As part of the process of evaluating how they’re doing with various programmes, the California-based Hewlett Foundation’s staff meets regularly to look at their ‘spectacular failures’ and ‘the least successful grant of the quarter’. Caroline Hartnell asked Hewlett President Paul Brest what the Foundation gains from this, and what prevents this sort of information being shared more widely – something that would surely benefit both foundations and their non-profit grantees.

The language of success and failure

‘You need to be really clear on your objectives and on what would constitute progress or success in order to talk the language of success and failure. If we’re not clear about what we’re trying to accomplish, how will we know whether we have achieved our goals?’ This is Paul Brest’s first point. His second point: ‘I think one reason why foundations don’t share as much information about successes and failures as they might is because they don’t have the information.’

So how do you get to being able to talk about success and failure? ‘I think a lot of people find it easier to talk about apparent successes and pat themselves on the back.’ But the goal isn’t just to talk about individual grants, but about series of grants, or portions of programmes. ‘We don’t look at it just at the grant level. We are fairly strategic in organizing our programmes to achieve certain objectives.’ Through asking whether these have succeeded or failed, ‘we begin to get comfortable within the staff with the language of failure’. The next step is to begin talking to the board in that language.

Why is this important? ‘My own sense,’ says Brest, ‘based on my life history, is that we probably learn as much, perhaps even more, from our failures as from our successes. So this really is a learning opportunity.’ Nor is it just spectacular failures they look at. Each quarter, Hewlett staff look at the least successful grants in the quarter. ‘That’s part of a broader project of evaluating on a regular basis how we’re doing in respect of components of various programmes.’

‘What we’ve been trying to do is try to identify where the failure arose,’ he explains. The explanation could be poor due diligence, or it could be that the theory of change underlying the grant, or the whole cluster of grants, is not a strong one. One pattern that is beginning to emerge – ‘and this is no big surprise’ – is how much organizations depend on their leadership. ‘A change in leadership, even in a fairly large organization, can quite often lead to, if not catastrophic failure, much less success, when the grant was his or her project.’ But Brest also emphasizes that a grant, or a whole programme area, can fail when nobody did anything wrong. ‘We and the grantees together took a risk, and if you’re taking risks then there has to be a failure rate.’

The advantages of sharing information more widely

Always assuming that they have this sort of information available, what would be the advantages to foundations and NGOs if foundations shared what they learn more widely? The advantages are obvious: ‘The whole field becomes better at doing its work.’ He draws on the analogy of hospitals: ‘They do their morbidity and mortality rounds each week, and increasingly this information is made available nationally, so they can learn what works and what doesn’t. Obviously most of what foundations do doesn’t have this precision of outcomes, that is to say the patient gets better or dies. But surely this sort of process would improve the work of foundations, just as it has advanced the medical profession.’

And the barriers? One major barrier is egoism. ‘It takes a very strong ego – or the absence of any ego whatsoever – to acknowledge failures. That’s obviously not a justifiable reason for not sharing information.’ However, things become more complex when you’re sharing information not about your own failure but the failure of the grantee. ‘I think we all feel more comfortable talking about what we wish we could have done and didn’t, than about somebody else, at least in public.’ There are of course issues to do with honouring the relationship with the grantee, ‘and just making sure that you’re not singling out the grantee in a way that’s unfairly disadvantageous, especially when the people to whom you might make this information available may not know how to read the failure.’

After only three years in the world of philanthropy, Paul Brest is not willing to guess how important this motive might be in preventing people sharing their failures as opposed to ego problems. ‘I mean after three years I feel I understand only one foundation.’

Risk and failure

I’ve been told that the Ford Foundation reckons that it would itself be failing if it had a failure rate of less than 17 per cent, because this would show it was not taking enough risks in its grantmaking. If foundations were generally more comfortable with the idea of taking risks, if they saw it as something they should be doing, wouldn’t acknowledging failures and learning from them follow naturally?

In answer, Brest first returns to an earlier point: ‘You really can’t talk the language of risk-taking if you’re not clear about what you’re trying to accomplish, because unless you do that you don’t know if you’ve succeeded or failed.’ And no, he doesn’t think that risk-taking is by definition something foundations do. ‘Different foundations define their mission in different ways. You can imagine a foundation primarily supporting very safe organizations, such as universities, theatres, operas, orchestras. Then you can imagine them supporting experiments where the outcome is not at all clear. My guess is that the big foundations – Ford, Rockefeller, Carnegie – do both.’

Nevertheless, for those that do accept it as at least part of what they should be doing, they should surely be more comfortable with the idea of failure? Brest agrees. ‘I think that’s right. Unless you have a spectacular run of luck – like flipping a coin and getting heads a thousand times – there will be quite a few failures.’

He also admits that foundations are uniquely capable of taking risks – at least among American institutions. ‘Businesses take risks but they’re financial, not social risks. Governments are on the whole too politically timid to take risks. And that leaves it to the non-profit sector, foundations and their grantees, to take significant risks to achieve social change.’

How are foundations accountable?

The conversation then turned to foundation accountability more generally. I start by asking what sort of accountability Hewlett looks for in its grantees. His answer harks back yet again to being clear about what you’re trying to achieve. ‘The first thing we do, as part of the application process, is ask them to make clear what their objectives are and how they expect to accomplish them. We want to understand their theory of change, how they plan to get from here to there.’ This is the document against which progress during the course of the grant will be measured. ‘After a year, three years, you look at it and ask, has it happened? If it hasn’t happened, why hasn’t it? Did we achieve our shared outcomes?’

And what about foundations, how are they accountable, and to whom? Fiscal accountability he sees as the easy part. ‘You are accountable fiscally in the public realm; you file yourtax return, and those who are interested can read it.’ But accountability in terms of achieving your goals is more difficult. ‘Evaluation is the core of that accountability, and it’s the evaluation of your own worth, which ultimately in most grantmaking foundations is accomplished with the grantee.’

The future for foundation accountability

What changes in foundation accountability does he anticipate in coming years? As far as voluntary accountability is concerned, his hope is clearly that foundations will move in the direction of sharing more information and greater transparency about what they’re trying to achieve and how they’re getting on. ‘But it’s not just information about successes and failures. Foundations have huge amounts of information, sometimes systematically collected, sometimes not, about the substance of what they’re doing. They commission reports from consultants. I think we’re moving in the direction of sharing that information, but foundations haven’t been very systematic about keeping it or organizing it. A lot of it resides in the head or the drawer of a programme officer.’

For Brest another interesting questioning in the US now is whether there is likely to be more government-required accountability. His view is that this is likely to happen on the financial reporting side, in both the corporate and the foundation sector. On the social investment side, ‘what we actually do with the money we have’, he feels there may be some kind of crude accountability requirement in terms of disclosure of, and perhaps some limits on, administrative costs, executive compensation, etc. ‘But I would be very doubtful about accountability for outcomes. Outcomes are sometimes very difficult to measure. Also, they really depend on what your objectives are, and foundations have vastly different objectives.’

Supporting general operating costs

The Hewlett Foundation has recently become involved in an effort to get other foundations to be more willing to fund general operating costs. A meeting in New York in June, which Brest initiated, brought together 50 or 60 people, a mixture of foundations and non-profits and some other organizations. The goal, says Brest, is to come up with some guidelines. They’ve formed a working subgroup, which is now close to distributing to the general group for comments a proposal that a foundation should have a presumption in favour of giving general operating support. ‘That doesn’t mean that’s the only thing they should do. Some foundations, even whey they think about it, will decide it’s not for them. They should also have a presumption that when they fund projects they should pay the full overhead.’

At this point, Brest stresses, nobody can speak for anybody else. ‘But my hope is that we can develop a consensus within the foundation world around the need for general operating support. It’s easy to get a consensus among non-profits – every non-profit wants general operating support!’

The Hewlett Foundation itself allocates approximately half of its grant budget for general operating support – ‘and that makes us a real outlier among large foundations. Most foundations don’t do that much.’ He says that it’s not at all unusual for Hewlett to renew a grant again and again. But he does offer a caution here: ‘If we change direction, as foundations inevitably do, we think it’s important that a good grantee is given a soft landing rather than just finding out one day that a grant won’t be repeated. With a grantee who’s doing poorly, that’s obviously a different story.’

Trustee compensation?

Given the current controversies raging in the US, I couldn’t resist asking Paul Brest his views on trustee compensation and foundation payout rates. As regards the first, Hewlett’s own policy is not to have compensation for its board of directors. But Brest is careful to emphasize that it depends what trustees do. ‘There are some very small foundations where the trustees are in effect the management. That seems to me quite different from a foundation like ours, or like most of the big private foundations, where there are professionals managing the grants and the trustees have oversight responsibilities.’

What Hewlett does have, however, is a very generous gift matching programme. ‘For us this is very appropriate. What we’re trying to do is encourage philanthropy, so up to a limit any gift that a trustee makes to a non-profit will be matched four to one.’ The maximum they will provide in a year is $40,000.

Foundation payout?

Here Brest feels people are asking the wrong question. ‘The fundamental issue is a normative one, and then there’s the empirical one. The normative issue is should foundations be allowed to exist in perpetuity, if their founders want them to? Potentially, a payout rate can decapitalize the endowment, which cannot then exist in perpetuity. I happen to be the president of a foundation whose founder wanted it to exist in perpetuity, but I think there are arguments to be made on both sides. What’s interesting to me is that in so much of the discussion about payout, that issue hasn’t been central, and that seems to me to be the question.’

The empirical question is a highly technical one, the issue being what payout rate will permit a foundation to exist in perpetuity. ‘Most of the analysis I’ve read suggests that in the long run – that is over a period of 50 or 100 years, which I guess is as close to perpetuity as we get – 5 per cent is the most you can pay out without decapitalizing the endowment.’

But in Brest’s view this empirical question, which is the one most people are looking at, doesn’t really come into play until you’ve answered the normative question. ‘The empirical question’s the tail, and the dog is the question of perpetuity.’

Paul Brest has been President of the William and Flora Hewlett Foundation in Menlo Park, California since 2000. He received an LL.B from Harvard Law School in 1965 and practised with the NAACP Legal Defense and Educational Fund in Jackson, Mississippi, doing civil rights litigation before joining the Stanford Law School faculty in 1969. At Stanford, his research and teaching focused on constitutional law and problem solving/decision making. From 1987 to 1999, he was dean of Stanford Law School. He can be contacted at p.brest@hewlett.org

The Hewlett Foundation was established in 1966 by the late Palo Alto industrialist William R Hewlett, his wife Flora Lamson Hewlett, and their eldest son, Walter B Hewlett. It concentrates its resources on activities in conflict resolution; education; environment; children, family and community; performing arts; population; and US-Latin American relations. The Hewlett Foundation is wholly independent of the Hewlett-Packard Company and the Hewlett-Packard Company Foundation.

For more information
http://www.hewlett.org