The cross validation with finance networks is a better idea and I think the next step is figuring out how to target non/low-unionized industries with strikes in unionized ones. As an aside, if you checked out my post on centrality, did I use the right I/O table from BEA? (https://apps.bea.gov/iTable/?reqid=150&step=2&isuri=1&categories=Io#eyJhcHBpZCI6MTUwLCJzdGVwcyI6WzEsMiwzXSwiZGF0YSI6W1siY2F0ZWdvcmllcyIsIkdkcHhJbmQiXSxbIlRhYmxlX0xpc3QiLCI2MDA0Il1dfQ==)

Tomas' point about strikes not creating value makes me think if theirs any objective function to judge a strike by, it's if the gains from it are sufficient to construct the commons (i.e can gains from strikes and corresponding boost in dues revenue suffice to construct co-operative groceries, housing etc. for the union members).]]>

I want to move towards the planning part in Adversarial planning, more akin to tomas' framework, https://www.xn--hrdin-gra.se/blog/2021/02/24/planning-complexity-for-model-economies/

But instead of computing optimal inputs, it'll compute an optimal supply shock.

Right now, my centrality analysis is a bit blind, sectors are not distinct, it's simply finding a sector that where the economy is most vulnerable to cascade failure regardless of which sectors are failing. However, we'd like to minimize the downstream effects on workers and maximize downstream effects on bosses.

I ended up finding Ian's work on super-integrated labor coefficients:https://www.youtube.com/watch?v=jROxFYv1bks

https://ianwrightsite.files.wordpress.com/2017/04/general-theory-labour-value2.pdf

I think this might be the special sauce; if a strike can be planned to induce a supply shock that maximizes the losses in the capitalist consumption sub-graph and minimize the losses the rest of the system we hurt the bosses with the least 'friendly fire'; "Better to sink a yacht than a cargo ship"

I wanted to use this thread to brainstorm.

There needs to be a notion of a 'supply shock vector', a vector that when subtracted from present input renders an output that is optimal in the manner described above, which is the tricky part, any ideas?

Ian's super integrated labor coefficients are awesome, a real breakthrough in my internal model of LTV when I read about them. I'm not sure if they have special advantage for this purpose over other models - I think all you need is some way to get better disaggregated input-output tables. One thing you might do is try to look at what industries are most fundamental in BEA tables, then zoom in on major companies in that industry, looking at what companies have correlated financial data. Use that to build an index, add research on production techniques as you zoom in on a target.]]>If we define some notion of damage then we can seek to maximize the ratio between the damage done to the bourgeoisie to the damage done to workers. The obvious measure of damage is to the capitalists' money purse. We also know that a long strike can be better weathered if food supplies and housing are guaranteed.

I've used the term "rate of cheating" for the amount that workers claw back in the post Shadow plans and ghost shifts. I think it may bear repeating that strikes have the problem that they produce no value. It would be even better to, rather than stopping work, to keep working, squirreling away goods produced for direct consumption by workers. This also takes away profit from the capitalists. That requires a system of planning to already be in place.

]]>But instead of computing optimal inputs, it'll compute an optimal supply shock.

Right now, my centrality analysis is a bit blind, sectors are not distinct, it's simply finding a sector that where the economy is most vulnerable to cascade failure regardless of which sectors are failing. However, we'd like to minimize the downstream effects on workers and maximize downstream effects on bosses.

I ended up finding Ian's work on super-integrated labor coefficients:https://www.youtube.com/watch?v=jROxFYv1bks

https://ianwrightsite.files.wordpress.com/2017/04/general-theory-labour-value2.pdf

I think this might be the special sauce; if a strike can be planned to induce a supply shock that maximizes the losses in the capitalist consumption sub-graph and minimize the losses the rest of the system we hurt the bosses with the least 'friendly fire'; "Better to sink a yacht than a cargo ship"

I wanted to use this thread to brainstorm.

There needs to be a notion of a 'supply shock vector', a vector that when subtracted from present input renders an output that is optimal in the manner described above, which is the tricky part, any ideas?