Data on house prices for two/three-bedroom dwellings in the fourth quarter of 2003 is compared with gross household earned incomes to produce robust estimates of affordability across England. Some of the main findings and conclusions:
- the average ratio of house price to household gross earned income ratio in England was 4.11 to 1;
- affordability issues now affect the South West region (4.66 to 1) and not just London (4.69 to 1) and the South East (4.61 to 1);
- on this basis the decision to concentrate new affordable housing on London and the 'wider south east' should be reviewed;
- there are also affordability concerns in smaller rural districts across England, raising questions about the adequacy of government programmes designed to promote affordable housing in rural areas.
This study provides a range of analyses of the difficulties working households aged 20 to 39 face in accessing home ownership in every area of England at the end of 2003. It is based on local prices for two- and three-bedroom dwellings, and draws on a range of data sources. The new analyses, by Steve Wilcox at the University of York, demonstrate that affordability issues now affect the South West region and not just London and the South East; these analyses update the 2002 analysis.
- Access to home ownership is most problematic in London, where the ratio of average house price to household gross earned income is 4.69 to 1. Ratios exceed 5.0 to 1 in nine areas; in Westminster and Brent the ratio exceeds 6.0 to 1. Ratios are lower than 4.0 to 1 in five areas. All ratios should be considered in relation to the ratio of mortgage advance to household gross earned income, which rarely exceeds 3.5 to 1.
- Outside London, access to home ownership is most problematic in the South West region, where average house price to household gross earned income ratios are 4.66 to 1. Average ratios exceed 5 to 1 in 21 areas; the ratio exceeds 6 to 1 in West Somerset, Christchurch, North Cornwall, Torridge and East Devon, though samples in the first two are particularly small. Ratios are lower than 4 to 1 in five areas.
- Access to home ownership is nearly as problematic in the South East region, where average ratios are 4.61 to 1. They exceed 5 to 1 in 17 areas, including all of East Sussex. While it is only in Chichester that the ratio exceeds 6 to 1, ratios are lower than 4 to 1 in four areas.
- Average ratios exceed 5 to 1 in four areas in the East of England and in two northerly 'hotspots' (Derbyshire Dales and South Lakeland).
- Over the past year, affordability issues in the South West have worsened relative to London and the South East, due to greater house price rises. This reinforces the case for the Office of the Deputy Prime Minister (ODPM) to review its current plans to concentrate increased funding for affordable housing in London and the wider south east.
This analysis provides an update to one element of the affordability analyses published by the Foundation last year in the first 'Can work: can't buy' report.
It is based on house prices for two-/three-bedroom dwellings in each area in the fourth quarter of 2003, derived from the Survey of Mortgage Lenders. This data provides a consistent measure of house prices for a modest size family dwelling in every local authority area in England. (It is preferred to Land Registry house price data, as that is based on a variable size mix of dwellings in each area, and cannot therefore provide a standardised measure of the difficulty of accessing owner-occupation as between one area and another.)
The analysis is confined to working households with a representative aged between 20 and 39. The great majority of first-time home-buying households are drawn from this age group.
House price to household income ratios
High ratios of price to earned income generally make homes unaffordable for first-time buyers. This analysis is based on gross household earned incomes; not the incomes of individuals. This is because it is households that buy houses, and in the majority of cases this involves households with two working adults.
Table 1 shows that there are marked regional differences in the relationship between individual and household earned incomes; these are overlooked when ratios are constructed solely on levels of individual earnings.
|north east||north west||Yorks and the humber||east midlands||west midlands||east||London||south east||south west||England|
|Numbers of workers in working households (000's)|
|Annual earnings at 2003 levels|
|Ratio of earnings between multiple and single earner households|
Gross earned income data, at the regional level, is taken from the Expenditure and Food Survey (EFS). Local income data is only available for individuals, and for this analysis it has been derived from Inland Revenue Survey (IRS) data. It is combined with Labour Force Survey (LFS) data, showing the numbers of working households with single or multiple adult earners, in order to compute local household gross earnings figures consistent with the regional earnings figures taken direct from EFS. The earnings figures are all then uprated to 2003 levels. A detailed note on data sources and methodologies is set out in the Appendices at the end of this document.
The average ratio of house price to household gross earned income ratio in England was 4.11 to 1, while regional average ratios range from 4.69 to 1 in London to 3.19 to 1 in the North East (see Table 2).
|Region||Numbers of working households||Household earned income||mean house prices||house price to income ratios|
|East of England||572,729||£37,353||£159,393||4.27|
|Yorkshire & Humber||511,539||£28,957||£99,642||3.44|
It is of particular note that house price to income ratios are now higher in the South West region than in the South East. The average ratio in the South West is 4.66 to 1, while in the South East it is 4.61 to 1. This is despite the average house price in the South East (£183,175) being 21.2% higher than those in the South West (£151,147). However, the average income of working households in the South East (£39,734) is 22.5% higher than those in the South West (£32,444).
The forty individual areas with the highest ratios are ranked in Table 3. In nine areas the ratios are over 6.0 to 1. Altogether, in fifty-four areas the ratio exceeds 5.0 to 1. The ratios for every local authority area in England are listed online in Appendix 1, by region.
|Local authority||Numbers of working households||household earned income||mean house prices||house price to income ratios|
|1. West Somerset (SW)*✝☦||1,802||£24,165||£164,245||6.80|
|2. Christchurch (SW)✝||1,983||£28,594||£191,945||6.71|
|3. North Cornwall (SW)||5,487||£25,161||£164,529||6.54|
|4. Torridge (SW)*||4,782||£24,142||£157,209||6.51|
|5. Westminster (Lon)||23,154||£67,387||£415,558||6.17|
|6. Chichester (SE)||8,616||£38,110||£233,324||6.12|
|7. Adur (SE)||5,303||£28,198||£172,312||6.11|
|8. East Devon (SW)||8,824||£30,013||£180,675||6.02|
|9. Brent (Lon)||20,768||£37,870||£227,851||6.02|
|10. Lewes (SE)||6,315||£30,776||£180,475||5.86|
|11. Purbeck (SW)*✝||2,739||£30,134||£176,662||5.86|
|12. Eastbourne (SE)||7,536||£26,988||£157,711||5.84|
|13. Torbay UA (SW)||12,517||£25,140||£145,954||5.81|
|14. South Lakeland (NW)||10,042||£27,573||£159,641||5.79|
|15. Mole Valley (SE)||5.793||£43,140||£247,102||5.73|
|16. Bournemouth UA (SW)||21,218||£29,298||£166,733||5.69|
|17. Kerrier (SW)||6,203||£26,111||£148,362||5.68|
|18. Penwith (SW)||5,661||£28,769||£163,040||5.67|
|19. Carrick (SW)||8,093||£29,159||£164,983||5.66|
|20. Brighton and Have UA (SE)||32,149||£34,221||£191,860||5.61|
|21. Arun (SE)||10,495||£32,081||£179,377||5.59|
|22. Cambridge (E)||12,252||£34,502||£191,396||5.55|
|23. West Devon (SW)*✝||3,221||£30,903||£167,697||5.43|
|24. North Devon (SW)||6,877||£27,356||£148,390||5.42|
|25. Waltham Forest (Lon)||24,639||£34,756||£188,359||5.42|
|26. Southwark (Lon)||36,122||£42,431||£229,490||5.41|
|27. Oxford (SE)||12,759||£39,712||£214,275||5.40|
|28. South Hams (SW)||5,963||£32,239||£173,529||5.38|
|29. East Dorset (SW)||5,536||£37,978||£203,191||5.35|
|30. Ealing (Lon)||25,544||£45,681||£244,075||5.34|
|31. Hertsmere (E)||12,783||£44,559||£237,901||5.34|
|32. Newham (Lon)||28,394||£32,252||£170,967||5.30|
|33. Rother (SE)||7,926||£34,175||£181,155||5.30|
|34. Restormel (SW)||9,083||£27,318||£144,474||5.29|
|35. Islington (Lon)||25,414||£54,579||£288,114||5.28|
|36. Epping Forest (E)||11,403||£45,579||£236,421||5.25|
|37. Derbyshire Dales (E Mid)*||4,957||£32,365||£169,722||5.24|
|38. Weymouth and Portland (SW)||5,560||£29,449||£153,635||5.22|
|39. Salisbury (SW)||11,763||£32,574||£168,680||5.18|
|40. Weaden (SE)||10,243||£37,242||£192,679||5.17|
* House price data samples fewer than 50
✝ Households and employment data samples fewer than 30
☦ Income data samples fewer than 50
At first sight, these ratios may not appear to be exceptionally high, given the more relaxed attitudes of lenders to mortgage advance ratios in the context of interest rates that are at historically low levels. However, it should be borne in mind that these ratios are based on household earnings, not the earnings of one individual.
The results for the areas where the income, employment and house price figures are based on small samples must be treated with caution (see Table 3 here and Appendix 1 in the online version). This applies, in particular, to the three areas where the house price, employment and income figures are all based on small samples (Berwick-upon-Tweed, Teesdale and West Somerset).
Analysis of mortgage advances in 2002 shows that it is only rarely that the ratio of mortgage advance to income exceeds 4.0 to 1 when advances are based on a single earned income. Ratios rarely exceed 3.5 to 1 when based on multiple earned incomes. In both cases, these levels of ratios only occurred for just over 10% of the advances to first-time buyers in the three most expensive regions in the country. They were even less likely to be found in other parts of the country.
Implications of the analysis
These findings contrast with the assertion in the ODPM's 2003 'Sustainable Communities' plan, that affordability issues are more acute in the South East region than the South West. There are three main reasons for these different results:
- The ODPM analysis was based on 2001 house price data, and did not take account of the more rapid rise in house prices in the South West in subsequent years.
- The ODPM analysis is derived from earnings data based on an individual's place of work. Place of work data fails to take account of the impact of the predominantly high earners who commute into London from parts of the greater south east. It thus overstates the average earned incomes of individuals residing in London, and at the same time understates the earned incomes of individuals living in those parts of the South East.
- The ODPM analysis is also based on the earned incomes of individuals rather than households, and thus fails to take into account the marked regional differences in the relationship between individual and household earnings (see Table 1).
The new 2003 analysis suggests that ODPM should review its previous policy decision to focus the growth in the provision of new affordable housing on London and the 'wider south east'.
The analysis also highlights the extent of affordability concerns in smaller rural districts, not just in parts of the south of England, but also in popular tourist and second home areas such as South Lakeland and Derbyshire Dales. This raises questions about the adequacy of government programmes to support the provision of affordable housing in rural areas.
About the project
The analysis was undertaken by Professor Steve Wilcox of the Centre for Housing Policy, University of York. It draws on Survey of Mortgage Lenders house price data, and data from the Expenditure and Food Survey, Labour Force Survey and Inland Revenue Survey to compute local household incomes.
This analysis is the first step in a wider study, which will define and identify the scope of 'intermediate housing markets' in each local authority area in England.
How to get further information
Details of house price to gross household income ratios, by local authority, county and region and of the data sources and the methodology applied in the construction of the 2003 house price to income ratios are available in the appendices below.
Appendices: Affordability differences by area for working households buying their homes - 2003 update
|Local authority||number of working households||household earned income||mean house prices||house price to income ratios|
|North East Derbyshire||10,745||£29,274||£102,909||3.52|
|Hinckley & Boswoth||9,352||£37,954||£119,932||3.16|
|North West Leicestershire||8,707||£39,634||£117,080||2.95|
|Oadby & Wigston||5,462||£32,895||£118,607||3.61|
|Newark and Sherwood||12,132||£36,715||£112,269||3.06|
|East of England||572,729||£37,353||£159,393||4.27|
|Kings Lynn and West Norfolk||11,099||£29,305||£115,065||3.93|
|Barking and Dagenham||16,871||£38,001||£150,615||3.96|
|City of London||1,219||£154,818||£354,295||2.29|
|Hammersmith and Fulham||22,950||£62,093||£307,796||4.96|
|Kensington and Chelsea||18,037||£120,209||£537,113||4.47|
|Kingston upon Thames||17,644||£51,425||£237,588||4.62|
|Richmond upon Thames||20,561||£74,923||£281,178||3.75|
|Redcar and Cleveland UA||12,149||£28,548||£85,957||3.01|
|Tyne and Wear|
|Newcastle Upon Tyne||25,405||£31.270||£113,725||3.64|
|Crewe and Nantwich||11,720||£35,974||£105,104||2.92|
|Ellesmere Port & Neston||8,912||£35,407||£100,960||2.85|
|Bracknell Forest UA||102,63||£42,097||£178,969||4.25|
|West Berkshire UA||15,009||£45,230||£198,057||4.38|
|Windsor and Maidenhead UA||10,559||£59,154||£244,180||4.13|
|Milton Keynes UA||27,411||£38,085||£146,155||3.84|
|Brighton and Hove UA||32,149||£34,221||£191,860||5.61|
|Basingstoke and Deane||21,033||£42,586||£183,839||4.32|
|Isle of Wight UA||12,377||£27,355||£140,630||5.14|
|Medway Towns UA||27,203||£33,325||£136,607||4.10|
|Tonbridge and Malling||11,053||£42,176||£178,325||4.23|
|Vale of White Horse||11,582||£42,360||£192,246||4.54|
|Epsom and Ewell||6,982||£45,375||£233,212||5.14|
|Reigate and Banstead||13,393||£47,245||£224,304||4.75|
|Bath and North East Somerset UA||17,401||£36,660||£184,631||5.11|
|North Somerset UA||20,613||£36,034||£156,093||4.33|
|South Gloucestershire UA||26,272||£35,221||£146,831||4.17|
|Weymouth and Portland||5,560||£29,449||£153,635||5.22|
|Forest of Dean||8,572||£29,469||£136,183||4.62|
|Hereford and Worcestershire|
|Shrewsbury and Atcham||12,549||£29,702||£139,727||4.70|
|Telford and the Wrekin UA||20,405||£31,157||£101,493||3.26|
|Nuneaton & Bedworth||10,888||£31,836||£101,336||3.18|
|Yorkshire and Humber||511,539||£28,957||£99,642||3.44|
|East Riding of Yorkshire UA||28,442||£30,263||£110,629||3.66|
|Kingston Upon Hull UA||27,020||£25,746||£63,563||2.47|
|North East Lincolnshire UA||15,873||£23,203||£78,303||3.37|
|North Lincolnshire UA||15,757||£28,031||£90,652||3.23|
* House price data samples fewer than 50
✝Households and employment data samples fewer than 30
☦ Income data samples fewer than 50
Appendix 2 Data sources and methodology for the construction of the 2003 house price to income ratios
This note provides a brief outline to the data sources and the methodology applied in the construction of the 2003 house price to income ratios. The ratios are based on gross household earned incomes and the prices of two- and three-bedroom dwellings.
However, the results for the areas where the income, employment and house price figures are based on small samples must be treated with caution. This applies, in particular, to the three areas where the house price, employment and income figures are all based on small samples (Berwick-upon-Tweed, Teesdale and West Somerset) and in the areas with the lowest number of working households in the country (West Somerset, Christchurch, Teesdale and Berwick-upon-Tweed).
Two of these areas, West Somerset and Christchurch, were found to have the highest ratio of house price to household gross earned income in the country, so the figures should be treated with particular caution.
The house prices are lower quartile prices and mean average prices for two- and three-bedroom dwellings, taken from the Survey of Mortgage Lenders. The data, for the fourth quarter of 2003, has been weighted on the basis of an even mix of both sizes of dwelling in every local area. This provides a consistent measure of prices for modestly sized family dwellings. In most cases, the Survey of Mortgage Lenders provides a substantial sample at the local level. However, in a few cases, the sample for two- and three-bedroom dwellings fell below fifty. The areas concerned (with sample sizes in brackets) are: Teesdale (20), Berwick-upon-Tweed (27), South Shropshire (27), Ryedale (28), West Somerset (30), Castle Morpeth (31), Eden (31), Alnwick (34), Craven (37), Rutland (38), Purbeck (43), Torridge (45), West Devon (47) and Derbyshire Dales (48).
The analyses use local income data for 2000/1 and 2001/02 from the Inland Revenue Survey (IRS). This covers the earned incomes of all individuals aged between 20 and 39 inclusive. This age range has been chosen because the great majority of first-time buyers fall within it. The 2000/01 figures were inflated by 5.2% to 2001/02 values to derive average local earned incomes from the combined years of data. The combined average figures were then inflated by 8.0% to 2003 values. In all but three areas, the IRS local data was based on a local sample of more than 50 cases for the two years combined. The three small samples were Teesdale (37), Berwick-upon-Tweed (43) and West Somerset (47).
Regional household incomes
Robust data on household incomes is not available at the local authority level. Regional data on gross household earned incomes, analysed by the numbers of adult workers in each household, has been obtained from the Expenditure and Food Survey (EFS) for households with representatives aged between 20 and 39. The data is based on three years of the survey, from 2000/01 to 2002/03, in order to ensure a robust regional sample. For the affordability analysis, this data has been uprated to 2003 levels (by 8.0%). The regional EFS gross household earned income figures provide control totals for the local level estimates of household incomes within each region.
Households and employment
With its enlarged sample size, the Labour Force Survey (LFS) can now provide data on the employment status of households at the local level. LFS data for 2001, 2002 and 2003 has been averaged to show the numbers of households in each local authority area, broken down into categories based on the numbers of people in each household in employment. The data shows the numbers of households with nil, one or two or more workers aged 20 to 39. However, even using three years' data, the samples for some areas are very small. The data for twelve areas are based on less than 30 cases: Christchurch (14), Berwick-upon-Tweed (16), West Somerset (18), Teesdale (19), Purbeck (24), Rutland (25), Castle Morpeth (25), South Shropshire (26), West Devon (27), Alnwick (27), Oswestry (28), Melton (29).
Local household incomes
Within each local authority area, it is assumed that the relationship between the earned incomes of one earner and multi-earner households corresponds with the regional relationship shown by the Expenditure and Food Survey (EFS) data. Factors are then applied to the local IRS data so that the regional sums of the computed local household earnings, based on the IRS and LFS data, are consistent with the regional EFS figures.