Fleeing Millennial Zombies

We are plagued with zombie factoids. You can kill them over and over again with evidence of their falsehood, but they keep resurrecting and coming back because they fit prevalent narratives.

The “fleeing millennial” is one of Vancouver’s key zombie factoids. The basic idea is that the city is getting so expensive that it’s driving out millennials. Vancouver is sacrificing its own future! The detritus of uneaten avocado toast will ruin our tidy streets! (I kid, that’s what seagulls are for). The problem with this factoid is that it’s false. I’ve previously attempted to kill this idea here, here, here, and here. But it keeps coming back. It’s a zombie; undead and on the lookout for brains to eat.

As factoids go, this one highlights a real and pressing problem: the local unaffordability of housing. But it does so in a misleading and harmful fashion. It’s misleading because there is NO data anywhere suggesting that we’re selectively losing people in the age groups reasonably corresponding to millennials. It’s harmful because every time it re-appears, the fine print reveals that it necessarily valorizes some millennials over others.

The most recent resurrection of the fleeing millennial zombie factoid arrives via the folks at Better Dwelling, running their headline: See Ya! Local Millennials Are Abandoning Toronto and Vancouver. The valorization of a particular set of millennials receives a nod in the headline: we’re dealing with “local” millennials. Within the body of the piece, this is fleshed out: we’re dealing with intraprovincial migration patterns, between Metro Vancouver and other parts of BC, and between Metro Toronto and other parts of Ontario. Intraprovincial migration leaves out international immigration and interprovincial migration, the two other main components of migration overall (as described by Statistics Canada). Given that Toronto and Vancouver are Canada’s premier gateway cities, this is a pretty big oversight! Immigrants flood into these cities, with some staying while others move on to other parts of Canada.

I’ll come back to this problem, but let’s start with Better Dwelling’s focus, on intraprovincial migration. Can I replicate their findings drawing upon Statistics Canada components of population change data (StatCan 051-0057)? Nope.


I can get the same pattern, but not the same numbers. Here’s what I get looking at intraprovincial migration (summing net change for the years and age groups covered and adding the historical context from 2002-2007 just because it’s there!) This worries me a bit, and I asked Better Dwelling for clarification, but they have yet to respond (I also checked my numbers with the ever-helpful Jens over at MountainMath, who also couldn’t reproduce Better Dwelling’s numbers).

At any rate, the pattern is the same. But the pattern is misleading. After all, we’re only looking at a single component of migration. What happens if we add them all in?


Wow. That looks quite different! If anything, net migration from millennials (age 20-34) is soaring to all new heights in the latest five year period estimation. It’s just that intraprovincial migration has a slight drag on this pattern (compared to a more steady drag across other age groups). What to make of this?

As a gateway city, Vancouver’s population growth is driven first and foremost by international migration. Both interprovincial and intraprovincial migration play a role at the margins. We are definitely not losing millennials. But what about the claim that we’re losing “local” millennials? It’s a weird claim. It’s likely the case that many of the millennials showing up as international migrants in one period might later show up as intraprovincial or intraprovincial migrants in another. Ideally that’s what gateway cities do: take in huge numbers of migrants, keep some, and send others toward the broader set of communities in the province (and the rest of Canada) looking for workers and growth. Personally, I happen to know at least one Chinese immigrant to Vancouver now working and residing in Kelowna. It might be fun to break down intraprovincial migration by place of birth or the like, but absent that kind of careful analysis there’s no reason to imagine intraprovincial migration patterns identify “locals” or are uniquely informative about the state of millennials in Vancouver or Toronto.

As a methodological aside, it also feels a little odd to sum up the components of growth from Statistics Canada models in the time periods specified without acknowledging that they’re calibrated to the census years (2001, 2006, 2011, 2016). So 2012-2017 provides an estimated breakdown of components of change drawn from the combination of a 2011-2016 comparison with various sources of administrative data. The 2016-2017 estimations are especially subject to change! Still, it’s neat to see how Stats Can does this work. As a bonus, here’s what you get in terms of Statistics Canada’s model of population change by age group in Metro Vancouver moving forward from the the Census year (2016) to the next (2017), but setting births aside.


What really jumps out is the contribution non-permanent residents are making! They don’t show up in net migration figures here, but wow! Look at how much they’re contributing to growth, especially in the late teens. See the CMHC’s write-up about the effects non-permanent residents might be having on housing demand, but keep in mind that many will transform (like I did) into permanent residents.

And in the meantime: watch out for zombies! Don’t let them near your brains!


[UPDATE: May 25th, 2018. This blog post was combined with a couple of others in its own The Tyee article! No, Vancouver Is Not Losing Its Millennials. Thanks to awesome local reporter Christopher Cheung for reaching out and pulling that together. After the Tyee piece went up, I’m told that Better Dwelling contacted them to report that they’d used the CANSIM table 051-0060 rather than CANSIM 051-0057, as above. This seems to explain the disjuncture in specific numbers, which is good! But the problems with ignoring interprovincial and international migration of millennials remain. Also, CANSIM 051-0060 covers a HUGE amount of territory. Rather than just looking at Metro Vancouver, it looks at an economic region defined as the “Lower Mainland-Southwest, British Columbia.” Unlike a metro regions, which build their integrity largely around daily commute-sheds, economic regions seem to work more like ways of dividing up the entire country into manageable, if somewhat arbitrary, contiguously defined units. So the economic region of “Lower Mainland-Southwest” combines the Census Divisions of Greater Vancouver, the Fraser Valley, the Sunshine Coast, and Squamish-Lillooet. Full Statistics Canada map here (it’s big!). The Immigration Settlement and Integration Program (ISIP) has put up a smaller map illustrating the economic regions of BC which I borrow below, showing what nice things you can do when you work to recognize and welcome international immigrants!]


BC Regions Map



Visualizing Succession

After my talk at the Vancouver Historical Society, I’ve been playing around with some of the old fire insurance maps for Vancouver. As it turns out, fire insurance maps are super-awesome resources for historical research. Here I want to explore ways of using them to visualize succession as an urban process.

Succession, meaning the gradual outward expansion of an urban core, was thought by many observers of the early 20th Century to be the key process driving neighbourhood change. For sociologists like Ernest Burgess and Roderick McKenzie, succession was a process of “invasion” from the centre outward through a series of concentrically organized ecological habitats. The central business district (Habitat I) was always attempting to move into the surrounding industrial and commercial areas (Habitat II), which in turn were always invading the areas of dense working class housing beyond (Habitat III), which were always trying to muscle in on the middle class and wealthy residential suburbs on the outskirts of the city (Habitat IV).*

“Succession” was a science-y and neutral sounding concept (having been drawn directly from plant ecology), while “invasion” was quite threatening, especially to the middle class and wealthy suburban neighbourhoods hoping to hold the urban core at bay (very much including its working class & minorities). With the invention, legalization and rapid spread of single-family detached zoning, the middle class managed to mostly surround and contain the urban core, as I describe in my recent talk and book. As a direct implication of halting the process of urban succession, constrained urban cores mostly grow up instead of out, and now we mostly talk about gentrification instead of succession as a key process affecting our cities.

But let’s go back to succession. Can we see it in action? Let’s start with an early residential development consisting of two blocks (subdivided by alleyways into six) on the outskirts of a rapidly growing city in the late 19th century.


The base layer here is an old fire insurance map from the Vancouver archives, though I’ve faded it out a bit to draw over top of it. Let’s highlight all of the houses.



We can see this is pretty much just a residential neighbourhood at this point, with the only non-residential use a church (lower-left). Most of the other buildings are sheds. But wait a minute, are these all single-family detached houses?



Trick question! Those weren’t invented yet! So nobody was recording whether or not single families lived in these houses, nor how they might have been subdivided. That said, it’s clear that four houses in the upper-left block were all connected to each other, making them recognizable as rowhouses. Meanwhile, there were at least two houses connected side-by-side on other lots, making them semi-detached houses (by modern Canadian census categories). Also notable: at least four lots had both houses fronting onto major streets and houses fronting onto alleyways – with the latter now identified as “laneway houses” in Vancouver. Lots of “missing middle” housing!

Importantly, these residential blocks were located on the outskirts of a growing urban core in 1889. What will become of them? Let’s jump ahead 25 years and see…



I’ve highlighted in red all of the buildings that disappear over the course of 25 years. Oh no! There goes our church! Over half of the houses are also gone. Devastation in Habitat IV! But at least the townhouses remain. Let’s fill in the neighbourhood…



Some of those old houses remain, and new ones have been added too, meaning that we don’t see a complete replacement of seemingly middle-class residential uses (Habitat IV). That said, for the most part torn-down houses have been replaced by much larger buildings that take up their lots and add abundant industrial, commercial, and office uses to the neighbourhood. At least three hotels (Alcazar, Cadillac, and Canada) join numerous rooming houses, with the latter being an especially important source of housing for the working class and poor (Habitat III). Industrial uses (machine shop, workshop, numerous printers, warehouse) also crowd the block, along with commercial uses and services (second-hand shop, steam laundry, tailor, restaurant). There’s even an undertaker with lodging above! Lots of what Burgess termed a zone in transition (Habitat II). And offices… so many offices. We can make a strong claim for central business district uses (Habitat I). In effect, the entire urban core has moved into a former middle-class residential district over the course of 25 years, give or take, without completely destroying it. On a first pass, pre-zoning urban succession looks more like urban mixing than invasion and replacement. But maybe we just didn’t give it enough time.

What does this marvelous neighbourhood look like now, some 100+ years later?**



It looks like the two blocks bounded by Pender, Hamilton, Dunsmuir & Richards, right smack in downtown Vancouver. The 19-story BC Hydro Centre Tower dominates these blocks. But remarkably, at least nine of the old buildings that replaced the original houses in the neighbourhood during the 1906-1914 era remain intact! Alas, the townhouses in the upper-left have been replaced by a parking lot. Overall, the neighbourhood still has a mix of low-income housing, commercial and service space, and office space, though probably not a lot in the way of middle class housing (have to seek out some condos nearby for that).

I grabbed the dates on all the buildings I could find from VanMap‘s Assessment data. Old Fire Insurance Maps can be found in the Vancouver Archives.

Here’s the Fire Insurance Map from 1889, with zoomed in panel below:


Here’s the Goad’s Atlas Map from 1910-1920, with zoomed in panel below:


This is a first-pass, of sorts, at drawing upon Vancouver’s old fire insurance maps to get at patterns of urban succession in the days before zoning. I’d welcome suggestions, collaborations, and better visualizations! (This was pretty much all done with powerpoint on the fly). In the meantime, here’s my cheap-o little animated gif of the slides above. Enjoy!




*- The Chicago School of Sociology modeled this process on Chicago itself – though McKenzie, it should be noted, also detoured to Seattle in describing “The Ecological Approach” to the study of The City (1925), in a volume for which Park & Burgess are usually given first billing.

**- The maps date from 1889 and somewhere between 1910 and 1920, which is why I say “give or take” in assessing the length of time that’s passed! Though the Goad’s Atlas dates itself as 1910, it was reproduced (with consolidated updates) in 1920, and includes buildings up to at least 1914, which is why I can narrow it down to around 25 years after 1889.


Talking cool old maps

Here’s my talk from February 2018 for the Vancouver Historical Society, now preserved for posterity! I was really delighted to receive the invitation, and the crowd was fun. I’d really encourage others to check out their series of talks.

I concentrated on the earlier sections of my book, and found an awesome old fire insurance map of land use for downtown Vancouver in 1889, available at the Vancouver Archives. I used it as an illustration of how land uses mixed in Vancouver’s early years, much to the concern of reformers and the evolving middle class. Now, of course, it’s the block of Chinatown bounded by Main St (then Westminster) between Keefer and Pender (then Dupont). I’m hopeful I can track down more of these maps soon. Tips welcome!



The Great Wait: Changes in Timing in BC’s Birth Rates

While putting together slides for my life course class I returned to BC Stats data on age-specific birth rates. It’s really nice data, broken down by local health area. I’ve played with data on the Total Fertility Rate before. This time I wanted to highlight a far simpler transformation in birth rates that I’ll call the Great Wait!

What is the Great Wait? Basically, it’s the transformation in age-specific patterns of childbearing, whereby most women are having children later and later in the life course. When I was playing around with the BC Stats data I accidentally produced a chart illustrating the Great Wait, and I just thought it was too beautiful not to share.


Notice the gradual shift from peak childbearing in ages 25-29 (in 1989) to peak childbearing in ages 30-34 (in from 2003 onward). By 2005, more 35-39 year olds were having children than 20-24 year olds (so called “geriatric pregnancies” – which is like seriously a total FAIL in medical terminology). By 2010, the birth rates for 40-44 year olds began exceeding those of 15-19 year olds. We have fewer and fewer teen moms, and more and more new parents in their forties.

There are many interesting causes and implications of this shift. On average women are taking longer to develop their education and careers before having children than ever before, facilitated by improved contraception and assisted reproduction technology. It may also be that women just don’t feel as ready to settle down into motherhood as they used to – either because the alternatives remain too interesting or because they don’t feel prepared for the job of being a parent yet (I’ve explored this latter explanation with respect to the role of acquiring housing as a stage prop for the role of parenthood here in my academic work).

With respect to the implications, some of the childbearing delayed will inevitably be childbearing denied, as later-life pregnancies are biologically less certain for women, and some new risks are entailed. But on the whole, having children later means parents tend to be more committed and more prepared, with more resources at their disposal to help care for their children. Not a bad thing. On a technical note: the ongoing shifts in the timing of when women have children somewhat artificially inflate the magnitude of recent fertility declines. This is to suggest that 1.4 children (our estimate of the number of children women in BC have on average based on TFR measurement) is likely somewhat lower than the number of children the average of any given cohort of women will ultimately end up with. It’s kind of a demographer fixation.

If the Problem is Speculation, then Why Focus on Foreigners?

Ok, I’ll admit it. I’m an immigrant to Canada. And from the US, which has recently seen a surge in anti-immigrant and anti-foreigner rhetoric and mobilization. We saw the same, of course, in the UK leading up to Brexit, and there are outbreaks in many other places around the world as well. Anti-foreigner rhetoric is having its populist moment. And I don’t like it.

So I’m especially sensitive to anti-foreigner rhetoric. Combine that with my research interests in housing and immigration and my love for my adopted city of Vancouver, and well, you can probably see where this is going: I think we should stop our destructive focus on blaming “foreigners” for our housing problems.

This doesn’t mean we should stop paying attention to the impact of immigration policy and global flows of capital (and there are many good reasons to oppose wealth-based immigration policies like the investor program still operating out of Quebec!)* But it does mean we should stop using “foreigners” as our go-to explanation for sky-high housing prices.

Why? First, it’s dangerous. It should be patently obvious at this point that anti-foreigner rhetoric is the handmaiden to fascism. If I need to explain why that’s bad news, go visit some other blog. Second, it’s sloppy. The concept of “foreignness” is not well-defined in rhetoric, readily lends itself to racism, xenophobia and related “other”-blaming, and remains inconsistent with rhetoric when applied to policy. Let’s take Andrew Weaver’s recent foray into advocacy for a ban on foreign ownership as an example. He defends the ban this way: “We are delighted for people to come and work and live and own property in B.C. but it’s not okay for people to park capital here with no intention of living here.” I’m on-board with carefully tracking flows of capital, but does “foreignness” now apply to folks from Alberta and Ontario? And how do we square “foreignness” with the complexity of immigration? (I worked here on a visa for three years prior to becoming a permanent resident and then citizen, fully intending to live here all the while). Analyst Andy Yan offers a different, but related take for his focus on foreign-buying data: ““It comes down to a question of fairness. As people struggle with keeping their existing home or even having a home in Metro Vancouver, is it fair to treat them in the same way as someone who has a secondary or tertiary home here?” This is an entirely legitimate question, but it has little to do with foreignness, and everything to do with speculation and investment. To put the matter differently, if we’re trying to counter the possible negative effects of speculation and counteract wealth inequality, then I’m in, but if all we’re doing is coming up with policy responses to favor Canadian speculators over foreign speculators, then count me out.

And just what is the balance between “foreign” speculation and Canadian speculation? This is a tricky question to answer, in no small part because we DON’T track global capital very well. Like, at all… I’m all on-board at doing better with this – let a thousand Panama Papers blossom and the CRA take notice! But for now, instead of tracking capital, we track bodies and assign them to residence and rights of citizenship in complicated fashion. We have legislation in BC distinguishing “Foreigners” on the basis of PR and Citizenship status, regardless of their actual place of residence (our “foreign-buyer tax“). And now we have Andrew Weaver extolling the virtues of New Zealand’s proposed “foreign-buyer ban“, which is a little more complicated, but linked to assigned residency (e.g., temporary residents on visas can buy a home, but have to sell it when they leave). The latter issue of residency is what was kinda-sorta measured by Statistics Canada in its recent release of data on non-resident (“foreign”) property ownership (CHSP). This is the data Andy Yan’s been playing around with. But to date most write-ups of this data that I’ve seen, including that from Statistics Canada, have simply focused on the comparison between non-resident (outside Canada) and resident (in Canada) property owners. When we use this as a proxy for talking about speculation, the implicit assumption seems to be that everyone outside of Canada is a speculator/investor/vacation home owner, and everyone living in Canada owns only the property they live in. We lose sight of domestic speculation and investment. How do we fix this?

One answer is we combine this information with Statistics Canada’s Survey of Financial Security (SFS). This handy little survey differentiates between principal residence and other real estate for participants, including those in metropolitan Vancouver and Toronto. Here I’m going to combine these datasets together in a very simple and replicable fashion to divide up the value of residential real estate into principal residence, other Canadian residence, and cross-border (Non-Canadian) residence. Effectively I’m just taking the total value of resident real estate holdings from CHSP and subtracting the Principal Residence real estate holdings value estimated from SFS to get the Other Canadian Residence category. This enables us to compare domestic speculators/investors/vacation home owners with foreign speculators/investors/vacation home owners. So what does total property value look like broken down into Principal Residence, Other Canadian Residence (a.k.a. “domestic investors”), and Non-Canadian Residence (a.k.a. “foreign investors”)?

Something like this:


Principal resident owners are separated out in purple, leaving only the properties held by investors / speculators / vacation home owners and the like in green. These are further distinguished between owners residing someplace else in Canada and owners residing someplace else across the border. In Toronto, these two groups are closely matched, though domestic investors look just a little more prominent than international investors and both are dwarfed by principal resident owners. In Vancouver, overall investment in the real estate market by non-owner occupiers is much, much larger, and domestic investors account for the lion’s share of the investment. Many of these investors, no doubt, live in Vancouver (where the SFS data estimate around 1 in 5 families own a second property besides their principal residence). Others live in Toronto, Calgary, and elsewhere across the country.  But with respect to Vancouver’s place in Canada the following seems clear: If speculation is the problem, then its largely home-grown.

There are some decent arguments out there for favoring the buyers in purple over the buyers in green. We’ve got lots of policy options for reducing the value of BC properties as investment vehicles – many having to do with tax policy. There are even better arguments, I think, for focusing on providing lots of affordable rental and non-market housing options for people so that ownership isn’t the only decent game in town. But what arguments do we see for favoring the folks in dark green over the folks in light green? If the important thing is to prop up the middle class in their entry into home ownership, why pit middle-class home ownership against foreigners instead of against the far broader class of investors and speculators as a whole? It’s for BC’s Green Party to answer that one, I guess…


*- I’m proud to have elicited the approval of at least one prominent academic I admire by ending my recent co-authored piece on Chinese-Canadian immigration with the line: “Investigations into the work of home-making suggest how Canada’s immigration policies contribute to rather than ameliorate global inequality, revealing an invitation that reads: give us your energetically leisured, your wealthy, your elites yearning to breathe freely.” Yes. We can do better, including taking in lots more refugees instead of fast-tracking “investor-class” immigrants!

NOTE: Inspired by the transparency of Jens von Bermann’s gitHub kits, but not knowing how to do that, please see my following spreadsheet if you’re interested in the data extracts I worked with and how I produced the above! Feel free to check my work and by all means let me know if I missed something!


The Things I Teach

I’m archiving my syllabi for current and recent undergraduate courses here on the blog, both for (ungated) student use and for public consumption. My courses all combine interactive lectures with student-led reading group discussions and some form of sustained research or building project.

Built Environments (2018) UBC SOCI 364: Syllabus-BuiltEnv2018

Sociology of the Life Course (2018) UBC SOCI 324:  Syllabus-LifeCourse-2018

Urban Sociology (2017) UBC SOCI 425-A:  UrbanSoc-Syllabus-2017

In the recent past, I’ve also taught graduate level courses (especially in Urban Sociology) and our undergraduate course in Research Methods. For other teaching scholars out there, please send me any suggestions for improvement! I’m especially interested in keeping my readings updated and interesting.

What Do Canadians Do All Day?

While preparing to teach my sociology of life course class, which is kind of like a sociology of time, I recently stumbled across the data tables from the Canadian Time Use Survey for 2015. This allows me to answer a very important question: what do Canadians do all day?

Of course different Canadians do different things.* But we can break them into a couple of broad groups to get some sense of the average number of hours they spend doing various things. It’s important to keep in mind that here that the averaging extends across both different people and different days (e.g. of the week, month, and year). Many people don’t record carrying out some of the activities described below at all, but they get averaged in with those who do. Others’ activities (e.g. paid work) may cycle so that active and inactive periods (like weekdays and weekends) average out (so a 40-hour work week averages to 5.7 hours a day).

Since I’ve got this life course theme and we’ve got data on age groups, let’s look at that. How does time use change between young adults, more middle-aged folks, and elders? We can split this out by gender too. I’ve color-coded activities so that sleep and personal care are in purple shades, work and study in green, travel in gray, unpaid homemaking work in orange, and leisured activities in blue. There are usually between one and two hours unaccounted for by this schema, coded in shades of off-white at the top.



Not surprisingly, young adults (age 15-24) spend more time studying than any other group. They’re the ones most likely still in school! But this time is balanced against working. Young men have strikingly more leisure time than young women, and they appear to spend a lot of it using “technology.” I suspect that means they’re playing video games for, like, an average of two hours a day, but there may be other interpretations (see footnote 24 in the original).

Folks closer to middle age spend a lot more time working, both for pay and at homemaking tasks (esp. childcare!) They also appear to spend the most time in personal vehicles. And they get less leisure time and less sleep than at other ages. But once again, men appear to get more leisure time than women.

How about those in retirement ages? Both men and women get a lot more leisure time (which they appear to spend largely watching TV). But do women catch up? Nope. Men still get the most leisure time, and the difference largely seems to come down to how much time is spent devoted to household chores instead. It’s certainly possible that some of those chores might also qualify as hobbies (e.g. cooking), and other things, like “personal care,” also help account for differences in time left over for leisure between men and women. But household chores and carework definitely remain gendered.

So, if you’re a Canadian man looking for a New Year’s resolution, here’s an easy one: take on more housework! (Note: this absolutely includes the author). The other big message for me is: I am not getting enough sleep! I mean seriously, I’m doing better than when my kids were really little, but 8 hours is still entirely aspirational.


*-An important caveat. As is so often the case, the General Social Survey doesn’t cover or claim to represent all Canadians.

The 2015 GSS collected data from persons aged 15 years and over living in private households in Canada, excluding residents of the Yukon, Northwest Territories and Nunavut and full time residents of institutions.