The center of this question seems to lay in the nature of experience. As far as stimuli are involved, Regret has it. I wouldn't go as far as saying we behave in "Skinnerian" ways, but that can just depend on what you define as the nature of stimuli. Anyways, this isn't a discussion of behaviorism.
So, what is experience? Well, I'm going to make some assumptions. The first is that experience in this case is limited to conscious perception. And secondly, that by "experience" we are talking about reportable and evaluated memories or thoughts, ie, what you are thinking about. Meaning that our experience is based upon the conscious perception of external stimuli and our couscous evaluation of it. This isn't a scientific definition by any means, as "experience" isn't really a scientific idea. Experience could mean so many things really, it is almost useless as a concept for rational discussion. In fact, even the term "conscious" runs into this same problem, but not so far as this post is concerned.
Alright then, conscious perception of stimuli is really interesting, and poses what is called the "binding problem" in cognitive psychology. Basically, we have evidence of how all the stimuli gets into our brains via receptors, and where certain values are attached to it for various evaluative properties, but we can't pinpoint any location where all the stimuli "meet" to become our overall "experience". Stimuli does not become conscious until it has already received emotional and evaluative processing. This is show by the existence of conditions like agnosia and in the effects of optical or other sensory illusions. Agnosia is a break in the pathway that processes emotional information of stimuli. Once that is done, the victim is unable to attach emotional value to objects. And in the case of optical illusions, we are unable to make ourselves "not see" them, even when we know the processes behind them consciously. Both of these points are proof that our conscious perception of stimuli is built upon underlying unconscious processes that we have no control over (this is a lie, we can have some priming effects, but I am being very general).
alright. Moving on we deal with the way we think about the stimuli that have just come in. Well, to begin with, what ARE these stimuli now? the best estimate (in my opinion) is that they are represented as a pattern of neuron activity. This pattern of activity must be processed through a part of the brain aptly named "the interpreter" that is responsible for our semantic and linguistic (reportable) experience of an event. In epilepsy patients, sometimes a part of the brain known as the corpus callosum is cut to prevent seizures. This unfortunately prevents any information from the right side of the brain from being processed by the interpreter, as the corpus callosum carries information from one side of the brain to the other, and the interpreter is on the left side of the brain. Subtle experiments in a laboratory can elicit various arousal states based on the subconscious processing of information presented only to the right side of the brain, and the subject will have no explanation at all for their arousal, simply because the interpreter has no idea what information is being given to the brain. For instance, women were shown erotic images to their right brain, got all hot and flustered, but couldn't consciously conceive why they had, attributing it to what was available to their left brain.
This poses a major issue for memory and where I think your question may be better answered. I really want to stress that this is not my area of expertise, and I am making a couple of assumptions that I am 90% sure of. However, I digress. Generally speaking, there are two kinds of memory, working and long term. Working deals with what we are currently experiencing and how it relates to our goals, long term is more to what we remember and how we feel about things. A very important point to make here is that Working memory is based more on evaluative and stimuli based processing whereas longterm is based more on emotion. An example of this would be in the type of error made by each type of processing, in short term it is more likely to be stimuli based errors (so, when trying to say the word six you say the word sex ) whereas longterm memory errors are more along the lines of misjudging the recency of favorable events or remembering events that support various ways you "feel" about an object, idea, or person.
The real significance of this is that, for memory to move from the working memory, where it is forgotten very soon, to the long term, it must have an "interpreter" interpretation to go along with the emotional arousal state. This means that, in the future, the "interpreter" will associate brain activity that you had previously given the linguistic label "happy" to as you being "happy". The more often this occurs, the more readily you will associate those patterns with "happy".
This sort of leads to how the "binding problem" of experience is solved, imho. Basically, I don't believe that the mind does bind things anywhere (specifically because there is no evidence of this) but instead forms very complex associations between each of its senses and emotional processes based on certain predetermined genetic dispositions. Basically, at a basic physiological level, we think pain is bad but soft touch is good, and other weird instinctive things. As we develop, those predispositions to certain behaviours allow us to form bonds between what we see and what we hear, and where we are in the world. To the interpreter, this would seem a continuous experience, not segregated functions because our memories would be such that certain visual stimuli would always be associated with certain smell stimuli. This is clearly evolutionarily advantageous because creatures that saw themselves as an individual would be able to make much better long term plans than those who had to respond to each receptor channel individually, without associating sight and touch.
So... happy....
There are certain "things" that can fall upon our receptors in any of our "senses". These receptors will produce natural genetic reactions that have, through evolutionary processes, been hardwired into us to have some idea of what will or wont be good for ourselves (ie, those organisms that were not able to dictate "pain" as "bad" naturally were less likely to pass on their genes than those who could). Those things that have, over time, proved to be beneficial to humans will release things that make us "happy".
This is the best answer I can give you. However, I think you are asking the question in the wrong way. You seem to have a top down approach to the problem at hand saying "I experience happiness, therefore where is it?". That question assumes a couple of incorrect things about the brain. To begin with, the brain is not the liver. You can look at the actions of a liver and pretty much figure out what is going on. You can't with a brain. Every "experience" is based upon billions of neurons firing in specific patters to elicit specific memories based on past experiences with the perceived stimuli. You can't just look at a brain and say "Oh, that lad is happy" and more than you can say "that guy is good at math". Thought is based on patterns of firing, that for each individual will be specific and unique.
Lastly, you are presuming that because we have a "word" or "symbol" for something that is used in colloquial, that is how the brain works. "Happy" is not a measurable phenomena in the brain just because we have a word that seems to insinuate that it should be. Just because we use terms like "consciousness" or "self" or "experience" does not mean that they are accurate terms for what is occurring. You can't start with the assumption that humans "experience" "happiness" without proving that they do. To say they do because we all feel happy is a tautology, making it a logical fallacy. Humans to not experience "happiness" as we would commonly use the word on the street.