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.

Anything to say on the productivity statistics?

“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.

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.

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!

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.

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.