This seems like the most reasonable take. Kinda sad the current consultation doesn’t really allow for this option 
Some people will interpret it as: I’m not voting for XRD tied up in defi, because it should be added later and is too difficult. And others will as: I eventually want XRD tied up in defi to also count, so I will vote for it.
Letting customers (defi users) into the vote is like letting the fatties that buy 20 Big Macs a day have a vote on how McDonalds should run things.
Those fatties will vote for BigMacs with 8 patties of beef each (vs. the current 2 patties) and price them at a $0.05 per BigMac.
Find me a real-world example of where customers = owners and those customers have no skin in the game and I’m willing to change my mind.
I have checked a couple of details and I now convinced of the following: Validator Staking LSUs can’t be the only asset accepted, we need to accept free floating XRD along to avoid those LSUs being targets for ancillary markets and vote-power trading.
That being said, only the active Validators Staking LSUs should be considered, not the LSUs on non-active validators. And, most importantly, not any LSU from an unregistered validator.
It’s those LSUs that potentially represent the biggest risk/incentive to creating ancillary market opportunities and vote-renting or vote-selling economics - I can go into details if needed, but just think on having an LSU that’s 1:1 with XRD and never accrues or decreases any value and can be made ready available using a mechanism similar to “instant unstake”. And at the same time, only such LSUs allow you to vote.
All such cases have a huge detriment if free floating XRD is given voting-power, as no one is going to engage in crafting any such system, it becomes useless.
Other types of assets, like the ones representing lock-up XRD in Radlock could be considered as “skin in the game”, to use the expression - but then they open a pandora’s box as we’ll need to constantly be aware of every type of instrument made available in the ecosystem and assess it to determine if it should be included or not.
I argue that this is not practical for our immediate needs and will need much complex and careful coding, if we want to implement it somewhere in the future
Except the fatties have a lot of McDonalds stock.
Token holders are not customers. Any holder of XRD should expect to have a say in governance. Staking is barely different then holding simple XRD. LSU can be as liquid as XRD. I don’t get the analogy. You are just drawing an arbitrary line on what should represent a vote. If any line should be drawn it could be around what(XRD/LSU) is in your voting address at snapshot time verses XRD you still technically own but do not have direct control of as it maybe staked in a dapp for example.
So because “all other projects do it that way” is your reason for why we should do it that way?
Not much of a selling point.
So in your view the people who are trying to quickly (short to medium term) profit off of the network should have the same vote power as someone who holds a longer term view that securing the network is more important?
The first group will sell and run as soon as a better opportunity (Layer1 + Defi Strategy) comes up. They could care less what Layer 1 that is. e.g. The fatties will run to whichever fast food chain has the cheapest burgers.
If we give them voting power they’ll burn through our treasury to ensure our burgers are cheapest and once the treasury is gone they’ll swoop down on the next project like vultures.
Just curious: WHO do you think are Radix’s customers? According to you we either have none at all, or you think that everyone is a customer. I’m not sure which. Can you clarify your position?
I think the best argument for doing xrd held in account, lsus held in account, stake claim NFTs held in account is they are all natively understood tokens on the ledger and can easily give a simple xrd total from redemption values. Anything beyond this will inevitably cost more to maintain. Whether its liquidity pools popping up that must be added, new vesting contracts where xrd/lsus are locked, or new defi protocols to deposit xrd, there will be increasing overhead for development time to add things in to make it “fair”.
Regarding using other non native tokens, I don’t know how it would increase our setup, legal, or compliance costs. I can argue that using xrd, lsus, and stake claim nfts are using a single token since the ledger understands and can return their XRD equivalents.
I think we have bigger fish to fry and we should revisit this when budgets aren’t as tight as they are and chart looks healthier. Until then we want this process as “set and forget” as possible.
Trigger warning: THis-like post coming up. , see a TLDR on the bottom. Also, I do not claim to know it all, even though this post is written like I do.
There’s currently a consultation ongoing to decide on which token(s) should count for DAO membership. It is quite the controversial one, it seems.
It’s easy to argue for any of the options intuitively (as in, “this feels fair”), but I reckon we would benefit from going back to first principles and reason our way to the end, step by step.
So, where do we start?
Imo, the central question is “what is the goal of our DAO?”
If we know the answer to this question, we can answer the question of which group of people is most aligned with this goal - or in other words “who does our DAO represent?”
Intuitively, there are two possible answers here. Our DAO’s goal is to make:
Now, while the first might be the more principled one - and the one I’m sure will be more in line with the DAO’s official purpose - it is a vague goal. How do we even quantify success?
Ironically, I would argue we largely do so through XRD’s price.
XRD’s price is not a perfect measure of Radix’s success, but it is by far the best objective proxy we have for it over the long term. Therefore, making Radix successful is closely tied to making XRD’s price go up.
So, whichever of these two goals you deem superior, the group that aligns most with it is everyone that wants XRD’s price to go up, especially over the long term.
Now, who wants XRD’s price to go up? Congratulations Sherlock - it’s people that own XRD.
So how should membership of our DAO be determined? Ownership of XRD.
Attentive readers would now interrupt and point out that I’m trying to trick them. Because, aren’t there people that want XRD’s price to go up, but don’t own XRD?
Yes, there certainly are. To include them would be the ultimate utopia. Still, I’m discarding them on the grounds that there is no objective way of quantifying how much they want XRD’s price to go up. There is no good way to proceed here without just ignoring this group.
That said, I do not mean to imply that this group is unimportant - only that it cannot be represented cleanly in a voting system without introducing subjectivity. I’m not opposed to allocating voting power to this group in the future. If we ever find a robust and objective way to accurately represent users that want XRD’s price to go up without having direct exposure to it, I think it could absolutely be useful to include them.
Yes, actually.
If you’ve been paying attention, you’ll probably guess my ideal scenario would be to allocate voting power by XRD ownership, or more precisely, exposure to XRD price movement.
The best way to measure this would be to add up a user’s net XRD, LSUs, and LSULP. This of course includes any of these tokens locked up in DeFi (ask the tax man if you own these and have exposure to them price-wise, the answer is most certainly yes).
A little earlier we discarded utopia, but I’d call this an almost-utopia.
Now, I’ve seen some people argue that calculating voting power using LSU holdings only would be better, because it is better for the network if everyone would stake. This seems like a great idea, I agreed with it at first too.
Imo, that logic is faulty though.
First of all, one could argue that the premise “more XRD staked is better” is subjective, and therefore shouldn’t be used.
But secondly - and this is the real fallacy - it confuses:
The former would be a democracy (which I assume we want), and the latter would be a meritocracy (which I assume we don’t want).
People don’t get a say in our DAO because they are contributing (staking). People get a say because they are aligned with the mission (having XRD exposure).
If the people aligned with the mission deem staking to be so important that stakers should be rewarded - which they already are, by the way - the proper way to do so would be to incentivize staking even more, e.g. via a proposal to increase staking rewards.
Anything with the word “utopia” in it makes some alarm bells go off. In this case rightfully so, I’m afraid.
Why?
Because sadly, I think for now our almost-utopia is not fully feasible.
It will be a major pain to also include all DeFi positions in XRD exposure calculations to determine voting power. Therefore, I’d argue - which I feel I’m often doing lately - for the pragmatic approach:
We count LSUs, XRD, and LSULP.
No other tokens for now.
Truthfully, my aim with this post is not to decide on exactly the degree of pragmatism we should adopt.
The aim is to agree on the premises underpinning my almost-utopia.
If we agree on those, we can try our best to get as close as possible to it, while staying in the realm of what is realistically possible.
So, THis and THat was your alt? 
Thanks for the write-up, I don’t necessarily agree with all your reasoning but the end result is the same as I feel we should go with for now.
On detail: you also need to account for any exposure under the form of unstaking, so claim badges for any outstanding unstaking should also be included - as they represent a transient state between two different forms of exposure that are accounted and valid for voting purposes.
Don’t forget to cast your vote on it, btw.
Are LSULPs simple to calculate/track?
Yea, very simple. Easier than LSUs.
Read only the TLDR but… Agreed!!!
this is great but consider also that Root + Weft are storing about 650k usd in Xrd + LSULP so not counting this will certainly decrease this number that are important for lending protocols.
More they are not really difficult to track
Just in: Root dev thinks we should track Root!
Messing with ya 
, you’re right. But there is obviously the issue that if you support 1 protocol, you kind of need to support all of them.
And in that lies the complexity of it all, as we don’t have
a) a proper registry of ecosystem accredited/accepted solutions/platforms/whatever
b) an easy way to seemingly “upgrade” the process as such solutions are born and deployed and/or retired
Pragmatism and strong time constraints demand us compromises … smtg that not all may be aware nor willing to accept, as each votes to whatever best fits its own strategy for exposure :-/
I like to think I can generally be very pragmatic, and I believe (and sincerely hope) many here can be too. If not, governance is going to be rough.
To add some context, a model I think we could follow is the one DefiLlama uses:
This process puts the burden of implementation on the protocol that wants to be added to the vote counting itself (and a little bit on the maintainer, they have to review PRs, but it’s a fairly small task).
If we start with only LSUs and XRD (maybe not even LSULP!), we could use this kind of model, and nobody will have received preferential treatment.
I am concerned about added compliance costs or additional legal exposure for the dao or whatever the next governing entity is if we add additional tokens or defi positions with the accepted tokens. Would we not be adding at the very minimum smart contract risk into governance decisions?
Do you know any examples of other daos that track defi positions to count toward governance votes?
If my concerns are unfounded, I would be open to supporting at least lsulp.
yes that could really great, each protocol puts its own effort to helps its supplier having voting rights
yes sure but you can always start from the one that has the most TVL and proceed in cascade with the others, the dApps are few. Otherwise the other solution to let the protocol owners to prepare PR is good. Also only xrd/lsu/lsulp is great… most important is to have the ball rolling.. some details will be fixed on the go
I agree integrating a new Dapp, the burden should fall on the Dapp creator.
But for now, being prahmatic, XRD, LSU, LSULP and HLP are relatively easy to implement and should be included ASAP.
I also agree with Leonets “start from the one that has the most TVL and proceed in cascade with the others”