Wednesday, 15 March 2017

Richard Murphy, GERS Denier

Twitter has been been rife with “GERS denial” over the last 24 hours, triggered by a bizarrely ill-informed blog written by "the man behind Corbynomics", Richard Murphy.

It's clear if you read his blog that he hadn't read the GERS methodology statement before he wrote it and - being brutally honest - he appears not to understand the basic principles of economic and statistical analysis. But as this blog will show: you don't have to take my word for that.

His core argument is that because estimates are used in GERS that means there's no data involved. I know, right? But seriously - he took to twitter to make his position on this extremely clear (follow the links if you don't believe me)
What is extraordinary here is that Richard appears not to understand that almost all economic analysis relies on estimates and that the estimates in question here are very much based on actual data. Suggesting the term "estimate" is somehow synonymous with "no data" shows that - frankly - he's either an idiot or that he assumes the people following him are.

Nearly all analysis of national statistics relies on estimates. That's why there are codes of practice that civil servants have to follow, why we have independent bodies to assess and qualify reports and why statistical calculations are used to determine what confidence intervals are appropriate to any findings.

Richard didn't stop there though - he went on to malign the integrity of those involved in producing these figures
We're now getting into conspiracy theory territory.

But who am I to question the opinions of the Professor of Practice in International Political Economy at City University? I am, after all, just a businessman and blogger - albeit a reasonably well informed one who has a hard-earned reputation for knowing what I'm talking about on this subject. So predicting the usual ad hom responses1, I decided to ask for help from a couple of the experts I'm fortunate enough to know.

I asked Professor Ronald MacDonald for his thoughts on the subject. Professor MacDonald is Research Professor in Macroeconomics and International Finance at the Adam Smith Business School; he has acted as an advisor on currency and exchange rate issues to the European Commission, IMF, World Bank, European Central Bank and a number of other central banks. He was previously Bonar Macfie Chair of Economics and Adam Smith Chair of Political Economy at the University of Glasgow and Professor of International Finance at the University of Strathclyde. I think we can fairly say that Professor MacDonald knows his stuff .

He offered the following comment:
"It is important to note that that GERS is a national Statistics publication and assessed by the independent UK Statistics Authority. The statistics are produced by civil servants, and not by a partisan group, and are best practice in the sense that they meet the Code of Practice for Official Statistics, a code that is consistent with the European Statistics Code of Practice.
As in practically any statistical exercise the GERS statistics depend on estimates and there is nothing unusual about that. In that regard it is noteworthy that the statistics produced and reported in GERS come with standard confidence intervals indicating the uncertainty with which the central estimates are held. An examination of these confidence bounds demonstrates that the generally accepted position on Scotland’s fiscal and trade positions are unchanged. This is why mainstream economists, statisticians and commentators will continue to use these statistics in their work."
Professor Ronald MacDonald
That really should be an end of it, the very foundation of Richard Murphy's "they're just estimates" case is shown to hopelessly naive and fundamentally flawed.

If you're struggling with how come estimates are acceptable: your watch estimates the time, your speedometer estimates your car's speed, your scales estimate your weight. You still know with reasonable confidence the time, how fast you're driving or how much you weigh.

In case you still need convincing, I also asked Professor Angus Armstrong for his response.  Angus is Director of Macroeconomics at the National Institute of Economic and Social Research (NIESR) and was previously Head of Macroeconomic Analysis at HM Treasury. It would be fair to say he too knows this subject better than most.

He offered me the following reply
All economic statistics involve sampling and estimates. But when the UK Statistics Authority designate figures as ‘National Statistics’ that’s hugely significant. This is a kite-mark showing they meet international statistical standards. Anybody who says these figures are “easily rigged” or “nonsense data” frankly doesn’t deserve to be taken seriously. The people who work to create these statistics are honest, hard-working and dedicated public servants who aren’t allowed to answer back to defend themselves. Anyone who questions our national statisticians’ honesty and integrity should take a hard look at themselves.
Professor Angus Armstrong


I shouldn't really need to go on, should I?

Just in case some of you think by picking on Richard's tweets I'm ducking what he actually wrote in his blog, I'll quickly cover the detailed points. His is a blessedly short blog, so here's a very quick pass at some of the more obvious gaffes he made

"So forget Scottish GDP data: we just don’t know what it is."
This will come as a shock to those who believed the SNP during the independence referendum when (because of oil) they were able to say (of the then recent past): “Scotland is the 14th richest nation in the world” and "Scotland’s GDP per head is £2,300 higher than the UK as a whole”. According to Richard we can't possibly know that.

"The allocation of government spending to Scotland will be arbitrary: how much defence should it pay, for example? Or interest? The arbitrary areas will be too great for this number to really be reliable."
Notice how he says "will be" twice in this paragraph? That's a clue to the fact that he hadn't actually bothered to check; he was asserting what he assumed "will be" the case.
Anybody who has taken even a passing interest in the debate will know that what is the case is GERS apportions the two specific examples he quotes on a per capita basis. "International Affairs" is the other category allocated this way - if he'd read GERS he'd have probably thought to include that example too.
It's widely understood that while we're part of the UK we pay our population share of expenses like these that are "incurred on behalf of the UK as a whole". More importantly: when making the case for an independent Scotland these three figures are normally the first to be looked at precisely because we can see them clearly identified in GERS - that's why everybody (including the SNP with their White Paper and as we'll soon see with their Growth Commission) start with the GERS figures and then make assumptions about what would replace them if we were independent.
Richard says of all these figures "to base debate on them would be a serious mistake". That would only be true if you were to make the serious mistake of not understanding how the figures are compiled and what they tell us. In fact the data provides a perfect platform on which to base a debate about an independent Scotland's finances - that's why everybody from the SNP through to the IFS do precisely that.
Oh: and the three areas mentioned above? The only ones so crudely allocated and the one's always considered separately in debate? They represent less than 10% of Scotland's Total Managed Expenditure. In fact in 2015-16, fully 63% of TME was fully devolved expenditure; is Richard claiming Scottish Budget figures are arbitrary? Of the balance the biggest element is social protection, mainly pensions; maybe Richard doesn't believe that "London" is really paying these (hell I don't know - it's not easy trying to follow his thinking) 
"Let’s be blunt: no one has a clue what crosses the borders from Scotland to England and Northern Ireland. These numbers are literally made up in that case."
So here we're moving away from GERS to something else Richard clearly hasn't looked at: Export Statistics Scotland (ESS). This is data compiled by the Scottish Government, primarily using the Global Connections Survey run by the Scottish Government. Yes it's an estimate - but one triangulated with other sources and which qualifies for designation as National Statistics (which as we hopefully now understand, really means something). I suppose at some level all numbers are "made up", but to suggest the extensive work carried out by the Scottish Government in this area amounts to "no one has a clue" is as witless as it is insulting to those involved.
"Westminster could pretty much manipulate this data at will."
An astonishing statement, and one that could only be made by somebody who has never communicated with those responsible for this report, namely the Scottish Government's Chief Statistician and Chief Economic Advisor. The idea that Westminster could have been consistently pulling the wool over the eyes of the Scottish Government when it comes to our national finances is not only ridiculous, it's deeply insulting to the professional integrity of the honest, hard-working civil servants involved.

If you are one of those who believes that there is some Westminster conspiracy which has been so brilliantly performed that the Scottish Government haven't noticed, I've honestly no interest in debating with you - feel free to not read this blog, I won't miss you.

"if there is to be meaningful debate on this issue then the SNP have a lot of work to do to produce best possible data. The last thing they should do is trust that from London"
Well it's easy to see that Richard is new to this debate.  What on earth does he think the SNP were doing during the years they prepared for the independence referendum while they were in government? I imagine there are more than a few exhausted civil servants who will be thrilled to hear that someone who demonstrably hasn't read any of their rigorous, detailed and professional work thinks they have "a lot of work to do". In fact I'm sure a few SNP politicians will find it amusing that Richard thinks they blindly "trust that from London"
I could go on, I really could but - like me - life is too short.

I'll leave with this simple table that clearly Richard hadn't seen before writing his blog. It appears on page 47 of the GERS report itself, so you don't even need to go to the separate Methodology statements to find it. This table explicitly deals with the question of statistical significance and confidence with respect to each of the survey based apportionments

So to recognise the fact that there are of course estimates used in compiling these figures, instead of saying "Scotland's GERS deficit is £14.9bn" we could say "we can be 95% sure that Scotland's GERS deficit is in the range of £14.3 - 15.5bn"2. It would be tedious if we said that every time though, I'm sure you'll agree.

To finish with another example of how confidence intervals work: when Richard Murphy repeatedly claims "there is no data", we can say be 100% sure that he doesn't have a clue what he's talking about.


1. apparently the fact that my little £20m turnover online business (employing people in Scotland, of course) sells - among other things - dog food means I can't have any credibility in this debate. Who knew?

2. as many of us have pointed out, there is one other significant allocation uncertainty on the revenue side which is not survey based but assumption based and that's corporation tax. After a few years of disagreement, on this figure HMRC and GERS make very similar assumptions, basically assigning £2.9bn or 7.3% of corporation taxes to Scotland. As a sanity check, Scotland account for 8.3% of the UK's population. The simple truth is that given companies are not currently required to report profit split between Scotland and the rest of the UK, nobody knows what these figures would be were Scotland to be independent - changes to tax rates and corporates' decisions around where to base activities and how to report profits make this figure a moving feast anyway.


Iaing said...

So, how much will the Pedigree Chum be after independence?

Libertarian said...

I wish Mr Murphy's carer would put him to bed as quickly as you did his ridiculous conspiracy theory.
I suppose we can excuse Mr Murphy just a wee bit, after all Trumpian conspiracy theories and ludicrous nationalist outbursts are all the rage in the U.S.
Nice work chokka.

Anonymous said...

I hereby nominate Kevin and Ronald MacDonald to head the official opposition to the next installment of the once-in-a-generation referendum.

flash said...

Iaing, how can anyone give an answer when we don't know what currency it would be in.

Anonymous said...

Did you get your degree in Economics at Trump University by any chance ?

Peter said...

Just a few thoughts.

Well argued, and deep flaws convincingly exposed, Kevin.

You indicate that R Murphy has wandered into the field of GERS uninformed - or at least severely underinformed - and started pontificating as if an authority.

I think this is the core of the problem you address here.

Wasn't it Pope (not the self-appointed, self-enthroned one in Bath) who wrote:

"A little knowledge is a dangerous thing"?

Being a Voltairean sort, a beacon of caustic enlightenment, the poet was by no means suggesting ignorance is bliss. But that those who acquire a thin patina of learning in any new field are (all too often) tempted to sound off with a false new authority - and expose a greater ignorance, at least to anyone who has got to know that field better,at closer quarters, by dint of harder work.

Murphy is no fool.

But I believe he is two things: a) an accountant and tax expert, b) committed to a left-progressive line especially on social issues. He does not generally give the impression of a grasp of large, more complex and multi-determined economic details.

As Pope might have continued, too little knowledge is no excuse.

You do not mention this in your preamble, but I think it is worth mentioning here, for at least two reasons: originally just about Corbyn's most prominent economics expert/champion has become thoroughly disillusioned by Corbyn as a "leader", and I think, also by what is oddly known as "Corbynism", as a cause. If not also the latter, so much the worse.

On GERS, I fear it may be once again a case of "commitment determines all". That he has not studied the details sufficiently, and worse: has committed himself to setting himself up as some sort of authority on something he does not understand.

Tim Worstall said...

It's worse than this I'm afraid. Our professor in practice of international political economy is failing to realise that GDP can be calculated in three ways. All production, all income, all consumption.

He's right, for example, that corporation tax isn't separately recorded, which makes income a tad difficult to determine on a sub-UK basis (profits are a part of income in GDP accounting).

But we can do much better with the production numbers, and profits are going to be a minority of the economy under the income approach too. And we must note that all three measures should add up to the same number, production, income, consumption. We have an internal check on this.

Which is why ONS is entirely happy calculating GVA (which is really GDP but applied to a sub-national unit) down to the level of local authorities and similar statistical areas. Numbers I've seen Ritchie use to talk about how oppressed the North is by London, I'm sure.

And there's yet another test we can do. If you add up the GVA numbers from ONS you do actually get to the same as reported for GDP. The local and national numbers are self-consistent. They're consistent across the two measures they use, production and income, and local summing to national (that second is not true of the Chinese figures for example). Similarly the US provides GVA numbers for states and I think for counties, or at least one statistical area smaller than the state - maybe metropolitan area.

If all that's true then it's not too difficult to get from GVA to GERS - just add up the GVA for all smaller areas within Scotland.

No, it'll never be perfect but then of course what economic number is?

And then it really gets worse. Because Murphy is making his usual theoretical mistake. He's a huge believer in how government should be planning everything. And to do that of course govt needs to know what is happening now. Hayek told us this wasn't really possible but Spud insists that a Courageous State does have the knowledge and should use it. But if our econ stats are just guesses then of course we don't have the knowledge to do Corbynomics, do we?

It's not possible to both know shit about the economy and direct it. But that's the dual insistence the Murphmonster ends up with.

Bit sad that he ended up as a professor but then it is Islington Technical College after all.....

Anonymous said...

I broadly agree with you but need to correct you on one point. Murphy is no tax expert, that's for sure. Ask anyone who actually works in tax.

Anonymous said...

It's remarkable that someone who *should* understand numbers apparently cannot understand their application in the area of forecasting and probability. Murphy argues that GERS figures are either wildly wrong, because some source data is estimated and so not 100% accurate, or are being manipulated by (presumably No-motivated) political forces to give a particular (and again false) economic picture. Either way they cannot be trusted, and so can be entirely discounted. His ludicrous case would be marginally more credible if he provided just one example of a wild data inaccuracy leading to an erroneous conclusion, or of a flaw in the GERS methodology, or of a proven case of Westminster meddling in figures produced FOR the Scottish Government BY the Scottish Civil Service IN Scotland.

It's equally remarkable that such conspiracy theorists and alt-truthers all assume that if the facts and figures can't be trusted (the point of the exercise) the "real" facts must by definition be better for their cause. Nowhere in Murphy's blog does one detect the possibility that if GERS is seriously wrong the figures might actually be a whole lot worse for an independent Scotland than they indicate. Don't trust that dodgy debt figure? Okay I won't, but does that mean that there's no debt or could it be twice as big as stated, or what? Murphy has no interest in getting to a more accurate economic picture, it seems, only in attacking the credentials of figures that don't suit him. A fact-free world is his end game, leaving the battle to be fought on ignorance and emotion only. RabCH

Anonymous said...

Tim Worstall, do you have a source for government spending as a % of Scottish GDP?

Tim Worstall said...

"Tim Worstall, do you have a source for government spending as a % of Scottish GDP?"

Not directly, no, never bothered to look it up.

Andrew Carey said...

Brilliant by chokka

Unknown said...

Interesting points. We can only hope for a civilized and informed debate; today's indignant hyperbole from FM at FMQ does not bode well. Nevertheless, please keep it up. I admire your tolerance of the zealots on twitter, but we need to get the debate away from the emotional soundbites that twitter encourages and into discussion of substance.
There are sensible arguments on both sides of this debate (eg some of what Common Weal says is interesting; but at least its civilized). The real question is whether leaving or remaining in UK is more likely to lead us to a peaceful and prosperous future for our children.

In regards to this matter I must disagree with a point in your penultimate paragraph, where you suggest that it would be better to use the short form of stats, such as "Scotland's GERS deficit is £14.9bn". I actually think that proponents of the economic case for both sides should cite their figures with a view to probability, so in the form "we can be 95% sure that Scotland's GERS deficit is in the range of £14.3 - 15.5bn". Tedious yes, but it would highlight the extreme optimism implicit in so many of the claims and hopes pushed by the seccessionists. For instance, if Salmond had said that oil would be at ~$130/barrel, he might have been within the distribution of possible outcomes, but that scenario would have had a probability of occurrence below ~5%.

Conversely, your points about trade, deficit, etc are all founded in actual data from our current known situation, thus can be proposed with a high likelihood of occurrence. If probability estimates were put against many of the brave claims for the outlook for an independent scotland, as with oil price, they may be possible, the question is: are they likely?

Anonymous said...

Ben, while I agree with your points about likelihood, I suspect that the only people who will appreciate the significance of probability are those already minded to take the numbers seriously. The rest will ignore, misunderstand, or gleefully interpret "in the range of" statements as uncertainty or inaccuracy. Sadly, Kevin's short, simple £14.9bn is more likely to carry the weight of conviction, and (happily) is almost impossible to refute. RabCH

theambler said...

If Murphy was correct, one would think that the SNP would instruct our civil servants to carry out an separate data collection exercise. After all, the fiscal situation of an independent Scotland could in theory be even worse than those presented in GERS. As it is, they do not.

Drew said...

While I can see why you think it is important to correct inaccuracies regarding economics, I think the EU referendum proved putting all your eggs in that particular basket is risky. Let's be honest, the vast majority of the population don't know much about economics, No voters included.

Personally I think there is more emotion to the No vote than Unionists let on (maybe the famous British stiff upper lip is in play?). In fact I think most of the 55% of No voters did it for emotional reasons as much as or more so than economics.

Sure, they have the numbers on their side at this particular moment in time and the economics of the Union have always been important since 1707.

But when Scotland was capable of generating large surpluses from oil revenues in the 1970s and 1980s which could have supported a Scottish currency, I'm guessing most No voters wouldn't have supported independence at the time.

And being told about the risks of independence and fearing them is an emotional response in itself. My mum is a staunch No voter and one of the reasons she gave was she didn't think we could afford it. I asked her on what evidence and she couldn't name a single figure or source to back up her argument. She's never heard of GERS. She doesn't read a daily paper so clearly it was a gut feeling based on what she had heard from the debates or in the news. But she is also very pro the Royal family.

For me, the Union has always been built on 3 things:

The Crown/Christian values (more specifically Protestantism)
The Empire
Trade Unionism

In the 18th century, most of Scotland, like the rest of Britain, was focused on keeping Catholics from re-taking the throne. In the 19th century the focus was on economic opportunity through the Empire. And finally in the 20th century, the labour movement and trade unionism helped create Labour, which created the Welfare State and the NHS.

Almost all of these things are far less important or even non existant than they were.

But I'm pretty sure a more emotional Unionist campaign, based on the British armed forces and the Queen who is still very popular in Scotland would have won a bigger margin of victory in 2014 and caused a more serious, potentially fatal blow to the SNP.

My guess would be Theresa May will lead that type of campaign if it comes to it, more so than David Cameron could have.

Ferdinand said...

I think Drew has a point, there.

However in polling after the 2014 Referendum, the economy and undertainty on the new currency were significant reasons why people votes no.

The current human wave style attacks being launched on GERS by Nats are a political attempt to marginalise their own authoritative information that sunk their own case last time round.

Matt Wardman said...

On Professor Richard Murphy it is straightforward - the man is a dipstick.

His habit is to wade into random areas without doing any homework to make sure he has any credibility on the topic. In diving into Sindy politics he entered a shark-infested custard, and has ended up looking as he normally does.

Murphy's normal approach to dealing with people pointing out inconvenient things such as facts is to ignore the evidence and rubbish the person. There is an excellent example in the comments after the article you have replied to.

Professor Murphy acknowledges Wings' clain about GERS being questionable because it was created by Ian Lang in 1992, in ignorance of the SNP overhaul in 2007-8, then goes ad hominem when the fact is posted.

>ian m says:
March 14 2017 at 10:37 pm
Just to add to the we don’t know what we don’t know theme.Well now we know even less

From The Herald Dec 18 2006

Gers was conceived as a political, not as a statistical, exercise. We know this because the original correspondence from the then Secretary of State, Ian Lang, was leaked some years ago – he wanted it to “undermine the other parties”, saying “this initiative could score against all of them”.

>Richard Murphy says:
March 14 2017 at 11:09 pm

>Marty says:
March 15 2017 at 3:16 pm
What’s ‘telling’ is you not knowing that in 2007, GERS was overhauled completely by the SNP, who invited the Cuthberts into the mix (qualified and respected economists, who funnily enough also support independence) to better reflect our finances. The work they did was not only needed, but revolutionary.

She wrote a good piece on Brexit and GERS back in 2016:

Anyone who knows anything about GERS and its history would know this. They would also know that due to the work carried out post 2007, prior reviews of GERS accuracy can be safely thrown out the window. For the record, previous years’ data were amended to reflect the new model and methodology. That obviously eliminates you, since you are ignorant of this fact.

I would genuinely like to know how you (as an economist) would feel if you published researched data for then another economist to basically flush it down the toilet with no evidence or source to support it?

>Richard Murphy says:
March 15 2017 at 3:35 pm
If and when you attempt objectivity I will try replying


Apologies for the length.

Papko said...

"Personally I think there is more emotion to the No vote than Unionists let on (maybe the famous British stiff upper lip is in play?). In fact I think most of the 55% of No voters did it for emotional reasons as much as or more so than economics. "

Impressive post @Drew
You articulate very well, what I have always thought.

Stouty said...

Part of the problem is that the vast majority of the population have real difficulty with abstract financial figures and just switch off when it comes to in-depth debate (such as this post/thread). So much easier to go with opinion of an 'expert' that aligns with your own views. It's a difficult problem to overcome.

Huge numbers such as £15bn are also largely meaningless to the average man on the street and it's difficult to grasp the reality of that number. I know it's often couched in terms of tax rises / public service cuts, but we never see a vision of what would be actually be required to cut the deficit (say by 50% in 10 years for example). How many business would have to be created, of what size, and what increase would be required in the working population to support them? The faithful might then have the reality check needed for a proper debate on the realities of independence.

David said...

Funny how Richard Murphy mentioned none of this in his response to the 2016 GERS consultation. Instead he cut and paste a Common Weal campaign asking for a breakdown of PFI debt to be included. In the response form, asked if he had any comments on the methodology of the whole thing, he just replied with a dash.

Anonymous said...

Fraser of Allander have also published a paper on GERS which doesn't back up Richard at all.