Winner of the New Statesman SPERI Prize in Political Economy 2016

Tuesday 5 June 2012

Unfunded pension schemes and intergenerational equity

                In two earlier posts (here and here) I looked at issues involving debt and intergenerational equity. The second attracted a lot of interesting comments. Some had difficulty with the idea of debt as a burden, and I think one way to help here is to look at pension schemes. (I hope to return to other comments later.)
                When I talk about intergenerational equity to students, I go through all the ways that the current generation is exploiting future generations, like climate change, rising house prices, and rising government debt. I also say that of course the older generation could just get the younger generation to pay them directly. I then reveal, to the surprise of some, that that is exactly what happens in many countries because these countries run unfunded pension schemes.
                An unfunded scheme is where the working generation pays social security contributions, and that money goes straight into paying the pensions of the old, rather than buying some kind of asset (hence unfunded). When the scheme starts, the current old receive a windfall: a pension without having contributed anything. The young pay contributions, but then receive a pension when they get old from the new younger generation. So the older generation when the scheme starts are clearly winners, but are there any losers? The answer depends on two things.
                The first is whether the scheme ever stops. If it stops, the young in the period beforehand are clear losers. They paid contributions (which went straight to the then old), but get nothing when they get old. Obviously the scheme is a huge burden on this ‘final generation’. However if the scheme goes on forever, there is no last generation, so there is no loser on that account.
The second issue is whether those who are forced to save by contributing to the pension lose out because they could have done better saving for themselves. That in turn depends on whether the rate of interest (r) is greater than the rate of growth (g). Those who contribute to the pension scheme get back more than they put in because of economic growth. The income of the young as a whole will be higher either because of technical progress which raises income per head (assuming contributions are a fixed proportion of income), or because there are more of them which raises the number of heads. In either case the current young generation contributes more than the old did when they were young, so the old benefit from that extra money. However if they had saved themselves their return would be the rate of interest. So if r>g, each generation of young lose out a bit by being forced to save in a way that produces a return equal to g rather than r.
All that is happening here is that money is passing from one group to another, with the government acting as an intermediary. Society is ‘paying itself’. But this statement says nothing about intergenerational equity. Suppose the scheme only lasted one period. Money passes from the young to the old, but it would be absurd to say that this implied that the young were not losing out.
Those who read this earlier post will see the parallel with the case of government debt. With pensions the young get an implicit promise that they will receive their money back with a return equal to g, whereas those that buy debt get an explicit note saying they will get their money back with some rate of interest r. The implicit/explicit thing is not crucial here – after all the government can default! The key difference is that, if people are to buy the debt voluntarily, they get a return equal to the return on other forms of saving (=r). If r>g, taxes have to rise to make up the difference between the two when debt is issued.
An unfunded pension scheme has the same macroeconomic costs that I discussed before in relation to government debt. In particular, if the young save at least part of their pension this way, they will save less in productive capital, and if the amount of capital in the economy as a whole is less than the optimal level, this will be an important cost. In addition, to the extent that the contributions act like an income tax, it can distort work effort.
Does this mean that an unfunded pension scheme is a bad idea? You can see why it might be introduced even if it was a bad idea – its introduction is great for the current old, so they will always vote for it. There are two potentially important benefits, if the alternative is funded private schemes. First, it takes away a lot of the uncertainty in funded schemes. If you buy your private pension when the stock market is high, and retire when it is low, you could lose out in a big way. You might also want insurance against labour income risk. (See, for example, this paper by David Miles.) Second, it provides a safety net for those who were misguided enough not to contribute to a private pension. (For a useful reference on pension issues, see Oxford Review of Economic Policy Vol 22, No1, Spring 2006.)
                I think this nicely illustrates why intergenerational equity can never be the overriding concern when it comes to issues involving government debt or unfunded pension schemes. There are many other factors to consider that may be more important. If it was decided that the costs of unfunded pension schemes exceeded the benefits (or, more realistically, that the balance should move to funded schemes), then the generations involved in any transition would almost certainly lose out. They would become like the final generation in my discussion above. That issue of intergenerational equity should be important in any decision, but it should never be the only consideration.


  1. What do you make of this:


    Looks like either:

    1. Just another drop in the ocean/wheeze (like "credit easing")

    2. Osborne has suddenly heard of this Keynes fellow and is trying to pretend he isn't uturning by giving the scheme a s*xy name ("growth bonds") and keeping everything off balance sheet PFI style (i.e. we already have "growth bonds - they're called "gilts")

  2. Readers who want concrete examples could look at Australia and New Zealand.

    NZ introduced a funded pension scheme in 1974 and, on a change of government, promptly replaced it with an unfunded scheme, for exactly the reason Simon gave. In 2007 NZ belatedly started to move back towards funding superannuation, a policy which has again been diluted by a change of government.

    Australia had various funded pension schemes in the 1970s, and progressively moved to consolidate and expand them.

    Comparisons of savings rates, growth rates (a function of savings in this, the short-medium run) and real wages in the two countries since 1970 illustrate the theory nicely.

  3. I find it helps to think of money as a "tool to mobilize and direct human effort." If I save for my old age, I trust that my savings will be sufficient to mobilize farmers, builders, caregivers, etc to do for me after retirement what I will not be able to do for myself.

    Money is not the only tool for that purpose -- in a society where there is little money paid out to the largest fraction of the society, effort can instead be mobilized through pity, charisma, threats, shaming, fraud and theft, among others. But these are less useful, requiring continuous luck and effort, and cannot be cleanly stored like cash can.

    However, in a society where cash has pooled in a small enough reservoir, two things happen -- you get inflation in the oasis and deflation in the surrounding desert and, if the pooling is severe enough, the cash ceases to be a universally accepted medium of exchange.

    Are Americans at the oasis/desert point? We're certainly approaching it. Only the fact that people here cannot live off the land, cannot withdraw from the cash economy, prevents the development of a parallel society of people abdicating altogether. I wonder how long it might be till lay-monasteries and beguinages begin to pop up? A generation, I would guess -- in the USA the pattern doesn't yet exist to be adopted easily (I imagine the memory of hippie communes has a lot to do with that.)

    Noni Mausa

  4. Your one period, two generations model suffers from a bad case of the Doctrine of Immaculate Conception. Old and young generations do not just spontaneously arrive on the scene with capital endowments. This is obvious when you move forward in time because we know there is not last generation. But it works the same going in the other direction. In the old days adults would have a bunch of kids, hope half of them lived to adulthood, they provided the young with human capital skills (e.g., how to pull a plow), and then eventually the young would inherit some shack of a home where mom & dad would live out their days to the ripe old age of 42. So it was the older generation that provided the capital endowment to young workers in exchange for support in their old age. A more efficient way to handle support for retirees is to take some of the higher returns from labor mobility and offer mom & dad a pension. More efficient still is to have the govt handle that pension. Looked at in this light it's actually the last generation that reaps the windfall because they can consume all of their capital.

    Also, taking money out of a govt funded pension and diverting it to a private pension obviously increases risk, but less obviously it also lowers the return on the private pension and increases the return on the public pension.

  5. Dear professor:

    - you are assuming government is simply transfering resources and that pensions adjust accordingly. However, the situation is more complex, with pensions adjusted slowly and therefore with room for not transfering all money, even in the medium term. This may help the government to build a better financial position, or in the converse case, to issue more debt. This should be considered as well. The assumption of full transfer seems to me very extreme.

    - with the same quickness that you consider uncertainty in funded pension system, you should consider it for unfunfed pension system since incomes are also affected by the cycle (particulrly employment, as wages seems to be less cyclical). If you keep assuming that all money collected is transfered to old generation, then there will also be volatility in pensions. Very likely, risk insurance is easier to get for financial assets (of a funded scheme) than for wages or employment!

    - funded schemes may also be compulsory, as it was since the beginning in Chile, first country to introduce it as a universal and unique system, in 1981. The problem however is that if a country has a very large informal market then people will not be on the scheme, linked to formal contracts (partly corrected with reforms in 00's by also adding a solidary pension for poor).

    - funded schemes boost investment and the development of capital markets (provided good regulation), which is relevant for less developed countries. This may move the economy closer to the point of dynamic inefficiency, but still not on it, hence improving all generations.

    - finally, is the issue of sustainability, key in european societies nowadays. It is politically inviable and welfare decreasing to keep unfunded schemes when g and n are so low, since that means pensions amount to decrease, and as you say, less optimal since r tend to be higher than both g and n (g+n?). And government debt or taxes cannot permanently be used to close the gap. There has to be another solution, in particularly considering that g or n will very likely not revert their tendencies in Europe.

    Best regards.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.