Noahpinion has some good thoughts on this question that Bryan Caplan and Karl Smith are going to debate soon. Some of it waxes utilitarian: "When I witness the urban blight, violence, drug abuse, and other social ills that poverty may be causing, as a non-poor person I have an interest in preventing these social ills from affecting me, regardless of whether the ills are "deserved."".
This is a good point too: "That said, I think the Caplan definition of "deserve" is not as "uncontroversial" a moral premise as Caplan declares. The reason is that it is a partial-equilibrium definition, not a general-equilibrium one. If we live in a society in which X percent of the populace must be poor, then no matter what set of actions is taken by the population, some people will wind up in poverty. To see this, imagine that we lived in a society in which the hardest-working 50% of people get to be spectacularly rich, and the other 50% are forced to live in squalid poverty. In this society, if everyone raises their effort by 1000%, the number of people in poverty stays exactly the same. I doubt that most people would say that the lower half of the population "deserved" to stay in poverty after raising their effort by 1000%! But that is exactly what Caplan's definition implies."
Persistent underutilization of the factors of production - among them labor - is usually considered to be a macroeconomic question requiring macroeconomic answers. For some reason, though, a lot of poverty gets discussed in microeconomic terms. I've never been quite sure why.
One concern of mine about this talking about "deserving" poverty is that whatever decisions people make that put them in poverty aren't made in a vacuum. When you're a kid you are crucially dependent on the investments that society makes in you and that your parents make in you. If there are youth populations that parents and society systematically underinvest in, those populations are going to end up with higher poverty rates. Now, we probably could trace that to decisions they make in almost every case. Even factors out of their control can ultimately ascribed to their own decisions. Depressed local labor markets? Why didn't you just set out on your own - why didn't you move? But that seems to miss the point. Would anyone else, put in the same circumstances growing up, have made different decisions on average? Probably not. There's plenty of variation in human behavior and that variation does affect outcomes - but it's still constrained by these circumstances.
This distinction between "equality of opportunity" and "equality of outcomes" is largely made up. If you give a poor black kid and a middle class white kid a "color blind" college admissions test or a "color blind" employment test, you can't call that "equality of opportunity". Opportunity goes considerably deeper than the immediate circumstances of any given choice. Opportunity is something that builds up over years of investment (or lack of investment) in a kid.