From the TUC

Public sector pay: some unstraight statistics from the Sunday Times

03 Jan 2010, by in Labour market, Public services

The Sunday Times has some shock horror statistics about the public sector today – and I suspect because they know they are dodgy – they have got Straight Statistics to validate them. The Sunday Times says:

The analysis was validated by Straight Statistics, a group that campaigns for the accurate reporting of official data.”

Straight Statistics say:

Straight Statistics is a pressure group whose aim is to detect and expose the distortion and misuse of statistical information, and identify those responsible.

Yet this article contains very few straight statistics – it both distorts and misuses official statistics. I hope that the group has been guilty of naivety by allowing their name to be used in an article whose final text they  did not see. If they did approve the final text, then they have some serious questions to answer.Let us look at the problems one by one.

The first error is in the opening sentence. It says:

Public sector workers earn 7% more on average than their peers in the private sector.

But the figures they use are the average for all workers in the private sectors compared to all workers in the public sector. If they were comparing peers, they would be looking at people doing the same job in the public or private sector.

Next, the figures they use are taken from the 2009 Annual Survey of Hours and Earnings produced by the ONS.  The Sunday Times uses Table 13.7A to get their figures for average pay. This is the figure for gross annual pay. It takes no account of how many hours people work as it mixes up full and part timers.  A legitimate way to compare pay between different groups of workers is to look at hourly pay. This is not hard as it is Table 13.5a on the same page (pdf).

Producing figures for peers is hard because the jobs done in the public and private sector vary such a lot. But one way of approaching this is to compare pay by educational qualification. You can read my post that does this here but to summarise briefly:

  • Public sector graduates are paid 3.4 per cent less than in the private sector
  • Public sector workers with higher ed short of a degree are paid 6.2 per cent less than in the private sector
  • Public sector workers with A levels are paid the same
  • Public sector workers with lower skills get paid more than in the private sector.

Pay in the public sector is flatter than the in the private. The low paid do better in the public sector because there are few minimum wage jobs in the public sector, and the rank exploitation of vulnerable workers that happens in parts of the private sector is missing. At the other end pay for skilled workers in the public sector is lower, and there are fewer public sector fat-cats and they are a lot slimmer than their feline counterparts in the private sector.

Studying historic changes in this way is even more prone to error.

If you look at the difference between average annual pay in either sector from one year to another, there are lots of different factors that will contribute to a change in pay in addition to any pay increase.

  • the ratio of part-time to full time workers (and the hours breakdown of the part-timers)
  • non-basic pay such as overtime and bonuses can vary from year to year.
  • the composition of the workforce (ie if the number of well-paid skilled staff increases and the number of low-paid less-skilled workers decreases then average pay will increase without anyone getting a pay rise.)

As the article concedes there has been a big increase in the proportion of graduates working in the public sector over time. This reflects both increased employment of staff such as doctors and teachers, but also the continuing trend to contract out unskilled jobs such as cleaners.

The article says:

The (pay) lead of the typical state employee stands at 7%, compared with 3% the year before. Until 2005, private sector workers received more on this measure and as recently as 2002 enjoyed a 5% lead.

But this again looks at the annual pay figure (source table 13.7a here). In my post here I looked in detail at the notion popular with the small state right that public sector pay has overtaken private sector pay. The figures show that the median hourly pay for full-time public sector workers has been higher than that for private sector workers since at least 1984. This is simply down to the higher number of professionals in the public sector.

Next:

Last year the average public sector worker laboured for 35 hours a week — a fall of an hour on the previous year and 2.5 hours less than the typical private sector worker.

This figure is taken from Table 13.9a here. This is for all employees whether part-time or full-time. It also only shows paid hours.  The change may be due to more part-time workers being employed in the public sector, or cuts in overtime in the public sector. This table simply doesn’t tell us. Many workers in both the public and private sectors do unpaid overtime. So the Sunday Times figures tell us nothing about the effort put in by private and public sector workers.

Let’s finish with a Pinteresque non-sequitur:

Most private sector workers have to work until they are 65 to claim their company pensions; the average public sector retirement age is 58.

Pension age is not the same as retirement age. Your pension age is one of the figures used to calculate your pension in a defined benefit scheme. If you retire before your pension age, you can still claim a pension but it is reduced. Retirement age is when you stop working – and is below 65 in the private sector too.

Most private sector workers are not in defined benefit pensions in any case – in fact nearly two thirds are not contributing to any employer backed pension.

Of course in some ways public sector worker have done better under this government. After all it was elected in part to put right the crumbling and neglected public services. Pay which had been held down did catch up to some extent (although there have also been pay caps in subsequent years.) There are more nurses, doctors, teachers and other, just as vital, public servants than there used to be.

During a recession it is not surprising that pay does badly in the private sector, particularly when there has been short-time working and sharp falls in paid over-time. Typical public sector pensions are better than private sector pensions. There are many rich and detailed official statistics that can no doubt be cut in various ways.

But this Sunday Times article distorts and misuses official statistics. I thought Straight Statistics was set up to stop this, not validate it.

Our comments box is open to Straight Statistics (and indeed anyone else).

15 Responses to Public sector pay: some unstraight statistics from the Sunday Times

  1. Tweets that mention Public sector pay: some unstraight statistics from the Sunday Times | ToUChstone blog: A public policy blog from the TUC — Topsy.com
    Jan 3rd 2010, 1:25 pm

    […] This post was mentioned on Twitter by ToUChstone blog, Nigel Stanley. Nigel Stanley said: RT @touchstoneblog: Public sector pay: some unstraight statistics from the Sunday Times http://bit.ly/8HedCA […]

  2. Nigel Hawkes
    Jan 3rd 2010, 3:59 pm

    We can all do special pleading, few better than Nigel Stanley. But let’s not get carried away.
    The first “error” he identifies is the claim that public sector employees earn 7 per cent more on average than their peers in the private sector. This is not an error: it is a fact. It may be justified by higher qualifications in the public sector, which the Sunday Times acknowledges. But it is not wrong.

    If you prefer, compare weekly earnings (conveniently collected by the then National Statistician Karen Dunnell in response to a question in Parliament to Kevin Brennan from Francis Maude at http://www.parliament.the-stationery-office.co.uk/pa/cm200809/cmhansrd/cm090127/text/90127w0046.htm) . This shows the gap has widened from £40 a week in 1997 to £62 a week in 2008. Updated with the 2009 figures from ASHE, the gap is now £74 a week.

    Now, there may be perfectly rational explanations for this, including the higher proportion of graduates in the public sector. I have already discussed this briefly on the Straight Statistics website (http://www.straightstatistics.org/blog/2009/10/13/whos-fattest-cat-private-or-public). The Sunday Times makes the same point.

    Nigel then goes on to argue that public sector pay has been higher than the private sector since 1984, quoting median hourly pay. As this shows the gap is now 14 per cent, not the 7 per cent cited by the Sunday Times, I am not sure what point he is trying to make.

    As for pensions, don’t get me started. As somebody recently retired from the private sector I am all too well aware of the differential. The Sunday Times calculates that the benefits of public sector pensions double the gap between private and public sector annual pay, based on methodology developed by the Institute for Fiscal Studies.

    “Of course in some ways public sector workers have done better under this Government” Nigel Stanley concedes. Thanks, at least, for that! But he is simply wrong to argue that the Sunday Times has distorted and misused official statistics. It has simply used them in ways he finds distasteful or unpalatable. He fails to point to a single error.

  3. Nigel Stanley

    Nigel Stanley
    Jan 3rd 2010, 5:36 pm

    Nigel Hawkes has graciously accepted my offer to respond to my post, and gives as good as he gets.

    If you read my post carefully you will see that I source every stat used in the Sunday Times article that I mention. My charge is not that any of the figures are inaccurate, but that they are presented in a misleading and distorted way, and should not therefore have been verified by Straight Statistics.

    (Nor by the way do I dispute that median and mean pay – whether weekly, hourly or annual – is higher in the public sector than the private sector. Indeed on what I think is the best measure for doing long-term historical comparisons it has been since 1984 at least.)

    There are too many Nigels here – so I’ve put our initials at the start of each para, so that you can follow the argument more easily.

    NH>The first “error” he identifies is the claim that public sector employees earn 7 per cent more on average than their peers in the private sector.

    NS> No, my objection was to the statement that: “Public sector workers earn 7% more on average than their peers in the private sector.” The word “peers” here has been carefully chosen to mislead. It implies that people doing the same job in the public sector get more money than they would doing the same job in the private sector.

    NS> I understand the word peer to mean “Somebody or something who/that is at an equal level” (see http://bit.ly/8e0heY for dictionary support.) Higher public pay is only true for public sector workers with lower skills. Public sector graduates earn less than their private sector peers. As there are more graduates in the public sector than the private sector, it may be that if you compare equivalent jobs – more people earn less in the public sector than their private sector peers. But I wouldn’t make that claim without a lot more analysis.

    NH> If you prefer, compare weekly earnings (conveniently collected by the then National Statistician Karen Dunnell in response to a question in Parliament to Kevin Brennan from Francis Maude at http://www.parliament.the-stationery-office.co.uk/pa/cm200809/cmhansrd/cm090127/text/90127w0046.htm) . This shows the gap has widened from £40 a week in 1997 to £62 a week in 2008. Updated with the 2009 figures from ASHE, the gap is now £74 a week.

    NS>My complaint here is that you cannot compare the wages of two groups of workers over time unless you allow for the fact that the proportion of full and part-timers may change. Using weekly pay repeats exactly the same mistake as using annual pay. That is why I prefer hourly pay.

    NS> If you did make these adjustments I suspect that the gap has still increased since 1997 as public sector pay had been held down by the last Conservative government – it may even have grown more. I also think you need to look at time series data, not compare two arbitrary snapshots, to get a proper picture. But my complaint is not that the article wrongly says that public sector pay has risen since 1997, but that the wrong statistics have been used to show this, and this should not have been verified by Straight Statistics, unless the name is changed to rough ball-park statistics.

    NH> Nigel (S) then goes on to argue that public sector pay has been higher than the private sector since 1984, quoting median hourly pay. As this shows the gap is now 14 per cent, not the 7 per cent cited by the Sunday Times, I am not sure what point he is trying to make.

    NS> My point is that Straight Statistics verified the wrong statistics, including the statement that private sector pay was higher in 2002 than in 2009. It wasn’t, not on the more meaningful measure of median hourly full time pay. I don’t dispute that median pay is higher in the public sector, but that there is nothing new about this.

    NH> he (NS) is simply wrong to argue that the Sunday Times has distorted and misused official statistics. It has simply used them in ways he finds distasteful or unpalatable. He fails to point to a single error.

    NS> Let me refer you to a Straight Statistics blog post by David Lipsey on public sector pensions. (http://www.straightstatistics.org/blog/2009/07/08/nonsense-about-public-sector-pensions). He describes a report that says “Cost of Public Sector Pensions equal to 85 per cent of GDP” as spin worthy of Joseph Stalin. This is not because any of the “facts” are wrong, ie an estimate pf the liabilities of public sector pension commitments (though I would dispute that figure) and the size of the UK’s GDP. It’s the way they are put that is wrong.

    NS> In particular this applies to the para on working time: “Last year the average public sector worker laboured for 35 hours a week — a fall of an hour on the previous year and 2.5 hours less than the typical private sector worker.” It’s clearly meant to imply that public sector workers are getting lazy – and is factually wrong as the figure are based on paid hours.

    NS> All this stat means is that there are more part-time workers in the public sector.

    NS> I used to think that Straight Statistics was against the kind of misrepresentation of facts in today’s Sunday Times. Now I am not so sure.

  4. John Gray
    Jan 3rd 2010, 10:32 pm

    Good stuff Nigel S – why didn’t the Sunday Times just run a more honest article on “why doesn’t the public sector pay poverty wages for working people just like the private sector?” or “Minimum Wage for All – Sunday Times Journalists and Executives will set an example to the Nation”.

  5. Tax Research UK » There are lies, and statistics
    Jan 4th 2010, 8:13 am

    […] Public sector pay: some unstraight statistics from the Sunday Times | ToUChstone blog: A public poli…. […]

  6. Tim Worstall
    Jan 4th 2010, 10:09 am

    “But the figures they use are the average for all workers in the private sectors compared to all workers in the public sector. If they were comparing peers, they would be looking at people doing the same job in the public or private sector.”

    Oooooh! Exactly! Just like we do when we calculate the gender pay gap.

    What was that? Sorry? We *don’t* do that when calculating the gender pay gap? You mean Harriet Harman, the EHRC, the TUC, all say that we should only compare like with like when we look at the public/private sector pay gaps but we should compare all with all when talking about the gender pay gap?

    Is this not a cake and eating situation? That we should vary our methods according to what supports a specific political position?

    No, tell us it ain’t so!

  7. Nigel Stanley

    Nigel Stanley
    Jan 4th 2010, 11:48 am

    I welcome Tim’s intervention. We have discussed this (rather civilly) at the comments to an earlier post (see http://www.touchstoneblog.org.uk/2009/12/more-about-public-versus-private-sector-pay/.

    I make clear that the TUC would always use hourly pay rates to compare gender pay, and do not see direct discrimination at the point of employment as the only source of the gender pay gap. But I have no wish to repeat these points here as they have already been explored.

    But I welcome Tim’s intervention because he is right to say that you should take care when comparing the pay of two different groups. For example I doubt Tim would allow anyone to compare pay by gender simply by looking at men’s and women’s annual take home pay, without taking account of hours worked. This, I think, is what ministers were criticised for by the stats watchdog.

    Yet this is what the Sunday Times has done – and had it approved by Straight Statistics – for public and private sector workers.

    I must assume therefore that Tim Worstall shares my critique of the statitistical methodology used here. I certainly wouldn’t use it to compare pay by gender.

  8. Tim Worstall
    Jan 4th 2010, 12:31 pm

    “For example I doubt Tim would allow anyone to compare pay by gender simply by looking at men’s and women’s annual take home pay, without taking account of hours worked.”
    Absolutely. Although that is the way the Americans come up with their “women earn 80 cents to every dollar earned by men”.

    “This, I think, is what ministers were criticised for by the stats watchdog.”

    No, they were criticised for two things. Confusion between mean and median: Stats people say use one, the whole discussion of the gender pay gap has always used the other as it makes the gap look larger.

    The second was obfuscating between part and full time pay. All part timers, male and female, get less per hour than full timers male and female. Thus the “female part time pay gap” should be a comparison of female part time pay to male part time pay: not, as it is usually quoted, female part time to male full time.

    But yes, of course, the S Times are very naught boys for not distinguishing between part time and full time positions and hours worked.

  9. Jamie
    Jan 4th 2010, 8:27 pm

    Good stuff – the means will also be skewed by the fact that the lowest wages in the public sector (cleaners/cooks etc.) have been pushed out to the private sector.

    Anything to say on the productivity statistics?

  10. Nigel Stanley

    Nigel Stanley
    Jan 6th 2010, 9:16 am

    Jamie, I wrote about this at http://www.touchstoneblog.org.uk/2009/06/public-sector-productivity-a-tricky-concept/. Chris Dillow at Stumbling and Mumbling had some interesting points about this that you might want to dig out. (too busy now to find the link!).

    I seem to remember that he said that it might well be inevitable that there are diminishing productivity returns from extra public spending for better public services. In other words it’s easy and cheap to educate the brightest or cure the only slightly sick, but when public services become more comprehensive it gets more expensive to educate the not so bright and heal the very sick.

    Of course that’s no excuse for inefficiency in public services, but productivity can never be the only criteria for judging public spending.

  11. Straight Statistics change their tune as Ben Goldacre joins the criticism of the Sunday Times | ToUChstone blog: A public policy blog from the TUC
    Jan 9th 2010, 11:47 pm

    […] of the Sunday Times Posted at 11:47 pm on 9 Jan 10 by Nigel Stanley Last Sunday I wrote a critique of a Sunday Times article about public versus private pay.  Frankly I do not expect very much from […]

  12. More misreporting on public sector pay | Left Foot Forward
    Jan 21st 2010, 11:33 am

    […] one. As has been regularly highlighted the earnings differential can be accounted for by the composition of the workforce (there are more professionals in the public sector), the fact that the lowest paid […]

  13. Alex Brummer attacks public sector pensions in the New Statesman | ToUChstone blog: A public policy blog from the TUC
    Jan 29th 2010, 6:19 pm

    […] public sector pay was recently lower than that in the private sector but has now overtaken it. See here, here and Ben Goldacre to  find out why this isn’t […]

  14. IFS have important data on public sector pay | ToUChstone blog: A public policy blog from the TUC
    Feb 9th 2010, 5:17 pm

    […] Ben Goldacre’s critique of the Sunday Times analysis that Nigel Hawkes of Straight Statistics defended on this blog. Comparing public and private sector remuneration is always a tricky business. The two sectors are […]

  15. How many public sector workers will be working late tonight? | ToUChstone blog: A public policy blog from the TUC
    Feb 26th 2010, 5:24 pm

    […] for this – just as the stories about public and private sector pay recently weren’t quite as clear cut as they may have appeared. The types of work done in the public and sector aren’t directly comparable, and in any case […]