Wednesday, May 12, 2010

Starts / Completions per New Person

Relating to the conversation in the previous post comments, here is a chart showing the ratio of housing starts and completions to new residents.


Van Housing Blogger said...

And further, the five year absolute (not percent) growth of population for 2000-04 was the LOWEST of any 5 year period in the data. We gained 119,138. The 2nd lowest was 1980-84, with 171,359 people. The peak was 90-94 with 390,629.

So, population growth did NOT spawn this boom.

It is true that pop growth did better in 2005-09, at 276043. But as Mo's graph above shows, we built more than enough for those guys!

RentingSucks said...

Also household size is one of those 'leveraged' numbers. When times are good and money cheap people might buy or rent a condo for themselves but when things get bad they start to double up (roomates, moving back home etc.) A small change in this number can have a big effect on whether there is enough supply or not.

Dave said...

276,000 is a huge number anyway you slice it. That's more than half the population of Vancouver added in that period! Think of all the housing downtown and around Vancouver. Now imagine having added half that in just a five year period. That's a lot of demand!

vibe said...

Dave, I was referring specifically to this part of your comment:

"Let's compare 1958 to our recent construction peak in 2008. Population went from 1.5million to 4.5million. Construction starts in 1958 were 13,000. To compare, we need to triple that number, which gives us 39,000."

You are saying that because the population is 3 times larger in 2008 that we would need 3 times the housing starts. That is nonsense. You need to compare starts to population growth because these are both rates of change (whether in absolute or percentage terms), apples to apples. Math is more fun when you do it right.

Dave said...

Vibe, good point. But, we also can't ignore replacement construction. I think my math still stands. 35,000 starts per year is in the ballpark for a sustainable rate. If starts were up to 50k or 60k, then I would say it's too many. If we are plus/minus in the range of 35k, then I don't think builders have overextended themselves.

RentingSucks said...

Looking at your 35000 number another way. 35000 x 2.6 (GVRD members per household) = 91000 people housed each year. This means we would add another million people in 11 years. Seems a little high to me.

patriotz said...

But, we also can't ignore replacement construction

There is virtually no "replacement construction", properly speaking. Almost every time a building is torn down a larger building replaces it, i.e. supply increases.

Dave said...

Not for Single Family Homes. Every new single family home west of the Port Mann replaced a prior home.

Not much of a name said...


One item that you seem to be ignoring is that the vast majority of replacement construction will have suites.

I will use my neighbourhood as an example. In the past two years, there have been four new home constructed. Not one SFH had a suite prior to the new home being built. All four now have legal suites. That is a 100% increase in housing units.

vibe said...

Are legal suites not included in housing starts? If a larger building is built do they count that as one start or do they count the number of units in the building?

Either way I think the biggest flaw in Dave's replacement estimate is the assumption of a constant level of starts. If houses last 100 years, then we should expect houses built 100 years ago to be replaced now. Call me crazy but I really don't think they were building 20,000 houses a year in 1910.

All of this analysis is completely unnecessary of course. Clearly the ratio of new residents to starts should remain fairly constant. We are currently well below the mean and that means we are over building. VHB and Mohican already did all the work, why are we wasting our time with these back of the envelope calculations?

JimTan said...

“Not one SFH had a suite prior to the new home being built. All four now have legal suites. That is a 100% increase in housing units.”

On the other hand, there are more apartments/condos being built these days compared to 20 years ago. Apartment/condos have fewer rooms than SFH. This means that the people capacity for a year's housing start is lower.

Let's think this through carefully.

Not much of a name said...

Jim don't confuse the issue. Dave was making the point that for every SFH that is replaced with new construction there is zero net increase in the housing stock.

I'm saying there is an increase due to the fact that the replacement SFH are not SFH but multi family homes. The four in my neighbourhood, which is a very small sample, all have suites. The housing stock has increased with the new construction.

JimTan said...

Not much,

I suggest you do the research before jumping to conclusions. Condos are a larger percentage of housing starts. Some of the condos are one bedroom. So, there are several issues at play.

The bottom line shows up in the vacancy rate. That's the excess capacity.

Not much of a name said...


Leave the condos out of the argument for a moment. You are entirely missing the point. This comes down to Dave stating that there is no net increase in housing when you look at SFH replacement. I'm saying that he is wring in that many new houses contain suites that previous houses did not have.

Condos have never entered into this point.

Anonymous said...

Just something for the bears to chew on. I'm not saying the bulls are right or the bears are right, just some points that I don't see mentioned often. I constantly see that inflation adjusted prices are way overextended, the issue with that argument is judging inflation is not an exact science by any means. To be honest the best measure of inflation is probably gold prices, and Vancouver real estate is NO WHERE CLOSE to record highs priced in gold.. so how extended are we really? Also with the situation in Europe it's going to be difficult for central banks to raise rates while Europeans may look to move to a more stable region of the world and Canada is one of the best.. what city pops into mind if moving to Canada from Europe? Vancouver.... Furthermore, gold prices are suggesting rampant inflation, if we have high inflation along with a currency contagion in which central banks have trouble raising rates dramatically, you have a perfect storm for continued higher real estate prices. Just food for thought, not saying any of these scenarios will play out.

johnnyrent said...


You're smoking your laundry.

Cheers, JR

jesse said...

Population growth is a bit of a red herring. Talk to someone who lives in Melbourne about population growth and a truely crimped housing supply. Rents there are significantly higher than Vancouver.

There is absolutely nascent dark condo supply out there. I was in Asia recently and there are huge tracts of vacant housing being held by locals and foreigners alike. They treat it no different to a bank account with an annual service charge. It is unclear to me much of Vancouver's dormant housing supply will fall on the market all at once.

The true gauge of whether high prices are justified is based solely on what rents are doing. There is no shortage of housing in Vancouver or rents would be increasing to levels seen in Melbourne or other "world class" cities.

That housing starts don't seem to compute with population growth looks odd at the beginning but I think there is significant capacity being added by the advent of multi-dwelling suites where there used to be a true "SFH" before.

In Vancouver proper the term "SFH" can only truly be applied to old stock and luxury residences. These days the term "detached" is rightly used by the real estate associations and economists alike. The vast majority of new detached housing are not much more than freehold townhouses. The prices vis a vis stata townhouses -- that arguably have more utility and privacy than the standard detached Vancouver special -- pretty much confirms this.

JimTan said...

Ha! The dark condo myth again. See here for research published May 2009

patriotz said...

Melbourne doesn't look much more expensive to me for comparable properties. Certainly not NYC league. Note the rates are weekly:

Melbourne city rentals

RentingSucks said...

Ha! The dark condo myth again. See here for research published May 2009

If your read to the end of the thread someone points out that it was only 2400 units in 13 buildings. I wouldn't call this definitive especially since you don't know whether the conductor of the survey was impartial or not. It is also quite a small sample size. Ignoring that criticism 5.5 to 8.5 percent of condos seems significant to me. That's a lot of condos. If even 1 percent of all condos came on to the market rapidly the market would be flooded. So I think there is still some validity to this observation and I wouldn't say it was definitively busted.

JimTan said...

"If your read to the end of the thread someone points out that it was only 2400 units in 13 buildings... "


It is the only empirical study thAT I know of. That's better than opinion-bias-opinion.

2400 units downtown is a reasonable sample. Opinion polls sample 1/10,000 or less of the population.

JimTan said...


Make that 1/1000 of the population

buff_butler said...


"To be honest the best measure of inflation is probably gold prices"

No... This is a common misunderstanding. Gold does well durring _currency volitility_. Increase in gold != increase in inflation.