Friday 1 August 2014

Planning supermarkets away, for less convenience and variety, higher prices and lower productivity

[By Paul Cheshire and Christian Hilber]

One of the 'joys' of putting together a serious evidence based analysis of the effects of our planning system is to have planners turn round and dismiss the results because the analysis has not evaluated the benefits of planning. This is especially true since one of us was the first – and still one of the very few – to attempt rigorously to evaluate the net effects of restrictions on land supply; and found them to be substantial and negative in terms of their welfare effects.

The issue is in reality that very powerful and dirigiste planning policies are introduced with no attempt to analyse the value of either the benefits they might generate or the costs they impose. Because it is absurd to try to claim that planning policies do not have any costs it is surely better to have a reasonable measure of what those costs are so when we evaluate benefits – whether identified quantitatively or qualitatively – we can set the benefits against their costs. Otherwise it is rather like the ‘ladies’ menu in a posh restaurant. Great claims may be made for the dishes but it is not possible to make an informed choice of what to eat unless you know that the foie gras costs £45 while the excellent artichokes are a snip for a fiver.

Just such a policy is Town Centre First (TCF). The strict version was introduced almost on a whim in 1996 with the aim of concentrating new retail development on particular sites in central locations. One of the outcomes of this policy is highly visible: small local stores – such as Tesco Expresses or Sainsbury Locals – have mushroomed in locales that are deemed to be 'town centres' according to planners. At the same time, very few large scale supermarkets – built after 1996 – can be found out-of-town, where English households increasingly decide to live (ironically, in part this suburbanization is driven by planning restrictions in urban areas that make housing in those places increasingly unaffordable). This is because of the so called ‘sequential test’ that was designed to rule out all possible sites before allowing an out-of-town site even to be considered. The ultimate outcome of this has been that it became all but impossible to develop large format out-of-town stores in England after 1996.

There are a number of obvious adverse consequences of forcing retailing into small and often awkward sites in ‘town centres’: these locations will be less convenient for suburban shoppers (an ever growing share of the population) and the smaller stores will, compared to large format out-of-town stores, be able to offer less variety, at higher prices.

Another potential cost of TCF policy could be that it lowers efficiency by forcing stores onto more awkward and difficult to manage sites in intrinsically less productive locations. A major reason for their intrinsic lower productivity is likely to be the difficulty of supplying them efficiently. Supply depots remain located with respect to the motorway system and local planners and politicians (rather than retailers) choosing store sites in ‘town centres’ is not a recipe for efficiency. It is issues such as these that we explore in an article that is forthcoming in the Journal of Economic Geography.

Using unique store-specific data from one of the four largest supermarket chains in the UK and exploiting useful variation in TCF policies between England and Scotland and Northern Ireland – where TCF policies were introduced later and much less rigorously, especially in Northern Ireland - we identify the loss of output imposed by the implementation of TCF policy in England. Also we have data on how planning restrictiveness has varied across English Local Authorities (LAs) since 1979.

Our findings are staggering. The first is that the most restrictive LAs more or less plan supermarkets away from the communities they serve. A one standard deviation increase in the restrictiveness of a LA (which would move the LA from about average to be just in the top 15% in terms of restrictiveness) reduces the probability of their being a supermarket in it by 26%. Since more restrictive LAs also restrict the size of stores, the same increase in restrictiveness reduces the chain’s floor area in the more restrictive LA by 42%. The resulting scarcity of supermarket space does mean more sales per sq ft of floor area so the chains’ sales are reduced by ‘only’ 32% for a one SD increase in local planning restrictiveness.

Turning to the direct impact of TCF policy, according to our most conservative estimate, the implementation of strict TCF policy in 1996, combined with the initial effect of tightening controls on out-of-town stores in 1988, caused a total loss of sales of some 32%. This is the total loss of output, all else controlled for, observed in an English store opened after 1996 compared to stores that opened up prior to 1988 (when retailers in England were still pretty free to choose optimal locations) - equivalent to more than a lost decade of growth in retail output.

Our analysis suggests that the gross cost of constraining retail to sites and locations chosen by planners and local politicians rather than by retailers and shoppers, has a staggering price tag attached to it, up there with the £45 foie gras. Whether the price tag is too high depends of course on the benefits forcing retail to ‘town centres’ may deliver. The declared aim of TCF has been to make cities more ‘sustainable’ and retain access to shops for those without cars. To deliver any final verdict on TCF, these alleged benefits would also need to be rigorously quantified.

As it is they are no more than claims because there is no evidence they exist. This is what we are trying to do in an ongoing project. Our still provisional findings do not suggest that TCF policies ‘deliver the goods’. What we find is that, since TCF was strictly imposed in England, adding new stores in a local shopping area reduces distances travelled for shopping in both England and Scotland, but it has done so much more in Scotland. This is not really surprising. Whereas in Scotland retailers built stores where households increasingly live and want to shop (in suburban and ‘out-of-town’ locations), in England planners and local politicians have deliberately made this much more difficult; new stores are disproportionately on sites that are less convenient for shoppers. To be sure; some shoppers will be better off. One of us does not have a car and lives near a ‘town centre’ high street. He appreciates the additional stores. However, not many households own no car and live near ‘town centres’. So, while TCF may be the equivalent of a nice meal for a few, for most of us it is probably an empty plate: or to mix our fables - emperor’s clothes. Either way it seems to have a very high price tag attached to it.