If I had a policy tardis...

Imagine it was possible to travel back in time, and tell the policy makers of the past everything that we now know about pay as you go pension (PAYG) plans. Life expectancies will increase, birth rates will fall, and the contributions required to sustain the schemes will grow. PAYG pensions will be blamed for discouraging savings, and dampening economic growth. 

Imagine it was possible to go back to 1951, when Canada's Old Age Security programme was introduced. Could Louis St Laurent be persuaded to introduce a fully funded pension plan instead?

I doubt it. In 1951 Canada had a serious problem: elderly folks without the wherewithal to live a decent life. Introducing a fully funded pension plan might make a difference some time in the future, but would have no immediate impact on the hardship people were facing right then and there.

O.k., suppose it was possible to go back to 1941? War time. 1931? Great depression. 1921? Would that work? 

An 80-year old widow in 1951 would have been 50 in 1921. Imagine that with our policy tardis we could go back to 1921 and persuade the Canadian government to set up a fully funded pension plan, along the lines of the Canada/Quebec Pension Plan. Employees and employers would jointly contribute 10 percent of the employee's earnings to the plan. The contributions would be placed in a government-administered fund, and invested in Canadian and international stock markets, government bonds, and real estate. The plan would pay benefits equal to 1 percent of inflation-adjusted pre-retirement earnings for each year of contributions. So, for example, a person who contributed for 20 years would enjoy a pension of 20 percent of pre-retirement earnings. 

Would such a plan have been financially viable? It would have taken a huge hit in the 1929 stock market crash, and would have seen poor returns throughout the 1930s, but perhaps it might have worked - it would be helped by the fact that not many people would live long enough to collect benefits.

Would it have been enough to allow our hypothetical widow to escape poverty?

In 1919, Statistics Canada estimated that the weekly cost of a family budget of staple food, fuel, lighting and rent was $23. Over half of the wage earning population was, at that time, earning under $20 per week - and not everyone earned a wage, 38 percent of the Canadian male labour force was employed in agriculture. Would four 1921 dollars a week - that's 20 percent of a somewhat above average wage - be enough to live on in 1951? Perhaps - but there would still be people who survived poverty and unemployment to live to a grand old age.

To relieve poverty among the elderly, any pension plan, even a fully funded one, must have a redistributive component. Nick Rowe argued in a recent post that the best way to design a pension plan is on a cohort by cohort basis. Each generation has its own, fully funded plan, one that guarantees an adequate standard of living in retirement for everyone, redistributing from the rich to the poor in each generation.

Imagine that Nick Rowe was the policy doctor, travelling back in time to persuade the politicians of 1921 to set up a fully funded, redistributive pension plan, one that would solve the problems of elderly poverty in the future. Heck, we could send him back to 1901 or 1881. Whatever it takes. He could give Arthur Meighen or Mackenzie King the ECON 1000 treatment, and convince them that a fully funded pension plan was the only way to go. 

But would the general public ever have voted for such a plan? Would it ever have been politically viable? Not before 1918, as women didn't have the right to vote, and women would benefit disproportionately from a redistributive pension plan.

In 1921? I find it hard to imagine that such a plan would pass. The contributions required would be unacceptable to many. Remember that an average wage was less than $20 per week, and a family's living costs were $23 per week. (Then, as now, a typical family needed more than one wage earner to have a decent standard of living.) Contributing 5 or 10 percent of earnings to a pension plan could cause some people real hardship.

The redistributive component would also be unacceptable because of the identifiable victim phenomenon. Here's an old quote from Schelling:

“There is a distinction between an individual life and a statistical life. Let a 6-year-old girl with brown hair need thousands of dollars for an operation that will prolong her life until Christmas, and the post office will be swamped with nickels and dimes to save her. But let it be reported that without a sales tax the hospital facilities of Massachusetts will deteriorate and cause a barely perceptible increase in preventable deaths—not many will drop a tear or reach for their checkbooks.” 

People are willing to relieve the actual poverty of friends and neighbours. But I doubt they would be willing to up their pension plan contributions in anticipation of the suffering of some unknown person decades from now. 

I suspect our ancestors would also perceive and decry the moral hazard and incentives for dependence created by a redistributive fully funded pension plan.

If a fully funded plan wasn't acceptable to voters then, what grounds is there to believe that a fully funded plan would be acceptable to voters now? And if it isn't, why do we keep on going on about it?

Comments

[david]

What kind of political landscape would demand both downward redistribution and social insurance, whilst permitting these to be operated separately? Presumably societies where these demands evolved at different times, with raw downward redistribution coming first and the social insurance padded on later. Unfortunately in Western industrial societies pensions (with some kind of implied social pressure) came before redistributive politics (and its helpful 'mixed-economy' socialist notion of nationalized development banks hoarding people's savings) became a thing.

[Livio Di Matteo]

Frances:
Interesting post. Fully funding social security or health for that matter is really pre-funding. There are three payment options for public programs: pre funding, pay as you go and post funding (that is, running a deficit and paying for it later). I think pay as you go and deficit financing are more popular due to simple rate of time preference. For most people, a dollar in hand today is preferable to one tomorrow. Taxes down the road to pay for deficits today are preferable to taxes today to spend on programs tomorrow.

[Jim Sentance]

I don't know that with a population age structure that is fairly stable that there's much point to a fully funded plan. Whatever advantages there might be to my mind are wiped out by the vagaries of the stock market.

The real problem arises when you have a population bulge like the baby boom, or more to the point a baby bust such as followed it. So realistically I can't see any period of our history before perhaps the late sixties seeing a need to be prudent like that. At that point we were already busy setting up a partly funded plan in the CPP/QPP. National debt was slowly still melting away, we were running surpluses and the economy was still growing at a handsome rate. None of the issues that have hit us since were apparent.

If Nick came back from the future to warn us, I imagine they'd have put him in a straight-jacket and locked him up.

[Frances Woolley]

Jim nails the central point, which is what would happen "If Nick came back from the future to warn us."

"I imagine they'd have put him in a straight-jacket and locked him up." Realistically, that should happen more often than it does on Dr Who, shouldn't it?

Livio - you're right on the politics of PAYG. It's true, people would vote for free ponies for everyone. However politicians don't give people a chance to vote for free ponies for everyone, because it's a stupid idea. So the question is not "would people vote for a PAYG pension scheme."

This is our thought experiment: the Doctor convinces politicians that PAYG=free ponies and trouble down the road. Politicians give people the choice between fully funded pensions + within-cohort distribution or nothing. Do people prefer FF pensions, or no government pension plan at all? I don't think they'd go for a plan with within-cohort redistribution.

[DavidN]

Australia has a fully funded scheme with a safety net provided by a means-tested public pension system. So politically, yes, it is possible.

[Livio Di Matteo]

Politicians do not need a vote by the public on whether they want free ponies. They can just go ahead and do it on their own by running up deficits. Moving difficult decisions into the future if they can is one manner in which politicians can get themselves elected.

[Frances Woolley]

DavidN: "Australia has a fully funded scheme with a safety net provided by a means-tested public pension system. So politically, yes, it is possible."

The Australian means-tested safety-net component of public pension system works like the Canadian one, doesn't it, in that it's funded out of general tax revenues? So there is redistribution across cohorts, which doesn't work well when the retired cohort is big and the working-to-support-them cohort is small.

Does the Autralian system have a fully funded safety net? If so, the politics of that getting through the electorate would be really fascinating.

[DavidN]

Frances: I’m not sure if the means-tested public funded safety net works the same, if I was to hazard a guess, it probably is. And no, Australia doesn’t have a fully funded safety net but the fully funded part of the pension system, Superannuation, has been highly effective in reducing the strain on the public pension system. The Super system is a combination of the CPP and RRSP. Contributions are compulsory but you can make voluntary contributions (with tax benefit for doing so) as well. You get out 100% (minus taxes) of what you put in so incentives are closely aligned. Effectively people only get on the public system once their Super runs out.

Does the CPP work the same way (in the sense that you only get out what you’ve put in)? If it is, then my apologies for jumping the gun.

[Zithromax]

Great article Professor. I'm from Argentina. I love Friedman point of view. I think the only thing he missed was that Humans are the worst missery on earth. The problem isn't the Free Market System. The problem is the ambitious people who make people work for pennys, waste natural resources just to make themself wealth bigger. I think that the key in economic is in the Human psychology. Free Maket is a perfect system: Adaptable, Extensible, that can get over most difficult crisis. Sorry for my english! Best regards.

[DavidN]

Having thought about it some more, I don't think you need a fully funded safety net (PAYG system would be sufficient) if you had a compulsory private retirement plan (which works as a fully funded system) that was sufficient to meet the majority of consumption needs post-retirement. I'm not sure if it's different in Canada, but in Australia the public pension is a backup, not a first choice.

[Frances Woolley]

DavidN - it appears that the Australians have gone where Canadians fear to tread. From how you describe the Australian plan, it seems that CPP/QPP has a greater redistributive element, in that some people get a much higher rate of return than others - the first people to collect got a huge windfall gain, and current young workers get a pretty mediocre rate of return from the plan, especially if they contribute for more than 25 years.

The other thing that's interesting about the Australian plan is that (according to Wikipedia) the bulk of contributions are made by employers, unlike the Canadian plan that has 50/50 sharing. I don't see Canadian employers going for that - there's more pressure to align policy with the US here.

But (again according to Wikipedia) the Australian means tested old age pension is financed out of general revenues. So there's still an element of intergenerational redistribution, with all of the issues that this raises when some cohorts are big and some cohorts are small.

Would the Ozzies have voted for a superannuation scheme that also financed an income guarantee?

[DavidN]

I don't know the political history of super reform but I suspect compusory employee contributions may have been one of tactics to gain popular support. There are more employees than there are employers. It was a Labor government which passed the reform so I don't think it would've hurt them as much than it would've for a Lib/Nat government (business/bosses aren't Labors natural constituency). Econ101 speaking it doesn't really matter who contributes, employers contributing just means lower wages for workers.

With respect to your question, I'm not sure, it was before my time. ;)

[Determinant]

Australian Superannuation reform was a policy decision to move away from a PAYGO system. It is privately managed so the usual problems about fees apply. Curiously union funds managed on a non-profit basis have better performance and lower fees.

Australia has a crucial difference from Canada in that there was a Constitutional referendum in 1945 about the Welfare state which places Superannuation, disability insurance, public health insurance and most other income programs under the authority of the Commonwealth Government, not the States. So Alberta or Quebec could not torpedo a "Big CPP" plan under the Australian regime, provincial support is needed for CPP changes.

The problem that everyone has with PAYGO is really the degree of redistribution. OAS is not that generous so I have no problem with its current level of redistribution. Canada could implement a Big CPP policy and I fully support that, alternatively Saskatchewan could legislate mandatory contributions to its Saskatchewan Pension Plan which is a DC plan that is a low-cost RRSP with a single fund.

In Canada provinces have constitutional authority for pensions, that's why Pension Benefits Acts are provincial. So mandatory pension contributions is a provincial issue. Quebec has made Pooled Registered Pension Plan contributions mandatory, Ontario can too.

Lastly the difference between PAYGO and fully funded is less than most people think. In a fully funded plan the pension obligation has to come from investment returns. Those returns have to be generated by current income and economic activity. Funding does not per se protect you from a falling working population, the returns on capital will decrease and benefits will be placed in jeopardy. What funding does do is allow you to invest overseas in other national markets with better demographic characteristics. This is the really reason CPP is now fully funded.

Many economists make the mistake of treating capital as exogenous to current demographics and income conditions, it isn't, capital is thoroughly endogenous.

By the way moving OAS to age 67 while maintaining CPP at 65 was nonsense and turned a defensible policy position into an act of crass politics.

[genauer]

Initial comment: I think most people here watch most threads here. Therefore I dont repeat myself, and just relate to the relevant links)

Why not go back to when this all started ?

1883, when our first iron chancellor, Bismarck started all that.
First with universal_health_care, then disability, then public pensions.

Originally started with just 1.7% of wages, 50:50 employer, employee, fully funded. Very small, below the "magic" 2.0% (see inflation). Not as a tax, but as an investment, a "Lock Box", LOL.

It always starts that way. It makes it a much easier to sell to the people.

Problem is, if you look at the LCH model, as in the Language Paper thread (Modigliani reference), you would need about 500 % GDP or more stashed away, and there is simply no room for that amount of investment.
(My reference to OECD data of working capital / GDP = 3 years, there).

Therefore it never stays that way. You will always end up with a full PAYGO system.

As in the very most large, mature, slowly or not (population) growing countries, now. As a large country you can not run persistent high Currenct Account surplus, you will not get the money back. You can not overinvest in a significant way. Cobb-Douglas, Solow is not true, but this is another story for another day. Japan is the best example.

You can not even promise people a fully inflation adjusted, zero real interest, pension.

The only realistic, and therefore credible, promise you can make is:

We can tax people at 50 % or a little more. From this we take (absolute) 20 % for pensions, 15 % for health care, and the rest for the other things.

German detail:
The 20 % get distributed, according to a simple linear scheme: you collect points, relative to your payments / median per year. And your payout later is according to what you payed in. if this is not enough, you get the standard social minimum.

Nobody falls through the cracks, nobody evades payment. Universal, long term stable.

Realpolitik, Ordnungspolitik (as formulated in the Greece thread), or as the anglos call it: ordoliberalism. The Irish (economists) are clever, and now show some interest in understanding it (via the iceland thread)

You can deviate from that for a long time, if you are a little statelet like Liechtenstein, Luxembourg, Singapore, Iceland. To a lesser degree as Belgium, Switzerland, Ireland. Young, growing, resource-rich Canada, Australia, Norway, too, until saturation, which might be away 50 years or more.

But in the end, you will all end up with the German Model, not because you like Germany, or some arrogant me, but because:

There is no alternative.

In the long run, it simply does not matter, what kind of monetary plans politicians draw up in the short run.

I am perfectly aware that this now sounds extremely arrogant, typical German.

Keynesanism, MMT, NGDP are just some monetary, psychological gimmicks from this long term view point.

Please be a little patient with me, and my verbal limitations.

And attack these sound bites, sentence by sentence, fisk me !

I thought about it, and believe that some radically different "grand scale" point of view, could be helpful here, before diving into Canadian details.

[Determinant]

Canadian and British welfare programs don't derive from Bismark, they derive from the Beveridge Report of 1942. It wasn't a response to Bismark but a response to the Great Depression. Anglophone welfare politics has always been different than Germany's. The Bismarkian model doesn't have much traction here. Check the internet for the comparison of the Beveridge vs. the Bismark model.

Canada isn't that young either because of our large "Baby Boom" bulge in our demographic profile.

Average children/family in Canada has been on the decline for most of the late 19th and 20th Century except for a 20-year "blip" that began in 1948 when we demobilized. Yet that blip coincided with the implementation of the Welfare State and skewed views about that the long-term population age profile would be. Further nobody predicted the massive increase in life expectancy in the 20th Century, Canada even outpaces Germany in that regard. I had five great-grandparents alive when I was born. My niece, just shy of a year old, has seven great-grandparents alive today and all four of her grandparents. One of my grandfathers died early from an unpredictable, non-lifestyle related disease. He was expected to live quite a while as he didn't smoke, didn't drink and had a slim waistline with no potbelly at all.

Deciding how to divide the Longevity Dividend is the economic, social and political problem of our time.

[The Least Popular Kid in Class]

"In 1919, Statistics Canada estimated that..."

Should be

Statistics Canada estimates that in 1919...

StatCan is not that old.

[Stephen Gordon]

Ah, if we could only bring back the Dominion Bureau of Statistics...

[Frances Woolley]

Least popular - actually it should be "the dominion bureau of statistics estimates...." If you click, you'll see that those numbers come from the 1921 Canada Year Book, so they aren't present estimates of past wages/prices, they're past estimates of past wages/prices.

Determinant: "Deciding how to divide the Longevity Dividend is the economic, social and political problem of our time." I wonder. Interesting thought.

[Determinant]

@Frances:

Consider a life annuity, whose microeconomics are well known. In an environment of increasing longevity, the provider will face increasing costs if the annuity begins at a fixed age, e.g. the age 65 limit which was standard in Canada. The annuity is unstable. But look at the Scandinavian countries where the public pension age moves with longevity. Now the annuity has much less longevity risk and becomes stable.

Think of this as longevity indexing, just like inflation indexing. In fact you would want Statistics Canada to have a yearly Life Expectancy Index just like the Consumer Price Index and it should be watched as just as widely and eagerly. Nick extols the benefits of inflation targeting with CPI, why can't we use a longevity index to fine-tune pensions the same way?

In fact it occurred to me that just as inflation indexing benefits beneficiaries, age indexing benefits sponsors. It actually seems to be a fair trade in my left-wing opinion.

I watched a lecture on Youtube from Gresham College, a public lecture institute in London, UK. The lecture was on pensions and longevity was a huge risk. In fact men have improved their mortality faster than women, they are nearing equality whereas men used to die significantly sooner. The difference is statins and cholesterol control. According to the lecture cardiac medicine has dramatically improved its ability to treat heart attacks, including recognizing different kinds. Cholesterol control is doing wonders for men's health.

Fine, wonderful, we all want better medicine and to live longer. But who should pay for the increased general lifespan? Longevity indexing seems a fair way to do it, because the window between expected retirement when you started working and when you actually retire should be five years and can be covered by direct individual savings, the annuity features of pensions can take over after that gap where the annuity model is a better fit.

[tyronen]

In 1951, OAS was a replacement for the Old Age Pensions scheme that had started in 1927. That provided a means-tested pension for those over 70. It was not pre-funded either. About half of seniors received benefits.

So in 1927, the public may not have been willing to pay for a large pre-funded program, but they did pay for a limited unfunded benefit that redistributed wealth from the working-age population to the elderly poor. OK, there were not as many elderly then, but by 1951 the complaints weren't that the program cost too much, but that the benefit was too little. The 1951 reforms doubled the cost of the program, and taxpayers cheerfully paid it. By 1957, the Tories were trying to outflank the Liberals on the left, accusing them of increasing the pension by a mere "two cups of coffee a day".

[Jacques René Giguère]

Determinant: five years... I am at the top of my game and wouldn't mind teaching 10 or 12 more years. I wouldn't mind getting back my now unnecessary contributions and those my employer put into my DBPP.
But can we start now a supplication session to the gods of economic policy to insure a smooth transition to a world with a 20% larger workforce where a bunch of newcomers will be shut out of the labor market and where a larger worker pool will decrease wages?

Frances: Nick wants a fully funded pension plan which will invest in debt instrument as we are in a monetary economy. Nick thinks debt impose a burden on the next generation... Does Nick thinks of himself as the two-headed main character in Jacques Godbout's novel "Les têtes à Papineau"? ;-)

https://www.google.com/search?hl=en&q=les+t%C3%AAtes+%C3%A0+papineau

[Frances Woolley]

Jacques Rene - "But can we start now a supplication session to the gods of economic policy to insure a smooth transition to a world with a 20% larger workforce where a bunch of newcomers will be shut out of the labor market and where a larger worker pool will decrease wages?" Ouch! Add to that an older workforce is, on average, a less healthy (+ all the other stuff that happens with getting older) workforce and think about the effect of this on average productivity thus average wages...

On the two-headed main character - I'll ask him if I see him at work today.

[Jacques René Giguère]

Frances: which head will you ask and which one will answer? That should start a nice thread...

[Determinant]

@Jacques:

The problem has started already. Pension posts like this one ignore the young cohort, mine, that just wants a job, please. We see beating up on hold people as gratuitous and unnecessary, and the middle cohort as engaging in unrestrained self-absorption and self-interest.

I find it hard to imagine a demographic shortage when we have so many unemployed.

[Nick Rowe]

Jacques Rene: "Nick wants a fully funded pension plan which will invest in debt instrument as we are in a monetary economy. Nick thinks debt impose a burden on the next generation... Does Nick thinks of himself as the two-headed main character in Jacques Godbout's novel "Les têtes à Papineau"? ;-)"

It all depends on whether the national debt rises and falls with cohort size. The initial decision to create a lot of debt to "fund" the pension plan does nothing to remove the burden from future cohorts (it's just a way of recognising that burden). It's what happens afterwards that matters.

[Larry Willmore]

Interesting post and comments, but more information is needed on Australia. It so happens that I covered this quite thoroughly in a 2007 publication - http://larrywillmore.net/UniversalWD.pdf - from which I copy and paste two paragraphs:

Australia provides its elderly with a means tested Age Pension equivalent to 29% of per capita GDP for a single person, or 48% of per capita GDP for a couple. (These amounts exclude supplements for rental assistance and health care.) The age of eligibility for women has increased gradually from 60 years beginning in July of 1995, and will reach 65 years, the same as for men, in July of 2013. As in South Africa, the intent of the means test is not to tightly target the poor, but rather to eliminate the wealthiest from benefits of the Age Pension. Prior to 1992 contributions to pension plans received tax relief, but were voluntary and never mandated. In recent years, twothirds of the covered population has qualified for a noncontributory Age Pension, twothirds of these for a full pension. After a generous ‘free’ income allowance, the means test reduces the pension at the rate of 40 cents for each additional dollar of income. ...

The hope of Australian policymakers is that mandated retirement saving will reduce the future costs of Age Pensions. This was, in fact, the specific reason for adding Pillar 2 to the pension retirement system. Workers are expected to save more for their own retirement, and rely less on Age Pensions. This is not likely to happen, because there is a strong incentive for workers to retire early, live for a while on the sums accumulated in occupational pension plans, and arrange their financial affairs to maximize benefits of the Age Pension. This strategy, known in Australia as ‘doubledipping’, is possible because Superannuation Guarantee benefits can be withdrawn as a lump sum as early as age 55 (increasing to age 60). Government could effectively end this behavior by becoming more intrusive, forcing everyone to convert their mandated retirement savings into a life annuity. This is politically unpopular, so is not likely to happen.

[anonymous]

http://fairlysafedelusions.blogspot.com/2012/03/at-zlb-central-bank-should-monetize.html