Economic review: May 2015

• The economy grew by 0.3% in the first quarter of 2015, a little more slowly than on average over the past two years, largely due to the services sector where quarterly growth eased to 0.5%.

• Unemployment has fallen substantially over the past 18 months, but the labour market’s response to the economic cycle has also been channelled through both changes in hours worked and in hourly pay.

• The decline in inflation to zero reflects not only falling prices for items such as food and energy but also a sharp drop in the proportion of the Consumer Price Index (CPI) exhibiting moderate and high price increases.

• Growth in unit labour costs has continued its long term decline despite the weakness in productivity growth. The weakness in productivity has added to unit labour cost growth since 2007, especially in financial services and in mining & quarrying.

• Households’ net interest payments (payments net of receipts) have edged up slightly as a share of disposable incomes since 2013, in part reflecting lower rates of interest available on savings deposits.

• The UK current account deficit was the largest among the G7 economies in 2014 as a proportion of Gross Domestic Product (GDP). A deteriorating primary income balance is a key factor behind this development.

The preliminary estimate of Gross Domestic Product (GDP) indicated that growth in the UK economy slowed from 0.6% in Q4 2014 to 0.3% in Q1 2015, and from 3.0% to 2.4% compared to the same quarter a year earlier. This edition of the Economic Review finds that the monthly profile of services industry growth reveals fluctuations masked by the quarterly data, notably in business services and finance.

A feature of the economic recovery in the UK has been the resilience shown by the labour market as measured by the rate of unemployment. This Review also notes that the labour market has responded through changes in weekly earnings growth, resulting from variation in both hours worked and hourly pay.

Despite the economic recovery, inflation remained at zero in March. While attention has focussed on those goods and services in the Consumer Price Index (CPI) basket that have seen price falls, this Review highlights the contribution arising from a smaller proportion of the CPI exhibiting moderate and high price rises.

Low inflation coupled with positive rates of growth in output and employment potentially offer conflicting signals to policy makers about the extent of underlying spare capacity. As a result, the behaviour of unit labour costs, and the impact of weakness in productivity growth, has also received attention from commentators. This Review notes that key contributions to productivity weakness have come from financial services and mining & quarrying.

The Bank of England’s Base Rate has remained unchanged since 2009, assisting both the economic recovery and the financial position of some households. However, this Review finds that interest received on savings by households has fallen since 2013, in contrast to the relative stability of interest paid on debt over the same period.

Finally this Review compares movements in the current account balance for the UK and other major economies, and finds that the deterioration in the UK’s external position contrasts with some improvements elsewhere internationally. A deteriorating primary income balance is a key factor behind this development in the UK.

The preliminary estimate of Gross Domestic Product (GDP) indicated that the UK economy grew by 0.3% in the first quarter of 2015, slower than the 0.6% growth recorded in both the third and fourth quarters of 2014. Compared to the same quarter of the previous year, GDP growth also slowed from 3.0% to 2.4%. However this extends a run of 8 consecutive quarters of positive quarterly growth during 2013 and 2014, such that GDP has risen by 10.7% compared to the trough of the economic downturn in Q2 2009, and is now 4.0% higher than the pre-downturn level of output in Q1 2008.

The lower preliminary estimate reflected falling output in the production and construction industries, as well as slower output growth in the services industries. Among the latter, slowing growth was most marked within the ‘business services & finance’ industries, where output growth fell to 0.1%, the weakest since Q4 2010, down from 1.3% in the previous period. This fall was driven by a broad range of sub-industries including a 1.5% fall in ‘architectural, engineering & technical testing’ services that have historically been a strong contributor to GDP growth, as well as a 0.6% fall in ‘financial & insurance’ services.

However, the latest quarterly growth figures mask fluctuations in the underlying monthly data for services sector output. To show this, Figure 1 plots the monthly change in total services output, along with the contributions made from major sub-industries.

Figure 1 shows that total services output fell by 0.2% in January, following relatively robust growth of 0.6% in December 2014. This fall was driven by a broad decline in the ‘business services & finance’ industries, with the largest contribution coming from a 1.4% fall in ‘professional services’ (such as architectural, management consultancy and legal services). However, February saw a return to growth in services output at a similar rate to that seen in much of 2014. This was partly due to a recovery in the same ‘professional services’ sub-industry, but it was also driven by rising output in ‘other services’ (including the activities of membership organisations, the repair of computers and personal and household goods).

Figure 1: Contributions to total services output growth from headline sub-industry groupings

(month on month, percentage points)

Source: Office for National Statistics

Notes:

1. Contributions may not sum to the total due to rounding

Looking over 2014 as a whole, services output grew in 11 of the 12 months, and by 3.7% in the 12 months to December 2014. The output of the production and construction industries has been more volatile. Figure 2 compares the path of output across production, construction and services from 2007. This shows that output in the services industry has risen consistently since 2009, with the level of output in Q1 2015 growing by 3.1% compared to the previous year and 5.7% compared to Q1 2013. In contrast, construction output has fallen for two consecutive quarters, while the level of monthly production output has remained broadly stable over the last year.

The relative strength of the labour market has been one of the most interesting characteristics of the UK’s recent economic performance. Higher unemployment accompanied the sharp fall in GDP in 2008 and 2009, but the rise in the number of people seeking work was less than might have been expected given historical experience. The output contractions which began in 1979 and 1990 of 5.6% and 2.2% respectively, resulted in unemployment rates of more than 10%, yet by contrast the 6.0% fall in GDP between Q1 2008 and Q2 2009 was accompanied by an unemployment rate of just over 8%. The fall in the unemployment rate since the start of the economic recovery has also been relatively sharp (Figure 3). Over the past eighteen months, the unemployment rate has declined almost 2 percentage points to 5.6% in the three months to February 2015 – among the fastest reductions in the past forty years – and is approaching its 2001 to 2007 average.

Both of these recent features of the UK’s economic performance are partly a consequence of flexibility in the labour market, which operates through several mechanisms. First, following an adverse shock, firms may seek to limit their costs of labour to remain in business by reducing the size of their workforce. Secondly, firms may reduce the number of hours worked – either by increasing the number of part-time workers that they employ, or by reducing the number of hours that existing employees work. Finally, if firms cannot reduce the quantity of labour they employ, they may seek to reduce the price of that labour by reducing wage growth. Each of these three channels appears to have played a role – but the latter two mechanisms may have helped to limit the extent to which firms responded to the economic downturn by reducing the size of their workforces.

While some of the labour market adjustment was mediated through the unemployment rate (Figure 3), there was also an adjustment to the number of hours worked. Figure 4 shows the first of two ways by which this was achieved: by increasing the share of employment accounted for by part-time work. This rose relatively rapidly from around 25% to 27% between 2008 and 2010, and has remained at this higher level over the last four years – only gradually drifting downwards since mid-2012. The share of employment accounted for by full-time workers experienced an equal and opposite move: easing from around three quarters of employment prior to the downturn, to just over 72% by the summer of 2012.

Alongside these changes in the share of employment accounted for by part-time workers, there were also relatively large changes in average hours worked within the full- and part-time categories (Figure 5). Average full-time hours fell from around 37.2 hours per week in the two years preceding the economic downturn, to around 36.6 hours per week at its lowest point in 2009. Average part-time hours also fell – albeit to a lesser extent – from around 15.6 to around 15.3 hours per week during 2009. Both the growth of part-time employment and the changes in average hours worked within full- and part-time statuses represent a substantial adjustment to the economic downturn.

Figure 5: Average weekly hours of work (Full-time, part-time)

Office for National Statistics

Taken together, Figures 4 and 5 suggest that this adjustment has started to unwind since the onset of the economic recovery. The full-time share of employment has since edged back up to around 73%, although it remains below its pre-2007 average. Average full-time hours have also risen from their low-point in 2009, reversing a long-term downwards trajectory, while part-time hours are at historic highs. These trends suggest that firms have responded to the economic recovery by increasing the volume of hours worked for current employees, and by starting to shift the ratio of full- to part-time workers back towards its historical average.

The third mechanism by which firms can adjust their costs of labour – through changes in the price of labour – is traditionally considered to be a more limited channel. Workers are thought to be unwilling to accept nominal pay cuts in response to economic conditions, which in turn limits the ability of firms to reduce their costs in response to a negative shock. However, variable elements of pay – including overtime, shift pay, bonuses and other performance related payments – can play an important role in these circumstances. Figure 6 analyses how changes in hours, basic pay and these elements of variable pay contributed to earnings growth between 2013 and 2014. It divides those who were in continuous, full-time employment (although not necessarily in the same job) over this period into 20 quantiles – each containing 5% of the workers – based on their earnings growth. Workers with the lowest (highest) earnings growth are groups on the left (right), and the circle marks average earnings growth in each quantile. The bars show the contributions of changes in hours (holding hourly pay constant), basic pay and variable pay (holding hours constant) to average growth in each quantile.

Figure 6: Contributions to mean weekly earnings growth in 2014 for those employees in continuous full-time employment, by earnings growth quantile (%)

Source: ASHE microdata, OCEA calculations

Notes:

Notes:

1. This chart uses individual level data from ASHE to calculate the growth of nominal weekly earnings for individuals observed in 2013 and 2014. Weighted summary statistics for the distribution of nominal earnings growth rates are calculated using their population weights in 2014, as longitudinal weights are not available for the ASHE survey at present. ONS is examining the feasibility of producing a set of longitudinal weights for ASHE that would permit more thorough analysis. Note that the ASHE methodology is not specifically designed to model earnings growth for individuals over time. The mean presented here measures the mean earnings growth rate for all employees in the relevant quantile.
2. Quantiles 6 and 7 contain employees who experience no change in weekly earnings. As more than 5% of workers experience no change in weekly earnings over this time period, these quantiles are grouped together

Figure 6 replicates the findings of previous analysis which indicated that average weekly earnings growth varies broadly, ranging from close to 20% for the 19th quantile to -14% for the second quantile. It also indicates that the sources of earnings growth vary across the distribution. At the higher end, where earnings are growing strongly, all three elements make a contribution to average earnings growth. Although longer hours of work and stronger variable pay together account for a large fraction of the growth of earnings for the 19th quantile, higher basic pay – perhaps as a result of promotions and job changes – accounts for more than half of the growth in earnings for this group.

At the lower end of the earnings growth distribution, by contrast, basic pay made a much smaller contribution. Reduced hours worked and lower variable pay accounted for 12.3% of the 13.9% fall in earnings for the second quantile. Basic pay – which is thought to be relatively sticky downwards – contributed 1.3 percentage points to the fall in earnings growth for this group – the only quantile for which basic pay made a negative contribution. Taken together, these results appear to provide evidence that while firms find it difficult to adjust basic pay downwards, flexibility in hours of work and variable pay may enhance the ability of the labour market to respond to changes in economic activity, by enabling employers to reduce costs while limiting the extent to which they may have to lay off workers.

Do these trends differ for part time workers, and how do they influence the earnings growth distribution? Figure 7 combines the recent trend towards more full-time work and longer hours in an analysis of how the volume and price of labour adjusted between 2013 and 2014. It presents a similar picture to Figure 6, but includes both full- and part-time employees. The blue bars measure the contribution to average earnings growth from full-time employees, the red bars measure the contribution from part-time employees, and the purple bars measure the contribution of those who move between full- and part-time employment. The contributions of the first two groups are further subdivided into the contribution from changes in hours worked (holding hourly earnings constant) and the change in hourly earnings rates (holding hours worked constant).

Figure 7: Contributions to mean weekly earnings growth in 2014 for those employees in continuous employment, by earnings growth quantile (%)

Source: ASHE microdata, OCEA calculations

Notes:

1. This chart uses individual level data from ASHE to calculate the growth of nominal weekly earnings for individuals observed in 2013 and 2014. Weighted summary statistics for the distribution of nominal earnings growth rates are calculated using their population weights in 2014, as longitudinal weights are not available for the ASHE survey at present. ONS is examining the feasibility of producing a set of longitudinal weights for ASHE that would permit more thorough analysis. Note that the ASHE methodology is not specifically designed to model earnings growth for individuals over time. The mean presented here measures the mean earnings growth rate for all employees in the relevant quantile.
2. Quantiles 5 to 7 contain employees who experience no change in weekly earnings. As more than 5% of workers experience no change in weekly earnings over this time period, these quantiles are grouped together.

Figure 7 indicates that changes in weekly earnings for full-time employees are more likely to be influenced by changes in hourly earnings rates (dark blue bars) than hours worked as they have less scope to increase their working hours than part-time employees. Conversely, for part-time workers, variations in weekly earnings are driven more by hours worked than hourly pay, and this comes increasingly into play towards either end of the distribution where the biggest adjustments in earnings – up or down - take place. Together with the increased prominence of those who moved between full- and part-time employment at the ends of the distributions, this indicates that changes to hours worked are a key aspect of labour market flexibility.

Despite continuing growth in the economy, the annual rate of consumer price inflation remained at a record low of 0.0% in March 2015. To assess the factors that have been driving low inflation, Figure 8 breaks down the annual rate of CPI inflation since 2003 into the contributions from six broad expenditure categories (detailed in the notes to Figure 8). This highlights how the recent drop in inflation has been led by falling prices for food and energy.

Table 1 shows that in March 2015, falling prices in the electricity, gas & fuel and food & drink categories reduced the annual rate of CPI inflation by 0.8 percentage points. The same two categories alone accounted for almost two thirds of the fall in inflation from its peak of 5.2% in September 2011 to zero in March 2015, as well as the drop from 2.9% in June 2013.

Core inflation – a measure that attempts to capture underlying inflationary pressures in the economy by excluding the more volatile energy, food, alcoholic beverages and tobacco components – has therefore fallen less sharply. While the headline rate of inflation and the core measure both stood at 1.6% in March 2014, they now stand at 0.0% and 1.0% respectively. This highlights the role that these more erratic components have played in pulling down inflation in recent months: excluding these components from the overall price index would cause inflation to remain within one percentage point of the Bank of England’s 2% medium term target.

Although the headline rate of inflation is a good summary of price pressures across the economy, it can mask a wide variation in the inflation rates of individual goods and services. For this reason, information regarding the range or distribution of price pressures can help understanding of the drivers of inflation.

To demonstrate this, Figure 9 shows the distribution of inflation rates experienced by the 85 class level components of the CPI’s basket of goods and services, on a monthly basis from January 2001. Red shaded areas toward the top of the chart show the proportion of the CPI that experienced price growth (inflation) greater than 0.5%, with darker shades of red indicating higher rates of inflation. The blue shaded areas toward the bottom of the chart show the proportion of the CPI that experienced price falls of more than 0.5%, with deeper blues indicating a greater rate of price fall. Grey shaded areas indicate price change of between -0.5% and 0.5%.

Figure 9: Proportion of the CPI basket of goods and services experiencing price falls or price rises, %.

Source: Office for National Statistics

Figure 9 shows that the underlying distribution of price inflation was relatively stable prior to the recent economic downturn but has since then become more volatile. Changes in the rate of inflation since 2011 have been driven by a combination of two factors: a marked increase in the proportion of the CPI exhibiting price falls of greater than 0.5%, and a sharp reduction in the proportion exhibiting substantial price increases. For example, the proportion of the CPI showing annual price inflation of greater than 2.5% fell to just over one quarter in the year to March 2015 from over three quarters in the year to September 2011.

The recent fall in the inflation rate to 0.0% cannot therefore be wholly attributed to an increasing proportion of the CPI exhibiting price falls, but also reflects the decreasing share that is experiencing significant price rises.

Information on developments in productivity and unit labour costs (ULC) can provide an insight into underlying inflationary pressures in the economy. Faster productivity growth could indicate that the economy can continue to grow without bidding up the price of factor inputs, while the absence of a recovery in output per hour may indicate the opposite.

To assess how these data interact with respect to labour, ONS publishes estimates of ULC alongside those of labour productivity. ULC conceptually measure the change in the total costs of labour incurred in the production of each unit of output. They can be decomposed into two core elements: changes in labour costs (including wages and non-wage benefits such as social security and pension contributions) per unit of labour input, and productivity growth. All else equal, rising labour costs act to increase ULC. However higher levels of productivity act to offset this effect, as businesses are able to produce more output with a given unit of labour input.

To assess how ULC have changed over time, Figure 10 plots the annual growth rate of ULC, divided into the respective contributions from the growth of hourly labour costs and labour productivity (output per hour). Negative green bars represent positive productivity growth that has acted to reduce ULC growth.

There has been a general decline in the growth of ULC over the past three decades, falling from an average annual rate of 5.6% over the 1980s to below 5% for much of the period since the early 1990s, and in the 0% to 2% range in the most recent years. This reflects in particular the decreasing rate of average earnings growth over time. In the period since 2008, this has been augmented by the weakness in productivity.

The influence of productivity on ULC since 2007 varies between industries, as shown in Figure 11. The mining & quarrying and utility industries have suffered from a large fall in productivity, partly due to falling North Sea oil production over the period. As a result, these sectors have seen big increases in ULC. Similarly, weak productivity growth in financial services has led to increasing ULC despite a slight fall in the hourly cost of labour for the industry. By contrast, strong productivity growth in professional, scientific and administrative services industries has led to falling ULC. The effects of productivity are more muted in other industries, where increasing labour costs have largely dictated the course of ULC.

Figure 11: Change in unit labour costs between 2007 and 2014, and contributions from the change in output per hour and hourly labour costs (%)

Source: Office for National Statistics

Since 2008, households have benefited from sharp cuts in the cost of borrowing and continuing historically low interest rates. More recently, improvements in labour market conditions, as well as the aforementioned fall in inflation, are also likely to have assisted the financial position of some households. However household finances are also influenced by levels of household debt and assets via flows of interest payments and receipts. Information on this can help understand the relative vulnerability of households to future tightening of monetary policy.

Figure 12 shows net interest payments, as a proportion of gross household disposable income (GHDI), calculated as the total interest paid by the household sector less the total interest received. While the difference between interest paid and received may differ widely between households, this gives an insight into the potential exposure to rising interest rates faced by the sector. Net interest payments dropped substantially following the downturn, but have edged up a little since 2013, as households’ receipts of interest weakened by more than their payments.

Although the Bank of England Base Rate has remained unchanged since 2009, Figure 13 shows how the interest rates on various deposits (solid lines) and loans (dotted lines) have moved since 2012. While the rate of interest received on savings from deposits with financial institutions has declined since the beginning of 2013, the rate paid by households on loans has remained relatively stable. Savings rates in some cases have fallen by over a third since 2013.

The current account of the UK’s balance of payments improved slightly in the final quarter of 2014, the deficit falling from £27.7 billion (6.1% of GDP) in Q3 2014 to £25.3 billion (5.6% of GDP) in Q4 2014. However looking over a broader time period shows a general deterioration in the current account; the deficit over the 2014 calendar year as a whole was £97.9 billion (5.5% of GDP), which was the largest figure since comparable records began.

The change in the most recent quarter was driven by an improvement in the trade balance, which reflected a £4.1bn rise in exports and a £0.2bn fall in imports. Previous analysis has shown that the deterioration in the current account has instead been driven by a large deficit in the primary income balance (mainly income earned by UK residents from investments overseas, less income earned by non-residents on their UK investments). In Q4 2014 the primary income deficit improved slightly, to £11.7 billion from £12.2 billion in the previous quarter. But the 2014 calendar year figure of £38.8 billion also represented the largest deficit on record.

There have been striking divergences in the relative performance of current accounts among the major developed economies. Figure 14 compares the current account balance as a proportion of GDP in the G7 economies (US, Japan, UK, Germany, France, Italy and Canada) for the two most recent calendar years with that in 2007. This highlights how the external position of major economies has evolved following the economic downturn in 2008 and 2009.

Figure 14 shows that four of the seven economies (including the UK) have experienced deteriorating current account balance relative to 2007; however the UK recorded the largest current account deficit among these economies in 2014 at 5.5% of GDP. Along with Japan, this also represented a worsening position relative to 2013. In contrast, Germany experienced the largest current account surplus in 2014, which was also a small improvement on the surpluses recorded in 2007 and 2013. Similarly, the US experienced an improvement on its current account balance from a deficit of 5.0% in 2007 to a deficit of 2.4% of GDP in 2014.

Figure 14: Current account balance in the G7 economies (% of nominal GDP)

Source: ONS, Eurostat, US Bureau of Economic Analysis, Japan Cabinet Office, Statistics Canada, INSEE

Notes:

1. Data for France was sourced from Eurostat for 2007 and INSEE for both 2013 and 2014

Figure 15 shows the contributions of the trade, primary income and secondary income balances to the current account balance in 2007 and 2014. The economies that have experienced the largest improvement in their current account balance since 2007 have achieved this through improvements to their trade balances, notably Italy and Germany. Along with Canada and Italy, the UK experienced a negative contribution from the primary income balance, but was the only economy where the deterioration in the current account was driven by primary income. The underlying causes of this have been examined in previous analysis, and include the relatively strong performance of the UK economy (in particular relative to the euro area, where a large fraction of the UK’s overseas assets are based).

Figure 15: Contributions to the current account balance in the G7 economies (% of nominal GDP, 2007 and 2014)

Source: ONS, Eurostat, US Bureau of Economic Analysis, Japan Cabinet Office, Statistics Canada, INSEE

Notes:

1. Data for France was sourced from Eurostat for 2007 and INSEE for 2014