1. Main points

  • The UK unemployment rate in the three months to February 2018 was 4.2%, the lowest since 1975.

  • The employment rate was at a record high during the same time period, at 75.4%.

  • There has been an increase since the economic downturn in the responsiveness of employment to economic growth.

  • Industries with high employment intensity of growth help to contribute to growing employment and low unemployment.

  • A large proportion of employment growth happens in industries where the average number of weekly hours worked is below the overall mean for the economy.

Back to table of contents

2. Unemployment

Latest estimates from the Labour Force Survey show that the number of unemployed people in the UK fell by 16,000, leading to a fall of the unemployment rate by 0.1 percentage points to 4.2% in the three months to February 2018 when compared with the previous quarter.

The unemployment level for men fell by 17,000 in the three months to February 2018 when compared with the three months to November 2017 and the rate fell to 4.2%. For women, unemployment increased by 1,000; although this rise was not large enough to have a material impact on the female unemployment rate, when rounded, the female unemployment rate constituted 4.2% in the three months to February 2018, compared with 4.3% in the three months to November 2017.

Back to table of contents

3. Employment

As the level of unemployment fell, the level of employment in the UK continued its rise, leading to a record-high employment rate of 75.4% in the three months to February 2018. The growth in the number of people in employment was driven by employees (increase of 103,000), with the number of self-employed contracting by 18,000 compared with the previous quarter. In the March 2018 commentary, we looked at the longer-term trends in self-employment and noted that the recent slight decline came after a prolonged period of growth, where the rate of growth for part-time self-employed exceeded that of their full-time counterparts.

Employment for both sexes increased, by 8,000 for men and by 47,000 for women when compared with the previous quarter. This increase brought employment among women to an unprecedented high of 15.17 million.

Changes in employment levels by age group were fairly consistent across the board, with every age group except those aged 18 to 24 years experiencing an increase in employment. Of some interest is the highest-recorded employment for 50 to 64-year-olds, which increased by 28,000 on the quarter to 8.9 million.

Broken down by working pattern, full-time employment increased by 25,000 compared with the three months to November 2017, which is less than an increase in part-time employment (29,000) that brought part-time employment to 8.58 million, a level that has not been seen before. Persistent employment growth in recent periods is a major factor contributing to the low rates of unemployment. However, the increased headcount may conceal the compositional changes in the labour force. We touched upon the changes in the working patterns in the February 2018 commentary and showed that part-time working is gaining more prominence particularly among those in self-employment in the March 2018 commentary. In this commentary, we will consider the responsiveness of employment to economic growth and its industry composition.

Back to table of contents

4. Employment intensity of growth

Figure 3 looks at the developments in real gross domestic product (GDP), employment and total hours worked in the economy since the start of 2008. Following a contraction in output during the downturn, the total number of hours worked fell further than employment. One reason for that is “labour hoarding”, a situation where companies continue to employ more workers than they need to produce the diminished amount of output. This phenomenon occurs for reasons such as fixed costs associated with redundancies (typically because of employment protection legislation) as well as expectations that skilled labour will be required in the recovery phase.

The longer time series suggests that compared with the earlier downturns, employment appeared less responsive to the fall in GDP and exhibited a relatively moderate contraction in the wake of what turned out to be a major economic downturn. Generally, hours worked tend to grow faster than employment headcount in the initial phase of the recovery, whereby the existing “hoarded” workers can meet the demand for extra output. It is therefore interesting that in 2010, employment started growing as soon as hours worked started recovering. Since 2013, both the hours worked and employment have followed a very close trend.

In the March 2018 commentary, we briefly looked at the quarter-on-quarter growth of the real GDP and employment. We explained that employment growth in percentage terms is normally slower than the growth of GDP, consistent with a flatter than GDP line slope of the employment and hours work lines. The stronger-than-before growth dynamics of employment in comparison with GDP raises the question of the sources of this expansion. The responsiveness of employment to GDP growth is often expressed in terms of employment elasticity. Since the last economic downturn, for any percentage point growth in GDP, there is now a higher growth in employment than prior to the downturn. In other words, employment elasticity has increased.

As we discussed in Section 3, one dimension of this conundrum is that not all jobs are created equal. Post-downturn years saw a substantial growth in the share of part-time workers, mostly among the self-employed. Many of these new part-time jobs are in fact found among the higher-skill end of the occupational spectrum (although the rise in share of higher-skilled occupations can be partly attributed to the increased general standard of education and the so-called “job title inflation”). Additionally, the degree of implicit underutilisation of labour may still be higher now than before the downturn, despite a material improvement in recent years. This means that people classified as being in employment may not be working as much as they would want to.

A decomposition of economic growth also provides a useful insight into employment growth. An increase in employment elasticity would be indicative of the economic growth shifting to more employment-intensive activities1.

Notes for: Employment intensity of growth

  1. Employment intensity of growth and labour intensity correspond to different concepts. Although both measure the amount of labour input used in production, in the case of labour intensity, the labour share is often expressed as labour compensation. Hence, a higher labour share does not necessarily mean higher headcount. Although the two tend to be correlated, scientific and legal services are the examples of activities where high labour intensity may in fact be associated with relatively few jobs.
Back to table of contents

5. Industry decomposition of growth

Employment intensity of growth is closely related to the notion of labour productivity – any given amount of output produced requires a certain level of employment (or, more generally, units of labour), multiplied by labour productivity (the value of output per unit of labour input). Figure 4 compares the cumulative change in gross value added (GVA) and employment by industrial sector (Standard Industrial Classification (SIC)) in 2014 to 2017. For illustrative purposes, it also shows the four-year period before the last economic downturn.

One can see that between March 2014 and December 2017, just under one in three new jobs were filled in the distribution, transport, accommodation and food sectors (SIC sections G to I)2. Back in 2004 to 2007, only about one in ten jobs and 17% of real GVA were created in the distribution, transport, accommodation and food sectors (SIC sections G to I), the major components of which are categorised as less knowledge-intensive by Eurostat.

Administrative and support activities (SIC section N) were the second-largest contributor to workforce jobs growth between March 2014 and December 2017, having a far more modest share of real GVA increase and being also considered a less knowledge-intensive industry. Together with activities such as arts, recreation and household services (SIC sections R to U), these three broad categories of services contributed to over a half of cumulative jobs growth but only to a third of the GVA increase.

This observation should not be interpreted as an indication of slow output growth in these sectors. In fact, taking the example of distribution, transport, accommodation and food sectors (SIC sections G to I), the real GVA growth constituted 12%, compared with the 7% jobs increase between March 2014 and December 20173. Instead, the data suggest that for a given percentage gain in GVA, employment growth was higher than average.

Notes for: Industry decomposition of growth

  1. The chart shows industries that contributed to positive growth in GVA and/or net workforce jobs. As such, Agriculture, forestry and fishing (SIC section A) is not shown separately as its share did not exceed 1% of either GVA or net jobs growth. Similarly, production (SIC sections B to E) is not visible on the 2004 to 2007 segment of the chart as it saw contraction in employment (therefore not contributing to job creation) and negligible cumulative contribution to GVA at around 0.01%.

  2. These statistics describe the net increase in the number of workforce jobs. For any industry, this corresponds to the number of workforce jobs created over the course of the three-year period less the number of destroyed jobs.

  3. For real GVA, values corresponding to Quarter 1 (Jan to Mar) 2014 and Quarter 4 (Oct to Dec) 2017 are used.

Back to table of contents

6. Changes in average weekly hours worked

Employment and workforce jobs measure the number of workers and filled jobs respectively, not the amount of labour used. Leaving aside the multi-factor effects of productivity, for any given amount of labour input, an increase in the number of part-time workers, relative to their full-time counterparts, would lead to higher headcount without necessarily increasing the output.

The data show that on the one hand, the industries attracting higher employment growth are not necessarily the industries with the growing share of part-time employees, or decreasing average weekly hours worked per person. However, they do tend to be the industries where average weekly hours worked are already below the average across the entire economy. Services with traditionally higher than average hours per worker, such as information and communication (SIC section J), and financial and insurance (SIC section K), contribute to a smaller share of the overall employment growth.

The economic growth driven by services with high employment intensity and changes in the composition of the labour force itself, may mean that employment elasticity will remain at a higher level than before. In other words, employment may be responding to output growth more strongly.

Back to table of contents

7. Authors

Ed Moskalenko and Aidan Beresford, Office for National Statistics.

Back to table of contents

Contact details for this Article

Eduard Moskalenko
Telephone: +44 (0)1633 456738