Editing Executive governance (section)

== Policing motivation ==
The most basic and strongest direct incentive a DAO member has for participating in the policing process is that they are rewarded with new REP tokens when they participate in the Validation Pool. Based on the [[Validation Pool#Vote|policing parameter <math>c_3</math>]] (which has default value <math>c_3=1/2</math>) and the rate of total fees [[Validation Pool#Vote|<math>f'</math>]] the DAO earns, we can determine how valuable it is to participate in executive governance. 

An indirect motivation for an individual to police is that it keeps the DAO, and therefore the individual's REP tokens, secure and valuable. A higher policing parameter motivates more participation, but it also makes it more difficult and expensive to achieve a 51% attack arbitrage.

However, a higher policing parameter makes finite lifetime REP tokens last longer, as new REP is automatically minted with fresh tenure. This leads to 

Such motivations need to be weighed carefully when a DAO decides how to set policing parameter. The decision is clarified with a rigorous valuation of policing tokens.

It is assumed most members will not leave their reputation tokens unused, merely passively earning the reputation salary. At a minimum, most members can be expected to also use their REP for executive governance, that is, to automatedly police all actions in the DAO by participating in each Validation Pool. This will be typical when the reward for policing is sufficient to motivate members to participate in policing. This means the reward is more than the price to run the machines needed to participate.

'''Proper Policing''' means continually using the most recent version of the canonical front end, which runs the automated algorithms that vote in accordance with governmentally established soft protocols. We distinguish Proper Policing from general policing, because in a decentralized system Byzantine behaviors are possible. Anyone can choose to directly vote in any way, e.g., against the accepted standards of the DAO.

An example of policing is given in the first Bitcoin node software. Running a full node means your computer will store the entire information of every transaction on the blockchain and run the full software package for participating in blockchain consensus. This software package runs automatically and deploys the following protocol: Listen for new users’ proposed transactions ('''TXs'''), send those TXs to the network (this is the '''gossip network'''), collect all known TXs for the last 10 minutes into a '''block''', calculate hashes to try to win the hash lottery for PoW ('''hash mining'''), submit your block for approval if you win the lottery, or check the other submitted block(s) are made according to protocol if you lose. The vast majority of participants in the Bitcoin blockchain skip most of these steps. They instead choose to run an edited version of a full node which limits their computing resources to the task of has mining so they can maximize their chances of winning the bitcoin reward. Policing, in this context, means the other steps in the process, which secure the network—participating in the gossip network and especially checking that the proposed blocks are formed correctly according to protocol.

How much is policing rewarded? In the Bitcoin network, there is no immediate incentive for policing. In more sophisticated PoS (proof of stake) consensus algorithms, such as Ethereum’s LMD-GHOST, policing rewards are built into the protocol. When a DAO implements a reputation mechanism, we need to account for policing rewards explicitly, so we can optimize the incentive. The formulas given previously show how much a token is worth for passively collecting the reputation-weighted salary. But in this section, we need to determine how much more a token gains by policing. Specifically, how much extra REP is gained automatically by participating in the non-contentious, binding validation pools that should result in unanimous decisions? If this value is not more than the price of running the machines necessary to execute the full version, then most users will eventually opt to execute simplified versions, which is a security threat.