It makes sense that house prices increase above the risk-free rate due to their risky nature and historical evidence and conventional wisdom seems to back up the idea that real house prices tend to increase over time.

However, the ultimate consequence of upward-trending real house prices would seem to be that, eventually, nobody on Earth could afford to buy a house.

What is the resolution to this paradox? Is it simply a mistake to assume that house prices tend to rise? Is it tied to increasing population?

In other words: Over the past 50 years, house prices have risen significantly in real terms. Is this true of the past 500 years? Can we expect it to be true of the next 500? Are we in a 50-year housing boom, or is there some fundamental reason why real estate prices tend to rise?

  • 1
    $\begingroup$ Welcome to the site. It would be helpful to indicate (by editing your question) the length of the long term you have in mind. Your linked chart covers a 50 year period, but your statement about the "ultimate consequence" possibly suggests a much longer period. $\endgroup$ Commented Sep 10, 2021 at 10:09

4 Answers 4


As I read your question, you actually ask two distinct question, one whether house prices can be assumed to have positive trend, and one about whether population or something else causes prices to rise.

Is it simply a mistake to assume that house prices tend to rise?

Yes, this would be a mistake. There is no inherent economic reason why real house prices, ceteris paribus (i.e. all else equal), should increase or have increasing trend.

This does not hold just for housing but virtually any good, service or commodity. Prices are ultimately determined by supply and demand, and in turn supply and demand are determined by various factors such as peoples tastes, incomes, expectations, availability of substitutes (i.e. examples of things affecting demand), or factors like cost of materials, labor or other inputs and again expectations (i.e. examples of things that affect supply). If you are unfamiliar with how supply and demand work, you can also consider having look at first few chapters of Mankiw's Principles of Economics.

In principle there will always be some combination of the supply and demand that can result in housing price increase or decrease. For example, if tomorrow suddenly co-living with roommates would become hip and desirable, more people would live together, demand for houses would be lower and ceteris paribus price as well. Or alternatively, some new technology could allow houses to be built more cheaply it would expand supply and thus reduce price, ceteris paribus, as well.

Consequently, it would be grave mistake to just assume real housing prices can just go up. There is no economic rationale for such assertion. You have to first examine economic fundamentals and trends in economic fundamentals (like the costs etc mentioned above), and based on that you can make some forecast (but forecasts get progressively less accurate the more into future you want to forecast).

Is it tied to increasing population?

This is one of the factors, but not the only one or even a necessarily the most important factor.

Increase in population increases demand for housing, so this definitely pushes the real price of housing up, but demand is always just a half of a story, there is a whole cornucopia of examples where increases in demand were offset by even larger increases in supply leading to lower prices. In this case a lot of the increase in real prices of houses can be explained by supply factors.

  1. Construction is one of the most labor intensive industrial jobs (Yang et al 2019). As real wages of workers increase it simply becomes more expensive for firm to build houses so they shift their supply to the left leading to higher prices.

    Rigorous research aside, even if you just look at the cost of materials typical house takes to build they are not that expensive, but the labor that goes into them is. As real wages typically tend to rise over long-run (see FRED data here) you would expect prices of labor intensive goods to also increase in real term (all else equal). Of course, it is always possible that there is some sort of robot or 3D printing revolution just around the corner, and if that would happen the real price of housing would fall.

  2. Part of the high price of housing is caused by bad government policies which restrict supply of housing. Many countries have on their books zoning laws or other policies that restrict the supply of housing. Such laws prevent firms from increasing supply of housing even if they would like to and thus lead to higher prices (see for example The Impact of Zoning on Housing Affordability Edward L. Glaeser, Joseph Gyourko).

    Zoning laws are typically popular with voters because people who are already established in some area and own home have incentive to keep new house production down to protect their own property values. Also, some restrictions are made for environmental concerns (or at least under the pretense of such concerns like green belts in the UK). However, again as in the previous case, these laws might be changed in future, they are ultimately not very sustainable as new generations of voters will find harder and harder to become house owners and thus voter base for such policies will shrink.

  3. Houses improve in quality over time. This can lead to sustained increase in real price, and even though typically price indexes are constructed in a way that at least a little bit tries to correct for this. Indeed if you click through the data sources you will learn the Fred data come from BIS which uses the same methodology as eurostat here, which does quality adjustments. Nonetheless, it is generally agreed that price indexes typically overstate rise in prices at least a bit (e.g. see examples of literature on this issue Baker Get Prices Right or discussion in Hwang et al 2004). Also, house prices are only recorded for sold houses but the houses people sell are not random. You will see new dwellings or newly renovated dwellings overrepresented, which skews the statistics.

However, the ultimate consequence of upward-trending real house prices would seem to be that, eventually, nobody on Earth could afford to buy a house.

No, this is incorrect assessment for several reasons.

First, even if there would be deterministic upward trend in house prices that would continue forever without any end the statement would not hold as long as real wages would increase faster or at the same pace as house prices. So the statement is logically flawed, rising housing prices do not necessarily mean people would not be able to afford houses at some point.

Second, economics is an endogenous system. The higher price of housing the higher pressure for people to change their taste (learn to cope with roommates, preferences for tiny houses etc), and more pressure for firms to come with innovation to be the first one bring cheaper houses and beat the competition, and at the same time the more pressure on bad policies to be either abolished or replaced by policies that promote house construction.

Historically people often worried that some commodity, like food, water or oil etc will eventually become so expensive nobody will be able to afford it and so far that fear is mostly unfounded thanks to innovation and productivity growth over time.

  • $\begingroup$ Thank you for your thoughtful answer! 2 questions: "You will see new dwellings or newly renovated dwellings overrepresented, which skews the statistics." I'm sure this skews house price data, but how would it affect house price increase? And why do people invest in real estate if we expect only a tiny rise in house price over the long term (e.g. ~0.4% in-line with wage increase)? Investors receive the risk premium mostly in the form of excess rent? $\endgroup$
    – Zaz
    Commented Sep 10, 2021 at 20:51
  • $\begingroup$ @Zaz 1. If there is a lot of technological improvement in housing (e.g. houses with nice open areas, smart houses etc) then prices of brand new houses would be expected to rise faster (as that extra technology/quality commands premium) then what would be actual increase if you could measure price of all dwellings at the same time. 2. I never said that increase in price of housing will just perfectly copy real wage increase. For example, those zoning laws can easily lead to more than 50 or 60\% prices than there would otherwise be, and then if that couples with those other factors $\endgroup$
    – 1muflon1
    Commented Sep 10, 2021 at 20:59
  • $\begingroup$ you will definitely see higher than 0.4 return in the short run. But in the long run anything can happen. There is no 100\% safe investment. It is completely possible (I would not be surprised at all) if someone who buys house today in some areas would, in real terms, even loose money on getting a house. For example, here in Amsterdam where I live I personally think there is a house market bubble, plus local politicians are here starting to worming up to the idea of letting more houses be built, not directly in Amdam but in very close areas that adjunct to Amdam like Amstelveen etc $\endgroup$
    – 1muflon1
    Commented Sep 10, 2021 at 21:03
  • $\begingroup$ I also see a lot of new container housing units being made in some areas (which are cheap for students). So I think that anyone who buys house in Amdam now will likely loose money but that is part of investing, you always risk that you will loose part of your investment, nothing is 100\% safe, not even things like gold. But I might also be wrong, point is no self respected economist would think that investing into housing is 100\% safe investment - there is no such thing $\endgroup$
    – 1muflon1
    Commented Sep 10, 2021 at 21:04
  • $\begingroup$ I find it strange that you have not mentioned the main cost of housing which, in urban areas at least, is the cost of the land the house is built on. $\endgroup$
    – Mick
    Commented Sep 12, 2021 at 9:07

"conventional wisdom seems to back up the idea that real house prices tend to increase over time"

Not clear that's correct. See, for e.g., "House Prices and Fundamentals: 355 Years of Evidence" by Brent W. Ambrose, Piet Eichholtz, Thies Lindenthal. It was covered in this story

  • 1
    $\begingroup$ I don't think the OP means that 'conventional wisdom' is true, merely the view of most people. In that sense I think he is right that most people tend to think house prices will always go up. $\endgroup$ Commented Apr 21, 2022 at 22:08

First, notice that comparison of the change in the house prices with risk-free rate doesn't tell you anything about the change in the real house price. You need to compare the change in nominal house prices with the inflation rate to see if the real house price increases.

There are many reasons why supply and demand forces will probably not let the real house price increase forever (many of them were eloquently summarized by 1muflon1's comment). But even if you assume a deterministic monotone increase in the real house price, still, it doesn't mean housing would eventually become unaffordable for people. This is because people are also, on average, getting richer (due to a positive GDP growth rate), and their real wages are believed to increase in the long run.

Just one caveat. We should be clear about what you mean by the real house prices. We probably mean the average of house prices in different parts of the country (or world). But, the mean disguises a huge variation in house prices. It might be the case (and seems to be) that the average price is increasing because housing in some top cities is skyrocketing, while housing prices in rural areas or less modern neighbourhoods are even going down. In this scenario, the average price would go up, while housing in many small towns is going down. This would result in a divide and the society, meaning wealthy individuals can go and live in big urban cities, while the rest of the population can only afford to live in small towns.

  • $\begingroup$ Very good point about mean/median difference; however, it seems that mean and median growth are similar: theoretically, I presume inequality can't diverge due to threat of revolution, so mean/median ratios should remain roughly constant in the long-term; practically, over the past 60 years we see 5.5% vs 5.4% (FRED » ASPUS, MSPUS) $\endgroup$
    – Zaz
    Commented Sep 11, 2021 at 20:53

They cannot rise in the long term. Indeed, as objects that will deteriorate over time, they should decline in value continuously in the absence of alterations. They not only deteriorate but, even if maintained well, also become obsolete. A late 18th-century log cabin would be nearly impossible to sell except as a campsite or for its purely historical value as an attraction.

Now successive generations of home can see a change in value to the extent the home could be the source of goods and services not previously available. For example, adding interior plumbing is a large gain in service. So would having a home that generated its own power supply, particularly if it were stable and consistent.

As homes are not investments and rental properties are not homes, you want to be careful in applying the risk-free rate as a concept. If you plan to purchase a home, live and ultimately die it in, the risk-free rate is of little importance except to benchmark opportunity costs for other uses for the money to buy the home. Conversely, if you are buying property to rent to others, then since money is fungible, the risk free rate is very important.

However, since rent is a cash flow, it isn't the house that is growing in value. It is the cash flows with reinvestment. The rented home is still deteriorating. The rents should be maintenance costs plus recovery of principal plus risk premia plus the compensation for the patient recovery of cash.

If you are buying a home, you are paying the rent to yourself. The maintenance costs are the same, but rather than recover principal through cash flows, you are enjoying it through consumption. The risks are different. With a renter, the risks are related to damages, non-payment and resale risks. If you will never resale or damage the property on purpose, the risk is the loss of the service value of the property. For example, your favorite neighbors could move away or the neighboring farmland with a gorgeous view could become an industrial complex.

Historical nominal increases in home prices have been due to changes in the inflation rate, local population changes and migration, technology changes, preference changes, and differences in funding costs and methods as well as taxation rules. The past will not repeat itself.

It would be dangerous and probably foolish to use time trends to value real estate and attached structures unless carefully conditioned on other data. Up is exactly one-half of the possible directions an asset's price can go.

  • $\begingroup$ but up is 100% of the possible prices :) $\endgroup$ Commented Sep 10, 2021 at 9:51
  • 4
    $\begingroup$ "They cannot rise in the long term". But the OP has linked to a chart showing a rising trend in real prices over a 50 year period. At the very least that statement needs some qualification. $\endgroup$ Commented Sep 10, 2021 at 10:13
  • 2
    $\begingroup$ It's a defining feature of residential properties that they are an investment and a home! $\endgroup$
    – BrsG
    Commented Sep 10, 2021 at 10:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.