.

Unallocated Fuel Combustion - Who knows what this is? Not me. It's unallocated. But since it involves fuel combustion and a chunk of fuel combustion is related to food, there's going to be some food connections here too.

Iron & Steel - At first I thought there wasn't much to be said here for food. Then I remembered that half of all office buildings are related to food. And big boats ship food. And trucks transport food. So there's a food connection here.

Aluminum / Non-Ferrous Metals - My first thought that most of this would be things like aircraft which have a relatively small relation to food. And then Paul pointed out pop cans. Sure they are small, but millions of Americans buy mounds of these things every year. So there is a relation to food.

Machinery - Combines, tractors, food processing machines, etc. And then there's the machines that make those machines, and the machines that make those machines, etc. How grotesque.

Pulp, Paper, and Printing - Books, newspapers, and magazines. Not a lot of food there. But for every book most people bring into their house there are recycling bins full of cardboard boxes that used to be filled with food.

Food & Tobacco - Umm... well... food's in the title so I'm gonna say there's a relationship. Not really sure what to say about tobacco.

Chemicals - The amount of chemicals used by the food industry is huge, both directly and indirectly.

Cement - Oh cement. If it's used in a building, some of those are related to food. If it's used for a road, those are used to transport food. Definitely some relationship to food here.

Other Industry - Ah, the old "other" category. It might have nothing to do with food. On the other hand it might. Hard to know.

T&D Losses - A strange category. Partly related to weird power conversion loss and partly to people stealing electricity. And probably some other things in there too. Maybe some relation to food but who knows? Not me.

Coal Mining - Generally to create electricity eventually, some of which is used for food and food-related activities.

Oil/Gas Extraction, Refining, Processing - Some of the end product is eventually used for food or food-related activities.


Okay, so nearly all of these things have some relation to food, direct or indirect. Some of them are going to have high numbers for food, like chemicals, while others might be slightly lower. Paul and I debated this one for a while. Eventually we agreed to say that we think 40% of these remaining emissions are going to be directly or indirectly related to food. It could be higher, but it doesn't feel like it would be much lower. 40% of the remaining 36.7% is 14.7%.

Adding it all together, it comes out to just over 40%.

---

I warned you in advance. There was a lot of guesstimating here. And there are so many ways that some of these numbers could be off in either direction. As we were hammering this out I felt very strongly that I wanted to be conservative in my estimates so as to not be accused of trying to make up and skew these numbers to make the food footprint seem far bigger than it really is. But somehow we still ended up at 40%. I still have trouble believing it.

As I mentioned earlier, for our book we're going to go with 35%. The main reason for 35% instead of the 40% we came up with is that I am trying to compensate for any overzealous estimating that we may have done, even though we tried really hard not to.

Whether the true exact number is 34.21%, 41.08%, or 38.42%, I hope that this exercise illustrates the immense role that food plays in our overall footprints. And shows the incredible impact that we can have on the world just by growing our own food in a way that heals the land instead of destroying it. We have a job to do. Let's get to work.

Hi Shawn, great job trying to put numbers to something so vague.  I have an engineering background so I like the mud as well.  Sorry in advance but here are some things to possibly help bring down your numbers:

Family and Errands includes a lot of schlepping the kids around to stuff.  While they might have orange slices at soccer practice, I'd think the reasonable number for many of those trips is 0.

I think the supposition that since 85% of people's vehicular shopping is food means that 85% of all shopping is food may be off a bit.  Most of the shopping I do on-line is because it isn't food.  All those Amazon trucks are primarily moving around non-food items.

For the Agriculture = 95% one, don't forget about ethanol.  Tons of corn is diverted from the food stream in order to make fuel for cars.

I think a better way to approach this problem is to segment out the carbon footprint of the agriculture industry (incl. transportation, emissions, etc) and divide by the average American's footprint. It doesn't include stuff like refrigerator use, but I'm willing to bet all of those things add up to a lot less than 10% of a person's footprint.

If the average American's footprint is 20T, and a meat eater's food footprint is 3.3T, the number is a lot closer to 17%.

More importantly, I think the real lesson is this is a place that averages tell less information than individual data points. I don't drive to the grocery store, so… and if you don't eat factory farmed meat then… etc etc. The variance in each footprint is so high that an average only hides the true distribution.

So we think that half of personal and half of commercial transportation is directly or indirectly related to food. What follows then is that half of all transportation is related to food.

I think your numbers here could use some work. The most important is that official EPA carbon emissions from transportation explicitly do not include agriculture operations (tractor-trailers moving goods from farm to distribution centers qualify here). From the EPA:

Note: Totals may not add to 100% due to rounding. Transportation emissions do not include emissions from non-transportation mobile sources such as agriculture and construction equipment. “Other” sources include buses, motorcycles, pipelines and lubricants.

The other problem is that the emissions don't correlate to tonnage of goods moved. Ships and boats represents a small portion of total emissions, but represent a huge percentage of food moved, especially since many population centers in the US are on ocean ports. I think more research needs to be done here before it's fair to say half of all transportation emissions go toward food.

My understanding is this book is for people levels 0-4 on the Wheaton eco-scale, with a huge focus on levels 0 and 1.  

I don't know if I'm correct, but that's the thought in my head when I write this.

I'm a pretty solid level 3.5 on the eco-scale.  One thing I hate is how the carbon footprint gets broken down and complicated.  I don't need it to be complicated.  I just want to glance at it and see big general things - here is something I can change in my life and make a big difference.  

When I talk to eco-scale 1 people, they want things even more general.  They are at the "tell me what to do" stage.  They want to make a difference, but they are too busy to research all the details.  They just want 5 simple things they can change in their life to help themselves feel good.  Thanks to advertizing, they usually choose 1. ride a bike (saves money and is faster commuting), 2. LED lights, 3. Eating local, 4. bring their own grocery bag, and 5. buy lunch at whole foods once a month.  All these have the appearance of being green and saving money.  

When communicating to these people, they gloss over when things get complicated.  They don't make any changes.  Footprints are designed to be simple concepts that people can use to change their lives without having to get depressed by how big and complicated the problem is.



That said, I do enjoy geeking out about datasets from time to time.  There are some neat ones out there that break down global and US agriculture into the different products it is used for.  Agriculture isn't just food production.  One data set I saw had textiles as half of industrial agriculture carbon emissions.  But I have a statistician in the family and he always said to his clients "don't give me the questions you want asked, give me the conclusion you want proved, and I'll prove it".  One time, he had two clients with the same data set, each one wanted the opposite thing proved.  Easy.  There's a lot of different ways to interpret data sets, most of which are accepted in the scientific community.  I think the public is beginning to understand this - we get one study saying coffee is the best thing in the world for us, then next week, coffee is the most deadly substance ever.  The populus is number-numb.  

When you start talking about food you have to include the emissions of CO2 from the plants that grew to provide the food, the soil organisms also emit CO2 that has to be accounted for in calculations.
Then you add in the emissions from the tractors that till, spray, harvest, the Big rigs that transport to the Silos the tug boats that move the filled barges, then the ships that take product overseas and then you can add in the domestic portion of those too.
(I believe the USDA did this type of study not to many years ago, but don't quote me on that)

Good job of doing a statistical analysis of this problem area that is really world wide.