The British System of Land Use Regulation: Key features and (unintended) economic consequences

Centre for Cities Research Seminar, London, 4 December 2013 The British System of Land Use Regulation: Key features and (unintended) economic consequences Christian Hilber London School of Economics December 2013

Overview 1. The policy issues some stylized facts 2. Features of the British system of land use regulation and implications 3. Empirical evidence a) The casual impact of local regulatory constraints on house prices in England b) Putting the evidence in an international context c) Impact on retail and office markets 4. Conclusions 2

Overview 1. The policy issues some stylized facts 2. Features of the British system of land use regulation and implications 3. Empirical evidence a) The casual impact of local regulatory constraints on house prices in England b) Putting the evidence in an international context c) Impact on retail and office markets 4. Conclusions 3

Stylized fact 1 House values in England particularly in London and SE are amongst highest in world Mean price of single detached house (all transactions in 2008): 1) Kensington: 4.3M Richmond: Hackney : Cotswold: 1.2M (greenish London suburb) 770k (rather distressed London borough) 470k (rural West of England) Buying price per square metre second highest in the world (topped only by Monaco) 2) Sources: 1) Land Registry; 2) Globalpropertyguide.com (last accessed 3/2013) 4

5 Real house price growth in %, average 1970-2006 Source: Bank for International Settlements (BIS) 4 3 2 1 0 5

Stylized fact 2 House prices in UK (and particularly England) are also extremely volatile UK as a whole substantially more volatile than single most volatile market in US 1980s/90s cycle: boom/bust in real terms UK: +83% / -38% 1) Los Angeles: +67% / -33% 2) Sources: 1) Nationwide; 2) Glaeser et al. (2008) 6

Stylized fact 3 Volatility has increased in recent decades Real Land & House Price Indices (1975 = 100) 600 500 400 300 200 100 0 Land Price Index Source: Cheshire (2009) House Price Index Note: House and Land data for war years are interpolated. 1892 1896 1900 1904 1908 1912 1916 1920 1924 1928 1932 1936 1940 1944 1948 1952 1956 1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000.A 2004.Ju 2008.Ju 7

And there is spatial variation in volatility Source: Nationwide, FT 8

Stylized fact 4 Housing units in UK are not only extremely expensive and volatile but also extremely small by international standards A new-build house in UK is 38% smaller than in densely populated Germany and 40% smaller than in the even more densely populated Netherlands And there are very few new-build homes Source: Statistics Sweden (2005) 9

Stylized fact 4 (cont.) 20 18 Completions per 1000 inhabitants: period 2004-2007 16 14 12 10 8 6 4 2 0 Source: Euroconstruct 10

Stylized fact 5 Not just housing office space in UK is also extremely expensive (and volatile) Total office occupation costs per m 2 in Birmingham in 2004: 44% higher than in Manhattan NY (KingSturge, 2004) Construction costs about half (Cheshire & Hilber, 2008) How can we make sense of this? 11

Derived research questions What factors cause the high level and volatility of prices and corresponding space shortage? Might the British system of land use regulation be a (the main) culprit? 12

Some background: The British system of land use regulation Supply constraints and Greenbelts have long history...origin dates back until at least 1580 Subjects of Queen Elizabeth I were commanded to desist and forebare from any new building of any house or tenement within three miles of any of the gates of the City of London where no house hath been known But was never fully enforced and disappeared following Fire of London in 1666 13

Some background (cont.) Today s planning system established in 1947 through Town and Country Planning Act Key features expropriated development rights of land owners Designated use classes, whereas any change of use requires development control permission (granted at local level) Aim is development control or containment 14

What Greenbelt containment looks like 5km Reading 60km west of London 15

And in London Source: Barney s blog (http://barneystringer.wordpress.com/2013/11/21/londons-green-belt) 16

The planning system does affect urban form Dutch concentrated dispersal Reading Similar densities Less restrictive planning associated with more sprawl Wider South East green belt constraint Source: Echenique (2009) Flemish region dispersal 17 17

Who decides in UK? Local Planning Authorities (LPAs) grant or reject planning applications Problem: Since 1947 virtually no fiscal incentives at local level to permit development (costs far exceed benefits) Government reforms since 2010 not (yet) biting Underlying causes? UK = highly centralized country, virtually no fiscal power at local level Political power tilted towards homeowners (NIMBYs or better: BANANAs) Local long-run supply curve nearly vertical 18

Theoretical prediction House prices Long-run supply Demand tomorrow Demand today Housing stock 19

Theoretical prediction House prices Long-run supply Demand tomorrow Demand today Housing stock 20

How to test? Hilber and Vermeulen (2010, 2013) Exploit spatial variation in three different types of supply constraints (regulatory, scarcity of developable land and topography) Interact supply constraints with demand shifters (local earnings) Use instrumental variable technique to identify causal effect of local supply constraints measures on local house prices 21

Average refusal rate (major residential projects) 1979-2008 Share developable land developed, 1990 Elevation range Source: Hilber and Vermeulen (2013) 22

Main findings Tight local planning constraints in parts of England (in conjunction with strong demand) are to a good extent responsible for extraordinarily high house prices Local scarcity of developable land matters but very non-linearly (only in most developed locations) Topography matters in statistical sense but very little in economic (quantitative) sense 23

Quantitative effects (based on IV with all instruments) If planning were completely relaxed in average LPA: House prices in average LPA: -35% and developable land were abundant: House prices in average LPA: -45% (Δ= -10%) and LPA were completely flat: House prices in average LPA: -48% (Δ= -3%) Note: These are likely lower bound estimates for a number of reasons Intro (see Hilber Stylized and Facts Vermeulen Features 2010 of and British 2012 system for details) Empirical evidence Conclusions 24

What would house prices in average English LPA be if 250000 200000 150000 100000 226k 147k 124k 117k 112k 50000 1974 1980 1990 2000 2008 Year Predicted real house prices in average English LPA Prediction with refusal rate set to zero - and share developed set to zero - and elevation range set to zero - and earnings assumed constant 25 25

North East vs. South East & 90 th vs. 10 th percentile 300000 250000 Had the SE the restrictiveness of the NE, house prices in the SE would be 25% lower! 200000 150000 100000 50000 1974 1980 1990 2000 2008 Year Predicted real house prices in average English LPA Prediction with refusal rate as in NE / SE Prediction with refusal rate as 10th/90th percentile 26

Log of real house prices (1974 = 0) Log of real house prices (1974 = 0) But large variation across locations 1.5 1.5 1 1.5.5 0 0 -.5 -.5 1974 1980 1990 2000 2008 Year Predicted log real house prices in Westminster Prediction with refusal rate set to zero - and share developed set to zero - and elevation range set to zero - and earnings assumed constant 1974 1980 1990 2000 2008 Year Predicted log real house prices in Reading Prediction with refusal rate set to zero - and share developed set to zero - and elevation range set to zero - and earnings assumed constant 1.5 1.5 1 1.5.5 0 0 -.5 1974 1980 1990 2000 2008 Year -.5 1974 1980 1990 2000 2008 Year Predicted log real house prices in Darlington Prediction with refusal rate set to zero - and share developed set to zero - and elevation range set to zero - and earnings assumed constant Predicted log real house prices in Newcastle upon Tyne Prediction with refusal rate set to zero - and share developed set to zero - and elevation range set to zero - and earnings assumed constant 27

Deviation from Long-Run Trend Price Deviation from Long-Run Trend Price Evidence from another country with tightly and little regulated cities San Francisco, CA (inelastic supply & volatile demand).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 San Francisco 1982q1 1988q1 1994q1 2000q1 2006q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) Columbus, OH (elastic supply).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Columbus, OH 1982q1 1988q1 1994q1 2000q1 2006q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) 28

Deviation from Long-Run Trend Price Deviation from Long-Run Trend Price And another example Los Angeles (inelastic supply & volatile demand).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Los Angeles 1982q1 1988q1 1994q1 2000q1 2006q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) Chattanooga, TN-GA (elastic supply).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Chattanooga, TN-GA 1982q1 1988q1 1994q1 2000q1 2006q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) 29

Deviation from Past-50 Quarter Moving Trend Price Deviation from Past-50 Quarter Moving Trend Price San Francisco, CA And based on deviation of HP from 50q past-trend HP (inelastic supply & volatile demand).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 San Francisco 1992q1 1996q1 2000q1 2004q1 2008q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) Columbus, OH (elastic supply).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Columbus, OH 1992q1 1996q1 2000q1 2004q1 2008q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) 30

Deviation from Past-50 Quarter Moving Trend Price Deviation from Past-50 Quarter Moving Trend Price Deviation of HP from 50q past-trend HP Los Angeles (inelastic supply & volatile demand).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Los Angeles 1992q1 1996q1 2000q1 2004q1 2008q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) Chattanooga, TN-GA (elastic supply).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Chattanooga, TN-GA 1992q1 1996q1 2000q1 2004q1 2008q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) 31

But boom and bust also in places with elastic supply Ireland for example underwent massive house price and construction boom Followed by extensive price bust, high vacancy rates etc. Unlikely caused by restrictive long-run supply Bust in Ireland characterized by massive over-supply and high vacancy rates More severe bust than boom Rather resembles boom and bust in Dallas and Houston in 1980s or Las Vegas and Phoenix during the 2000s 32

Deviation from Past-50 Quarter Moving Trend Price Deviation from Past-50 Quarter Moving Trend Price The Puzzle Las Vegas (elastic supply).6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Las Vegas 1992q1 1996q1 2000q1 2004q1 2008q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) Phoenix (elastic supply) Chattanooga, TN-GA.6.5.4.3.2.1 0 -.1 -.2 -.3 -.4 -.5 -.6 Phoenix 1992q1 1996q1 2000q1 2004q1 2008q1 2012q1 Year and Quarter Source: Own calculations based on FHFA all-transactions HP index, Hilber (2013) 33

Low tier sales prices in Las Vegas and Phoenix and 15 other US MSAs Source: K.E. Case, in Land Lines (Lincoln Institute), pp. 8-13, http://www.lincolninst.edu/pubs/1743_land-lines-january-2010 34 34

So what might have happened in Ireland? Extraordinarily strong demand boost during 1990s and until mid 2000s ( Celtic Tiger years ) Supply not sufficiently responsive in short-run (due to planning and construction lags) House prices start rising significantly Myopic agents start forming unrealistic expectations about future price rises If supply is elastic in the longer-run (unlike in UK): Construction boom Then great recession hits and triggers bust phase with massive declines in house prices, vacancies, defaults & follow-on effects Similar stories in LV & Phoenix (DeFusco et al., 2013) 35

What about Dublin? Dublin appears to have tight regulatory constraints (including restrictive height controls) Political-economical equilibrium probably tilted towards owners of developed land who are dominant in Dublin and who have incentive to oppose new development (similar to London, SF, LA) (Hilber and Robert-Nicoud, 2013) Rest of Ireland appears to have elastic long-run supply Owners of undeveloped land/developers of such land (who benefit from permission to develop) are arguably politically very influential relative to owners of developed land (Hilber and Robert-Nicoud, 2013) This in combination with the common occurrence of corruption arguably lead to construction boom outside Dublin 36

A stylized explanation P STS 0 = STS 1 LTS e P 1 * P 0 D 0 e = D 1 Q 37

Unexpected initial demand boost Price increase due to inelastic short-run supply (lags!) * P 1 P STS 0 = STS 1 LTS e P 1 * P 0 D 0 e = D 1 D 1 Q 38

Myopic (and/or) exuberant agents & elastic long-run supply cause construction boom P STS 0 = STS 1 STS 2 LTS * P 1 e P 1 * P 0 D 0 e = D 1 D 1 Q 39

Myopic (and/or) exuberant agents & elastic long-run supply cause construction boom P STS 0 = STS 1 STS 2 LTS * P 1 e P 1 * P 0 e D 2 D 0 e = D 1 D 1 Q 40

Bust phase (triggered by negative shock) can lead to over-supply/ high vacancies P STS 0 = STS 1 Vacancies (sticky prices) STS 2 LTS * P 1 e P 1 * P 0 e D 2 D 0 e = D 1 D 1 D 2 Q 41

and eventually massive price decline (unlike in UK!) P STS 2 LTS * P 1 e P 1 * P 0 Price Falls e D 2 * P 2 D 0 e = D 1 D D 1 2 Q 42

What about impact of British planning system on retail markets and office markets? Two particularly interesting policy reforms 43

Town Centre First policy Town Centre First strictly implemented in England in 1996 Needs test : Need to prove that more shopping space is needed locally Sequential test : Need to prove no more central site is available Made major out-of-town retail shopping in England difficult after 1988 and all but impossible after 1996 Put differently: Location and site-selection effectively micro-managed by planners rather than supermarket chains 44

Economic consequence? Exploiting a DiD-type setting, our estimates suggest that TCF policies imposed loss of output of some 32% on stores that opened in England after 1996 compared to stores that opened prior to 1988 (Cheshire et al., 2013) 45

Last example: Nationalisation of business rate In 1990 Thatcher s government converted the commercial property tax from a local to a national basis Removing any fiscal incentives at local level to permit commercial development Economic consequences? Removal of fiscal incentives made supply of office space inelastic (no more incentives to approve developments) Estimates suggest regulatory tax imposed on commercial firms in form of higher office space prices far exceeded revenue from business rate (Cheshire and Hilber, 2008) The law of unintended consequences is powerful indeed 46

Conclusions Planning serves important purpose - in principle it can improve welfare through correcting market failure such as externalities and public goods But difficult to design system that strikes right balance British system far too restrictive, but more flexible systems also have their downsides Irish system seemingly did not get balance right either So, how should ideal system look like? 47

Some guiding principles: The ideal planning system should 1. Focus on correcting market failures (externalities, public goods) that are endemic in land markets 2. Work with the grain of the markets Planners ought not micro-manage location choices or specific site selection (Cheshire et al., 2013) Planners ought to take into account price signals (rather than ignore them) (Cheshire and Sheppard, 2005) 3. Align incentives Those who bear the costs of development should also reap the benefits (Hilber and Vermeulen 2010, 2013) Possible tools: impact fees (reflecting marginal social costs), genuine local property tax (& get rid of stamp duty) (Hilber and Lyytikäinen, 2013; Mirrlees et al., 2011) 48

One last point In order to tackle affordability problem, policy makers tend to endorse policies that boost housing demand, especially demand for owner-occupied housing Help to buy (in UK) Mortgage interest deduction (in US and elsewhere) No capital gains tax and no inheritance tax on principal owner-occupied dwelling (almost anywhere) In places with tight regulatory constraints like in Britain, this merely increases house prices further making owner-occupied housing less not more affordable for young-would-be buyers (Hilber and Turner, forthcoming; Hilber 2013) 49

Q & A Thank you! Presentation with references & hyperlinks will be downloadable from: http://personal.lse.ac.uk/hilber/ 50

References (and further readings) Cheshire, P. and S. Sheppard (2005) The Introduction of Price Signals into Land Use Planning Decision-making: A Proposal, Urban Studies, vol. 42, No. 4, 647-663. Cheshire, P. and C. Hilber (2008) Office Space Supply Restrictions in Britain: The Political Economy of Market Revenge, Economic Journal, 2008, vol. 118, Issue 529, F185-F221. (Latest discussion paper) Cheshire, P. (2009) Urban Containment, Housing Affordability and Price Stability Irreconcilable Goals, SERC Policy Paper No. 4, September. Cheshire, P., C. Hilber and I. Kaplanis (2013) Land Use Regulation and Productivity Land Matters: Evidence from a Supermarket Chain, London School of Economics, SERC Discussion Paper No. 138, August. DeFusco, A., W. Ding, F. Ferreira and J. Gyourko (2013) The Role of Contagion in the Last American Housing Cycle, Wharton School, mimeo, June. Echenique, M. (2009) Sustainable Cities, Presentation given at the Spatial Economics Research Centre Policy Seminar, London, 15 October 2009. 51

References (cont.) Glaeser, E.L., J. Gyourko, and A. Saiz, (2008). Housing supply and housing bubbles, Journal of Urban Economics, vol. 64(2), pp. 198-217. Globalpropertyguide.com (http://www.globalpropertyguide. com/most-expensive-cities; last accessed: 28 March 2013) Hilber, C. (2013) Help to Buy will likely have the effect of pushing up house prices further, making housing become less not more affordable for young would-be-owners, British Politics and Policy at LSE Blog, June 25. (http://blogs. lse.ac.uk/politicsandpolicy/archives/34388) Hilber, C. and T. Lyytikäinen (2013) Housing Transfer Taxes and Household Mobility: Distortion on the Housing or Labour Market?, VATT (Government Institute for Economic Research) Working Paper No. 47, June. Hilber, C and F. Robert-Nicoud (2013) On the Origins of Land Use Regulations: Theory and Evidence from US Metro Areas, Journal of Urban Economics, vol. 75, No. 1, 29-43. 52

References (cont.) Hilber, C and T. Turner (forthcoming) The Mortgage Interest Deduction and its Impact on Homeownership Decisions, Review of Economics and Statistics. (link to latest paper version) Hilber, C. and W. Vermeulen (2010) The Impacts of Restricting Housing Supply on House Prices and Affordability Final Report, London: Department for Communities and Local Government. Hilber, C. and W. Vermeulen (2013) The Impact of Supply Constraints on House Prices in England, IEB Working Paper No. 2013/28, July. (Older version: SERC DP #119) KingSturge (2004) Global Industrial and Office Rents Survey, London: KingSturge. Mirrlees, J., S. Adam, T. Besley, R. Blundell, S. Bond, R. Chote, M. Gammie, P. Johnson, G. Myles and J. Poterba (2011) Tax by Design: the Mirrlees Review. Oxford University Press. Statistics Sweden (2005) Housing Statistics in the European Union 2004, Karlskrona: Boverket, Publikationsservice. 53