There's been much discussion recently, amongst finance journalists, about whether the employment data suggests that the economy has turned and perhaps the GDP figures are wrong. Obviously that's a possibility. But there are much less hopeful interpretations available.
Suppose we are in an economy in which there are old-fashioned technologies (like CDs and physical shops) along with new technologies (such as online downloads and internet retailing). And let's suppose that although the new technologies could expand, there were a limit to how fast they could expand, so that the new technologies couldn't service all the demand (e.g. some people would still have to buy their music as CDs and their retailing from physical shops, even if the online variants were better). Then suppose you were a businessman in charge of an old technology business, whilst the new technology were still small. (To make the point concrete, imagine it were 2006 in Britain.)
As such a businessman, you might properly reason as follows: Although the new technology will eventually take over, I still have time to engage in one last phase of investment in my old-techology business before it becomes obsolete, because there will still be demand for my old-techology products for a while, since the new technology can't expand fast enough.
So let's suppose you do invest – it's 2006 and you invest one last time in some old technology. But now let's suppose there is a bad recession and demand doesn't pick up quickly after it – as in 2008 onwards. What will happen then is that there won't be the same "excess" demand that the maximum feasible rate of expansion of new technologies can't serve. What seemed like a good investment in 2006, based on projections of technology change and demand from GDP growth, will now look like a very bad investment. Indeed, the capital invested will have become obsolete faster than previously expected.
I think the above process is one possible way to explain the puzzle of GDP contracting whilst employment is rising. In any economy, the amount of GDP depends upon how much labour there is (employment), how much capital there is ("capital" being what investment produces), and the shares in output of each. So if GDP is falling whilst output is rising, that could be because labour is getting a lower share of the total cake (e.g. because salaries are rising more slowly than prices). This is probably happening, but may not be the whole story. The other possibility is that the amount of capital is falling. That might happen for the sort of reason sketched above – capital is becoming obsolete more quickly than originally anticipated. That could be for the technology-related reasons above. Alternatively, investments could become obsolete for previously-unanticipated policy-driven reasons (e.g. if the UK were to leave the EU within the next few years, as now seems increasingly likely).
If the real output-generating value of the UK's capital stock is falling, it may be less surprising that employment can rise whilst we're in double dip recession. But it would also mean we shouldn't expect a rise in unemployment to be put off for long. Because as that obsolete capital – those misguided investments from around 2006 – finally becomes exhausted, it won't be replaced, and the staff whose work is supported by those obsolete investments will be laid off. In other words, it could well be that before the economy can recover, we need to shift both capital and employment away from capital created by now-obsolete investments. A big rise in unemployment could therefore be a sine qua non for recovery.