I'm going to revisit it, as the idea of power being an artifact of free consensus of belief in a group, and not intrinsic to violence is hard to reconcile with some of her earlier observations, and indeed, physics.
Your "two levers," I think are interesting in that coercion is, in my experience of popular political argumentation, allowed to sneak in via a societal focus on physical violence ("force" in your terms?).
Example: some leftists claim property violence is not violence because, paraphrasing, is does not harm the human body. Leaving aside the validity of this claim, the coerciveness of the act remains but is somehow able to pass the smell test in certain circles. I think this depends on one's perception of conflict, but that is a long discussion.
Meanwhile, as you imply above, deception is clearly coercion, but much more effective in the longrun than physical violence. Deception generates belief congruent with the deceivers' attempt to acquire (more) power, which, once established, requires less maintenance than fear.
Finally, there is nothing in the description of power as cited above that is necessarily free from coercion. If power simply means both individual and collective agency, then capturing agency via coercion is still an acquisition of power.
Oddly enough, they want to specifically use the word "power" because of its existing reputation, i.e. its connotations with acts of non-consensual control. They may say something like "real power does not stem from violence, real power comes from a group and does not exist without group coherence." However, the only reason they're focusing on the word "power" is because they know what it implies and are using its reputation as a concept in order to lend credibility to their own ideas, which themselves usually contradict the agreed-upon meaning of the "power".
To be clear, this may not strictly apply to Hannah Arendt, but I see other people playing with the word "power" so frequently that I thought it was relevant to mention.
I agree with her that there's a distinction to be made between, on one hand, a group of people agreeing to follow a set of rules to accomplish some goal and, on the other, a slave driver forcing his slaves to accomplish that goal.
The idea that there is no real distinction here, that all power is based on coercion, was popular when she wrote the linked essay and it's popular today (as evidenced by the replies here). I used to think that way but I no longer do and, I have to say, it's quite a relief.
I'm not saying this necessarily makes you wrong, but this is explicitly not what she thinks. In the version of this essay published in Crises of the Republic, she writes:
> It is, I think, a rather sad reflection on the present state of political science that our terminology does not distinguish among such key words as power, strength, force, authority, and, finally, violence – all of which refer to distinct, different phenomena and would hardly exist unless they did.
In other words, power is not reducible to some combination of "force and deception". In the edited version of this essay, she even defines each of these categories:
> Power corresponds to the human ability not just to act but to act in concert. Power is never the property of an individual; it belongs to a group and remains in existence only so long as the group keeps together.
> Strength unequivocally designates something in the singular ...
> Force ... should be reserved, in terminological language, for the "forces of nature" or the "force of circumstances," that is, to indicate the energy released by physical or social movements.
> Authority can be vested in persons – there is such a thing as personal authority, as for instance, in the relation between parent and child, between teacher and pupil – or it can be vested in offices, as, for instance, in the Roman senate or in the hierarchical offices of the Church. (A priest can grant valid absolution even though he is drunk.) Its hallmark is unquestioning recognition by those who are asked to obey; neither coercion nor persuasion is needed.
> Violence, finally, as I have said, is distinguished by its instrumental character. Phenomenologically, it is close to strength, since the implements of violence, like all other tools, are designed and used for the purpose of multiplying natural strength...
Taken from http://www.fsmitha.com/review/arendt.html because I don't have a text version of this essay, but you can find it here https://monoskop.org/images/7/70/Arendt_Hannah_Crises_of_the... (154 of the pdf).
The point is that, for Arendt, power is not about coercion and it's not about deception or force. It's about consensus, shared opinions and judgements, and the proof of this consensus, which is a group of people's ability to act in concert.