Economic review: June 2015

• UK economic growth eased to 0.3% in the first quarter (January to March) of 2015, reflecting slower growth in household spending and a sharp fall in the contribution of net trade.

• The proportion of students who were successful in finding work fell markedly in the downturn, and those who were successful took longer to find a suitable match: both these trends have started to reverse in recent quarters.

• Firms are increasingly hiring new workers from other firms, in a trend which is closely related to recruitment difficulties. One measure of potential labour supply from among the unemployed also appears to be closely related to movements in industry earnings growth.

• The Consumer Prices Index inflation rate fell to -0.1% in the year to April, while a marked weakening in “core goods” inflation has contributed to the recent decline of core inflation.

• Investment has recovered relatively strongly since the downturn but its composition has changed in a manner which suggests large capital outlays on buildings were the first to be cut back and have been the slowest to return.

• While total hours worked have increased strongly, growth in the quality of labour input has waned since 2012, turning negative in 2014, contributing to weak productivity growth. This may reflect a changing composition of the workforce towards lower-skilled roles.

The second estimate of Gross Domestic Product (GDP) confirmed that quarterly output growth in the UK economy moderated from 0.6% in Q4 2014 to 0.3% in Q1 2015. Compared to the same period a year earlier, GDP growth slowed from 3.0% in Q4 2014 to 2.4% in the first quarter of 2015.

Despite the recent economic recovery, the Consumer Prices Index (CPI) annual inflation rate fell to -0.1% in April – the lowest rate on record. A measure of “core inflation” – which excludes the volatile prices of food, drink and energy – has also fallen in recent months. This edition of the Review examines the factors which have contributed to this moderation. While there has been some slowdown in inflation for “core services”, much of this recent movement is as a result of lower price pressure from a set of “core goods”.

The fall in the rate of inflation notwithstanding, this Review finds little evidence as yet that the fall in the oil price has either led to households deferring large amounts of expenditure in anticipation of future price falls, or increasing their total spending as a result of an “oil dividend”. Expenditure on durables – the timing of which households may have greater discretion over – made a broadly similar contribution to quarterly spending throughout 2014. Aggregate household expenditure growth moderated in the first quarter of 2015.

Following a brief fall at the end of the year, investment by firms and government grew by 1.5% in Q1 2015, returning the level of Gross Fixed Capital Formation (GFCF) to broadly its pre-downturn level. While investment has recovered to its broad 2007 level, this Review finds that its composition has changed in a manner consistent with firms gradually returning to large capital outlays after cutting back at the start of the downturn.

While weak investment by firms has been cited as one possible explanation for recent low productivity growth, both the strength of employment growth and the composition of that growth may also have played a role. This Review uses recent ONS analysis to consider how improvements in the quality of labour input have moderated in recent quarters as the portion of the workforce defined as low-skilled has edged up, possibly reflecting prior weakness in investment immediately following the crisis.

Finally, this Review examines signals of a tightening labour market and finds that companies are increasingly recruiting labour from other firms as well as from the unemployment pool. The proportion of new hires which are drawn from other firms is closely related to a Bank of England measure of recruitment difficulties, while one measure of potential labour supply from among the unemployed also appears to be closely related to movements in industry earnings growth. Both trends indicate a tightening labour market and the potential for rising wage pressures. This Review also finds that the proportion of students who were successful in finding work fell markedly in the downturn, and that those who were successful took longer to find a suitable match: both these trends have started to reverse in recent quarters.

The second estimate of Gross Domestic Product (GDP) indicated that the UK economy grew by 0.3% in the first quarter of 2015, unrevised from the preliminary estimate. Following output growth of 0.6% in each of the final two quarters of 2014 (Figure 1), the pace of GDP growth moderated slightly during the first three months of 2015. Compared to the same period a year earlier, output growth is estimated to have slowed from 3.0% in Q4 2014 to 2.4% in Q1 2015. However, this moderation notwithstanding, Q1 2015 was the ninth in a run of quarters of continuous GDP growth and aggregate output is now estimated to be 10.7% above its Q2 2009 trough.

The moderation of GDP growth in Q1 2015 reflected slower output growth in the production and services industries, as well as further falls in construction output. Production output grew by just 0.1% in the first quarter of 2015, while construction output fell by 1.1% over the same period. Although the outturn for both these industries was stronger than estimated at the time of the preliminary estimate, their performance was offset by weaker than previously estimated growth in the services industries, leaving the estimate of GDP growth unchanged. Aggregate services output growth fell from 0.9% at the end of 2014 to 0.4% in Q1 2015, reflecting a weaker performance of the business services and finance industries in particular.

The second estimate of GDP also contained the first information on the expenditure components of GDP growth in Q1 2015. Investment and government consumption added 0.7 and 0.1 percentage points to aggregate expenditure in Q1 2015 respectively, while strong growth in imports resulted in net trade making a negative contribution to GDP growth of 0.9 percentage points in the first quarter – offsetting a positive contribution of 0.8 percentage points in Q4 2014. Aggregate household consumption – which has been one of the key drivers of the recovery – grew by 0.5% on the quarter.

Despite a marked weakening of inflationary pressure over the last two quarters, any impact on household spending appears to have been limited. On the one hand, the fall in the rate of consumer prices index (CPI) inflation to -0.1% in April – partly as a consequence of the fall in the price of oil – has strengthened the purchasing power of households. This in turn has led to expectations of stronger spending from this ‘oil dividend’. On the other hand, if households anticipate further falls in prices, some have argued that they may choose to postpone some spending – resulting in the opposite effect. In this latter scenario, durable goods expenditure (including on items such as furniture and vehicles, which is at households’ discretion) would grow more slowly, while spending on non-durable goods (including on food and fuels, which is more difficult to postpone) would be less affected.

However, household spending data so far provides little evidence of either a slowdown in durable goods purchases, or of a substantial acceleration of household consumption (Figure 2). Consumer spending on durables and semi-durables continued to grow at broadly the same pace as in previous quarters in Q4 2014 – accounting for almost half of the total growth in household spending – showing no sign that households are deferring large purchases. Aggregate household spending growth also moderated in Q1 2015 – providing little evidence that lower price pressure in the first quarter led to stronger household consumption. Spending on non-durables remained lower than it was a year ago in Q4 2014, consistent with the relative weakness of aggregate household consumption per head.

The retail sales index also shows relatively little sign as yet of an ‘oil dividend’. While the volume of retail sales was 4.7% higher in the three months to April 2015 than in the same period a year earlier, retail sales growth has been relatively strong since the start of 2014 – a period before the oil price started to fall (Figure 3). While the lower oil price may have supported the growth of retail sales in recent months, there is little definitive evidence as yet that a sharp fall in the price of petrol in particular has prompted stronger aggregate household spending – perhaps indicating that these effects may work through with a substantial lag.

Figure 3: Retail Sales growth, 3 months on same 3 months a year ago, chained-volume measure, seasonally adjusted, % and change in the US Dollar price of Brent crude, %

Source: Office for National Statistics

Despite the recent economic recovery, the rate of Consumer Prices Index (CPI) inflation fell to a record low of -0.1% in April 2015. This is the first time that annual inflation as measured by the official CPI has been negative, and the first time that comparable estimates of consumer prices have fallen since March 1960. Annual CPI inflation fell from 0.3% in the year to January 2015, to zero in February and March, before slipping to -0.1% in April. While much of the recent weakness has been as a consequence of recent falls in the oil price, the recent appreciation of Sterling against the Euro and price competition among retailers, the fall on the month was at least partly driven by air and sea fares related to the timing of Easter.

However, while a range of temporary factors appear to be pushing down headline inflation, “core inflation” – a measure that attempts to capture underlying inflationary pressure in the economy by excluding the more volatile energy, food, alcoholic beverages and tobacco components – has also weakened in recent months (Figure 4). While this measure of inflation has been more stable than the headline CPI, it has moved gently downwards since the end of 2011. Core inflation fell from 1.4% in the year to January 2015, to 0.8% in April: its lowest level since March 2001.

The recent weakness of core inflation is partly explained by falls in the prices of “core goods”, which have only been partly offset by a relatively stable set of “core services” (Figure 5). This set of services – comprising housing, catering and education services – have together added 0.8 percentage points to core inflation in every month since October 2013. This recent consistency is a long-term feature: since 2003, their combined monthly contribution to core inflation has varied between 0.6 and 1.3 percentage points. Transport & package holidays and recreation & cultural services prices also make relatively large contributions and account for both some of the fall in inflation relative to its peak in 2011, and for some of the weakness in April 2015 in particular.

However, the largest contribution to the recent fall in core inflation is from the “other” category, reflecting the impact of core goods prices in particular. The contribution of this other group has varied much more broadly than the other components over the last decade and accounts for much of the variation in core inflation. It also accounts for much of the recent weakness of price pressure, as its contribution has fallen from 1.5 percentage points in April 2011, to 0.7 percentage points in April 2014, to just 0.2 percentage points in the year to April 2015. Within this other group are a large number of products which have experienced moderated price pressure in recent months. Core goods – including clothing, footwear, cars and furniture & furnishings – account for around 56% of this group, and account for a large fraction of the recent reduction in inflationary pressure; other core services inflation has fallen in recent months, but to a lesser extent.

Differences in the rates of consumer goods and services inflation are a persistent feature of the CPI, and are partly reflected in the evolution of producer price series – the prices that producers receive for their products from other firms. While the headline CPI measure fell to -0.1% in April 2015, the prices of goods included in the CPI fell by 2.0%, and the prices of consumer services grew by 2.0%. This divergence is particularly marked, and is consistent with trends in the producer prices series (Figure 6). Manufacturing firms have reported receiving lower prices for their output since early 2014 – dropping by 4.8% in the year to Q1 2015. Services producer prices – which are close to recent lows – have been more resilient over recent periods, a difference which likely indicates the degree of supply chain price pressure in the two industries, and in particular which captures the relative importance of labour, goods and services in their input mix.

Figure 6: Gross Services and Manufacturing Producer Price Indices, annual change, %

Source: Office for National Statistics

Recent movements in exchange rates can explain some of this weakness in price pressure, but are likely to affect different goods and services prices to differing extents. Movements in exchange rates affect both producer and consumer prices, as they affect the price of both direct imports and products that use imported components. If sterling appreciates (depreciates) against a foreign currency, the price of imports from that country will fall (rise), with a corresponding impact on domestic inflation.

However, over the past twelve months, the pound has moved in different directions against the currencies of its two major trading partners: the Euro-area and the US. Between April 2014 and April 2015, the pound appreciated by 14% against the euro and depreciated by 11% against the US dollar. Partly as a consequence of these exchange rate movements, the prices of imports from these different regions have moved in different directions. In particular, the prices of goods imported by producers from the European Union (EU) fell in price over the year to April 2015, while the prices of goods from outside the EU – among whom the US is the largest player – have been more resilient.

These changes in relative prices notwithstanding, net trade was a drag on aggregate GDP growth in the first quarter of 2015, only partly offset by a slight rise in investment growth over the same period. Gross Fixed Capital Formation (GFCF) grew by 1.5% in the first quarter of 2015 – up from a contraction of 0.6% in the final three months of 2014 – while business investment grew by 1.7% – up from a contraction of 0.9% in the previous three months. Partly as a consequence, the level of GFCF has now returned to its pre-downturn level (Figure 7), and has grown at a compound average quarterly rate of 1.1% since the trough of the economic downturn in Q2 2009. This compares favourably with the average of 0.7% per quarter in the 1997-2008 period, and with the recoveries from previous economic downturns. Over a comparable period following the contractions in the early 1980s and 1990s, aggregate GFCF is estimated to have grown by 1.2% per quarter and 0.1% per quarter respectively.

Figure 7: Gross Fixed Capital Formation (GFCF), chained-volume measure, seasonally adjusted, £ billion

Source: Office for National Statistics

While investment has recovered to broadly its pre-downturn level, the composition of that investment has changed. Figure 8 shows that the recent recovery in investment has been driven by a combination of a recovering investment in dwellings and other machinery & equipment – which accounted for a large fraction of the fall in GFCF following the economic downturn – and intellectual property products (IPP). This trend appears to suggest that following the onset of the economic downturn, large capital outlays on dwellings, buildings and structures and other machinery & equipment were cut back first, while firms continued to invest in transport equipment and IPP. Since the onset of the recovery, this pattern appears to have started to unwind: higher investment in transport equipment and IPP has been followed by a recovery in machinery & equipment, dwellings and, to a lesser extent, in buildings investment. This is consistent with firms waiting for clear signs of a recovery before restarting large capital projects.

The reluctance to increase capital investment immediately following the downturn has been cited as one cause of the recent weakness in productivity growth, as have changes in the composition of employment. Capital deepening – the amount of capital available per unit of labour input – which contributed on average 0.6% per year to labour productivity growth prior to the downturn – reduced labour productivity growth by 0.1% in 2012 and made no contribution in 2013. However, changes in the composition of the workforce may also have been important. Recent ONS analysis has shown that the fraction of ‘low-skilled’ employee jobs in the economy rose to 25.9% in 2014 from 25.5% in the previous year, while the share of ‘high-skilled’ jobs fell by 0.4 percentage points to 44.7%¹. Recent Bank of England analysis also indicates that employment growth has been more concentrated among low-skilled occupations.

These analyses are reflected in new estimates of Quality Adjusted Labour Input (QALI). This measure of labour input differs from the traditional measure of total hours worked (“unadjusted hours”), as it weights hours worked by different types of workers according to their share of total labour income. For example, if hours worked by a highly skilled and highly remunerated type of labour (such as brain surgeons) increased, then the volume of labour input as measured by QALI would rise by more than the observed increase in hours. The difference between the growth rates of QALI and total unadjusted hours therefore captures changes in the skills composition of the workforce or the ‘quality of labour’, that should to some extent affect growth of labour productivity (Figure 9).

Notes for investment

1. Low-skilled occupations comprise elementary occupations, sales & customer services operators and process, plant & machine operatives. High-skilled occupations comprise managers, directors & senior officials, professional occupations and associate professional & technical occupations.

The marked rise in the quantity of labour input has come about partly as a result of a sharp fall in the unemployment rate, which has been on a downwards trend since the beginning of 2013 and fell to 5.5% in the 3 months to March 2015. The fall in the unemployment rate – which has been particularly marked among young people – is one of a number of indicators pointing towards a tightening labour market for some time, raising the potential for wage pressure. As the unemployment rate approaches its pre-downturn average, strong demand for labour and a smaller number of unemployed workers may put upwards pressure on earnings. If employers cannot find available workers with appropriate skills, they may also seek to hire workers from other companies, which may involve higher wages being offered as an incentive to move. As a result, both the unemployment rate and the skills-mix of the unemployed can have important consequences for inflationary pressures in the labour market.

While objectively measuring an individual’s skills – unemployed or otherwise – is difficult, one approach involves examining the industry in which the unemployed previously worked. To the extent that some of the skills which these workers acquired in their previous jobs were industry-specific, existing firms may find these workers relatively more attractive than other workers, who might need more training and investment. A larger number of former construction workers in the unemployment pool, for example, may mean that construction firms could draw on a larger number of potential entrants who are already trained, reducing wage pressure in the industry. Figure 10 analyses the unemployment rate based on the industry in which these workers were formerly employed, relative to their long-run averages. Points above (below) the long run average indicate that workers from a given industry make a greater (lesser) contribution to the unemployment pool than they did over the 2002 to 2007 period.

Figure 10: Contributions to the change in the unemployment rate relative to its 2002 to 2007 average, 12 month moving average

Source: Labour Force Survey, cross-sectional datasets

Notes:

1. ‘Other’ contains those who report having previous work but do not state an industry, non-responses, as well as inconsistent responses (for example, those who report no previous job, but report their last job as being in construction).
2. Short (long) term unemployment is defined as durations of less (more) than two years.
3. Data in this analysis is not based on, and will not be comparable to, published LFS National Statistics on unemployment by previous industry. This analysis instead uses quarterly LFS micro-datasets in order to analyse those with no previous job.

This analysis suggests that former services workers and those who have never had a job accounted for a large fraction of the rise in the unemployment rate above its pre-downturn level. Former services workers accounted for much of the initial rise in the unemployment rate, while the proportion who had never worked before increased more gradually during 2010 and 2011. Workers from the construction and production industries also pushed up the unemployment rate in 2009 and 2010, but this effect has waned relatively quickly. In recent months, the fraction of the unemployed that were previously employed in these latter industries has fallen below its long run average, possibly presenting the potential for upwards wage pressure. This leaves two groups to account for the remaining degree of elevation in the unemployment rate: long-term unemployed services workers, and those who have not previously had a job.

The contribution to unemployment of those who have not previously had a job is primarily made up of young people and appears to be related to changes in the success of students in finding work. People aged 16 to 24 who have not previously had a job contributed 0.80 percentage points to the elevated unemployment rate in 2012, compared with a contribution of just 0.11 percentage points from those with no previous job from other age groups. Figure 11 presents the proportion of economically inactive students (aged 16 to 24) who enter the labour market and find work. It suggests that a smaller fraction of students found work following the economic downturn, and that it took longer on average for those who were successful to find a match. The transition rate between education and employment fell markedly between 2008 and 2009, reflecting the fall in the aggregate hiring rate and the unwillingness among firms to take on untested labour in a period of substantial economic uncertainty.

Figure 11: Proportion of students entering the labour market who find work (ages 16-24), by the length of transition into employment, 4 quarter moving average, %

Source: Labour Force Survey, longitudinal datasets

Notes:

1. Due to low data coverage during a change in survey questioning, these following quarters are imputed for each categories; Up to 3 months series (2005 Q2), Up to 6 months series (2005 Q3), Up to 12 months (2006 Q1);
2. Due to the five-quarter LFS being currently re-weighted, data for 2014 Q3 and beyond for these two categories are not displayed: Up to 6 months series and up to 12 months series.

However, in recent periods this transition rate has risen towards its pre-downturn average, suggesting that demand for labour is recovering and drawing in students at its fastest rate since 2008. The proportion of students who find work within three months is close to 30% – only slightly below its long-term average – and there has been a smaller rise in the fractions finding work at longer intervals. This effect also appears to be reducing the “stock” of inexperienced workers – ameliorating the accumulated impact of lower transition rates in recent periods (Figure 10). Taken together, these trends suggest stronger labour demand for students.

As well as drawing on former students, firms have increasingly drawn from the inactive and from other firms to source additional labour, with implications for future wage pressure. The proportion of new hires from other firms increased to around 54% of new recruits in the year to Q1 2015 – below its pre-downturn average, but the highest rate since early 2009 (Figure 12). This appears to reflect rising recruitment difficulties, suggesting employers may have turned to labour from other firms as the supply of experienced labour among the unemployed and inactive has gradually fallen.

Figure 12: Proportion of new hires from employment, recruitment difficulties, one year moving averages, 1998-2015

Source: Labour Force Survey, longitudinal datasets, Bank of England

Notes:

1. Prior to Q4 2001 the portion of new hires from the employed does not include women over 59.
2. ‘Recruitment difficulties’ represents the Bank of England’s Agents’ general assessment of recruitment difficulties across the economy.
3. Prior to 2005 ‘Skill shortages’ is used to represent recruitment difficulties.

The sharp fall in the unemployment rate, the rise in job to job flows and the uptick in recruitment difficulties all point to a tightening labour market, and suggest that the supply of suitably qualified, available workers has fallen over recent periods. Building on the approach above, Figure 13 presents the relationship between a measure of potential labour supply and the median growth of earnings by industry. The former measure expresses the number of the unemployed who formerly worked in an industry as a fraction of the total number of current and former, now unemployed, workers in that industry. The larger (smaller) this ratio, the larger (smaller) the potential supply of workers with relevant skills and experience on which an industry can draw.

Figure 13: Median growth of earnings and potential labour supply, selected services industries, %

Source: Labour Force Survey, longitudinal and cross-sectional datasets

Notes:

1. ‘Median growth in earnings’ refers to the median value of earnings growth, and is taken from LFS five-quarter longitudinal data.
2. ‘Potential labour supply’ measures the number of unemployed workers whose last job was in a particular industry as a proportion of the number of people currently and formerly employed in that industry.
3. Industries are given on a SIC92 basis.
4. ‘Other’ contains public administration and defence, education, health and social work, and other community, social & personal activities.
5. Please note that this chart/table was updated at 12:00 on 03-06-2015 to correct a minor labelling error. The data are unaffected, but references to potential labour supply have been corrected to median growth in earnings and vice versa.

The trend lines presented in Figure 13 suggest both a negative relationship between the median growth rate of earnings and potential labour supply and that this relationship differs between services sub-industries. In general, a lower potential supply of experienced labour for a particular industry appears to be associated with higher earnings growth for those currently employed. For industries that tend to have a relatively high rate of turnover – including hotels & catering, wholesale, retail & motor trades – this relationship appears to have shifted to the right, possibly indicating the higher degree of frictional unemployment associated with these industries. Taken together, these trends are consistent with a tightening labour market and the recent return to positive real earnings growth, as firms increasingly draw on labour in other firms, students and the inactive to meet their demand for labour.