Wednesday, 5 October 2016

Let's Talk about Immigration: Young People, Jobs and Training

The Home Secretary on the Today programme said that she was happy to talk about immigration in the context of suggesting that there may be a link between immigration and lack of jobs and training opportunities for young people.

So let's talk. It is very hard to find any link in the UK between immigration, jobs and training opportunities for young people. And conversations that suggest there may be a link are probably not very helpful.

The chart below graphs the change in the  NEET rate (not in employment, education or training) for UK-Born 16 to 24 year olds alongside the change in immigration rate for each of around 200 local authorities in the UK between 2008 and 2015, a period in which immigration continued to grow rapidly  The red line  summarises the relationship between the two variables. There isn't much of one. - but if anything the NEET rate for young UK-born people fell more in areas that received more immigration.

(technical note: the source is the Annual Population Survey and the data that make up the red line are weighted to reflect the different sample sizes in each area).

Does it matter if we use EU or the eastern half (A8 and A2 migrants)? No. Equally we can do the same type of graphs for training or apprenticeship rates or different time periods - available on request - for young people and the lack of association emerges each time.

What does probably matter more for young people's chances is not immigration but the state of the local economy.  The next graph plots the association between the young person NEET rate and the changes in the local area employment rate for adults 25 and over. The association with local employment opportunities for older workers is more important than changes in immigration

Get the local economy going and NEET rates will go down and opportunities for young people will go up. That is probably a better conversation to have

Jonathan Wadsworth
CEP & Royal Holloway College

Wednesday, 11 May 2016

SWOB 10. EU-turn if you want to. Brexit & Immigration

Immigration has for some years been the uppermost worry among the issues thought to be facing Britain in many opinion polls so it - or rather people's perceptions of its extent and its effects - is almost certainly one of the key issues that will influence the upcoming vote on whether to stay or remain in the EU.  A new report from the CEP  looks into this. Workers have had a rough ride in recent times. Real (inflation adjusted) wages fell by around 10% in the years after the global financial crisis of 2008 and the ensuing austerity. Such a sustained fall in pay is unprecedented in British post-war history. 

Alongside falling wages, immigration from the EU has gone up a lot. Between 1995 and 2015, the share of EU nationals as proportion of working age adults more than tripled from 1.8% to 6.3%, and most of this happened after 2004 when the eight East European countries – the A8 – joined the EU. 

So it is understandable that many people believe this migratory wave hurt UK-born workers and hence leaving the EU would make things better.  

At face value this might seem plausible. Since immigration increases the total number of people in work or looking for employment, doesn't that mean that UK workers unemployment must have been harmed by this increased competition for jobs? 

The short answer is no. There would be harm only if the total number of jobs was fixed. But since immigrants also consume local services and goods, this increases demand and so raises job prospects for all.

But even if there is no reason to think that immigration should increase unemployment, is it not obvious that an increase in the supply of workers must drive wages down? No. Alongside the increased demand that a rising population brings, greater movement of labour allows countries to specialise in what they are best at, just like increased trade. Firms will change the mix of their products to account for the new skills available to them. Immigrants, especially if they are more skilled, can boost productivity. All these effects will tend to increase wages.

Consequently, the impact of immigration on UK-born workers is an empirical question and not a foregone conclusion. We need to look for evidence.
The graph above plots the unemployment rate for individuals born in the UK alongside the changes of EU immigration between 2008 and 2015 (one dot for each of 200 local authorities). The solid red line summarises the relationship between immigration and UK-born unemployment rates. If immigration increased unemployment, we would expect a strong upward sloping line – more EU immigrants would mean more unemployment for local workers. But it is very clear from the graph that there is no significant relationship (negative or positive) of EU immigration on unemployment rates of those born in the UK over this period.

Why is this? Just because immigration and unemployment both go up in certain areas on the graph does not mean that EU immigration causes more unemployment for UK-born workers, since it is quite easy to find areas on the graph where rising EU immigration is accompanied by falling unemployment. Something else must underlie the ill fortune of areas with rising unemployment.

The next graph does the same analysis of the impact of EU migration on pay. Again, there is no apparent link between change in the real wages of UK nationals and changes in EU immigration. Wages of UK-born workers changed at much the same rates in areas with lots of EU immigration as in areas where the change in EU immigration was low.

The upshot? It is hard to put the blame the EU immigration for any adverse experiences of UK-born workers with regards to difficulties over jobs and falling real wages. Rather the finger should point to the largest economic crash for more than 80 years and ensuing austerity

More details on this can be found in a new report from the CEP . We will return to this in a future

Jonathan Wadsworth
CEP & Royal Holloway

Monday, 18 April 2016

SWOB 9. Is it Worth It? Are there too many graduates in the UK?

Welcome back. The Bank of England has a regular labour market commentary in its quarterly bulletin in which it looks at issues that may influence productivity, wage pressure and hence inflation (if wage pressures are passed on to consumers in the form of  higher prices). In their latest post there is a chart which suggests that the wage benefits of being a graduate have receded over the past 20 years, which they suggest may be because the supply of graduates has risen faster than demand or that the "signalling" value of a degree is more muddied now because so many people have one.  Yet the IFS published a study this week saying there were large benefits to being a graduate - though some graduates are more equal than others.  So what's going on?
Well partly the studies use different data and so focus on different aspects of graduate employment. As such it may be worth looking at some basic facts about education, work and pay. The graph below shows that the share of graduates in the population is indeed rising quickly. Nearly one quarter of the UK working age population now has a degree, up from one in fourteen in the mid-1990s. The shares of those with A level (or equivalent) and GCSE qualifications are broadly stable, while that of those with no qualifications is falling rapidly. (This doesn't mean that all school leavers are now becoming graduates, rather that shifts in educational attainment  manifest themselves this way). There are now more than 3 times as many graduates in the working age population (11.5 million) than those with no qualifications 3.3 million). This means that the average graduate is not the same as the average graduate 20 years ago. Neither is the average person with no qualifications. For example, the typical no qualifications is 2 years older with more work experience on average now compared to 1995. The typical graduate is 2 years younger now with less work experience - but the variation in experience for graduates is also now so much bigger than in the past - largely because there are so many more of them. So it is becoming harder to summarise the experience of gradates by the average.  An issue we will return to in a future blog.

With that caveat in mind here is a graph of pay differences by education over time.  Measuring pay differences is a little tricky since a simple comparison of wages across groups at any point in time, takes no account of differences in pay that might be caused by other things (like experience). So each line below plots the average  approximate % average difference in pay of groups with differing levels of educational attainment after netting out the influence of age, gender and region of residence on pay. (future blogs will look at how sensitive these gaps are to netting out different effects). While falling by around 10% over 20 years, the graduate premium is still around 100% higher relative to no qualifications and 50% higher than the average A-level or equivalent worker (the graph is in logs so these are only approximately % differences, 70 log points is around 105%).   On a full-time salary, that's an average yearly pay difference of around £20,000 and £10,000 respectively in favour of graduates. And while that is an average across all ages, those pay differentials are more or less sealed after the age of 30

Source: LFS

The other issue with regard to education is that cushioning impact it has on job prospects. If graduates were not in demand we might also expect to see their relative unemployment rates falling over time. The last graph tracks the chances of being unemployed for each qualification group relative to those with no qualifications (net of other influences on unemployment like age gender and region). It can be seen that a graduate education was the best means, on average, of securing employment over the downturn and in a recovery.

So what are we to conclude? Is there an over-supply of graduates?  Is it still worth going to university? If measured by the extent that being a graduate - despite there now being so many more of them - still generates an average pay premium and employment protection way in advance of other educational types then the answer is almost certainly still yes, even net of the now considerable cost of financing a graduate education. It's just a shame about all those fridge sharing issues that will ensue.

Friday, 18 March 2016

SWOB 8: Employment Performance: UK versus the USA. Jury Out

Lost in the Budget fog this week was the latest set of UK labour market performance statistics among which (buried in an increasingly impenetrable government website ) were international comparisons of employment.  The USA is often held up as an exemplar of what the UK labour market might be (for better or worse) and so it can be useful to try to benchmark the relative position of the two economies with regard to jobs.

The graph below plots four lines. The top two lines are the non-employment rates for individuals of working age (living in households where the head is below retirement age & removing any students). Blue for the UK. Red for the USA.  This is a metric often used to measure labour market performance - it is just one minus the employment rate.  On this basis the UK labour market is doing rather well. Non-employment is at its lowest level for at least 40 years. It is also doing much better than the USA using this metric and has been since well before the last recession.

The bottom two lines, however track the share of all households that are workless - counting the number of individuals in each household not in work using the same definition of employment as on the individual counts. For the UK, this measure of joblessness is 3 times as high as in the mid-1970s. Around 15% (1 in 7) of all working age households are now workless -  despite the improvement in the individual based jobless count the workless household rate has improved but mot as much as the individual count. Not so good.  And worse than the US where despite a relatively poor individual employment rate the US managed to avoid some - though by no means all - of the concentration of joblessness in households. In other words, the USA has a lower workless household rates despite having fewer jobs per head of population than the UK

What this means is that access to jobs is still unevenly distributed across households more so in the UK than in the US. So UK employment performance looks good when using individual based counts, less good when using household based counts. Which is better? Neither. Both say something important about the state of the labour market. We should get used to judging performance by more than one yardstick.

What does all this mean for the government's targets of children living in workless households and long-term workless households? We will return to this in future blogs.

Jonathan Wadsworth

Friday, 11 March 2016

SWOB 7:International Women's Day: Reasons to be (a little bit) cheerful about women in the UK labour market, 1, 2, 3

On Tuesday, International Women's Day, the Fawcett Society published a review of the the progress (or otherwise) of women in the UK labour market. Their central finding was that women - in full-time work - earn significantly less than men which can amount to some £3000,000 over a working lifetime. Clearly while the gender pay gap has narrowed over the years, something still needs to be done.Here, however, are 3 features of the labour market where women have done relatively well recently (according to the Labour Force Survey data).

1. Women are now more educated than men. It's official
The graph below shows the number of men and women (aged 16-64) by educational attainment. There are now more women educated to degree or further education level than men. Successive waves of undergraduate entries in which women have been in a majority mean that there are now 6 million female graduates in the population and 5.5 million male graduates. Similarly for further education. This can only help narrow pay differentials in the future (assuming pay for education skills is rewarded equally).

2. Women are now a majority of employees in several areas of the UK
Rising participation for women and stagnant participation of men have combined so that women are the majority of employees (but not all employed, self-employment is still male dominated) in South Yorkshire, East Yorkshire, Merseyside and Scotland. Indeed women now make up the majority of all the employed in Merseyside. Some of this is undoubtedly due to years of industrial decline and the loss of male dominated sectors, but the consequences have - so far - been to feminise the workforce.

3. The Gender Pay Gap has long since disappeared in Part-Time Work
Indeed there is a premium to being female and working part-time. (though this goes to zero when differences in part-time working by age, qualifications and region across gender are accounted for) So the main issue here is now no longer gender based,  but instead why part-time jobs are paid so poorly relative to full-time jobs.  Progress (of sorts)

Jonathan Wadsworth
CEP & Royal Holloway

Friday, 4 March 2016

SWOB 6: Record Employment in the UK? Well yes but.......

Common Mis-Perceptions About Recent UK Labour Market Performance No 1. A Record number of people in work

The opening sentence of the latest DWP press release on the state of the labour market trumpets  that there are "a record 31.4 million people in work". While this is true, it should not be taken as a measure of employment performance (and beware of anyone who claims it is) for the simple fact that the UK population is also at a record high and growing. Just as it would not make sense to compare the sizes of the workforce in Norway and Spain and conclude that Spain was doing better simply because it has more people in work. The respective sizes of the populations make such a comparison of employment levels meaningless. While not on the same scale, UK population growth (currently around 250,000 a year - caused now principally by immigration, though in the past caused by baby boomers entering the labour force) is a major contributor to growing numbers in work. Rising populations tend to generate more demand (more people means more food, clothes etc) so it should not be too surprising that the numbers in work are rising when the population is.  The UK working age population (16-64) has grown by around 1 million since the end of 2007 (source: ONS labour market statistics).  If the employment rate was 70% then we might expect 70%  (700,000) of that extra 1 million to be in work, notwithstanding the performance of the economy.

So we should be cautious about pronouncements of record levels of employment.  (just as we should with comments about GDP growth rather than GDP per capita growth for the same reasons)

The numbers in work are also rising partly because the retirement age (for women) is rising and partly because more people are in work beyond pensionable age. This makes cross-time comparisons harder. In the years when women retired at the age of 60 we should not expect the numbers in work to be as large as now, other things equal.

Conversely, most people entered work at age 16 in the 1970s and 1980s. Fewer people now leave school or find work at age 16. This would tend to reduce numbers in work (but then part-time work among students is counted).

So...... it is all a bit tricky to compare employment levels over the long-term.

What is to be done?

A better - though not perfect - test of performance is to look at the employment rate (as the DWP also do). This is the number in work divided by the population of working age. Much of the change in employment caused by changes in population will be netted out by doing this. Most but not all.  Cross-time comparisons of the employment rate will still be influenced by things like changes in the retirement age and staying on rates at school and college over time.

Possibly one way forward is to compare how long it took for the employment rate to get back to its level before any downturn.  In other words look at the relative performance of the employment rate across different recessions and subsequent recovery. That way the problems of comparing the rate itself over longer periods of time are reduced and it is easier to see how this recovery compares with previous ones.

The figure below looks at the relative performance of the employment rate (16-64 year olds) in the last 4 downturns. So the employment rate is now, after 8 years, around 2% higher than it was before the downturn of 2008.

On this basis it is indeed true that a) the time taken to get back to where things started was quicker this time round but that b) the fall in employment was not as large as in previous downturns and c) the 3 years of stagnation at the bottom of the peak is unusual in the latest downturn compared to the downturn of the 1980s & 1990s  (the 1970s recovery in employment rates never got going much before it ran into the 1980s recession)

If we do the same with the unemployment rate then the 1990s recovery looks to be stronger than the current one (though the 1980s and 1970s look much worse on this basis)

Of course the nature of the jobs being created is another matter as is the issue of whether there was a trade off of jobs for lower pay growth. We will return to this in a future blog

Jonathan Wadsworth
CEP & Royal Holloway

Thursday, 25 February 2016

SWOB 5: Rumours of the end of the job for life are a little exaggerated (but not if you're male)

For more than 20 years commentators have commented as they are wont to do on the end of the job for life. So here's a thing. Long-term job stability in the UK is on the up. - if you are female.
If you are male, however, secure job prospects are not as prevalent as they once were.

Look at the graph below. It shows the share of  all employees who have been with the same employer for 10 years or longer. This is a proxy for long-term stability. Since younger workers cannot expect to have been in a job for more than 10 years, the sample is restricted to those aged 35 and above. The middle line shows a small fall in the share of workers in a long(ish) term job - down from around 46% of employees in the 1980s to around 43% now.  Note how the 2008-2012 downturn led to a fall in long-term jobs but that the subsequent recovery has almost offset this drop.

Source: LFS
Now look at the trends for men and women. Very different. The long-term job share for men has fallen a lot since the 1980s (though even this recovered a little in recent years). In contrast, the long-term job share among women has risen - not withstanding the recession.  Once again this is another manifestation of the rising feminisation of the workforce. Women are staying with employers longer, largely because family friendly working enables them to stay with an employer  (an issue to be discussed in future blogs).

Another way to look at the issue is to measure the share of short-term jobs (those that have lasted for less than a year). If everyone were changing jobs a lot more than in the past then this short-term job share would be higher than ever. It isn't. While the graph below is highly cyclical- hiring falls a lot in downturns - it is clear that short-term jobs are much less prevalent now than in either the 1980s or in the 2000s.  So it does seem that rumours of the end of the job for life are still a little exaggerated.

Jonathan Wadsworth
CEP & Royal Holloway