Employment income is critical for most Americans and for the US economy. The above chart shows how the key factor – the participation rate – has changed since records began in 1948. It shows July data, to avoid the need for seasonal adjustment, split by Men, Women and Total:
□ Last month’s Total rate was 63.4%, versus the peak of 68.1% in July 1997, and was back at the 1977 rate
□ The Men’s rate was 70.3%, versus the peak of 88.7% in July 1948, equal to the 2014 – 2015 all-time lows
□ The Women’s rate was 56.9%, versus 1999′s peak of 60.3%, and was back at 1988′s level
This highlights how the workforce is now missing 12m people, due to the fall in the participation rate. 173m would have been working today if the peak rate of 68.1% had been maintained. Instead, only 161m were working. Inevitably, therefore, US GDP continues to disappoint.
Yet policymakers ignore this critical fact. Instead, they pretend that monetary policy can somehow fill the gap created by these missing 12m people. It would be nice if true, but common sense tells us this is wishful thinking.
The chart also highlights the key reasons why (a) the economy boomed during the Boomer-led SuperCycle and (b) is now seeing slower growth:
□ The key issue driving the SuperCycle was the dramatic growth in female BabyBoomer employment. Only a third of women were working in the early 1950s, but the rate then nearly doubled as a paradigm shift took place in the status of working women. As a 2005 National Bureau of Economic Research Study noted – they were no longer “perceived as secondary earners within the family”
□ Instead, women began to routinely return to the workforce after having children. This not only added to total employment, but also led to a rise in their earnings relative to men – from 62.3% in 1979 to 81.8% in Q2 today. In turn, this phenomenon of dual-income households (for the first time in history) turbocharged economic growth
□ This growth in female employment also helped to counter-balance the ongoing loss of higher-paying, mostly male, jobs in manufacturing, which were being steadily replaced by technology. More than a quarter of US employees worked in manufacturing in 1970, but this percentage had fallen to only one in ten by 2010.
□ US manufacturing employment fell from 19.6m in 1979 to 13.7m in 2007 – but the majority of this loss only took place after 2001, as companies realised the potential of IT-led automation
These are not difficult trends to reverse, if policymakers were willing to do this. They would, however, involve initially difficult conversations with voters – which is why neither Presidential candidate are likely to raise the key issues.
US GDP COULD RESUME GROWTH IF POLICYMAKERS ADDRESSED THE DEMOGRAPHIC ISSUES
The US is not like Japan, where the population is now actually falling. Instead, its population has been growing quite strongly. There were 267.8m people in July 1997 when the total participation rate peaked, compared to 324m in July.
But the demographics of the population continue to change quite significantly:
□ 15% of the polulation is currently aged 65 or over, compared to 11% in 1979, whilst life expectancy at age 65 has risen to 84.3 for men and 86.6 for women. Yet there has been no discussion about raising the retirement age, even though the Social Security trust fund will be exhausted within 13 years
□ Similar myopia dominates the racial make-up of the workforce. Bureau of Labor Statistics data shows that unemployment is far higher amongst Blacks (11.3%) and Hispanics (7.4%) than for Whites (5.3%). Even more disturbing is that Blacks earn only 80% of the average wage, and Hispanics just 77%, compared to Whites at 103%
□ This matters because Blacks are 12% of the 18+ population, and Hispanics are 16% – and, of course, these percentages are rising all the time as 92% of all US population growth since 2000 has been in Minority communities
These problems won’t go away just because policymakers choose to ignore them. They will simply get worse, and make it even harder to reach an agreed solution. Historians will not be kind when they come to discuss this failure.
WEEKLY MARKET ROUND-UP
My weekly round-up of Benchmark prices since the Great Unwinding began is below, with ICIS pricing comments:
Brent crude oil, down 59%
Naphtha Europe, down 60%. “Naphtha market still bearish”
Benzene Europe, down 53%. “Spot numbers drift lower. Weak outlook for oil/energy pricing”
PTA China, down 42%. “G20 shutdowns unconfirmed, producers remain in discussions with authorities”
HDPE US export, down 33%. “Weak demand for PE in China seemed to continue in the near term, hence the price gain might be limited”
S&P 500 stock market index, up 12%
US$ Index, up 18%
There’s something very wrong with the US jobs market, as the slide above confirms. Commentators professed to be surprised by the disappointing May report last Friday, but its hard to know why:
- The overall participation rate has been in decline since July 1997, when it reached 68%: today it is only 62.7%
- Male participation is at an all-time low of 69.1%; female participation is back at 56.8%, the 1988 level
- The total civilian population was 253m, so 172m would be working today if the participation rate was still 68%
- Instead, only 159m were working – 13m fewer people, over 5% of the civilian population
Equally worrying is that there is no sign of any improvement. If we look at data for the month of May, it shows that overall and female participation has been in steady decline since 2008, when they were 66% and 59% respectively. Every single year since then has seen a decline. Male participation has seen almost exactly the same trend, with the exception of last year – when it rose by 0.3% to 69.5%. This year, however, it fell to a new low.
The striking thing about the post-2008 downturn is that it parallels the launch of the US Federal Reserve’s major stimulus programmes. The Fed has just two targets – employment and inflation. It has boosted its balance sheet by over $4tn dollars since 2008 to try and meet these targets, and cut interest rates to near-zero. But the policies have clearly failed, given the participation rate decline and that CPI inflation is virtually half the target level at 1.1%.
One might think that policymakers would feel some embarrassment about such a very expensive failure. But it seems not:
- Instead, they prefer to avoid focusing on these key trends. Their focus is on indicators which disguise the key trends, such as the jobless rate. This is a far less reliable indicator, as it only measures people actively looking for work, not those who have given up. So it is always likely to give a more favourable view of the position. Similarly, they prefer to talk about so-called “core inflation”, rather than the CPI itself
- This self-delusion has a purpose, of course. If they focused on the key data, they would soon have to explain why they were ignoring the demographic issues that are causing the participation rate to fall, and inflation to turn to deflation. They would also be forced to have difficult discussions with the voters about the sheer unsustainability of current policies – which have failed to adapt to a world where the average 65-year old American now has 20 years’ life expectancy ahead of them
The simple fact is that an extra 13m people would be working today, if the participation rate had remained at 1997′s level. There would then have been no need for the Fed to have created today’s vast debt levels. And the whole world would be in a better state as a result.
The UK economy appears to be recovering well from the financial crisis. But appearances can be deceptive.
Certainly employment has risen for both men and women since 2009, and the jobless rate has fallen. But new data yesterday from the Office for National Statistics highlights how, despite these achievements, total incomes have been falling in real terms since 2009, as the chart shows:
- Male employment has risen by 900k (blue shading) and female employment by 700k (red)
- Male and female employment have also hit all-time highs of 16.8m and 14.1m respectively
- But total male and female earnings have continued to decline in real terms as £2015 (black line)
- They peaked in 2009 at £1.1tn ($1.67tn), but have since fallen 9% to £1tn today
The reason, of course, is the factor that governments prefer to ignore, namely demographics. The UK is an ageing society, and earnings peak by the age of 50 as the second chart highlights (again based on ONS data):
- There were 5.8m people working in the 30-39 age group, with median earnings of £490/week ($750/week)
- 6.4m were working in the 40 – 49 cohort, with median earnings of £493/week
- But increasing life expectancy meant there were 5.4m working in the 50-59 cohort at £458/week, and 1.9m working in the 60+ cohort at £339/week
- Whilst falling fertility rates meant there were only 4.4m working in the 22-20 cohort for £383/week, and just 1.2m in the 18-21 cohort for £201/week
Of course, governments prefer to focus on achievements when they talk about employment. And with major spending cuts due to be announced next week, no doubt ONS felt it was better not to highlight the issue in their summary report. As far as I can tell from a Google search, none of the major media has mentioned this key fact in their reports.
However, it is clearly critical for both companies and investors. Consumption is around 2/3rds of the UK economy, and it is the 5th largest economy in the world. The data clearly highlights the fact that, like other major economies, the UK faces a future where there will be a declining number of people in the peak earning and spending Wealth Creator 25 – 50 age group, and a rising number of people in the lower-income and spending 50+ age group.
Companies can’t expect to sell products and services to people who can’t afford them. There is therefore an increasingly urgent need for them to develop new business models focused on providing low-cost essentials to the growth area of the over-50s.
Demographics is destiny, after all,
The US population reached 320m this year, an 11.35m increase versus 2010, according to the US Census Bureau:
“The U.S. is expected to experience a birth every 8 seconds and one death every 12 seconds, whilst net international migration is expected to add one person to the U.S. population every 33 seconds. All these factors mean a net gain of one person every 16 seconds to the U.S. population”.
Over the same period, the number of working age people has also grown, from 237.8m in 2010 to 250.3m last month. Yet sadly, the percentage of people in employment has fallen, according to Bureau of Labor Studies data, as the chart shows:
- 64.6% of the working age population were working in January 2010, versus 62.6% last month (black line)
- 70.9% of men were working then, versus 69% last month (blue)
- 58.7% of women were working then, versus 56.6% last month (red)
These rates are also slightly lower since November, when I last analysed the jobs market. So it is puzzling, to say the least, that most reporting of the US job market currently suggests it is steadily improving . Jobs are certainly being created, but not in sufficient numbers to maintain spending power and thus revive economic growth.
A simple calculation highlights the issue:
- Another 5.1m people would be working today, if the economy still had the same percentage of employed people as in January 2010 – 2.3m men, and 2.8m women
- This would provide major support for the economy, with median male earnings at $895/week in Q1 2015, and median female earnings at $730/week
Clearly the US, like other Western nations, needs to find a solution to this problem. But unfortunately it is currently using the wrong tools. Common sense, as well as the experience of the past 6 years, tells us that it is impossible for the Federal Reserve to create jobs in today’s economy via the use of monetary policy:
- One major problem today, for example, is that only 47% of those without a high school diploma have jobs
- This is far below the 75% employment rate of those with bachelor’s degrees
- Data for Black and Hispanic employment shows similar gulfs versus White employment rates
How can printing money change these fundamental causes of unemployment?
Hopefully next year’s Presidential election will include debate on this critical issue. Otherwise, on current trends, the US will have suffered a further 2% decline in participation rates by 2020, when the 45th President comes to the end of his/her term in 2020.
Apparently Friday’s US jobs numbers disappointed the experts. The consensus forecast was that 250k jobs would have been created in March – yet only half the forecast actually appeared. Even more tellingly, hiring estimates for January/February were revised down. Separate data also showed weak growth in wages and spending.
None of this was really a surprise, however. There was plenty of evidence that US employment had simply seen a temporary boost from the shale gas bubble.
My 8 January post was even headlined “US jobs growth at risk with end of the shale gas advantage“, and was followed a week later by a post titled “US economic recovery at risk as energy bubble bursts“. All the necessary data was easily available in the public domain, if anyone wanted to look.
But as during the subprime era, the consensus simply didn’t want to know. It was much easier to pretend to believe that somehow printing money could change the fundamentals of the US jobs market. But at the risk of repeating myself, the key data continues to be found in charts 16 and 17 of the Bureau of Labor Statistics monthly report, as shown above:
- US employment rates depend on race (chart 16) and educational level (chart 17)
- The jobless rate for Blacks (10.1%) is double that for Whites (4.7%) and Asians (3.1%), and 50% higher for Hispanics (6.8%)
- The rate for those without a high school diploma (8.6%) is 3x that for those with a bachelor’s degree (2.5%)
- The rate for those with a high school diploma (5%) is still double that of those with a degree
The issue is rather, as I noted back in September, that politicians prefer to ignore these structural problems in the economy. It is much easier to instead simply tweet about the need for more stimulus, and then deliver a sound-bite on the subject for the evening news bulletin.
The news is also embarrassing for the US Federal Reserve. They have spent nearly 2 years preparing to celebrate the success of their stimulus policies, since the ‘taper tantrum’ – when then Fed Chairman Ben Bernanke suggested in May 2013 the Fed would soon be able to “normalise” its policies.
But now that oil prices are returning to their historical relationship to natural gas, in terms of energy value, the bubble is ending. And unfortunately, it is taking with it the highly paid jobs that the bubble had created.
We now have full US Census Bureau data for housing starts in 2014, which shows:
- Starts returned to the 1m level for the first time since 2007
- They were also nearly double the low of 554k seen in 2009
- But at 1.006m, they were less than half of the 2.068m peak in 2005
The data also confirms the dramatic swing away from single-family homes towards multi-family apartments. These were one-third of total starts in 2014, around double the average seen between 1989 – 2007, and back to levels last seen around 40 years ago – before the BabyBoomer home boom began.
The data is also a sign of the overall decline in home ownership levels, which at 64.4% are back to 1995 levels, when records first began. As with employment, there is also a major divide between ownership rates for the relatively wealthy White population at 72.6%, and those for the poorer Hispanic (45.6%) and Black populations (42.9).
Equally significant is the data for the major regions in the US since 2009, as the chart above shows:
- The largest gain has been in the South, which averaged 500k starts in 2014 versus 260k in 2009
- The West also saw a large percentage gain, with starts doubling to 235k in 2014 versus 115k in 2009
- Gains in the North East and Mid-West were more modest at 50k and 65k respectively
As the map from the US Energy Information Agency on the right shows, 3 of these regions have also seen strong growth in oil/shale gas-related activity.
It highlights the 7 most prolific areas, responsible for all domestic natural gas production growth, and 95% of all domestic oil production growth, between 2011-2013.
Clearly it would need more detailed study to directly link this data with growth in housing starts. But we do know that workers have flocked to these regions in search of jobs.
Separate Census data shows that Houston had the 2nd fastest growth in population in 2013 (after New York), and Texas had 3 of the top 10 cities in the list of fastest growing cities.
And according to the Census, Texas added more housing units than any other state as a result.
But now, the boom is turning to bust, and companies are laying off workers – particularly in labour-intensive areas such as drilling and support services. Major job losses are already underway at Schlumberger and other key employers.
We won’t know till March or April just how bad the hit will be to housing. But it seems more than likely it will end the recent recovery in housing starts, taking 2015 levels back below the 1m level again.