RTB Rent Register

I started using this data as a project to learn Jupyter Notebooks with my AI assistant at my side. Here are a few graphs:

1) Boxplot for the 1-bed properties in Cork South Central. The box shows the 25th to 75th percentile, with the horizontal lines showing median, max, min.

1773682072549.webp


2) Rent distribution for all the 1-bed houses and apartments together. With such a large spread, it really begs the question, what is a "market rent"!

1773682207793.webp


3) Rent distbution for apartments in Pembroke EA of Dublin 4 (Donnybrook, Ballsbridge, etc)
1773682231439.webp


4) Again Pembroke EA, rent levels for 1,2,3 bed apartments.
1773682349976.webp

5) Tenancy Start Date for all the areas I surveyed. As noted, these are parsed out of the RT number, so may be flawed. But there appears to be a clear pattern of underreporting of 2021-2023.
1773682511528.webp
 
What on earth is going on with 2021 - 2023 that there are so few tenancies?

Here's a suggestion, put in larger context. Perhaps people who like data could check it.....:

RTB recently started publishing the "Profile of the Register". with 'cleaned up' data. What they mean by that is that, over the years "phantom" tenancies gradually built up due to tenancies often not being removed from the register after a tenancy ended or, as they refer (iirc) to it "inflation" in the registered tenancy numbers. Once annual re-registration started it provided a way to get rid of the phantom tenancies - so they started this new "clean" publication.

One problem with it is it starts in Q2 2023.

1773686469573.webp


You will see the rising trend in private tenancy numbers. This has been key in convincing the Housing Minister that skeptics of the idea landlords were leaving the market were right - either that talk of landlords leaving the business was either mistaken or a deliberate attempt to mislead politicians and public in order to protect landlords' interests.

RTB Annual Reports show that having been steadily rising, private tenancies peaked in 2016 at about 320,000. They then, slowly, then more rapidly fell to about 220,000 in 2021. But there are various caveats in the data releases to the effect that 2021 figures are estimated to try to take account of a shift in part 4 expiries due to a 2016 chance in the length of them from 4 to 6 years. The falls in the 2022 and 2023 figures were partly due to:

a) actual tenancies going
b) phantom tenancies being removed
c) (common sense suggests) actual tenancies being wrongly removed, only to appear as registrations later - see below

Number of private tenancies, per RTB annual reports.

2012 264,434

2013 282,918

2014 303,574

2015 319,609

2016 319,822 ('temporary' rent control introduced)

2017 313,002

2018 307,348

2019 303,023

2020 297,837

2021 276,223

2022 246,453

2023 223,979

2024 240,946

General inflation in private tenancy numbers would have gradually accrued as every year some ended tenancies would fail to be removed from the register. RTB indicates they reached a peak of about 20k in 2021. The RTB took out it's estimate of an accumulated 20k extra caused by it in the numbers reported that year - 2021. Had that not built up in the first place then presumably there would have been a steadier decline in tenancy numbers from 2016 to 2021 rather it being concentrated in the drop in the 2021 figure.

Likely the effect of that was to mask the fall in the tenancy numbers after rent control was introduced in 2016. Had the data been clean of that accumulation the declines reported up to 2021 would have been even steeper, but ended at the same figure. It was a loss of about 45,000.

Where there is a real break in the data is between those reported for 2021 and 2023. That's when the RTB made mess of introducing a computer system that landlords were legally obliged to use to register their tenancies (and do so annually). It was a fiasco and didn't work (penalties for non-registration were not applied as a result) until March 2024 (penalties reintroduced). Landlords often just gave up as it was often impossible to register a tenancy and the RTB cancelled all late fees.

So throughout 2022 up to early 2024 it is likely unwise to rely on figures an any particular point in time. Although common sense tells us registration numbers in 2022 resulting from the earliest chaos will have been lower than reality, and that the numbers for second half of 2023 and 2024 should be boosted by delayed attempted registrations from earlier eventually getting done.

Another likely problem, that the RTB does not seem to acknowledge anywhere, is that they introduced the Annual Registration requirement at a time when landlords could not actually perform the annual registration since the computer system did not work. But they did (according to what they say in the Annual Report) use the lack of a re-registration to indicate a phantom tenancy and 'clean up' the data. It seems reasonable therefore to assume the trough of 223k tenancies they report for 2023 was at least in part an consequence of deleting tenancies which were real, and got re-registered later on. That would mean the Profile of the Register series from Q2 2023 would have an initial rise just due to registrations which could not, or were not, made earlier finally getting done.

That initial increasing trend in the Profile of the Register was interpreted as evidence of renewed investor enthusiasm for the Irish Rental market and a tailing off (or to those who prefer "proof it was all just lies") of existing landlord exits.

Somewhere between the figures for 2021 and those for 2024 there will have been a (quite significant) number of genuine 'phantom' tenancies flushed out of the system, but we don't know how many came out when because of the systems fiasco, so the movement in figures 2022 & 2023 are regarded as not terribly reliable outside RTB. The end figure of about 240k for 2024 tho is thought now much more accurate than the 2022 ones, but the data before 2022 can sensibly be used for trend analysis.

But that pre-2021 data shows a significant and politically inconvenient trend, so it is ignored. Instead the post Q2 2023 showing an initial rise and then mere flattening-out is preferred.

Here's how the RTB described the 2021, 2022 and 2023 figures.

1773689203935.webp
 
Last edited:
Excellent analysis @polecon - you explain really well what has been going on.

As many of us suspected here, there wasn't a massive surge of additional tenancies, nearly 30k in total, between Q2 2023 and Q4 2024. If nothing else, no one noticed this huge increase on the ground. Posters here correctly deduced that it was landlords finally being able to register after the RTB's annual registration fiasco.

The RTB, who must have known the truth, used the 'increase' to demonstrate that the strictest rent controls in the world weren't causing a fall off in supply. When numbers started to fall again they used cost rentals and unregistered landlords who they caught to keep the numbers up. That they can only maintain the number at 240k with this massaging shows how strong the underlying trend of landlords leaving is.
 
Yes @polecon great work. FWIW I have a tenancy (one remaining) which in the RT number is 0223 (ie 2023) even though the start of the tenancy was one year earlier. This difference might well be due to the reasons you outline. However I have never been able to get this RT number to appear in a search on the RTB Rent Register, even where it should be a 100% match to the search criteria, which leads me to also suspect some filtering in the query results.
 
Great work @patfert1 !

With such a large spread, it really begs the question, what is a "market rent"!
I agree, its a reasonable question to ask having seen the nonsense register. At least the RTB have a definition on their website (worrying that I even need to go look for it)

What is market rent?
Market rent is the amount a tenant would reasonably pay a landlord for a similar home in a comparable area at the time the tenancy begins.
 
What is market rent?
Market rent is the amount a tenant would reasonably pay a landlord for a similar home in a comparable area at the time the tenancy begins.


The term 'reasonably' stands out to me in the legal definition. Why not just say "the amount a tenant would pay a landlord for a similar home..."?

My fear is that the rent register will be used as the test of what is reasonable.
 
My fear is that the rent register will be used as the test of what is reasonable.
I think that you are correct there. The whole purpose of the Rent Register is to impose what the RTB view as the 'correct' rent.

Your tenant agrees to rent your house for €2000. He waits 6 months and asks the RTB to investigate. He says that he had to agree the rent because he'd applied to hundreds of properties and yours was the only offer he got. The RTB then agree with him that the rent charged is not 'reasonable'. You're stuck now with a lower rent possibly permanently, you're fined and you have to repay the tenant the overcharge.

The legislation doesn't say that the rent is what the tenant would pay, it is what the tenant would reasonably pay which is something quite different. It is very easy for the RTB to rule that the rent charged is unreasonable. What are you going to do - take the RTB to the High Court. The High Court will very likely say anyway that the RTB are the experts here.
 
The term 'reasonably' stands out to me in the legal definition. Why not just say "the amount a tenant would pay a landlord for a similar home..."?

My fear is that the rent register will be used as the test of what is reasonable.
Or one could look at it as an attempt to distinguish between what is reasonable and what is unreasonable.

If so it's crude attempt but not with out some merit, insofar as its trying to protect some tenants who maybe in a difficult situation and could be exploited.

The excellent and generous work of patfert1 means that mining the register is easier such that in most cases we can identify meaningful comparables.
 
If so it's crude attempt but not with out some merit, insofar as its trying to protect some tenants who maybe in a difficult situation and could be exploited.
Such protection is important. There are vulnerable tenants who could be exploited. It is important to have a mechanism to prevent that from happening.

That said, based on what I have seen to date with the Minister, the RTB and the Department of Housing I don't believe for a second that that is the only reason for the inclusion of the word reasonable.
 
Or one could look at it as an attempt to distinguish between what is reasonable and what is unreasonable.
Per the above "Legal language uses "the reasonable" (and the "reasonable person" standard) to provide a flexible, objective, and fair benchmark for evaluating human behavior across countless, unpredictable situations. It bridges the gap between rigid, absolute rules and the practical need for justice based on the specific circumstances of a case. "
 
What is market rent?
Market rent is the amount a tenant would reasonably pay a landlord for a similar home in a comparable area at the time the tenancy begins.


The term 'reasonably' stands out to me in the legal definition.
Where does it say reasonably? Is that just on the RTB website? I can’t find it in the legislation.
 
I agree, its a reasonable question to ask having seen the nonsense register. At least the RTB have a definition on their website (worrying that I even need to go look for it)

What is market rent?
Market rent is the amount a tenant would reasonably pay a landlord for a similar home in a comparable area at the time the tenancy begins.
WHen I wrote my comment after seeing such a large spread of rents in that set of data, I was aware the RTB has a definition. As others noted, the problem is that you have to back-up your rent level with 3 from their 10 matches. And as I can find by trawling the data, there is actually a really wide spread of rent levels, in many cases far beyond what you can see in 10 matches. In some cases the wider set of data I get has no clear curve, in other cases a nice bell curve... but what is a "market rent" even if you have visibility to such data (which the RTB doesn't give in its 10 matches)... is a market rent (or "reasonable" market rent) the median, the mean, the 75th percentile, average of the top 3, average of the rentals from the past 3 months, .etc ... it's all completely open to interpretation and of course large funds managing entire apartment blocks will be much better prepared to argue their case than the small landlord with 1 or 2 properties in the countryside.
 
Not that it could not be an issue if contested but it was made very clear at the webinar that any 3 property could be chosen, even the top 3.
I'm not sure I'd rely on what the RTB say in an unrecorded webinar especially where possible criminal prosecution is involved.
 
What is market rent?
Market rent is the amount a tenant would reasonably pay a landlord for a similar home in a comparable area at the time the tenancy begins.

That isn't the legal definition. That is what the RTB want it to be. By making it "reasonable", they make it subjective, so they can use their discretion.

Act said:
In this Part “market rent”, in relation to the tenancy of a dwelling, means the rent which a willing tenant not already in occupation would give and a willing landlord would take for the dwelling, in each case on the basis of vacant possession being given, and having regard to ...

It is how much would a tenant (who doesn't already live in the house) pay to a landlord to rent the house.

Market rent isn't subjective, it is where supply and demand balance.
 
Back
Top