I've blogged a few times before about levelling down in respect of pension provision, because I find the arguments put forward by people like Ros Altmann wholly unconvincing. So I though I would try and explore the logic of the argument in reference to Personal Accounts.
The common claim is that when Personal Accounts come into force (or perhaps ahead of it) with the requirement of a 3% contribution from companies if employees don't opt out, many employers will "level down' the provision they offer. (Incidentally, note that depending on strength with which this point is expressed it falls into one of the two classic reactionary forms of argument as set out by Albert Hirschman. If, taking the stronger line, it is argued that Personal Accounts will result in less saving, it's perversity. If on the other it is argued that the reform will have no impact (because levelling down offsets the gains) it's futility.)
In practice levelling down by companies would presumably mean either reducing their contribution to a non-Personal Accounts DC pension scheme to the 3% level (from a higher level), or simply enrolling their staff into Personal Accounts and closing their existing scheme (which again must have been more generous than 3% in order to count as levelling down). Just to be clear - the only other possibilities are first, that the employer already has a scheme that is better than Personal Accounts and sticks with it, or that the employer did not have a scheme, or pay into a scheme, and now is forced to pay into Personal Accounts. In the first case there is no levelling down, in the second case there is levelling up.
So, why would an employer level down? Well one reason is because they can save money. Fine, but this does first lead to a question over why they don't level down now, rather than when Personal Accounts go live in 2012. In fact they could level down now to zero. So why don't they do so? There are at least three possible (not necessarily credible) reasons. First, they see some value to providing a pension, perhaps because it helps recruit and retain staff. Second, they fear employees' reaction if they do level down. Third, they are unaware that they could contribute less (nothing, in fact).
Obviously in the first case the introduction of Personal Accounts could actually reduce the value of providing a pension (because the comparative advantage to the potential employee is reduced) which could be argued to provide a reason to increase the level of company contribution. The second case is possibly more plausible. Companies may wish to reduce their pension commitments and Personal Accounts could provide the cover to do so. Companies could argue that their pension costs will go up (if everyone ends up auto-enrolled) and thus they need manage this. I think the final case is less plausible, but there may be small businesses out there which have a GPP or stakeholder set up with a contribution of over 3% and perhaps their adviser may make them aware of the incoming regime and urge them to respond.
But we can see that in practice most employers are not levelling down now. So they are not sufficiently concerned about costs to take pre-emptive action. And if companies do reduce contributions now, how much weight do we place on the introduction of Personal Accounts, which don't go live for another 3 years. To what extent is the company simply saving money, rather than being pushed into an action by impending reform? It simply is not clear that the reform 'causes' the company decision. Critics of Personal Accounts need to present evidence if there is any.
A more realistic version of the levelling down argument is that companies wish to maintain their current pension costs rather than reduce or increase them. What do I mean by this? Well, I think the main cause of any any adjustment to companies' pension provision will not actually be Personal Accounts at all, it will be the requirement to auto-enrol. Think about it - this is where the cost pressure really is. If I am running a scheme with a 6% company contribution maybe only 40% of my employees actually join it. But if they are auto-enrolled and only 1 in 5 decide to opt out (ie I now have a membership rate of 80%) then my pension costs have risen by 100%. If I then decide to 'level down' to the 3% minimum I keep my total pension costs the same, but the amount contributed to each employee's pension has halved.
This seems to me to be a not unrealistic (if very simplified) picture of how some companies may think and act. But it is clear then that the real issue for companies will be auto-enrolment (whether into Personal Accounts or existing schemes) not the new scheme in itself. It also means that - if companies adhere to my stylized version of responding to auto-enrolment - the 'levelling down' refers to the amount contributed to each pension, not the amount contributed in aggregate.
This is important, because it means that what is actually implied by critics of auto-enrolment (not Personal Accounts!) is that it is better for employers to contribute more to fewer employees than it is to contribute less to more. This is not an unreasonable position (for example because they may think that the minimum contributions are too low) even though I don't share it. But it is a position that is not made explicit - either because those advancing the argument don't want to be explicit, or because they haven't thought through the logic of their position.
A final point - obviously a company facing auto-enrolment could choose to level down to an extent that their aggregate commitment is lower as well as their average contribution rate. This would be levelling down in both senses. But once again it is the company's decision to pitch their commitment at this level, and again it is not clear how the reform 'causes' them to take this step. Once again, it's over to the critics to provide the proof.