First off, what am I even talking about? Happiness is going to be defined ostensively for a second, but by "the priority" thereof I mean the intention of setting as one's first priority being happy. It is often remarked, I've seen, that the only reasonable course of action is to (try to) be happy. That is, "There's no rational reason to aspire to anything else," goes the claim.
On a very high level of description, Immanuel Kant defines morality in terms of principles with unconditional priority over others. (This is a simple translation of the frequently convoluted phrase "categorical imperative.") Likewise, John Rawls writes:
The entire question of "morality over happiness" is just the question as to whether anything has priority over happiness, even. This is where the confusion arises for some over whether the absence of divine retribution in the world would license the loss of moral concepts. The truth that should be evident enough is that this is not the case, for it is not even the case that the goal of being happy is logically structured so as to allow it to have unconditional primacy in one's decision-making.A further condition is that a conception of right must impose an ordering on conflicting claims... The fifth and last condition is that of finality. The parties [to the transcendental election in the original position] are to assess the system of principles as the final court of appeal in practical reasoning." [A Theory of Justice, sec. 23]
As Kant elsewhere writes:
Kant, Rawls, and Bernard Williams have all remarked that the concept of being happy is just the concept of having achieved all of one's goals, more or less (I think Williams puts it in terms of "projects" or something). Considered in this light, it is clear that becoming happy is not able to directly pertain to one's choices as a reason to act one way or another, for it is only acting on a prior reason (a prior goal) that one can attain the state of having a goal fulfilled. That is, something else has to be elected to the role of one's first priority (one's ultimate goal) in order for one to be happy in having fulfilled this goal.The commonest intelligence can easily and without hesitation see what, on the principle of the will's autonomy, requires to be done... the moral law demands the most punctual obedience by everyone; it must, therefore, not be so difficult to judge what it requires to be done.
... If only it were equally easy [compared to technical decision procedures] to give a definite conception of happiness... [such a conception] is not an ideal of reason but of imagination, resting solely on empirical grounds, and it is vain to expect that these should define an action by which one could attain the totality of consequences which is really endless.
2. A concrete reason
Here's a thought experiment designed to show that it is logically possible for something to divest happiness of its priority over at least one other thing. Suppose you had a disease whereby every time you were happy, you suffered an extreme migraine. If you were in this situation, would it remain true that you would have a reason to try to be happy? Note that even setting yourself to abandoning your goals as a means to avoid being happy would, as a goal, even if otherwise achieved, cause you acute suffering. It seems likely that you might end up with chronic, if fairly dull, pain.
You might object that in this case, it is not that being happy is no longer one's goal in the sense that you would be happy if you could, but since you no longer can, you make no effort to achieve your priority in this case. I think this objection mistakes the nature of prioritizing. In a system that is even partly deterministic, as even libertarian free will would have to be (I am an indeterminist about this topic, but only in recognizing some small pockets of reality in which free choice exists), there must be some functional structure to that system. Practical reasoning, inasmuch as we execute our planning by its output, is realized in how we act. Akratic we may be, yet this is only in falsifying our priorities in the context of what provokes our incontinence. What we give priority to just is how we immediately determine ourselves to act; it's the mechanical form of the shove we give ourselves in our directions. The simple fact is that what we do is give avoiding pain preference over high or low pleasures when we reject the latter for the former in the sum of our choices. That is our character in this case.
The negation of agony may be passive or aggressive. In a sense, Buddhist thought involves a theory of emotional passivity as the solution to the problem of rational priority. Nirvana is attained not by forcing our desires to disappear but by falling somehow into the transcendental mists behind them and ceasing to will the energy that strains them. Now there is something complex about the antinomy between inaction as free will and action as the same, and Buddhism arguably resolves it somehow, but I'm not competent to say whether it does. I'm just making a suggestive remark about the issues, here.
Happiness is an emotional state, though the abstract nature of these states in general is still up for debate. David Hume seems to have thought that sympathy was the emotion superior to happiness; Kant said it was respect; I would vote for saudade not as the only one better, but nonetheless the best. It would, I think allow for an elegant solution to the problem of moral/rational decision-making for someone infected with the hypothetical sickness I described earlier.
Plus romantic love.
A much misunderstood subject. See...the whole idea of happiness as a goal or objective or whatever that everybody must attain is very very new. This book on the subject looks very interesting: 
Everything is about perspective. 
