The term 'root cause' is problematic because it assumes a simplistic cause-effect model. However, a problem may be the consequence of a complex 'chain reaction' of events. Most charities identify a single event leading up to a problem and focus on eliminating/mitigating that.
E.g. a problem is: people in country X dying unnaturally every year. The immediate cause is starvation, which is indirectly caused by famine as a result of multiple factors like poor farming technology and government policy. There is no 'root cause', and attempting to solve the farming problem might be much less resource efficient harder than just providing emergency rations every year.
Another problem with your wording is 'change for the good'. If famine death is reduced by 100 deaths for each 1000USD donated, is this a 'change'? Or would Alfred only take action if it would possibly result in entirely stopping the deaths 100%?
I agree with your point about incentives in philanthropy though, there are some organizations that do a better job than others about that.
Say there is a situation that Alfred have the power to effect. To reliably cause change for the good Alfred must:
(1) Identify a problem and its root cause
(2) Change the root cause
(3) Confirm that the problem is solved.
There are people who claim you can leap straight to (2) and skip (3). These are people who (if they ever honestly checked) would discover that they are not responsible for very much positive, meaningful change. They are flailing, looking busy and have the best of intentions.
For anything reported in the news step (1) represents a huge investment of time because the news has probably got the situation at least somewhat wrong. To say Alfred has a moral responsibility to do something is equivalent to saying Alfred has a moral responsibility not to read the news; because Alfred doesn't have enough time or resources to confirm what the problem is for all the sad stories he can read about.
In your specific example - 'sending money to NGOs with workers on the ground' - Alfred would probably divert a lot of money to organisations who are good at emotive advertising campaigns with administrative costs verging on embezzlement. We have good historical evidence that Alfred's best bet is to go all-in on capitalism, make a fortune, and then do some philanthropy to maximise his impact.
Giving to a worthy cause creates incentives that are terrible. I can't be convinced that it is a good idea. I have a lot more respect for people who volunteer time, but even then I'm suspect of their ability to correctly diagnose the problem they are trying to fix.