Shadow plans and ghost shifts

Lately I've been thinking of ways to perform "dress rehearsals" for a future democratic planned economy. As I have said on here before, it is important that planning penetrates down to the shop floor and into all communities, that all workers are able to directly engage in formulating the one continuously changing plan.

I am convinced that it is important to build institutional knowledge/know-how around planning even before any kind of revolutionary scenario takes place. To this end I propose two mechanisms: shadow plans and ghost shifts.

Ghost shifts

Ghost shifts or rogue operations are one way to seize partial control over the means of production without employers noticing. A strike produces no value, whereas engaging in rogue production for the commune produces value/use-values "in the opposite direction". From the workers' point of view we therefore have three scenarios:

  • Business as usual. Workers are exploited. I call this state of affairs minus (-)
  • Striking. No value is produced, but exploitation may reduce slightly afterwards as a result. Because such gains are eventually undone via mechanisms like inflation, I call this the neutral or zero result (0)
  • Covertly working for the benefit of the commune. Here value is withheld from the employers and funneled to the commune instead. Therefore I call this a plus (+)

Rate of communalization and rate of cheating

If we take an average 40 hour work week and a 100% rate of exploitation[1] then 20 hours per week are spent generating one's wage (v, four hours per day) and the remaining 20 hours generating profit (s, also four hours), like so:

[vvvvssss vvvvssss vvvvssss vvvvssss vvvvssss]

Each letter represents one hour. I use the same symbols as in the English translation of Capital, so v stands for variable capital and s stands for surplus value. The surplus value is the amount of time spent per week performing unpaid labour for your employer. Or put another way, how much you pay your employer for the privilege of working for them!

Using the symbol κ for the time spent working for the benefit of the commune, I extend the definition of the rate of exploitation to s=s/(v+κ) . If just one hour per week can be shifted over to communal production then the rate of exploitation falls to s=19/2190% :

[vvvvssss vvvvssss vvvvκsss vvvvssss vvvvssss]

Notice how a mere 2.5% shift in the work week results in a 10% reduction in the rate of exploitation.

I see two ways of looking closer at this value transference. The first I will call the rate of communalization (kommunaliseringskvot) denoted κ , which is the fraction of the working week spent working for the benefit of the commune:


where κ ranges from 0% to 100%. In the example above κ=1/40=2.5% .

The second way I will call the rate of cheating (avlurningskvot) denoted χ , which is the fraction of the working week spent "cheating" the employers out of their profits:


where χ ranges from zero to infinity. χκ . In the example above χ=1/392.6% .

If it's one hour every workday then the numbers become s=15/25=60% , κ=5/40=12.5% and χ=5/3514.3% :

[vvvvκsss vvvvκsss vvvvκsss vvvvκsss vvvvκsss]

Notice that the amount of value taken home by each worker has now risen from 20 hours to 25 hours per week. This means we could in the best of worlds shorten each workday by one hour:

[vvvκsss vvvκsss vvvκsss vvvκsss vvvκsss]

In the above example I presume the politics play out such that the workers accept a lower wage, knowing that they can still withdraw an amount of goods equivalent to what we had in the very first example. Due to the pay cut the rate of exploitation goes up: s=15/20=75% . But the rates of communalization and cheating also go up: κ=5/3514.3% and χ=5/3016.7% .

Notice that in all three cases the rate of exploitation is lower than what we started with, and the wage share[2] is higher, which is why I call it a plus. A strike by comparison looks like this:


This assumes the strikers don't go and do other productive labour. I also assume that the situation isn't a general strike but a limited strike within a trade union, which is the typical situation here in Sweden.

It might not be immediately obvious why this is beneficial to workers, but the benefit is directly tied to the rate of exploitation. By working for their employer the worker takes home only ½ hour's worth of value per hour worked, before tax, whereas when working for the commune they receive one hour's worth, minus deductions for the common fund (roughly tax, around 33%, see the section "On notation" below)[3]. The workers therefore get better "leverage" on their time by working for the commune. This means a radically shorter work week in the long run, a situation like this:

[κκκκ κκκκ κκκκ κκκκ κκκκ]

A 20 hour work week with a rate of communalization of 100%.

Ghost shifts are nothing new, but the application of them towards building socialism might be. At present ghost shifts are done to sell more of some commodity under the table. It is still commodity production, and it only benefits the workers at one particular workplace. It cannot benefit from economics of scale, nor benefit from co-operation with other workplaces, nor does it benefit society as a whole.

Shadow plans

Much like how opposition parties in bourgeois parliaments may formulate "shadow budgets", the commune could formulate "shadow plans". By this I mean a kind of "eyeing the prey" of useful means of production that the commune does not yet control. Such tentative plans could be computed and compared to figure out where best to direct amalgamation efforts. Workplaces that are easy to incorporate, have dissatisfied workers and that produce the most useful goods and services are prime targets.

Concretely this means attempting to reverse engineer how different workplaces operate, to glean technical coefficients from people outside and inside these workplaces. The thread Towards Adversarial Planning for Industrial Action on CASPER Forum goes into some ideas similar to this. I have drawn the figure below to hopefully help illustrate the concept:

Figure demonstrating the concept. A smaller polygon labeled 'currently feasible plans' sits inside a larger polygon labeled 'potentially feasible shadow plans'

Adding more productive methods means the polytope that describes the set of feasible plans grows larger. I have represented this in the figure by the gray polygon being larger and differently shaped than the inner green one. The actual polytope is not 2-dimensional but more like 1,000,000,000-dimensional so some imagination is necessary regarding its actual shape.

This is a concept I may develop further in the future. For now even in this early form it felt appropriate for the spooky theme of this post 👻

On notation

I have used a more "Marxian" κ and χ in this text, but we could use the same strategy as in the post I wrote on notation. We have the following instantaneous rates of communalization and cheating:


Obviously the more communalization and the less exploitation the better. The corresponding regulating depreciations look like this:


We can interpret these two as a high rate of depreciation of wage labour and exploitation being beneficial to the commune. Depreciation in the amount of communalized production on the other hand is detrimental.

Note that κ breaks down into its own constant (fixed + circulating) and variable parts. We can call these κc=κf+κ and κv respectively. Unlike capital the commune does not benefit from depreciating constant "capital" (means of production), what we might call κ(c,-) . When c- increases p¯- increases. Capital prefers to get rid of "dead weight", and does not grow weaker from factories and people being bombed, but stronger. In contrast an increase in κ(c,-) leads to a decrease in χ¯- . The commune does not want its stuff destroyed. The same goes for depreciation in the labour force itself, or κ(v,-) . The commune does not want its people dead.

For completeness the full expressions for the rates of cheating when κ is broken down this way looks like this:


I am not sure whether it makes sense to speak of surplus value production here ( κs ), or whether the value demanded to maintain the non-working population (the young, sick and pensioners) should be accounted as a tax on κv . Going by the OECD average of 65% employment-to-population ratio[5] and the aforementioned 100% rate of exploitation, we have the following result: In capitalism each worker supports two non-workers: ½ actual dependent and 1½ dependents' worth of capitalist exploitation. Or in terms of time: 13⅓ hours for the worker, 6⅔ hours for dependents and 20 hours exploitation per 40 hour week. In socialism each worker only needs to support the ½ dependent, a number that is hard to reduce without cheaper healthcare, later retirement or ättestupa.


[1] Rate of exploitation being defined as the ratio between surplus labour performed for your employer divided by the time spent generating your wage, or s=s/v . It currently sits around 100% on average worldwide[4].

[2] The wage share is another way to look at the rate of exploitation. It is defined as the ratio between the time spent generating your wage over the length of the working week, or w=v/(s+v)=1/(1+s) . For the 100% figure in [1] we get w=50% .

[3] K. Marx, Critique of the Gotha Programme, 1875, (accessed 2022-11-21).

[4] E. D. Farjoun, M. Machover and D. Zachariah, How Labor Powers the Global Economy: A Labor Theory of Capitalism, Springer Cham, 2022, doi:10.1007/978-3-030-93321-0.

[5] (accessed 2022-11-21).