Why is microwave better than visible light for point to point data communication? In the electromagnetic spectrum, it appears that the shorter the wave length, the better the bandwidth due to higher frequency. Some communication systems utilize microwave to transmit data to offer high-bandwidth line-of-sight data communication. My question is why isn't visible light (or any spectrum with even higher frequency: UV, X-Ray, Gamma) in widespread use? Shouldn't these be carriers of even higher possible bandwidth?
 A: The problem is that the shorter wavelengths get absorbed by anything in the middle, but micro and radiowaves (depending on the freqency), are either transparent or bounce back from the upper athmosphere. The highest frequencies (such as x rays) can pass trough objects, but they are harmful to us. 
A: To have a communications link, you need emission, transmission, and reception.  For transmission, there are well measured windows in the atmosphere.  The atmosphere transmits visible very well, and some lower frequencies, but X-ray and gamma not so well, so those are out.  For emission, you need some way to modulate the signal, which so far has meant transistors that operate at that frequency, then some way to mix it onto a carrier.  Even if you carry the signal in the visible, the bandwidth is what the modulating transistor can do.  As it is easier to mix it with a carrier in the RF/microwave, that is what we do comercially.
A: There are two basic problems with visible light communications: (1) atmospheric absorption and scattering, (2) antenna directivity/beam width and its size relative to wavelength. The absorption is easier to understand, which is insignificant below 10GHz, and then progressively increases with huge discrete lines, see http://upload.wikimedia.org/wikipedia/commons/3/34/Atmospheric_electromagnetic_opacity.svg
The antenna problem is a bit more subtle: to overcome the propagation spreading proportional to $\lambda ^2$ one needs narrow beam antennas that are many wavelengths across. Unfortunately you cannot have it both ways: both narrow beam and wide beam simultaneously. To communicate with visible light you need directional beams and then you must acquire and track it and that is a difficult thing to do for a mobile system. In contrast, your car radio does not need to any of that. Even for fixed sites, point-to-point you would need to keep the radiators clean, dirt on the optics would scatter the light while your rooftop antenna works even when covered with ice.
While there are no transistors for visible light there are for far infrared, and lasers themselves can be modulated quite well. Neither modulation nor detection would be limiting  communications (or radar) below frequencies of UV and above 150GHz but almost everything else would.
A: Carrier wavelengths close to the visible part of the spectrum are used in optical fibre links. The precise wavelengths are chosen to minimise attenuation in the fibre. But as Ross points out, the bandwidth is then limited by the modulation/demodulation electronics.
A: First of all using visible light to transmit data is definitely possible. We can argue as much as you want on the fact that the transmissions are not optimized and that the efficiency of such techniques are limited to very small distances (because of absorption and attenuation in the air).
About UV I would say that there would be a lot of interference given by the fact that the sun is a major source of noise in that portion of spectrum. Regarding X and Gamma rays, I can only agree with some people here who pointed out the danger for any living organism. Their energy is also very high that they can change the atomic structure, therefore possibly damaging (cheap) electronic equipment that receives those signals (this last sentence is not totally true though, since it is actually possible to receive those signals with expensive technology ensuring a decent "life" to the electronics involved)
