In the past I have had two main problems with failed ligation reactions.
The first appeared to involve a seasonal decrease in the quality of
MilliQ water (associated with filters which were not changed often
enough). This would'nt appear to have affected you, as your
phosphorylated vector was religated.
The second problem I have encountered, involves over CIPing your vector.
Cheaper preparations of phosphatase have traces of exonuclease activity,
these will nibble the ends of DNA. The more common problem is the carry
over of CIP into the ligation reaction. This effectively removes the
phosphates from your insert, thus no ligation and no transformants.
I routinely use no more than 1unit CIP/pmole DNA ends (routinely
0.2unit/pmoleDNA). I heat inactivate the CIP (65C/20mins) in the
presence of 5mM EDTA, phenol extract and gel purify the linearised
vector. This method maybe a bit over the top, but I have had no problems
cloning when my vector is purified in this way. Alternatively, one can
use shrimp phosphatase which can simply be heat inactivated.
Best of luck