As mentioned below, one way to think about it is that EM only affects charged particles (and depends on their charge), whereas because gravity is acting on the underlying space-time it has a universal effect on everything (including light).
We can pretty much boil EM down to: like charges repel, unlike attract, strength is charge1*charge2/distance^2. What about magnetic field, photons, QFT etc?? None of this exhibits effects which could be described as stretching space-time either.
But we cannot do the same with gravity. An explanation like the above but for gravity (which is traditional Newtonian) leaves out many, now observed effects such as:
-> time dilation (GPS relies on this calculation) (measures time stretching and contracting)
-> gravitational waves (LIGO) (measures space stretching and contracting)
How do we _know_ any of this? People propose theories, those theories are then tested against experiment. AFAIK to date there is no experimental evidence suggesting EM stretches space, and no theory proposed that includes such an effect and correctly matches experimental data. That's the most holistic answer (but unfortunately one you just have to believe unless you have a lot of spare time!)
We can pretty much boil EM down to: like charges repel, unlike attract, strength is charge1*charge2/distance^2. What about magnetic field, photons, QFT etc?? None of this exhibits effects which could be described as stretching space-time either.
But we cannot do the same with gravity. An explanation like the above but for gravity (which is traditional Newtonian) leaves out many, now observed effects such as:
-> time dilation (GPS relies on this calculation) (measures time stretching and contracting)
-> gravitational waves (LIGO) (measures space stretching and contracting)
How do we _know_ any of this? People propose theories, those theories are then tested against experiment. AFAIK to date there is no experimental evidence suggesting EM stretches space, and no theory proposed that includes such an effect and correctly matches experimental data. That's the most holistic answer (but unfortunately one you just have to believe unless you have a lot of spare time!)