Humans of all kinds can breed and make new humans easily. Same for dogs... easily can breed with each other and increase diversity. So in both cases same species. But humans and chimpanzees? Probably not (albeit there are rumours? :) ), so different species.
It's my opinion that if aliens visited earth 30k years ago and used our definition of the word, they would classify us as different species. To be clear I'm not advocating that we should refer to humans as different species, just that because we're capable of such advanced thought we should abide by different rules than we hold other animals to.
All human beings from all over the world can interbreed and their children will be fertile themselves (with the same rate of accidental infertility they would find in their own populations). As such, it would be completely preposterous to claim that there is more than one species of human on the planet today.
Two populations definitely do not belong to the same species if they generally can't interbreed even if they are morphologically identical.
Many accepted species will interbreed in captivity or rarely in the wild.
*: not taking a position on human species
Probably the better example you could give is bears, where brown and polar bears are always considered separate species, but they do successfully interbreed, even in the wild occasionally.
The species concept is a little too vague, and essentially the difference between two closely related species and two distant varieties of the same species is fuzzy and often historically biased (if the two populations were considered separate species for a long time, there is large inertia against re-categorizing them as a single species when evidence of genetic similarity and interbreeding arises).
If Aliens landed on our earth 20k years ago and used our definition of the word there's no doubt in my mind they'd consider us different species, we'd be like bears. As far as I can tell the only reason we don't classify people as different species is for the sake of society, as people would naturally then try to determine differences and rank them, which has an obviously bad long-term outcome.
However, few of these arguments apply to human populations. With the exception of dogs, cats and other domesticated animals, few or no other mammal populations that have diverged as recently as humans are categorized as separate species (for example, brown and polar bears have diverged ~600k years ago, before our species even existed). Human populations only began spreading out of Africa some 150k years ago, which is not a lot of time in evolutionary terms at all, for a mammal that only reproduces around 1/year and needs more than a decade between generations.
Not to mention, the separation between all human populations is merely geographical. When different human populations came into contact, even ones that had been isolated from each other for tens of thousands of years, they virtually always interbreeded in significant numbers, contrary to, say, wolf and dog populations. This is also a strong argument against considering distant human cousins different species.
Additionally, morphological variation among isolated human populations is actually quite small, compared to intra-population variation. Skin color is the single biggest morphological difference, by far; followed by the size/hardness of our hairs, especially the ones on the head. Apart from these, virtually all other differences are more about statistical prevalence of certain traits (e.g. blue eyes are more or less common in various populations, body hairyness is more common in some populations than others, tallness is more common, etc).