The Housing Affordability Crisis as a Generational Externality: A Time-Series Analysis of UK Housing and Income Dynamics (1995–2024)
Atreyi Roy
For the past three decades, the United Kingdom has experienced a relentless crisis regarding affordability of housing, posing severe implications on distribution of wealth, labour mobility and intergenerational divide within British society. In 1995, the median house price in England and Wales was around 3.6 to 4 times the average annual salary, a historically affordable level. However, by 2024, the average home now costs 8-9 times an individual’s average earnings, more than double the 1990s ratio (Bengtsson and Lyons 2015). This thus results in a structural gap in affordability, explained by a plethora of factors working in tandem, such as the tools of the monetary policy, rental market dynamics, supply-side factors, and shifting labour-market conditions.
The article adopts a time-series approach, looking at data from 1995 to 2024 in an effort to analyse the effects of the above factors on the affordability ratio, and hence quantify their contributions to the same.
A formal definition of an externality states it to be an additional cost imposed on a third party as a result of an economic decision they were uninvolved in, resulting in them being cross-sectional. This aligns with Paul Samuelson’s Overlapping Generations (OLG) model, as well as Thomas Piketty’s work on capital accumulation in highlighting how long-run returns on assets outgrow growth in income. In Samuelson’s OLG model, each generation’s decision-making processes rely heavily on their lifecycle needs, such as the working-age accumulating assets while the older generations draw down on them. However, when the price of these assets outpace income growth, an intergenerational gap is formed as older generations enjoy rising wealth whereas the former face steeper barriers to entering the housing market (Weil 2008, 115-34). Piketty’s theory further suggests that when returns on capital exceed the rate at which wages and output grow, wealth becomes especially concentrated within pre-existing owners of assets (Summers 2015).
In the UK, policy has only amplified the above dynamics. For instance, Thatcher’s post-1980s council house sell-offs reduced social housing stock, with the ‘Right to Buy’ scheme privatizing minions of council homes without any equivalent replacements, hence raising rents and increasing the pressure on the market-priced housing stock (Disney and Lao 2017, 51-68). Furthermore, restrictive Green Belt policies constrained developable land, as it limited housing construction around high-demand cities, specifically London. This major supply constraint resulted in demand-side shocks being translated into higher prices (“Green Belt Planning Loopholes” 2023). Alongside this, the sustained low interest rates after the 2008 financial crisis made mortgages cheaper while increasing the present value of long-lived assets such as housing. However, with supply being unresponsive, cheap credit immediately resulted in price inflation, through which existing homeowners gained substantial capital appreciation while the gap for new entrants was widened.
Before estimating the full structural model, it is useful to examine how the affordability ratio moves in isolation with each of the macro-drivers identified in the literature. These simple regressions provide a diagnostic overview: they reveal which variables closely co-move with affordability over time, and which appear only weakly related. This step reflects standard empirical practice, allowing us to observe raw correlations before introducing controls.
For each regression, the dependent variable is the Affordability Ratio, defined as (Median House Price)/(Median Annual Earnings). Each explanatory variable is entered into the model one at a time, producing four separate estimations of the form:
In considering the base rate, we get the equation Affordability(t)=8.2751-0.1202(BaseRate)(t). The coefficient is -0.1202, indicating that a higher base rate results in better affordability. The sign is in line with the expected relationship: tighter monetary policy should cool housing demand and ease price pressures. However, while textbook theory posits base rates as a predominant driver of affordability, the regression generates a p value of 0.292, suggesting that the above relationship is not very statistically significant. This is largely because the post-2008 period is dominated by an unusually long stretch of near-zero interest rates, creating too little variation for the regression to detect meaningful relationships. Moreover, the effect of monetary policy is transmitted indirectly through house prices and incomes rather than directly through the rate itself. Once these channels are included in other regressions, the base rate loses explanatory power, making the empirical results appear inconsistent with the simplified theoretical story. A further reason the regressions may diverge from textbook predictions is the limited sample size. The analysis uses annual UK data from 2005 to 2024, giving only twenty observations per regression. With so few data points, statistical power is inherently low, making it difficult to detect subtle economic relationships even when they exist.
With the supply of housing, quantified through the number of net additional dwellings in the UK, the regression obtained is Affordability(t)= 3.9943+0.00002053(HousingSupply)(t). The coefficient for housing supply obtained is positive, indicating that greater supply of housing results in a higher ratio, hence housing affordability worsens. The p value is less than 0.05, indicating this to be a highly statistically significant relationship. Standard supply-and-demand theory suggests that increasing supply should improve affordability. In contextualizing the housing supply scarcity, according to DLUHC data, the UK has consistently underbuilt since the 1990s, averaging ~155,000 new homes per year, far below the estimated need of 300,000 (Housing Digital 2025). Yet the regression yields the opposite sign. This occurs because UK housing supply is pro-cyclical, as developers build more in economic booms when prices are already rising quickly. As a result, higher supply is observed during periods of worsening affordability, producing a misleading positive correlation. In addition, “net additional dwellings” includes conversions, student accommodation, and high-end developments that do not expand the stock available to first-time buyers. These measurement issues result in housing-supply data that does not capture the true responsiveness of supply to affordability pressures, breaking the aforementioned theory.
In examining renting vs buying dynamics, the private rent index has been taken as the independent variable, producing the regression Affordability(t)= 4.2961+0.0443(RentIndex)(t). The equation, with a statistically significant p value of 0.007, hence explains that higher rents result in a greater affordability ratio, hence indicating worsening affordability. The ONS Private Rental Price Index shows steady rental inflation since 2005 (Office for National Statistics 2025). When rents rise faster than earnings, saving for a deposit becomes harder, trapping younger households in long-term renting. High rents simultaneously suppress savings and push some renters prematurely into buying, adding further demand-side pressure. The regression on the private rent index reveals a strong and theoretically consistent relationship between rental-market pressures and housing affordability, as increases in private rents signal tightening supply conditions, heightened competition for rental units, and broader structural scarcity, all of which spill over into the owner-occupier market.
Lastly, in considering the unemployment rate in the UK, Affordability(t)= 10.8438-0.5080(Unemployment)(t). The regression on the unemployment rate produces a strong and theoretically intuitive relationship between labour-market conditions and housing affordability. The negative and statistically significant coefficient indicates that higher unemployment is associated with improved affordability over the 2005–2024 period. This aligns with macroeconomic theory: when unemployment rises, household income expectations weaken, credit constraints tighten, and effective demand for housing falls, leading to a slowdown in house-price growth relative to earnings. The UK housing market, which is highly sensitive to demand-side fluctuations, reflects these dynamics clearly in downturns such as 2008–2012 and 2020, when affordability temporarily stabilised or improved despite broader economic stress. However, this effect should not be interpreted as structurally beneficial; improvements in affordability under high unemployment represent temporary cyclical corrections rather than long-term solutions, and they occur precisely when households are least able to purchase property, hence underscoring the inherently unstable and pro-cyclical nature of UK housing affordability.
To incorporate broader structural factors such rental pressures, supply limitations, and labour market conditions, we specify the full and final multivariate model, allowing the effect of each variable to be interpreted conditional on all the others:
Affordability(t)=0+β1(HousePrice(t)) +β2(Earnings(t)) +β3(RentIndex(t)) +β4(HousingSupply(t)) +β5(Unemployment(t)) +Ɛ(t)
The full multivariate model provides a broader picture of housing affordability by jointly estimating the effects of house prices, earnings, private rents, housing supply, and unemployment, with the equation being
Affordability(t)=2.3883-0.000009379(HousePrice(t)) +0.0002(Earnings(t)) -0.0724(RentIndex(t)) +0.00002633(HousingSupply(t)) +0.6729(Unemployment(t)).
However, the overall explanatory power is modest (R² = 0.318), and most variables are not individually significant at conventional thresholds, reflecting the complexity of affordability dynamics and potential multicollinearity within the UK housing system. House prices carry a negative but statistically insignificant coefficient (–9.38e-06, p = 0.304), suggesting that once earnings, rents, supply, and labour-market conditions are accounted for, annual variation in average house prices does not independently predict year-to-year changes in the affordability ratio. Earnings display a positive effect (0.0002) and are close to significance (p = 0.093), implying that rising wages modestly improve affordability, though not strongly enough to offset house-price inflation across the sample. The private rent index shows the expected negative sign (–0.0724) and is marginally insignificant (p = 0.079), indicating that upward pressure in the rental sector slightly worsens affordability, consistent with textbook theory on tenure-spillover effects. Housing supply has a small but positive coefficient (2.63e-05, p = 0.058), meaning increased net additions tend to slightly improve affordability, but the magnitude is extremely small, reflecting the UK’s chronic undersupply relative to demand. Finally, unemployment carries a positive coefficient (0.6729, p = 0.070), which at first appears counterintuitive but likely reflects recession-driven house-price slowdowns and suppressed demand during downturns rather than genuine affordability improvements.
Model diagnostics point to important limitations: the condition number (1.5×10⁷) signals strong multicollinearity, especially between housing supply, rent index, and house prices, which inflates standard errors and weakens coefficient significance. The Jarque–Bera test (p = 7.67e-05) and high kurtosis also indicate non-normal residuals, likely arising from structural breaks such as the 2008 financial crisis and the 2020 pandemic housing shock. Moreover, the sample size of only 20 annual observations (2005–2024), may not be indicative to estimate a five-variable model, especially when the underlying relationships evolve slowly and policy shifts occur discretely. Taken together, the multivariate regression suggests that while supply, rents, and earnings show signals consistent with theory, the UK housing market’s structural rigidities, data limitations, and collinearity between variables constrain the statistical clarity of causal effects. The results should therefore be interpreted as indicative rather than definitive, highlighting the need for deeper structural modelling or higher-frequency data to capture the true drivers of affordability.
Taken together, the regression results indicate that structural factors, particularly supply constraints, rental market pressures, and labour-market fragility, play a more consequential role in shaping affordability than short-term monetary adjustments. Accordingly, policy should prioritise expanding effective housing supply by reforming binding planning constraints, accelerating approvals in high-elasticity regions, and scaling investment in social and intermediate housing. Parallel reforms in the rental sector such as strengthening security of tenure, improving rent-setting transparency, and incentivising institutional build-to-rent could dampen rental inflation and ease saving constraints for prospective buyers. Finally, while monetary policy interacts with asset prices, the evidence suggests it is an imprecise tool for housing affordability, reinforcing the need for coordinated fiscal, planning, and labour-market interventions rather than reliance on interest-rate adjustments.
References:
1. Bank of England. 2020. “Bank Rate reduced to 0.1% and asset purchases increased by £200bn - March 2020.” Bank of England. https://www.bankofengland.co.uk/monetary-policy-summary-and-minutes/2020/monetary-policy-summary-for-the-special-monetary-policy-committee-meeting-on-19-march-2020#:~:text=On%20March%2019%2C%202020%2C%20the%20Bank%20of,%C2%A3645%20billion%20*%20Enlarge%20t.
2. Bengtsson, Helena, and Kate Lyons. 2015. “Revealed: the widening gulf between salaries and house prices.” The Guardian. https://www.theguardian.com/uk-news/2015/sep/02/housing-market-gulf-salaries-house-prices.
3. Disney, Richard, and Guannan Lao. 2017. “The Right to Buy public housing in Britain: A welfare analysis.” Journal of Housing Economics 35 (March): 51-68. https://doi.org/10.1016/j.jhe.2017.01.005.
4. “Green Belt Planning Loopholes.” 2023. Clear Architects. https://cleararchitects.co.uk/green-belt-planning-loopholes.
5. Housing Digital. 2025. “The UK housing crisis: Causes, impact, and solutions.” Housing Digital. https://housingdigital.co.uk/the-uk-housing-crisis-causes-impact-and-solutions/.
6. Office for National Statistics. 2025. “ONS, Private rent and house prices, UK, July 2025.” Private rent and house prices, UK - Office for National Statistics. https://www.ons.gov.uk/economy/inflationandpriceindices/bulletins/privaterentandhousepricesuk/july2025.
7. Summers, Lawrence H. 2015. “Piketty and the inequality puzzle.” Institute of Economic Affairs. https://iea.org.uk/blog/piketty-and-the-inequality-puzzle/.
8. Weil, Philippe. 2008. “Overlapping Generations: The First Jubilee.” Journal of Economic Perspectives 22 (4): 115-134.
