Ireland's Supposed Housing Deficit

[schmittel]

I've asked this question elsewhere and never really got to the bottom of the conundrum, so wondered can anybody here explain.

If you were trying to work out the deficit or surplus relating to the required quantity of anything the first thing you would do is count what you already have in stock. This is not only instinctive and logical, it's one of the first things we learn in maths eg:

Jonny has 5 sweets. He has 3 friends coming to play. He wants to make sure he and his friends have 2 sweets each. How many more sweets does Jonny need to buy?

It's simple. He needs 8 sweets in total. He already has 5, so he needs to buy 3.

5, what he already has, is a very significant number in this calculation.

Amidst all the different experts who have calculated Ireland's supposed housing stock deficit nobody has ever taken in to account either (a) the existing habitable housing stock or (b) the existing population.

If you take the data from Census 2022 housing stock and population, there is no deficit.

This appears to be the inconvenient truth that explains why nobody is willing to use these figures as the basis of their calculation on our supposed apparent housing deficit.

Or am I missing something?

[NattyO]

If you use the 2022 Census numbers, Ireland has a total of 2,112,121 permanent dwellings. According to the same census, the population was 5,149,139. This, by my calculations gives us an average of 2.44 people per dwelling. However, according to the ESRI, we have an average of 2.74 people per dwelling - they actually cite the 2022 census for this figure, so I’m not sure where the discrepancy between their number and my number comes from:
https://www.irishtimes.com/ireland/hous ... 2%20census.

Anyway, either number puts us on the high, to very high side in a European context - the average number of people per dwelling in Europe is between 1.9 and 2.3, with the more advanced countries like Germany, Denmark etc. having the lower figure.

All this presupposes that everyone in the country at any moment in time is housed - either in a dwelling or in homeless/emergency accommodation, so in reality the only figure that matters is the number of dwellings vs the increase in population. Ireland is currently building less than 30,000 homes per year (it will be well under that this year, regardless of what fictional numbers they publish), while our population is growing by nearly 100,000 per year. That would give us a number between 3.3 and 5 people per dwelling - more than twice the European average.

[schmittel]

If you use the 2022 Census numbers, Ireland has a total of 2,112,121 permanent dwellings. According to the same census, the population was 5,149,139. This, by my calculations gives us an average of 2.44 people per dwelling. However, according to the ESRI, we have an average of 2.74 people per dwelling - they actually cite the 2022 census for this figure, so I’m not sure where the discrepancy between their number and my number comes from:
https://www.irishtimes.com/ireland/hous ... 2%20census.

The Census figure for average household size is 2.74. The discrepancy from 2.44 is due to the fact that household size is calculated from number of occupied dwellings rather than total housing stock. Obviously for a variety reasons a number of houses will be unoccupied at any given time. Ireland has a very high number of empty houses.

Anyway, either number puts us on the high, to very high side in a European context - the average number of people per dwelling in Europe is between 1.9 and 2.3, with the more advanced countries like Germany, Denmark etc. having the lower figure.

The ESRI study cited in the IT article you linked specifically found that overwhelming driver of high household size in Ireland relative to European peers is due to the fact the we have more children per couple here. All the media chat narrative would have you believe that our high household size is caused by a large number of adult children living with their parents, but ESRI's findings are it is in fact because of the large number of children under 18 per household.

https://www.esri.ie/publications/househ ... try-trends

All this presupposes that everyone in the country at any moment in time is housed - either in a dwelling or in homeless/emergency accommodation, so in reality the only figure that matters is the number of dwellings vs the increase in population. Ireland is currently building less than 30,000 homes per year (it will be well under that this year, regardless of what fictional numbers they publish), while our population is growing by nearly 100,000 per year. That would give us a number between 3.3 and 5 people per dwelling - more than twice the European average.

This is the point I am getting at with Jonny and his sweets. All the focus is on the number of new houses built relative to the population growth, totally ignoring how many existing houses we already have.

That doesn't make any sense. It's equivalent to Jonny saying "The only figure that matters is that I have 3 friends coming who need 2 sweets each, so obviously I need to buy 6 extra sweets."

It totally ignores the fact that he already has 5 sweets. It doesn't make any mathematical sense, so why are we so wilfully squinting to find methods that ignore the basic mathematics of the question?

[BrianD3]

I don't know much about this topic but well remember the era of ghost estates which isn't that long ago. Have we still got houses in the wrong places? is it a planning issue or a housing issue. If Jonny has 5 sweets but they are located 2 hours away, does he still have 5 sweets

Regarding NattyO's post above, I thought that Denmark and Germany would have smaller dwelling sizes than Ireland but according to stats we have both smaller average dwelling size and higher occupancy than those countries, a double whammy. Denmark's average size is around 50% higher than ours.

People like myself skew averages, I currently live alone in a 4 bedroom house with lot more underutilised space in the dormer, Lots like me. To try to continue the sweet analogy, this could be like Jonny having 5 sweets but keeping 4 of them in a safe and deciding that he needs to buy 7 sweets.

[NattyO]

The Census figure for average household size is 2.74. The discrepancy from 2.44 is due to the fact that household size is calculated from number of occupied dwellings rather than total housing stock. Obviously for a variety reasons a number of houses will be unoccupied at any given time. Ireland has a very high number of empty houses.



The ESRI study cited in the IT article you linked specifically found that overwhelming driver of high household size in Ireland relative to European peers is due to the fact the we have more children per couple here. All the media chat narrative would have you believe that our high household size is caused by a large number of adult children living with their parents, but ESRI's findings are it is in fact because of the large number of children under 18 per household.

https://www.esri.ie/publications/househ ... try-trends



This is the point I am getting at with Jonny and his sweets. All the focus is on the number of new houses built relative to the population growth, totally ignoring how many existing houses we already have.

That doesn't make any sense. It's equivalent to Jonny saying "The only figure that matters is that I have 3 friends coming who need 2 sweets each, so obviously I need to buy 6 extra sweets."

It totally ignores the fact that he already has 5 sweets. It doesn't make any mathematical sense, so why are we so wilfully squinting to find methods that ignore the basic mathematics of the question?

Interesting point re the average.

I think your analogy with Jonny and his sweets is incorrect for one reason - unless we believe that a large proportion of the new population growth will move into houses with other people who are already in situ, then the number of houses already existing doesn’t matter - the analogy would be more accurate as something along the lines of “Jonny had 5 sweets. His friends have eaten these sweets already, so how many does he need……”

[schmittel]

Interesting point re the average.

I think your analogy with Jonny and his sweets is incorrect for one reason - unless we believe that a large proportion of the new population growth will move into houses with other people who are already in situ, then the number of houses already existing doesn’t matter - the analogy would be more accurate as something along the lines of “Jonny had 5 sweets. His friends have eaten these sweets already, so how many does he need……”

Fine, ignore Jonny and his sweets! And forget about the existing population/housing stock for a minute. Assume that existing stock/population is in equilibrium and there is no deficit or surplus.

If you knew for sure that the population growth in the next 12 months would be exactly 100k, how many houses do you think we would need to build accommodate that 100k?

[NattyO]

Fine, ignore Jonny and his sweets! And forget about the existing population/housing stock for a minute. Assume that existing stock/population is in equilibrium and there is no deficit or surplus.

If you knew for sure that the population growth in the next 12 months would be exactly 100k, how many houses do you think we would need to build accommodate that 100k?

Depends!

If we want to accommodate people at the “ideal” of the lower end of the European average, then 100,000 / 1.9 =52,632
If we want to accommodate people at the higher end of the European average, then 100,00 / 2.3 = 43,478.
We could add those two numbers and divide by 2 to get an average of the average, but for arguments sake, lets say we need between 43,000 and 53,000 new houses for that level of population increase.

[schmittel]

I don't know much about this topic but well remember the era of ghost estates which isn't that long ago. Have we still got houses in the wrong places? is it a planning issue or a housing issue. If Jonny has 5 sweets but they are located 2 hours away, does he still have 5 sweets

In the first instance I was trying to simplify at a national level. Do we have sufficient houses in the entire country to house the entire population. If the answer is yes then we can drill down and find if there are more local problems.

People like myself skew averages, I currently live alone in a 4 bedroom house with lot more underutilised space in the dormer, Lots like me. To try to continue the sweet analogy, this could be like Jonny having 5 sweets but keeping 4 of them in a safe and deciding that he needs to buy 7 sweets.

This is kind of my point. I think all the data is pointing to the fact we have a sufficient number of existing housing units to house the existing population, and thus the problem is an inefficient allocation of the existing housing stock both in terms of location and occupancy.

That problem cannot be solved by simply building an extra 10k/15k/25k houses per year, and we'll never solve it if we are ignoring what the data is telling us.

And I think for some reason we are deliberately ignoring the data.

Consider if you'd never read a media article about the lack of new build supply, or listened to a radio show or a politician on the subject and somebody asked you to work out how many houses Ireland needed at a national level.

To solve that problem you start with a count of the total number of housing stock and a count of the total population.
Then you would factor in AHS and the number of vacant houses.

It's totally logical and exactly the way the experts used to do it. But if you any of the estimates of what we need to build in the future, absolutely none of them reference the current count.

[schmittel]

Depends!

If we want to accommodate people at the “ideal” of the lower end of the European average, then 100,000 / 1.9 =52,632
If we want to accommodate people at the higher end of the European average, then 100,00 / 2.3 = 43,478.
We could add those two numbers and divide by 2 to get an average of the average, but for arguments sake, lets say we need between 43,000 and 53,000 new houses for that level of population increase.

Fair enough it depends! I think 1.9 is unlikely because a reasonable amount of the 100k will be new born children! But I take your point.

But can we agree that 43k houses for 100k people would not represent crisis levels of shortage?

[NattyO]

Fair enough it depends! I think 1.9 is unlikely because a reasonable amount of the 100k will be new born children! But I take your point.

But can we agree that 43k houses for 100k people would not represent crisis levels of shortage?

It doesn’t matter if they are new born children or not - the average is the average regardless of age.

Nobody said 43k houses for 100k people would represent crisis levels of shortage, it is at the higher end of the European average, nothing crisis about it. In fact if we were building at that rate, it would represent a huge improvement from where we are now. We’re not though.

[NattyO]

Talking about a “logical” use of housing stock is, well, illogical.
It is similar to the argument that flooding Ireland with thousands of foreigners to fill roles in American MNC’s is the right thing to do for the economy - yes it may be good for “the economy” but is it good for the Irish people?
Housing distribution and allocation is not a science, it is based on people’s personal feelings, comfort and upbringing. If elderly people choose to live in a 3 bedroom house that they bought, raised children in, and want to die in, it may be “illogical” but the “logical” alternative - forcing them out so that a family can move in, is hardly desirable from a social and humane point of view. Similarly, it may be logical that everyone should be forced to live in whatever large town or city needs the most new workers this month, but that would make for a lot of unhappy people - many want to live in or near the place they were brought up, where their families and friends are etc.

Humans are illogical (as Spock might say) and trying to impose logic on them usually ends in disaster - as history has shown many times.

[schmittel]

It doesn’t matter if they are new born children or not - the average is the average regardless of age.

Nobody said 43k houses for 100k people would represent crisis levels of shortage, it is at the higher end of the European average, nothing crisis about it. In fact if we were building at that rate, it would represent a huge improvement from where we are now. We’re not though.

I agree, nothing crisis about 43k houses for 100k people - a ratio of 2.3

But the question is if 43k houses for 100k people does not represent crisis of shortage why does 2,112,121 houses for 5,149,139 people, a ratio of 2.4, represent a crisis level of shortage? Particularly when the ESRI say that when you control for our higher number of children, we are not out of kilter on AHS.

Because whilst I agree nobody said 43k houses for 100k people is a crisis, there are plenty who say the 2.1m for 5.1m is a crisis, not least the housing commission who claimed we had a crisis level deficit of 250k houses in April 2022.

If you read all the experts who publish figures about new build numbers required, they all say we need to build more because we have a deficit in the existing housing stock. But the maths doesn't stack up.

It's true we may not be building enough per year to accommodate the population growth per year, but it is not true that we do not have enough existing houses for the existing population.

[Headshot]

I been saying this for awhile now but Ireland desperately needs to introduce residential rules

A work migrate should not be allowed to barely be 5 mins into the country and allowed to buy a house

[knownunknown]

It's true we may not be building enough per year to accommodate the population growth per year, but it is not true that we do not have enough existing houses for the existing population.

I’m finding it a little difficult to follow, but just as an aside you need more houses than is required by the population for housing in order for people to be able to move around. In a healthy market there always must be a percentage of empty houses(or empty rooms). Not sure exactly what this percentage is but it’ll have a mathematical solution.

Think of it like musical chairs, if there are five people and five chairs no one is ever able to move around.

If we fill all the vacancies we encounter more problems than just how much accommodation is there in total.

Edit: a quick google puts this number at about 5%. https://www.nccor.org/tools-econindicat ... sidential/

[schmittel]

I’m finding it a little difficult to follow, but just as an aside you need more houses than is required by the population for housing in order for people to be able to move around. In a healthy market there always must be a percentage of empty houses(or empty rooms). Not sure exactly what this percentage is but it’ll have a mathematical solution.

Think of it like musical chairs, if there are five people and five chairs no one is ever able to move around.

If we fill all the vacancies we encounter more problems than just how much accommodation is there in total.

Edit: a quick google puts this number at about 5%.

Yep, it's generally agreed that between 5 and 6% is optimum level of vacancy to allow for turnover in sale and rentals, probate, renovations etc - basically just providing essential liquidity of stock in the market.

A vacancy rate under 5% would represent an undersupply of total housing stock. A vacancy rate above 6% would represent a housing stock in oversupply.

Census 2022 found the national vacancy rate to be nearly 8% at the same time the Housing Commission said we had a deficit of 250k houses! Go figure.

The vacancy rate (excluding holiday homes) fell to less than 8% in 2022 from over 9% in 2016 and 12% in 2011.
Vacancy rates varied by property type.

Vacancy rates were lowest among semi-detached properties at 5% in both the 2016 and 2022 censuses.

The vacancy rate among terraced houses fell from 10% in 2016 to 7% in 2022.

Flats in converted buildings and bed-sits were most likely to be vacant at 18% in 2022, down from 28% in 2016.

Vacancy among apartments in purpose-built blocks dropped from 13% to 9%.

https://www.cso.ie/en/releasesandpublic ... dwellings/

[knownunknown]

Yep, it's generally agreed that between 5 and 6% is optimum level of vacancy to allow for turnover in sale and rentals, probate, renovations etc - basically just providing essential liquidity of stock in the market.

A vacancy rate under 5% would represent an undersupply of total housing stock. A vacancy rate above 6% would represent a housing stock in oversupply.

Census 2022 found the national vacancy rate to be nearly 8% at the same time the Housing Commission said we had a deficit of 250k houses! Go figure.



https://www.cso.ie/en/releasesandpublic ... dwellings/

Can’t argue with any of that but to put some context, in 2022 we were coming out of lockdowns and tourism had been closed for a long time. We’ve since since a huge increase in tourism and a massive influx of refugees from Ukraine and other places.

What happens to a housing market when the vacancy levels drop too low? I guess we won’t know exactly for sure until the next census but you could predict the trends.

[knownunknown]

This is a worrying trend surely,

“ When house vacancy rates drop too low, demand for housing exceeds supply, leading to higher rents, increased competition, and potential displacement of tenants who cannot afford the rising costs. This situation reduces tenant bargaining power, potentially forcing people out of their homes and creating instability, as the housing market becomes less accessible to new residents and existing ones struggle with affordability”

The national residential vacancy rate in Ireland reached a record low of 3.8% in the final quarter of 2024, according to GeoDirectory and EY figures released in February 2025. The west of Ireland had the highest vacancy rates, with Leitrim at over 12% and Mayo at 10.6%, while Dublin had the lowest at 1.2%.

https://murphyoconnor.ie/2025/02/19/res ... ber%202023

[schmittel]

This is a worrying trend surely,

“ When house vacancy rates drop too low, demand for housing exceeds supply, leading to higher rents, increased competition, and potential displacement of tenants who cannot afford the rising costs. This situation reduces tenant bargaining power, potentially forcing people out of their homes and creating instability, as the housing market becomes less accessible to new residents and existing ones struggle with affordability”



https://murphyoconnor.ie/2025/02/19/res ... ber%202023

The measure of vacancy from GeoDirectory and the Census is very, very different, and the Census has a much higher count of vacancy.

The GeoDirectory figure is what they consider as long term vacant, so they exclude anything that might be vacant for instance for rent or for sale. When the Census includes all vacancies it is over double the GeoDirectory rate from same time period.

The problem is that people, such as those in the link you posted, quote the GeoDirectory figures and then sound the alarm because 3.8% is worryingly low.

Worryingly low relative to what?

Of course it is relative to the benchmark 5%, but that benchmark includes all the vacancies that GeoDirectory ignore. If you include those the vacancy rate is actually very high relative to the benchmark.

Again it is just another example of people squinting to make the narrative fit the data.