The UK plans to Spend! Spend! Spend!

There is something of an irony today as the UK faces a Spending Review which is not called a Budget but is set to be more significant than nearly all of the latter. Also we are reminded that previous to this phase we were in uncertain times but that has been squared or even cubed this year. Perhaps the biggest example of that affecting the public finances came with Lockdown 2.0 as the government announced this on Bonfire Night.

Today, we are extending the CJRS until the end of March for all parts of the UK. We will review the policy
in January to decide whether economic circumstances are improving enough to ask employers to
contribute more. The Job Support Scheme is postponed.
Eligible employees will receive 80% of their usual salary for hours not worked, up to a maximum of £2,500
per month.

Interestingly they switched to telling us the cost when the scheme for the self-employed was announced at the same time.

This is £7.3 billion of support to the self-employed through November to January alone, with a further
grant to follow covering February to April. This comes on top of £13.7 billion of support for self-employed
people so far, one of the most comprehensive and generous support packages for the self-employed
anywhere in the world.

The Resolution Foundation has calculated the costs this year as this.

The largest AME components of these increases are the estimated £56 billion spent on the Job Retention Scheme (JRS) and £23 billion on the Self-Employed Income Support

Having checked the numbers on Friday which covered the period until October some £61 billion or so has already been spent to the danger in those numbers looks set to be from the upside. In terms of a total they think this.

We estimate that in the region of £250 billion of additional Covid-related spending will take place in 2020-21. This, and the much smaller economy, combine to mean that the
size of the state relative to GDP is set to sky-rocket this year, from 40 per cent of GDP to around 60 per cent of GDP.

So there is an element of today being a bit after the Lord Mayor’s Party so let me lighten the atmosphere with some examples of the first rule of OBR Club.

GDP growth in the third quarter of 2020: the level of GDP was 7 per cent higher than the OBR had expected in July

That is a pretty spectacular fail and there is another.

Since that forecast, unemployment has risen
only slightly, as shown in Figure 4: unemployment in 2020 Q3 was 4.8 per cent, less than half that expected in the OBR’s central scenario.

There are two issues here which in my opinion the Resolution Foundation miss. They treat OBR forecasts seriously and hang their view on the future off them when as you can see the future is very unlikely to be as forecast. Also the unemployment definition has failed us and we should be looking at underemployment measures such as hours worked to get a much better view of the state of play.

What about today?

The Financial Times gives us an example of government by leak.

Rishi Sunak will on Wednesday set out a £4.3bn plan to tackle the threat of mass unemployment as the chancellor braces Britain for the brutal economic fallout from the coronavirus crisis. Mr Sunak will tell MPs in his spending review that his “number one priority is to protect jobs and livelihoods”.

What does this mean in practice?

Mr Sunak will announce £2.9bn of spending over three years on a “Restart” programme to help Britons find jobs, plus £1.4bn of new funding to increase the capacity of the Jobcentre Plus network to help more people back to work. The Restart scheme, offering regular and intensive “tailored” job support, is particularly aimed at older workers who are most likely to be left facing “the scarring effects” of long-term unemployment.

Let us hope that this works although it relies on there being jobs to go to. The Jobcentre Plus scheme has seen famine after 2015 but now is back to feast so I wonder how effectively it can be expanded? Sadly the FT continues the media obsession with the fairly useless unemployment numbers.

The latest official statistics show that an estimated 1.6m people were unemployed in the three months to September — 318,000 more than a year earlier. The unemployment rate stands at 4.8 per cent of the workforce.  With many companies pressing ahead with redundancy plans, unemployment is set to rise further in the coming months.

The BBC takes a wider view including other measures some of which have already been announced.

These include an extra £3bn for the NHS in England to help tackle the backlog of operations delayed due to Covid, an increase in defence spending and a £4.6bn package to help the unemployed back to work.

So whoever leaked this to the BBC has added some £300 million to the unemployment plan compared to the leak to the FT. Also there is something of a difference into the issue of future austerity. The FT suggests it is a can to be kicked into the future wheres the BBC gives examples of it already beginning.

The government is expected to announce a cut in the UK’s overseas aid budget to 0.5% of national income, down from the legally binding target of 0.7%……There have also been reports that the chancellor is considering a pay freeze for all public sector workers except frontline NHS staff.

There are even reports that this will extend to Members of Parliament.


The main issue here I think is what is the role of government? I am not particularly thinking of the size of it here. What I mean is what can it do about employment and unemployment? It can make a major difference if it can pock out which are the viable jobs that need support for say a year and can then thrive. We win out of that via future tax payments before we get to other issues. The problem is that the credit crunch was far from the best example of this as we ended up protecting the banks with a The Precious! The Precious perspective only for them to then retrench anyway and have a zombie business model. Along the way inflating the housing market was a consequence too, although that has become an international game.

The S&P CoreLogic Case-Shiller U.S. National Home Price NSA Index, covering all nine U.S. census divisions, reported a 7.0% annual gain in September, up from 5.8% in the previous month.

As to whether we can afford it then as I pointed out as recently as Monday we can borrow very cheaply. We are paid to borrow at the shorter end and even the fifty-year yield is a mere 0.73%. So it has completely ignored the expected spending increases. That requires a so far,as back in the Gordon Brown days it used to wait until late afternoon on the day. Our reputation may be damaged by the announcement on the RPI that I reported on last week.

Massive day for the UK index-linked gilt market. Today we get the government’s response to the RPI Reform Consultation: likely that RPI will be aligned to CPIH from 2030, with no compensation for investors. Some even think this might be moved forward to 2025. ( @bondvigilantes )

If I was in charge I would scrap that plan and I would look to strengthen our position by issuing some one hundred year bonds. As Steve Winwood so aptly put it.

While you see a chance
Take it


UK hours worked have fallen 12% since the Covid-19 pandemic began

This morning has brought the focus back on the UK and the labour market release has brought some better news. Sadly the unemployment numbers are meaningless right now so we need to switch to the hours worked data for any realistic view.

Between April to June 2020 and July to September 2020, total actual weekly hours worked in the UK saw a record increase of 83.1 million, or 9.9%, to 925.0 million hours.

Average actual weekly hours worked saw a record increase of 2.7 hours on the quarter to 28.5 hours.

This is our first real look at a fullish set of data for the third quarter as we do not get the Gross Domestic Product or GDP numbers until Thursday. Will they also show a bounce of around 10%? Our official statisticians seem to have lost a bit of faith in their own figures as they quote the Markit PMI as back up.

The IHS Markit states that the recovery in business activity, which continued across the manufacturing and service sectors in September 2020, reflects the record increase in total hours worked on the quarter to September.

Perhaps they are unaware of the reduction in credibility for that series. However we can sweep this section up by noting that whilst we have much better news we are in a situation described by Foreigner.

But I’m a long, long way from home

That is because the numbers are still 12% below the pre pandemic peak of 1,052.2 million hours.


We had feared a rise in these, and sadly they have been coming.

Redundancies increased in July to September 2020 by 195,000 on the year, and a record 181,000 on the quarter, to a record high of 314,000 (Figure 3). The annual increase was the largest since February to April 2009.

In terms of what they tell us? We have an issue because we were seeing rises ahead of the further wind down and then end of the Furlough scheme which then saw a U-Turn extension to March. So much for another form of Forward Guidance. So the real message here is somewhat confused.

Using the tax system

This is a new innovation designed to give more timely data and to that extent it helps as we get a signal for October.

Early estimates for October 2020 indicate that the number of payrolled employees fell by 2.6% compared with October 2019, which is a fall of 763,000 employees……..In October 2020, 33,000 fewer people were in payrolled employment when compared with September 2020 and 782,000 fewer people were in payrolled employment when compared with March 2020.

These numbers have proved useful for a direction of travel but again due to the furlough scheme are much too low in scale. Also the wages numbers are best filed in the recycling bin.

Early estimates for October 2020 indicate that median monthly pay increased by 4.6%, compared with the same period of the previous year.

What they are most likely telling us in that job losses have been concentrated in the lower paid which has skewed the series.


Sadly the BBC seems not to be aware that these numbers are way of the mark and so are actively misleading.

The UK’s unemployment rate rose to 4.8% in the three months to September, up from 4.5%, as coronavirus continued to hit the jobs market.

The reason for that is the furlough scheme.

Experimental estimates based on returns for individual weeks show that the number of people temporarily away from work rose to around 7.9 million people in April 2020 but has fallen to around 3.9 million people in September 2020. There were also around 210,000 people away from work because of the pandemic and receiving no pay in September 2020; this has fallen from around 658,000 in April 2020.

Following international guidelines has led us up the garden path.

Under this definition, employment includes both those who are in work during the reference period and those who are temporarily away from a job.


We can now switch to the price of labour where according to out official statisticians there has also been some better news.

Annual growth in employee pay continued to strengthen as more employees returned to work from furlough, but pay growth was still subdued as some workers remained furloughed and employers were paying less in bonuses…..Growth in average total pay (including bonuses) among employees for the three months July to September 2020 increased to 1.3%, and growth in regular pay (excluding bonuses) increased to 1.9%.

As you can see below there were hard times still for some sectors.

During the early summer months, the industry sectors accommodation and food services and construction had seen the largest falls in pay, down more than 10% in April to June; in July to September, both recovered some loss although their average total pay growth remained down, at negative 1.8% and negative 3.9% respectively.

Actually the construction numbers seem curious as in my part of London it all seems to have got going again, but as ever London may not be a good guide.

We can see who is doing relatively well by switching to the most recent single month numbers which are for September. Here we see public-sector total pay was up 4.4% on a year ago. Also that the services sector has risen to 3,5%. Switching to manufacturing we see that annual growth has finally become positive but is at a mere 0.6%.

The improvement has followed through into the real wages data at least according to the Office for National Statistics.

In real terms, total pay in July to September grew at a faster rate than inflation, at positive 0.5%, and regular pay growth in real terms was also positive, at 1.2%.

In terms of actual pay those numbers mean this.

For September 2020, average total pay, before tax and other deductions, for employees in Great Britain was estimated at £553 per week in nominal terms. When expressed in real terms (constant 2015 prices), the figure in September 2020 was £509 per week, notably higher than the £488 per week estimated in June 2020.

It may be notably higher than June but is still below the pre credit crunch peak of £522 for the constant price series from February 2008. Actually that number looks a bit of a freak or more formally an outlier but even if we discount it we are still below some of the others from around then.


We find ourselves again mulling the way that conventional economic metrics have failed us. To be specific we see that underemployment measures are much more useful that unemployment ones as a 12% fall in hours worked gives a much more realistic picture than a 4.8% unemployment rate. In the short-term the improvement in the situation will clash with the November lock down and thus get worse. Although with the Hopium provided by the positive vaccine news from Pfizer there are now more realistic hopes for a better 2021.

Switching to the wages numbers I think there is a compositional effect making them also unreliable or rather more unreliable than usual. We even have an official denial to confirm this.

 that is, if the profile (percentage within each industry) of employee jobs had not changed between July to September 2019 and July to September 2020, the estimates of growth in total pay and regular pay would have been 0.1% lower than reported in this bulletin.

In my opinion the numbers are not accurate enough to claim that. So we know more but much less than some try to claim.

By the way those pushing the 4.8%  unemployment rate ( and thereby believing it) surely they should be pushing for the Bank of England to raise interest-rates as it is well below the levels it was supposed to?


UK Wages are falling again as we go back in time to 2006

The pace of UK economic data releases is relentless at this time of the month as we have several “theme” days. Officially they are to highlight areas but in fact the role is to hope that any bad data is quickly replaced by good and also to swamp us with too much information. For example UK trade is worth a day on its own but rather conveniently tends to get ignored on GDP day. This morning brings the labour market which is in crisis and I shall first look at the numbers which are providing some insight and then move onto the ones which are failing us.


We have been both fearing and expecting  a drop here and sadly that has arrived.

Growth in average total pay (including bonuses) among employees slowed sharply in March to May to be negative (at negative 0.3%) for the first time since April to June 2014; regular pay growth (excluding bonuses) slowed to 0.7%.

As you can see total pay has been dragged into negative territory by quite a plunge in bonuses, which is hardly a surprise in the circumstances. This means that those who concentrate on regular pay are missing the bus. Whereas we note that bonuses have gone -2.3%, -15.4% and then -23.5% in the latest 3 months. Weekly bonuses started the year at £34 in January but were only £25 in May.

This means that the wages growth we were happy to see this time last year has gone like this.

The rate of growth has been slowing since April to June 2019, when it stood at 4.0% for total pay and 3.9% for regular pay, the highest nominal pay growth rates since 2008. It had slowed to 2.9% in December 2019 to February 2020 immediately prior to the coronavirus (COVID-19) pandemic.

It was slowing anyway but now someone has stamped on the brakes.

We do get a breakdown for the last year as we see the public sector did much better than construction which is a shift as I recall it being the other way not so long ago.

Between March to May 2019 and March to May 2020, average pay growth varied by industry sector (Figure 3). The public sector saw the highest estimated growth, at 3.8% for regular pay, while negative growth was seen in the construction sector, estimated at negative 5.4%, the wholesaling, retailing, hotels and restaurants sector, estimated at negative 2.1%, and the manufacturing sector, estimated at negative 1.6%.

Have you noticed how the official release concentrates on the better regular pay series in the same way we are presented CPIH inflation? Let me help out by pointing out that in May the public-sector did even better for total pay growing by 4.8% on a year before. Whilst weekly bonuses have fallen there they are small ( £3 to £2). Construction total wages have fallen by 9.8% on May last year driven by a fall in bonuses from £30 to £19. Quite a shift to say the least.

Did Furlough Impact This?

Yes as you can see below 60% of those on furlough were only on it so 80% of previous wages.

The Office for National Statistics (ONS) has published estimates of approximately 30% of employees being furloughed in the last two weeks of May, and a little over 40% of furloughed employees having their pay topped-up above the 80% pay received under CJRS ( Coronavirus Job Retention Scheme)

This pulled pay lower.

The combined impact of this is a downward drag of a little over 3%.

So we are now at 169 on the total wages index compared to the recent peak of 174.2 in January

Real Wages

Here are the official numbers.

In real terms, total pay growth for March to May was negative 1.3% (that is, nominal total pay grew slower than inflation); regular pay growth was negative 0.2%, the difference being driven by subdued bonuses in recent months.

They have a favourable inflation number ( CPIH) but the impact of that is lower right now. There is of course the caveat that the inflation numbers are missing quite a bit of data due to the pandemic.

The perspective is this and the last sentence does some heavy lifting here.

For May 2020, average regular pay, before tax and other deductions, for employees in Great Britain was estimated at £504 per week in nominal terms. The figure in real terms (constant 2015 prices) fell to £466 per week in May, after reaching £473 per week in December 2019, with pay in real terms back at the same level as it was in March 2019.

Pay in real terms is still below its level before the 2008 economic downturn.

As it slips their mind to do this let me help out using total pay and indexing to 2015 Pounds. The previous peak of February 2008 of £522 per week seems a statistical aberration so you can either use it or the £507 of May 2008 to compare to the £490 of this May, and yes this is flattered by the woeful inflation number used. A lost decade of twelve years and counting…..

Thirty years of hurt
Never stopped me dreaming ( Three Lions)

However not everyone is losing and thank you to Lynn Lewis and Ben McLannahan for this.

pay at @GoldmanSachs  up by almost a fifth in H1

Time to remind ourselves of this one more time.

The first thing you need to know about Goldman Sachs is that it’s everywhere. The world’s most powerful investment bank is a great vampire squid wrapped around the face of humanity, relentlessly jamming its blood funnel into anything that smells like money. ( Matt Taibbi)


So having sorted out the price of work how much was actually taking place? The best guide comes below.

Between March to May 2019 and March to May 2020, total actual weekly hours worked in the UK decreased by 175.3 million, or 16.7%, to 877.1 million hours (Figure 4). This was the largest annual decrease since estimates began in 1971, with total hours dropping to its lowest level since May to July 1997……..Average actual weekly hours fell by a record 5.5 hours on the year to a record low of 26.6 hours.

Indeed even this is an understatement it would seem

Experimental work with adjusted methodology suggests the use of the existing methodology has understated the reduction in the actual numbers of hours worked by approximately 5% to 6%

So the real fall looks to be of the order of 22%.

Another perspective is provided by the analysis of the Pay As You Earn ( how many are paying tax) figures.

In June 2020, 74,000 fewer people were in paid employment when compared with May 2020 and 649,000 fewer people were in paid employment when compared with March 2020.


We see that the wages situation is grim with both nominal and real wages falling again. That means that the journey to the previous peak looks ever longer. A more positive view is that there is a small flicker in the May figures so there may be signs of a recovery from the lows. On the other side is the furlough scheme which in a broad sweep is responsible for the wages drop in return for keeping people employed. When it ends though we will see unemployment rise and whilst some will return on normal wages we have already seen wage cuts applied. I expect more of them.

“Following intensive negotiations between Balpa and Ryanair a package of cost savings was put together,” Balpa said. “Pilots have agreed to accept a 20% pay reduction in order to save 260 of the jobs that were at risk, ( The Guardian)

Shifting back to conventional measures they are failing us as you can see.

The UK employment rate was estimated at 76.4%, 0.3 percentage points higher than a year earlier but 0.2 percentage points down on the previous quarter.

Really? Still at least we avoided a form of La Dolce Vita where unemployment supposedly fall, but even so this is hopeless.

The UK unemployment rate was estimated at 3.9%, 0.1 percentage points higher than a year earlier but largely unchanged compared with the previous quarter.

The Investing Channel







The Euro area unemployment rate is much higher than the 7.4% reported today

A clear feature of the economic landscape post the Covid-19 pandemic is mass unemployment. We should firstly note that this is and will continue to create quite a bit of suffering and angst. Also that all the easing policies of the central banks over the past decade or so were supposed to avoid this sort of thing. But if the system was a rubber band it had been stretched towards breaking point and now all they can do is pump it all up even more. But for our purposes there is another issue which is that we have little idea of either how much unemployment there is or how long it will last. Let me illustrate by looking at the numbers just release by Italy.

The Italian Job

As you might expect employment fell in May.

On a monthly basis, the decline of employment (-0.4%, -84 thousand) concerned more women ( 0.7%, 65 thousand) than men (-0.1%, -19 thousand), and brought the employment rate to 57.6% (-0.2 p.p.)…….With respect to the previous quarter, in the period March – May 2020, employment considerably decreased (-1.6%, 381 thousand) for both genders.


Also unemployment rose.

In the last month, also unemployed people grew (+18.9%, +307 thousand) more among women (+31.3%, +227 thousand) than men (+8.8%, +80 thousand). The unemployment rate rose to 7.8% (+1.2 percentage points) and the youth rate increased to 23.5% (+2.0 p.p.).

Now the problems begin. Firstly I recall that last time around we were told the unemployment rate was 6.3% which has seen a substantial revision to 6.6%. There my sympathy is with the statisticians at a difficult time. But for the next bit we have to suspend credulity.

In the last three months, also the number of unemployed persons decreased (-22.3%, -533 thousand), while a growth among inactive people aged 15-64 years was registered (+6.6%, +880 thousand).

If we look further back we just compound the issue.

On a yearly basis, the decrease of employed people was accompanied by a fall of unemployed persons (-25.7%, -669 thousand) and a growth of inactive people aged 15-64 (+8.7%, +1 million 140 thousand).

As I pointed out last month the issue is how unemployment is defined.

Unemployed persons: comprise persons aged 15-74 who:
were actively seeking work, i.e. had carried out activities in the four week period ending with the reference week
to seek paid employment or self-employment and were available to start working before the end of the two
weeks following the reference week;

The definition fails when you have a lockdown as some cannot go to work and others quite reasonably think that there is no point. If we assume that the rise in activity is all a type of hidden unemployment then we get an unemployment rate of 12.4% in Italy. Our estimate will be far from perfect so let us say we think it has risen from ~11% last in April to more like 12% in May.

An even grimmer situation is shown by youth unemployment. The official reading is bad enough.

the youth rate increased to 23.5% (+2.0 p.p.).

But if we apply the same methodology we get to a rather chilling 46.3%. The inactivity category here is huge at 4.6 million which I hope is pretty much students. I have to confess that I am reminded of the Yes Prime Minister quote from the 1980s that education was mostly extended to reduce the unemployment numbers. Anyway it is a blunt number but frankly will be much nearer than the official one. Also there will be many young Italians who have had little hope of a job post credit crunch as it was and it just got worse.

What we do learn is how few people are surveyed for these numbers.

The number of interviewed households for May 2020 is about 17,000 (almost equal to 35,500 individuals) and is
approximately 10% lower than the average number of interviews used for the production of estimates related to a
four-weeks month.


If we switch to the Ministry of Labour we get a barrage of numbers.

Unemployment is reduced in all sectors except agriculture and among claimants “without previous employment”
There are fewer unemployed registered in ten autonomous communities
In June 308,985 more contracts were signed than in the previous month
Almost six million people received SEPE benefits in May.

These numbers look both more useful and realistic. Things started to get better last month with around 309,000 new jobs but the Furlough scheme count in May of 6 million gives a perspective. Also unemployment edged higher.

The registered unemployment in the offices of the State Public Employment Service (SEPE) has increased by 5,107 people compared to the previous month. This represents an increase of 0.1%, which deepens the trend of slowing down the growth rate of unemployment that began in May.

So we end up with this.

The total number of unemployed persons registered in the SEPE offices amount to 3,862,883.

There is an irony in using registered unemployment numbers as they fell into disrepute due to the way they can be manipulated and fiddled. But right now they are doing better than the official series. El Pais summarises it like this.

The total number of jobseekers in Spain has risen to 3.86 million, the highest figure registered since May 2016……The rise in unemployment for June is the first increase seen since 2008, just months before the fall of Lehman Brothers and the year of the financial crisis. The increase in contributors to the Social Security system for the month is also the smallest since 2015.

So we see that there are also still around 2.1 million people on the furlough scheme. In total these benefits were paid out.

In May, the SEPE paid 5,526 million euros in benefits, of which 3,318 million were dedicated to paying ERTE benefits and 2,208 million to unemployment benefits, both at the contributory and assistance level.

If we use these numbers are plug them into the official unemployment series we end up with an unemployment rate of 16.8%.

Euro Area

This morning’s official release tells us this.

In May 2020, a third month marked by COVID-19 containment measures in most Member States, the euro area seasonally-adjusted unemployment rate was 7.4%, up from 7.3% in April 2020……..Eurostat estimates that 14.366 million men and women in the EU, of whom 12.146 million in the euro area, were unemployed in May 2020. Compared with April 2020, the number of persons unemployed increased by 253 000 in the EU and by 159 000 in the euro area.

Unfortunately we do not have an update on inactivity so we can have a go at getting a better picture. We are promised more but not until next week.

To capture in full the unprecedented labour market situation triggered by the COVID-19 outbreak, the data on
unemployment will be complemented by additional indicators, e.g. on employment, underemployment and potential
additional labour force participants, when the LFS quarterly data for 2020 are published.


As you have seen earlier this is a “Houston we have a problem moment” for unemployment data as it rigorously calculates the numbers on the wrong football pitch. It creates problems highlighted by this tweet from Silvia Amaro of CNBC.

#unemployment in the euro zone came in at 7.4% in May. At the height of the debt crisis it reached 12.1%. #COVIDー19

That creates the impression things are much better now when in fact they may well be worse. Without the furlough schemes they certainly would be. What we fo not know is how long it will last?