In a recent piece, DICE versus RACI, I riffed on a recommendation by Clay Parker Jones about using DICE (decide, informed, consulted, executes) as a decision-making framework. Clay wrote,

RACI is vague, hard to use, and reinforces the "what the hell is happening here" status quo. DICE is specific, easy to use, and shines a bright light on dysfunction.

As I commented:

Clay thinks 'responsible' is too vague, and leads to decision-making floating away from those implementing change. ‘Decides’ is crisp. Likewise distinguishing who executes something also makes things more clear than who is accountable.

It seems that others have similar issues with RACI, as reported by Aaron De Smet and colleagues in If we’re so busy, why isn’t anything getting done?:

We recommend a simple yet comprehensive approach for defining decision rights. We call it DARE, which stands for deciders, advisers, recommenders, and executors:
Deciders are the only ones with a vote (unlike the RACI model, which helps determine who is responsible, accountable, consulted, and informed). If the deciders get stuck, they should jointly agree on how to escalate the decision or figure out a way to move the process along, even if it means agreeing to “disagree and commit.”
Advisers have input and help shape the decision. They have an outsize voice in setting the context of the decision and have a big stake in its outcome—for example, it may affect their profit-and-loss statements—but they don’t get a vote. :: turn that around: just because your are strongly impacted by a pending decision -- like your P&L will be hit -- doesn't mean you get a vote.
Recommenders conduct the analyses, explore the alternatives, illuminate the pros and cons, and ultimately recommend a course of action to advisers and deciders. They see the day-to-day implications of the decision but also have no vote. Best-in-class recommenders offer multiple options and sometimes invite others to suggest more if doing so may lead to better outcomes. A common mistake of recommenders, though, is coming in with only one recommendation (often the status quo) and trying to convince everyone it’s the best path forward. In general, the more recommenders, the better the process—but not in the decision meeting itself.
Executers don’t give input but are deeply involved in implementing the decision. For speed, clarity, and alignment, executers need to be in the room when the decision is made so they can ask clarifying questions and spot flaws that might hinder implementation. Notably, the number of executers doesn’t necessarily depend on the importance of the decision. An M&A decision, for example, might have just two executors: the CFO and a business-unit head.

I feel this attempts to clarify, like Clay, who makes the decision.

I was mulling this, and I recalled that I had written about this from another angle in On Consent-based Decision Making [emphasis mine]:

Decision-Making Framework

I think about organizations as a network, with those farthest from the customers, partners, and markets are at the center of the organization (like conventional CEO's, C-suite, senior execs, back office folks, etc.).
Top-down decision making is characterized by decisions made at the center of an organization, and by a solitary autocratic centroid, at that (or a small management team, like a CEO and the company's VPs).
Consensus decision making is often thought of as being more empowering for employees, but in fact it is more like the tyranny of the group blocking innovation and experimentation, since everyone has to agree to the decision. This is due to the natural conservatism in groups, and aversion to risk. Consensus decision making is quite similar to Top-down, in that regard.
Distributed decision making retains the solitary decision maker like Top-down, but shifts the decision out toward the edge, closer to where the decision is being applied. This is likely to improve decisions since those involved are most knowledgeable about the issue. But like Top-down, there is the likelihood that those impacted by the decision may feel that they weren't involved sufficiently.
Consent-based decision making shares characteristics with Consensus and Distributed. The decision is moved out to the edge as in Distributed, but the group impacted by the decision have a voice. Once the decision has been spelled out -- by an individual or a subset of the larger group -- each member of the group must provide their consent, meaning 'no significant objection' to the decision. Note that this is significantly different from consensus, partly because any such objection must be raised, and its significance established.

The Consent approach shares similarities with De Smet et al’s comments regarding ‘Disagree and Commit’.

My bet is that Consent -- and not the other three styles of decision making -- will increasingly become the atom of 'fast-and-loose' organizations, from which the molecules of the new work culture will emerge. By allowing those closest to a problem to take the lead in resolving it, but balancing necessary speed and proximity with the need of those impacted to raise objections when warranted, Consent is perhaps the best general approach to organizational decision making.

Work Futures

DICE, RACI, DARE, and Consent