If the UK built 1 million homes, what would happen to house prices?

Alastair Parvin
15 min readJun 6, 2021

A thought experiment

A few years ago, a paper by the Government Office for Science brilliantly described the UK housing debate as one “bedevilled by rival simplifications”. Of no aspect is this more true than the ‘supply and demand’ debate. The question at its core is: are inflated house prices the result of too little supply, and therefore could they be reversed by simply building more homes?

There is actually an official answer to this question, and that answer is sort of. Economic analysis by the Ministry for Housing, Communities and Local Government (MHCLG) in 2018 concluded that a 1% increase in the housing stock (that’s about 270,000 homes) for a fixed population would cause house prices to fall by around 2%. So, by that logic, building 1m homes should cause a 7% drop in house prices. But this hardly figures against the 200%+ increase in house prices we’ve seen since 1996, during which time the number of homes per head of the population has actually increased, not fallen. Meanwhile the Bank of England’s analysts have acknowledged that “scarcity of housing has played almost no role” in increasing prices since 2000, but rather, the spiralling house prices were largely caused by the overabundance of cheap mortgage credit. Think The Big Short.

And yet somehow, however much evidence we might be shown, there is a part of our brain that simply refuses to believe that house prices aren’t primarily a problem of supply and demand. It just feels true. It fits our mental model.

This may be because, at some level, it is true. But what is true at the scale of a single neighbourhood may not be true at the scale of a country. What is true over the course of a month may not be true over a year. What is true in one location may not be true in another.

This is the problem with simple mental models. They are necessary, but also dangerous. As humans, we like the idea of something being either true, or false. And we like to form tribes. This bias towards binaries blinds us to nuance: it distracts us from the possibility that two apparently opposing ideas might both be true, and part of a larger, more complex picture.

Supply: it’s complicated

For example, let’s start with supply. We talk about ‘building more houses’. But what do we actually mean by that? Most of what makes a property unaffordable is not the house – but rather the land underneath it. If you own a house worth, say, £400,000, I’ve got news for you: almost all of that value is not in the house, it’s in the land underneath it. The house is just a heap of old bricks. If you want to test this theory, try putting your house onto the back of a lorry and see how much you get for it on eBay.

The value of a specific piece of land relative to another is driven by its desirability. That largely depends on three things: it’s location (proximity to jobs, beaches, a high status neighbourhood full of cool people, quietness, clean air etc), the value of infrastructure and services from which it benefits (roads, power, schools, police etc) and whether or not it has planning permission. So a plot of land in Mayfair with no chance of getting planning permission for anything other than use as a public park has a market value of something close to zero. Meanwhile, a plot of land in the middle of Siberia with planning permission for five hundred homes, but no connection to infrastructure or jobs will also be worth nothing. On the other hand, being within the catchment area of an Oftsed ‘Outstanding’ school will typically add £40k-100k per home to the value of a permissioned plot of land.

So fundamentally, when we talk about increasing the ‘supply’ of homes, it’s not really about the construction of the actual houses themselves. It’s about increasing the number of serviced plots of land with planning permission for homes, in places where people want to live. That is the thing that makes the value of a property go from £0 to £N,000,000.

The rate at which physical, inhabitable structures get built onto those ‘plots’ and by whom, is a separate question; one that brings its own layers of complexity.

In the UK today, our land market is dominated by a small number of companies called ‘housebuilders’ whose business model is roughly similar to that of ticket touts at a concert: that is, they buy land before it has planning permission, then wait for its value to rise (not least by capturing the huge increase in its value when we give our collective consent for it to be developed). Then, provided prices are rising, they build-out and sell, trickling new properties into the local market at a steady rate to avoid exceeding the ‘absorption rate’ (the rate at which the number of homes on the market in any one week might exceed the number of prospective buyers, thus giving buyers more power and causing prices to fall). The housebuilders themselves have been keen to prove how good they are at this. Barratt homes even commissioned a study by LSE to demonstrate to local homeowners that their developments would not reduce local house prices. This steady-trickle approach is obviously very different to the way that, for example, self and custom-builders behave if you release serviced plots directly to them. They build as soon as they can, because they want somewhere to live.

If a developer were to foolishly drop too many homes onto a local market too quickly, prices might indeed drop, but only temporarily. Sooner or later they would re-correct back up to their underlying value. And to understand why that is we need to look at the other side of the equation: demand.

Demand: it’s even more complicated

If the concept of ‘supply’ in housing is a bit tricky, the concept of ‘demand’ is a philosophical minefield.

In our neat mental model, we tend to think of demand as being finite; satisfiable. For example, I am writing this on a laptop. I do not need more than one laptop every ten years or so. So, at any given time, the demand for laptops is exhaustible for a given population. If Apple, HP and Microsoft were to suddenly flood the market with more laptops, the price of new laptops would fall. Classic supply and demand curves.

But land is different. It is compulsory (you can’t choose to not need it), finite in supply, immobile and it lasts forever. For most of us it is also the single most expensive thing we ever buy, and most of us stretch ourselves to the financial limit to get the best bit of it we can. As a result, for any given home, there are actually three demand-markets. And they don’t behave the way that Econ 101 textbooks would like them to.

Buy to live

The most obvious demand market for homes is people who want somewhere to live. In theory, that should be a finite market — because at any given time the population is only a certain size. So there should, in theory, be a point where you can have ‘enough’ homes. That’s to say, as many homes as there are families; thus creating a buyers market, and driving prices down.

But in practice that point is very hard to find – for two reasons.

New household formation. Demand for homes is a bit like demand for roads: the more lanes you add, the more people travel by car. So you can build more homes in a desirable location, but if better-off families can afford to, they will buy them as a second home in the city, or in the country, or by the sea or wherever. Or children will move out, or wealthy couples get a divorce etc. Houses are not rationed to one each, so even if we were to create so many new homes that prices begin to fall, a tier of wealthier families would still get first dibs on second homes before anyone on lower incomes got a look-in for their first home. Just ask anyone who lives in Cornwall.

All locations are not equal. In reality, demand is never ‘satisfied’ even once you have a home. There is always a better place.

Imagine if we were to line-up every residential property in Britain in order of desirability, with highly desirable locations on the left and less desirable locations on the right.

The first thing we’d notice might be a surprise to us: there are already enough homes for everyone, it’s just that some are in very low income, low desirability areas.

Land prices are society’s way of organising itself in space, along a ladder of relative wealth. Rich people want the best spots, and the prices adjust accordingly. Everyone wants the best place on the map that they can afford, and what you can afford is determined by how much other people can afford. This is what Adam Smith and David Ricardo recognised as the ‘Law of Rent’. The price of a plot of land is not defined by what it cost to produce (it didn’t cost anything to produce – it was just there), but by what the richest potential buyer can afford to pay for it. That’s why we saw such huge inflation in the 2000’s. Everyone could borrow more, so everyone could afford more, so prices went up.

The fact that the value of land and property starts with what people can afford and works backwards is somewhat counterintuitive (John Stuart Mill rather snobbishly referred to it as the Pons Asinorum of economics – meaning a concept that only intelligent people can grasp) but most of us are actually already familiar with it. If you have ever bought or rented a house, the first thing you do is work out your budget, based on what you can afford, then you go and look for what you can get for that amount. If you have ever been outbid on a house, it was almost certainly by someone who could simply afford to pay more than you could. You couldn’t have outbid them even if you had wanted to.

But it gets even more complicated. In any country that is larger than a one hour commute in size, there is not just one property market, but many local property markets. Each local area is different in terms of its desirability, infrastructure, jobs, services and income levels.

So when asking ourselves a question like ‘what would happen if we were to build one million new houses?’ it is meaningless to talk about it at the scale of a whole country. We can only really look at the impact on a particular local market.

Within any given local frame of reference, demand is always non-finite – because there is an effectively unlimited number of people who might want to move into or out of that area, driven by a whole range of factors – both push and pull.

The most important of these push / pull factors is access to jobs, wages and business opportunities. In other words, local rents and property prices are a function of local incomes (or more precisely, what someone can earn while living there). This is why you’ll hear people talk about ‘giving up their job and moving to the country’ in the same sentence (home-working via zoom is now changing this, resulting in an explosion of property prices in rural areas).

It also helps explain why rate matters. Only a certain number of people are moving into (or within) a given area in any given month. Since buyers’ budgets are set by the maximum they can afford to pay (or borrow), it makes commercial sense for a seller to wait for the next wealthy buyer next month, instead of selling more homes this month at a temporarily lower price.

Buy to let

There is also, of course, another demand market for homes: landlords. People who are rich enough to own not just their home, but also to own other peoples’, and capture the lion’s share of their income as rent. So long as someone, somewhere with enough cash believes that a given location is a place that people will want to live over the next ten years or so, they’ll buy it and rent it out. It’s basically free money. In a globalised, increasingly unequal economy there is a limitless pool of capital looking to buy property, and huge amounts of UK land is now owned by overseas investors. The obvious problem with this is that it crowds-out families who want to buy homes for themselves – as illustrated by stories such as this. You will often hear shills claim that landlords ‘contribute to supply’, but when you think about it logically, it’s an absurd idea. In the same way that every empty plot is a plot that could have a house on it, so every home that is owned by a landlord is a home that could be owned by the family that live in it, if they had the upfront capital.

In truth, the Buy to Let market neither adds to overall housing capacity, nor takes capacity out of it (except for the growing trend of turning full time homes into AirBnB holiday / temporary lets) – it just inserts itself between plot supply and Buy to Live demand in ‘hot’ areas, in some cases providing a service (for example, providing reduced rent homes to those in need, or renting homes to those who don’t want to buy), but in most cases increasingly just capturing economic rent and crowding-out families who would like to buy their own home in those areas.

Buy to leave

The third demand market overlaps with the other two, but it also sometimes breaks out on its own – much to people’s fury and indignation. It comprises investors who buy property purely in order to speculate on its rising value, or to store money: not even with any intention of living in it or renting it out. In an era of global instability and monopoly consolidation, where savings and other forms of investment (for example in productive goods and services) seem to offer relatively high risks and low yields, using property as a speculation asset – like Bitcoin or gold – offers a tempting alternative. The UK offers a particularly acquiescent regulatory environment for this kind of speculation activity. In fact, in recent decades, successive UK governments of different stripes have demonstrated their willingness — intentionally or not — to use policy and taxpayer money to inflate land prices. For example subsidies, loans, Stamp Duty holidays and Help to Buy. The most recent 2020-21 Stamp Duty holiday is estimated to have cost the taxpayer £3.3bn, and helped drive a 10% increase in house prices. Whilst on the face of it these policies might appear to help first time buyers, in reality they do the exact opposite: inflating prices, locking out even more people from ownership. They also send a signal to speculators that the taxpayer will effectively underwrite the risk of their bet, despite the economic and social damage that land and property speculation does: keeping properties empty, pricing-out families in ‘hot’ areas, and driving a boom and bust cycle in the economy that wipes out businesses and savings every 18 years or so.

The way that these three demand markets interact with each other is complicated, but the takeaway is clear: at any given time, the number of permissioned plots is finite but demand is effectively infinite. For as long as lenders and speculators believe that rents and property in a particular area will retain their value over the mid term, property there will remain valuable, no matter how many you build. When that confidence collapses, house prices won’t fall gradually in a neat, linear relationship with supply. They will plunge dramatically.

When we think about it, we probably already knew this. To test this, let’s do it as a thought experiment. Let’s ask ourselves what would happen if we were to build one million new homes in the UK tomorrow.

To make things slightly simpler for ourselves, let’s remove the filtering effect of housebuilders’ business model from the equation and imagine that the state is going to both permission and build these homes directly, almost overnight.

Experiment 1

In the first scenario, let’s imagine they are all built in say, Middlesbrough, where it is currently possible to buy a house for as little as £60,000. What would be the impact on house prices in London? Probably none. Unless the development was so beautiful and also included such spectacular investment into schools, trains, internet and economic infrastructure and jobs that young families in London were tempted to move north en masse. In which case we can say it wasn’t really the one million new houses wot dunnit at all – it was the investment into infrastructure, jobs and placemaking.

Experiment 2

For a second scenario, let’s do something different. There are 1.8m postcodes in the UK. Let’s say that we put one new house in 1m of them. What would that do to the house prices in any given street? Again, we’d probably say not that much. Demand is still infinite, and the inherent value of each location hasn’t really changed. The additional capacity would soon be absorbed.

Experiment 3

Ok, so let’s focus on a more obvious scenario. You will often hear politicians talk about ‘building homes where they are needed’. What they mean by this is building homes in desirable areas, where prices are high and rising.

Let’s put all 1m new homes within metropolitan London (which is entirely possible, by the way). Increasing London’s housing capacity by 28% overnight in this way would almost certainly create a sudden glut of choice for buyers and renters, causing a heavy drop in prices. But what would happen after that? Would prices stay low forever?

It may actually have the opposite effect and increase house prices. If you think about it, today there is no longer a particular geographic reason for London to be in London. People only want to be in London because… other people want to be in London. Its gravitational pull is a self-fulfilling prophecy. Employees want to be where businesses are. Businesses want to be where employees are. Creatives want to be where other creatives are. Young people want to be where young people are. We are, basically, a social species. That’s why cities exist, and that’s why most high value land has high value.

And we would have just added even more people to London, adding even more economic activity, labour capacity, creating jobs; reinforcing even further London’s gravitational pull. This is what’s known as the ‘agglomeration effect’. Hot areas get hotter, and cold areas get colder: the opposite of levelling-up. Indeed, on the long run, I’d suggest that the only way that adding 1m homes to central London in this way would reduce property values is if those homes and neighbourhoods were of such poor quality, and so underserved by infrastructure and services that people want to leave as soon as they can. Sure, we would have succeeded in lowering property prices, but in the worst possible way.

Experiment 4

Similar reasoning can be applied to another hotly-debated idea: building on the greenbelt. What if we were to build the 1m homes in new dormitory neighbourhoods around the outskirts of London? Would that lower the value of homes in central London? Again, probably not. It’s like trying to reduce the value of front row tickets to a concert by putting extra chairs out in the car park. Indeed, it may even increase the social cachet attached to owning a house in the centre.

To be clear, none of these are arguments that we should not build a million new homes in the UK. We should. But we are doomed to failure if we see it purely as a game of numbers at any cost; lowering the bar to get developers to build tiny, dim, carbon-intensive, socially-isolating, unaffordable boxes, in barren, treeless, tarmac neighbourhoods, with little or no community infrastructure — all in the mistaken belief that if we build enough, it will fix the affordability problem.

The debate we should be having instead is about what kind of places we want to create, and under what terms of ownership or tenure; looking for ways to keep homes affordable as places to live, without families being crowded-out by private landlords or the property speculation market. We need to reshape the systems and incentives that drive development; so we’re building beautiful, generous, healthy, zero-carbon, zero-waste homes in dense, leafy, walkable low and medium rise neighbourhoods, with great social infrastructure, thriving local businesses and resilient communities. Great places to grow up and to grow old. We need to capture some of the inflated land values in the South East, and invest it into building green infrastructure, schools and business opportunities in the North and West, creating ‘location value’ across the whole country using an infrastructure-first approach, then letting as many communities and families as possible buy land or serviced plots to build their own zero carbon homes. This is the kind of distributed, purpose-led development that is likely to be popular, creating places across the UK that are first and foremost built as fantastic places to live and work: hundred-year platforms for successful, resilient, inclusive communities and economies.

Sometimes, the best conclusion we can take from a thought experiment is that we were asking ourselves the wrong question in the first place.