Acid dissociation is usually exothermic, and it has an increase in entropy, so the change in Gibbs free energy should be negative. So why isn't always $K_\mathrm{a}>1$ in accordance with $\Delta G^\circ = -RT\ln K$? For example, when $\ce{H3PO4}$ dissolves, the change in enthalpy is $\pu{-14.2 kJ/mol}$, and the entropy presumably increases, but $K_\mathrm{a} = \pu{7.6E-3}$ (less than 1).
Edit: Apparently acid dissociation doesn't usually increase entropy, so my new question is why that is.