I totally disagree with your frankly wild assertions. Here’s why. In Limerick, for example, I estimated that a three-bed home should cost about €200,000, based on closing prices being 10% below asking prices. When I look at real-world examples here – http://bit.ly/2ymft7 or in particular here http://www.daft.ie/1447740 (if it’s still live) – I see that’s about right. A 92% mortgage on a €200k home translates into a repayment of €612 per month. On the rental side, look again at real world examples here – http://bit.ly/oqL3J – which suggest that €700 to €850 is the range for the exact same type of house, same quality etc. The estimated rent I used was €787.

Unfortunately for your argument, the rent and the house price are for similar types of property. This is because all property prices are calculated through hedonic price econometric regressions, which control for all measurable attributes, including specific area, bedroom/bathroom number, house type, etc. The method in question was reviewed and described in this paper, http://bit.ly/DioOJ, by the ESRI.

While you certainly talk a good game, you’ll have to concede that you do not have a monopoly either on scientific methods or on real-world examples. In fact, so far, you’ve failed to give any rebuttal on either, other than some vague talk about correlation not being equal to causation. I have no prior beliefs on this: as explained elsewhere, I had no idea what the answer was going to be when I started out my analysis and I’ve no incentives at all to produce biased research one way or the other. Frankly, that would undermine all the rest of my work.

This post isn’t an apology or encouragement for buying a house – it’s showing how the maths has changed and changed back so dramatically. It’s not taking account of everything by a long stretch; a follow-up post will extend the cashflow well past the myopic viewpoint of just one year. I thought the post was interesting precisely because some features of the 2006 market have returned, but sales haven’t reverted to anywhere near their previous levels (and nobody’s arguing that they should).

R

I’m a bit pressed for time so this may appear to be direct and blunt which is not my intention.

I will categorically state that your conclusions are simply invalid.

Your argument you use to draw your conclusion is simply invalid. You’ve taken two disjoint populations and calculated an aggregate for each (a mean). The problem is you’ve drawn a conclusion about equivalent individuals in each population from these aggregate figures.

This is as scientifically valid as concluding that drinking tea causes cancer from the observation that 98% of cancer patients had consumed tea in the year before being diagnosed.

In other words, it is not valid at all. It would not pass any sort of critical analysis. In academia, this would be laughed at if used even in a working paper.

There is simply NO SCIENTIFIC basis on which you can draw the conclusion you did from the evidence you have.

As a result, I am mildly suggesting that perhaps you are interpreting evidence to support something you believe in rather than presenting objective findings from the data you have.

Economics is supposed to be a science (albeit often a soft one)! The very least you should do is look at testing your conclusion empirically. Turn off Excel for a minute and go out and arrange a couple fo viewings of secondhand properies for sale and rent in Dublin.

It should be easy for you to pick out lets say equivalent (in terms of utility) properties for sale and rent in each of the 4 geographical areas. I’d suggest picking 3 equivalent pairs in each area (low, medium and high end) AND ONLY AFTER MATCHING THEM in terms of quality and utility, compare the corresponding costs of renting versus buying.

If you can do this in any sort of fashion, I will concede you may have a point.

I dunno how to say this subtly but you are just wrong here. Your reasoning is faulty and unscientific, you’ve presented no examples to provide any sort of empirical verification of your conclusions and in fact I have a small selection of empirical examples which flatly contradict your findings.

The problem, as with all “popular” science is that the general public do not have the background to spot logical flaws like this. Even separating correlation from causation is frustratingly difficult to do when attempting to discuss things in an accessible manner. In fact, in my experience, some people never full understand that correlation does not imply a causal relationship. The logical flaw in your argument is more subtle than this classic but is just as invalid.

Thanks for your comment. On (1), I take your point and perhaps a distinction between different segments of the market is appropriate, e.g. between the couple with kids and (presumed) secure employment planning to live in an area for a generation, who one would expect to place a premium on security and who would eveything else equal prefer to own their accomodation for the same amount as not owning it, and the young, single, mobile worker, who presumably places a premium on mobility. By focusing on three-bedroom homes, I was trying to isolate the former, but am happy to concede this line of thought may need more empirical foundations.

Points (2) and (3) likewise taken – on (3), you’re getting to the nub of things now, IMO. What this post was trying to do is start a discussion which might end in an estimation of the relative importance of the various factors leading to property purchase decisions. This post clearly shows, again IMO, that affordability is not the primary – or perhaps even a major – factor. It’s unfortunate that the discussion on thepropertypin has gone down an entirely different route.

R

Fair question. Simple answer is that this is the safest year of budgeting for anyone (an FTB couple themselves or an economist looking at them) to do. A few grand cash in hand after year 1 is a lot more tangible than a potentially larger but much less tangible amount after 20 years. After that, assumptions become a lot more important and therefore open to criticism or counter-assumption. Given the wealth of discussion about these calculations on thepropertypin, I will be speeding up my next round of calculations – which will be a lot more assumptions-led – and will look at a 10 year-cash flow also.

Thanks for stopping by and taking the time to comment,

Ronan.

1. The assumption that it should be cheaper to rent than buy. I don’t know whether it should or not, but my intuition seems to be opposite to yours. To me unless buying was cheaper there would be no such thing as property investors, and everyone would want to rent because you get flexibility and less hassle for a lower price.

But who would they rent from if there were no landlords?

Renting should be more expensive.

2. Are people really willing to live as tenants in conditions they wouldn’t put up with if it were their own home?

People who were willing to buy in Carlow and commute to and from Dublin on a daily basis were clearly willing to suffer more to be an owner than I would be willing to put up with as a renter.

If billboards in Carlow aimed at people sick of commuting is anything to go by, we’re hitting reality on that plan right about now.

Leave students out of it, they have very little to teach you since they are rarely making the buy vs rent choice.

3. You point out that the benefits of ownership include the ability to sell and/or borrow against the property.

Hardly benefits that apply at the moment.
It would be fairer to say that a downside of owning at the moment is the inability to sell.

-Rd

Thanks for coming back, and excellent comment, easy winner of the ‘Most Thorough Comment’ award in my book… and don’t worry, I know exactly what you mean about time constraints!

Thanks for going through things step by step. Your overall point is well-made, namely the one about individual properties vs. averages. I may not be able to get down to housing estate level, but I can hopefully get down to a set of characteristics (e.g. certain area, certain house type, certain bathroom number, etc.) that makes it reasonable to assume the properties are equivalent. I’ll see what I can do about the new properties too.

On the deposit issue, you’re right, I’ve been a classic economist and just assumed away that problem. Which makes the couple I’m looking at one that thought about buying in early 2007, having saved up their deposit, but one that’s held off since then.

I’m less convinced by the quality argument – again, just anecdotally, one could easily make the opposite argument that people are prepared to live as tenants in conditions they wouldn’t put up with if it were their own home. Or maybe I’m just stuck in student mentality! Perhaps the Census might have some proxies for this, e.g. presence of TV, broadband, etc., occupancy status. Something for me to think about anyway. Thanks again.

I am not a professional economist but I have some background in finance and have an amateur interest in this area. I was drawn to your original article during a discussion on askaboutmoney where I’ve been trying to argue against the prudence of the idea that people should put the vast majority of their wealth and excess income into Irish residential property from as young an age as possible. It is not directly relevant but if you’re interested, I use darag as a username there also so you can have a look at how your study got dragged into the discussion.

Hands up, yes I shot from the hip there and didn’t fully read your introduction. I looked at the conclusion, didn’t like it and blasted away. In particular, I missed the fact that the study was restricted to 3 bedroom houses. In fact I’d be happy enough if you simply ignored the points I made in my earlier message.

I’d like to back up and start again.

First of all, yes – as you correctly guessed – I had made up my mind before reading your report. However my position on this this is not based purely on whimsy. I have a number of grounds to suspect your findings.

First of all, in a personal capacity, during a period of about 6 months in 2008 I tried kicking tires on houses for sale as I considered purchasing (for the second time – I was an owner up to about 3/4 years ago). Don’t laugh yet. Every house I looked at, I made a deliberate effort try to find a comparable neighbour for rent. I had built a spreadsheet to attempt to capture all the costs of the two modes of living. In addition I had 2 direct examples of a landlord offering to sell to a tenant (late 2008 to a friend of mine and early 2007 to me) as data points. I didn’t find a single case where the cost of renting exceeded that of ownership. I found the opposite in fact; in every single case (bar one where owning was only slightly more expensive) ownership would cost significantly more. This samples include: four 2 or 3 bed houses (two in Phibsboro, one in Ranelagh and one near Island bridge) and three 2 bed apartments (one in D2 and two in D6). All were secondhand.

Obviously my set of data points is completely biased and unrepresentative and is tiny. It would be completely unreasonable to draw any general conclusions from such a sample. Despite this, I believe this sample to represent a challenge for your results. The chances are minuscule that of a sample size of 6 or 7, I just happened to get outliers for all accept one.

Next, the fact that landlords have deserted the market is also telling me that there is something wrong with your findings. Yes, your calculations are based on a model for first time buyers but the differences in terms of costs are not big enough by my calculations to dismiss this point as being completely irrelevant. Landlords have lots of ways to save costs also and can offset all interest from a tax point of view. The simple fact is that rational landlords are out of the market simply because rents are too low. Rents in Dublin have fallen in real terms between 1997 and 2007 while house prices have multiplied in the same period. If it was cheaper in 2007 to buy, then extrapolating back to 1997, you should have been able to afford to buy at least two properties for what it would cost to rent a single equivalent. I never remember such yields being available on Irish property but I could be wrong.

So where does that bring us? I cannot say where exactly the problem is with your analysis is but I am 99% certain that there is something wrong given the above.

My initial knee-jerk reaction was that you hadn’t included all the costs which the landlord bears when you rent thus underestimating the cost to own. I’m talking about things like bin charges, furnishing and maintenance. I still think there may be millage in this line of attack but I’m not sure. I might come back to it if I need to but for now it’s not what interests me.

Another cost I modelled which you seem to ignore (correct me if I’m wrong) is the cost of the deposit. Most buyers simply cannot access money for a deposit at mortgage rates. 10% down at the start is a big cost – many first time buyer spend a few years saving for a deposit but your model doesn’t include this cash-flow. OK we know that people fiddle the system by using undeclared credit union loans or some get money from parents but I suspect most scrimp and save for a few years. But again, I don’t feel this point has much chance of delivering the killer blow.

Currently, what screams out at me is that I have compared INDIVIDUAL houses for sale with equivalents (or in two cases the exact same property) for rent. You have used the AVERAGE price of a 3-bed house for sale in an area against the AVERAGE rent for a 3-bed in the same area. It is not at all evident to me that the average property available for rent is equivalent or comparable to the average house available for sale. An extreme example will illustrate the danger of what you’ve done here. Imagine a semi-feudal country called Absurdistan; on http://www.daft.astn there are 20 labourers cottages for rent and two landlord’s mansions with swimming pools. A straight comparison of average rents against average house prices is going to be completely misleading here.

When this first stuck me I thought bingo but to be honest I still need a plausible explanation for a bias of this nature (if this is the reason why we reached different results); i.e why would more of the superior class of house be available for rent while inferior properties were being offered for sale? One idea was that, perhaps most of the houses being sold during the period in question are newly builds in the outer suburbs while most people rent in the city centre or close to it where property is more expensive.

A more promising observation is that every property I looked at was secondhand and every single one required some investment to bring them up to the standard of an equivalent rental. In 3 cases it was negligible – maybe a couple of grand for a lick of paint and some flat-pack furniture; but at a rough estimate two required over 60k of work (installing central heating, roof replacement, rewiring, replumbing, kitchens or bathrooms installation, that sort of thing) and one required about 120k. Properties in this state can be put on the market for sale but are rarely rented.

As I was thinking about this I was wondering whether you’d be interested in looking at this discrepancy further? The most interesting insights often occur when reality contradicts a conclusion reached by deduction which I believe is happening here.

I don’t have access to the data you have and this data has great value as it is extensive and representative. The value of my couple of samples is that I’ve compared directly comparable properties not averages across regions and this gives a truer picture of the real choice facing a purchaser between buying and renting (albeit only if the purchaser is interested in living close to the centre of Dublin in decent 2 bed apartment or a small 2/3 bed house). I’m not sure how rich the data is that you have but one idea I had was that perhaps you could repeat your comparison but use as small a geographical divisions as possible. If you could get down to a granularity of say housing estate size, then I’d have much more confidence that your averages were in fact comparable but I don’t know whether you have the data to do something like this. Another line would be to add a general adjustment to the price of secondhand houses for sale to account for my observation that it cost anything up to 150k to bring a house for sale up to the average standard of houses to rent. This weight could be calibrated with a few manual comparisons. Alternatively, maybe try only considering NEW properties for sale as at least you know they won’t require roof replacement or rewiring to be habitable. There are lots of interesting approaches here.

Unfortunately I am scheduled to do a lot of travelling over the next 3 weeks so it may be difficult to dedicate big blocks of time to this but I’m certainly interested in exploring this further.