...against fictions and other tall tales

Saturday, 25 February 2012

Sense and nonsense about the aging of the population

Earlier this week, the federal minister responsible for overseeing Canada's public pension and old age security programs, Diane Finley, suggested that the aging of the population and the future cost of social programs targeted to retirees and seniors will lead to massive increases in taxes, crippling debt and a huge debt burden on future generations.  To remedy the situation, Minister Finley is proposing to raise the eligibility age to Canada's old age security program as a way to reduce future costs and preserve the "sustainability" of federal budget costs.

The rationale for the proposed program changes is based on the fact that the ratio of working-age people to seniors is projected to decline in the next decades or, as Minister Finley puts it, "as we go forward, we’re going to have three times the expense in Old Age Security as we do now, but we’re only going to have half the population to pay for it".

But is it really correct to say that the government is headed for a demographic shift that will jeopardize the sustainability of the federal budget in years to come?  I am not convinced. And here's two reasons why I think the problem of population aging is currently overblown.*

The ratio of workers to seniors

First of all, it's important to understand that the so-called "ratio of workers to seniors" is, by itself, a fairly uninformative concept for analyzing the issue of population aging.  The reason for this is simple: the ratio of workers to seniors tends to distort the true burden associated with the aging of the population.  Focusing on the ratio of workers to seniors, as most commentators and policymakers are currently doing, obscures the fact that seniors are, and will remain, a relatively small share of the total population.  Also, it's important to keep in mind that the working population must also "support" itself and the youth, in addition to supporting the senior population.

Instead, a more useful concept for analyzing the impact of population growth on the economy is the ratio of the total population to working-age population.  By using this ratio, one gets a much better sense of the real impact of population growth on the economy and, as a result, on future government budget outcomes.

Figure 1, Source: OCA, 2010
Figure 2, Source: OCA, 2010
Figure 1 provides a good snapshot of the increase in population projected (broken down for each age category) between 2011 and 2030 in Canada (see Figure 2 for the period 2011-2050).**  As shown in Figure 3 below, if the focus is solely on the population over age 65, the picture looks worrisome: over the next two decades (2011 to 2030) the number of people 65 and older will rise 78% relative to those 20 and 64.  When adding those under 20 to those 65 and older, the dependency ratio rises by 36% over the next two decades.  And when the entire population is considered relative to the working-age population, we note that the ratio rises from 159% to 180%, a much more manageable 13% increase over the next two decades (note: 13% of 159 is 21).

In other words, the actual "burden" of aging based on current projections consists of an additional 13% more people per working-age person.  This is much smaller than the 78% that is currently being mentioned by commentators and politicians.

Figure 3, Source: OCA, 2010 and author's calculations

Figure 4, Source: OCA, 2010 and author's calculations

Now, some people may argue that this amount is still quite high and, as a result, the government should nonetheless intervene to reduce future outlays to seniors.  This brings me to the second reason why I'm skeptical about the argument that population aging will have deleterious effects on the economy and public sector budgets: productivity growth. 

Productivity growth

Discussions about the aging of the population rarely, if ever, highlight the critical role that productivity growth plays in enabling the economy to afford the cost of programs destined to retirees and seniors.  Yet, the role that productivity growth plays is actually very important because as people become more productive at work, more income is generated to support those who aren't in the labour force such as the youth and seniors

In fact, if we look at the average rate of productivity between 1981 and 2011 for Canada, we find that productivity grew at a rate of approximately 1.3 percent per year (Martel et al., 2011).  This productivity gain greatly contributed to helping the Canadian economy shoulder the increased burden of aging during previous decades.  In 1981, the ratio of workers to seniors was approximately six to one whereas today it is approximately four to one. In 2030, it is expected to be below three to one (Statistics Canada, 2011).  Therefore, assuming that the average rate of productivity will remain at this level until 2031, we find that productivity will have nearly doubled between 1981 and 2031.***

Thus, once you consider the impact of productivity growth, the picture doesn't look so bleak anymore:  three workers in 2031 are expected to produce approximately the same level of output that six workers produced in 1981, fifty years earlier.  In other words, because of productivity growth, the worker in 2031 will generate almost twice as much output per hour as the worker from 1981.

A simple rule of thumb is that population aging remains "sustainable" as long as productivity rises faster than population.  Therefore, assuming an average productivity of 1.3% per year between now and 2031 (the same level as for the period from 1981 to today), we find that productivity will grow 28% whereas population will grow by 13%, as shown in Figure 3.  This is a noticeable difference that would enable Canada's economy to shoulder the burden of population aging while also increasing the population's standard of living.  Comparing this amount to the projected increase in population of 13%, we see that productivity growth will more than make up for the future increase in population.

Now, it is possible that future gains from productivity may not be entirely reflected in increased income for workers (via rising real wages).  Productivity gains may end up being disproportionally absorbed by businesses through increased profits.  However, it is important to understand that this problem is an entirely different one from that of the sustainability or solvency of public pensions and retirement programs.

In the event that future productivity gains get absorbed disproportionally into business profits, the remedy would be for government to ensure that a fair share of the gains from productivity be diverted toward real wage growth for workers.  Certainly, such a scenario would not warrant making drastic changes to the federal government's old age security program by raising the program's eligibility age from 65 to 67, as Minister Finley is proposing to do.

To conclude, I am not saying that taxes will not need to be raised by some amount to cover the future cost of public pensions and other retirement benefits.  The point here is that increased costs to taxpayers and workers should not be overly onerous.  The resources will be there to support the senior population in the future since productivity growth should more than make up for the 13% growth in the total population that working-age people will have to support in the coming decades.

The FRB blog invites your comments. Please share your thoughts below.

* This analysis is based on the excellent article by Spriggs and Price (2005).
** All figures are based on OCA, 2010, p. 96 and author's calculations. See Figure 5 below.
*** An average rate of productivity of 1.3 percent between 2011 and 2031 is a conservative assumption. Some economists are suggesting that Canada's future rate of productivity will rise in the coming decades. See Arlene Kish's IHS Global Insight dated October 2011, as well as the October 2011 edition of the Bank of Canada's Monetary Policy Report, (p. 19) regarding the expected increase in productivity the next few years. Note: a quick and easy way to approximate the number of years it takes for a variable growing at a constant rate to double is to use the "Rule-of-70" or dividing 70 by the chosen growth rate.

Figure 5, Current population and projections,
Source: OCA, 2010 (in thousands) and author's calculations
Hoc dicatur meum filium Vincent, cui futurum quasi electa ut sol. Ut non factus hostiam logica. 


OCA (Office of the Chief Actuary), The 25th Actuarial Report on the Canada Pension Plan, November 2010

OCA (Office of the Chief Actuary), The 10th Actuarial Report Supplementing the Actuarial Report on the Old Age Security Program, August 2011

Martel, L. et al., Projected trends to 2031 for the Canadian Labour Force, Statistics Canada, August 2011

Spriggs, W. and Lee Price, Productivity Growth and Social Security's Future, Economic Policy Institute Issue Brief #208, May 2005.

Statistics Canada, Revisions to Canada and United States Annual Estimates of Labour Productivity in the Business Sector 2006-2009, March 2011 (Table 4)

Friday, 10 February 2012

Fiscal policy vs monetary policy to control inflation: MMT perspective, by Joseph Laliberté

In the comment section of a previous post, I suggested that Modern Monetary Theory (MMT) was a form of "quantity theory", and that it would be to MMT's advantage to develop on this point. The following is an excellent piece on this topic by fellow blogger Joseph Laliberté. This article is cross-posted in French on Joseph's blog, Défricher l'économie.

Scott Fullwiller once pointed out that MMT is also a quantity-theoretic model of changes in the price level:
Interestingly, MMT is also a quantity-theoretic model of changes in the price level. The differences are (1) net financial assets of the non-government sector, rather than traditional monetary aggregates, is MMT's preferred measure of “money,” and (2) desired leveraging of the non-government sector is akin to what one might call “velocity.” In MMT, the two of those together (net financial assets of the non-government sector relative to leveraging of existing income) set aggregate demand and ultimately changes in the price level, at least the changes that are demand-driven.
Net financial assets of the non-government sector are equivalent to past accumulated government deficits (a government deficit is a surplus for the non-governmental sector, see here for the accounting demonstration). As per MMT formulation, we have: (I think Warren Mosler was the first to come up with the MMT interpretation to this famous economic equation)

1) M*V = P*Q

Where M is net financial assets of the non-government sector and V is the desired leveraging of the non-government sector (V could also be seen as the inverse of the desire to save of the non-governmental sector); P is the price level and Q is the output.

Assuming that the economy is operating at capacity (denoted by Q’) at a given price level (P’), any increase in M or V would be inflationary (i.e. would increase the price level beyond P’). Therefore, if the government decides to increase net financial assets of the non-governmental sector by running a deficit, in such scenario, inflation would result. However, mainstream economics would argue that the increase in net financial assets of the government sector need not be inflationary to the extent that the Central Bank could influence the desired leveraging of the non-governmental sector (denoted by V) by manipulating the interest rate. In a nutshell, the Central Bank could offset the inflationary effect of an increase in M with a corresponding increase in the interest rate (denoted by r) so that V decreases. Therefore V is a function of the interest rate, or V(r).

Assuming that we have a "super Central Bank" that is always able to set its policy interest rate at a level where monetary policy always offset the inflationary effect of fiscal policy, we have in period 0:

2) M0*V(r0) = P’*Q’

Since the economy is operating at capacity, the government should normally aim at balancing its budget in period 1. Instead, let's say that the government decides to run a one-time deficit in period 1, which adds to the net financial assets of the non-government sector. The Central Bank would then increase the interest rate in period 1 to decrease V in order to make sure that the government deficit is non-inflationary. We would then have:

3) (M0 + ΔM1)*V(r0 + Δr1) = P’*Q

Where ΔM1 is the government deficit in period 1, and Δr1 is the increase in the interest rate in period 1 necessary to decrease V and keep inflation in check.

In period 2, assuming that the government withdraws its economic stimulus and goes back to a balance budget, then M would still rise as a result of the increase in interest that took place in period 1 (note: assuming that government debt in circulation is all short-term, then r is also the interest rate on government debt, therefore it is the interest rate at which net financial assets of the non-government sector compound). To make sure that this increase in M is non-inflationary, the Central Bank would need to raise the interest rate further by Δr2. (I'm assuming here that the Central Bank adjusts r even it means an increase of a fraction of a basis point). We would then have:

4) M2*V(r0 + Δr1 + Δr2) = P’*Q’

Where M2 = (M0 + ΔM1)*(1 + Δr1)

A similar logic would prevail in period 3:

5) M3*V(r0+ Δr1+ Δr2+Δr3) = P’*Q’

Where M3 = (M0 + ΔM1)*(1 + Δr1+ Δr2 )

Again, a similar logic would prevail in period 4:

6) M4*V(r0 + Δr1 + Δr2 + Δr3 + Δr4) = P’*Q’

Where M4 = (M0 + ΔM1)*(1 + Δr1+ Δr2 + Δr3 )

And so on.

One can see clearly from the demonstration above that the Central Bank’s action is both the solution and the source of the problem: its increase in the interest rate in period 1 expands the net financial assets of the non-government sector (M) in period 2, which renders necessary a further increase in the interest rate to decrease V in period 2 in order to keep inflation in check, and this further increase in interest rate further expands M in period 3, which commands still a further increase in the interest rate to decrease V in period 3...and so on, and so forth.

The only definitive solution to this vicious cycle would be to use fiscal policy rather than monetary policy to eliminate the inflation threat originally caused by the government deficit in period 1. This would mean generating a budget surplus in period 2 that would exactly offset the budget deficit of period 1 plus the interest payment. The size of the budget surplus relative to GDP necessary in period 2 could be expressed as follow:

(M0* Δr1) + ((ΔM1)*(1 + Δr1)) / (P’*Q’)

Flowing from this analysis is the following critical policy implication: the longer the country tries to fight inflation with monetary policy, the bigger the size of the budget surplus relative to GDP that is necessary in the future to eliminate the inflation threat. For example, in period 4, the size of the budget surplus relative to GDP necessary would be:
(((M0* (Δr1 + Δr2 + Δr3 )) + (ΔM1 *(1 + Δr1 + Δr2 + Δr3))) / (P’*Q’)

This demonstration above explains why MMT holds that monetary policy is really an ambivalent tool when it comes to fighting inflation as increasing the interest rate could be both expansionary and contractionary with regards to aggregate demand. This observation holds even if one assumes that a "super all-knowing central bank" exists that is always able to adjust the interest rate perfectly in order to keep inflation in check. (i.e., the Central Bank is able to perfectly control the desired leverage of the non-governmental sector through monetary policy). (Note: most MMTers would argue that a central bank is incapable of doing such a thing to begin with)

Does this demonstration have any implications in the real world? In the Canadian context, I would say yes. To the extent that one deems the budget surpluses of the 1990s in Canada necessary, the high interest rates of the early 1990s likely made these “required” budget surpluses even larger.

Furthermore, one could speculate that a quid pro quo took place in the mid-1990s between the government and the Central Bank whereby the Central Bank accepted to slowly reduce interest rates if the federal government started tackling its deficit.

Tuesday, 7 February 2012

A note on the increase in Canada's unemployment rate

Canada's rate of unemployment rose for the fourth consecutive month in January. Now at 7.6 percent, the rate of unemployment is a mere 0.2 percent higher than it was 12 months ago. According to Statistics Canada's January edition of the Labour Force Survey, last month's increase in the rate of unemployment was the result of growth in the labour force outstripping that of employment.

Unfortunately, this trend probably won't end any time soon. With business investment slowing markedly, the growth in personal expenditures declining and Canada's public sector poised to cutback on hiring during budget season, it's doubtful that things will improve in the short-term.

But what about the longer term? Here too, there isn't much to be optimistic about. The reason for that is because, within the next few years, households will embark on a period of deleveraging during which households will seek to pay down debt and increase savings, as recently highlighted by the Governor of the Bank of Canada. And when households reach that point, the economy will likely encounter a considerable setback that will lead to weakening employment growth. This will be especially apparent if businesses don't increase their level of investment and exports stay at their current levels. Public sector austerity and deficit reduction will only worsen the situation.

To illustrate this point, I thought it would be useful to include the following charts showing the rate of unemployment along with Canada's household sector balance (click on charts to enlarge)*. As you can see, the rate of unemployment and the household sector balance tend to move in tandem: whenever the household sector balance trends upward (downward), the unemployment rate rises (falls).

Chart 1, Source: Statistics Canada

Chart 2, Source: Statistics Canada

Now, it should be pointed out that this relationship runs both ways. On the one hand, higher (lower) unemployment reduces (increases) household spending, and, one the other hand, increased (reduced) consumption by households helps to boost (weaken) the economy and increase (decrease) employment.

But given that growth in consumer credit is at the lowest it's been in over two decades and that owner's equity as a share of real estate is declining rapidly (which reduces households' ability to finance consumption using equity in their homes, see chart 3), it seems reasonable to believe that it's the rise in the household sector financial balance that will create upward pressure on unemployment rather than the other way around.**

Chart 3, Source: Statistics Canada

To conclude, the implication here is that the federal government's intention to move forward with austerity at this time is ill-advised. Unless the decline in the household sector's participation in economic growth is not offset by that of another sector (government, foreign or business), weak employment growth will likely last longer than most people expect.

* A sector's net financial balance is the difference between sectoral savings and investment. Reduced household consumption and investment, as well as higher savings increase the financial balance of the household sector.

**For those who are interested in this type of analysis, I strongly recommend the work of economist Josef Steindl. According to Steindl, the examination of sectoral financial balances provides a useful framework for the analysis of macroeconomics. Also, Steindl viewed the household sector's propensity to save as an important determinant of economic growth.


Steindl, Josef. 1982. “The Role of Household Saving in the Modern Economy”, Banca Nazionale del Lavoro Quarterly Review, Vol. 35, no. 140 (March), pp. 69-88