DAO, Technological Advancement and Contract Structure

Editor's Note: This article comes fromEthereum enthusiastEditor's Note: This article comes from

(ID: ethfans), author: A Jian, released with authorization.

I. Introduction

Editor's Note: On May 16, 2019, EthFans, Orange Book, and Spark Mining Pool hosted the fifth session of the "Insights" event in Beijing, with the theme of "Blockchain, Ledger, and Organization." Today, with new forms of DAO emerging in an endless stream, discussing DAO will seem a bit overwhelming anyway, and it is difficult to find a suitable balance between being too conservative and irrelevant. In this speech, A Jian used the concept of transaction costs to explain the contract structure in life, and also discussed how technology will change the contract. He is undoubtedly conservative, but being conservative also lights up another imaginary space.<1>I. Introduction

The concept of "smart contract" has a subversive meaning from the beginning. Nick Szabo mentioned in his 1997 article: The essence of smart contracts is to control all types of assets in a digital way

, in other words to use machines to enforce (at least some of) the rules of property rights. Letting our imaginations run wild a little, we immediately imagine the possibility of "reinventing corporations" by replacing contracts and the legal system with such machines.<2>However, the premise of running a "smart contract" is that its environment is safe enough, so safe that the power of a single group can hardly intervene and control, and then endow electronic contracts with strict machine attributes, so "smart contracts" have not found a suitable operating environment . Until 2009, the Bitcoin mainnet was launched, and even until 2015, the Ethereum mainnet was launched.

By 2014, the concept of Decentralized Autonomous Organizations had taken off

In the beginning, this concept actually had a strong utopian color, but later the idea of ​​applying it to coordinate human cooperation and create value emerged. The DAO is an obvious example, but it was actually set up for investment.

In 2016, The DAO was hacked and Ethereum was hard forked. As a result, the concept of DAO became cold. At the same time, there was another concept: Code is Law.

In 2017, when ERC-20 tokens became popular, many people began to talk about a pair of new terms: "Token Economy" and "self-organization". Although no one seems to say that Token Economy has anything to do with DAO, but I can vaguely see its reference to DAO, such as using tokens to specify governance permissions.

The above quotation can be regarded as a report, telling us how a group of concepts with very similar cores appeared in different appearances in different periods, and at the same time, how these different appearances changed people's understanding of it. angle. In my opinion, they are all answering the same question: Can a trusted execution environment of digital rules become a tool for coordinating human cooperation and creating new types of organizations?

2. Introduction to DAO, self-organization and token economy

Here I may spend time doing some boring things first: summarize our existing practices, point out where they are not ideal, and then lead to what I think is a more reasonable direction.

(1) Definition and practice of DAO

First define DAO:

1) DAO must have identifiable boundaries, we must be able to clearly see who is and who is not a member of this organization, otherwise DAO becomes a meaningless concept. For example, does the entire Ethereum community count as a DAO? I don't think it counts because there is no clear boundary.

2) The main rules governing the relationship and authority among DAO members must be expressed as a set of executable codes that run on a decentralized system. This is the so-called "decentralization", that is, the relationship between members is not mainly constrained by laws in reality.

3) The DAO must be able to generate common actions based on those enforceable rules, including changing (parts of) its own rules, while people outside the organization are excluded from the formal decision-making process.

The third point is actually controversial. Because the largest DAO in history, that is, The DAO, had a problem in the process of transforming rules into executable code, which gave hackers an opportunity, and finally everyone participated in The DAO’s joint action . Of course, it can also be said that this is not a problem, because The DAO at that time has indeed existed in name only. In other words, whether the third point should be included in the definition is actually related to your belief in "Code should be Law".

Another less important controversy is whether participants can withdraw from DAO with their own property at any time, whether it is the definition of DAO or not. The DAO has supported this from the beginning, arguing that if it is not guaranteed, people with more wealth can "legally" use other participants' money for purposes they do not support. However, if we relax the definition of DAO and think that the "Private DAO" proposed by Aragon can also be regarded as DAO, this kind of DAO implemented by multi-signature cannot be withdrawn at any time. Your withdrawal means the termination of the contract and the Re-establishment, which also requires a joint decision to implement.

Finally, let's look at a success story: MakerDAO.

MakerDAO uses MKR to mark the identities of its members and define their governance authority over DAO, allowing the voting results of MKR holders to determine Maker's joint actions. The main role of these joint actions is to determine the risk-taking of MKR. As we all know, Maker issued the famous decentralized stablecoin DAI. Although DAI has excess ETH as collateral, it is not impossible that ETH suddenly plummets and CDP is too late to liquidate. Who will bear this risk? It is the holder of MKR. In extreme cases, the contract may issue additional MKR to supplement the collateral of DAI, thereby maintaining its price. In MakerDAO’s governance choices, whether it is increasing the stable fee rate or increasing the type of collateral, it is actually to determine the degree of risk of MKR and require an appropriate price.

MakerDAO can indeed give you a lot of inspiration: first, they made a contract that can be automated (financial, homogeneous), second, they identified a risk factor and let a certain asset bear it, and finally the owner of this asset With governance authority, it opens up new possibilities for the entire system. This model is actually quite worth learning.

(2) Tokens and self-organization

Another new type of organizational time that we often hear is the so-called "self-organization". The core aspiration is that participants can be motivated by automatically issuing tokens, so that these organizational members can work hard spontaneously without explicit coercion Disciplinary execution, and this process will spontaneously produce certain results.

Although each has its own emphasis on ideas, there are several common points in practice.

1) Tokens have been issued, and the tokens will have certain functions in the future ecology of the project. The purpose is to align the interests of community members with the long-term interests of the project.

2) The work of the motivators is often not simple and easily verifiable work. Therefore, it is difficult to issue tokens completely with machines (contracts). For example, in translation, translation work should not only be based on quantity but not quality.<3>

At least as far as I have observed, so far, there has not been a particularly successful implementation of this model (that is, it has really achieved the "let a hundred flowers bloom" as the issuer wanted). So we can't help but ask, what went wrong? Is the incentive mechanism not working, or is it too easy to be plucked?

Ms. Jade from BESLab summed up the dilemma of this kind of token practice as a pair of paradoxes: if such behaviors can be incentivized, then token incentives are obviously not as good as BTC, ETH, RMB, so the effect is very weak; if such behaviors It cannot be motivated by currency at all, so how can tokens work?

I admire Jade very much. She is a top thinker in this industry, and she has the courage to expose the illusion and point out that the king actually has no clothes. But here I wish to express a small disapproval of her, in my opinion that the insufficient monetary nature of tokens does not fully explain the failure of such practices.

(3) Why did the token experiment fail?

(1) Currency and Token Incentives

Why do most of these experiments not last long? Why was everyone so excited at the beginning, but then slowly became less motivated?

The most obvious reason is that these tokens are currently useless. It constitutes an incentive purely because of everyone's expectations or trust in the project party. As this trust changes, the incentive effect will also change. In most cases, it is still a question of whether these tokens can capture value in the future ecology. After everyone realizes it, the incentive effect will gradually decay.

Second, even if they are useful, they are not suitable as money for paying wages. Think about it, will workers in household chemical companies accept washing powder as wages? They may be willing to take it as a benefit, but it is difficult to take it as a salary. The reason is that accepting an item, or an item with a relatively high quality judgment fee, as a means of payment will turn the contract between the two parties into a more primitive kind of barter exchange, and both parties must pay high fees to judge the quality of the counterparty. Commodity quality, this is destined to be a form of transaction with high friction.

This is Alchian's insight: why does money exist? In order to reduce the judgment fee in the transaction. Why haven't diamonds become currency? Before the advent of diamond technology, diamonds were also physically scarce. However, when diamonds are traded in the market, the four aspects of size, color, flaws, and cut must be judged and priced separately. If you are given a diamond, you will have to spend a lot of money to find out its true price. Willing to accept diamonds. This is why some people think that the less the use of currency, the better. In fact, it is not a matter of use, but the lower the quality judgment fee, the better. However, if there are more uses, the quality judgment fee will inevitably rise.

(As for the kind of tokens that are a bit like residual claims, they cannot be issued in this form of continuous issuance.)

(2) Entrepreneur

Another interesting question is that if the so-called ideal of self-organization is to design a contract that can be automatically executed on the chain, automatically evaluate the work of each resource contributor, and then give rewards, this process is impossible unless DAO is introduced. There is no entrepreneurial role, since all contracts are self-executing.

But this is very strange, can you imagine a business without entrepreneurs? Entrepreneurs don't just pay money, they judge market demand, determine the combination of factors, evaluate the quality of work, and finally pay remuneration and obtain benefits. One of the most critical aspects is to evaluate the quality of work, because different results will get different prices in the market. If wages are paid without properly evaluating the quality of work, it is tantamount to encouraging the production of low-quality goods, and the result is insufficient income and loss of capital.

Therefore, when the quality of the work of the participants varies greatly, and it is difficult to have a clear standard that can be executed by the machine to judge its quality, the result of using the contract to issue token incentives must be low-quality output, which is what we call wool party. This is the fate of most "self-organizing".

Combining the above two points, I will summarize the dilemma of this kind of token practice as follows: it cannot become a business, because there are no entrepreneurs, but it cannot become a two-sided market, because the market needs money, and the utility of the token economy Token does not meet this need, it only increases friction. Therefore, organization is possible, but "self-organization" is difficult.

Next we have to enter more complex issues, what is an organization, why do organizations have different forms, and how will technology change organizations?

3. Transaction fees and contract structure

Organizations, I think, are really groups of human beings linked by contracts. Although the contract is actually subject to multiple constraints of laws, traditional customs and morals, making the concept of the contract more or less like a black box, it is undeniable that the formation of an organization is more or less based on the voluntary conclusion of the participants. Then, the so-called organizational structure is the structure of the contract. For the study of organizations, if there is a theoretical basis, I think it is still necessary to look for it in contract economics or property rights economics. Of course, there are differences in the degree of explanation. The explanation of individual phenomena can only turn to other disciplines.

So the question is, why are there different types of contracts?

First of all, it can be determined that the contract structure is not determined by the economic input (quantity) of both parties. Different economic inputs only require different returns, and the returns can be reflected in the price. The amount of investment undoubtedly only determines the market they get. What is the price, but cannot decide what kind of contract they will use to coordinate each other's behavior.

Therefore, we want to turn to a very important phenomenon in the transaction, that is, in the transaction process, both parties to the transaction often have to pay a part of the fee, which has nothing to do with how much they are prepared to invest and how much resources they have invested, and it has not been determined by any party. get. For example, in order to make a transaction, we have to find a transaction partner, either on Taobao or walking to the supermarket. This part of the fee is not obtained by any party in the transaction, but it must be paid in order to complete the transaction. This part of the fee is the so-called "transaction fee".

This part of the transaction cost, however, is not captured by both parties to the transaction, so this is exactly the cost that both parties want to reduce. This is why there are different types of contracts - because different contracts pay different transaction fees. (Interested readers can refer to the second chapter of Professor Zhang Wuchang's "The Theory of Shareholders", in which the professor proves that if transaction costs are not considered, there is no difference in resource allocation between fixed rent and shared rent.)

(1) Classification of transaction costs

• Conceptually speaking, there are any number of classification methods for transaction costs, which is also a topic of debate in the academic circles. But in my opinion, the use of the concept of transaction costs is for the purpose of studying contracts. Therefore, logically, as long as there are no omissions, it should be as clear and direct as possible.

• So the classification method I am used to is this:

• The cost of finding a trading partner;

the cost of entering into the contract;

After the contract is signed, the cost of supervising the other party's performance as agreed.

This classification method is divided according to chronological order. There is another classification method that does not exactly correspond to the above three types of fees, but is conceptually clearer: information fees (negotiation fees), measurement fees, and judgment fees. Because in the transaction, we all inevitably have to go to the market to obtain information as evidence for bargaining. At the same time, we must determine the quantity of goods provided by the counterparty according to certain indicators, and the choice of indicators will also affect the form and quality of the transaction. Convenience, and finally, we have to pay a certain fee to determine the quality of the goods provided to us by the counterparty.

As mentioned above, the classification of transaction costs is nothing more than hoping that it will be clearer in concept and more convenient in analysis.

(2) Contract type

We will encounter many different phenomena and different contracts in life. To give a simple example, everyone knows that there is a difference between piece-rate wages and hourly wages in wages. So why are some jobs (such as factory work) allowed to pay by the piece, while others (office work) can only be paid by the hour?

Because the content of office work is often more diverse, the results are also very different. In this case, if piecework wages are adopted, although the cost of measurement is relatively low, the corresponding supervision cost will become higher (one job must be carefully looked at), or even unattainable; and these jobs are relatively trivial , It is difficult to set prices one by one, and the cost of negotiation will therefore become very high. These difficulties make enterprises prefer to adopt hourly wages and bear higher measurement costs, but in order to save bargaining costs. In the same way, in addition to hourly wages, we will also see various bonuses and welfare arrangements, such as full attendance awards, performance appraisal awards, etc. These incentives will undoubtedly increase the cost of measurement, but they can reduce supervision costs (no need for the boss to spend all day watching you work).

In the same way, when working in a factory, since the results of labor are standardized, or the quality is relatively easy to determine, there is a huge market outside the factory, and the prices of various standardized products are constantly flowing, so piecework wages are a bargaining fee. Measuring costs are low. When Hong Kong's ready-to-wear market was developed, the sleeves of shirts had clear prices.

In addition to the labor market, in terms of land and brand management, there are two common types of rent contracts. One is fixed rent, which requires a fixed price to be paid when resources are used, and the other is shared rent. A certain percentage of the output is charged as rent (in the tenant farming system, it is a share of the output of agricultural products). The chain service industry often uses sharing contracts. In fact, it treats the brand as a piece of land and lets the store operators operate and use it. In this case, the store operators directly face consumers and have more information than the headquarters. , in order to obtain this information as a reference for subsequent leases, and to restrain store operators (their behavior can directly affect the brand image), the headquarters often adopt sharing contracts.

There is a spectrum between full sharing and full fixed rent. There may be countless types of contracts on the spectrum. The specific contract will depend on the transaction costs.

(3) Changes in transaction costs and changes in contract structure

As mentioned above, in contract economics, we believe that the specific transaction costs determine what kind of contract we will adopt, and the marginal change in transaction costs also determines which direction the contract will move. As long as we can independently identify how transaction costs will change, subsequent economic inferences are easy.

For example, online shopping. Think about how the transaction costs of online shopping and offline shopping are distributed? With Taobao, thousands of products are displayed in front of you just by searching, so the information cost is very low; but the disadvantage is that you can’t see or touch it, and after you place an order, you may not be able to tell whether the seller will send you a second order. Goods, so the cost of supervision is relatively high. If the cost of such supervision is too high, it will be difficult for everyone to accept this kind of contract. In fact, this is the main problem faced by e-commerce in the early days. Later, Taobao invented a method to solve this problem, that is, after you place the order, you transfer the money to Alipay, and after confirming the receipt, let Alipay send the money to the seller, because the seller knows that you can reject it, and you are satisfied. He can really get the money, so he dare not send you defective products, which reduces the supervision cost and makes online shopping really popular.

The same reason can explain why books and digital products are the first to become popular in online shopping, because these products are standardized and the supervision cost is relatively small (especially for digital products, the information cost of offline shopping will be relatively high). In the same way, the later rise of Tmall and JD.com can also be partly attributed to this reason. The supervision cost of such self-operated products will be relatively low, because the cost of platform breach of contract is relatively high.

Armed with these concepts, we can easily understand the economic principles (money and entrepreneurs) mentioned in Section 2, and think tentatively about how technology changes contracts and changes organizations.

4. Miscellaneous talk about technology and organizational structure

(1) Company system and salary contract<4>Have you ever thought about a question, how long is the company system we see now?<5>According to Wikipedia, the earliest documentation of the limited company can be traced back to 1250 in France, even if we say later, it is also in 1600 in England, the East India Company

. The partnership system has a longer history, it is said that it can be traced back to ancient Rome, even if we say later, it was Italy in the tenth century

At that time, "commenda" was often used by people who went to sea to engage in business to form a partnership to pool money (in this kind of partnership, there must be at least one unlimited liability bearer (general partner), TA's commitment to the company's debts is not based on TA's contribution On the contrary, it is limited to the entire property of TA; while the limited partner only bears the debt of the enterprise only limited to the amount of capital contribution; the relationship between limited partners can be bought and sold).

Have you ever thought about how long is the salary contract we are familiar with now?

Why do you say that?

Going by what we've learned in contract economics, contracts for piece wages are likely no more than 500 years old, and hourly wages may have a larger history than that, but not by far.

Why do you say that?<6>Regarding piecework wages. First of all, before the emergence of productive machines, there was no good means to standardize the products of human labor. Since they cannot be standardized, there is naturally no way to talk about unified pricing. In other words, the use of piece-rate wage contracts will make the supervision costs of enterprises very high, and it will take a lot of effort to ensure that the quality of laborers' products is within a certain range. However, the machine can very efficiently ensure that the quality of the product will be stable within a range, thus greatly reducing this part of the supervision cost. In addition, the development of transportation technology is also a key factor. Because of the development of transportation technology, the industry has agglomerated geographically, and then there is the agglomeration of labor product price information, which reduces the bargaining cost of piece wages.

About hourly wages. The premise of adopting hourly wages is that time itself is easy to measure. If there are many obstacles even to timekeeping, what about hourly wages? In Europe, the escapement did not appear until the 14th century, and the watch industry did not begin to develop until the 15th century.

. Without these technologies, the cost of measuring workload would not have really come down, and piece wages would have been impossible.

The implications I hope to illustrate here are somewhat complex: (1) technology undoubtedly affects organizational form, as it affects transaction costs and further affects contract forms; The adoption of permanent machinery not only increased production capacity, allowing us to make factories bigger, but also led to the emergence of piece-rate wages, so that those who are willing to take risks and those who are not willing to take risks can get what they want; (2) Regardless of these technologies No matter how it progresses, it does not change the basic structure of "risk-taking", but only allows the resource investment of those who do not take risks to be measured more accurately. In other words, it makes the components of benefits clearer and allows The input of non-entrepreneurial talent is further stripped from the actual work of the entrepreneur.

1000 years ago, if an entrepreneur wanted to do something, he had to be a laborer, provide technical knowledge, and do legal affairs by himself. Which part of his income should belong to labor remuneration and which part should belong to knowledge remuneration , which part belongs to entrepreneurial ability, it is not clear. But with the development of technology, we can distinguish more and more clearly.

(2) Marketization of fair services

Finally, I offer a fantasy here.

Going back to the previous inference, judging the impact of technological progress on the contract structure mainly depends on the impact of technology on transaction costs. So what exactly has smart contracts changed?

It reduces the cost of contracting. You and I have never met before, but sending a transaction on the spot to make a 2/3 multi-signature contract is considered a contract, which is very simple. Whether there is money in the contract is easy to discuss, and we will pay each other after we have negotiated. But at the same time, the cost of supervision will increase, because it is a digital contract, it is difficult to form substantial constraints on the other party.

Now what we want to think about is, is there any way to reduce the supervision cost? Are there any activities that are sensitive to contracting costs?

The thing that comes to my mind is that this reduction in contracting costs will lead to the popularity of arbitration services and the "marketization of justice services". Because the contract is extremely simple, when we cannot trust each other, we can easily bring in a third-party arbitration institution, and resolve the dispute through the third-party institution when there is a disagreement (thus, these third parties are providing "fairness" service, as we expect from the judiciary). Some people may wonder, what if the other party colludes with a third-party arbitration institution? They can indeed collude, but this lack of access will lead to competition among arbitration service providers, and the market will punish those arbitrators who neglect their duties, forcing market participants to improve their service quality.

Not all trust must be established through technology, and many trusts are established through market competition. That's the most daring thing I can imagine.

Note 1: https://ethfans.org/posts/the-idea-of-smart-contract-nick-szaboNote 2: https://ethfans.org/posts/visions-of-ether-part-1

Note 3:

Token Economic Paradox — Incentive Mechanism, Social Production, Post-Capitalism

Note 4: https://zh.wikipedia.org/wiki/%E8%82%A1%E4%BB%BD%E6%9C%89%E9%99%90%E5%85%AC%E5%8F% B8