In focus - The labour market

3.1 What do surveys of firms tell us about the labour market?

Surveys suggest businesses are increasingly reluctant to increase headcount…

A number of surveys suggest firms’ appetite to expand employment has fallen since mid-2018 (Chart 3.2). The decline has been particularly sharp in the REC survey. This may overstate the fall in demand for labour as it only captures demand for new recruits. The Bank’s Agents’ score measures a different concept: firms’ expected change in headcount, which takes into account the number of employees leaving. That score has declined to a lesser extent, but has still weakened. It is now consistent with roughly no change in headcount over the coming year. The BCC survey suggests stronger demand for labour, but is still consistent with a softening.

The REC survey suggests the demand for new permanent staff has been weaker than that for temporary staff recently (Chart 3.3). This could reflect a reluctance to commit to permanent hires because of uncertainty about the economic outlook. Uncertainty stemming from the Brexit process has caused firms to delay or cancel capital investment (see Section 4 in the November 2019 Report). There could have been a similar effect for hiring, with firms increasingly valuing the ability to reverse the hiring decision (Broadbent (2019)). Despite this, the share of employees on temporary contracts has fallen slightly this year.

Chart 3.2 Surveys of employment intentions have weakened

Surveys of employment intentions ^(a)

Sources: Bank of England, BCC, IHS Markit/CIPS, KPMG/REC Report on Jobs and Bank calculations.

(a) Surveys from the Bank’s Agents (employment intentions over the next 12 months), BCC (employment expectations over the next three months), IHS Markit/CIPS (PMI composite employment index) and KPMG/REC (index of demand for new staff). Agents’ scores are monthly until August 2016 and six weekly thereafter. BCC data are quarterly.

Chart 3.3 Demand for permanent staff has been weaker than for temporary staff

Difference between permanent and temporary staff demand indices ^(a)

Source: KPMG/REC Report on Jobs.

(a) Three-month moving average.

…and the number of job vacancies has fallen sharply.

The decline in demand for new employees is also apparent in the official vacancy figures from the ONS. Although the number of vacancies remains high, it has fallen by 7% over 2019, the biggest fall for a decade (red line in Chart 3.4).

The decline in vacancies has been widespread across the private sector, although the manufacturing and transport, storage and communication sectors have accounted for a disproportionate share (Chart 3.5). Together, these sectors account for just over half of the decline since January. The decline in the manufacturing sector is consistent with the sector’s falling output over the past year, which is in part the result of a slowdown in global growth (Section 2).

The number of vacancies per unemployed worker remains high though.

Although the decline in vacancies is a signal that demand for labour has fallen, it is less clear if the labour market has loosened. The ratio of vacancies to unemployed workers (the ‘V/U ratio’) is often used as a measure of labour market tightness as it shows how many jobs are available for each unemployed person. Although the number of vacancies has fallen, the pool of unemployed workers has also shrunk. As a result, the V/U ratio remains close to a record high (blue line in Chart 3.4).^[1] Survey measures of recruitment difficulties also remain elevated, consistent with a tight labour market.

Chart 3.4 Vacancies have fallen, but the ratio of vacancies to unemployed people remains high

Vacancies and unemployment

Sources: ONS and Bank calculations.

Chart 3.5 Two sectors account for over half of the decline in vacancies

Change in the number of vacancies since January 2019 by sector

Sources: ONS and Bank calculations.

3.2 What has happened to employment and flows around the labour market?

Employment growth has slowed slightly, which could reflect both demand and supply drivers.

Employment growth has been volatile this year. Quarterly growth was briefly negative in September, although it has since recovered to 0.6% in the three months to November. Looking through the volatility, quarterly growth has averaged 0.2% in 2019. That is a little lower than recent years, and the lowest since 2011.

Lower employment growth could reflect weaker demand for labour as a result of weaker demand for goods and services. Slower global growth and domestic Brexit-related uncertainties weighed on UK GDP growth in 2019 (Section 2). Historically, employment growth has tended to be highly correlated with GDP growth, albeit with a short lag. Lower demand would be consistent with the evidence from surveys and vacancies.

However, lower employment growth could also reflect limits to labour supply. Employment growth might be expected to slow as spare capacity in the economy is eroded. As the pool of unemployed workers shrinks, there are fewer people left to move into employment. Unemployment is currently below the MPC’s estimate of its equilibrium rate (Section 4).

Labour market flows can help us identify turning points and the nature of shocks hitting the economy.

Labour market flow statistics offer an alternative perspective (Chart 3.6). On average, every quarter around 6% of working-age people move between the three labour market states: employment, unemployment and inactivity. A further 2% move between jobs. The size and direction of these flows can reveal the first signs of turning points in the labour market. They may also shed light on the nature of the shocks hitting the UK economy.

The number of people switching jobs has fallen.

In 2019 Q3, the proportion of people moving between jobs fell to its lowest level since 2014 (top panel, Chart 3.7). These figures are quite volatile, and the latest move might just reflect sampling. Nevertheless, the size of these job-to-job flows tends to be negatively correlated with unemployment (Gomes (2012)) as a tight labour market encourages growing firms to poach workers from other firms (Haldane (2019)). So the UK’s recent experience of low and stable unemployment with subdued job-to-job flows is somewhat unusual.

The job-finding rate is also low given the tightness of the labour market…

The proportion of unemployed people moving into employment — the job-finding rate — has recently picked up (middle panel, Chart 3.7). But it has been unusually low over the past two years given the tightness of the labour market (Chart 3.8).

Chart 3.6 The flows between labour market states offer an alternative perspective

Employment, unemployment and inactivity as a proportion of the working-age population in 2019 Q2–Q3 ^(a)

Sources: Labour Force Survey and Bank calculations.

(a) Bank staff estimates using the Labour Force Survey. Stocks and flows are expressed as proportions of the 16–69 year old population. All figures are seasonally adjusted by Bank staff.

Chart 3.7 Job-to-job flows have fallen recently; the flow out of employment into unemployment has remained low

Job-to-job flows and flows between unemployment and employment ^(a)

Sources: ONS and Bank calculations.

(a) Two-quarter flows. Job-to-job flows are based on total employment of people aged 16–69. Flows between employment and unemployment are based on total employment and unemployment of people aged 16–64. Dashed lines are averages from 2002 to 2007.

…but unemployment has remained low because few people have been leaving employment.

The low rate of unemployment over the past two years has been sustained by relatively few people leaving jobs, rather than large numbers starting jobs. The quarterly flow of workers moving from employment to unemployment — the job destruction rate — is close to a record low (bottom panel, Chart 3.7). This is partly the result of relatively few workers being made redundant. The redundancy rate hit a record low at the end of 2018, although it has picked up a little in recent months (Chart 3.9). The low redundancy rate is consistent with the labour market remaining tight, despite the recent softening in demand for new workers.

Chart 3.8 The job-finding rate has been low given the tightness of the labour market

Job-finding rate and labour market tightness

Sources: ONS and Bank calculations.

(a) The job-finding rate is the percentage of unemployed people that move into employment, as shown in Chart 3.7. It is a two-quarter flow based on total employment and unemployment of people aged 16–64.
(b) Measured using the vacancies to unemployment ratio, as shown in Chart 3.4.

Chart 3.9 The redundancy rate has picked up in recent months, but remains close to its record low

Redundancy rate

Sources: ONS and Bank calculations.

Overall, the flows suggest unemployment will remain low…

Although job switching and job finding are lower than we might expect, low job destruction suggests that a pickup in unemployment is not imminent. If the flows were to remain at their current levels, unemployment would also remain around its current rate.

…but they do show that churn has fallen, which could reflect caution on the part of firms and workers.

However, the flows do suggest some softening in the demand for new workers, corroborating the employment surveys and vacancy numbers. Lower labour market churn could reflect economic uncertainty discouraging firms from engaging in costly, hard‑to‑reverse recruitment. Uncertainty may also have encouraged labour hoarding, contributing to the low rate of job destruction (Section 4). Low churn could also reflect low confidence on behalf of workers, who might be reluctant to take risks when the outlook is weak. The proportion of households expecting unemployment to go up over the coming year has increased over the past two years (Chart 2.21), although measures of personal job security have been relatively resilient. Other factors may also have contributed to the fall in job-to-job flows, however. For example, the workforce has been ageing in recent years, and older workers tend to move jobs less frequently (Saunders (2018)).

Although the flows data do not point to an imminent increase in unemployment, there may be other implications of lower churn. Some studies have found job-to-job flows to be a useful predictor of pay growth (Moscarini and Postel‑Vinay (2017)), in which case the recent fall may suggest pay growth will weaken. But Bank staff work suggests the unemployment rate — the traditional determinant of pay growth in macroeconomic models — is at least equally important for pay growth in the UK, and that has remained low.

Unemployment has remained low among groups that are most sensitive to a downturn…

The unemployment rate remained at 3.8% in the three months to November (Chart 2.1). Bank staff’s estimate of the claimant count rate^[2] — a timelier indicator of unemployment — fell to a record low of 0.7% in December.

Youth unemployment tends to be more procyclical than total unemployment. This is because young people are often entering the labour market for the first time and are therefore dependent on firms’ willingness to hire. Younger people in employment may also have more informal, less secure working arrangements. Although unemployment among 18–24 year olds drifted up slightly over 2019, it fell back in October and remains below the average over the past two years (Chart 3.10).

…and among groups that tend to lead total unemployment.

Certain types of unemployment tend to lead the total. The rate of short-term unemployment leads overall unemployment by around six months as it contains newly unemployed people. Short-term unemployment has picked up very slightly over the past few months, but remains well below average at 2.4% (Chart 3.11).

Chart 3.10 Youth unemployment is more procyclical, but has not picked up materially

Unemployment rates by age group

Chart 3.11 Short-term unemployment has picked up slightly over the past year, but remains low

Unemployment rates by duration ^(a)

Sources: ONS and Bank calculations.

(a) The number of people unemployed in each category, divided by the economically active population. Dashed lines are averages from 2002 to 2007.

3.3 What does the recent fall in pay growth signal?

Pay growth has fallen back slightly in recent months.

Pay growth has slowed in recent months. Private sector regular pay growth was 3.4% in the three months to November, down from a peak of 4.0% earlier in the year (Chart 3.12). Whole-economy total pay growth — a broader but more volatile measure of pay growth — was 3.2%, down from a peak of 3.9%. New experimental statistics from the ONS and HMRC using payroll data suggest the slowdown has been felt across almost all parts of the UK and across most of the pay distribution.

Slower pay growth might reflect weaker demand, but the decline is relatively small…

Pay pressures reflect the balance of supply and demand in the labour market. The decline in pay growth may therefore corroborate a fall in labour demand. However, the tick down in pay growth is small relative to the steady increases seen in recent years. Growth has averaged over 3½% on both measures over 2019, the highest annual averages since 2008. That strength reflects stronger pay growth for both workers moving jobs and workers staying in jobs: 2019 was the first year in which ‘stayers’ received an average pay increase of 4% since 2007 (Chart 3.13). That is consistent with the Agents’ latest pay survey in which the ability to recruit and retain staff was cited as the largest upward pressure on settlements (Box 2).

Chart 3.12 Pay growth has slowed over the past few months, but is still higher than recent years

Measures of pay growth ^(a)

(a) Three-month average growth on the same period a year earlier.

Chart 3.13 Pay growth picked up in the year to April 2019 for those staying in jobs as well as those moving jobs

Median annual growth rates of pay ^(a)

Sources: Annual Survey of Hours and Earnings and Bank calculations.

(a) Median annual growth rate in April. Based on hourly gross earnings obtained by dividing gross pay in the reference week by total hours worked. Workers moving jobs are defined as workers in employment in consecutive years but in a different job.

…and pay growth was probably boosted in mid-2019 by temporary factors.

The slight slowdown in pay growth also reflects the impact of temporary factors unwinding. The official pay growth figures can be temporarily affected by the changing composition of the workforce. Employment has increased in higher paying occupations and industries over the past year. This boosted average pay growth in the first half of 2019 (Chart 3.14), but only does so as long as the compositional shift continues. Such effects have tended not to persist in the past, and the total positive impact had already begun to reduce in size in 2019 Q3. That has brought pay growth back into line with the prediction of a simple model using productivity, inflation expectations and a measure of slack in the economy, having been above the prediction in mid-2019.

3.4 What are the implications for the MPC’s forecast?

The softening in labour demand is consistent with developments elsewhere in the economy.

The softening in labour demand appears to be consistent with the wider slowdown in the UK economy over the past year. The slowing in employment growth over 2019 is about the right size to bring the levels of output and employment into line with their usual long-run relationship, given assumptions about trend productivity. That, along with the rest of the analysis in this section, suggests developments in the labour market are not signalling that a prolonged slowdown is imminent.

The unemployment rate is projected to fall a little further.

In the MPC’s central projection, GDP growth recovers over the forecast period and unemployment falls a little further (Chart 3.15) (Section 1). Pay growth falls slightly over the coming year, partly reflecting the continuing unwind of temporary factors. However, it picks up over the latter part of the forecast period as unemployment falls a little further below its equilibrium rate and productivity growth rises. Section 1 sets out the risks around these projections.

Chart 3.14 Compositional effects have pushed up pay growth

Estimates of the contribution of employment characteristics to four-quarter pay growth ^(a)

Sources: Labour Force Survey and Bank calculations.

(a) Estimates are shown relative to their averages over 1995–2010. The effect of individual and job characteristics are derived from a regression of these characteristics on levels of pay. The total compositional effect is obtained by combining these estimates with changes in the composition of the labour force.
(b) Other includes age, tenure, gender, region of residence, whether full-time or part-time, in permanent or temporary employment, and in public or private sector employment.

Chart 3.15 Unemployment falls a little further in the MPC’s central projection

Unemployment rate and four-quarter pay growth ^(a)

Sources: ONS and Bank calculations.

(a) See Table 1.C for more information about the indicative projections consistent with the MPC’s forecast.