Sunday, February 07, 2010

Construction employment

Housing starts have really fallen off a cliff in the past year. However, we have not yet seen the last of the construction lay offs. Why not? Because construction employment relates to units under construction, not starts. Units under construction has fallen, but still has some way to go--we're maybe half way there. I predict it will bottom out below 10K/year. This means we still have a lot of construction jobs to be lost.


As for construction employment, it has dropped by around 20 percent since the peak. I think it has a long way to go--I would not be surprised at seeing 2001 levels again by the end of 2010.

That alone, all else equal, would be enough to push the unemployment rate back above 10 percent.

These rates do appear to have plateaued at current levels. However, once the Olympics are done and we see the continued completion of construction projects through the spring and summer, I would not be surprised at all to see 10 percent unemployment in BC by the end of the year.

17 comments:

Van Housing Blogger said...

Hey, can anyone tell me where I can get monthly REBGV total inventory numbers? Jesse and Mohican must have these numbers to form their so-informative MOI v. price change calculations, but I can't find the inventory count anywhere.

M- said...

Hi VHB, Agent Will posts the REBGV stats package on his website: http://agentwill.com/statistics/january-2010-vancouver-real-estate-stats-released/

Van Housing Blogger said...

M:

The REBGV reports do not, as far as I can tell, have the total inventory. They have the flow of new listings, but not the stock of total existing listings.

jesse said...

Try here for data back to 1999.

There has been some debate about how fast previous construction is now being absorbed by population growth. If I get time I'd like to analyze how possible it is to have completions run below the population growth rate for many years and not have a housing shortage. To wit completions through the late 1990s and early 2000s were very low yet people were not sleeping on the streets. There must have been overbuilding before then.

A sort of real estate dark matter conundrum.

RentingSucks said...

It's a little weird that if housing prices are still going up units under construction are going down. Shouldn't new projects have ramped up again if everything is fine?

Could this be a factor in propping prices up? Supply is winding down and we could undershoot again like we did in the late 90's. I was really thinking that the construction overhang would help in a rapid correction but a bunch of that supply is being absorbed by low interest rates. So will this ameliorate any correction?

Drove by motif in Burnaby and there are still real estate agents there despite line ups when the units came up for sale several months ago. Another weird data point that somethings a little off in this market but I'm not sure what it means.

jesse said...

"Supply is winding down and we could undershoot again like we did in the late 90's."

RentingSucks, to give a little perspective, dwelling formation in the boom between 2003 and 2008 was the same as in the mid '90s when population growth was twice as high.

Further, in the '90s, when population growth was twice as high, there was over-building that led to a lull in construction in the late '90s and early 2000s. Imagine what levels of population growth would be required to mop up the current oversupply in a year, or even 5 years for that matter.

One important note: demand for housing can never exceed supply or there would be people on the streets. The argument that the construction boom from 2003 onwards is "mopping up" the previous dearth is flawed.

From what I observe it is not the case that entire buildings or neighbourhoods sit vacant due to overbuilding as in many cities in the US. In Vancouver there HAS to be over-building based on the population and construction data. Units are therefore being absorbed by decreasing occupancy per dwelling. The income pie hasn't changed but now has to support many more plates.

For how long do you think that can go unchecked?

david said...

"One important note: demand for housing can never exceed supply or there would be people on the streets."

Not necessairly. What if people wanted their own place but could only find a place with roommates? Or they couldnt find a place to still lived at home. Or people would like to move here but cant find a place at a price they can afford.

I think it is possible for there to be more demand than supply, but eventually supply will catch up.

I agree that there is no supply shortage now and we are probably not in any danger of seeing a major housing shortage for several years. It will be interesting to see how long completed units are below population growth before we start to see people lining up for presales. I know that the lines didnt really have anything to do with low supply, more about increased demand through speculation.

JimTan said...

"RentingSucks, to give a little perspective, dwelling formation in the boom between 2003 and 2008 was the same as in the mid '90s when population growth was twice as high.

Further, in the '90s, when population growth was twice as high, there was over-building that led to a lull in construction in the late '90s and early 2000s. Imagine what levels of population growth would be required to mop up the current oversupply in a year, or even 5 years for that matter."

Numbers from Jesse please.

I would point out that the rental vacancy rate for metro Vancouver only rose +1% during the greatest crisis since 1930s.

That includes 6% in outlying areas like Surrey.

JimTan said...

P.S. Have you noticed that 'under construction' is consistently below 'completion' in the 1990s and reversed in 2000s.

Any idea what that is about?

JimTan said...

Note that Conference Board is calling for 4.5% growth for the province this year. Unemployment rate of 10%? How long will be last?

jesse said...

"What if people wanted their own place but could only find a place with roommates? "

But they would then have more income per dwelling as well. Real rents barely increased in the '90s. If there were a housing shortage requiring roommates, we would expect rising real rents or falling incomes.

More likely there was a population shift away from multi-bedroom housing towards single or double bedroom housing (i.e. condos). We know the data from areas with high % SFH dwellings show falling populations.

There was no housing shortage in the '90s. Near the end of the decade and into the 2000s, there was willingness and ability for people to move into separated accommodation, leaving bedrooms empty elsewhere. The increased supply manifested itself as underutilized housing, a condition still prevalent today.

There may be conditions where people are living in living rooms or being literal roommates (as opposed to flatmates) but I doubt this is any more prevalent today than 15 years ago. There were, though, more children at home 15 years ago. The shift into smaller accommodation could likely be from demographic shifts as children leave home. That mode eats up available dwelling supply but, as I mentioned, leaves parents with houses too big for their needs.

jesse said...

JimTan,

Sauder rental data

Vancouver CMA population data from BC Stats

jesse said...

JimTan to explain your question about the under construction versus starts/completions, remember that starts and completions are recorded as instantaneous events whereas under construction is active for a period of time. The numbers above are 12 months worth of data. It is intuitive that under construction can fall below the sum of starts from the previous year in certain circumstances, especially if projects are fast to complete

JimTan said...

Dear Jesse,

Forgive me if I should criticize your work methods. You need to compare the cumulative population increase to the cumulative housing starts/completion.

For example, population increased by 400k between 1990 and 1999. How does the cumulative housing data look? I think that there was a housing shortage in 2001. The CMHC report mentioned that Abbotsford rental vacancy was only 2.5%; below long term average of 4.3%.

It's true that population growth (300k) slowed significantly in 200-2009. At the same person/dwelling ratio of 2.5, our need for housing dropped from 16k/year (1990-9) to 12k/year (2000-9).

The problem may be that the housing mix has changed. In the 1990s, there may be more landed property with higher people capacity. In the 2000s, we saw a very large increase in condo building with fewer people capacity.

So, the issue cannot be resolved without a detailed analysis. And, it's the RE professionals who have the data and know the market best.

For myself, I place greater emphasis on recent data. We know that rental vacancy has increased by only 1-1.5% from the boom period. The question is how bad the housing starts shortfall affects the rental vacancy.

From the numbers, it looks bad. That could be a shortfall of 10k from 2009/10. That is twice the entire current rental stock (2009).

Note that housing starts was the same in 2001 and 2009. Therefore, housing starts has to jump in 2011/12 to catch up. Will it? That's the key to the medium term price trend.

patriotz said...

Units are therefore being absorbed by decreasing occupancy per dwelling.

Here's an anecdote for you. Couple I know with 20 something daughter, had lived in a bungalow in Burnaby for years. Didn't use all the space and rented out a bedroom to ESL students.

Guess what they did at the top of the market in 2008? Sold the bungalow and moved into a $800K house with basement suite in Cloverdale.

What on earth for? (Not for work, that remained in Burnaby/New West). Only conclusion is that they are "hoarding" dwelling space which is speculative by definition.

I don't think this kind of thing is an isolated occurrence. Combine these trade-ups with the boomers who are having their kids leave home (eventually!) and you are looking at massive pent-up, and unreported, supply for decades to come.

Dyugle said...

Home reno tax credit has ended. Completions should accelerate now.

jesse said...

"The problem may be that the housing mix has changed."

I agree however the mix works both ways. If existing stock is underutilized it will lead to a prolonged period of below trend construction. In addition detached properties are using space more efficiently -- the number of bedrooms is higher than in years past given multiple basement suites being common now where the norm was only one.

I disagree there needs to be a more detailed analysis. I see enough evidence that there are more bedrooms being produced than the population can handle, resulting in people spreading out. The result: low vacancy rates with low rental appreciation, which is exactly what we are observing today.