Notes on Social Benefit and Cost

Hi all,

I’ve been doing some deep thinking about social cost and benefit in Parecon.

Prof. Hahnel and I have discussed some of this already. I’ve collected my thoughts so far in a PDF which you can access here:
Password: sbsc

The document is 14 pages and quite technical. However, I think it is useful and raises some important questions. It is focused, not a ramble. If anything is unclear, I’m willing to explain what I meant.

I think these issues are of fundamental importance to the system even though they can appear abstruse and finicky.

I’m open to any comments, including constructive criticism. In the forum, email, video call, whatever.

1 Like

Thanks for sharing, it’s very well put together! I’ve read this a few times, and I’m not sure I’m in the right headspace to respond. I struggle to understand formally written math, so it’s likely I’m not fully understanding. However, I’ll try to give a response.

If I’m understanding right, I think the core issue boils down to the how the income is being distributed to workplaces. Workplaces are given a lump sum of money that they will distribute among workers. If the workplaces are given an average based on worker statistics, it’s possible this sways the numbers in some weird ways when you look at this in combination with how the social benefit/cost ratios are calculated. Is that correct?

If so, I’m probably not the best person to respond. The way I’ve implemented this in my own project is on an individual basis instead of how you describe. Each worker receives income by multiplying the hours worked by the global hourly income for their category of labor (which is decided politically and is separate from the opportunity cost that’s calculated during the procedure). I don’t use averages or even have the workplace distribute the income. The hours are logged at the workplace and then income is sent directly to the worker. I personally think this is the best way to implement the concept, but as you know there’s a lot of discussion regarding how to implement this and even talk about it to the public.

In regards to the SB/SC ratio, I think the main purpose of this is to provide a quick metric for assessing the plans. I don’t think the goal is to maximize total benefit, but rather to have a way of assessing if the proposals are responsible in regards to the inputs they’re asking for.

Again, I may not be on the right page here, but hopefully this is contributing something to the discussion haha.

1 Like

Hi AfterTheOligarchy.

I am not sure I fully understand the issues and problems you are raising either but I just quickly want to note two points that I think are important to remember in the context:

  1. One main question is: what sum or ratio measures differences in members’ effort or sacrifice best given the annual planning procedure. I think one can argue that SB/SC is the best candidate. But one option is of course to make no adjustments for differences in SB/SC ratios, i.e. assume similar levels of effort in all WCs in the economy

  2. Adjustments to individuals’ compensation for work based on an SB/SC ratio in their WCs should not change the total (or average) compensation for work in the economy in a year in the aggregate. (At least not if you start from a calculated base compensation aiming to reflect a standard claim on the total production in the economy). All such adjustments will add up to zero in the aggregate. Adjustments should be a percentage reflecting how much a WC’s SB/SC ratio deviates from the average in the industry or the economy.

Thanks for this work. Here are my thoughts.

question 1. SB/SC ratio as a metric of socially responsibility.

“At present, in a worker council the average income per worker wcap (i.e. the avg. effort cap) equals the base WC income per worker wbase times the social ratio:”

As I understand it, Robin has suggested two ways for how a Participatory Economy could decide to set the total income cap of a workplace:

Method 1: The first is as you have it in your document that the total income each worker council gets is: economy base hourly income x total hours worked x SB/SC ratio. In this way, workers in workplaces that deliver a greater SB/SC would get a proportionally higher income than other workplaces.

Method 2: The second suggestion is not to correlate the income cap with workplace’s SB/SC ratio at all, i.e every workplace gets the same total pot: economy base hourly income x total hours worked. In this way, it doesn’t matter if the SB/SC ratio of your workplace is higher or lower than another workplace. You get the same (and any differences are only from internal redistribution).

It might be interesting to ask Robin more about this in your next interview if you get a chance.

I think it’s worth reminding ourselves that the goal in a PE is for any differences in income from work to be due to differences in effort or sacrifice, and not to other factors beyond one’s full control, including the value of the contribution of one’s output.

With this in mind, the advantage of the first method is that it also aims to take into account any differences in effort BETWEEN workplaces, not just within workplaces. It also provides more of an incentive to increase one’s workplace SB/SC ratio. However, it only fulfils this goal IF the differences in SB/SC between workplaces are due to differences in effort or sacrifice and not due to other factors. For example, workplace A has a better machine than workplace B which means it can produce more output. If the pricing of capital is accurate than workplace A will be charged more for their more advantageous machine and the advantage will level out, and if they produce more it will be down to greater intensity, not capital advantage. So it depends on whether pricing is accurate enough to make differences in SB/SC a level playing field.

In your document, if I understand it correctly, you raise that the SB/SC ratio may not always be an accurate indicator of greater levels of effort for other reasons, for example that this indicator is a relative ratio, and doesn’t consider the total absolute figure of SB-SC.

You give the example:

“So a couple of children selling muffins could have a profit margin m of 50% but only make a profit of €10 (silly example but the point is made). Whereas an enterprise producing medical imaging equipment might only make a profit margin of 5% but make a total profit of €10,000,000.

Firstly, I think you need to factor in time here and capture worker Income per hour in your equations. Total hours worked will vary from workplace to workplace. So the income the children received (assuming they are above the legal age to work) for selling muffins at a school fair for a couple of hours one weekend would be for the hours they worked, i.e if base hourly income is $40 per hour: $40 x 1.5 SB/SC x 4 hours worked = $240. That’s the comparison to make versus the workers receiving much higher income in total for working full time hours for the medical imaging equipment enterprise.

My next question would be: why would their SB/SC ratio be so high? They don’t set the price for muffins. This is socially set during the planning procedure. Like anyone else, they need to pay for the costs to make the muffins and the margin should be similar to other muffin makers, unless they try to increase their margins by lowering costs and making seriously small muffins, in which case who is going to buy them? I think this also goes back to the above point: If pricing is accurately set during the planning, then differences in SB/SC ratios should be due to differences in effort. The question is: can this be done so accurately during the planning procedure to achieve this situation? I think this can only be answered empirically in the future, but my guess would be that it cannot be done perfectly and why I would favour method two.

The advantage of method two, is that you don’t need to worry about the above and every workplace just gets the same pot for same hours worked, regardless of SB/SC, and as long as they are socially productive, i.e SB>SC. The downside is that it doesn’t attempt to factor in potential differences of effort between workplaces, but for me, I’m not so confident that this will ever be able to be done accurately enough.

There are advantages and disadvantage with either method and ultimately this is one of those questions for people in a future PE to decide which one to go with.

question 2. Separation of the opportunity cost of labour from income from work.

With your second question, I’m afraid I didn’t follow it entirely. You write:

“Because that means, by design, that within a given enterprise there will be workers receiving a wage that exceeds the opportunity cost of their labour (measured as a user fee). And surely, also, there will be enterprises where the total wage bill exceeds the total opportunity cost of labour (user fees).”

As I understand it, the opportunity cost of labour and income from work are independent. In this sense, a workplace doesn’t have a ‘wage bill’. The workplace has a total opportunity cost of labour bill, i.e the opportunity cost for each labour category multiplied by the hours. Similar to a profit and loss account in today’s economy, each workplace would have a SB/SC account. Labour costs would be an expense, like any other expense. However, the key difference is that workers are not paid from this account. The income from work comes from a completely separate accounting entity, which Anders in his book calls the Society Account. E.g. If Sarah works 30 hours a week, her workplace is charged 30 hours times times the opportunity cost of her labour. However, Sarah is credited (30 hours x base hourly income x WC adjustment) in her personal consumption account (that she uses to consume stuff) for the work she is done from the society account, not from the workplace account.

I don’t know if this is helpful or if you could clarify your question?

Dear All,

I’ve been out of touch at a conference at the Murphy Center at Tulane University in New Orleans for the past week. I’ve got to say: Interacting with philosophers on the subjects of economic justice and democracy is much more enjoyable than interacting with economists on those subjects! And I may have recruited a rising philosophy star to our camp! If you see a philosophy professor named Suzy Love visiting our website in the future, you will know where she came from. Anyway:

Ferdia and I have talked about this some. I’ll post some further thoughts when I have time to mull it over more. But for now:

  1. It IS technical.

  2. Ferdia raises some issues which are DIFFERENT from the issues we all have been aware of for some time about whether average effort ratings should be set equal to SB/SC ratios for WCs, or whether average effort ratings should simply all be equal to one another regardless of any differences in SB/SC ratios among WCs.

  3. I SUSPECT that the issue Ferdia is raising can be expressed in a far more simple way than he has in his posting… which I’m sure everyone will be happy to hear if I prove to be right in that regard.

  4. I believe Ferdia is concerned that when WCs maximize their utility from work (which is the same as minimizing their disutility from work) subject to the constraint that their SB/SC ratio must be greater than or equal to one, that WCs may not be using scarce productive resources efficiently. In other words, he’s concerned with efficiency not equity. In any case, that is what I am mulling over for now.

Hi everyone,

Thank you Michael, Anders, Jason, and Robin, for actually reading that document and for your thoughtful replies. I have read and reflected on all of the above and I’m continuing to digest it. I won’t respond to each of you point by point because that would be too verbose. But I will try to cover all the main issues raised.

Firstly, yes, my apologies for the very dense exposition. I was trying to gather my thoughts most precisely, which for me means putting it into a mathematical formalism. And the document should be split in half (Actually, originally I was going to send this thing just to Anders to ask an accounting question about the very last part, but I decided to post it here for general interest). I’m going to put the second question aside for the most part.

For now, on the second question, suffice it to say that (1) Anders and Robin are correct to say that even though an individual firm can pay less in opportunity cost of labour than its worker members claim in consumption (wages), in the aggregrate this should all cancel out, (2) that this ‘imbalance’ between labour’s recorded cost and claims on consumption is a design feature.

I think the more important question is the first one.


Yes this is a good way to summarise it concisely.

I suppose I’m thinking about the efficiency of the comprehensive plan. If we think about the activity within a parecon, do we think it is operating in the direction of something we consider optimal in the long run? For example, in central planning the optimisation of the plan is done explicitly. Mathematical techniques are applied to maximise some objective function, then that plan is implemented by different production units carrying out their part of it. In Parecon, the optimisation of the plan is done implicitly. Production units and consumption units make proposals, and these proposals are reconciled so that supply equals demand. But just because supply equals demand doesn’t mean that this plan is as efficient as it could be or as we want it to be. And by ‘efficient’ here I mean making the best use of our limited human and non-human resources.

I’m not necessarily saying that the plan in Parecon won’t be efficient. I’m raising a question about how we characterise an efficient plan in Parecon, and what basic economic rules can be instituted so that is realised at the macro level.

SB/SC Ratio

The essence of what I wrote in equations in the first section is basically this. In the Democratic Economic Planning version of Parecon, workers in an enterprise receive income which is some Base Wage X (SB/SC), where SB is basically recorded revenue and SC is recorded cost. To me, the basic idea makes sense: there is some monetary connection between the economic performance of the production unit and the private narrow economic interest of those who work there.

The question I’m raising is this: is the SB/SC ratio the best metric for that?

What I showed in the document above is that the SB/SC ratio is basically a proxy for the ‘profit margin’ of the enterprise, i.e. (revenue - cost) / (revenue). The SB/SC ratio is just a function of that. So really, wages are basically Base Wage X Profit Margin.

Profit margin is of course not the same as profit. An enterprise could have profit of €1mil and profit margin of 10%, or profit of €10k and profit margin of 10%. Profit margin is not necessarily a bad metric to use, I’m just pointing out that it is distinct and has distinct attributes.

Net Social Benefit: SB-SC

My essential point was that if we can truly consider enterprise revenue as ‘social benefit’ and enterprise cost as ‘social cost’, then surely we want to maximise net social benefit in the plan (within constraints, such as environmental).

In that case, to begin with you want to institutionally encourage enterprises to maximise their SB - SC, or their ‘profit’. This is related to, but not the same as, the profit margin. To illustrate the practical difference simply, you could imagine two enterprises: both have the same profit margin, but one has a way higher quantity output than the other. So in relative terms they are equally efficient (same SB/SC), but in absolute terms the second enterprise is producing a lot more ‘net benefit’ (SB - SC).

The idea would basically be that if prices are functioning properly (in particular, not excluding real costs from monetary costs, and not including real costs in monetary revenues), that ‘profit’ would actually represent a net social contribution of the enterprise. (And of course, it’s not really a ‘profit’, because it can’t be accumulated and used to buy capital).

So that over time, the economy would move in a direction where enterprises are producing more of what people want and need, while using less productive resources to accomplish it.

Could that be achieved using the SB/SC ratio instead of SB - SC? Maybe. I’m not sure. It’s a ‘work in progress’.

There’s more I would like to write here but I’m going to leave it there because it’s very long. That is a overview of what I’m trying to analyse though.

1 Like

Let’s imagine a scenario where we have two bakeries that make bread. For simplicity, let’s say they only need to use wheat. Let’s use the following statistics in our scenario:

Inputs Required for 1 Unit of Bread
Bakery One: 10 units of wheat
Bakery Two: 1 unit of wheat

Bread: 80
Wheat: 5

Benefit/Cost for 1 Unit of Bread
Bakery One: 1.6 (80 / 5 * 10)
Bakery Two: 16 (80 / 5 * 1)

As you can see, bakery two produces bread more efficiently and has a higher benefit/cost than bakery one. Now, let’s imagine a scenario where the inefficient bakery is the main one supplying the bread.

Scenario One
Society Demand: 200 units of bread
Bakery One Supply: 150 units of bread
Bakery Two Supply: 50 units of bread

If I’m understanding correctly, this is the kind of scenario you’re concerned about, right? The inefficient bakery is supplying most of the bread, but they’re using a lot more wheat to get the job done. Both bakeries have a beneifit/cost greater than 1, so from that perspective they’re both seen as responsible.

I don’t see this scenario as a bad thing. If we’re not able to get enough supply out of bakery two, I’d say it’s a good thing that bakery one is picking up the slack. Ideally, it would be nice if these numbers were flipped, but there could be all kinds of reasons for why this isn’t able to happen.

Now let’s imagine a new scenario where both bakeries are proposing to meet society’s demand.

Scenario Two
Society Demand: 200 units of bread
Bakery One Supply: 200 units of bread
Bakery Two Supply: 200 units of bread

As you can see, we have a little competition going on between the two bakeries! As a whole, society is proposing to produce too much bread, so the price of bread is going to decrease. For simplicity, let’s say the price of bread is cut in half since there’s quite a bit of excess supply.

New Prices
Bread: 40
Wheat: 5

New Benefit/Cost for 1 Unit of Bread
Bakery One: .8 (40 / 5 * 10)
Bakery Two: 8 (40 / 5 * 1)

If we’re automatically rejecting plans that have a benefit/cost less than 1, bakery one has lost the competition and the more efficient bakery will be responsible for meeting society’s demand for bread.

If this isn’t covering the points you’re making, it would help me if you could frame it in a similar way. From there, I’ll do my best to respond!

Hi Anders,
I’m currently reading your book ‘Anarchist Accounting’ and really struggle to understand some technical aspects of capping a workplaces average-effort-rating via the SB/SC-Ratio (Rule 1). I already had the same comprehension-problems in Robin Hahnels ‘Democratic Planning’ (p. 117). Here a longer extract form your book (I hope it is ok to post it here). The part I have the most problems with is highlighted:

"While an individual workplace is free to design its own internal procedures for rating its members’ relative efforts and, thus, allocate compensation between its members, its total average compensation must, one way or another, be related to other workplaces’ total average compensation in a way that takes into account differences in effort and sacrifice between workplaces. The relation between different worker councils’ average compensation can be decided in two different ways, each with their own advantages and disadvantages.

Rule 1: based on social benefit/social cost ratios

The first possibility is to base the interrelation between workplace average compensations on their social benefit/social cost ratio (SB/SC). For example, if a workplace’s SB/SC ratio is 105% while the average for the industry (or some other possible reference group) is 100%, this means that the workplace average compensation should be 105% (calculated as 105/100) of the industry average. This rule is based on the assumption that a higher SB/SC ratio actually indicates a larger effort by the worker council’s members compared to the other workplaces, and that a better outcome cannot be attributed to other factors, such as more efficient machines. The annual planning procedure is designed to achieve this by pricing access to more efficient equipment and resources higher than access to less efficient equipment and resources. Thus, the denominator of the SB/ SC ratio will be higher if more efficient resources are used, leading to a lower SB/ SC ratio at unchanged production quantity. If the annual planning procedure, to a sufficient extent, succeeds in this regard, the capping of the average compensation based on worker councils’ SB/SC ratio will be fair. If there is a risk that the pricing of user rights fees for capital, resources, and labour in the participatory planning procedure not quickly or accurately enough reflects differences in efficiency and quality, it can be argued that relating worker councils’ total average compensations based on SB/SC ratios is unfair. Worker councils using relatively inefficient machines and resources would then be unfairly disadvantaged.
In this context, SB/SC ratios are always calculated based on actual outcome, and not on the production proposals that are prepared during the annual planning for the year to come. When a year is over and the final annual results are known, members’ compensations (before adjustments for desirability), for the completed year, may need to be adjusted retroactively so that the average compensation for every workplace corresponds exactly to its SB/SC ratio relative to its reference group. A reference group consists of all the worker councils whose average compensations are determined based on their relative SB/SC ratios and may consist of all worker councils in one or more industries or a specified part of an industry, for example, based on geographical divisions if there are regional differences that affect the circumstances and costs of the worker councils. The reference groups are defined by the industry federations.
Conditions may vary across industries in ways that hamper or impede the possibility for a fair comparison of SB/SC ratios between workplaces in different industries, even though comparisons between workplaces within the same industry are fair. The industry federations can, therefore, adjust the rules in different ways to deal with the fact that the SB/SC ratios may not always exactly reflect differences in effort but still work satisfactorily as an indication.

Rule 2: all workplaces are equal

A second possible rule for relating total average compensations in different workplaces is that all workplaces are considered to be equal in terms of average effort and no workplace can differ from any other. No retroactive adjustments of the workplaces’ average compensations are then necessary. Using this alternative, one instead risks disadvantaging workplaces whose members are sincerely committed to more than average efforts. This rule could be combined with a requirement for worker councils to demonstrate a certain minimum level of SB/SC ratio in order to be assigned the economy’s average compensation. If they fail to reach the required minimum SB/SC ratio, their average compensation could retroactively be adjusted downwards by a certain percentage." (p. 37-38)

  1. The first question I have is: Does the highlighted part above imply, – given the assumption that the only difference lies in efficiency-differences of the inputs – that the SB/SC-Ratio of a council, which uses more efficient means of production (machines/technology/resources), should equal the SB/SC rating of another council in the same sector (and with the same abilities/effort), which uses less efficient means of production, after the planning procedure is accomplished?

  2. Thereupon the second question: Assuming this is correct, how is it exactly achieved, that in the planning process, the SB/SC rating is equalized to the degree that every later difference in SB/SC reduces to the level of effort and sacrifice? Is this because:
    In the first planning-step every council will initially apply for the use of the most efficient means of production to achieve the highest possible SB/SC-Ratio. If these means of production are scarce, in the next round the iterative price will rise (due to excess demand) and the price of less efficient production-factors will correspondingly fall (due to excess supply). Thence the SB-SC of both councils will equalize, although they are using technology/resources which differ concerning their efficiency (assumed the process works accurately enough). Is this correct?

  3. Assumed all this is true: Isn’t there a problematic incentive concerning technological innovation/advance and the use of (i.e. demand for) efficient resources? Where is the point of applying for the best machines and the most effective resources (and therefore the most expensive ones), if in the end of the planning process, every council (given all the other capabilities/factors are the same) will have an equal SB-SC-Ratio, no matter what machines/resources one uses?
    In ‘Democratic Planning’ Hahnel writes – addressing ‘allocative efficiency’ –, that the councils want to increase their SB/SC-Ratio during the planning process to be considered in the distribution of the productive resources in the planning process and therefore have the natural incentive for applying for the best productive materials/resources/machines so that they can increase their SB/SC-Ratio (p.119). How does this fit together if in the end it really doesn’t matter if you use the most efficient productive resources, when less effective ones will eventually even raise your SB/SC-Ratio?

  4. Concerning the 2nd Rule, that the average effort-cap is equal to all councils: Doesn’t this reduce the incentive to adjust the production according to the (changing) demand: I think Schweickart claimed, that the councils would have no interest in finding out, what the consumers exactly want (i.e. what their preferences are). Hahnel replied in his discussion with Wright (‘Alternatives to Capitalism’) that they indeed have an interest to identify and respond to the wishes of the consumers, because the effort rating for the next year is bound to the SB/SC-Ratio. Sure there is still an incentive, since you suggested a certain minimum level, but the incentive would be reduced if all would have the same average, right?

I am sure, I missed something (or a lot of things) and hope you could help me with this jumble of thoughts.

Greeting and thanks in advance,

1 Like

Hi Askush.
Glad o hear that you are reading my book.
I will answer briefly and hope that AfterTheOligarchy doesn’t feel that we are hijacking his thread here:

Re q1: Yes, the idea, and aim, is that any differences in SB/SC ratios between workplaces should reflect only differences in effort. If the same level of effort, the SB/SC should be the same regardless of resources.

Re q2: Yes, I think that is correct. But I would add that some WCs, when applying for “most efficient means of production”, may also be motivated by wanting to achieve a positive SB/SC ratio, which is needed for approval, with a minimum of effort.

Re q3: This is a fair and common question. I tried to answer this exact question in a separate article » Innovation.

Re q4: WCs still need a positive SB/SC ratio and they presumably want to achieve this with a minimum of effort. So they have no incentive to waste resources.


1 Like


I didn’t forget this I’ve been thinking about it, because it’s a huge topic and has many facets.

The general question is: ‘what is a good plan for the economy?’ and more importantly ‘what is a good situation in actual production, distribution, and consumption, for the economy?’.

I’m going to write a shorter reply for now, and maybe write a blog post where I can properly organise my thoughts.

Using your examples of bakeries. Basically there are two questions:

  1. Do we want bakeries to produce more bread for the same cost?
  2. Do we want bakeries to produce the same quantity bread for lower cost? These are two different ways of improving ‘efficiency’ (more for same, or same for less).

In a snapshot of time, (1) leads to higher Revenue for the same Cost, (2) leads to same Revenue for lower Cost. Both result in a higher difference between Revenue and Cost.

So the question is: is a given bakery incentivised to do this? If the only constraint is Revenue = Cost, how will bakeries produce? Will they produce as ‘efficiently’ as we might like?

There is another part of it. In the Democratic Economic Planning version of Parecon, the Sacrifice Rating Cap (or wages cap) for an enterprise is proportional to some metric of enterprise perfomance, given as Revenue/Cost (the “SB/SC” ratio). In that document I show that the Revenue/Cost ratio is basically a function of the Profit Margin (profit margin is: [Revenue - Cost]/Revenue). So if overall wages for an enterprise are proportional to the Profit Margin (loosely speaking), the enterprise is incentivised to maximise Profit Margin. This is different to being incentivised to maximise Profit (Revenue - Cost). Both have advantages and disadvantages.

For example:

Bakery X
Revenue = 100
Cost = 90
Profit = 10
Profit Margin = 10%.

Bakery Y
Revenue = 100,000
Cost = 90,000
Profit = 10,000
Profit Margin = 10%.

As in, both bakeries record the same Profit Margin and therefore get the same wages. But Bakery Y has a Profit of 10,000, which is much larger than Bakery Y’s Profit of 10. In Parecon, this ‘Profit’ (Revenue - Cost) is supposed to be the measure of net Social Benefit. So according to that argument enterprises should be maximising Profit (not Margin). It’s mostly an issue of scale of production.

1 Like

Yeah, I think there’s different ways to go about this. I’m probably a little more “anarchist leaning” than most, so I feel like the aim should be to let workplaces do whatever they want as long as the inputs are seen as responsible / efficient in the grand scheme of things. We don’t want to waste natural resources or labor, and I’m convinced the current SB/SC method will prevent this from happening.

When it comes to the income part of this, I don’t really have any comments on that. If someone asked me to describe the ideal economy, I would say one of the main goals should be to stop using money as an incentive to coerce people. :stuck_out_tongue: There are all kinds of “natural incentives” that I think will flourish in a system like this, but I understand from an economist’s perspective you want to assume everyone is lazy and greedy to prove the theories and so on. Regardless, I think the worker’s income cap / distribution side of this is a technical issue that can easily be changed once an economy like this is implemented. If we had to decide this today I would advocate for just paying the workers directly like in the way I described above.

Also, to everybody:

After much reflection and also discussion (with Prof. Hahnel, and particularly with Anders Sandström), I would recommend ignoring the second part of the PDF I shared in the original post. As in, ignore everything after page 8. It’s not necessarily ‘wrong’ but I think it is mostly confusing and unhelpful. Good news is the PDF’s effective length is now half.

The first part is good though.

1 Like