Unemployment

Every major development scheme is accompanied by a stream of plausible mis-information, as the applicants strive to convince the public that these are completely innocuous proposals which will bring unparalleled benefits to every sector of society. Such claims will be carefully tuned in the weeks before the application is submitted, to exploit local grievances and aspirations, and to prey on local fears. Usually there will be a "cherry" - some cheap but superficially attractive feature which can be portrayed as a major public benefit. Eventually will come the "bums' rush" - a skillfully timed suggestion that unless the planning authority accepts this unrepeatable special offer without further delay, something terrible will happen to them, or at the very least the developers will take their valuable business elsewhere.

From the developer's viewpoint, the need for speed is paramount. If there is undue delay then people may begin to question their claims. The last thing a speculative property developer wants is a meaningful dialogue with the public.

The Berkshire Plan was a marvellous illustration of the developers' skills, a text-book example that repays careful study.

The application was a departure from the statutory development plan, which allocated the area for playing fields and public open space. This meant that it must be advertised, and the minimum period for a decision would be three weeks. It would be difficult to guarantee responses from the statutory consultees within such a short period, so the developer and the LDC aimed for seven weeks, from 1st April 1993 when the application was submitted, to the LDC board meeting on 20 May.

The developer made no attempt to show his actual planning application to the public. Those who took the trouble to visit the city centre planning office (about 4km distant), and knew what to ask for, would discover that this was an outline planning application which sought approval only for the means of access (a large gyratory system on the busy A65 radial commuter route) and the total floor areas of the supermarket, shopping mall and multiplex cinema. These were the key features which affected the economics of the development. Almost everything else was undefined, and a matter for future negotiation with the planning authorities. There were illustrative drawings, and a great deal of supporting text, but these had no legal significance.

It is important to realise that under UK law the outline application is the definitive planning approval. Once this has been granted, it is pointless objecting to the reserved matters, because the developer will always win. You might influence the details but you cannot overturn the principle of development by arguing about the reserved matters.

Most people don't realise this, and imagine that because the developer shows a few pretty pictures this is what he must actually build. He might do, but only if there's money in it.

The developer's public exhibitions followed the three-week statutory consultation period, leaving the minimum interval for discussion and debate. The displays made effective use of colour and illustrative drawings, and the oral presentation was extremely plausible. Some visitors mistook the supermarket car park for an orchard, because it was completely covered with a regular grid of bright green trees. There was no scale on the plans, and it was not immediately obvious that most of the 300 trees on the computer-generated drawings would be about 12.5 metres in diameter, while the largest were about 20 metres across.

The trees were a smart move. Most people shown a picture of a tree will relate it to their own domestic experience. Very few city trees are more than 6 metres in diameter (the size of a small house) and most are much smaller. The effect of the drawings was to reduce the apparent area of the development by at least a factor of four. In reality the green "orchard" represented hard-surface parking for about 1500 cars.

Kirkstall Valley Properties might well have got away with it, but for one fatal flaw: their gyratory system didn't work. It would have been unsafe and the Council's highway department refused to build it. The unexpected nine-month delay gave objectors the chance to counter the developer's other claims. Developers often make planning applications without actually owning the site, and in reality many of the alleged "benefits" consisted of giving away other people's property. The public eventually submitted over 1000 written objections to the proposals.

It is of interest to examine the employment claims in greater detail. Supermarkets undoubtedly employ large numbers of staff. Many of these are part timers working for low pay with poor conditions of service, but the most important thing is that these are not new jobs. Total sales of food and supermarket goods are more or less constant, because there is a limit to how much food people can eat, or how many bars of soap they can use. If a new store opens and does well, competing establishments will lose trade and shed an almost exactly equivalent number of staff.

This point is made in a National Audit Office (NAO) report Creating and Safeguarding Jobs in Wales, which has been published as House of Commons paper number 664 from the 1990 - 91 session. The NAO recognise job displacement effects, and use the concept of net versus gross job creation. They also discount those changes which would have happened anyway, even if nothing had been done.

The developer made the following claims on page 5 of the "second draft" Retail Impact Impact Study commissioned from DTZ Debenham Thorpe in May 1993:

(There never was a "final" report. These retail impact assessments are often demanded by planning authorities considering major development schemes. The consultant is commonly selected and paid by developer, and almost invariably reports that the scheme will trade profitably, creating hundreds of new jobs without any adverse effects whatsoever.)

It is possible to make independent estimates of how many people would really work at a proposed retail development such as the Kirkstall Valley. Our original 1993 calculations have been set out in detail below to help others challenge inflated jobs estimates often quoted in support of development schemes. To skip around these calculations, click here.

One method is to assume that the new development would reflect the existing trading pattern in the locality. No distinction is made here between comparison goods (furniture, fridges etc) and convenience goods (food and small household items) which have very different staffing, turnover and profit margins. We merely assume that the range of goods on sale would be broadly typical of Leeds as a whole. It is important to distinguish between the gross floor area of the shops, and the net sales area (between 60% and 80% of the gross) after subtracting storage, administrative and welfare accomodation. The gross retail area in the1993 Kirkstall Valley proposals was 225,000 sq ft. The total Leeds retail sales area could be estimated in three ways:

a) Until 1985 the DoE published a regular analysis of commercial floorspace, based on rating valuation officers' returns. The last set of figures show 8000 establishments with a gross area of 1,089,400 sq m or 11,743,732 sq ft, increasing by 0.7% per year. This total includes shops with living accommodation and restaurants.

b) A 1987 Leeds Shopping Study by Bernard Thorpe & Partners, which was commissioned by Leeds City Council, revealed a net area 5,215,501 sq ft in major retail centres. It is likely that many smaller outlets (and at least one big store) were missed in this survey, which did not include restaurants.

c) A study by the Building Services Research & Information Association (BSRIA) which was published in January 1992 contains an estimate of 8,700,000 sq ft net for Leeds, excluding restaurants. This figure may include some non-sales areas, but seems consistent with the earlier DoE statistics and is just below the national average value of 13.6 sq ft net sales area per capita.

The 1991 analysis of the September 1989 census data shows a total retail workforce of 31874 in the Leeds area, including motor vehicle sales. This included 12387 part time workers (mostly women) some of whom were very badly paid. The most recent census is still being analysed, but the situation has not changed greatly.

In answer to a written parliamentary question from John Battle M.P. (Leeds West) on 8 June 1993, Michael Forsyth M.P. indicated that full time employees in retail distribution averaged 39.9 hours and earned £234.4 per week gross but part time workers averaged 18.2 hours and earned only £68.60 per week. Thus the 12387 part-timers equate to 5650 full time equivalents (fte) on an hours basis, but only 3625 fte on a wages basis.

All of the above figures are based on categories 64/65 from the Standard Industrial Classification. The earnings data do not include part-time or casual workers below the PAYE threshold, and therefore over-estimate the part-time to full-time conversion factor. Thus the Leeds retail workforce averages at most 2.06 fte per 1000 sq ft gross, based on hours, or 1.89 fte per 1000 sq ft gross on a wages basis (based on the DoE gross floor space figures, projected forward to 1991, since the restaurants make a relatively minor contribution).

If the Kirkstall development proposed in 1993 performed at the average Leeds levels, then the expected retail employment would have been approximately 350 fte, less than half the developer's estimate. This does not allow for the inevitable job losses at competing retail establishments.

The developer's job creation estimates are also inconsistent with the retail impact study by DTZ Debenham Thorpe which was submitted in support of the planning application. There are many problems with this retail impact study, and Leeds City Council officers have challenged both its methods and its conclusions. However, it contains the developer's own figures, and even these do not support his job creation claims. For example, tables 8, 9 & 10 purport to show that, as a result of anticipated growth, the 10 minute isochrone could support (by 1996) some 38,000 sq ft net additional DIY floorspace, trading at £96 per sq ft per year, plus some 20,000 sq ft net additional furniture & carpets trading at £86 per sq ft and some 10,000 sq ft net additional electrical goods, trading at £355 per sq ft. This division reflects the existing pattern of comparison goods shopping in the surrounding area, and assumes that no other major traders open in competition with the Kirkstall centre.

These figures equate to a weighted average turnover of £131 per sq ft per year for mixed comparison goods shopping, at 1990 prices. The planning application itself envisages a 225,000 sq ft gross development, made up of 80,000 sq ft gross convenience goods, 100,000 sq ft gross comparison goods and 45,000 sq ft gross leisure and fast food. The DTZ Debenham Thorpe study estimates that net sales areas are 81% of gross for mixed comparison goods, although the 1987 shopping study for Leeds City Council gave much lower net sales areas, between 60% & 75% of gross, consistent with DoE statistics. However, it must also be admitted the 1987 study gave slightly higher turnovers per sq ft, after allowing for inflation. If we stay with the DTZ Debenham Thorpe study, the comparison goods area of 81,000 sq ft net would generate an annual income of £10.6 million. Assuming a retail profit margin of around 33% this represents a usable income of £3.5 million, after meeting the wholesale cost of the goods. It is not clear how VAT has been treated in the Debenham Thorpe report. Assuming that the operator pays VAT at 17.5% on the majority of comparison goods sales, this would leave tax-paid receipts of perhaps £3 million.

Out of this the operator must pay fixed overheads: loan charges, business rates, heating, lighting & ventilation, advertising and so forth. The developer estimated his construction costs at £10 million in the planning application, plus land acquisition costs of about £8 million, plus (say) £2 million for furniture and fittings. An overall figure of £20 million was repeatedly quoted in the Press, and was not challenged by the developer. With long term interest rates around 10%, annual loan charges should be about £2 million for the entire site. If these are allocated in proportion to gross floor areas then the comparison goods outlets would have find £0.88 million per year for this item alone. They must also pay interest on their current stock. Business rates on similar retail warehouses in Leeds are at least £120,000 so it is difficult to see how the site operator could find more than £2 million for salaries and wages in the comparison goods sector. The 1992 new earnings survey indicates that full time retail employees earn around £234 per week gross, so that a development of this size would support a maximum of 164 fte (full time equivalents) on a generous assessment of the likely operating parameters.

This is half the developer's estimate of 338 fte for the comparison goods sector, and makes no allowance for any job losses at competing retail outlets, nor for any increase in the efficiency of the distribution network expressed as turnover per employee. Moreover, these figures are based on anticipated growth, and are heavily weighted towards the expanding electrical sector. If we simply take the actual floor areas and turnover figures from table 5 in the DTZ Debenham Thorpe report for major retail warehouses trading in the 10 minute isochrone, then the average turnover drops to £103 per sq ft net per year, with a huge reduction in employment prospects. Indeed with a gross turnover of £8.4 million, the anticipated profit is only £2.4 million. Fixed costs are little changed, so on this rather pessimistic basis the comparison goods sector would support only 32 fte!

It is possible to make a similar analysis of the convenience goods figures from table 7 in the DTZ Debenham Thorpe report. The estimated turnover was £18 million, there is little VAT, but profit margins are slimmer and the business rates and energy costs are higher. Annual wages for the developer's estimate of 365 fte would have cost £4.44 million, plus loan charges of £710,000, business rates of £310,000 and energy costs of about £130,000. The developer's staffing estimates would require a supermarket profit margin of 31% of gross turnover merely to fund the above items, without making any allowance for transport, insurance, pilfering, stock wastage, maintenance, advertising and other unavoidable costs. It could not possibly be done.

There is another way to look at the problem. Company balance sheets identify staff costs in relation to gross turnover. For example, in 1991 J. Sainsbury plc reported turnover of £7,813.3 million, with wages & salary costs of £748.4 million for 41,816 full time employees and 67171 part-timers. These figures are entirely consistent with the government statistics mentioned above:

52 weeks * (41816 full-time @ £234.4 + 67171 part-time @ £68.6) = £749.3 million

The agreement is so close it is almost uncanny!

Converting the Sainsbury figures into fte hours leads to the conclusion that one fte requires a turnover of £107,836. On this basis, if the Kirkstall development were to match the UK market leader, a turnover of £18 million would support only 167 full-time equivalent staff. If one converted to full time equivalent wages there would be only 142 fte staff.

This does not support the developer's estimate that 365 full time equivalents would work at the Kirkstall Valley supermarket. The Sainsbury data are not unique, but are typical of the entire industry. The following table was compiled partly from company annual reports, and partly from the 1992 Key Note Report on Supermarkets and Superstores. The fte figures are calculated solely from wages data, because some companies do not reveal their workforce structure.

The calculations above demonstrate by a variety of independent means, and using the developer's own figures as a basis, that the developer's employment calculations were probably exaggerated by a factor of two or three. However, even the reduced totals do not represent net job creation, because they would be obtained at the expense of workers in competing retail outlets, who would lose trade (and therefore staff) to the new development.

This point is recognised implicitly in table 7 of the DTZ Debenham Thorpe report, which purports to show the retail impact in the convenience goods sector. However, this table contained fundamental errors, which cause the true impact to be seriously under-estimated.

This consultant accepted that convenience goods spending was almost static, but then proceeded to add the projected turnover of two new Netto establishments 20 & 21 to the total turnover in 1996, before calculating the retail impact of the Kirkstall Valley Properties scheme. The correct procedure should have first increased the total 1991 turnover by 1.7% from £182.5 million to £185.6 million (apportioned pro rata between the various shops) and then calculated the impacts of the unfinished Netto stores and the new scheme, alone and in combination.

Table 7 allocated £3.9 million trade diversion (22%) to other locations. The effects of this new development on the existing neigbourhood shops in Armley, Bramley, Headingley and Kirkstall were likely to be severe.

Despite the revolution in shopping patterns in Leeds over the last twenty years, and the large number of new shops, the retail workforce has remained more or less constant. The total "head count" has increased because shops are employing more part-time staff, but when these are converted to full time equivalents the changes are small:

The LDC response to this detailed critique of the job creation figures was curious. Although it was entirely feasible to check our calculations for themselves, they decided to employ consultants and commissioned a report from Roger Tym and Partners.

Roger Tym adopted a similar approach to the Kirkstall Valley Campaign (except they did it less thoroughly) and based their conclusions on company annual reports. They only analysed two companies (Asda and Morrisons) but for some inexplicable reason concealed their identities in their finished report. Their conclusions generally agreed with those from KVC, which is unsurprising since they were using the same company reports. Their gross totals were somewhat larger because they inadvertently counted some of the leisure employees twice, and they assumed that the supermarket turnovers would be considerably bigger (£24,370,000 or £27,160,000) than the developer's own estimate of £19,000,000 for the Kirkstall Retail Centre. The last column of the table below shows the Roger Tym figures after correction to remove the double counting and using the supermarket turnover expected by the developer.

It is important to be consistent about turnover. The supermarket may well perform better than the developer predicts, but if that is the case then the retail impact and traffic generation should also be revised to reflect this. It is unacceptable to follow the Leeds Development Corporation, and to use a low turnover for the impact assessments, and a high one for employment figures!

Roger Tym were also asked to report on the net job creation, after allowing for job losses at competing establishments. Their conclusion was 45 full time jobs net, although little evidence was given to support this. It is quite possible that there would be a net loss of jobs, if a large modern store achieved a more "efficient" distribution of produce using fewer employees. Nobody has yet refuted the KVC position that the Kirkstall Valley retail and leisure development would have a negligible effect on unemployment.

Throughout its life the Leeds Development Corporation systematically inflated its job creation statistics. About 4000 new jobs were claimed in the first year, from ongoing projects which could not possibly be attributed to the LDC's activities. Their final total of 8100 is wildly exaggerated, and completely out of line with Leeds City Council and central government statistics for the LDC area. The erroneous data reflects a defective method of calculation employed by consultants Ecotec Ltd, who were commissioned by LDC to collate employment data, but are now themselves in liquidation.

In 1993 a report was published by the National Audit Office (NAO) on The Achievements of the Second and Third Generation Development Corporations. This states on page 12:

• unreliable output data. For example, jobs claimed were often based on estimates used when projects were first appraised rather than actual jobs created. And at Bristol, Leeds and Tyne and Wear jobs were calculated on the basis of density assumptions applied to available floorspace, rather than jobs in place ...

The treatment of the Roger Tym report was entirely consistent with the usual LDC attitude towards accuracy and truthfulness. Their most important conclusion (that the Kirkstall Valley development would have no discernible effect on Leeds employment totals) was, to our knowledge, never formally reported to the LDC board.