Hosting arrangements
REC Management Note No. 1
For many small organisations a hosting arrangement, whereby your staff are employed by the host organisation, may be beneficial. The host takes care of providing an office with all auxiliary services, payroll, travel and financial management. All this is for a fee, and it may compromise your identity as a separate organisation. In recent years many of the roles played by a host can be purchased from separate service providers (e.g. shared workspace and payroll services), which presents a third alternative to hosted versus independent organisation.
This note is reflection on hosting arrangements, from the perspective of the hosted organisation, which I prepared in response to a request from Hewlett Foundation. I have anonymised the content somewhat, but since people know, or can find out, where I have worked, I have not put too much effort into that. All the people concerned have moved on from their positions since the circumstances I describe below.
I had had three experiences of hosting arrangements when I prepared this note. When I set up 3ie it was first hosted by the Global Development Network (GDN) in Delhi and LIDC/LSHTM in London. When I became CEO of Campbell it was hosted by the Norwegian Institute of Public Health (FHI). I negotiated the first two, the hosts being identified by a competitive process, and inherited the third, which existed as the Norwegian government gave funds to FHI which were earmarked for Campbell.
Common purpose and inevitable tensions
Your host should be an organization with whom you have a common purpose. That may extend to joint events or other joint activities. This does not preclude tensions in the relationship. Staff members represent the interests of their own organisation. That is their job, so you can remain perfectly civil so long as both sides are behaving reasonably. But where resources are concerned, disagreements are inevitable. As economists say ‘complete contracts’ are rarely possible, so there will be unanticipated situations which is where disagreements may arise.
The common purpose counted for less than expected. In principle 3ie and GDN have overlapping missions. The agreement spoke of exploiting complementarities and we regularly discussed this. But it didn’t happen. It didn’t happen as agencies just stick to what they do in their mandate, and they guard their turf so weren’t willing to share opportunities that came their way. 3ie was always a much more sharing organization. That is generally my approach – I don’t see working together as a zero sum game. We do have joint activities with our London host, LIDC, which I think is one reason they have been so lax about other aspects of the agreement. Here the relationship has mattered more than the formal agreement.
Project or legal entity?
An important question for a small organization, especially a small research organization, is whether to be a project or a legal entity. Universities run many multiyear projects which have larger budgets than small research organizations, and so can readily accommodate your organization as a project. There are pros and cons of both sides. I would favour the independence being a legal entity gives you, which includes the ability to leave the host when you feel the time is right. If you do want to be an independent legal entity, then you need to be sure that the host has the same understanding. One bidder to host 3ie was rejected on the grounds that their proposal was very much to treat 3ie as a project.
Management issues in hosting
Once you are hosted, if all staff, including management, are staff of the host organisation then in principle they report to the host management and its board, not the board of their own organization. In the early days of 3ie the 3ie Board was very exercised about the issue of ‘who runs 3ie’ for this very reason. In principle, I, as a GDN employee, in my capacity as 3ie Executive Director, reported to the GDN President.
The situation is different, but still potential problematic, if management are not employed by the host, which was the case for 3ie in London and Campbell in Oslo. Management are not officially the line managers of the people below them.
But none of this was an issue for 3ie at LIDC although the DD there nominally reported to Director LIDC, he de facto reported to me although I was not a LIDC employee. Although the 3ie donors and board were not keen on the gentlemen’s agreements which worked with GDN and LIDC, and they are no substitute for formal agreements, a good relationship at the managerial level seems key to a successful hosting arrangement.
Financial management
When 3ie started hosted by GDN. I had thought GDN financial systems would provide the financial reporting required. It was not true. The hosted organization needs its own separate financial management: if possible, at least one finance officer, an accountant, and for sure their own audit.
Even in the first couple of years when 95% of 3ie expenditure was via GDN, 3ie needed its own separate check of 3ie expenditures being processed by GDN, even though this involved some duplication of effort. When we first started our own financial section I was concerned of capture by GDN’s own finance team but that didn’t happen. Our finance staff were also clear that they represented 3ie.
The hosted organization needs to have control over its own budget, which means (1) a separate bank account, (2) be responsible for its own budgeting, (3) be responsible for approving expenditure, and (4) have your own financial staff. None of these were the case in the hosting arrangement I inherited at Campbell, and it proved very problematic. At the first meeting I had in Oslo I asked ‘whose money is this?’. There was not agreement on that,
A final point on finances is that if the organization starts small, as 3ie did, and as Campbell is always was, it is difficult to achieve the required separation of duties required by an auditor. Being hosted can help with that if the host is part of payment processing, so that is an aspect to take advantage of, though you would want to be sure that the host won’t question payments on substantive grounds or hold up your payments. That was never an issue for 3ie, except for persistent late and under-billing from LSHTM which required building up substantial accruals in the accounts which the auditors didn’t like. This may seem to run counter to my recommendation to control finances. What I mean is that the hosted organization has to have effective approval. The host should not hold up or refuse payments on substantive grounds, but can and should do so if the paperwork is not in order. So when considering a possible host you really want to look the audit reports of the host organization – not the audited accounts but the audit letters for the last three years to be sure their own financial systems are sound.
Assess if the host can deliver the required services
3ie was a grant-making organisation (REC Management Note 2 will be on grant management). We planned to issue dozens of research contracts in our first year. I wish I had known that GDN actually had so little experience processing contracts. Since GDN was supporting research mainly through regional research hubs, it did not have the experience of itself making large numbers of contracts with research teams. The contracts GDN proposed after quite some time were woefully inadequate. 3ie had to draft the contracts to be issued, which it then took GDN a long time to approve as they had no in-house capacity in such matters. The general lesson here is to clearly lay out the expected functions of the host and have some way of ensuring that they have the capacity to deliver what you expect them to deliver e.g. we could have asked GDN to show us sample research contracts they had issued, and a list of such contracts for the last couple of years. Once we had the contract templates and processing system in place, all this went well.
But there were other problems which arose from different understandings and an incomplete contract. The problems were: (1) delays in contracts, already mentioned, (2) staffing, (3) staff relationships and support and (4) finances.
Whilst the hosting agreement identified staffing requirements for 3ie, GDN initially tried to meet these obligations by allocating existing staff part time to 3ie tasks. This didn’t work as the staff were not suitable for the tasks and the 3ie jobs were not part time jobs. Since the staff didn’t report to me they always gave GDN work precedence over 3ie work. We negotiated one by one to get these posts changed to be full time 3ie positions with new hires to fill the positions. The general lesson is again to be very clear in the hosting agreement (and during negotiations) as to the expectations as to the set up including staffing. More specifically, don’t leave staffing to the host and be very clear on reporting lines. Avoid any joint appointments between host and hosted.
Issues around staff management were always contentious at GDN. Since 3ie staff are de facto GDN staff then there is a need for equitable treatment. For example, GDN routinely gave all staff the maximum performance rating each year, making it very difficult for 3ie to link pay rises to performance because GDN weren’t doing it. 3ie staff always complained that GDN HR didn’t treat them the same as GDN staff proper, e.g. with respect to staff transport (the office location required that female staff be provided transport to leave in the evening). Again, this is something to be clear about in the contract.
As has already been made clear, finances is a further instance of clarity, you need to be clear on what is covered and what is not. That worked reasonably well with GDN and no problems at LIDC. But 2016 FHI were slow to tell us the overheads they planned to charge for each year.
A specific point was that the structure of the agreement was that GDN would only make payments for funds paid for that specific purpose. This arrangement allowed GDN to hold large balances of 3ie funds. It was a simple issue to solve but took some time for GDN to agree (interest rates weren’t 0 per cent then so these balances generated healthy interest income). Avoid such problems!
Have a Plan B
We always intended the hosting by GDN to be temporary, and we left after six years. We could have left one, or possibly two, years earlier, but before that would have been premature. The further delay was because of the amount of time it took the Indian government to give us branch office status. This was our Plan B, we made the application early on to have in our pocket in negotiations with GDN, but GoI approval took over 4 years.
We always planned to leave GDN as Delhi was 3ie’s main office and we wanted to be able to have our own clear identity. That was never a problem in Delhi itself – 3ie is far better known locally than GDN which has made little effort to engage in India. But 3ie contracts are GDN contracts, and grantees acknowledge GDN as well as 3ie. One donor (CIDA) insisted on making its grant to GDN not 3ie. All good for GDN not 3ie.
We also broke away to save money. GDN was taking a large annual fee. They always said that was the actual cost, though we knew it was at least double true cost. When they realized we were serious about leaving they offered a substantial cut off the fee. Without a credible Plan B we were of course locked in to the host making it difficult to negotiate. Also for the annual fee increase the hosted organization is in a week bargaining position, so you want to have a clear formula in the agreement to limit that increase.
We never considered moving the London office out of LIDC whilst I was there. It is a relatively small branch office of 3ie and we require very few services from the host office than office space and hiring our staff. It is not worth the hassle of acquiring the status of a legal entity in the UK.
19th August 2024
(Based on a note prepared in June 2016)
