We can calculate recoil easily. Let's assume we want a laser gun with the same amount of recoil as a pistol. Picking a bullet at random, I find a weight of 8.04 grammes and a speed of 360 m/s.
The formula for the momentum of light is easy. P=e/c.
The result is 860 MJ of energy, 6 orders of magnitude more than that of the bullet. Suffice to say, recoil would not be a problem.
http://m.wolframalpha.com/input/?i=8.04g*%28360m%2Fs%29*speed+of+light&x=0&y=0
https://en.m.wikipedia.org/wiki/9×19mm_Parabellum
At those energies, you start to bump heads with the restraints I mentioned on page one. That's why I mentioned them. Basically, when you start emitting that much laser energy that you get bullet like recoil from the radiation pressure, your beam will self-refract and other nasty things. Basically, you cannot make a laser weapon that will do this.
Certainly not in a handheld form factor.
Here's a link to a pulp science article which provides both a link to the source research and give the derived hard limit on laser energy before spontaneous QED breakdown happens and the beam instantly becomes incoherent.
http://theastronomist.fieldofscience.com/2010/08/limits-on-lasers.htmlAccording to the article, that hard limit is 10 to the 25 W/cm.
IIRC, the bloom limit in atmosphere is substantially lower, closer to 10MW. (there's a reason why LaWS is a 50KW pulsed laser!!)
In short, a laser pistol capable of producing recoil equal to a conventional pistol will cause the air in front of the gun to bloom with astounding energy, equal to a small explosion, and will probably cook/kill the soldier firing it. Only a tiny portion of the beam energy would make it to target.
For practical reasons, such a gun is not useful as a sidearm, and can be discounted from being reasonably considered.
Again, that was the point I was delicately trying to convey.