Sometimes I have to remind myself that the sort of short run macro I do doesn't really matter much. Things like the recent global recession don't make much difference in the big scheme of things. So I'm writing this as an antidote to my own narrow perspective. Or, maybe I'm just a typical boomer seeing everything through the lens of my own age-group.
The first big macro shock was the invention of agriculture. Productivity rose, then fell again for Malthusian reasons. The second big macro shock was the agricultural/industrial revolution. Productivity started growing so quickly it outran those Malthusian reasons.
I think the third big macro shock will be the retirement revolution. Poor people, on the Malthusian margin, retire when they die (or die when they retire). Rich people retire before they die. The world population is ageing. But age per se has no macroeconomic implications. Retirement does have macroeconomic implications. The fact that there will be a greater percentage of old people doesn't matter. The fact that there will be a greater percentage of retired people does matter.
Let's start with some simple accounting. The formula for the price of a life annuity P, where the annuitants have a constant risk of death d, and want a constant consumption stream C, with a constant rate of interest r, is:
P = C/(r+d)
People with a probability of death d per year will live on average for 1/d years. If people retire just before they die, 1/d will be a small number and d will be a big number, so the price of a life annuity for someone retiring will be a small number. If people retire a long time before they die, 1/d will be big, d will be small, and P will be big.
That's the first thing the arithmetic tells you. Here's the second thing:
P is the assets that pension plans will need per retired person. If R is the ratio of retired people to total population, and A is the assets that pension plans will need per head of total population, then:
A = R.P = R.C/(r+d)
Rearranging that formula, and assuming that retired people have the same level of consumption as the population average, we can solve for the ratio between total assets to total consumption as:
A/C = R/(r+d)
The retirement revolution will cause an increase in R as well as a reduction in d. Both will cause the ratio A/C to rise. For example, if r were near 0%, then if the percentage of retired people doubled, and retired people lived twice as long, then aggregate pension plan assets would need to nearly quadruple as a ratio of aggregate consumption. [Update: Damn! Am I double-counting here? My math-brain has stopped working.]. [Update 2: Yep. This quadrupling bit is wrong. As Alex notes in comments, I'm ignoring the assets penion plans hold for people who are saving for retirement but are not yet retired.]
Now suppose that land were the only asset, and that land is in perfectly inelastic supply. It is not possible for pension plans to quadruple the amount of land they hold. But a quadrupling of the price of land would have (almost) the same effect. The quadrupling of the demand for assets by pension plans would cause the price of land to quadruple.
But it's not exactly the same. Because there's a negative relationship between the price of land and the rate of interest, and a negative relationship between the rate of interest and the assets that pension plans need to hold. The rate of interest will equal the annual rents on land as a ratio of the price of land, plus the expected rate of appreciation in the price of land. Ignoring expected appreciation, a quadrupling in the price of land would mean interest rates fall to one quarter of their previous level. And if interest rates fall when pension plans demand more assets, that increases their demand for assets still further. Because a lower interest rate means higher annuity prices. See the equations above.
This does not mean there will be an explosively unstable spiral in land prices (because a halving of r will cause A to less than double if d>0), but it does create a "multiplier" effect. If the percentage of retired people doubled, and retired people lived twice as long, then aggregate pension plan assets might need to more than quadruple as a ratio of aggregate consumption.
Now land is not the only asset that pension plans can hold. They can also hold physical capital. Physical capital is exactly like land, except it doesn't have a perfectly inelastic supply curve. When the price of physical capital rises, more will be produced, and the stock will rise over time. But as the stocks of physical capital rise, relative to the stock of land and working population, annual rents on physical capital will fall. So there is a limit on the amount of physical capital that can be produced that can offer a rate of return at least equal to that of holding land. Investing in physical capital alleviates the effects somewhat, but it will not eliminate the effects.
Perhaps pension plans don't need to hold only land, or physical assets. They could also hold chain letters or other Ponzi assets, by operating as Pay As You Go plans. They tax the working young and make transfers to the retired old, and hold purely fictitious assets that would have no value if the next cohort of young refused to play the game and broke the chain.
Such Ponzi schemes can under some circumstances be sustainable, and make all generations better off forever. If alternative investment opportunities are bad, so that interest rates are less than the growth rate of the economy, such Ponzi schemes can be sustainable. But could such Ponzi schemes ever beat land as an investment opportunity? If land rents rise in proportion to GDP, land would always beat Ponzi schemes. Because land prices would grow at the same rate as GDP, giving a rate of capital gains equal to owning a Ponzi asset, plus land will yield rents in addition to that. And some land rents will rise in proportion to GDP, or more. All we need is a type of land that produces a service the demand for which is sufficiently price inelastic and income elastic. Cobb-Douglas preferences for the services of one type of land (so we spend a constant share of income on those services) would be sufficient to eliminate PAYGO pension plans as a sustainable competitive alternative. The prices of land like that would look like a bubble (which is another word for Ponzi scheme), and it would be very close to being a bubble. In the limit, as the rate of interest fell, relative to the growth rate of rents on such land, and the price/rental ratio rose, it would approach being a pure bubble asset. P=rents/(r-g).
Maybe, just maybe, recent short run macro events aren't just a blip on the long run big picture, but are a symptom of the difficult transition to that long run big picture. When I read about the price of agricultural land, it seems to be actually doing what prices of building land have been trying to do.
Maybe, when people a century from now look back on today, they will wonder why macroeconomists didn't pay much attention to the really important stuff. Will it be robots, or retirement?