Posted by Dr Christian Hilber, SERC and LSE
Joe Gyourko, (from the Wharton School) gave a great SERC seminar on Friday. The NBER paper - An Anatomy of the Beginning of the Housing Boom: US Neighbourhoods and Metropolitan Areas - attempts to figure out when the boom began in different US cities and neighbourhoods and whether any fundamentals were moving in the ‘right direction’ to help explain timing and magnitudes.
The research provides pretty convincing evidence -- using some amazing data -- that housing markets are 'local' (i.e., metro area) phenomena and are to a large extent driven by income as the main fundamental on the demand side and by regulatory and physical/geographical constraints on the supply side. Supply constrained cities boomed earlier and the research explains pretty convincingly that strong income growth occurred at the same time and offers a plausible explanation of house price booms in those cities. This is essentially ‘Economics 101’. Even more interesting, however, are the places that experienced substantial house price booms that were not explained by high income growth and tight supply; those in places such as Las Vegas or Phoenix. In these places Joe finds no evidence of the boom coinciding with a positive income shock nor is supply very tight. One further thing we know from Joe’s paper: The housing booms (or bubbles) in Las Vegas and Phoenix emerged much later and both the boom phase and the bust phase were very steep. So how can these phenomena be explained?
As Joe was careful to emphasize; the paper he presented does not provide an answer and at this point we can only speculate (although he is promising much more research to come). One plausible explanation to me is the following: Irrational exuberance needs some sort of ‘convincing story’ to emerge. Strong house price growth in places such as SF, NYC, Boston, LA etc. – that were supply constrained and at had good fundamentals – coupled with the extremely low interest rates created the “story” that the house price boom was being driven by historically low interest rates (or other changes in the economy). People started believing in this story. I would imagine that there may have been some spatial contagion effect. This is at least what the ‘spatial history’ of the US boom suggests – and is consistent with some maps on timing that Joe showed during his seminar. Essentially the belief in ever growing house prices may have spread from places such as SF and LA to places such as LV or Phoenix, even though the fundamentals in those places - apart from the interest rates - were extremely different. Now, in the places where it is easy to build (e.g., in Las Vegas or Phoenix) developers started to build like crazy. In places such as SF or NYC perhaps irrational exuberance started to contribute to the price increases originally driven by fundamentals. But how can prices increase at all in LV or Phoenix if supply is elastic? Perhaps the explanation is that supply is only very elastic in the long-run but not in the very short run; this is due to various planning, development and construction lags. So in the very short run if price growth expectations are (too) high this may encourage developers to add a lot of new housing stock. This pushes up house prices in the short-run but once supply adjusts in the medium and long-term prices come back to the pre-boom levels or may even fall below those. Because overbuilding can easily happen in LV or Phoenix but not in Los Angeles or SF, this can explain why the house price levels in SF and LA are still significantly higher than before the boom started, whereas this is not the case in LV and Phoenix. The downward adjustment was much steeper in LV and Phoenix.
Again, Joe’s research doesn’t yet prove any of this – so this is only informed speculation on my behalf. But SERC research shows that elements of this story – e.g. that planning constraints and physical/geographical constraints in conjunction with strong income growth explain the strong increase in prices – are certainly consistent with UK data. There are likely other explanations and the above may be too simplistic to explain everything. After all, on our Real Estate masters we spend several lectures on the fundamentals of housing markets and house price dynamics and there are numerous factors that contribute. Still the above may be a rough explanation of the fascinating picture Joe painted about the timing of the boom across different US metro areas.
CEP Urban and Spatial Programme Blog (formerly SERC)
Monday, 7 November 2011
Thursday, 3 November 2011
Falling house prices and the planning system
Posted by Philippe Bracke, LSE and SERC
In periods of economic distress it is natural to think in terms of cycles. Reaching the bottom of the cycle is a painful experience but at least gives some hope: from that point onward, things can only get better. UK house prices, however, look still far from their bottom.
In periods of economic distress it is natural to think in terms of cycles. Reaching the bottom of the cycle is a painful experience but at least gives some hope: from that point onward, things can only get better. UK house prices, however, look still far from their bottom.
In a recent research project for the IMF, I have studied the house price expansions and contractions of 19 OECD countries since the first quarter of 1970, and identified 55 expansions (of which 6 are ongoing) and 62 contractions (of which 13 are ongoing). On average, expansions last 6 years and produce a 60% house price increase in real terms; contractions last 4 and a half years and produce a 30% real price decline.
These numbers give the impression that real house prices are increasing in the long run, because expansions are longer than contractions and entail bigger absolute price changes. If one removes the most recent house price boom from the sample, however, this impression largely disappears. Other studies, which have a more restricted geographical focus but a broader temporal window, show that over the centuries real house prices are fundamentally flat. (Piet Eicholtz, for instance, has studied three centuries of house price data for the Herengracht canal in Amsterdam and has found their average real annual growth has been fairly low, at about 0.5%).
If there is such a thing as a pattern for house prices in the long term, it is not that they are increasing; it is that they are cyclical. All the countries I examine in my research have gone through multiple expansions and contractions of national house prices. These fluctuations are not just due to randomness – in the paper I show that the probability of ending a house price expansion increases with its duration. In other words, longer expansions are more likely to terminate: what goes up has to come down.
What does this mean for the UK? The figure below shows the real house price index and the corresponding peaks and troughs for three countries of my sample: Germany, the UK, and the US.
A couple of things stand out from the chart. First, the last house price boom did not involve all countries. There are a few cases of advanced nations, like Germany (or Japan), which did not experience any substantial price increase. For Germany in particular, over the last 40 years real house prices have stayed constant or declined slightly. Once again this is proof that real house price growth should not be taken for granted, even in productive and well-functioning economies.
Second, if history is any guide, countries like the UK and the US will continue experiencing real house price declines for some time. This adjustment process is already under way. According to the Land Registry, nominal house prices in England and Wales are down 2.6% on a year-on-year basis. Taking into account an inflation rate of 5%, real house prices have fallen by almost 8% in the last 12 months. However, there is surely potential for more substantial drops, especially considering that US house prices (which have grown less than in the UK during the boom) have already fallen by more than 30% from their peak.
Third, and most importantly, while some degree of up-and-down in house prices is unavoidable, the range of these oscillations should be carefully monitored. Ups and downs are an intrinsic feature of all economic series but booms and busts are not, and this is where the UK compares unfavourably with other countries. Even by US standards, UK house prices look like a rollercoaster: they more than doubled in real terms since the mid nineties; before that, they fell by almost 40% from 1989 to 1995.
For prices to vary so much, quantities must be very sticky. Indeed, a recent OECD working paper shows that the number of new housing units built in the UK is low compared to other nations. A report by the Department of Communities and Local Government suggests that strict planning regulations hold back housing supply and make prices more volatile. Let’s hope therefore that the current debate on planning reform will provide solutions that go in the right direction.
It might seem strange to advocate more house building in a period where house prices are falling. However, the current decline in house prices represents a cyclical adjustment that is not due to abundance of housing units. If this were the case, we wouldn’t see the current rent increases.
This post first appeared on the LSE's British Politics and Policy blog on 1 November. Follow them on twitter @LSEpoliticsblog
Wednesday, 2 November 2011
London's (shocking?) growth performance
Writing about the RGF earlier in the week, I noted "if growth is the absolute priority then you begin to wonder whether the government might be better off dropping the 'R' from the Regional Growth Fund. The economics of that are difficult. On the minus side it might be more difficult to find projects in the 'south' where employment generation is genuinely additional. Offsetting this is the fact that a Growth Fund would expect to be generating those jobs at relatively more productive firms. Of course, while the economics might be difficult, the politics of such a change are far trickier."
I was somewhat surprised, therefore to see Ed Balls making a similar argument in the Evening Standard: "the Government must also act with extra care to safeguard the London economy. For a start, that means making sure London is not excluded from action to support jobs." (He was writing about the National Insurance Holiday, but the same logic could apply to the RGF)
According to Mr Balls, London needs help because "a report last week found London is no longer the fastest growing part of the country and in the past year it has seen the biggest rise in unemployment of any region." Somehow, this manages to make things sound considerably worse in London than is the case. For some period now, London's relative performance (both compared to other regions and to predictions) has been pretty good. I don't think the position has changed that much. Assuming that Mr Balls was referring to the latest BRES numbers they show London as the second fastest growing region after the South East. Anyhow, according to my colleague Ian Gordon, "June-June annual comparisons show London as having the fastest growth rate of any region in just 4 of the last 15 years. More than any other, but hardly a shock when it’s not in the top spot."
What about the unemployment numbers? Again, these are not that surprising because London has large numbers of people (young, lower skilled) who are doing very badly in this recession. In the aggregate, better outcomes for 'higher skilled' workers tend to outweigh the poor performance of 'lower skilled'. There is nothing much new here - those kind of polarised outcomes have characterised London for a long time.
So the relative economic performance of London doesn't actually provide that strong a case for further intervention. Instead, the arguments depend on the extent to which policy in London is actually able to generate additional jobs and what are the ultimate objectives of policy.
I was somewhat surprised, therefore to see Ed Balls making a similar argument in the Evening Standard: "the Government must also act with extra care to safeguard the London economy. For a start, that means making sure London is not excluded from action to support jobs." (He was writing about the National Insurance Holiday, but the same logic could apply to the RGF)
According to Mr Balls, London needs help because "a report last week found London is no longer the fastest growing part of the country and in the past year it has seen the biggest rise in unemployment of any region." Somehow, this manages to make things sound considerably worse in London than is the case. For some period now, London's relative performance (both compared to other regions and to predictions) has been pretty good. I don't think the position has changed that much. Assuming that Mr Balls was referring to the latest BRES numbers they show London as the second fastest growing region after the South East. Anyhow, according to my colleague Ian Gordon, "June-June annual comparisons show London as having the fastest growth rate of any region in just 4 of the last 15 years. More than any other, but hardly a shock when it’s not in the top spot."
What about the unemployment numbers? Again, these are not that surprising because London has large numbers of people (young, lower skilled) who are doing very badly in this recession. In the aggregate, better outcomes for 'higher skilled' workers tend to outweigh the poor performance of 'lower skilled'. There is nothing much new here - those kind of polarised outcomes have characterised London for a long time.
So the relative economic performance of London doesn't actually provide that strong a case for further intervention. Instead, the arguments depend on the extent to which policy in London is actually able to generate additional jobs and what are the ultimate objectives of policy.
Monday, 31 October 2011
Regional Growth Fund (Round II)
The government's has announced the next round of projects receiving £950m from the Regional Growth Fund. We are told that this will directly create 37,000 jobs with a further 164,000 created indirectly ('in the supply chain').
As with round 1, with the details provided (severely curtailed by confidentiality requirements) it is impossible to provide any analysis of whether it will achieve this on the basis of the list of schemes agreed. Writing in 2005, SERC affiliate Colin Wren reviewed the available evidence on the impact of Regional Selective Assistance (a competitive scheme for allocating money to firms in depressed areas). The estimated cost per job ranged from £8,000-£21,000 (in 1995 prices). If the RGF of £950 million delivers 201,000 additional jobs that suggests a cost per job ‘created’ by the government of just over £4,700 (the same calculation for round 1 suggested 3,500 per job). In short, if these numbers played out, this would be a pretty effective intervention relative to existing schemes.
There are a number of reasons to think that these figures may be optimistic. First, with incomplete monitoring it is highly likely some of the 'leveraged' private sector funds ('£5 for every £1 of public money') would have been spent anyhow. To the extent that monitoring is imperfect, the RGF will only create additional jobs if it is being given to organisations that are credit constrained. Research that I have done with colleagues at the CEP suggests that this may only be true for smaller firms. We suspect this is because larger firms are better able to game the system (so monitoring is not as good) and are less likely to be genuinely credit constrained.
All of this suggests monitoring will be important for delivering additionality. Here, if I understood Nick Clegg correctly, the RGF is doing something different from the RSA. Specifically, when defending the amounts of money distributed so far he suggested that organisations that know they have the money coming have started activities. With RSA, my understanding was that usually firms need to receive the money first to demonstrate that public money is crucial to the project going ahead. This might suggest that additionality will be less for the RGF.
A separate issue is whether RGF will be more efficient than the Regional Development Agencies. Of course, it is impossible to tell at this stage. The RGF uses a different (competitive) mechanism for deciding on projects. This may lead to better decision making (or it may not). I would expect RGF to be more efficient per pound spent simply because it is spending less money. Civil servants may not be able to perfectly rank projects, but I don't believe that their selection is completely random, so the fact that the fund is smaller means it should achieve higher returns.
A final note of caution on the employment numbers - if all of government truly believed these numbers you might expect to see a lot more spending on RGF (unless they think that the smaller size of the scheme drives the high returns - as discussed above).
What about growth? Here I think there are further reasons to be cautious. In our work on RSA, we were able to find a causal effect of government money in increasing employment and investment, but not productivity. In addition, assisted firms are on average less productive, so RSA expands employment in less productive firms. This is still a 'growth' effect to the extent that these workers would have been unemployed (and we find some evidence, for RSA, that this might have been the case). But increasing the employment share of less productive firms may not be a good long run strategy for driving growth.
Indeed, if growth is the absolute priority then you begin to wonder whether the government might be better off dropping the 'R' from the Regional Growth Fund. The economics of that are difficult. On the minus side it might be more difficult to find projects in the 'south' where employment generation is genuinely additional. Offsetting this is the fact that a Growth Fund would expect to be generating those jobs at relatively more productive firms. Of course, while the economics might be difficult, the politics of such a change are far trickier.
As with round 1, with the details provided (severely curtailed by confidentiality requirements) it is impossible to provide any analysis of whether it will achieve this on the basis of the list of schemes agreed. Writing in 2005, SERC affiliate Colin Wren reviewed the available evidence on the impact of Regional Selective Assistance (a competitive scheme for allocating money to firms in depressed areas). The estimated cost per job ranged from £8,000-£21,000 (in 1995 prices). If the RGF of £950 million delivers 201,000 additional jobs that suggests a cost per job ‘created’ by the government of just over £4,700 (the same calculation for round 1 suggested 3,500 per job). In short, if these numbers played out, this would be a pretty effective intervention relative to existing schemes.
There are a number of reasons to think that these figures may be optimistic. First, with incomplete monitoring it is highly likely some of the 'leveraged' private sector funds ('£5 for every £1 of public money') would have been spent anyhow. To the extent that monitoring is imperfect, the RGF will only create additional jobs if it is being given to organisations that are credit constrained. Research that I have done with colleagues at the CEP suggests that this may only be true for smaller firms. We suspect this is because larger firms are better able to game the system (so monitoring is not as good) and are less likely to be genuinely credit constrained.
All of this suggests monitoring will be important for delivering additionality. Here, if I understood Nick Clegg correctly, the RGF is doing something different from the RSA. Specifically, when defending the amounts of money distributed so far he suggested that organisations that know they have the money coming have started activities. With RSA, my understanding was that usually firms need to receive the money first to demonstrate that public money is crucial to the project going ahead. This might suggest that additionality will be less for the RGF.
A separate issue is whether RGF will be more efficient than the Regional Development Agencies. Of course, it is impossible to tell at this stage. The RGF uses a different (competitive) mechanism for deciding on projects. This may lead to better decision making (or it may not). I would expect RGF to be more efficient per pound spent simply because it is spending less money. Civil servants may not be able to perfectly rank projects, but I don't believe that their selection is completely random, so the fact that the fund is smaller means it should achieve higher returns.
A final note of caution on the employment numbers - if all of government truly believed these numbers you might expect to see a lot more spending on RGF (unless they think that the smaller size of the scheme drives the high returns - as discussed above).
What about growth? Here I think there are further reasons to be cautious. In our work on RSA, we were able to find a causal effect of government money in increasing employment and investment, but not productivity. In addition, assisted firms are on average less productive, so RSA expands employment in less productive firms. This is still a 'growth' effect to the extent that these workers would have been unemployed (and we find some evidence, for RSA, that this might have been the case). But increasing the employment share of less productive firms may not be a good long run strategy for driving growth.
Indeed, if growth is the absolute priority then you begin to wonder whether the government might be better off dropping the 'R' from the Regional Growth Fund. The economics of that are difficult. On the minus side it might be more difficult to find projects in the 'south' where employment generation is genuinely additional. Offsetting this is the fact that a Growth Fund would expect to be generating those jobs at relatively more productive firms. Of course, while the economics might be difficult, the politics of such a change are far trickier.
Friday, 28 October 2011
Crime Maps
The government has added more crimes to its online crime maps.
The discussion on the Today programme centred around the extent to which the maps, together with new police commissioners might skew decisions on how to use police resources. Back in July, the worry was around whether this would skew incentives to report crimes.
The magnitude of both these effects is unknown. One thing that is certain, however, is that reported crimes have a big effect on house prices. To the extent that this is valuing the costs of crimes (at least to residents) then you would think it should have some bearing on the allocation of resources (independent of the mechanism through which this is achieved). Steve Gibbons interesting post from July has more details.
The discussion on the Today programme centred around the extent to which the maps, together with new police commissioners might skew decisions on how to use police resources. Back in July, the worry was around whether this would skew incentives to report crimes.
The magnitude of both these effects is unknown. One thing that is certain, however, is that reported crimes have a big effect on house prices. To the extent that this is valuing the costs of crimes (at least to residents) then you would think it should have some bearing on the allocation of resources (independent of the mechanism through which this is achieved). Steve Gibbons interesting post from July has more details.
Wednesday, 26 October 2011
High Speed Fail
The Adam Smith Institute's 'high speed fail', represents the latest effort outlining the anti-side of the HS2 argument.
The Campaign for High Speed Rail has already responded: According to the FT the Campaign portrays the ASI's opposition to HS2 is “purely ideological, as they are fundamentally opposed to large-scale infrastructure investment [... begging] the question as to why such groups failed to also dismantle the case for projects such as Crossrail and the Jubilee Line extension, which were based on far lower financial returns.”
My overall position on HS2 remains unchanged - the costs of the project are large and I think that the money could be better spent. I am not, however, ideologically opposed to large-scale infrastructure investment. Indeed, I am more sympathetic to the case for Cross-Rail (and previously for the Jubilee Line extension). This is partly because I think that the (narrow) user benefit case for these latter two projects relies on less extreme assumptions about the growth in passenger numbers (and I don't remember them having 'far lower' CBA figures). But I am also more sympathetic because I think that the wider economic benefits (not captured by traditional analysis) are likely to be larger for schemes freeing up bottlenecks within our more successful cities. In contrast, I am not convinced that the wider economic benefits of HS2 will be large (and consistent with this I would prefer to see the money spent on within city transport schemes with better benefit-cost ratios).
In short, while I am sure that the ASI are perfectly capable of defending their own position, it is not contradictory to be supportive of some transport schemes and not others.
The Campaign for High Speed Rail has already responded: According to the FT the Campaign portrays the ASI's opposition to HS2 is “purely ideological, as they are fundamentally opposed to large-scale infrastructure investment [... begging] the question as to why such groups failed to also dismantle the case for projects such as Crossrail and the Jubilee Line extension, which were based on far lower financial returns.”
My overall position on HS2 remains unchanged - the costs of the project are large and I think that the money could be better spent. I am not, however, ideologically opposed to large-scale infrastructure investment. Indeed, I am more sympathetic to the case for Cross-Rail (and previously for the Jubilee Line extension). This is partly because I think that the (narrow) user benefit case for these latter two projects relies on less extreme assumptions about the growth in passenger numbers (and I don't remember them having 'far lower' CBA figures). But I am also more sympathetic because I think that the wider economic benefits (not captured by traditional analysis) are likely to be larger for schemes freeing up bottlenecks within our more successful cities. In contrast, I am not convinced that the wider economic benefits of HS2 will be large (and consistent with this I would prefer to see the money spent on within city transport schemes with better benefit-cost ratios).
In short, while I am sure that the ASI are perfectly capable of defending their own position, it is not contradictory to be supportive of some transport schemes and not others.
Tuesday, 25 October 2011
Radical Solutions to the Housing Crisis
The BBC News Magazine suggests eight radical solutions to the housing crisis. Here are previous SERC blog posts on some of them (or other closely related issues):
My preferred 'radical' solution - build more houses.
My preferred 'radical' solution - build more houses.
Subscribe to:
Posts (Atom)