I think I understand what you're saying. So, to clarify, it depends on how you slice'n'dice the research areas. To give a simple example, e.g. in econometrics - small-sample estimation is a separate area (as compared to, say, asymptotics), but then you can use small-sample estimation in labour economics or some experimental economics or any other areas.
Similarly, with game theory - the theoretical results could be applied in many different topics (from political economics to micro-founded macro and more), but the theory is a separate research field.
In any case, the divisions are probably not as clear cut as one would like them to be and boundaries of research areas shift over time.