Everything people do is political to some extent, if we conceive of politics as the intentional influence applied to individuals / everyone as a consequence of outlooks or beliefs held by a majority or a group of individuals in a position of sufficient power.
The point of politics is to influence people. If it didn't work, people wouldn't bother with it.
That said, the sociology around science is awfully good at shielding the methodology required for generation of reliable and lasting knowledge from the pressures to get results of a certain sort that would be convenient or desirable (from the perspective of individuals or groups in a position of influence).
(Incidentally, I do not know why you say that science (as a process) is not a way of testing explanations. That is just about all it is.)
"Hard" sciences are very heavily data-driven, and no matter how much you might want data to come out a certain way, it is very hard to get those wants to reliably produce the "right" results especially across different labs and after taking the precautions that one is supposed to take (scoring done blind or double blind, appropriate controls, etc.). It is possible for enough pressure to be applied that normal processes are not followed, but it's almost always painfully obvious when you really look into it. (A lot of normal, uninfluenced science doesn't have proper methodology or controls also. Scientists need always be on guard for such things.)
Because one is not supposed to lie or fabricate data (and the consequences are supposed to be, and usually are, really horrible socially), and because one is supposed to report clearly one's methodology (and it is obvious when one has not), coming up with apparently very strong evidence that something is true when in fact it is not is incredibly difficult. That, if it could happen, would be the most damming kind of non-objectivity.
Instead, politics is limited to kind of a soft power to influence results: by allocating funding to only look in certain areas or by making one a pariah if one gets certain kinds of results. This may leave our knowledge incomplete or uncertain in politically-motivated ways, but it doesn't much taint whatever we conclude we know robustly.
Of course it's always possible for everyone to be so strongly culturally biased that nobody manages to see through a flaw in interpretation, or for nobody to challenge the dominant view on things, but because of the data-centrism you typically notice in such cases that we just don't understand the process very well: the science is not all that predictive. And that goes for practically all of the softer sciences. So whether or not they are influenced, we already know we shouldn't be too sure of these results.
So the hard sciences are tarnished a bit by politics, but not terribly much.