Should MNDs be allocated to "constituencies"?

AMIPP suggested no quotas for the reasons given below.  The Trust proposes 2 employees and 2 retirees for the initial four MNDs.  In the terminology of the Code of Practice, this makes two constituencies.

Should deferreds be discriminated against?

AMIPP suggested not.  The Trust proposals give deferreds no say.

How many MNDs should there be?

AMIPP noted the potential for six.  The Trust proposals do not acknowledge this prospect.

Should there be a minimum length of service to be eligible to be a candidate?

AMIPP suggested "No".  None is specified in the Trust proposals.

How should nomination work?

AMIPP made no recommendation.  The Trust proposals require a few signatures, although these "supporters" do not have to vote for the candidate.

What happens if there are not enough nominations?

AMIPP recommended measures to make the MND role more attractive. 

What happens if an MND resigns mid-term?

The Single Transferable Vote mechanism provides for replacements and additions.  The Trust proposals do not acknowledge this.

What happens if candidate or MND circumstances change?

The Single Transferable Vote mechanism provides for replacements and additions.  The Trust proposals do not acknowledge this.

What should be the term of office for an MND?

AMIPP recommended three years, for the reasons given below.  The Trust plans four years.

What information, education & training should be given to prospective candidates and MNDs?

AMIPP notes that each person will be different.  They should be given what they ask for, unless it is prohibitively expensive.  The Trust plans do not address the topic.

What information about the MND process should be given to the membership in general?

The principle of transparency is vital here.  The membership should be told in advance about recounts and replacements, when there would be new elections, etc.  The full STV count should be made available.  The Trust plans have some generalities but almost no facts which would allow the membership to know in advance what would happen.

When should the MNDs be appointed?

AMIPP suggested January 2008.  The trust plans say May 2009.

Below here discusses the position as it was in May 2007.

Several places in Newsletters have contained information about the regulations requiring Trusts to progress from member representatives decided under "Opt-out" arrangements (like our MEDs) to Member Nominated Directors (MNDs).   MNDs have protection and empowerment because of the regulations, which our MEDs will not have.   So it is in the interest of scheme members to get the status of our representatives upgraded as soon as possible.  Our Trust is currently working on this.

There is a Code of Practice developed by the Pensions Regulator which covers the issues Trusts will want to consider.  It encourages Trusts to act quickly and gives some figures for delays that will be regarded as unreasonably long.  This web page contains the views of the AMIPP officials and advisers on how the issues in the Code of Practice should be tailored to out Trust.  As always, your comments via the message board are welcome.

The Code of Practice applies to thousands of very different Trusts so it cannot be expected to be very specific.  However, for all the Trusts it advocates the principles of proportionality, fairness and transparency.

AMIPP suggests that these principles are best met by what is nearly the simplest of possible arrangements:

- Selection by election, using the Single Transferable Vote method of ranking candidates.

- All those with a pension paid from the Trust, and those that will get a pension from the Trust if they do not die too soon, and those saving with the Trust, should be eligible as candidates for election and as voters.

- A full report of the voting to be made available to all voters.

- The MNDs should take office at the start of 2008.

- Replacement MNDs, if required, to be determined from the preferences voters gave in the election.

Here is how those proposals relate to the Code of Practice:

The principle of proportionality covers the fact that what is appropriate for a scheme with a lot of members may be out of proportion (too elaborate or too expensive) for a scheme that has only a few members.  As our Trust is one of the largest in the UK, it is appropriate for it to use an election.  Elections run by Electoral Reform Services using the STV method are the choice of several large organisations and that is a suitable choice for our Trust.

(Some people have interpreted "proportionality" as indicating MND numbers should be proportional - ie if the scheme has predominantly employees it should have predominantly employee MNDs. - but that is not how the code is using the term.)

The principle of fairness dictates that none of the people with an interest in the good governance of the Trust should be discriminated against.  It is fair to exclude those who no longer have an interest, such as retirees from the M-Plan who get their pension (annuity) from the insurance company they chose at retirement.  Exclusion by age alone would be unfair.  However, it is not unreasonable to exclude children who are receiving from the Trust because a parent died; these children are not receiving a pension.  (A pension lasts for a lifetime.  The benefit paid to children ceases after their education.)  It might not be unreasonable to exclude those receiving pensions as "survivors".  Although they have an interest in the good governance of the Trust, they are not regarded as "pensioners" in the regulations.

It can be argued that fairness is lost if the characteristics of the MNDs elected are not well matched to the demographics of the voters.  The AMIPP view is that a match is a good thing, if the voters choose that way, but there should be no interference with the mechanics of voting in an effort to match the election result to "constituencies".  This is discussed below.

The principle of transparency dictates that the Trust should avoid any secrecy about what happened  in the election.  It also requires that the Trust makes clear what the votes are for, for instance how far the votes will be used to decide replacement MNDs.  This is discussed below.

The concept of unreasonable delay has some flexibility.  There are advantages to synchronising MND tenure with either the calendar year or the financial year.  Since the need for arrangements has been known about for years, and formal consultation on the Code of Practice started in July 2005, some people will feel that the delays are already unreasonable.  However, there should be no difficulty in appointment of MNDs by 2008, given that the election arrangements will differ from MED elections only in details.

Here are some specific questions:

Should MNDs be allocated to "constituencies"?

The Code of Practice, paragraph 61, allows for constituencies, where there is a quota of MNDs for each constituency.  Our scheme has had quotas in the past.  There are a number of reasons not to have quotas.

- It is disrespectful of an MND to suggest they should represent a particular constituency.  MNDs are required by law to act equitably (fairly) for all members of the scheme.

- It is disrespectful to the voters.  How can it be desirable desirable to constrain voters in getting the MNDs they want?

- As implemented for our scheme, quotas contradict fair voting.  The Trust releases few details on actual voting but it is known that in one election a person who did not get to be an MED got twice as many first ranked votes as somebody in another category who did get appointed.

- With only a few MNDs, it is not possible to force the election results to match all the characteristics of the voters.  The Code of Practice gives as examples quotas by site, category of member, or scheme section.  But these are only examples.  If quotas are good, why not a quota for sex?  Or for age groups?  Or for final-salary/money-purchase?  There will always be some decisions to be made by MNDs where such characteristics are a factor.  With our four (or maybe six) MNDs, there is no hope to match all the relevant characteristics, even if there was cause to attempt that.

- Quotas increase the risk of an inadequate number of candidates.  It could happen that there are more candidates than MND places but not enough candidates in some category.  This is unlikely, but there is no cause to risk it.

- Quotas complicate the issue of replacements.  If MNDs are not replaced when they change category (for instance an employee MND retires), why should a candidate from one category not be a replacement for a MND of another category.? (Employee candidate replaces retired MND who died, when there is no retiree candidate.)

There may be some case for quotas as an issue of fairness - perhaps employee candidates have an advantage because their communications with colleagues are easier than the communications of retirees with past colleagues - but there is no evidence of such an effect in the past.

AMIPP would prefer there to be no quotas, for simplicity and the reasons above. 

If there are to be quotas then the principle of transparency dictates that the voters should know what category they are voting, eg if there are to be two employees and two retirees then the voting forms should provide an opportunity to vote for employee MNDs  and a separate opportunity to vote for the retiree MNDs.

Should deferreds be discriminated against?

All members have an interest in the good governance of the schemes and in IBM World Trade's long term ability to provide fund contributions.  Active members have the extra privilege of voting for the "Employee Forum" members, who take an interest in pensions.  [Employer/employee consultation is regulated in the UK and the EU.]   Not allowing deferreds into the MND process would discriminate against deferreds.

There are a number of reasons why deferreds might be treated as second class members, although the reasons are weak.

- It is slightly more costly and difficult to maintain a list of addresses for deferreds, in comparison with retirees, since retirees have a motivation to keep their address info up to date so as to receive the pension.  However, the Trust has to catch up with deferreds eventually.

- Deferreds will tend to have smaller pensions.  This is hardly a reason at all.  If anything, those with smaller pensions have the greatest need to influence the Trust since they can least afford losses.  Those that have very small pensions are not a consideration; the law allows them to be given a lump sum in place of their pension.  (OK provided the lump sum is calculated fairly.)

- The regulations pay less attention to deferreds than to employees and retirees.  This was probably because MPs felt that deferreds were less at risk from Trust actions than the others.  Deferreds won't be affected by scheme changes like contribution rates and accrual rates.  During the deferment their pensions are revalued so as to retain value in the face of inflation, so they are insulated from trustee attitudes to inflation for that period.  However that was how MPs felt at the time of the 2004 Pensions Act.  The revaluation during deferment is now under attack.

- Deferreds are less likely to know the candidates personally.  A weak argument, because most voters will know little or nothing of the candidates from personal acquaintance.

- Deferreds are less likely to vote even if given the chance.  Nobody knows if this true - it may be that the deferreds get so little communication from the Trust that they will tend to take what they get seriously.

On balance, AMIPP believes the principle of fairness should override the justifications for disenfranchising deferreds.

How many MNDs should there be?

This question is largely out of the Trust's hands.  There have to be 4 (out of 12) by law.  The Minister can change this to 6 without going to Parliament for approval of that.   There are good reasons to go to 5 or 6 voluntarily;  50% MNDs works well for Trusts that have it and with more MNDs there is more room for the voting results to reflect demographics.  But the company influence on the Trust Board won't allow a voluntary decision to have more than four.

The 50% MNDs was agreed between government and the unions as part of the "Warwick Agreement" prior to the last general election.  The unions will be pressing for it to be implemented before the next election.

What is clear is that the arrangements should be prepared for six MNDs.  So the voters should be told they are voting for six, even though two might become "reserves".  (And voters should be told that every candidate could get to be an MND, if enough other candidates dropped out of the reckoning.)

Should there be a minimum length of service to be eligible to be a candidate?

It is hard to see any argument in favour of this.  An argument about having a small pension is unsound.  An argument about talent makes no sense - years in IBM training to be a computer salesman or programmer would not necessarily make for a better MND. There is a case for putting the data on the voting form for all candidates, but no case for making it a condition.

How should nomination work?

The disadvantages of requiring some signatures of supporters are that it is discriminatory (easier for employees to arrange than others) and no guarantee (supporters are not required to vote for the candidate).

However, since candidates can be elected by default when there are insufficient other candidates, requiring some signatures will be some measure of ensuring against those MNDs whom the electorate really does not want.

What happens if there are not enough nominations?

Each election so far has resulted in fewer candidates than the one before.  MEDs have not always stood for re-election, and the reason for that has sometimes been disillusion about the way the MED was treated.  So it is worth considering the scenario of too few candidates, even though it is unlikely.  Quotas make the scenario more likely.

The first need is for more recognition of what MNDs do, particularly in the case of employees who put themselves in the firing line (perhaps not literally:-) by expressing opinions contrary to those of senior IBM management.  There is also a need to give MNDs the opportunity to do their job, by not denying them information they regard as relevant.

Ultimately it is likely that calling for more candidates would succeed, if the alternative was to have appointments made by the company dominated Trust Board.

What happens if an MND resigns mid-term?

The STV voting system has the merit of answering this question.  The vote counting is rerun with the surviving MNDs marked as to-be-elected and the dead-or-otherwise-ineligible candidates marked to drop out of contention.  (The secondary preference votes of those voters who voted for the drop-outs are still taken into account.)  

[This is what should have been done in the appointment of Elaine Kirkwood.  Elaine would most likely have been elected on such a recount but the recount was not done.  Instead the Trustees chose her on the basis of more than just the voting, thus depriving her of a democratic mandate.] 

An alternative would be to hold a new election.

There should be no period when the votes from the previous election are deemed unsuitable for MND replacement purposes yet no new election is to be held.  The rules for when one election ceases to provide replacements and a new election is required should be fixed before the results of the first election are known.

What happens if candidate or MND circumstances change?

It is an advantage of not having quotas that this does not present any problem.  If the change leaves the candidate/MND eligible then nothing needs doing.  If the change leaves the candidate/MND ineligible this would be treated like a resignation.

[The Mike O'Sullivan fiasco arose from quotas.  Mike retired shortly after being elected.  It is almost certain the electorate wanted him to be an MED, despite that.  He could have been appointed even after his retirement but the Trust&Company decided against that because it would have upset the quotas.]

Because there have been IBM UK lay-offs leading to early retirements, there will be a number of retirees (and deferreds) who have taken up or will take up non-IBM jobs.  These are likely to be Information Technology jobs, and hence to some degree "in competition" with IBM UK.  However, it is extreme to argue that would make them unsuitable for election as MNDs.  They will still have an interest in the success of the Trust.  There is a possibility of any trustee learning "IBM Confidential" information but negligible chance of that being used to IBM's detriment.  (We know from the Trustee's reaction to one scheme degradation quickly following another that the level and timeliness of the information that IBM gives the Trustee would make such information of little value to a competitor.)

The principle of transparency is against barring people on the basis of their job.  What would the members be told about who decides what jobs are "in competition" with IBM UK?  Whoever decides, what would the members be told about the criteria for the decision?

What should be the term of office for an MND?

There is a strong reason for it being three years, to make it match the cycle of Actuarial Reporting, which is legally set at three years.

A longer period would also be worse because (a) it would increase the likelihood of the mechanisms for replacing MNDs having to be brought into play and (b) it might lead to disillusioned MNDs staying in office longer (if they felt resignation was too big a step).

Six years would match the actuarial cycle but is too long. 

What information, education & training should be given to prospective candidates and MNDs?

Each person will be different.  They should be given what they ask for, unless it is prohibitively expensive.

What information about the MND process should be given to the membership in general?

The principle of transparency is vital here.  The membership should be told in advance about recounts and replacements, when there would be new elections, etc.  The full STV count should be made available.

When should the MNDs be appointed?

Operating without MNDs has no advantages for the members, while operating with MNDs has advantages.  The Trusts obligation to act in the interests of the members would lead to early appointments.  The Code of Practice describes what delays are reasonable - a total of a year between the regulations being relevant and the actual appointments.  Appointment dates of the calendar year 2008 or financial year 2008-2009 would be reasonable.

There might appear to be some cost-saving merit in using the views expressed by the electorate in 2005 for a longer period but the government has rejected this view - the "Opt-out" arrangements are obsolete and the people appointed under them have no status derived from those arrangements.