In the context of the adage "give a man a fish and he'll be fed for a day, teach him to fish and he'll be fed for life," one can expect cash given to the first group to be spent on fishing rods, while cash given to the second to be spent on fish.
I find giving cash to the first group to be morally different than giving cash to the second group, and I suspect that poor Kenyan villagers are likely to belong to the first group, while beggars in wealthy developed countries are likely to belong to the second.
One can also reasonably judge that on the balance of probability, aid money will almost certainly have a significant positive effect on the lives of Kenyans. The effect of a comparable amount of cash on the lives of beggars in Western countries is much less unequivocally positive (potentially a net negative if its spent on substances that harm the beggar's health, or the income disincentives them from seeking programs that aim to help them with more than just their next meal)
I'm surprised you think there's no moral difference. There would be a moral difference between saving someone's life with $20 and treating someone to a meal at Starbucks for $20, right?
While I would like those beggars to receive money, I don't want them to be the exclusive recipients of cash aid, excluding the many poor people who don't happen to be actually begging on the street. I want an organized program that takes the available cash (donations or taxes) and distributes it in a "fair" way among the poor. What constitutes fair is difficult to answer, but I certainly don't have enough information available to make that decision right when someone asks me for money.
Giving money to a beggar in the US would not do nearly as much good, both because the US has excellent safety nets (hospitals that will treat you, food banks) and because everything in the US costs more (so you can provide less of it to someone in need per dollar).