There could be many reasons. I can provide a case which shows that incompetence and low work ethic of an examiner and USPTO offices could be among the reasons.
My application numbered 09682315 was filed in 2001 and processed for 6 years and rejected. It was about a software development system. If you know a little bit about software development then you will laugh at the rejection reasons.
The application has a claim of “software development without a compiler”. The examiner got a Java tutorial book (JAVA Database Programming with JDBC 2nd Edition, Pratik Patel & Karl Moss), hastily went through several pages (pages 265 – 26 to search for the word “compile” and not found one. Thus the examiner got his conclusion that this tutorial author had made a software development system which did not involve “compiling” and thus my system was not new and hence not patentable (office action page 4: “Patel makes no reference to compiling”; office action page 5: “again that no reference is made to compiling”). I said the examiner “hastily went through several pages” because if we went through these same pages a little more carefully, we could find the word “compiling”. The author Patel actually did say “compile Java code”: “The BeanBox will generate and compile an adapter class – code that will be executed when the actionPerformed event occurs.”. The examiner should have searched the text more carefully in his reference material. But a more sad fact was his using of word-matching to examine a software patent application. Even if the word “compile” could not be found in the whole book, anyone knowing a little bit about Java knows that Java code must be compiled for it to run.
The examiner uses the word-matching technique quite extensively in various ways. For example, on page 5 of his office action, he picked the following words out of my claim context “the input devices controllable by a user to create visual representations on the display screen,” and said that other people already used input device and already provided visual representation. It is as if saying “ already exists” when denying a claim of “a car with 4 wheels that runs on road and flies in the sky”. If you find such out-of-context rejection amusing then unfortunately such rejections formed 80% of all rejections reasons.
There is also much room for the examiner to improve his work ethic. My application was touched basically once a year by the examiner to issue an office action for a few issues and then another year for some other unrelated issues, and so on. One office action was about drawings not meeting the USPTO requirements. I sent the same drawings again and told him that those drawings were the same and were made by a professional drawing company, and the company guaranteed 100% USPTO compliance. I told the examiner that if he insisted the drawings were not acceptable then I’ll sue the drawing company. Guess what? The examiner accepted the drawings.
USPTO offices were also quite sloppy. Out of my 4 office action responses, 3 of them were discarded for no reasons. I called after long time inactivity and the examiner told me that I had passed the response deadline because they did not receive my response. I told him that I had post office receipt to prove that USPTO received my mail. The examiner said that it was not his fault and asked me to send it again. It did not happen just once, and not twice, but 3 times. I sent a letter to USPTO complaining it. I got a response from some higher officer telling me that this was normal due to the large number of correspondences USPTO had to process every day. 75% of my office responses were discarded and it was normal by USPTO standard!

http://ipspotlight.com/2008/01/21/us...

Be well