Reducing emissions and saving lives

In this post I will look at food production and distribution under capitalism and its problems, and how a cybernetic planned economy might address both.

Food production under capitalism

Food production as it is done currently suffers from two major problems: overproduction and greenhouse gas (GHG) emissions.

Food waste varies somewhat between countries, but the global average seems to be between 30-50% (1, 2). If we take the number 33% then this means that workers in the food sector work 50% more than what is actually necessary. If food production were optimized then the working week for food workers could shrink from 40 hours to 30 hours or even 20 hours. This doesn't happen currently because it is more profitable to let food go to waste.

Food production also accounts for between 24-26% of greenhouse gas (GHG) emissions (1, 2). Since part of these emissions is purely waste, there is room for improvement even without anyone having to change their diet.

Food distribution under capitalism

Currently between 9-36 million people die due to starvation every year, depending on how you count (1, 2). This is not because we're not producing enough food, clearly, but because the food that is being produced is not getting to the people that need it. Beyond starvation, around 800 million people are malnourished, one ninth of the world population. The amount of suffering this entails is hard to fathom. From a purely economic perspective it is an enormous waste of labour and talent. Starving bodies means brains and muscles not working as they should. The labour that goes into raising children and adults, only for them to starve to death, is clearly an immense waste. Again, this loss of life is not because we don't have enough food, but because people cannot afford it.

The food situation has recently gotten worse due to the COVID-19 pandemic, with an estimated 6,000-12,000 extra deaths per day (2-4 million per year, 1). These numbers are also bound to get worse as the climate crisis evolves.

How we could do better

I should hopefully have made it clear in these posts that one important mechanism in cybernetic planning is constraints. Constraints allow us to learn whether a proposed plan is feasible or not. Once computed, it is in everyone's interest to work according to the plan, since doing so means plan goals are met with a minimum of labour. I will note here that the plan may well change from day to day, depending on how work is progressing and on the amount of computational resources available.

Food is a rather thankful good to plan for, since we know how much of it is necessary in any geographical location based to how many people live there. People have preferences of course, and these preferences vary by location, but a basic requirement is that people shouldn't starve. Only after this is ensured can we worry about having more varied diets.

I have previously shown that meeting basic macronutritional needs is relatively straightforward. There is also clearly enough food being produced to maintain a surplus in case of bad harvests or disaster. We have two problems to solve: distribution and GHG emissions.

Planning distribution

Distribution is a productive process where a set of goods is transformed from being in location A to being in location B. A and B might not be directly linked, so distribution will entail intermediate stops at more locations. This amounts to solving a multi-commodity flow problem, something which we have become quite good at as a species in recent years. Conveniently, flow problems can be formulated as linear programs and therefore can be plugged in to the same sort of equations as I have elaborated on in these posts. The problem then becomes knowing how much volume and mass each type of food occupies, how long before it spoils, how fast each mode of transportation is etc. Ensuring that everyone has enough food therefore means knowing that a feasible solution to the linear program exists, and that this solution is followed by everyone, at least roughly.

Planning GHG reduction

As I hinted at in the start of this post, one straightforward way to reduce GHG emissions is by ensuring only as much food as is actually necessary is produced. Since we know how much food people need, their preferences and how much we have in reserve, we can compute every year how much food should be produced. There is no need to produce more than this, because reserves have already been planned for. Based on the numbers presented earlier, potential GHG reductions can be estimated as follows:

  • assume the 1/9 of the world population that currently goes hungry is currently getting zero food (increase demand by 1/8th)
  • assume that only 30% of food is wasted
  • assume that agriculture accounts for only 24% of GHG emissions

Putting all this together we can expect a reduction of 24%*(1-(1-0.3)*9/8)=5.1% . In addition, the work week in the food sector would shrink to 40*(1-0.3)*9/8=31.5 hours.

Dietary changes

So far I have assumed that no changes are made in people's diets. But as most of us know, meat production, especially beef, accounts for a huge amount of emissions (1). Even greater GHG reductions can be had by regarding meat as a luxury rather than a staple as it is now. It is not unreasonable to enact a rationing system for meat, like many nations do in wartime. Unlike current attempts like cap and trade, where the burden of GHG reduction ultimately falls to the consumer (the working class), a rationing system would be fair, since it affects everyone equally (except of course vegetarians and vegans).

Stopping meat production entirely would work even better, but is likely politically impossible. In such a situation, hunting, which is necessary, would be the only source of meat.


9-36 million lives can be saved, 800 million do not need to go hungry, food's work week can be shortened to 32 hours, and a 5% reduction in GHG emissions is possible.