Home delivery of groceries produces far lower CO2 emissions than driving to the supermarket: Erica Wygonik and Anne Goodchild find this in a recent study of the Seattle area (thanks to Tanya Snyder at Streetsblog for the reference). Wygonik and Goodchild cite similar findings from Sally Cairns in the UK, Hanne Siikavirta and colleagues in Finland, and Tehrani and Karbassi in Iran.

These studies find that the CO2 savings can be as high as 80-90%, if the deliveries are scheduled to minimize driving, which means the customers don’t pick the times; if customers do pick their own times, the savings are lower, though still considerable. Also, it’s 80-90% not of the transport cost in the whole supply chain, but just of that required to get the groceries from the supermarket to your house.

Now think of three different logistical systems for the delivery of groceries:
(1) Car to supermarket (big trucks take groceries to a supermarket, customers drive to the supermarket in their own cars).
(2) Home delivery (big trucks take groceries to a supermarket or depot, small trucks take them to customers’ homes).
(3) Walkable retail (medium-sized trucks take groceries to small shops, customers walk to the shops).

Wygonik and Goodchild and the studies they cite are all comparing (1) and (2); I’m more interested in the comparison of (1) and (3), but knowing something about (1) and (2) helps with that.

I made the assumption, in this paper (forthcoming in Spatial Economic Analysis), that walkable retail (3) saves a lot of CO2 in comparison to car-oriented retail (1); my paper is not about CO2 savings – it is about the conditions under which walkable retail can also be price competitive, but one of the reasons that is important is that when people walk to shop they produce less CO2. I never really doubted the validity of this assumption, but from a purely logical standpoint it is not watertight: walking to shop of course means lower direct CO2 emissions by shoppers, but the medium-sized delivery trucks going to a lot of small walkable shops will surely produce more CO2 than a smaller number of bigger trucks going to a few supermarkets. Is it possible that the apparent logistical inefficiency of small shops – all those deliveries by medium-sized trucks – would offset the CO2 saving from reduced driving by shoppers?

What home delivery and walking to shop have in common is that the consumer doesn’t drive to get groceries, so when we compare (2) and (3) we just need to ask which produces more CO2: delivery vans going from house to house (2), or a smaller number of delivery trucks making a smaller number of stops at local shops (3)? The answer, certainly, is that home delivery produces more CO2 than walkable shops:

CO2 from Walkable Retail < CO2 from Home Delivery
and, as the studies cited show,
CO2 from Home Delivery < CO2 from Car to Supermarket

In other words, walking better than web, web better than driving.

Notice, too, that walkable shops should produce less CO2 than even the most carbon-stingy home delivery variant, where deliveries are scheduled by the retailer with the aim of minimizing transport costs – the mode that saves 80-90%. When customers can request delivery times, the CO2 costs from home deliveries rise, motorized deliveries to each doorstep will produce more CO2 than deliveries to walkable shops.