Urban versus rural living

In my post yesterday I talked about the reasons why simplistic comparisons of rural and urban living lead to concepts of a 'rural penalty' that are misleading. Here's a slightly more meaningful thought experiment to get at the differences between rural and urban living. Imagine yourself moving across different areas in Britain. In each area you'll live in roughly the same type of house (number of bedrooms etc) and you'll get paid the wages consistent with your experience, education etc. If you are going to stay living in areas where house prices are high relative to income, you better be getting compensated by some nice local amenities (e.g. pretty countryside, lower crime, etc). If you are going to stay living in areas which aren't very nice (e.g. polluted, run down, etc) you better be compensated by house prices that are low relative to incomes. In other words, the gap between house prices and incomes provides important information on the amenities offered by different locations.

Here's what that gap looks like across different areas of Britain (taken from a 2011 SERC DP). I've coloured the rural areas in green and the urban areas in red:

What you can see from this is that it makes no sense to talk about 'a rural penalty'. There are a group of rural places that are high housing cost, low pay - but these are places we would associate with good amenities (e.g. West Cornwall). There are a group of urban and rural places (in the middle of the picture, below the 45 degree line) that offer higher incomes given the cost of housing. These places generally offer lower amenities (e.g. they are very remote or are old industrial towns). Finally, to the right of the picture there another group of (mostly urban) places where incomes are high, but housing costs are even higher. These places must offer some kind of amenities (e.g. good restaurants, easy access to good jobs and schools) to offset these low real wages. If they didn't, why would so many people live there?

The dashed line summarises the trade-off that we face in Britain as we move from low to high wage places. We can have high amenities, expensive houses and low income. We can have poorer amenities, cheaper houses and higher incomes. Or we can have high amenities, the highest incomes but the most expensive houses. Take your pick.

Country Dwellers and the 'Rural Penalty'

Lots of coverage today for the Commons Rural Affairs Committee report suggesting that country dwellers pay a rural penalty in access to services.

I accept that this is true for some services (e.g. broadband). However, the same report also tells us that many rural areas have very high house prices relative to income. This tells us that the countryside must provide a lot of amenities that make people willing to accept low real wages (as measured by the difference between house prices and incomes). In short, the data tell us that overall the countryside is a pretty nice place to live. I'm not sure in what sense this constitutes a rural penalty.

Of course, this is not to deny the fact that the countryside may not be such a nice place to live if you are poor. But then that's also true of most of our cities. What this teaches us is that, once again, simple comparisons between urban and rural areas are misleading and fairly pointless. We need to focus on individuals and families and understand the implications for different groups of different levels of public service provision.

[Related posts: Rural Broadband; Rural Cost of Living; Rural Housing]



 

Economic Impact of the Olympics

The government has announced a £9.9bn boost to the economy as a result of the Olympic Games.

The 'detailed' UKTI report on which these numbers are based suggests this breaks down to include £5.9bn of additional overseas sales and a £2.5bn boost in inward investment. Unfortunately, it provides no information on how these figures are calculated. It is certainly not possible to figure out whether they are in any sense additional (i.e. would not have happened if the games had not taken place).

One thing we can do, to start to figure this out is to look what happened to the overall balance of UK trade in that period:

To my eyes, there is little evidence of a trend change around the time of the Olympics (individual components don't look much different as you can see from the original report).

In short, you should take these numbers with a pinch of salt. If anyone knows of more rigorous analysis I'd be happy to receive a pointer.

Mandelson and HS2

Three things I found particularly interesting about Peter Mandelson's conversion on HS2

1. He's changed his mind.
2. His description about how and why Labour decided to support the project (costs far in the future and we get to announce a big scheme) - very depressing for those of us who favour evidence based policy making
3. The fact that he highlights the opportunity cost of the spending and the uncertainty over costs and benefits, including the impact on the North-South divide. Precisely (I won't repeat the arguments here - you can follow the links)

One thing on opportunity costs: Every time I ask someone from 'the north' whether they would rather have, say, £5bn spent in each of 8 cities outside of London or one HS2 line they (nearly) always prefer the former to the latter. Unfortunately, they don't think the money would be invested if HS2 scheme abolished so they continue to support the scheme even though they don't believe it represents good value for money. Again, very depressing.

Anyhow, let's hope that Mandelson's intervention the thin end of a very big wedge. We need a much more sensible debate on whether HS2 is the right way to spend money.

On the road to recovery?

Posted by Dr Rosa Sanchis-Guarner, SERC

This morning the Government unveiled "the biggest road-building programme in 40 years", as part of a package of infrastructure schemes intended to drive the UK’s long-term economic development. Road-building is often opposed on environmental grounds, and those costs are clear. But does it produce any economic gains? Surprisingly, we have very little robust evidence - until now. 

SERC’s research suggests that roads can produce economic benefits – at least locally. In the first rigorous studies to look at the effects of UK road transport improvements on employment, output and wages, we find that road-related accessibility improvements  between 1998 and 2007 increased local employment and raised wages. Back-of-the-envelope calculations suggest that £1bn of road improvements could create around 2000 jobs in the affected areas.

Road networks dominate transport infrastructure in most countries, including the UK. According to official transport statistics, in 2010 91% of passenger transport and around 68% of freight transport was by road. Road traffic has increased steadily since the 50s, up to around 240 billion vehicle miles in 2011. And most of this traffic is concentrated in the major roads network. So it’s not surprising that a substantial amount of UK public spending is devoted to roads: around £1 billion, or 44% of total transport spending in 2010/11. An important slice of this expenditure is for new road links (since 2000, over 300 kilometres of new roads in England alone).

In theory, transport improvements decrease transportation costs, improve access to markets, foster economic integration, stimulate competition, generate agglomeration economies and a number of other ‘wider’ economic benefits. This is why transport improvements are frequently proposed as a strategy for economic growth, integration and local economic development. But for economies with well-developed transport networks like the UK, there is little good evidence on the extent of the gains that result from additions to the existing network. Although road improvements are routinely subject to appraisal - predicting the economic benefits before the roads are built - they have not historically been subject to any evaluation in order to work out whether these benefits actually materialised!

SERC’s research provides some of the first hard evidence. We link data on 31 major new road construction schemes between 1998 and 2007 to administrative data on businesses and workers. There are major challenges when evaluating the economic effects of road construction. 

The first one is how to capture the effect of new roads. We do this using an index of employment ‘accessibility’ that estimates the number of workers that can reach a location, per unit of travel time, using optimal routes along the major road network. When new links are added to the network, optimal travel times decrease and employment accessibility increases, but by different amounts according to where a place is in relation to the existing road network, the new road links and major centres of employment.

The second major challenge comes from the fact that new roads are not placed randomly, but are targeted specifically to improve traffic flows, decrease congestion or improve connectivity. For this reason, to measure the effects of increases in accessibility we cannot simply compare places that had new links to places that did not. Instead, we look only at places which that are close to a new road scheme, estimating the effects from the subtle local changes in accessibility that occur within a 20km radius. 

*

Linking these accessibility changes to firm level data on employment and output allows us to estimate how transport affected local production and employment. Firstly we find that places (electoral wards) that experienced accessibility improvements saw employment gains. A 10% increase in accessibility leads to a 3-4% increase in plants and employment. 

These gains come about through an increase in the number of businesses, particularly in the service sector, though surprisingly firms already operating in the area shed workers. An explanation for existing firms cutting employment is the observed wage increases in response to the accessibility changes (evident in other results based on worker data), coupled with improvements in productivity (output per worker). These wage and output effects are of a similar to the effects on employment: a 10% increase in accessibility results on an increase in labour productivity and average wages of around 2.5-3%.

*

So will road-building set the UK on the ‘road to recovery’? Overall, the economic benefits are relatively small. We also need to be careful in interpreting these changes as gains to the national economy. To some extent, jobs may be displaced from other areas, although our evidence indicates this is not the primary channel. Also, we are learning about the potential impacts of transport improvements from very localised changes, and abstracting from changes induced by local road schemes in more distant places. Given how much political attention infrastructure projects attract, though, our work opens the way for more research in an area where proper evaluation is badly needed.

Help to Buy

Posted by Dr Christian Hilber, SERC and LSE
 
I recently participated in a public debate organized by the Strategic Society Centre. The topic of the debate was the government’s Help to Buy scheme. The central question: Who will benefit from the scheme? My short answer: It is almost certainly not the young first-time buyers who are the supposed main beneficiaries of the scheme.

How can this be? Aren’t the equity loan scheme and the mortgage guarantee scheme directly targeted towards first-time buyers? Yes, true. And does the equity loan scheme not only apply to new build homes? Again, yes, true. It is true that Help to Buy will likely increase mortgage availability, especially for those attempting to buy with small deposits.

However, the effect of this is an increase in demand for housing – in economics language: it shifts the individual and aggregate demand curves outwards to the right. It does not shift the supply curve. But as long as supply is ‘responsive’ (i.e., the supply curve is not very steep), this should still have the desired effect of more new housing and perhaps only slightly higher prices.

So what can we say about the supply side? In the very short-run, housing supply does of course not respond at all to demand shocks because of planning and construction lags, so prices can be expected to rise in the short-run. The trouble is that in the UK, even in the very long-run, supply is incredibly unresponsive to demand shocks (i.e., the long-run supply curve is very steep). Why? The main culprit is the UK planning system, which is, in an international context, extraordinarily inflexible. Since 1947 there are virtually no fiscal incentives at the local level to permit development. Local planning authorities face most of the costs of development but have very few benefits from it. Moreover, local homeowners – in the UK perhaps best described as BANANAs (Build Absolutely Nothing Anywhere Near Anything) – oppose (virtually) any development in their backyards.

As a consequence of all this, Help to Buy will likely have the effect of pushing up house prices (and rents) further with very little positive effect on new construction. Housing will likely become less – not more – affordable for young would-be-owners! The beneficiaries of the scheme are the existing (typically wealthier) homeowners who benefit from the capital gain. However, not all homeowners benefit equally. Young expanding families may not benefit at all since their new larger house will also become more expensive. The true beneficiaries are those elderly homeowners who are thinking about downsizing (not many) and those wealthier retired homeowners who decide to pocket the capital gains from Help to Buy and move to Spain or Southern France, where housing is much more affordable and spacious. First-time buyers will arguably be no better off because the easier access to credit is offset by higher prices. And mobile young households who do not want to or still cannot own will be worse off. They will face higher rents and, as tax payers, bear the burden of the schemes.

The empirical evidence strongly supports these arguments. In recent work (here and here) Wouter Vermeulen and I provide strong evidence that local planning constraints – in conjunction with strong local demand – are indeed the main culprit of the extraordinarily high house prices and the cramped spaces in the Greater London Area and the South East of England. (The UK has the second highest house prices per square metre in the world – only topped by tiny Monaco – and new build houses in the UK are 38% smaller than in densely populated Germany and 40% smaller than in the even more densely populated Netherlands.)

In another paper, forthcoming at the Review of Economics and Statistics, Tracy Turner and I find for the US, that in tightly regulated cities mortgage rate subsidies have the effect of substantially raising house prices and reducing homeownership attainment – the opposite of policy makers’ intentions. The aggregate effect on homeownership attainment for the US as a whole is ‘zero’. Notably, this is achieved at an annual cost exceeding $100 billion.

Proponents of Help to Buy have argued that the equity loan scheme only applies to new build homes, so this should only increase the demand for new build homes and thus lead to more construction. But this is a bogus argument: existing homes and new build homes (and in fact rental homes) are reasonably close substitutes, so Help to Buy will affect aggregate demand for housing and not just new build. There have been, at long last, clear signs of a recovery in the construction sector – in fact, for a while now; talk to architects! But this has nothing to do with Help to Buy. It is merely a (welcome) coincidence (for proponents of the scheme).

I argued above that the main beneficiaries are mainly wealthy and older homeowners. But especially the wealthy arguably also pay much of the taxes. So one might think that Help to Buy does ‘no good’ but it also does ‘no harm’ – it is a ‘zero sum game’. Unfortunately, this couldn’t be further from the truth for at least two important reasons. Firstly, the taxes needed to finance the schemes have a ‘dead weight loss’ – they are a pure welfare loss to the society. Secondly, there are ‘systemic risks’. In contrast to, for example, the American and Canadian systems, there is no cushion in the form of private mortgage insurance companies that bear the risks associated with a house price bust.

In the US for example, even before the crisis, Fannie Mae and Freddie Mac only guaranteed and securitised mortgages up to a 80% loan-to-value (LTV) ratio; the portion of the loan that exceeded 80% of the house value had to be ‘enhanced’ by private mortgage insurance companies. These insurance companies either assumed those risks or – more likely – reinsured them, the reason why AIG got into such trouble. Should the British housing market(s) take a significant hit (perhaps not very likely, but who would have predicted the ‘Great Recession’?), then the situation for UK taxpayers would be much worse than that of US taxpayers.

This is because the government doesn't just provide an implicit warranty as a last resort (as was the case in the US) but directly takes the first hit. Private mortgage insurance companies won't provide any cushion, unlike with Fannie and Freddie (who only got into trouble because house prices in many metro areas decreased by more than 20%). The UK government – or rather, the British tax payer – will face huge costs, even if loans go only slightly ‘under water’. The fact that mortgages in the UK are ‘recourse’ helps in the sense that there would be fewer defaults (compared to the US mess) but would also arguably mean prolonging the crisis.

A last point: what is the exit strategy? If the scheme was indeed withdrawn abruptly after three years, then might this not cause a house price bust and yet quite possibly another recession. If this is ‘too big a political risk’, is it really plausible to think that the scheme will end in three years? My concern really is that we end up in a vicious circle with a policy that arguably does not only do more harm than good and is very costly but is also very hard to get rid of. If we are lucky, the two schemes will, for whatever reason, never really take off and the schemes will be silently discarded after a few years. My sincere hope is that the government does more to tackle the real problem on the supply side: the fact that there are too few incentives at the local level to permit residential (or commercial) development. I floated the idea of a genuine local property tax that would replace the current council tax and stamp duty (another harmful tax – see here and here) …it would be a (good) start…

A version of this post first appeared on the LSE Politics and Policy blog.

Time to Build

I'm still grappling with figures around land hoarding and completion times. Prompted by a colleague (who tells me that yesterday's numbers were 'hard to follow') I took a look at the US figures on completion times to provide a comparison to figures for England and Wales.

For England and Wales, average completion time is 25 months which breaks down as:

• 10 months from obtaining permission to starting on the site
• 15 months from starting to completion

For the US as a whole, average completion time is about 9.5 months which breaks down as:

• 1 months to start
• 8.5 months start to completion

So, we look quite slow. But we know that the US market as whole is very lightly regulated compared to the UK. What happens if, instead, we compare with the figures for the North East (which is the more heavily regulated market in the US - and also more metropolitan). For the North East,  average completion time is about 12 months which breaks down as:

• 1 month to start
• 11 months start to completion

12 months is a lot lower than 25 months so does this suggest developers here are slow? The problem is that the US numbers are presented by housing unit broken down by the size of the unit whereas the England and Wales numbers are broken down by scheme and size of scheme. Our average time to completion would look considerably lower if we looked at the completion rate for individual houses within schemes (rather than dating completion when the last unit in the scheme is built). It's also important to note that the time to completion various a lot depending on whether you are building single unit or multi-unit buildings (with the latter taking much longer in the US). If most England and Wales permitting is for flats, our completion rates would compare reasonably favourably with the North East.

In short, the international comparison doesn't seem to get us much further other than to show that i) we are slow to start and that ii) in the US at least, as regulations increase, it takes developers longer to complete. 

Either way, simply pointing to the headline figures for England and Wales of 400,000 permissions 'unimplemented' and average completion times of 25 months doesn't tell us much about the source of any problem and what we might do about it.

[NB: If anyone knows of something more systematic on this, I'd be happy for pointers.]