In 2012 and 2013 there were also effects from the fee increase: in 2012 student numbers were down for many subjects in part because of the poor results but also because students were deterred by the new fees, and many universities were desperately looking for AAB+ students in clearing.
In 2013 student numbers were back up, in part because of students who had deferred entry and in part because the grades were more in line with expectations. However, a number of universities were still looking to take extra ABB+ students in clearing to push their overall numbers up. In the university where I work engineering numbers were up around 20% in 2013, without dropping below AAA in clearing, and mathematics numbers were up from 170 in 2012 to 250 in 2013, again without dropping below AAA equivalent in clearing. Places were mostly filled before clearing but we accepted extra students in clearing. These students were mostly trading up using "adjustment" rather than entering clearing because they had missed offers.
Several social sciences subjects in the same university whose regular offers are AAA/AAB dropped to ABB in clearing to get extra students to make up for their numbers shortfall in 2012. They were mostly taking students who had missed grades rather than getting students who were trading up. (Social sciences has a lower reputation than maths and engineering at this university.)
As creamteas writes, neither 2012 nor 2013 should be viewed as good predictors for 2014. We have some sense of whether we will be entering clearing and how many places we might offer but we cannot know for sure until we see the A level results. And it's not just results in our subjects which matter - we (maths and engineering) would never need to enter clearing, we will have good student numbers regardless of results, but we will be forced by the university management team to enter clearing and take a few extra AAA+ students if other subjects cannot fill their courses with high level students.