Employment and pay – we delve into 5 years of data

19 February 2018

Employers signal information to potential recruits in a number of ways. Firstly, whether they are recruiting at all, and the methods they use to advertise jobs. Secondly, the pay offered.

The Office for National Statistics produces once a year a detailed analysis of the pay and hours of employees. The detail is enormous, which means that different users may quarry different things out of the data.

In this blog, we are drawing together three sources.

Firstly, the ONS release on occupational groups for April 2017.

Secondly, the equivalent release for 2011, the first time the same occupational groups were used.

Thirdly, our own analysis of the Labour Force Survey (a quarterly ONS survey) which records, among other things, the length of time someone has been in a job and the qualifications they hold.

There are 368 occupations in which there is at least some information in all three sources.

What we are comparing is:

  • The change in numbers employed in each occupation from 2011 to 2017 (on the chart, the horizontal axis).
  • The change in pay in each occupation adjusted for inflation between 2011 and 2017 (on the chart, the vertical axis)
  • The number of employees in each occupation (on the chart, the size of each circle)
  • The qualifications of job starters in each occupation (the colour of each circle – lighter blue is higher qualified)

Charting changes over six years

For each bubble, the information for that occupation is available if you move a mouse or pointer over the bubble. You can zoom in to parts of the chart by selecting an area. You can return to the overall view by double-clicking anywhere or there is a menu at the top right.

Each bubble (and the details that appear when you click on the bubble) are coloured according to the average qualification of job starters. The darker the colour, the lower the average qualification level. Qualification levels of job starters are related to pay, so darker blobs tend to be lower paid (there are exceptions, but they tend to be small groups).

We have four areas on the chart. These are:

  1. Top right. Growing jobs, rising pay. The most desirable area, where numbers in jobs are growing and pay, over six years, has risen faster than inflation.
  2. Bottom right. Growing jobs, pay falling. May be trading on earlier pay rises so the pay levels need looking at as well as the five-year change. Caution needed.
  3. Top left. Falling jobs, rising pay. In this area, employers are actively seeking to replace workers who leave or retire in otherwise declining jobs.
  4. Bottom left. Falling jobs, pay falling in value. Here be dragons.

Key findings

  • The weight of dots in the chart are on the right hand side – consistent with growing employment over the six years. Note we are measuring employees only so we are showing growth in employee numbers – not all employment growth is self-employed.
  • Many occupations affected by the National Living Wage are showing 5-10% real rises in pay. Where employers have reduced weekly hours, wage rises could be lower.
  • Most of the larger occupations affected by the National Living Wage show rises in employee numbers, with exceptions such as cleaning occupations.
  • The group of large occupations showing 7-10% real falls in pay and which are higher qualified are public sector occupations such as teaching (primary, secondary, FE) and nursing.
  • Construction employees tend to be in area 4 – falling jobs and falling pay. This applies to most contruction occupations of any size. The exceptions (e.g. plasterers) have small numbers so the estimates have wide confidence intervals.
  • IT occupations are well off to the right (programmers up 49% in jobs) while pay has done no more than keep up with inflation (in some cases not even that).
  • Employers faced with rising labour costs (National Living Wage, Apprenticeship Levy, etc.) can reduce hours worked and/or employment numbers. There is little sign of that here, which is consistent with flat productivity.

Cautionary notes

The changes in pay shown are affected by movements in the National Minimum Wage and the National Living Wage. This means that the six year change is likely to be higher than it would have been without these changes in legal minima. However, for these jobs, it is useful to look more closely at the changes in employment numbers. If the rises in pay were leading to job losses, this would be evident in the employment numbers. If employers are continuing to grow employment while having to pay higher earnings, this could be a good sign.

The earnings figures we are using here are gross weekly earnings, including both full-time and part-time workers.

Annualised, the 2017 earnings figures used range from Chief Executives and senior officials at £83,800 down to school midday and crossing patrol occupations at £3,032. These are the ‘median’ earnings – the point above (and below) which half the employees are paid. The average earnings for those occupations are £98,900 and £3,794 respectively.

Why do we use earnings including part and full-time workers equally?

  • We prefer to look at the whole labour market. Ignoring part-time jobs means ignoring a large (and substantially female) part of the labour market.
  • Employers shifting to a greater or lesser use of part-time work is an important part of the labour market story. If employers of an occupation are shifting to a greater use of part-time work, then median pay levels will drop in inflation adjusted terms and relatively to other jobs. The opposite applies if employers are moving away from part-time working.
  • Flexible working in terms of hours (which includes not only part-time but also zero hours and contracts with small numbers of guaranteed hours) is an important part of labour market information.

We are unable to include self-employed jobs in anything other than the job starts by skill level analysis. This is because the main data source is an employer-based survey conducted by the Office for National Statistics, which only covers employees.

Over the six years analysed, there has been a large increase in self-employment, but this is not in occupational groups where self-employment was strong before. Considered as a share of employment in the occupational group, the largest rise in self-employment was among managers, directors and senior officials, up from 20% to 24%, followed by process, plant & machine operatives – here, mostly drivers – up from 16% to 19%, and Associate professional and technical occupations, up from 15% to 17%. Skilled trades have the largest incidence of self-employment at 37%, but this is only up 1 percentage point.

The Labour Force Survey analysis covers the whole six-year period, because in each quarterly survey there are remarkably few people who have just started a job in many of the occupation groups. With 26 quarterly datasets, we amass 39,681 job starters (including those moving jobs). Even with this number of observations, there are 59 occupations with an estimated total number of job starts under 10,000 (in six and a half years). Most of these are not on the chart at all, as ONS refuses to provide estimates for either pay or employment as there are too few survey responses to provide good estimates.

This is enough to calculate an average qualification level but not a great deal more. We have counted those with a ‘trade apprenticeship’ and no other qualification information as Level 2.5.

ab