Friday, April 28, 2006

Buy Local vs. Shop Local

Buying local has become all the rage lately (see Tree Hugger, 100 Mile Diet,Grist, or Local Harvest). I think there are some good reasons to do it. I think you should buy local because it is fresher. I think you should buy local because it allows for crops to be farmed that are bred for taste rather than for longevity and therefore they taste better. I think you should buy local if having a relationship with a farmer is something that you want. But I am suspect whether you should do it for environmental reasons. You see lots of claims like the following:

Today, the ingredients for an average meal travel over 1,500 miles. As a result, a huge volume of oil is used to transport food that could be purchased locally. Buying locally decreases transportation and thus helps America to reduce its reliance upon foreign oil supplies.
So it made me curious, how much fuel does it really take to transport my food and how does it compare with the fuel that I use to drive to and from the store?

I first tried to determine how much fuel it took to move my food. In this report (.pdf), they take a look at how much food can be transported per gallon of fuel with 3 different types of trucks:

semitrailor = 6.1 mpg * 38,000 lbs of food = 231,800 lb mile/gallon
midsize truck = 8.5 mpg * 13,775 lbs of food = 117,000 lb mile/gallon
small truck = 17.2 mpg * 1,550 lbs of food = 26,600 lb mile/gallon

Although the semitrailor only gets 6.1 miles to the gallon, because it can carry 38,000 pounds of food, it is much more efficient in pound miles/gallon than a small truck that gets 17.2 miles to the gallon.

This report also says that the average distance food travels in the US is around 1,500 miles, and the average for locally grown food is 345 miles. I assume that the long distance food travels by semitrailor and the local food by small truck.

For each pound of "normal" food:
1,500 miles / 231,800 lbs mile/gallon = .0065 gallons per pound
For each pound of local food:
45 miles /26,600 lbs mile/gallon= .0017 gallons per pound

Buying local saves .0065-.0017= .0048 (or .005) gallons per pound.

I weighed my groceries this week and it came to 20lbs. I don't know what the average is for an American, but I bet it is somewhere around there. If I bought only local, each week I would save 20*.005=.1 gallons. For an entire year that take me to 5.2 gallons of gasoline. That is not nothing, but given the average American uses 450 gallons of gasoline a year, it doesn't seem like a whole lot to me. There seem like a lot of other ways that could save much more than that.

Then I wondered, how does the gas used to transport my food compare with that I use to drive to the store?

The "normal" food I buy took 20lbs*.0065 gallons per pound = .13 gallons to transport. My grocery store is 2.6 miles away, or 5.2 miles round trip. Lets assume my car gets 30 mpg, then it works out to 5.2/30mpg = .17 gallons. So it actually takes more gasoline for me to drive to and from the store than it does to transport it all 1,500 miles from the farmers to the store. Of course, if I was shopping for a family of four and getting 80lbs of food, then it would be .52 gallons to transport vs. .17 for the drive.

Then I wondered, it terms of saving fuel, how does buying local compare to cutting out a Trader Joes run, or a second run to the store a week, or buying on the internet and having it delivered?

For me, I would save .1 gallons a week buying local. My Trader Joes is 2.4 miles away or 4.8 round trip. That would use .16 gallons. So, I would be better off skipping TJs and just sticking to the closest grocery store even if that meant no local food. For the family of four, they would save .4 gallons buying local, so if TJs has the local food, it would be worth it. If on the other hand, they were running (and by running I mean driving a car) to 2 stores twice a week (lets say 20 miles round trip) and they could cut it down to 1 store once a week (5 miles) they would be saving .5 gallons, which is more than what they would save from the local food. I don't have numbers but I suspect that buying "normal" food via internet delivery (ala Webvan) would save more gasoline than driving to your local store and buying local, due to the efficiencies of delivery vans if you were buying 20 lbs of food.

So, then I wondered, if I could choose between a normal grocery store with "normal" food or local food at a farmers market, how much farther could I drive to the farmers market and still save gasoline in total?

Buying local saves me .1 gallon a week. At 30mpg, that goes 3 miles (for the round trip). I could only drive an extra 1.5 miles to a farmers market if I was trying to save gasoline (or 6 miles if I was getting food for 4). So much for the idea of driving cross town to the farmers market to help the earth.

My final "then I wondered" was: what if the food was shipped from China?

I found this webpage that took a look at how many ton miles you can transport per gallon with ocean barges (assuming no backhaul).
Size of shipNet ton-milesNet pound-miles
per gallonper gallon
30,000 dwt574.81,148,000
50,000 dwt701.91,404,000
70,000 dwt835.11,670,000
100,000 dwt1,043.42,086,000

So, lets call it 1,500,000 lbs miles/gallon (compared to 231,800 for the semitrailor or 26,600 for the small truck).

The 6,000 miles it takes to go from China would use 6,000/1,500,000 = .004 gallons/lb. This is actually less than the .065 that it took to transport 1,500 miles via semitrailor. So, assuming that there is limited use of semitrailor before or after being on the barge, it actually takes less fuel to ship from China than it does to ship inside the US. Of course if it was flown over (which you would have to do if it needed to be eaten fresh) then I think it would use much more fuel.

So my take away from the whole buy local thing is that it doesn't really save that much fuel. In most cases, it would be better to focus of shopping locally and infrequently (or walking to the store) than to buy local. But of course there are other reasons to buy local, like because those local blueberries are oh so scrumptious.


viral said...

interesting analysis! thanks ...

Anonymous said...

One point missed in this analysis is that the "local food" typically used in these calculations is fresh food. That is, fruits and vegetables.

You have to factor in more miles for anything that is processed. Tomato processors, for example, rely on regional processing facilities. Tomatoes can travel many miles in a variety of long-haul vehicles just to get to the processing facility. From there they are trucked again as completed products. It's a lot more complicated with "savory snacks" and cookies.

One wonders whether calculations include the CSA program, or local Farmers' Markets (typically within a very short drive of farmer and consumer).

You didn't itemize your groceries; it could have been all savory snacks. Without careful detail, the calculations you made are as good as drink-sodden scribblings on a cocktail napkin.

mping said...

Thanks for the comment. Couple of points in response.

First, you are correct that these calculations are rough so they could easily be off by a factor of 2 to 5. I wish I could have just referred you to a good study that was well designed rather than making an estimate myself, but there are none that I am aware of (please leave a comment if you know of any). If you go to the sites that claim a significant savings in oil by buying local, it is hard to find the underlying research that they base their claims on. That is why I wrote this, to give some perspective as to how much could be saved and how that compares to the gasoline used driving to the store.

Second, I am not sure what the 1500 mile number takes into account. I would assume it takes processed foods into account, but without seeing the underlying research (no link on the site to it) I am not certain. If it does, then I think the analysis still holds. Of course if you switch from processed to raw ingredients the weight of your foods probably goes up and you make other substitutions. I am not sure how this would affect the analysis.

If on the other hand, you were to keep buying your processed goods, and just substitute the non-processed goods like fruits, vegetables and milk (as I think many "buy localers" do), then the impact of buying local would be even smaller than I calculated.

Third, I think the conclusion that you should be just as concerned about the number of trips to the store you take as whether the food is local still holds. Making a special additional trip to the local farmers market to just buy a few fruits and vegetables is still likely to use more fuel in driving there than is saved by purchasing local goods. So I remain skeptical of any major oil savings that would occur from more people buying local.

Jonathan said...

This is a great back-of-the-napkin experiment and thanks for doing the legwork. I've commented on this over at Environmental Economics, but wanted to make sure you knew I appreciated your running through the hypothetical numbers. I'll quote some of my comment from there, if you don't mind:

One thing I haven't noticed mentioned is the question of what kind of travelling the ("local") farmer does under the two scenarios. The farmer will have to get their produce to market one way or another. If they sell through a farmers' market, then that travel impact has to be counted. But if they sell their produce through a grocery store, how's that produce getting there? I've got in mind a farmer Raleigh, where the orginal questioner lives. Unless we are going to argue that all farming should be done in California, and all land around Raleigh converted out of agriculture, the Raleigh farmer is going to sell their tomatoes somehow.

Where are the Raleigh area tomatoes to be sold?

Through what commercial establishment (farmstand, farmers' market, grocery store)?

Each of these involves travel by the consumer and/or farmer and/or third party transporter. Hey, J.S., is this something you could work into your estimates?


A related question: what's more greenhouse-gas efficient, canning my own tomatoes grown in my (Vermont) backyard or buying canned tomatoes grown and processed in California? J.S., can I convince you to take a crack at this one?

Anonymous said...

im confused w/ something . . . could anyone explain to me what is the difference between 'buying local' and 'shopping local'?

Fat Knowledge said...


Buy local refers to purchasing products, typically food, that were produced locally, typically within 100 miles.

Shop local refers to shopping at the closest store to you.

My point here is that if you are trying to minimize CO2 emissions, you might want to focus more on how far and how often you travel to go shopping, rather than on how far the products you purchase at the store have traveled.

Anonymous said...

oh.. i see!. . thanks for the information

Anonymous said...

1.) I think "Buy local" inherently encourages people to consume less packaged and more raw foods, so it's about changing the proportion of items, too.

2.) I think 20 lbs./wk. is too conservative - two bags of groceries per week for one person? How often do you eat out? I estimate 5/bags/week/person (~50 pounds), with a minimum of processed (cans/boxes).

3.) I think _most_ imported food is shipped in, but then has to be trucked cross-country from there.

4.) Many processed foods rely on multiple steps - ship from farmer to plant, ship ingredient from plant to processor, ship the product from processor to packager, ship to the regional distribution warehouse, ship to the store. Sugar packets travel 10,000 miles to get back to their source in Hawaii.

5.) I think a better analysis might include avg. mpg * avg. dist. to groc. store. (2.7 mi)

(I tried it, but with conflicting results).
pop: (Jul. 07 est) 3.01*10^8 ppl/US
#cars (04 survey, 05 est) 0.83334 cars/person = 2.51*10^8 cars/US
# gal. gas consumed (07 EIA data)
1.42*10^11 gal/yr
%commercial (05 data) 32% of all vehicles (most are diesel, though)
Avg. ann. est. miles/car (AAA 07) 15,000/yr. (Although I don't think this can include commercial vehicles, if you look at the result below.)
This would indicate average mileage is 26.7 mpg, which I think is high if the the median age of an auto is 8.9 yrs (06)
Avg. mpg/ _new_ passenger vehicle (04) 24.7mpg (from 23.1 in 1980).

6.) What about the potential fuel savings of converting "trucking" back to rail for long-haul?

Fat Knowledge said...

Hi Anon,

Thanks for the comments. Here is my response to your points.

1) I think everyone has their own definition of what "buy local" means. I think there are good reasons to buy local, I just don't think that reducing co2 emissions is one of them.

2) 2 bags of groceries per week is about right for me. I usually eat out 1 or 2 times a week. I don't buy fluids and I do buy quite a bit of processed foods, so maybe that explains the difference.

But, 50 lbs a week works out to around 8.5 lbs of food a day. That sounds like a lot to me.

3) Good point. I live on the west coast so for me it doesn't have to go far by truck, but for others that would be different.

4) Good point. If I had better data on how far food gets transported and how much co2 is emitted to make it, I would use it.

5) So, you would like to use 26.7 mgp rather than 30 and 2.7 miles to the store rather than 2.6?

Sounds fine by me, but it doesn't appear to change the result by much does it?

6) I don't know what the potential fuel savings of converting trucking to rail for long haul is, but that is a good question.

My original conclusion was that you should worry more about how often and how far you drive to the store rather than how far the food traveled to reduce co2 emissions. That taking an extra trip to go to a farmers market will not save any co2 emissions. I still think that is valid.

If you want to save co2 emissions, switching to local food is way down on the list of what you should do, behind getting an energy efficient car & house, and eating vegetarian.

But local fruits and vegetable are much more tasty, so I am all for it for that reason.

Matthew said...

eating vegetarian indeed. I would love to see an emissions comparison between a meat eating diet and a vegetarian diet. The use of fresh water in meat production might also shock you and is maybe more cause for concern than co2 emisions

Matthew said...

not to mention the abundant and stinky livestock 'emissions'

Fat Knowledge said...

Hi Matthew,

I compare the emissions of a meat eating diet and vegetarian diet in this post.

Post a Comment

Note: Only a member of this blog may post a comment.