You can’t always get what you want: ‘amount applied for’ in grants data

You can’t always get what you want: ‘amount applied for’ in grants data

This blog post was written by Nick Perks, Trust Secretary at the Joseph Rowntree Charitable Trust

In the spirit of an empowering approach, at Joseph Rowntree Charitable Trust (JRCT) we like to award the full amount that an applicant requests. “Don’t spoil the ship for a ha’porth of tar” has become something of a saying in our grant committees. However, despite these good intentions, it is not uncommon that we will in practice offer less than the full amount requested.

There are at least four (usually) acceptable reasons for offering less. If a project budget looks inflated, then a slightly smaller award may be perfectly adequate to get the work done. If a project comes in distinct phases or workstreams, a funder can – ideally after consultation to check viability – legitimately choose to fund one workstream or phase and not another.  If an applicant has several funding applications pending, then a smaller grant may help to lever in the matching balance. Finally, where a grant is for general core support, with no specific requirements on outcomes, most charities would cheerfully accept any level of donation.

There are also at least two (usually) bad reasons for offering less than the applicant requests. If a grant committee is split, or sympathetic but not entirely convinced, then offering less can be a compromise between offering everything and nothing. Alternatively, when facing acute pressure from a number of excellent applications, offering less than requested to several groups can be easier than turning some down outright. Both of these fudges, however well meant, can mean simply setting grantees up to fail.

As part of reflecting more generally on our funding model, JRCT trustees asked staff to look into how often we “trim” grants. Trustees and staff anecdotally felt we might do it too often, but we needed some data.

We started with in-house numbers. Between April 2014, the start of our current funding programmes, until the end of December 2017 we made just over 300 grants. Of these, 47% were given less than the amount that they requested in the application. So roughly half of applicants get what they requested, and half do not. This was higher than we’d expected.

We then compared this with a similar group of 350 grants made a decade earlier between April 2004 and December 2007. In that period, a whacking 60% of applicants got less than requested. The reduction since then is encouraging – either we are getting better at guiding applicants as to the likely level of funding, or we are getting better at avoiding fudges, or both.

360Giving data helped us start to understand how our practice fits with other funders. Most grantmakers sharing their data in the 360 Standard don’t give information about the level of request, but some do – including the Big Lottery which makes up a significant proportion of all records. In total, out of 263,897 grants listed on GrantNav (at the time I downloaded the data), 176,619 grants (67%) include the amount requested. Of those, 32,150 didn’t get the full amount requested – about 18%.

Every statistic invites a new question: Is this trend changing over time? What is the range of practice across different funders? How much are grants trimmed by? And so on. Anyone can go and interrogate the data further.

Already, I can say to trustees that – based on the data available – we are trimming less often than we used to, but that we still trim relatively often compared to the wider sector. Getting a clearer picture of our practice then helps inform any discussion of if and how we can improve it.

A final note of confession: We don’t yet complete the ‘amount requested’ field in our own open data, so the stats above for JRCT are based on internal analysis. Something we could aim to change in future. One of the joys of 360Giving is that you can get involved at whatever level you are ready for, and then add more information over time.

Credit for cover photo: