06How big is Britain’s backlog of missing homes?

The scale of underbuilding in Britain since the Second World War means the UK should have many more homes than it does today. Compared to other European countries, Britain has accumulated a large backlog of missing homes that need to be built if the UK is to end its housing crisis. Housebuilding needs to rise to address both annual need and clear this unbuilt backlog.

Britain has a backlog of millions of unbuilt homes

It is possible to produce reasonable estimates of how many homes the UK would have added had it seen its housing stock grow at a similar rate to other European countries in every year from 1955 to 2015, as Table 3 shows. The first row contains the number of homes actually built in the UK by tenure and number of homes per thousand people in 2015. Each row after can be thought of as the number of extra homes the UK would have if it had pursued the policy approach of each country, adjusted for the UK’s high initial ratio of homes per person as well as slower population growth.

Had the UK built houses at the rate of the average Western European country from 1955 to 2015, it would have added a further 4.3 million homes than it actually did – resulting in 15 per cent more homes than the 28.3 million dwellings that actually did exist in 2015. As the UK increased the size of its housing stock by 12.2 million homes from 1955 to 2015, these 4.3 million extra homes would have required new additions to be 35 per cent higher across the entire period.

Table 3: The missing homes from underbuilding from 1955-2015, by Tenure

Country Increase upon the UK’s housing stock in 2015 (%) Total increase in homes, 1955 to 2015 Private sector contribution, 1955 to 2015 Public sector contribution, 1955 to 2015 Total private : Public mix of gross housebuilding, 1955 to 2015 Homes per thousand people – 2015
United Kingdom 0 12,230,000 7,875,000 4,358,000 64 : 36 425
Increase to the UK’s added homes from 1955 to 2015 if the UK had built at the rate of…
Switzerland 6.0 1,647,000 6,005,000 -4,358,000 94 : 6 450
Sweden 7.7    2,137,000 2,054,000 82,300 68 : 32 457
Denmark 8.8 2,445,000 3,910,000 -1,465,000 77 : 23 462
Belgium 10.1 2,795,000 7,153,000 -4,358,000 96 : 4 468
Netherlands 10.3 2,836,000 833,800 2,002,000 59 : 41 468
Norway 12.1 3,349,000 5,821,000 -2,472,000 84 : 16 476
Germany 13.9 3,835,000 6,367,000 -2,533,000 85 : 15 484
France 19.5 5,393,000 6,270,000 -877,700 78 : 22 507
Austria 25.3 7,007,000 4,850,000 2,157,000 66 : 34 532
Ireland 25.6 7,076,000 8,888,000 -1,812,000 84 : 16 533
Finland 29.9 8,276,000 12,300,000 -4,019,000 94 : 6 552
Western European Average 15.4  4,254,000 5,859,255 -1,604,855 80 : 20 490

Source: Centre for Cities calculations. United Nations, Bulletin of Housing and ConstructionFrom 2000 Housing stock values are taken from respective national statistical agencies which are listed in the appendix.

Subsequent columns show how net additions by tenure would have changed had the UK adopted the approach of each country, the tenure mix, and the predicted number of homes per person. Building to the European average after 1955 would have improved housing outcomes in 2015 by increasing the ratio of homes per thousand people from 425 homes to 490, with a fifth of all the homes built being social housing.

For some counterfactuals the number of missing homes is even higher. Had the UK added homes at a similar rate to Finland, it would have added an extra 8.3 million homes from 1955 to 2015, a 30 per cent increase in the 2015 dwelling stock. Even the less impressive counterfactuals still point towards a deep shortage of homes. Even if the UK had adopted Denmark’s relatively subdued approach to housebuilding, it would still have an additional 2.5 million more homes than it does today, a 9 per cent increase in total dwelling stock in 2015.

Box 8 describes how these counterfactual levels of British housebuilding and missing homes were calculated.

Box 8: Calculating the backlog of missing homes

The backlog of “missing homes” that would have been built had the UK built has many as other countries, and the number of homes that the UK needs today to improve its housing outcomes are not necessarily the same number. The former can though indicate how large the latter is and provides a range of targets for future housing policy in the present to aim for.

Creating estimates for how many homes the UK would have built if it had adopted the policies of a different country must adjust for different populations, different levels of population growth, different initial numbers of homes per person, different levels of demolitions, different tenure mixes between countries, and how these factors affect each other.

As other European countries tended to have faster population growth than the UK and had fewer homes per person in 1955, comparisons based solely on gross housebuilding produce estimates for the missing number of homes that are too high, as other countries had greater need than the UK.

Comparisons based solely upon changes in the number of homes per person in 2015 produce more appropriate estimates but omit tenure and are inconsistent due to differences in the definition of primary dwellings in national statistics (e.g. some countries count vacant and second properties, others do not).*

The modelling in this section adopts an alternative approach. For each European country, a counterfactual UK is generated that is adjusted for the differences between the UK and that country in population growth and initial numbers of homes per person, and the changes to the rate of UK demolitions this implies. This counterfactual UK then adds homes every year from 1955 to 2015 for both tenures at the same rates as its comparator country.

After controlling for the differences in population growth and initial housing stock, the difference in cumulative rates between counterfactual UK and the actual UK is then applied to the actual UK housing stock from 1955, and produces the estimates in this chapter. More details on the technical aspects of the modelling are in the technical annex.

*This latter discrepancy explains why the UK’s housing outcomes had at least 7.8 per cent fewer houses per person than the European average in 2015, but the 4.3 million houses estimate of the backlog would increase the UK’s housing stock by 15 per cent. The 5pp gap between these occurs because the statistical offices of Austria, Belgium, and Norway in 2015 only count primary residences, and other European countries have different standards for habitable dwellings to the UK. In other words – the 8 per cent deficit does not account for the poor quality and inequality problems of housing in the UK, while a 15 per cent increase in stock would improve the availability, distribution, and quality of housing services to European levels.

The 4.3 million missing homes estimate is fundamentally conservative. It omits any underbuilding that occurred from 1947 to 1955 and from 2015 to 2023, and returns the UK to the European average, not the above-average housing outcomes Britain enjoyed in 1955. It also omits the geography of where these homes should be built. The lion’s share of these new homes would likely need to be concentrated in the most expensive parts of the country – London and the South East – to improve outcomes as much as they can.

The backlog is comparable to other estimates of unbuilt housing in the UK – 2.9 million in England in 2009 according to the then-Department of Communities and Local Government; 3.1 million social homes in England in 2019 according to Shelter; and 3.9 million in 2015 in Britain published by Crisis.57

As the backlog is calculated cumulatively year-by-year, its emergence can be dated. For the average European scenario, 29 per cent of the missing backlog – roughly 1.2 million homes – was accumulated during the post-war era between 1955 to 1979 (40 per cent of years of the entire 1955 to 2015 period).  This indicates that while the period since 1980 has seen particularly low housebuilding, British underbuilding predates it.

Britain could have built more public and private housing

As private housebuilding in the UK was so low, the UK had unusually high share of public housebuilding in its tenure mix.58 Most counterfactuals therefore show an increase in the total number of homes, but a decrease in the total and share of new social housing provided.

Table 3 shows that had the UK adopted a policy approach similar to the average European country, it would have built 4.3 million more homes from 1955 to 2015 to keep up with European housing outcomes. As amounts to 5.9 million more homes built by the private sector, and 1.6 million fewer homes built by the public sector. Accordingly, the tenure mix of new supply would have changed from 64:36 private:public to 80:20.

As the most important factor behind the UK’s unbuilt backlog is a low rate of private housebuilding, every single modelled scenario sees the private sector build more houses. However, there are a few countries that indicate the UK could have built more private and more social housing. Had the UK taken the policy approach of the Netherlands or Austria, the British public sector would have built between 2 to 2.2 million social homes beyond our actual social housing programme, alongside an additional 2.8 to 7 million new private sector homes.

England needs between 442,000 and 654,000 new homes every year to clear the backlog

Housebuilding in the UK needs to be high enough to both meet normal annual demand and make progress in clearing the backlog of 4.3 million unbuilt homes. Taking England’s share of this backlog – between 83.2-83.4 per cent of all the dwellings in the UK are in England – it is possible to estimate what this means for English housebuilding today.59

At present, England has a housebuilding target of 300,000 a year, which implies a rate of 1.3 per cent per year. If Europe continues building at its current rate of 0.9 per cent, England building 300,000 homes a year will take 65 years to close the housing backlog.60

Assuming there is a desire to end the housing crisis sooner than that, Table 4 shows the housing targets and housebuilding rates that England requires to clear its share of the unbuilt backlog from each counterfactual from the previous Table 3 within ten and twenty-five years. If the Government’s target of 300,000 new homes is taken to be an accurate estimate of annual and potential future need in England every year for the next two and a half decades, housebuilding above that number can be dedicated solely to clearing the backlog of missing homes from each counterfactual.

Ending the housing crisis in the next twenty-five years would require England to add 442,000 homes every year, double the current housebuilding rate of 220,000 a year, as shown in Table 4. Solving it in ten years, or two parliamentary terms, would require 654,000 new homes a year in England. Achieving housing outcomes of countries with above-average records, such as Finland and Austria, would require even higher amounts of housebuilding over the same periods.

Table 4: England needs to build more than 300,000 homes a year to end the housing crisis within the next few decades

Missing homes backlog to overcome 300,000 + backlog in 10 years Initial and final building rates over 10 years 300,000 + backlog in 25 years Initial and final building rates over 25 years
300,000 Target 300,000 1.3 – 1.1 300,000 1.3 – 1
Switzerland 437,000 1.9 – 1.6 355,000 1.5 – 1.1
Sweden 478,000 2.0 – 1.7 371,000 1.6 – 1.1
Denmark 504,000 2.1 – 1.8 381,000 1.6 – 1.2
Belgium 533,000 2.3 – 1.9 393,000 1.7 – 1.2
Netherlands 536,000 2.3 – 1.9 394,000 1.7 – 1.2
Norway 579,000 2.5 – 2 412,000 1.75 – 1.2
Germany 619,000 2.6 – 2.1 428,000 1.8 – 1.3
France 749,000 3.2- 2.5 480,000 2 – 1.4
Austria 884,000 3.75 – 2.8 533,000 2.3 – 1.5
Ireland 889,000 3.8 – 2.8 536,000 2.3 – 1.5
Finland 989,000 4.2 – 3.05 576,000 2.4 – 1.5
Western European Average 654,000 2.8 – 2.2 442,000 1.9 – 1.3

Source: Centre for Cities calculations. United Nations, Bulletin of Housing and ConstructionFrom 2000 Housing stock values are taken from respective national statistical agencies which are listed in the appendix.

The housebuilding rate that the ten-year goal requires is between 2.8 and 2.2 per cent growth in English housing stock per year to reach the European average. This is comparable to the interwar average of 2.4 per cent growth in the number of homes per year, the historic peak of English and Welsh housebuilding. The housebuilding rate required of the twenty-five-year goal is 1.9 to 1.3 per cent, which is below the post-war average rate of 1.9 per cent from 1947 to 1979 but above the entire post-1947 average of 1.3 per cent.61

Since 2015, England has added between 220,000 to 240,000 new homes a year – roughly 0.9 per cent growth per year. As the European average housebuilding rate today is also around 0.9 per cent, the current English housebuilding rate happens to be minimum number of new homes required to stop Britain’s current housing outcomes from getting worse relative to the rest of Europe. If the English housebuilding rate falls below 234,000 (or more accurately, 0.9 per cent a year) then housing outcomes in the UK will deteriorate compared to European countries.

Addressing the backlog and building the missing 4.3 million homes is essential if the UK is to achieve a similar standard of housing to other European countries. The next and final section sets out the conclusions for understanding the housing crisis and what needs to change to end it.



  • 57 Bramley, G. (2010), Estimating housing need, ‘Communities and Local Government’; Shelter (2019), ‘Quantifying the need for social housing’; Bramley, G. (2019), ‘Housing supply requirements across Great Britain for low-income households and homeless people: Research for Crisis and the National Housing Federation; Main Technical Report’ (2019).
  • 58 The model produced small negative results for Belgium and Switzerland that implied these policy approaches would have resulted in the UK demolishing 421,000 and 677,000 social homes respectively from 1955 to 2015. These results stem not from these countries actually demolishing social housing (no evidence of this occurring at scale in reality could be found, and Switzerland has high quality demolition data), but from how little social housing the two countries did build and the limitations of the modelling method when it adjusts for population growth. As a result, these UK counterfactuals have been adjusted upwards by the authors to result in zero net new social housing from 1955 to 2015.
  • 59 Neither Scotland nor Wales have housing targets as clear as England’s 300,000 figure, making it hard to benchmark for UK-wide future annual need. Northern Ireland has a housing target that amounts to 100,000 new homes from 2022 to 2037, roughly 6,666 a year – see Department for Communities (2021) Housing Supply Strategy – but separate Northern Irish annual housing stock data does not exist in the postwar period.
  • 60 Were England to build at the rate that 300,000 currently implies (1.27 per cent) and therefore experience compounding in absolute numbers, it would take 51 years.
  • 61 The housebuilding rate declines over the period as a fixed housebuilding target results in it slowly becomes a relatively smaller share of a growing stock. In contrast, compounding means a fixed housebuilding rate requires a larger number of homes to be built every year.