Demographic change isn't a long-term problem any more

It's now a short-term problem.

I spent the morning doing interviews on CBC radio morning shows in Corner Brook, Victoria and nine places in between. (For those of you who caught it, yes, I really do speak that quickly. I try to slow down when I'm on the air, but sometimes I forget.) One of the talking points was the PBO's concern that an aging population was going to slow economic growth and make it even harder to get rid of the deficit.

We've been talking about the problems associated with what happens when the baby boom generation retires for so long that it's become just part of the background noise in many policy discussions. Four years ago (one of my holiday reruns), I wrote

[E]veryone knows that the baby boomers are also approaching retirement, so the relative size of the working-age population is going to be decreasing. And it's going to happen soon: according to Statistics Canada's projections, the number of people between the ages of 15 and 64 will peak in about 5 or 6 years, and will then decline sharply. That extra 0.25% in growth is not only going to disappear, it's going to go negative: the decline in the size of the labour force will reduce per-capita growth rates by about 0.4%

Those projections were generated in 2005, and they may have been a bit optimistic. Here is a graph of the working-age population expressed as a share of the total:

Since 2005, the data have tracked the lower end of the 13 scenarios that StatsCan ran.

Yikes.

Comments

[Geoff NoNick]

A couple of questions/comments:

(1) Can you explain why the working population percentage was so low in the 1970s? Surely the baby boomers were all online by then?

(2) Can you discuss the dip that occurred in the 1990s?

(3) Is this an issue that's better addressed through increased immigration or policy changes that discourage labour-inefficient practices like, say, unionization (or are labour unions doomed regardless of policy)?

[Stephen Gordon]

(1) The front end of the boomer wave was on the market, but the boom didn't end until 1964 or so. The data start in 1971, so we don't see the initial surge in the graph.

(2) I think David Foot called this the 'Baby Bust'.

(3) Who knows?

[Patrick]

Yikes indeed. Does this imply inevitable contraction of the economy? And is it likely to cause deflation?

[Just visiting from macleans]

Good for you on the CBC thing. You've encouraged other academic economists to speak out/make themselves available to the media. How were you successful here, do you believe, in getting you opinions/expertise heard? Your teaching position? This blog? Referral? Just curious.

[Geoff NoNick]

To play devil's advocate, this may not be the end of the world, then. It seems we can look forward to an economic era in some ways similar to that of the 60s and 70s, when a large proportion of the population was not productive: in that case the non-productive portion of the population were children receiving costly education, rather than financially independent retired people.

It seems the questions we need to ask next are:

(1) How much more expensive is medical care for older people, versus education for children? To what extent can we ask terminal patients to subsidize their care using the assets they accumulated during their lifetime (for those that have such assets)?

(2) What were we doing in the '60s that enabled a relatively small workforce to support a large dependency-base? How can we leverage existing technology to do it even better?

[Matthew]

Which radio shows were you on? Would it be archived anywhere on CBC's website?

[Matthew]

"Is this an issue that's better addressed through increased immigration or policy changes that discourage labour-inefficient practices..."

The studies I have read seem to suggest that even significantly increased immigration wont alter the trends much.

[Doc Merlin]

Yikes indeed.

[Jim Rootham]

This is not a problem, this is a solution.

The fundamental driver causing systemic collapse modeled by "The Limits to Growth" was exponential population growth. To avoid that we need to get to population reduction, or at least stabilization. That implies having the kind of demographic distribution described above.

For those who will tell me the Standard Model of "Limits to Growth" did not turn out to be the case: a) I know b) please show me how the sensitivity analysis they did was wrong.

Hmm, fewer workers, lower labour supply, more non workers, increased labour demand, implies higher wages and lower unemployment. What's not to like?

[Andrew F]

"(1) How much more expensive is medical care for older people, versus education for children? To what extent can we ask terminal patients to subsidize their care using the assets they accumulated during their lifetime (for those that have such assets)?"

It's not just health care. We also have fairly generous PAYGO old-age transfers in the form of OAS. These payments will balloon as boomers pass age 65.

"(2) What were we doing in the '60s that enabled a relatively small workforce to support a large dependency-base? How can we leverage existing technology to do it even better?"

We had a lower standard of living in the '60s. It would also seem that the age of plentiful, cheap natural resources is drawing to a close.

[Stephen Gordon]

Jim, the issue is the composition of the population, not its size.

[Nick Rowe]

I hadn't realised the change in composition would be that big -- a drop of around 10 percentage points.

Right now we're at a sort of double-peak, because the leading edge of the boomers (born 1945) are just beginning to hit 65, and the trailing edge of the echo-boomers (kids of boomers) has all just entered working age.

My first thought was: "Right; simple solution is for people to retire later, especially since we all live longer now anyway, and especially women, who live about 5 years longer than men; so no retirement for you yet, Grandma!". But it would take another 8 years working to increase the percentage of "working age" by 10 percentage points (for an 80 years life expectancy). I can see people working a bit longer than they did in the past (especially women?), but I can't see another 8 years.

[RSJ]

What about immigration? If Canada really wants to change the composition of its population, it can do so.

[Stephen Gordon]

I dunno. I seem to recall from looking at this stuff a long time ago that immigration has more of an effect on population growth than on population aging.

[David]

It's interesting; I'm doing a population economics course right now so this is all stuff that's coming up in class.

Stephen is right; it's hard to alter the demographic structure dramatically through immigration. It takes a lot of immigrants to make a dent. And if you wanted to make the population younger through immigration, you'd have to change immigration policies to favour younger immigrants; current policies favour wealthier, more established immigrants, who tend to be more middle aged or older as opposed to young and fertile.

Also, that graph looks really dramatic, but ultimately you're talking about a 10 percentage point drop in the total dependency ratio over the next 50 years. It could be worse.

Finally, our prof made an interesting argument that people tend to focus on pensions and health care when talking about an aging population, but there are other services currently provided by government that are used less by older people than younger people (schools, roads, transit, welfare…). I don't know what studies he was referring to specifically, but he did say although population is expected to age, the projections suggest there won't be any effect on government expenditures if the per-capita age-adjusted spending stays the same.

Of course, public choice theory suggests it's easier to add more programs than cut existing ones, so whether governments have the gonads to close schools and run fewer buses to offset pension and health care expenses is a big question mark.

[Stephen Gordon]

Oh, those aren't dependency ratios. The aged dependency ratios start at 20% and go to 35-40% by 2030.

But those other points are worth remembering. Not all costs will go up, and some will go down.

[Mark Dowling]

"What about immigration? If Canada really wants to change the composition of its population, it can do so."

Get an MP to stand up in the HoC and propose the ending of family class sponsorship of parents over the age of 60 and ratchet down the maximum points age bracket for skilled workers (currently 10 points between 21-49, dropping to 0 by 54 which is a bit abrupt), and prioritising skilled worker applications from those aged 25-35.

See what happens.

[Geoff NoNick]

@Andrew F -

You're right - the difference is the welfare state benefits payable to elderly dependents that aren't payable to young dependents. OAS is an excellent example; it may be time to consider moving towards a needs-based OAS rather than the current semi-automatic OAS. It is, after all, "security". I would suggest that someone earning $60K in retirement isn't in need of state-sponsored security.

Most pension schemes aren't a problem, as long as they're of the defined contribution variety rather than defined benefit. It is unquestionably the time for public services to move toward defined contribution pensions (the private sector made the move a generation ago). It is argued that public servants "earn" these pensions in the form of reduced compensation while they are employed. That being the case, such a move should be accompanied by an increase in pay to public servants that will enable them to make additional retirement provisions for themselves, should they so desire. At least the actual costs of their employment are then known and crystallized in the current year, instead of being carried against a presumed growth in a future economy. Right now the public service pension is "long" GDP, when it should actually be "short".

I'm not convinced that lower living standard is the key to why we were able to support more dependents with fewer productive employees in the 60s. Bear in mind that women had not yet entirely entered the workforce, meaning that even more of the age demographic will be in the workforce in the future than in the past given comparable age structure. I would suggest that improved technology will allow us to support a comparable level of dependents at a higher standard of living than in the 60s. Unquestionably living standards will be affected, but they will come in the form of acquiring relatively fewer increases, not an absolute drop (anyone who has marveled at the primitivity of a European washroom will understand what I mean here).

I'm unconvinced that resource access will greatly affect our living standard going forward. I would argue that our heavy use of resources of a certain type in the past was largely wasteful in a way that reflected our lower technological level and the ready accessibility of them.

[Andrew F]

I just checked Budget 2009, and it projects "Elderly Benefits" (GIS and OAS, as far as I know) to grow rapidly from ~$32 billion in 07-08 to $42.6 billion in 13-14. I'm not sure if it will plateau then or continue a rapid rate of increase.

[Jim Rootham]

Ok, if we keep our eyes on the composition rather than the total I still see fewer workers, more labour demand and higher wages. Why is that not a good thing?

[Declan]

To the extent that there is an economic impact from a shifting percentage of the population that is working age, I wonder whether the greater effect is with the level or with the trend. I guess for now we'll just hope that the fact that the downward slopes on your graph match up fairly well with poor economic performance is just a coincidence.

[Nick Rowe]

Jim: from the perspective of us aging boomers, paying higher wages to you young people who are going to be providing us with the goods and services we need is definitely a bad thing!

But it's the diminished tax revenue from lower total output, plus the increased expenditures to keep us old folks alive, that are the fiscal problem.

[Frances Woolley]

I'm as worried about the composition of the working age population as the over 65 under 65 divide. Think about big innovative ideas - the PC and the mac, computers, software, stuff that really has a high value added - how many big ideas come from people over 50? Over 40? How many of us can honestly say 'my pay relative to my productivity has not increased as I've got older'? (Happily, I can't). Ending mandatory retirement in academia is a nightmare from this perspective.

From the budget balance stand point: raising the retirement age to 67 would make a huge difference and solve a lot of problems.

On education v. health care - the cost of education per person is fairly predictable and generally limited - how much are you ever going to spend per person? The cost of health care per person is unpredictable and we haven't figured out what the limit on it is yet.

[David]

Stephen, my bad. Your graph is basically an inverse of the total dependency ratio. So I meant to say dependency ratio going up, or your graph going down. Thanks for the catch.

[Nick Rowe]

Frances: "....raising the retirement age to 67 would make a huge difference and solve a lot of problems."

What would "raising the retirement age" mean, in practice? Would it mean you can't collect CPP and OAS (at full rate) until 67? Anything else? Or would it just act as a sort of focal point? Right now, 65 is still the focal point (it is for me anyhow), as the "natural" age to retire.

And, why/how does it make such a big difference? Wouldn't it just be like changing the working age percentage by (67-65)/(age till death) x100 percentage points? Say about 3%?

[Patrick]

Boomers vote, and there are lots of them. School age kids don't, and to a large degree neither do their parents (sadly). So I think education spending will be cut before health care spending. Especially higher education. Everyone knows (especially in AB) that those ivory tower dwellers are under worked and overpaid.

I think it would be sensible to raise taxes on accumulated personal wealth (e.g. estate taxes); the boomers are going to have to trade assets for health care. But it seems to me very unlikely that boomers will vote to increase taxes on themselves.

A two tier health system might help a little; make rich old people pay more. But good luck getting that through the HoC.

[Just visiting from macleans]

This will bring the two tiered health care debate to the fore - the growing part of the gov'ts budgets that is increasing faster than the rate of pop growth.

It would also lead to further prioritizing of services and delisting some that are too costly to publicly fund. How about the growth in hip/knee replacement surgeries? Private pay as you can clinics sprouting up moreso than now?

It seems to me I've seen studies that demonstrate a disproportionate amount of one's health care funding is in the latter stages of one's life - and to seniors who may have the means to pay out of their pockets(rather than pass on their wealth in inheritance).

Lots of Canadian values of universality may come into play.

[Just visiting from macleans]

Patrick and I think alike, and write at the same time I see.

[westslope]

So what's the problem?

More Canadian will be earning capital income than labour income. Those working will be older, more experienced and more productive.

What was the problem again?

[Just visiting from macleans]

A two tier health system might help a little; make rich old people pay more. But good luck getting that through the HoC.

Starve the beast at Provincial level for elective surgery or whatever(wait times increase) - where the jurisdiction lies for health care priorities and expenditures.

[Frances Woolley]

Nick, you cut right to the essential point as always. Raising the retirement age to 67 is indeed nothing more than a polite euphemism for restricting eligibility for OAS, GIS, CPP/QPP etc to 67.

Westslope, more experienced=more productive? Sure, that's conventional human capital theory, but is it true in your workplace? Not mine - and even for those one or two colleagues whose productivity has arguably increased with age, you'd be hard pressed to show the increase in productivity exceeded the increase in salary.

J.V.f.M, it's funny how people who think we have no two tier health care in this country can watch hockey games and just accept it without question when some hockey player is rushed off for an emergency MRI. You can go private for work-related injuries, which is the exception to the no two tier rule that's being exploited here (I wonder if Danny Williams could have claimed heart problems were work related?). The second tier of our health care system is Hospital USA. Without that escape valve, the political pressure for a two tier system would be irresistible.

Frances

[Patrick]

Frances : "more experienced=more productive? Sure, that's conventional human capital theory, but is it true in your workplace? "

If by 'more productive' you mean 'better at CYA and playing politics and spending all day in useless meetings', then yes, they are more productive.

;)

[Just visiting from macleans]

Yes, the existing two tier health system certainly was highlighted last year during the swine flu hysteria. It seems to me some medical person in Calgary even was disciplined/dismissed by the premier for allowing innoculation of the Flames while the public clinics were bottlenecked due to lack of delivery personnel/coordination. Same with companies that had private health practitioners. The whole debate made no sense as they were two separate yet parallel systems of flu shot/health care delivery.

[Jim Rootham]

Nick: I AM an aging Baby Boomer. I'm looking at this from a class perspective not an age perspective (and even there it's not so obvious where I sit).

How do you get lower output and increased expenditures at the same time? Even if you assert all the increases are on the government accounts there is still some natural balancing.

The taxes on those higher incomes will cover the costs.

The optional use of the US hospital system by Canadians accounts for about 1 hospital admission in 5000 (Michael Rachlis in the Star, the non optional uses are births while traveling and emergency room visits). Doesn't sound like much of a second tier to me.

[Stephen Gordon]

The problem is that with fewer people working, there won't be enough income being generated to redistribute to finance our current level of services.

Yes, wages will likely rise. But unless output per worker grows at a rate fast enough to compensate for the decline in the working-age population, per capital income will fall. Hence the preoccupation with productivity.

[Nick Rowe]

Jim: an increase in (real) wages per worker doesn't necessarily mean total wage income will rise, if there are fewer workers. And an increase in total wage income doesn't mean total income will rise, if all that happens is that the share of wages in total income will rise.

Plus, Stephen (and I) are coming at this from a long run aggregate supply perspective, where real income is determined by the supply side, and output=income must fall if there are fewer workers available. I may be misinterpreting you, but I think you are coming at this from an aggregate demand perspective, and implicitly assuming that increased government spending will increase income by reducing unemployment, and that total employment will rise even though the number of workers available will fall?

[himaginary]

Jim:"Hmm, fewer workers, lower labour supply, more non workers, increased labour demand, implies higher wages and lower unemployment. What's not to like?"

Dean Baker is saying almost same thing about Japan.

However, what everyone tends to overlook is that fall in working population means fall in aggregate demand. Yes, it may also mean fall in aggregate supply, but in developed countries that is not really a problem, as Nick noted here -- we have constant overcapacity in these countries. Overall economy size may shrink as labor force decreases, but it's not so big problem as imbalance between aggregate demand and aggregate supply (e.g. Japan is going to be surpassed by China in GDP this year, but that's not our biggest problem). On the other hand, retired people spend less than those in working age (see here), and this brings about overall demand shortage. This type of demand shortage may explain malaise of Japan since 1990 to some extent, as shown here.

[Jim Rootham]

Nick

Not quite, or at least, only partly.

I am mostly lookinjg at this from the position of the people at the bottom of the economic pile. I measure economic systems not by how high the highest can rise, but by how low the lowest get pushed. As you have demonstrated several times in this blog, the hightes can take care of themselves.

If labour shortages start constraining the economy I would expect that ultimately the response would be to increase wages. The wage increases would disproportionly benefit low paid workers, and I say, that is a good thing. An economy that paid more in wages and less in returns on capital than ours is likely to be a better economy, even if the the total per capita is smaller.

But yes, part of my thinking on this subject is conditioned by the fact that we are living in a demand limited economy, not a supply limited economy.

[Phillip Huggan]

Which healthcare services will get cheaper and which won't? Assuming equal importance to quality-of-living, the one's that don't could start to be funded now out of carbon taxes or undoing J.Flaherty's Oct/07 budget. The ones that do, stuff like active living and healthy foods, genome sequencing to find heredity-predispositions....who are we kidding? The goal of Canadian boomers is to consume carbon and live an obese suburban lifestyle. Maybe boomer children can send their parents to grade school to learn about AGW and cholesterol?

[Stephen Gordon]

Jim, what about retired people who aren't independently wealthy? I just don't see how the arithmetic would work.

[Jim Rootham]

Why are they any worse off under the new conditions than they are now?

We will be able to produce everything we need, so real shortages are not going to be a problem. The tax take from labour income will be higher than the current tax take from investment income (or at least it could be if we decided to) so OAS and GAINS will be funded.

I still don't see a problem.

[Phillip Huggan]

...and again, this can be part of industrial policy. The $10 billion taxpayer bailout could've been to retrofit bankrupt autos (at least perpetually wounded Chrysler portion) to building Honda palliative care robotics. There are amazing healthcare technologies coming out of U of T and St.Michaels hospital despite not enjoying tar and bank tax subsidies and indirect J.Flaherty big biz corporate tax reductions. Japan could use our investment and if/when it comes out of buying out the most carbon intensive properties in the country, boomer aging corporations would be better for Cgy/Edm/G.Prairie than failed 1970s oil industry copper investments. How come we aren't matching Obama's new home retrofit programme if we are apparently matching their cap targets? Cuckoo clock country.

[Stephen Gordon]

We will be able to produce everything we need

How is that possible with fewer people producing? (Unless, of course, productivity per worker increases enough to offset the reduction in the number of workers.)

[Phillip Huggan]

..sorry to spam the blog, but thinking about this, investments and laws that themselves turn future fixed costs into lower costs would be wise. Preventative healthcare, keeping fast-food 250m from schools (or whatever the magic distance is), subsidizing lean artificial meats, as you say Stephen, raising the mandatory retirement age...Kane Biotech is working on a process to dissolve biosludges that will plague hospitals. Every dollar of the $3M they've R+D'ed will pay dividends but because they make what boomers will *need* rather than the crappy CO2 cars boomers *want* (from some Cold War 1960s Hollywood pipedream?), they don't get bailout money. I wonder how much less stressful it would be to be a nurse if Medonyx got the $10B bailout or banker insurance to give all healthcare providers and first responders gelFAST (eventually a foreign supplier will pony up)? Medonyx would use the revenues to invent new health tech (car makers won't touch electric cars or public transit with their $$).
I'm not sure how much of the $2.3T-23T in needless melted Arctic costs will be borne by Canadians, but if the taxpayer choice is boomer palliative care or fixing the post-boomer climate mess...duh.

[Jim Rootham]

Because we can already produce way more than we need. We are demand limited not supply limited.
And yes, productivity will increase if it needs to to meet demand.

[Stephen Gordon]

Well, I'll leave it to you to make the distinction between what people want and what they need. Maybe you'll be the first central planner in history to get it right.

[Just visiting from macleans]

It can be easily modeled. Starting with a pyramid, how big is your aggregate trapezoid?

http://en.wikipedia.org/wiki/File:Maslow's_Hierarchy_of_Needs.svg

[Too Much Fed]

I agree mostly with Jim Rootham. I am glad someone else is bringing these issues up.

I want to add one question.

What happens if productivity and/or cheap labor from free trade allow for enough goods to be produced for the home country, while more workers need to be employed in the home country but there are fewer jobs available in the home country?

[Jim Rootham]

Well I seem to be being REALLY unclear. Central planning is the opposite from what I am arguing here.

My argument is that labour markets will operate better (from a socially desirable point of view) if demand for labour is high and supply is low rather than the other way around. Whether demographic change will actually bring that about is an open question.

[Just visiting from macleans]

Alberta is a good example in Canada of where demand for labour is high, and supply is low. And to partly meet the health care requirements of the growing population, inflated wages to draw skilled healthcare workers from other provinces ensues. Now for the belt tightening:

It's part of a cost-cutting exercise that has the province [Alberta]-- already facing a $4.3-billion deficit this fiscal year -- looking for potentially $2 billion in savings in the 2010-11 budget year.

The austerity measures could include direct program cuts, departmental efficiencies or protracting billions of dollars in capital spending.

Yet, the Tory government has already promised Alberta's $13-billion health system, one of the highest-funded per capita in Canada, another spending increase -- most likely at the expense of other departments.
...
Ken Boessenkool, an executive fellow with the school [University of Calgary's School of Public Policy], says Alberta coughed up more on program spending per person ($10,431) last year than all provinces but Newfoundland -- easily outpacing Ontario ($7,264), British Columbia ($8,127) and Quebec ($9,270).

http://www2.canada.com/calgaryherald/iphone/news/latest/story.html?id=2535483

Those natural gas prices better rebound soon, the royalty source of much of the Alberta gov'ts revenue.

[Too Much Fed]

Nick's post said: "Plus, Stephen (and I) are coming at this from a long run aggregate supply perspective, where real income is determined by the supply side, and output=income must fall if there are fewer workers available."

And there is the problem with almost all economists because they are still "living" in their post WWII world that IMO does not exist anymore. IMO, the "worldwide" economy is NOT running out of workers. Maybe china can make all the tradable goods the world needs by itself?

I also like this question. What happens if about 3 billion workers can provide everything about 6 billion people need, but about 4 billion need/want jobs?

[Nick Rowe]

Too much Fed: That's given me the idea for a new post. Still mulling over the title: "Of horses and men"? The demand for horses' labour has almost disappeared; how come the demand for human labour hasn't?