A lack of productivity growth has been a major problem for the UK economy over the past decade (click here for a PowerPoint of the chart). Is it possible that the new decade may see a pick-up in the growth in output per hour worked?
One possible solution to low productivity growth is to reduce working hours and even to move to a four-day week, but not to reduce total pay. If people work fewer hours, they may well be more productive in the hours they do work. In fact, not only may output per hour increase, but so too may output per worker, despite fewer hours being worked. What is more, the quality of output may increase with people being less tired and more motivated.
Several companies have experimented with a four-day week, including Microsoft in Japan, which employees 2300 workers. It found that, despite a 20% reduction in hours worked, output per hour worked increased by 40%, with total output thereby increasing. Workers were generally happier and more motivated and asked for fewer days off.
And it is not just a question of output: fewer hours can result in lower costs. The effect on costs will depend on the nature of new work patterns, including whether everyone has the same extra day off.
But a four-day week is only one way of cutting working hours for full-time employees. Another is to reduce the length of the working day. The argument is that people may work more efficiently if the standard working day is cut from eight to, say, five hours. As the first Thrive Global article article (linked below) states:
Just because you’re at your desk for eight hours doesn’t mean you’re being productive. Even the best employees probably only accomplish two to three hours of actual work. The five-hour day is about managing human energy more efficiently by working in bursts over a shorter period.
If people have more leisure time, this could provide a boost to the leisure and other industries. According to a Henley Business School study:
An extra day off could have a knock-on effect for the wider society. We found 54% of employees said they would spend their day shopping, meaning a potential boost for the high street, 43% would go to the cinema or theatre and 39% would eat out at restaurants.
What is more, many people would be likely to use the extra time productively, undertaking training, volunteering or other socially useful activities. Also family life is likely to improve, with people spending less time at work and commuting and having more time for their partners, children, other relatives and friends. In addition, people’s physical and mental health is likely to improve as they achieve a better work-life balance.
So, should firms be encouraged to reduce hours for full-time workers with no loss of pay? Many firms may need no encouragement at all if they can see from the example of others that it is in their interests. But many firms may find it difficult, especially if their suppliers and/or customers are sticking with ‘normal’ working hours and want to do business during those hours. But, over time, as more firms move in this direction, so it will become increasingly in the interests of others to follow suit.
In the meantime, should the government introduce incentives (such as tax breaks) or regulations to limit the working week? Indeed, it was part of the Labour manifesto for the December 2019 election that the country should, over time, move to a four-day week. Although this was a long-term goal, it would probably have involved the use of some incentives to encourage employers to move in that direction or the gradual introduction of limits on the number of hours or days per week that people could work in a particular job. It is unlikely that the new Conservative government will introduce any specific measures, but would probably not want to discourage firms from reducing working hours, especially if it is accompanied by increased output per worker.
But despite the gains, there are some problems with reduced working hours. Many small businesses, such as shops, restaurants and firms offering technical support, may not have the flexibility to offer reduced hours, or may find it hard to increase productivity when there is a specific amount of work that needs doing, such as serving customers.
Another problem concerns businesses where the output of individuals is not easy to measure because they are part of a team. Reducing hours or the working week may not make such people work harder if they can ‘get way with it’. Not everyone is likely to be motivated by fewer hours to work harder.
Then there is the problem if reduced hours don’t work in boosting productivity. It may then be very difficult to reintroduce longer hours.
But, despite these problems, there are many firms where substantial gains in productivity could be made by restructuring work in a way that reduces hours worked. We may see more and more examples as the decade progresses.
- Economics of a four-day working week: research shows it can save businesses mone
The Conversation, Miriam Marra (11/11/19)
- Less hours for work, more time on Earth: Why a four-day working week is good for you
Independent, Steve Taylor (16/12/19)
- Microsoft Japan Launched A Four-Day Workweek To Much Success: Is This The Key To Attracting Talent In The Tight U.S. Job Market?
Forbes, Jack Kelly (5/11/19)
- Will The Five-Hour Work Day Catch On In America?
Forbes, Jack Kelly (28/10/19)
- My Company Implemented a 5-Hour Workday — and the Results Have Been Astounding
Thrive Global, Stephan Aarstol (3/10/19)
- Why Does a Four-Day Work Week Achieve Better Results?
Thrive Global, Stephanie Lin (16/12/19)
- Four-Day Working Week Improves Staff’s Mental Health By 87%, Company Finds
Unilad, Niamh Shackleton (10/12/19)
- Hull business launches four-day working week as well as your birthday off and ‘Beer Fridays’
Hull Live, Phil Winter (5/12/19)
- A four-day work week? Sounds nice, but here’s the real deal
Sydney Morning Herald, Tony Featherstone (21/11/19)
- The four-day debate: Fantasy or feasible?
The Hindu, Business Live, Kamal Karanth (20/11/19)
- Finland’s ministry of transport floats tech-enabled four-day week
Computer Weekly, Gerard O’Dwyer (20/11/19)
- Should You Consider a 4 Day Work Week?
Small Business Trends, Rob Starr (5/12/19)
- Distinguish between different ways of measuring labour productivity.
- Give some examples (from the linked references) of employers which have tried introducing a four-day week or reduced hours for full-time workers. What has been the outcome in each case?
- In what ways may reducing working hours reduce a firm’s total costs?
- What are the advantages and disadvantages of the government imposing (at some point in the future) a maximum working week or a four-day week?
- What types of firm might struggle in introducing a four-day week or a substantially reduced number of hours for full-time employees?
- What external benefits and costs might arise from a shorter working week?
It’s been a while since I last blogged about labour markets and, in particular, about the effect of automation on wages and employment. My most recent post on this topic was on the 14th of April 2018 and it was mostly a reflection on some interesting findings that had been reported by Acemoglu et al (2017). More specifically, Acemoglu and Restrepo (2017) developed a theoretical framework to evaluate the effect of AI on employment and wages. They concluded that the effect was negative and potentially sizeable (for a more detailed discussion see my blog).
Using a model in which robots compete against human labor in the production of different tasks, we show that robots may reduce employment and wages … According to our estimates, one more robot per thousand workers reduces the employment to population ratio by about 0.18–0.34 percentage points and wages by 0.25–0.5 percent.
Since then, I have seen a constant stream of news on my news feed about the development of ever more advanced industrial robots and artificial intelligence. And this was not because of some spooky coincidence (or worse). It has been merely a reflection of the speed at which technology has been progressing in this field.
There are now robots that can run, jump, hold conversations with humans, do gymnastics (and even sweat for it!) and more. It is really impressive how fast change has been happening recently in this field – and, unsurprisingly, it has stimulated the interest of labour economists!
A paper that has recently come to my attention on this subject is by Graetz and Michaels (2018). The authors put together a panel dataset on robot adoption within seventeen countries from 1993 to 2007 and use advanced econometric techniques to evaluate the effect of these technologies on employment and productivity growth. Their analysis focuses exclusively on developed economies (due to data limitations, as they explain) – but their results are nevertheless intriguing:
We study here for the first time the relationship between industrial robots and economic outcomes across much of the developed world. Using a panel of industries in seventeen countries from 1993 to 2007, we find that increased use of industrial robots is associated with increases in labor productivity. We find that the contribution of increased use of robots to productivity growth is substantial and calculate using conservative estimates that it comes to 0.36 percentage points, accounting for 15% of the aggregate economy-wide productivity growth.
The pattern that we document is robust to including various controls for country trends and changes in the composition of labor and other capital inputs. We also find that robot densification is associated with increases in both total factor productivity and wages, and reductions in output prices. We find no significant relationship between the increased use of industrial robots and overall employment, although we find that robots may be reducing the employment of low-skilled workers.
This is very positive news for most – except, of course, for low-skilled workers. Indeed, like Acemoglu and Restrepo (2017) and many others, this study shows that the effect of automation on employment and labour market outcomes is unlikely to be uniform across all types of workers. Low-skilled workers are found again to be likely to lose out and be significantly displaced by these technologies.
And if you are wondering which sectors are likely to be disrupted most/first by automation, the rankings developed by McKinsey and Company (see chart below) would give you an idea of where the disruption is likely to start. Unsurprisingly, the sectors that seem to be the most vulnerable, are the ones that use the highest share of low-skilled labour.
- “The effect of automation on wages and employment is likely to be positive overall”. Discuss.
- Using examples and anecdotal evidence, do you agree with these findings?
- Using Google Scholar, put together a list of 5 recent (i.e. 2015 or later) articles and working papers on labour markets and automation. Compare and discuss their findings.
Latest resesarch from the independent American think tank The Conference Board paints a worrying picture about the growth of UK labour productivity. While global growth in labour productivity has weakened following the financial crisis, its weakness in the UK is singled out in the Board’s 2019 Productivity Brief. It finds that amongst large mature economies the decline in labour productivity growth rates has been greatest in the UK. This has important implications for the country’s longer-term well-being and, specifically, it peoples’ living standards.
The UK saw the growth in real GDP (national output) fall from 1.8 per cent in 2017 to 1.4 per cent in 2018. The Conference Board predicts that this will fall further to 0.8 per cent in 2019. In the context of living standards, the growth in real GDP per capita is particularly important. An increase in the population will, other things being equal, lower living standards because more people will be sharing a given amount of real national income. The growth in real GDP per capita fell from 1.1 per cent in 2017 to 0.7 per cent in 2018 and is predicted to fall to just 0.1 per cent in 2019.
Chart 1 shows the annual rates of growth in real GDP and real GDP per capita from the 1950s. The average growth rates are 2.4 and 1.9 per cent respectively. The other series shown is the annual growth in real GDP per person employed. This is a measure of the growth in labour productivity. Its average annual growth rate is also 1.9 per cent. This illustrates the intrinsic long-run relationship between labour productivity growth and the growth rate of GDP per capita and hence in general living stanadards. (Click here to download a PowerPoint copy of the chart.)
In the short term, rates of growth in output per worker (labour productivity) and GDP per capita (general living standards) can be less similar. For example, when unemployment rates rise labour productivity rates may be little affected despite GDP per capita falling. Nonetheless, the important point here is the close long-run relationship between the growth in labour productivity and GDP per capita. This then raises an important question: what factors contribute to the growth in output and labour productivity?
An approach known as growth accounting helps to identify four key contributors to the growth of total output. The first is the quantity of labour, commonly measured in labour hours. The second is the quality of labour, also known as labour composition. Third is capital services which are physical inputs into production and include machinery, structures and IT capital. Capital services are affected by quantity and quality, but, unlike labour, it is practically more difficult to separate out these dimensions. Fourth, is Total Factor Productivity (TFP).
TFP it is essentially the residual contribution to output growth that cannot be explained by changes in the quantity and quality of the individual inputs. Hence, in principle, it is capturing changes in how effectively the labour and capital inputs are being employed and combined in production. The Conference Board’s Productivity Brief describes the growth in TFP as providing ‘a more accurate picture of the overall efficiency by which capital, labour and skills are combined in the production process’.
Chart 2 shows Conference Board estimates of the percentage point contribution of these four sources of growth since 1990. Over this period, output growth averaged 2 per cent per year. The contribution of capital services and, hence, what is known as capital accumulation is particularly significant at 1.5 percentage points per year. This has been significantly larger than the contribution of labour hours which averaged only 0.3 percentage points per year since 1990. This evidences the importance played by capital deepening for output growth in the UK. (Click here to download a PowerPoint copy of the chart.)
Capital deepening captures the growth in capital services relative to the growth in the labour input. It takes on even greater significance when we think about the growth in labour productivity since, after all, this is the growth in output relative to the quantity of labour. It is significant though that since 2015 the growth of capital services has contributed only 1 percentage point to output growth while the growth of labour hours has contributed an average of 0.7 percentage points. This points to a slowdown in capital deepening and hence in the growth of labour productivity.
Chart 2 also illustrates the importance of TFP growth to overall output growth. It is also important (along with capital deepening and the growth in labour quality) for the growth in labour productivity. Interestingly, we observe significant fluctuations in the growth of TFP. This is thought to reflect fluctuations in the utilisation of inputs. For example, if the utilisation of inputs falls (rises) when output falls (increases) this will be mirrored by a disproportionately large fall (increase) in TFP. In the longer-term, however, changes in TFP capture aspects of technological progress and advancement that enable more effective production methods and techniques to be deployed. In other words, the growth of TFP captures the ability of production to benefit from the advancement in ideas, products, processes and know-how.
A decline in the growth in TFP growth following the financial crisis is found quite widely in mature economies. The annual rate of growth of TFP across mature economies fell from 0.5 per cent year in 2000-2007 to 0.2 per cent in 2010-2017. In the UK this fall was from 0.5 per cent to -0.1 per cent. Hence, the decline in TFP growth of 0.6 percentage points between 2010 and 2017 was double the 0.3 percentage point fall across all mature economies. In 2018 the Conference Board estimate that TFP in the UK fell by 0.1 percent further exacerbating the downward pressure on labour productivity.
As our final chart shows, it is the magnitude to which labour productivity has eased following the financial crisis that sets the UK apart. While across all mature economies the growth of output per labour hour (another measure of labour productivity growth) fell from an average of 2.3 per cent per year in 2000-2007 to 1.2 per cent in 2010-2017, in the UK the fall was from 2.2 per cent to 0.5 per cent per year. (Click here to download a PowerPoint copy of the chart.)
While the productivity problem facing the UK is not new, the latest figures comes as a very timely reminder of the extent of the problem. To some extent the uncertainty around Brexit and the negative impact on capital accumulation has only helped to exacerbate the problem. But, this may mask a more systemic problem facing the UK. Getting to the root of this problem matters. It matters most significantly for our long-term wellbeing and prosperity. The productivity gap with our major industrial competitors is a gap that policymakers need not only to be mindful of but one that needs closing.
- What do you understand by the term labour productivity. How could we measure it?
- Why is it important to look at the growth of output per capita when assessing the benefits of long-term growth?
- Why is labour productivity important for the long-term well-being of a country?
- What do you understand by the method of growth accounting?
- What is the distinction between capital accumulation and capital deepening?
- What might explain why the growth of labour productivity has been lower in the years following the post-financial crisis?
- What do you understand by Total Factor Productivity (TFP)?
- What does the long-term growth of TFP attempt to capture?
- If you were an economic advisor to the government, what types of policy initiatives might you recommend for a government concerned about low rates of growth of labour productivity?
The distinction between nominal and real values is an incredibly important one in economics. We apply the latest GDP numbers from the ONS to show how the inflation-adjusted numbers help to convey the twin characteristics of growth: positive longer-term growth but variable short-term rates of growth. It is real GDP numbers that help us to understand better the macroeconomic environment and, not least, its inherent volatility. To use nominal GDP numbers means painting a less than clear, if not inaccurate, picture of the macroeconomic environment.
The provisional estimate for GDP (the value of output) in the UK in 2018 is £2.115 trillion, up 3.2 per cent from £2.050 trillion in 2017. These are the actual numbers, or what are referred to as nominal values. They make no adjustment for inflation and reflect the prices of output that were prevailing at the time. Hence, the figures are also referred to as GDP at current prices.
The use of nominal GDP data can be something of a problem when we compare historical values. In 1950, for example, as we can see from Chart 1, nominal GDP in 1950 was a mere £12.926 billion. In other words, the nominal figures show that the value of the country’s output was 163.595 times greater in 2018 (or an increase of 162,595 per cent). However, if we want to make a more meaningful comparison of the country’s national income we need to adjust for inflation. (Click here to download a PowerPoint copy of the chart.)
If we measure GDP at constant prices we eliminate the effect of inflation. This allow us to make a more meaningful comparison of national income. Consider first the real GDP numbers for 1950 and 2018. GDP in 1950 at 2016 prices was £373.9 billion. This is higher than the nominal (current-price) value because prices in 2016 were higher than those in 1950. Meanwhile, GDP in 2018 when measured at 2016 prices was £2.034 trillion. This real value is smaller than the corresponding nominal value because prices in 2016 where lower than those in 2018.
Between 1950 and 2018 there was a proportionate increase in real GDP of 5.439 (or a 443.9 per cent increase). Because we have removed the effect of inflation the real growth figure is much lower than the nominal growth figure. Crucially, what we are left with is an indicator of the growth in the volume of output. Whereas nominal growth rates are affected both by changes in volumes and prices, real growth rates reflect only changes in volumes.
Consider now output growth between 2017 and 2018. As we saw earlier, the nominal figures suggest growth of 3.2 per cent. In fact, GDP at constant 2016 prices increased from £2005.4 trillion in 2017 to £2,033.6 trillion in 2018: an increase of 1.4 per cent. This was the lowest rate of growth in national output since 2012 when output also grew by 1.4 per cent. In 2017 national output had increased by 1.8 per cent, the same increase as in 2016.
To put the recent growth in national output into context, Chart 2 shows the annual rate of growth in real GDP each year since 1950. Across the period, the average annual rate of growth in real GDP and, hence, in the volume of national output was 2.5 per cent. In the current decade growth has averaged only 1.9 per cent. This followed falls of 0.3 per cent and 4.2 per cent in 2008 and 2009 respectively as the effects of the financial crisis on the economy were felt. (Click here to download a PowerPoint copy of the chart.)
By plotting the percentage changes in real GDP from year to year, we get a much clearer sense of the inherent instability that we identified at the outset as a characteristic of growth. This is true not only for the UK, but economies more generally. This instability is the key characteristic of the macroeconomic environment. It influences and informs much of what we study in economics.
The variability of growth rates that create the instability of economies again requires an understanding of the distinction between nominal and real GDP. Chart 3 illustrates the growth in GDP both in nominal and real terms. The average annual rate of growth of nominal GDP is 7.8 per cent, considerably higher than the average real growth rate of 2.5 per cent per year. The difference again reflects the effect of rising prices. (Click here to download a PowerPoint copy of the chart.
Chart 3 clearly shows the wrong conclusions that can be drawn if one was to focus on the growth in nominal GDP from year to year. Perhaps the best example is 1975. In this year nominal GDP grew by 24.2 per cent. However, the volume of national output contracted: real GDP fell by 1.5 per cent. The growth in nominal GDP reflects the rapid growth in prices seen in that year. The economy’s average price level (the GDP deflator) rose by 26.1 per cent. Hence, the growth in nominal GDP reflected not an increase in the volume of output – that fell – but instead a large increase in prices.
The importance of the distinction between nominal and real GDP is further demonstrated by the fact that since 1950 nominal GDP has fallen in only one year. In 2009 nominal GDP fell by 2.7 per cent. The 1.6 per cent rise in the economy’s average price level was not enough to offset the fall in the volume of output of just over 4.2 per cent. In other years when the volume of output (real GDP) fell, the effect of rising prices meant that the value of output (nominal GDP) nonetheless rose.
So to conclude, the distinction between nominal and real GDP is crucial when analysing economic growth. To understand the distinction gives you a truly real advantage in making sense of the macroeconomic environment.
- What do you understand by the term ‘macroeconomic environment’? What data could be used to describe the macroeconomic environment?
- When a country experiences positive rates of inflation, which is higher: nominal economic growth or real economic growth?
- Does an increase in nominal GDP mean a country’s production has increased? Explain your answer.
- Does a decrease in nominal GDP mean a country’s production has decreased? Explain your answer.
- Why does a change in the growth of real GDP allow us to focus on what has happened to the volume of production?
- What does the concept of the ‘business cycle’ have to do with real rates of economic growth?
- When would falls in real GDP be classified as a recession?
- Distinguish between the concepts of ‘short-term growth rates’ and ‘longer-term growth’.
- Why might the distinction between nominal and real be important when analysing changes in people’s pay? What would be the significance of an increase in real pay?
How would your life be without the internet? For many of you, this is a question that may be difficult to answer – as the internet has probably been an integral part of your life, probably since a very young age. We use internet infrastructure (broadband, 4G, 5G) to communicate, to shop, to educate ourselves, to keep in touch with each other, to buy and sell goods and services. We use it to seek and find new information, to learn how to cook, to download music, to watch movies. We also use the internet to make fast payments, transfer money between accounts, manage our ISA or our pension fund, set up direct debits and pay our credit-card bills.
I could spend hours writing about all the things that we do over the internet these days, and I would probably never manage to come up with a complete list. Just think about how many hours you spend online every day. Most likely, much of your waking time is spent using internet-based services one way or another (including apps on your phone, streaming on your phone, tablet or your smart TV and similar). If your access to the internet was disrupted, you would certainly feel the difference. What if you just couldn’t afford to have computer or internet access? What effect would that have on your education, your ability to find a job, and your income?
Martin Jenkins, a former homeless man, now entrepreneur, thinks that the magnitude of this effect is rather significant. In fact, he is so convinced about the importance of bringing the internet to poorer households, that he recently founded a company, Neptune, offering low-income households in the Bronx district of New York free access to online education, healthcare and finance portals. His venture was mentioned in a recent (and very interesting) BBC article – a link to which can be found at the end of this blog. But is internet connectivity really that important when it comes to economic and labour market outcomes? And is there a systematic link between economic growth and internet penetration rates?
These are all questions that have been the subject of intensive debate over the last few years, in the context of both developed and developing economies. Indeed, the ‘digital divide’ as it is known (the economic gap between the internet haves and have nots) is not something that concerns only developing countries. According to a recent policy brief published by the New York City Comptroller:
More than one-third (34 percent) of households in the Bronx lack broadband at home, compared to 30 percent in Brooklyn, 26 percent in Queens, 22 percent in Staten Island, and 21 percent in Manhattan.
The report goes on to present data on the percentage of households with internet connection at home by NYC district, and it does not take advanced econometric skills for one to notice that there is a clear link between median district income and broadband access. Wealthier districts (e.g. Manhattan Community District 1 & 2 – Battery Park City, Greenwich Village & Soho PUMA), tend to have a significantly higher share of households with broadband access, than less affluent ones (e.g. NYC-Brooklyn Community District 13 – Brighton Beach & Coney Island PUMA) – 88% of total households compared with 58%.
But, do these large variations in internet connectivity matter? The evidence is mixed. On the one hand, there are several studies that find a clear, strong link between internet penetration and economic growth. Czernich et al (2011), for instance, using data on OECD countries over the period 1996–2007, find that “a 10 percentage point increase in broadband penetration raised annual per capita growth by 0.9–1.5 percentage points”.
Another study by Koutroumpis (2018) examined the effect of rolling out broadband in the UK.
For the UK, the speed increase contributed 1.71% to GDP in total and 0.12% annually. Combining the effect of the adoption and speed changes increased UK GDP by 6.99% cumulatively and 0.49% annually on average”. (pp.10–11)
The evidence is less clear, however, when one tries to estimate the benefits between different types of workers – low and high skilled. In a recent paper, Atasoy (2013) finds that:
gaining access to broadband services in a county is associated with approximately a 1.8 percentage point increase in the employment rate, with larger effects in rural and isolated areas.
But then he adds:
most of the employment gains result from existing firms increasing the scale of their labor demand and from growth in the labor force. These results are consistent with a theoretical model in which broadband technology is complementary to skilled workers, with larger effects among college-educated workers and in industries and occupations that employ more college-educated workers.
Similarly, Forman et al (2009) analyse the effect of business use of advanced internet technology and local variation in US wage growth, over the period 1995–2000. Their findings show that:
Advanced internet technology is associated with larger wage growth in places that were already well off. These are places with highly educated and large urban populations, and concentration of IT-intensive industry. Overall, advanced internet explains over half of the difference in wage growth between these counties and all others.
How important then is internet access as a determinant of growth and economic activity and what role does it have in bridging economic disparities between communities? The answer to this question is most likely ‘very important’ – but less straightforward than one might have assumed.
- Comptroller, New York City, Internet Inequality
- Czernich, N., Falck, O., Kretschmer, T. and Woessmann, L., 2011, Broadband infrastructure and economic growth, The Economic Journal, 121(552), pp.505–32
- Koutroumpis, P., 2018, The economic impact of broadband: evidence from OECD countries, Ofcom
- Atasoy, H., 2013, The effects of broadband internet expansion on labor market outcomes, ILR Review, 66(2), pp.315–45
- Forman, C., Goldfarb, A. and Greenstein, S., 2009, The Internet and Local Wages: Convergence or Divergence? (No. w14750), National Bureau of Economic Research
- Is there a link between economic growth and internet access? Discuss, using examples.
- Explain the arguments for and against government intervention to subsidise internet access of poorer households.
- How important is the internet to you and your day to day life? Take a day offline (yes, really – a whole day). Then come back and write about it.