Ask why house prices rose or fell and you will get a different single answer from every person you ask. Each of them is usually describing one real force in isolation. The useful version is that a small number of forces act at once, on different timescales, and that most of the argument is really about which one is currently dominant.
Credit conditions, which usually lead
Most homes are bought with borrowed money, so the price of borrowing changes what buyers can offer before it changes anything else. When rates fall, a given monthly payment supports a larger loan, and competition converts that capacity into price. When they rise, the same arithmetic runs backwards.
The effect is not symmetrical. On the way up, extra borrowing capacity is bid straight into prices, quickly. On the way down, sellers resist, transactions fall instead, and the adjustment happens slowly through volume, withdrawn listings and, in real terms, through inflation eroding prices that never nominally fell.
Rates are not the whole of credit conditions either. Lending criteria matter as much: how income is assessed, what deposit is required, how affordability is stress tested, what proportion of a property's value a lender will advance. A market can tighten sharply with no change in the headline rate at all.
Supply, which moves slowly and matters more than it seems
New building adds to the stock very slowly. In an established area, annual completions are small against the existing stock, which is why new supply rarely moves prices in the short run and why it is such a poor emergency lever.
The more immediate supply question is how many existing owners choose to sell. That is a behavioural quantity, not a physical one. It responds to confidence, to the availability of somewhere to move to, to the cost of moving, and to whether people expect prices to be higher next year.
Forces, and the timescale each works on
- Interest rates and lending criteria
- Weeks to months. Changes buyer capacity almost immediately and is usually the first thing visible in offers.
- Employment and wages
- Months to years. Sets who can enter the market at all and how confident existing owners feel about committing.
- Willingness of owners to sell
- Weeks to months. Determines visible supply and therefore competition, independently of how many homes exist.
- New completions
- Years to decades. Small against existing stock in any one year, decisive over a generation.
- Planning decisions and land allocation
- Years. Determines where and whether future supply can appear, and reprices land long before any building starts.
- Transport access
- Years, then a step change. A new station or road reprices commuting time, which is one of the largest components of location value.
- School catchment and public services
- Continuous, with sharp boundaries. Produces measurable price differences across a single street where an administrative line falls.
- Seasonality
- Annual. Changes the mix and volume of what is on the market and can make year-on-year comparisons misleading if ignored.
Employment, wages and confidence
Housing demand is ultimately demand from households with income. An area's employment base determines how many households can form, how secure they feel, and how much of their income a lender will treat as reliable.
Confidence deserves its own mention because it is not the same as capacity. People with unchanged incomes and unchanged borrowing capacity will still stop bidding if they believe prices will be lower in six months, and that belief is self-fulfilling for as long as it lasts.
Planning, land and the thing being priced
It is worth being precise about what a buyer is paying for. A house is a structure with a fairly well-understood replacement cost, sitting on a plot whose value is essentially the value of permission to occupy that location. Most of the variation between an expensive home and a cheap one of similar construction is in the second part.
That is why planning decisions move prices so strongly and so far in advance. An allocation for development, a refusal, a conservation designation, a change to what may be built on adjoining land: each of these changes the supply of permitted locations, which is the genuinely scarce good.
Transport and time
Location value is largely a function of access, and access is measured in time rather than distance. A new fast service to a major employment centre can reprice an entire town, because it moves that town into a different commuting band. The effect typically arrives before the infrastructure does, as soon as the route is credible, and then flattens once the new journey time is established and priced.
Boundaries, catchments and micro-geography
Some of the sharpest price differences in housing occur over very short distances, at administrative boundaries. Where a school catchment, a council boundary, a parking zone or a flood designation cuts through a neighbourhood, otherwise identical properties on either side can trade at persistently different levels.
These effects are real but unstable. Catchment boundaries are redrawn, designations change, and any premium that depends on an administrative line is exposed to that line moving.
Seasonality, and how it fools people
Housing markets have a strong annual rhythm: activity clusters in spring and autumn, and thins around midsummer and midwinter. This changes not only how many homes sell but which ones, since family houses and small flats have different seasonal patterns.
Comparing a January figure with a June figure, or reading a quiet December as a market collapse, is one of the most common errors in casual analysis. Compare the same period year on year, and look at volume alongside price.
How the forces interact
The forces above rarely act alone, and they frequently offset one another. Rising rates with a strong employment base and very tight supply can produce a market that stalls in volume but holds its prices. Falling rates with weak local employment can produce activity without any real price movement. Rising supply in an area whose transport access has just improved can be absorbed entirely.
This is the honest reason forecasts of individual local markets are so unreliable: the interaction terms matter more than the individual coefficients, and much of the input is behavioural.