Thursday 13 March 2014

How Unbalanced is Infrastructure Spending?

[Posted by Prof Henry G. Overman]

Following BBC's Mind the Gap and my R4 Today discussion with Evan Davies and Ed Cox at IPPR North I've been thinking more about infrastructure spending and it's role in explaining urban economic performance in the UK.

As I discussed in a post last week IPPR North uses numbers for this that I think are misleading: "IPPR North's numbers which suggest we are set to spend £5,000 per person on infrastructure in London but only £250 per person in the North East. These figures are certainly striking, but they should be interpreted with considerable caution because they are far out of line with actual expenditure. Actual expenditure is reported in the Treasury's Public Expenditure Statistical Analysis Tables. According to those tables, in 2010-11 London received £800 per head (compared to an English average - including London - of around £400). But the second ranked region was the North West with £337 per head."

Following Ed's comment on that post I went back and re-read their 2013 update 'Still on the wrong track'. I remain to be convinced. Let me highlight three specific concerns (there are others) about the way in which these figures are misleading:

1) The £5,000 per head figure for London includes both private and public investment. If you look only at projects that involve the public sector as a funder the London figure is £2,500 per head. If you look only at projects where the public sector is the sole funder the figure is £770 per head. Of course, the figures for other regions drop too - the total for the North East is £250 per head, with some public sector or purely public sector it's £5 per head (presumably because there is no private sector investment in NE projects in the pipeline). I can see that the ratios don't shift much, but people who don't have time to think hard about these issues will focus on the headline number of either £2,500 or £5,000 per head for London. They will also tend to ignore any caveats about the role that private sector spending plays in driving these disparities. Which brings me to my second concern ...

2) Media coverage of these figures is highly misleading and this is misleading public debate. Here are three stories on the IPPR report - all courtesy of main stream media:

Transport Spending Skewed Towards London (BBC website)

The government spends more money on transport projects for Londoners than on those for the rest of the country combined, a think tank says. The Institute for Public Policy Research North says £2,700 is spent per person in London compared with £5 per head in the north-east of England.

Track changes: The North East is being Hobbled by Lousy Transport (The Economist)

Partly as a result, public expenditure on transport infrastructure projects is £2,595 ($4,340) per head in London but only £5 in the north-east, according to IPPR North, a think-tank.

Transport Spending skewed towards London (The FT)

The think-tank calculated that Londoners received £2,600 per head compared to £5 per head in the northeast. Of all planned capital investment, it said, 89 per cent went to London and the southeast.

Ed Cox, director of IPPR North, said the government was failing to tackle the issue of regional inequalities in infrastructure. “Skewed spending benefiting London and the southeast is nothing new but as we head towards new announcements at the spending review, these figures will strike most people as deeply unfair.”


Any one reading these reports (especially the first two) would conclude that these figures tell us how existing public expenditure on transport is hugely skewed towards London. This is simply not the case. The best figures we have for the regional distribution of existing public expenditure on transport come from the PESA tables that I quoted above. To repeat, these show "in 2010-11 London received £800 per head (compared to an English average - including London - of around £400). But the second ranked region was the North West with £337 per head.

Of course, it's hard for IPPR North to control how the media use their figures. But here is Ed Cox, tweeting just before our R4 debate: "Looking fwd to discussing why London is doing so well with & on tmrw - here's a clue " If you follow the link you'll see that the graph uses the £5,000 per head figure. Further, at least to my reading, Ed is suggesting that these figures explain why London is doing so well. I find that confusing. How can future projections of differences in public expenditure on transport infrastructure, that are out of line with historical public expenditure, explain the role public investment plays in explaining the current economic performance of London?

3) It could be, of course, that historical expenditure from the PESA tables isn't a good projection of the future. For example, the historical figures showed an increase from £600 to £800 per head for London since 2008-09; while expenditure in all other regions is flat at around £300 per head. Perhaps the next set of PESA tables will show a huge leap in London to, say, £1,500 per head while the North-East drops to £5 per head. I think this unlikely - the pipeline is biased towards expensive projects that take a long time to build. Simply dividing by population then gives a very distorted figure of annual expenditure. But even a more modest continuation of past trends would see London's per-capita public expenditure pulling away from the rest. Whether this matters depends in turn on the extent to which public expenditure on infrastructure explains London's current economic performance. Pointing to future spending doesn't answer that question.

In short, these numbers are hard to interpret and they are being interpreted in ways that mislead public debate. They do not provide an accurate picture of how public expenditure has been distributed across regions in the recent past.  They do not tell us how much this has contributed towards the gap in economic performance between London and the rest.  Finally, it's not even sure that they provide an accurate guide to the regional distribution in future.

Decisions about public expenditure on infrastructure clearly matter for urban economic performance. I think the IPPR report raises an interesting point about the role of both private and public investment (where PESA only tells part of the story). I also agree that better future projections of the regional distribution of private and public infrastructure spending (that are consistent with PESA) would allow for a more informed debate. I'm afraid, however, that I don't see these numbers as they stand helping much with the latter objective.