Evidence and the persistence of mistaken ideas: the case of house prices

Another paper, this time from the Bank of England written by former MPC member David Miles and Victoria Monro, shows that the rise in house prices we have experienced since 1985 is mainly the result of lower real interest rates. The other, less important, driver is household income. Those two effects together can account for all the increase in house prices relative to inflation. The increase in house prices is not the result of a shortage of new houses.

Those who remember two earlier posts of mine will know of my own conjecture along similar lines. More recently Ian Mulheirn has championed this theory: here he is commenting on an apparently contrary view from Paul Cheshire. The importance of real interest rates to house prices has been understood for a long time: the first time I came across it was when Steve Nickell wrote a paper when I think he was still on the MPC. Very recently, here is Paul Johnson making the same point.

Secular stagnation is used by most macroeconomists to describe the current era where real interest rates appear to be permanently lower than they were decades before. The uncomfortable conclusion would be that as long as this era lasts, house prices will remain at levels that are unaffordable for many young people. Building more houses on any reasonable scale is not going to change that very much.

The reasoning behind the theory is incredibly simple. Houses are an asset. Like any asset, its price depends on the return from holding them (in the case of housing rents) and the rate of interest. The demand and supply for housing services (i.e. a roof over your head) determines rents rather than house prices. Imagine choosing between investing in housing or in government debt (more specifically a perpetuity, so you never get the money back but the interest pays forever), Interest rates on government debt are 2%, so on every £100 K you invest in government debt, you get 2K a year in interest. Suppose the (net of costs) rent on every £100K of house was 2K a year. Then you are indifferent to whether you own either asset.

Now suppose interest rates fall to 1%, but rents stay the same. Everyone wants to become a landlord, and people with money to invest buy houses to rent, because before interest rates rose you are getting double the return you were getting on debt. With perfect arbitrage this will carry on happening until houses that used to be worth 100K are now worth 200K, so that the return to housing again equals the return to holding debt = 1%. House prices have doubled, but the demand and supply of housing services has remained unchanged. The suggestion is that this is the process behind rising house prices in the UK.

That does not mean building more houses (increasing the supply of housing services) has no effect on house prices. Raising supply pushes down rents, other things being equal, and that reduces the return from owning a house, so it will reduce house prices. But the stock of houses is very large, so even with large house building programmes the impact on rents is small. Here Ian Mulheirn shows what the paper by Miles and Munro says about the small size of that effect.

You might say that any reduction in house prices is welcome, but you are using a great many resources (and a fair bit of land) to produce a modest effect. You might get a similar impact on house prices if the government undertook a serious fiscal stimulus, leading to a rise in short interest rates which would have a modest impact on long interest rates, but a noticeable impact in reducing house prices.

My question is why this point is almost never made in the popular discourse on the house price problem? One answer is that housebuilders have a vested interest in suggesting a dire need for more housebuilding, in part because it adds to pressure on governments to free up greenfield sites. This is exactly what has happened since 2010. There is nearly always a vested interest in perpetuating incorrect economic explanations.

In this case, as in others like the supposed need for austerity, there is something else, and that is an apparently simple piece of economics that perpetuates this misconception. With austerity it is that the government should be like a household, which most economists believed before Keynes showed it was false. With house prices it is that prices reflect demand and supply.

The difference between austerity and failing to distinguish between house prices and the price of housing services is that the former is more difficult to challenge than the latter. The reason is that everyone also talks about housing normally being a good investment. That is seeing housing as an asset, so all you need to do to break the misconception is a bit of asset pricing theory.

With issues like these, there are two spheres of understanding, with precious few links between them. There is what I will call the knowledge sphere, where academics (including academic think tanks) and economists in central banks and elsewhere regularly exchange ideas and evidence within that group. There is a second group comprising most of the print media, the broadcast media, some (mainly right wing) political think tanks and most politicians, where again communication within the group is pretty good. 

Communication between the two spheres is sparse. Most political journalists in the broadcast media spend more time watching each other and reading the print media than they do talking to people in the other sphere. Despite many who work hard to package knowledge in accessible ways, often the best those in the knowledge sphere can hope for is an article in the Guardian, FT or Times. If politicians don’t want to access expertise, there is therefore little requiring them to be knowledgeable. The examples I have highlighted are from economics, but I think it is true for all the social sciences.

As a result, politicians can continue to propagate and pursue bad ideas, like austerity is necessary or house building is the answer to high house prices, with little or no challenge in their own sphere. This is not about experts forcing politicians to do what they suggest, but about the public and even politicians being aware of what the evidence suggests. The fundamental problem is not that those in the knowledge sphere don't communicate well, but that too many politicians and much of the media do not want to be well informed. 

House prices and rents in the UK

I am not a housing expert, but it seems to me that the public debate is completely confused because it fails to make the distinction between house prices and rents. If we are talking about the supply and demand for housing, the price that equates those two things is rent, not house prices.

I discussed why here, but let me summarise the argument. Rent reflects the cost of being housed, of having a roof over your head. If there are less houses to go around, rents will be higher: higher enough to make some people share flats, live with parents or whatever. Because houses to buy can quickly change into houses to rent, there are not really separate markets for buying and renting, but just one big housing market.

The price of a house is the price of an asset. The asset in this case provides a roof over your head for as long as you own it. This means that house prices depend on current and future rents. Crucially, however, like any asset, the price is the discounted sum of future rents, where the discount rate is the real rate of interest. If real interest rates fall but future real rents stay unchanged, housing becomes a more attractive asset, and so wealthy people will buy more houses, pushing the price up.

Below is a chart of the ratio of house prices to rents in the UK and France, from OECD data.

There are large swings, but no major trend before around 2000. (That may surprise people, but it reflects what has happened to rents which we will come to.) In the early years of this millenium the house price to rent ratio increased substantially in both countries, and has stayed higher. I have included France with the UK to suggest that there may be some common factor influencing their similar behaviour. [1]

That common factor is real interest rates. You can define real interest rates many different ways: here I’m just going to be very lazy and pull data from the World Bank.

Again ignore the details (I have no idea about 1995) and focus on the trend. Around 2000, real interest rates started falling, and falling substantially. As real interest rates fall, house prices rise.

This will only be true if the housing market is liberalised so that this kind of arbitrage works, and that there are no taxes that stop the arbitrage happening. That was not the case in the UK before the 1980s (mortgages were rationed when I bought my first house), which is just one reason why you would not expect this relationship to hold over that period. But in the last two decades, lower returns on other assets has seen the rise of the middle class landlord as a way of saving for retirement.

This substantial fall in real interest rates is a worldwide phenomenon, and it goes by the name of secular stagnation. Why it has happened and to what extent it is permanent is still the subject of lively debate, which is beyond the scope of this post. The key test will be when nominal interest rates begin to rise over the next few years: to what extent do real interest rates rise with them. All I can say for sure is do not rely on those who say house prices always rise over time.

Thus the rise in house prices in the UK and France since 2000 has got little to do with a lack of house building, a point that Ian Mulheirn has stressed. But what about rents, which is where we should look for any imbalances in supply and demand. Here is some IFS data from a recent paper by Robert Joyce, Matthew Mitchell and Agnes Norris Keiller.

Outside London, there has not been a rise in the proportion of income spent on rent. Essentially, and I suspect this applies before the mid-90s, housing costs (rents) have risen with earnings rather than prices, and at constant real interest rates that would mean house prices rising with earnings. This represents a very reasonable return on any asset, and is why we think buying a house is a good investment. Now you could argue that we should build enough houses so that this proportion of income spent on housing falls, as it has for food for example. What you cannot argue is that building too few houses has anything to do with why houses have suddenly become unaffordable to young people.

The situation for rents has clearly been different in London in recent years, and London house prices have also risen much faster than elsewhere. David Miles and colleagues have written an interesting paper on how house prices in cities can rise as more people work in them but transport costs do not fall. In recent years UK governments have been trying to reduce the subsidy for train travel, and higher rents are a natural consequence. One way to reverse this is to invest in new and improved transport links into cities. However I think the main reason that house prices have recently risen in major cities in many countries is the decline in real interest rates noted above. (Here is the same debate in Vancouver.)

Does secular stagnation (low real interest rates) mean that a whole generation has to rent rather than buy? The main problem is the deposit that first time buyers have to find. Low real interest rates mean a mortgage is easier to service once you have one, although low rates of nominal earnings growth mean that it doesn't get so much easier over time as it used to. But rising prices means rising deposits, which if parents cannot help means saving for a long time. Banks do not want to take the risk of lower deposits, particularly if there is a real chance that house prices could fall. Help to Buy is about the state taking over the risk that Banks will not take, but is that something we collectively want to do? That is the debate we should be having in an age of secular stagnation. Building more houses may or may not be fine, but if real interest rates stay low it is not going to make houses affordable again for the generation that can no longer buy a home.

[1] It is fascinating to look at the countries that are similar to the UK and France, and those that are not (like the US and the Netherlands, but especially Germany). If anyone can tell me why these countries have not seen a permanent upward shift in house prices I would love to hear it.

Household debt and house prices

In previous posts I have talked about why I am suspicious of (but not completely opposed to) the idea that the UK (or US) has a serious problem because there is too much personal debt. Too much popular discussion goes as follows: booms and busts are often caused by excess lending and borrowing, household debt to income ratios are currently high compared to a few decades ago, and so we must be on the verge of a new personal debt crisis. The first two points are true, but the third does not follow because of one thing: house prices.

I thought I would illustrate the key point with a graph, based on data from the OECD’s Economic Outlook.

The yellow line is house prices relative to income: the absolute level is arbitrary. The red line is mortgage debt as a ratio to personal income, and the blue line is the total debt to income ratio. The green line is the difference between the blue and red i.e. non-mortgage debt relative to income.

The key point is that most of total household debt is mortgage debt, and this follows house prices. That the two should track each other over the long term is not surprising, but the fact that mortgage debt seemed to fall exactly with house prices is. (If house prices fall, this changes the value of new mortgages, but not the value of existing mortgages.) The reason may be that in the short term the interaction is two way. A fall in the demand for house purchase (and hence mortgages) will impact on price. Non-mortgage debt is now a little lower relative to income than before the crisis.

The basic story is therefore very simple. The main reason people go into debt is to buy a house. The more expensive houses get, the more they have to borrow. If there is a problem, it is not that we have all gone on unaffordable spending sprees. It is that house prices have been rising. Rising house prices increase not only household debt but household wealth, which is a key reason why wealth was also rising rapidly before the financial crisis.

The picture for the US is similar, except that non-mortgage debt has returned to pre-crisis levels.

This suggests no near term risk of any private debt crisis. Indeed for the UK, as Chris Giles reminds us, 2008 itself was not a crisis about personal debt, but a crisis about UK banks overseas lending. As a result, talk about private debt nearing ‘2008 crisis levels’ in the future is highly misleading.

There are two reasons why house prices have been rising in the UK: not enough houses are being built and real interest rates have gradually declined (secular stagnation). As governments have relatively little control over long term real interest rates, you will only reduce mortgage debt by reducing house prices by building more houses. To put it very simply, the aggregate private debt problem in the UK is a reflection of our longstanding inability to build houses.

That is a serious problem, and not just because it prevents a lot of potential first time buyers from being able to afford to buy. It means that, if interest rates were to rise significantly, households with mortgages would be spending much more of their income paying off the mortgage, and they would be more vulnerable to shocks to income as a result. One of the problems with the recent relatively slow growth in nominal wages is that the real burden of a fixed nominal mortgage has not been falling much as the mortgage grows older.

Worse still, if real interest rates did start to recover (secular stagnation proved to be less permanent than many people currently think) this would in itself tend to reduce house prices. That could leave many relatively new home owners with a mortgage larger than their house was worth. In the UK people cannot walk away from this negative equity. Equally lenders could have loans that were no longer covered by the value of an asset. Deflation coupled with rising real interest rates is a toxic mix. But all of these problems reflect the fact that house prices are currently too high. [1]

I think the simple takeaway is this. Anyone who talks about the growing problem of total private household debt without also talking about what has and what will happen to house prices is missing the elephant in the room.

[1] It is tempting to write that high levels of private debt are a symptom rather than a cause of these problems. That is too strong: people choose to take out a mortgage to buy a house rather than rent. However as most people only own one house, it has an element of truth. It seems odd to argue that an irresponsible debt fuelled increase in the desire to own houses is pushing up house prices.

Why high house prices are partly down to austerity

Diane Coyle, in reviewing Rowan Moore’s book Slow Burn City: London in the 21st Century, focuses on the idea that forever rising house prices could gradually kill off what is now a vibrant city. As housing gets steadily more expensive, getting people to work there will get more and more difficult. In the meantime, young people who can afford to buy get more and more into debt. I wonder whether soon mortgage providers will become more interested in the wealth of borrowers parents than in the borrower’s own earning capacity. (This is not just a London problem: see here about New York for example.)

The reason for this that everyone focuses on, understandably, is stagnant housing supply. However, housing can also be seen as an asset. Just as low real interest rates boost the stock market because a given stream of expected future dividends looks more attractive, much the same is true of housing (where dividends become rents). Stock prices can rise because expected future profitability increases, but they can also rise because expected real interest rates fall. With housing increasingly used as an asset for the wealthy, or even as a way of saving for retirement, house prices will behave in a similar way. A shortage of housing supply relative to demand raises rents, but even if rents stayed the same falling expected real interest rates raise house prices because those rents become more valuable compared to the falling returns from alternative forms of wealth.

That is why a good part of the house price problem comes from the macroeconomy: not just current low real interest rates, but also low expected rates (secular stagnation). The idea that house prices are tied down by the ability of first time buyers to borrow (and therefore to real wages or productivity, modified by changes in the risks lenders were willing to take) seems appropriate to a world where the importance of the very wealthy was declining, and most people could imagine owning their own home. We now seem to be moving to a more traditional world (remember Piketty) where wealth is more dominant, and with low interest rates that may also be a world where renting rather than home ownership becomes the norm for those who are not wealthy and whose parents are not wealthy.

There may be factors behind secular stagnation (low long term real interest rates) that we can do little about, but there are things we can do right now that will raise interest rates, and thereby tend to lower house prices. The most important of those is to stop taking demand out of the economy through continuing fiscal consolidation (aka austerity). This boost to demand that comes from ending fiscal consolidation will allow central banks to raise interest rates more quickly. While central banks may only be able to influence real interest rates in the short term, because so much uncertainty exists about what this long term involves the short term may have a powerful influence on more distant expectations.

We can also have some positive influence on the longer term by increasing public investment, including forms of public spending (that may not be classified as investment) that encourage private investment. It should also include building houses where (or of a kind) the private sector will not build. That will have beneficial effects in terms of raising real interest rates in both the short and longer term.

Ever rising house prices lead to unprecedented high levels of private debt, and also destroy the dream of many young people to own their own home. One answer is to build more houses, but another is to run better macroeconomic policies. That house prices continue to rise during a period of fiscal austerity is not an anachronism. It is not a bug but a feature of an age of austerity.

House prices and secular stagnation

This post starts off talking about the UK, but then goes global

We are all used to seeing graphs of house price to income ratios. Here is Nationwide’s first time buyer house price to earnings ratio for the UK and London.

UK First time buyer house prices relative to earnings: source Nationwide

Housing is becoming more and more unaffordable for first time buyers. Yet prices are currently booming (at least in London), and demand is so high estate agents are apparently now holding mass viewings to cope. In the UK the media now routinely call this a bubble, and the term ‘super bubble’ is now being used. London may be a bit unusual (see this extraordinary research), but it can also be a leading indicator for UK prices in general.

Bubbles are where prices move further and further away from their fundamental value, simply because everyone expects prices to continue to rise. One of the earliest and most famous bubbles involved tulip bulbs in the Netherlands in 1637. Yet that bubble lasted less than a year. The dot-com bubble lasted two or three years. If you think there should be some underlying constant value for the house price to income ratio, then this UK housing bubble has been going on for much longer than that. Instead of being pricked by the 2009 recession, it merely seems to have paused for breath. 

Yet does it make sense to compare house prices (the price of an asset) to average earnings or incomes? A more natural ratio would be the ratio of rents (the price of consuming housing) to earnings, and this has been relatively stable over this period. Or to put the same point another way, the ratio of house prices to rents has shown a similar pattern to the ratio of house prices to incomes shown above. (The Economist has a nice resource which shows this, and covers all the major countries besides the UK.)

If we think of housing as an asset, then the total return to this asset if you held it forever is the weighted sum of all future rents, where you value rents today more than rents in the future. Economists call this the discounted sum of rents. (If you are a homeowner, it is the rent that you are avoiding paying.) So why would house prices go up, if rents were roughly constant and were expected to remain so? The answer is that prices would go up if the rate at which you discounted the future fell. The relevant discount rate here is the real interest rate on alternative assets. That interest rate has indeed fallen over much the same time period as house prices have increased, as Chapter 3 of the IMF’s World Economic Outlook for March 2014 documents.

Think of it this way. You believe that the return you get from owning a house (the rent you get or save paying) will be roughly constant in real terms. However the return you get on other assets, measured by the real interest rate, is falling. So housing becomes more attractive as an asset. So more people buy houses, and arbitrage will mean its price will rise until the rate of return on housing assets adjusts down towards the lower rate of return on other assets. As Steve Nickell pointed out in 2004, if the expected risk free real interest rate permanently fell from, say, 4% to 2%, this could raise real house prices by 67%.

It is the expected return on other assets that matters here. The fact that actual real interest rates have fallen in the past would not matter much if they were expected to recover quickly. A key idea behind today’s discussion of secular stagnation is that real interest rates might stay pretty low for a long period of time. That in turn implies that house prices will be much higher relative to incomes than they were when real interest rates were higher.

So what appears to be a bubble may instead be a symptom of secular stagnation. We can make the same point by looking at another measure of affordability, again provided by Nationwide.

UK First time buyer mortgage payments as a percentage of mean disposable income: source Nationwide

First time buyers are able to afford elevated house prices, because interest rates on mortgages are so low. Of course raising the deposit is a problem, but the government’s Help to Buy scheme has come to the rescue by effectively restoring the 95% mortgage that disappeared in the recession.

Secular stagnation is a global idea, so if this story is right then we should see similar patterns abroad. Using the Economist as a guide, I think we can split countries into three groups. The first group is the UK, France (which looks very much like the UK!), Belgium, Italy, Sweden, Canada, Australia and New Zealand. There the house price to income ratio rose sharply in the 2000s, and has stayed high. The second group is the US, Denmark, Ireland, Netherlands and Spain, which also show large increases in the 2000s, but where post-recession declines have been so large as to actually wipe out (or come close to wiping out) these gains. For these countries it could be a bubble, or it could be an underlying rise temporarily offset by the impact of the recession. The third group is Germany and Switzerland (and maybe Austria), where the ratio has been falling over time, but has picked up over the last five years. There is one outlier, Japan, where the ratio has been falling since 1990. In a nutshell, the data is not clearly consistent with the secular stagnation story but does not clearly reject it either (ever thus!)

Does this mean we should stop calling what is happening in the UK a bubble? The first point is that secular stagnation is just an idea, and it may prove wrong, and if it does house prices may come tumbling down. Second, even if it is not wrong, it is still possible to have a bubble on top of the increase implied by lower interest rates. Indeed one of the concerns about the lower real interest rates associated with secular stagnation is that, by raising asset prices not just in housing but elsewhere, it may encourage bubbles to develop on top. So all we can say with certainty, for the UK at least, is that the Financial Policy Committee will have their work cut out when they next meet in June.   

Misunderstanding macroeconomic models

The first half of this post is meant for non-economists, but it ends with a couple of points on OLG modelling

I recently wrote a post on the Eggertsson and Mehrotra paper on secular stagnation, because I thought the paper was interesting. A much more critical post from Unlearning Economics (UE) has just appeared in Pieria. UE says it “helps to illustrate the troubles faced by contemporary macroeconomics”. One of UE’s complaints seems to reflect a misunderstanding, often shared by non-economists, about what much academic macromodelling is designed to do.

UE objects to the fact that the model assumes that the amount the young can borrow (the degree of leverage) is exogenous, which means that there is no attempt to explain where this constraint on the borrowing of the young comes from. UE also complains that the model contains no banks, and no investment in physical capital. In other words, the model is much too simple. It is a natural enough idea: to explain what might be currently going on, you need a more complex model that includes everything that could be important.

There is certainly a place for this kind of more elaborate model. Christiano, Eichenbaum and Trabandt in this paper want to argue that a model based on New Keynesian theory can track what has happened over the last ten years. Their model has 40 equations. If I was trying to do a similar exercise, I would want to augment the standard New Keynesian framework with at least the following: nominal wage stickiness as well as price stickiness, a financial sector that endogenised both the cost and rationing of credit, a model of consumption which allowed for credit constraints and precautionary saving, a housing market, a model of the labour market that combined matching with rationing (as here), and something that allowed recessions to have long lasting (hysteretic) impacts on labour supply and technical progress. However large models like this will involve many macroeconomic ‘mechanisms’, and it will generally be unclear which mechanisms are important at driving particular results or explaining particular facts. We do not want to treat the elaborate model as a black box, but instead we want to understand its properties.

To understand complex models, we need much simpler models. (I once - in this paper - called the process of relating complex models to simpler models ‘theoretical deconstruction’.) In fact it is often sensible to start with the simpler model. For example, a particular issue with secular stagnation is to show how the natural real interest rate can be negative for decades rather than years (i.e. beyond the Keynesian short term)? What mechanism can do this? As I explained in my post, neither a standard representative agent model nor a standard two period overlapping generations model (OLG model, where the two generations are those earning and those retired) will give you that result. What Eggertsson and Mehrotra show is that a very simple three period OLG model (which adds a young generation that borrows) where borrowing by the young is constrained (they would like to borrow more but cannot) can provide just that mechanism.

That is a key point of the paper. The paper is not designed to explain where borrowing constraints come from: there is now a big literature on that. Thankfully the authors do not feel compelled to microfound these constraints. Instead the paper simply offers and explores a mechanism whereby an increase in these borrowing constraints could move the natural interest rate into negative territory, and for it to stay there. Having established that result, it is for subsequent work (which the authors intend to do) to see if that mechanism survives complicating the model, by for example adding investment.

Suppose the endeavour is successful, and a more complex but realistic model is able to provide an account of secular stagnation that includes other important mechanisms and which is based on a realistic set of parameter values. That would be a success, but those not familiar with all the work would ask: why does this model allow real interest rates to be negative when the standard models we know do not. The reply would be that the three period OLG structure was critical, and to see why have a look at the original, simple model.

Now you might say the authors should wait until they have built the more realistic model before creating what could turn out to be a research path that might fail to achieve its goal. That would be quite wrong, because the more debate there is within the academic community when ideas are at their early stages the better. I want to give an example of this, but here I will go into territory that will probably only interest macroeconomists.

It might be the case, for example, that the authors intuition that their results will survive introducing other assets like physical capital can be shown to be wrong very quickly. Indeed, Nick Rowe has already made such a claim, arguing that the presence of land as an asset ensures a positive real interest rate. If Nick was right this could be enough to kill the research programme, without any more time being wasted. Whether he is right is another matter: this paper by Rhee may be relevant in that respect.

Here I just want to add a final thought. Within an OLG framework, it may not be necessary to establish the existence of a steady state with negative real interest rates. The typical period in an OLG model lasts two or more decades. So if the dynamics of such a model involved some overshooting, it might be possible to generate prolonged periods (in years) of negative interest rates even if the steady state real interest rate was positive. To be honest I’m not sure what might give rise to overshooting of this kind, but that may just reflect my inadequate imagination.