Policies for 'mixed communities': a critical evaluation

Policies for 'mixed communities': a critical evaluation

Land Use Regulation and Retail: Space Constraints and Total Factor Productivity Paul Cheshire, Christian Hilber and Ioannis Kaplanis ERES Conference, Milan 24th June 2010 This paper: hypotheses & intended contribution Seems likely planning policy restricts land available for retail development: so increases costs of space: reduces retail TFP Try to quantify the impact by: 1) estimating production function - including space 2) Investigating connection to differences in planning restrictiveness 3) Quantify impact on TFP and retail prices Problem: Planning policy may negatively affect TFP via two distinct routes: 1)Restriction of land supply for retail raises prices and cause profit maximising retailers to substitute land out of production 2) Town centre first policies may force to locate on smaller and less productive, higher cost ( for logistics, labour, customers) sites At this stage not distinguishing Using microdata and detailed planning performance data The issues. Three factors of production: land labour and capital Forget land (unless agricultural economist) But land an input into production in retailing: think Ikea!! In 1980s land for retailing in prosperous SE of UK 250 X land for retailing in comparable US location (Cheshire & Sheppard 1986) UK Planning system imposes (intentional) restrictions on supply of urban land via containment & 60% brownfield And restricts for each (legally classified) use Not surprising increases cost of housing: reduces supply elasticity => so increases volatility Nearly all work so far on housing; But Hilber & Cheshire 2008 costs of office space Much higher in UK than continental Europe or New York

tax on space in London West End equivalent of 800% over 1999-2005 Town centre first + virtual prohibition on out of town large scale development => even higher cost for retail? Another peculiarity of British planning reliance on development control => more politicised, less planned The issues. Increasing support for idea that planning policies reduce productivity in retail: McKinsey Global Inst. 1998; Barker, 2006; Haskel & Sadun, 2009 Haskel & Sadun - first academic study: by preventing emergence of large format out of town stores estimates lost 0.4% p.a. TFP growth 1996-2006 Also Competition Commission 2000; 2008 Well worth looking at: access to store level micro data for 4 main supermarket groups Strong finding larger stores more productive and profitable More local competition reduces store prices (CC 2008) And land for retail in UK x 5 to 10 in France (CC 2000) Planning policy and its impact Prior to 1988 relatively relaxed approach to retail as such though clear evidence of overall space restriction via containment e.g. Reading 1984 1988 PPG6 tried to steer out of town to regeneration sites e.g. Bluewater but still not restrict competition PPG6 1993: attempts steer to in-town sites because of belief free market might under-provide in town shopping Big change PPG6 1996 -More or less prohibited out of town development for all town centre activities i.e. not just retail but offices, leisure, restaurants -Introduced Needs test + Sequential test Fear - mainly a development control tool ODPM (2004) Confirmed even reinforced by PPS6 2010 And implementation requires current local development plan estimated less than half LPAs have them 2 applications for retail- major 3 4

5 Figure 1: Number of Applications for Major Retail Developments, 1979-2008 1980 1990 calendar year 2000 2010 Figure 2: Applications for Extensions to Foodstores, 1990 to 2001 Figure 3: Big 5 Supermarket In- and Out of Centre Openings, 1990-2000 In-centre opening rise relative to out of centre. But note in- or ou of centre defined for planning purposes Merryhill Figure 4: Age of Building Stock by Use Category And an aging stock of retail buildings. Data, approach and some problems Store level data for all stores for major retailer mainly food Detailed development control data for all LPAs (so far collected only England): applications, refusals, delays & appeals Stores geocoded - so also data for store catchment areas population within given drive times, car ownership, competitor

stores x distance, etc Some summary statistics Table 2 Summary Statistics Variable Obs Mean Std. Dev. Min Max Sales/employment Sales Employment 357 357 357 4246 921115 213 544 406300 85 2349 5706 73978 2056014 32 471 Net floorspace (sq.ft.) Gross floorspace (sq.ft.) Food floorspace Non-food floorspace Net/gross floorspace (ratio) Density (empl/1,000 Sq.ft) 357

357 357 357 46710 81633 27819.6 18890.5 17352 31095 10144.7 9859.5 8313 15076 0 671 101091 180000 54290 52576 357 357 0.58 4.57 0.07 1.10 0.33 1.01 0.83 7.40 Non-food format (dummy) Mezzanine (dummy) Petrol station (dummy) Parking spaces Years since first opening Total opening hours 357

357 335 356 357 357 0.06 0.17 0.52 576 14.4 119 0.24 0.38 0.50 264 10.5 29 0 0 0 82 1 64 1 1 1 2000 43 168 Population within 10mins Car ownership share within 10 mins drive Competition variable 357 81226 43706 5532

229246 357 357 0.70 4.97 0.08 3.49 0.45 0.29 0.88 23.30 Data, approach and some problems How measure planning restrictiveness? Use refusal or delay rate? Problem of endogeneity developers behaviour may be influenced by LPAs the discouraged developer effect So need instruments to identify: 1. Exploit change in targets for delays more than 13 weeks 2002 separate for minor and major Expect more restrictive LPAs to both refuse more and delay more: not possible post-2002 =>So use change in delay rate pre- & post2002 2. Or use political make-up of LPAs (Cheshire & Hilber, 2008: explicitly Haskel & Sadun, 2009, Hilber & Vermeulen, 2010); rise of NIMBYism Figure 5: Plotting the coefficients from regressing refusal rate on delay rate: Residential (major) 1979-2008 .2 0 -.2

-.4 b coefficient (major residential projects) .4 435 1980 1985 Graphs by lacode_num 1990 1995 calendar year 2000 2005 2010 re 6: Plotting the coefficients from regressing refusal on delay rate: Retail (major) 1979-2008 0 .2 Nos of major retail low relative to major resid - so more noise -.2 b coefficient (major retail projects) .4 435

1980 1985 Graphs by lacode_num 1990 1995 calendar year 2000 2005 2010 But are Town centres actually town centres? The case of Merryhill; the comparative lack of current local development plans Town centre versus out of town may be planning definitions more than geographical, functional or economic! Test 1) does size of store vary with planning location? 2) does price of space vary with official locational classification? 3) are planning locations strongly related to distance from town centre e.g. major rail stations? Or do PPG6 1996 & PPS6 2010 really just more or less prevent all retail development and particularly large format retail development? Done 1) & 2) e 3a Number of stores and average floors by location type

Location Type No of stores Mean Net floorspace (sq.ft.) S. D. Town Centre 46 42609 15429 District Centre 41 45564 18053 Suburban Centre 25 44732 10202 Edge of Centre 63 43598 16527 Out of Town 123 50889 17459

Destination 13 63760 22824 Retail Park 25 52015 14063 Non-food Format 21 28279 5086 nly Destination stores clearly larger on average Table 3b Floorspace costs by location type 6.7 6.9 Rateable value 2005/gross floorspace 12.9 14.4 27.8 26.3 26.7 31.8 27.8 4.7 6.0

5.8 3.8 9.3 Non-food Format 13.8 All stores 25.7 Location Type Town Centre District Centre Suburban Centre Edge of Centre Out of Town Destination Retail Park Rateable value 2005/net floorspace 23.5 24.7 3.6 4.6 Rateable value 2010/net floorspace 33.5 37.1 15.3 15.0 15.4 17.6 16.2 2.5

3.8 3.5 3.8 6.3 4.5 9.9 6.8 14.8 S.D S.D. S.D. No of stores 8.9 9.5 45 39 35.9 36.2 37.8 41.2 40.6 6.7 6.9 6.6 4.9 14.4 21 60 112 12 21 2.8

17.4 5.6 13 4.1 36.1 9.0 323 it price of Destination stores highest: town centre est - contrast with distance decay of price in Reading 1984 Simple Cobb-Douglas production function Y = A F 1 L2 K3 eX eu lnY i = 0 + 1 lnFi + 2 lnLi + 3 lnKi + X i + X +u (RTS= 1 + 2 + 3) Y: sales of store i; or gross margins Y= PQ- PwQw or Y= PQ - PwQw PmM F: floorspace; L: labour; K: capital for store i Xi: vector of store specific controls X: vector of area specific controls etailed info on margins but assured they are consta em across stores. So using sales as measure of out 2000 3000 4000 5000

6000 Figure 7: Relationship of productivity (sales/employment) to net floorspace 0 20000 40000 60000 NET SALES AREA (SQ FT) Sales per employee 80000 Fitted values 100000 Table 4: Basic results from a TFP approach with Total Sales as output (1) (2) (3) (4) 0.0472 (1.407) 0.0972 (2.665) 0.128 (2.719) 0.118 (2.542) 1.083

(37.76) 1.043 (35.42) 1.000 (22.15) 0.974 (20.27) -0.0594 (-2.815) -0.0499 (-2.408) -0.0547 (-2.685) -0.0815 -0.0775 (-1.091) (-1.052) VARIABLES Net Floorspace Employment Mezzanine dummy Non-food format dummy Hours Constant Observations R-squared 0.000915 (3.246) 7.405 (29.64)

7.093 (26.25) 6.989 (23.45) 7.126 (24.54) 357 0.958 357 0.959 357 0.959 357 0.961 Findings. Clear evidence productivity rises with store size Elasticity 0.1 to 0.13 Productivity also rises with number of hours open and employment Falls with non-food format and if mezzanine Non-food format stores have different production functions Add controls: Competition Characteristics of catchment area Age of store (date of opening)

Test model only on English sample (availability of planning data) Table 5 Add further store & area controls; UK&England (5) (6) (7) (8) (9) (10) VARIABLES UK UK UK UK UK ENGLAND 0.135 (2.925) 0.140 (3.107) 0.102 (2.185) 0.103 (2.207) 0.115 (2.538) 0.144 (2.559) Employment

0.936 (19.39) 0.902 (18.86) 0.918 (19.29) 0.913 (18.77) 0.899 (18.94) 0.846 (13.79) Mezzanine dummy -0.0430 (-2.168) -0.0393 (-2.025) -0.0387 (-2.081) -0.0382 (-2.020) -0.0391 (-2.110) -0.0365 (-1.765) Non-food format dummy -0.105 (-1.433) -0.133

(-1.821) -0.135 (-1.839) -0.140 (-1.891) -0.145 (-1.958) -0.257 (-2.870) Hours 0.00106 (3.745) 0.00102 (3.653) 0.00101 (3.695) 0.00104 (3.787) 0.00103 (3.807) 0.000905 (2.541) Years since opening 0.00222 (3.402) 0.0106 (3.925) 0.00900 (3.335) 0.00910

(3.377) 0.00942 (3.529) 0.0123 (4.074) -0.0235 (-3.428) -0.0201 (-2.957) -0.0203 (-3.021) -0.0213 (-3.195) -0.0272 (-3.705) 0.0444 (3.742) 0.0491 (3.799) 0.0570 (4.164) 0.0509 (2.885) 0.0769 0.0945 0.0740 (1.050) (1.293) (0.835)

-0.00379 (-2.078) -0.00415 (-2.236) Net Floorspace Years since opening sq. Population within 10mins Car ownership share within 15m Competition variable Constant Observations R-squared 7.098 (24.78) 7.183 (25.56) 7.024 (25.42) 6.923 (22.57) 6.783 (22.21) 6.844 (19.33) 357 357 357 357

357 269 0.962 0.963 0.965 0.965 0.965 0.965 Figure 8: Productivity by year of opening of store age is interesting/suggestive using estimates model (9) =>Oldest stores least productive (no surprise) oductivity falls cet. par. in stores founded from late 198 lls strongly thereafter. Looks like PPG6 . productivity 6.9 6.88 6.86 6.84 6.82 productivity 6.8 6.78 6.76 2008 2005 2002 1999 1996 1993

1990 1987 1984 1981 1978 1975 1972 1969 1966 6.74 Role of planning?. Is store size influenced by restrictiveness of local LPA? Test against: 1. Refusal rate both major residential and major retail (note major retail numbers can be small and seem noisy) 2. Instrument 1 change in delay rate following new targets in 2002 - measured as change in mean delay rate 1994-98 & 2004-08 3. Instrument 2 % share of labour seats at the local elections (average over 2000-2007) Table 6: Regressing floorspace on planning restrictiveness (major residential projects refusal ratio); IV: share of Labour seats VARIABLES (1) All England (2)

>1980 (3) >1990 (4) >1997 Refusal rate (residential) OLS -0.485 (-1.508) OLS -0.642* (-1.818) OLS -1.058** (-2.255) OLS -0.900 (-1.583) IV -0.746 (-1.401) IV -1.024* (-1.782) IV -1.546** (-2.036) IV -1.371 (-1.466) -0.192 (-12.37) -0.191

(-12.81) -0.198 (-12.25) -0.190 (-10.34) 153.05 254 164.22 221 149.96 143 106.85 114 Refusal rate (residential) First Stage % of Labour seats F excl.instr. Observations Notes: The dependent variable is log(net floorspace). The sample excludes non-food formats. The sample is restricted to the stores that are located in England only regulation data collected The refusal rate is calculated as the ratio of declined major residential projects applications to the total number of applications and averaged over 1979-2008 ; t-statistics in parentheses Table 7: Regressing floorspace on planning restrictiveness- alternative measures OLS regressions (2) (3) (4) VARIABLES (1)

All England >1980 >1990 >1997 Refusal rate (retail projects) -0.0509 (-0.180) -0.132 (-0.441) -0.294 (-0.621) -0.223 (-0.426) Change in delay rate (major residential) 0.0688 (0.565) 0.0333 (0.255) 0.371** (2.082) 0.455** (2.029) 254 221 143 114

Observations Notes: The dependent variable is log(net floorspace). The sample excludes non-food formats. t-statistics in parentheses. The sample is restricted to the stores that are located in England only planning data collected. refusal rate: ratio of declined major retail project applications to the total number of applications and averaged over 1980-2008 (the period for which regulation data exist). delay rate: change in the average delay ratio of applications pending for more than 13 weeks between the period 1994-98 and the period 2004-2008. Conclusions 1. Strong confirmation that productivity rises with store size So - restricting stores sizes by either direct constraints on sites/formats, or restricting supply of land so raising prices =>Increases resource use in retail and raises retail prices Clear welfare cost: but not yet quantified (possible) 2. Clear evidence that more restrictive local planning policy causes stores to be smaller By implication planning policy responsible for lower retail productivity See impact of restrictiveness from late 1980s and esp. 1990s Since poorer spend proportionately more of disposable income in stores (esp. food) this is distributionally regressive Net costs? What are the benefits esp. of Town centre first? Concluding Discussion Benefits? Claimed Town centre sites most sustainable because most accessible by alternative transport modes + allow linked trips so reducing need to travel But need to distinguish between what people should do and what they actually do Continue to decentralise: use cars for shopping: car use continues to rise at about same rate just more congestion So town centre locations likely: 1. Separate households from shops lead to longer & more congested trips 2. Reduce shop sizes more trips plus more restocking 3. Increase logistics costs To test - but seem likely benefits = additional costs (+carbon)

Recently Viewed Presentations

  • Insurance accounting change FASB Exposure Draft

    Insurance accounting change FASB Exposure Draft

    Increase the decision usefulness of the information about the nature, amount, timing, and uncertainty of cash flows related to long-duration insurance contract assets and liabilities, and the effect on the financial statements. Improve the financial statement comparability of insurance entities...
  • France '14 - Kinlochleven High School

    France '14 - Kinlochleven High School

    France'14 31st August- 7th September Extra costs: Should the behaviour of a child not confirm with the signed behaviour agreement, the pupil in question might face sanctions.(e.g. activity ban, room relocation, word home)
  • Reaching for the sky with - TMCnet

    Reaching for the sky with - TMCnet

    Reaching for the sky with Asterisk and Amazon EC2. ... an email will be sent to you for final confirmation. ... To create a new certificate, simply log-on to your account and create the certificate from their. You will also...
  • Cubisme et futurisme - maupassant-lyc.spip.ac-rouen.fr

    Cubisme et futurisme - maupassant-lyc.spip.ac-rouen.fr

    « Nous chanterons les foules agitées par le travail, par le plaisir ou par l'émeute : nous chanterons les marées multicolores et polyphoniques des révolutions dans les capitales modernes ; nous chanterons la ferveur nocturne vibrante des arsenaux et des...
  • Weather and Climate  Weather atmospheric conditions at a

    Weather and Climate Weather atmospheric conditions at a

    1 NEXT Hurricanes • Huge storms called hurricanes, or typhoons: - form over warm, tropical ocean waters - hit land with heavy rain, high winds, storm surge Tornadoes • Tornado a powerful, funnel-shaped column of spiraling air: - born from...
  • USC Brain Project Specific Aims

    USC Brain Project Specific Aims

    Computational Architectures in Biological Vision, USC, Spring 2001 Lecture 12. Visual Attention Reading Assignments: None Several Forms of Attention Attention and eye movements: - overt attention (with eye movements) - covert attention (without eye movements) Bottom-up and top-down control: -...
  • Bamboo Bike Capstone Team Members Camrun Arkills Kyle ...

    Bamboo Bike Capstone Team Members Camrun Arkills Kyle ...

    Alexi Lamadrid. Owner, Bamboo'd Gear. Outline. ... Wood support board is less sturdy and requires holes to be drilled manually to attach jig components. ... Bamboo Bike Capstone Team Members Camrun Arkills Kyle Heitmeyer Fay Gibson Matt Brown
  • Dear PowerPoint®* Users,

    Dear PowerPoint®* Users,

    * Opportunity Overview Potential Earnings As an Herbalife Independent Distributor, you can become eligible to earn from six areas of potential income and incentives Retail Profit Wholesale Profit Royalty Overrides Monthly Production Bonus Annual Bonus (Mark Hughes Bonus) Promotions The...