Voluntary sector pay: Evidence is not the plural of anecdote

In my first post on the Inquiry into senior pay in the voluntary sector, in response to some comments I highlighted that the evidence brought to bear so far is partial. I think this is important and highly relevant; and the comment of an HR professional is that one of the starting points for setting pay is know your market.

Whilst the word ‘market’ is, for some, at the very root of the problem of senior pay, I think we do need to start gathering quantitative evidence on salary levels. Yet again, I don’t have all the answers, but I thought a blog might be a good place to start assembling the answers.

But first of all, we need some questions. My starting six questions are:

1. What sources of quantitative data on pay exist and how reliable are they?
2. What ‘descriptive statistics’ do we have about pay in the voluntary sector? What are the mean, median, minimum and maximums, and ranges of pay?
3. What does the distribution of salary levels look like – how many employees, for example, are in the top 10% of salaries? And how many organisations are within the pay bands?
4. What and who should we be comparing (or benchmarking) salaries with? Is it by sector, industry and/or job role? Is it just within these groups, or with others? And how relevant still are scales such as NJC (often used by smaller voluntary organisations)?
5. This probably then begs the question: is there a relationship between pay levels and other objective data we have for voluntary organisations, such as expenditure?
6. How is pay changing over time? Is it catching up with other sectors or falling behind? And how does change compare with benchmarks such as inflation?

I’ve been accused elsewhere of writing ‘sanctimonious waffle’ so I hope that these are the sort of objective questions that the panel will ask. What else do you think we need to ask?

Data, data everywhere, but not a drop to drink

Inevitably, a real challenge is going to be the availability of reliable data to answer these questions. The key word for me here is reliable: the drip, drip of examples in the media is a very good basis for selling newspapers, but less useful for serious discussion. Evidence is not the plural of anecdote.

I’m aware of the following sources…

Salary surveys and benchmarking exercises

Include xPert HR’s voluntary sector salary survey, Croner Reward’s survey (this one is a bit old),  and of course Acevo’s CEO salary survey. These sorts of surveys are useful, but they of course depend on who chooses to take part. They’re not random samples. My fellow anoraks will know this limits their usefulness if you want to generalise about the sector as a whole.

Charity accounts

These report total spending on employment, numbers of staff and, for larger charities, the pay, in bands, of the senior staff. NCVO’s is here, with salary details on page 39. These data are the source of numerous league tables – even the Guardian was doing this stuff a decade ago. It’s worth noting that in some charities, the CEO isn’t the most highly paid person, but you can’t tell that from the accounts. It’s also worth noting that sector-wide analysis of average salaries is difficult because some charities report head counts, others report FTEs. Nevertheless, we’ve got salary data for 10,000 charities from our work to produce the Civil Society Almanac. But it needs cleaning to address errors and other problems associated with it, which will take some time.

The Labour Force Survey

This is a very, very large survey (60,000 households every quarter) run by the Office for National Statistics. So, it’s independent. Years of testing and refinement mean its likely to be a valid and reliable measure of pay, though the main challenge is a lack of coverage of people working in the sector. This problem is exacerbated if we want to look at senior staff. What’s potentially useful about this is that its a survey of households, not organisations. Our existing analysis of the LFS is here.

Am I missing anything?

The point is that there isn’t a perfect source of data –  so, we’re just going to have to make sure that we’re aware of the problems and biases of the different sources. Over the next month or so we’ll be taking a look at these sources and trying to pull together a review of the evidence.

Questions and Answers

What with timing and all that, we could of course do with answers now. I’ll try and use this post to add in the answers to my questions.

A reasonable starter is this benchmark data from xPert HR. Here’s our summary of the state of pay in the voluntary sector. It suggests median pay for a CEO was £90k in 2011/12 – based on a sample of 159 large charities (with an average of 388 employees).

This paper from Alasdair Rutherford – Voluntary Sector Wages in the UK 1998 – 2007 – uses LFS data to compare salaries between the public, private and voluntary sectors. It concludes that, over a period of time, voluntary sector salaries as a whole have risen more quickly than in the public or private sectors, but that salaries as a whole are still lower in the voluntary sector. There are some very interesting gender differences.