Recently I received a question from Cindi, who is part of a poodle flickr group. A discussion came up because several members of the group had found that the microchips implanted in their dogs could not be read. I won't copy the entire discussion here, but this is a basic summary: Several chips from different manufacturers in different pets and different geographical locations could not be read by multiple scanners. Within a veterinary practice 2-3 different scanners were used and none of them could find the chip. So obviously people are concerned about the microchips in their pets. And this is a good and valid topic for discussion!
First, let me point you to a post I made in 2009 about the basics of microchips: read this and then come back. I won't repeat information from that post here, and it will give you good basic information about some of the current microchipping issues (which haven't changed since I wrote it).
Microchips function as passive transmitters. The microchip itself has no battery or other power source. Until activated it's basically an inert object that simply sits under the skin and does nothing. The scanners broadcast a radio frequency over a short distance. Some scanners broadcast a specific frequency while most newer ones will broadcast multiple frequencies to cover different types of chips. Each chip responds to only one specific frequency of radio wave. When that frequency hits the chip, it gives it enough power to broadcast a small amount of data (its number) over a short distance.
Now there is a problem with microchips in the US because there are a couple of different competing frequencies (see my previous post). Modern universal scanners should be able to read all frequencies. However, there is one company, Avid, that has encoded their chips so only an Avid-brand scanner will read them. There was even a controversy because all other microchip manufacturers agreed to unencode their chips so all scanners could read them, but in the US Avid refused to do so (even though they have different frequencies and open-read microchips outside of the US). Some scanner companies found ways to "crack" the code and can now read Avid chips. However, this isn't consistent or reliable, so in some cases an Avid microchip can only be read by an Avid scanner. I'll be nice and not give my personal feelings about Avid because of this issue (though you can probably guess that I'm not a fan of them).
In my opinion, microchip failure is very rare, and overall they are one of the best ways of identifying your pet if lost. I believe in them enough that I have microchipped all of my own pets, and will continue to do so. However, a microchip is an electronic device, and any such device does have a chance of failing. It's a fact of the laws of entropy that any mechanical or electronic object has the potential of breaking down. You can manufacture an object to be very reliable and have a low chance of failure, but it's impossible to make something that will never stop working.
Personally I haven't seen or read any reports of wide-spread microchip failures. There has been nothing reported in any of the US veterinary journals that I receive, and I haven't personally seen this as a wide-spread issue. I can think of a couple of cases over my career where the client said they had a microchip and I couldn't find it with my universal scanner. In those cases I recommended going back to the shelter or vet who implanted the chip and have them scan it with one of their scanners. I can't say that I ever heard back from those clients to see if the chip was found.
In my own practice I'm a big advocate of microchips, and we use the 134 kHz chips. We routinely scan new pets, especially any that are brought in as strays. I can think of at least a half dozen stray pets over my career that I have personally found microchips in and been able to track down the owner. I can also say that when a client says their pet has a chip and I scan it, I'll find at least 99% of them even if we didn't implant them. So I believe that as a whole, microchips are very reliable. Are they perfect and immune to failure? No. That's why I tell my clients to still use collars and ID tags, and both of my dogs have this form of identification on them. But when a collar or tag is lost and the pet gets away, a microchip is still your best chance of having that pet found.
Cindi, I hope this answers the questions your group had on this topic. There were some comments regarding tattoos as a method of identification, and I'm going to talk about that tomorrow.