Tiberius Brastaviceanu on Why We Need a Contribution Accounting System
- 1 Source
- 2 Text
- 3 More Information
This article, now titled, Why do we need a contribution accounting system?, was first published on 3 January 2014 and last modified on 8 January 2018.
Version without graphs, links and illustrations reproduced below.
With the advent of the Internet and the development of new digital technologies, the economy is following a trend of decentralization. The most innovative environments are open source communities and peer production is on the rise. The crowd innovates and produces. But the crowd is organized in loose networks, it is geographically dispersed, and contributions to projects follow a long tail distribution. What are the possible reward mechanisms in this new economy?
Our thesis is that in order to reward all the participants in p2p economic activity, and thus to incentivise contributions and make participation sustainable for everyone, we need to do contribution accounting: record everyone's contribution, evaluate these contributions, and calculate every participant's fair share. This method for redistribution of benefits must be established at the beginning of the economic process, in a transparent way. It constitutes a contract among participants, and it allows them to estimate their rewards in relation with their efforts. We call this the contribution accounting system.
For the rest of this article we will try to explain why a contribution accounting system is needed in a more decentralized economy, and unavoidable in a p2p economy.
Contribution accounting and exchanges
First, we need to make a distinction between a contribution accounting system and an exchange system.
Suppose that we have 3 individuals picking cherries using one basket. The contribution accounting system keeps track of how many cherries everyone puts in the basket, so that when they sell the basket on the market they can decide to redistribute the revenue in proportion to everyone's contribution. It describes how contributions from multiple individuals amalgamate into a product, during a co-production processes.
Once a product is created, i.e. once the basket is full and ready to go to market, it can be exchanged using an exchange system: barter, currency, etc.
The contribution accounting system is not a currency, not a barter system. It doesn't refer to an exchange between our 3 individuals who are picking cherries, or between them and another entity like a company. They are not getting paid a salary in exchange of their work. They are collaborating, they all add cherries into the same basket, which is their product to be. The exchange might occur at a later point in time, once their basket is full and ready to go to the market. Meanwhile, they all share the risk of having their cherries being eaten by birds, or of not getting a good price for their basket.
A production process that requires more than one individual can be based on the following 3 arrangements, or on a combination of them:
- stigmergic coordination- Participants don't have aligned goals, don't maintain a relationship other than being contributors to the same process. Ex. this is how Wikipedia is built.
- cooperation - The goals of participants are not necessarily aligned. Ex. in a corporation employees and business owners usually have divergent interests and goals.
- collaboration - Requires a large degree of alignment in goals. Ex. a group of individuals climbing a mountain together.
The traditional capitalist economy is mostly about cooperation, which doesn't require an tight alignment of interests and goals. Production is sustained through an exchange process, where workers exchange the time they spend on different tasks against wages. The exchange process transfers risk from workers to the owners of capital, but at the same time, the workers are stripped of their rights to the output of their labor. Workers cooperate (despite some inconveniences and misalignment in interests and goals) with the owners of capital in production processes because there exists an economic dependency between the two groups. Workers need money, which are by far the predominant means to acquire basic necessities. On the other side, the owners of capital need labor to generate more wealth. This economic dependency is not symmetrical and makes the system prone to abuse, which explains the existence (and necessity) of unions to counterbalance the tendency for exploitation.
In peer production we have a blend of the 3 arrangements mentioned above, mostly coordination and some stigmergic collaboration. In general, no one works for anyone else. Everyone involved is a peer, an affiliate of a peer production network. The p2p culture prescribes that the output of a collaborative and participatory process should not be owned or controlled by anyone in particular, but shared among participants in a fair way. Immaterial artifacts that are produced in such way (such as software or hardware designs) are usually released as commons (they are openly shared). Material goods can be exchanged on the market, and the revenue generated is shared among all the participants. Service-based models also exist, where services are exchanged on the market against some form of payment, which is redistributed to everyone involved in the providing the service. A good example of service-bases p2p model is the Bitcoin network. If we focus only on the mining aspect, minors form a open network of peer participants, they collectively maintain the hardware infrastructure of the entire network. Minors are rewarded in proportion to the computing power that they provide to the network.
The normal and the long tail modes of production
normal mode of production
In the traditional capitalist economy wages should be regulated by the free labor market, if we make abstraction of all sorts of mechanisms through which this market can be biased (labor unions and governmental intervention included). The market is responsible for the difference in salary between an engineer and a clerk. The notion of job implies that a salary is determined and agreed upon before the employee starts working (with the possibility modify the salary based on performance). Since the amount of $ per hours of work is pre-established, the capital owner needs to make sure that the employee produces enough during the work hours. Therefore, a new role is needed within the organization to guarantee this, the beloved project manager. Traditional organizations spend a lot of energy doing time management, because usually the interest of the worker is not perfectly aligned with the interest of the capital owner (see cooperative production above). Classical organizations operate on the normal mode of production (from the normal curve or bell curve), where the number of workers is minimized, and the majority of employees in a category of roles produce almost the same amount. Very few workers produce less than the norm, because they are eliminated (i.e. fired). Very few produce more, because there are no incentives to do so, the association with the mission of the traditional enterprise is weak, the sense of belonging is usually low (usually fabricated by the HR department), the sense of ownership is almost absent, etc.
long tail mode of production
The situation is very different in a peer production environment, which is open to participation, is decentralized in terms of allocation of resources, and uses a horizontal governance system. In peer production, we see a log tail distribution of contributions, which means that a very large number of individuals are involved in production, only a very small percentage of those contribute a lot, the great majority of them contribute very little, and most of the production is done by those who make small contributions. A prearrangement on revenue is impossible in this context. First, because the production process is very dynamic and relations of production cannot be contract-based. Second, the process involves a great number of individuals that are distributed all over the planet, therefore it is impossible to do time management. Moreover, no one can force anyone else to work more. In this mode of production we need to evaluate rewards after the fact, based on deliverables or based on the type of activity and its potential to increase the probability of production of valuable products. A system is needed to account for everyone's contribution, to evaluate these contributions and turn them into rewards.We call this an access to benefits algorithm.
In some sense, the access to benefits algorithm is a distributed solution to time management, which can be applied to large scale and very dynamic peer production processes. It embodies positive and negative incentives, and can contain parameters to influence individual participation and quality of contributions, it can regulate behavior, it gamifies production. For example, a reputation system can be tied to the access to benefits algorithm: a higher reputation results in a higher reward, all other things being equal, and vice versa. Moreover, it can also contain parameters to incentivise periodic and frequent contributions, and to prioritize important processes.
Contribution accounting and network resource planning
The long tail mode of production needs a contribution accounting system in order to allow fair redistribution of rewards. It allows participants to record contributions of various types and it uses an access to benefits algorithm to turn them into benefits. But this is only the first part of the story.
In the OVN model, contributions are attributed to the creation of resources, which can be documents, designs, parts or full prototypes, etc. (some contributions go into infrastructure of community development and they lack clear resource or deliverables). From the resource level, contributions aggregate at the project level. A project is an open venture, or a business unit. It is the smallest unit within the OVN that can generate all sorts of benefits, including revenue.
The fact that contributions can be attributed directly to resources (not projects) is very important for commons-based peer production (CBPP), which builds on open source. On Github, pieces of open source software (OSS) can be picked up by someone and remixed into something else. Open source hardware (OSHW) development follows the same path, i.e. designs (mechanical, electronic, optical) are forked and remixed. This ability to fork and remix parts of more complex systems makes open source development a very efficient process. This explains why modularity and interoperability are very important properties of OSS and OSHW. If rewards are envisioned for the work done, CBPP needs to find a way to account for contributions at the resource level and to track the way resources are put together in different contexts (projects are considered contexts). If contributions are only recorded at the project level, projects become silos of economic activity with a reduced possibility of benefits flows between them.
Taking into consideration the structure of OSS development, the solution to the benefit/reward redistribution problem is to attach some information to individual resources created that allows their reevaluation later, when they get remixed and integrated into larger systems, in other contexts. The metrics of evaluation can vary depending on the context. This is the role of the network resource planning system NRP, which allows benefits/rewards to propagate upwards through value streams and the creation of a single resource can generate rewards from many different sources (many projects), depending on how many successful projects are using it.
This goes even further, because this same NRP also provides a growth mechanism for CBPP networks. To illustrate this, imagine that members of a CBPP community decide to attribute equity to resources that are created by other communities. (Example: SENSORICA decides to integrate a piece of open source hardware developed by another OSHW community). First, why would SENSORICA affiliates decide to diminish their revenue by giving equity to other groups when they can just copy the open source design? The economic rationale is to reduce efforts required to internalize new capacity (new knowledge and know how around that piece of open hardware) and to increase the speed of execution (a first to market advantage). CBPP networks grow by affiliation. By offering equity to other CBPP communities they are essentially building bridges to innovate faster and improve production processes. This is the higher-level structure of networks-of-networks (see the Open Alliance).
We believe that in order to sustain the CBPP we need to create infrastructure that allows attribution of value-related properties to individual resources, to allow reevaluation of these individual resources in context, and to facilitate the formation of networks-of-networks that preserve the individuality of every community part of it, but at the same time brings them together on the same economic platform.
Contribution accounting in transition models
As the economy transitions to a networked state, existing organizations are trying to adapt. We already see traditional corporations going from in-house R&D, to outsourcing R&D and more recently to crowdsourcing R&D. This movement is forced by the need to innovate fast, and by the fact that open source lowers the price to a point where traditional high-tech corporations can be put out of business. Crowdsourcing R&D means utilizing all sorts of schemes to attract the participation of the crowd into innovation processes that are sponsored by these corporations. In early crowdsourcing practices corporations tried to control the innovation by signing non-disclosure agreements with the participants. Crowdsourcing platforms were created to match corporate projects with skilled individuals. The practice was competitive, i.e. the company would chose a winner among different proposals, and usually the winner was rewarded with money. This practice gradually became more open, since the first iteration of crowdsourcing platforms were not very successful in attracting highly skilled individuals. In order to attract innovation, in order to grow open innovation communities around them, corporations need to think seriously about the reward mechanisms they put in place. It is not so difficult to understand why the early crowdfunding platforms were not very good attractors. I would not compete in a call by a company to design something for a few bucks, with a good probability of losing the race, knowing that the company will monopolize the work and probably make a lot of profits on it. The trend is to go from closed crowdsourcing to truly open source innovation, which must be accompanied by a broadening of the reward system. Since companies are going to deal with the crowd more and more, they need a contribution accounting system to account for contributions. See this presentation by SENSORICA making the distinction between competitive crowdsourcing and collaborative crowdsourcing.
In parallel to the adaptation of traditional companies we also see the creation of hybrid organizations and models. For example, in the realm or hardware, we have the emergence of ecosystems like Arduino and 3D Robotics/DIY Drones. They are composed of a traditional for-profit organization surrounded by an open source community. This post describes the situation. The difference here is that in most cases the open source community pre-existed the traditional for-profit, the later being created to manufacture and to distribute the products that are based on the innovation created by the open community. These hybrid models, the ones that are sustainable and successful, maintain an precarious equilibrium between the profit motive that can arise within the centralized traditional organization the open and sharing culture within the open innovation community. In some cases, this equilibrium is not maintained and the synergy between the two entities disappears, destroying the ecosystem. This was the case of Makerbot and the RepRap community, well captured in the Netflix documentary Print the Legend." (http://multitudeproject.blogspot.ca/2014/01/why-do-we-need-value-accounting-system.html)