So I had this idea recently about how to deal with the problem of trade between different socialist enterprises within capitalism without money, which would also apply to trade between different socialist countries. Basically the idea is this: an enterprise creates a budget of final goods to export and a budget of manpower/other resources to export access to, they also create a set of contract solicitations available for other social enterprises to bid on. These solicitations would be for goods that they either need in the production process or have final demand for if they handle distribution.
An enterprise can put up a bunch of solicitations and multiple other enterprises can bid for a cost specified in their contract proposal. These costs can be expressed as a token which gives them access to a catalogue of the other solicitor's goods, or as a specific quantities of goods decided ahead of time. Naturally, specifying the goods ahead of time would carry less risk considering availability in a catalogue like that can change. Contracting officers of course can choose to consider things besides costs, just as they do in real life, such as credibility, quality, environmental impacts, ect.
Such a system uses comparative advantage to persist in some way, ensuring that there aren't big sinks on productivity because of big treaties between entities. And it gives two ways for a given enterprise to access another enterprise's resources - as payment on a contract and as fulfilment of a contract.
Here's a diagram I drew of it, sorry if it's rough.
>trade between different socialist enterprises
Why? Just use planning.
>different socialist countries
There cannot be different socialist "countries", only different nations within the one socializing economy. You can't have competing bases, that leads to conflict.
@thardin From a practical perspective, conflict is inevitable, I think it's very utopian to assume that political differences won't lead to heterogeneous systems. How do you deal with this when it happens is a very important question. You shouldn't just assume from the beginning that everything can be incorporated into a homogenous planning system, and it's an important issue today when talking about enterprises, communes and coops of all types which are not centrally integrated but might have aspirations of moving beyond capitalism.
There will of course be political strife. But the superstructure plays second fiddle to the base. If you have more than one base then you have laid the groundwork for there to be conflict between said bases. See for example the Sino-Soviet split. Unless you want to recreate such splits then you should work towards there being only the one base. Having two or more systems of planning means chaotic production. Unless you are very careful (and no theorists I've seen so far have been sufficiently, mathematically, careful), it means recreating the anarchy of the market. You will have something like the business cycle. You will have war.
It is true that we have heterogenous systems in place at the moment. These must be sewn into a single system, into a single base. Else you will have anarchic production and war.
I am not denying that one can have loose coupling in a system of calculation in kind. But if you want to do so then you must also provide the spectral analysis that shows that system is free of modes that can be excited.