Relation between the spin of a field and the way charges interact

When solving the problems of V. Rubakov’s "Classical Theory of Gauge fields" book, I encountered the following phenomenon:

• For a real scalar fields (spin 0) $phi$, if we consider the action with source $rho(overrightarrow{x},t)$,
  $$S=int d^4x left{ frac{1}{2}partial_muphipartial^muphi-frac{m^2}{2}phi^2+rhophiright}, $$
  and that we take the source associated to two charges $q_1$ and $q_2$ in $overrightarrow{x}_1$ and $overrightarrow{x}_2$,
  $$ rho(t,overrightarrow{x}) = -q_1delta^{(3)}(overrightarrow{x}-overrightarrow{x}_1)-q_2delta^{(3)}(overrightarrow{x}-overrightarrow{x}_2),$$
  then the energy of interaction of those charges (obtained by extracting the divergent terms of the Hamiltonian) decreases when $|overrightarrow{x}_1-overrightarrow{x}_2|$ decreases if $q_1q_2>0$. In other words, charges of the same sign attract each other.
• For a (real) vector field (spin 1) $A_mu$, if we consider the action with source $j^mu(overrightarrow{x},t)$,
  $$S=int d^4x left{ -frac{1}{4}F_{munu}F^{munu}-ej_mu A^muright}, $$
  with $F_{munu}=partial_mu A_nu-partial_nu A_mu$,
  and that we take the source associated to two charges $q_1$ and $q_2$ in $overrightarrow{x}_1$ and $overrightarrow{x}_2$,
  begin{align}
  j_0(t,overrightarrow{x}) &= -q_1delta^{(3)}(overrightarrow{x}-overrightarrow{x}_1)-q_2delta^{(3)}(overrightarrow{x}-overrightarrow{x}_2),
  j_i(t,overrightarrow{x}) &= 0,
  end{align}
  then the energy of interaction of those charges (obtained in the same way) decreases when $|overrightarrow{x}_1-overrightarrow{x}_2|$ increases if $q_1q_2>0$. In other words, charges of the same sign repulse each other. As in well-known electromagnetsim.
• For a spin 2 field (the metric for example), the behavior switch again: charges of the same sign attract each other. This is pretty intuitive when when it comes to gravity.

The pattern seems pretty clear.

1. What is the fundamental reason for this to happend?
2. What about half-integer spin fields?
3. Does it continue for higher (than 2) spin fields? I heard that higher spin fields are much more difficult and that is was still not much understood. Why is that?