Wednesday 12 October 2011

(Return of) The North-South Divide

The new unemployment figures certainly look grim. If you are clutching at straws, I guess that you could point to the fact that the rate of increase has slowed, but that provides cold comfort.

Centre for Cities has a great map highlighting the spatial differences. In my view, the North-South divide in terms of relative performance is pretty striking. With one or two exceptions, places seeing lowest increases are in the South, those seeing highest in the Midlands and the North.

This caused me to look back to my first post on the geographical impact of the recession, back in October 2008, which noted: "Although there is some academic debate on this, the south was probably unusually hard hit in the early 1990s recession as well. A different shock, may be, but it reminds us that we have been here before and that the south is very capable of bouncing back from recession."

By January 2009, the general pattern was already apparent: "Centre for Cities City Outlook received wide coverage yesterday of its finding that cities in the north are seeing the highest increases in JSA claimants [NB: they also point out that the north-south divide language I am using here hides some important details - sorry for that]" but some of my earlier optimism had vanished: "One final thing, the comment in my October post about "two quarters" now looks optimistic, but doesn't change my conclusion on the geographic impact. In short, it's grim everywhere."

The spending review in October 2010, clearly looked set to reinforce the overall spatial pattern: "Abstract from issues about the timing and speed of cuts, it is as interesting to think about the long run impact on the economic geography of the UK. A growing public sector has clearly propped up total employment in some areas that are very bad at generating private sector employment. The direct effects of public sector job cuts simply must be bad for those areas. But the tricky thing to predict is how the economy will adjust to this initial effect. For the private sector in these areas there are two offsetting indirect effects. The first is that public sector jobs create demand for local goods and services - so cutting them will be bad for the private sector. High public sector salaries also create distortions in local labour markets - particularly in competing for the best workers - so this cutting jobs will be good for the private sector. The coalition is banking on the second effect being larger than the first. Labour used to think that the latter effect dominated in the South East, but that the former dominated in the rest of the UK. In reality we simply don't know the magnitude of these two offsetting effects."

Finally, by January 2011, the spatial pattern was pretty well established, particularly for greater London: "First, in absolute terms and relative to expectations, London appears to have successfully gotten away with it." Although some commentators were still disagreeing with that overall assessment.

When I delivered a public lecture on how London 'got away with it' in January 2011, my broad explanation ran as follows: "The over-representation of the professional occupations partly explains this. The bailout may explain why these occupations did even better in London. But other explanations are possible. The shift in financial sector employment may be driven by the increased importance of timely information flows when things turn bad. Despite improvements in ICT, economists still think that spatial proximity play an important role in exchanging information. Alternatively, London may be benefiting from the depth and breadth of its labour market. Finally, Spatial Economics Research Centre research suggests that the most talented are highly concentrated in London." I suspect that these sort of reasons for the spatial differences remain valid.

In short, I don't find the spatial pattern of unemployment growth particularly surprising (there are a couple of idiosyncracies in terms of the performance of particular places - e.g Blackpool and Preston). Of course 'predicting' the general spatial pattern conditional on the overall performance is much easier than predicting that overall performance (as the government is finding it).