Effective strategies do not necessarily require lots of money to be thrown at specific technologies to increase productivity growth, says Thomas Aubrey.
Over the last few months a number of economic commentators[1] have argued that what matters most in driving up productivity in the UK is to focus on increasing aggregate demand. The idea is that running the economy ‘hot’ will force firms to innovate and automate as labour shortages emerge. This argument shifts policies away from supply-side strategies focusing on key sectors, which – it is argued – can play only a small part in redressing the productivity gap.
One question this raises is what the historical evidence has to say concerning a continuous boost to aggregate demand? The last time developed economies ran hot in this way was back in the 1960s. If we take the unemployment rate as the relevant indicator, France and the UK had unemployment around 2% then. The rate was slightly lower in West Germany and somewhat higher in the US, albeit on a downward trend. The logic of the argument that aggregate demand is needed to drive productivity growth ought to have led to lower total factor productivity (TFP) growth in the US and higher growth in West Germany. But the figures do not support a clear relationship. In fact, France had the highest rate of TFP growth, which over the decade was double that of the UK. Other factors are clearly relevant here.
Chart 1: Productivity performance 1960s
There is however evidence that for economies with flexible labour markets and a steady supply of low-cost workers, firms are less likely to invest in capital resulting in lower labour productivity growth.[2] This is particularly the case for sectors such as agriculture where investing in, for example, fruit picking robots – assuming the business operates at big enough scale to make automation worthwhile – would result in higher productivity growth. Hence higher productivity growth could potentially be achieved through a tighter labour market in certain sectors. But many non-tradable sectors (including many services) do not operate at sufficient scale so running the economy hot is less likely to be an effective strategy in those areas.
Nor is this to say that demand doesn’t play a role. Reduced demand has likely led to lower levels of productivity growth since the financial crisis in the UK as reported in the CFM Survey. This can be clearly observed in the banking sector as demand for financial services products shrank dramatically after the crisis. The collapse in oil prices from 2014 also reduced production of North Sea oil due to its relatively higher production costs. If the UK had not had these two sectors between 2010 -2019, UK productivity would have been 77% higher.
But even with this boost, productivity growth would still have been significantly lower than Taiwan as shown in chart 2. This raises the question why Taiwan has maintained much higher levels of productivity growth than the US and UK in recent years.
Chart 2: Productivity growth comparison 2006-2020
A disaggregation[3] of private sector productivity growth of close to 50% between 2006 and 2020, indicates that nearly a half of this was due to manufacturing, followed by the wholesale and retail sector accounting for 13% and financial services for 7%. The performance of specific sections of the economy with a competitive advantage – in this case manufacturing – therefore seems an important part of the productivity story.
Critically, this higher rate of productivity growth in one section of the economy drives up wages across the economy. While manufacturing saw the largest increase in output per hour in Taiwan, it experienced a smaller than average increase in wages (although manufacturing wages there are relatively high). The largest increase in wages occurred in other sectors, with the lowest productivity growth (Table 1). This supports Enrico Moretti’s argument in The Geography of Jobs that rising wages in high value added sectors drive up local aggregate demand, which raises wage levels in lower value added sectors. The direction of causality here is the opposite of that in the aggregate demand argument. This finding additionally implies that regions which do not have firms operating in sectors experiencing faster productivity growth are likely to have lower wage growth. This has obvious implications for geographic inequalities and Levelling Up.
Table 1: Comparison of increase in wages, growth in value added and levels of value added in selected sectors
Source: National Statistics, Republic of China (Taiwan)
The uneven nature of productivity growth across sectors, reflected in different national rates of productivity growth and higher wages is wider than a Taiwanese phenomenon. Another example took place in the US during the 1930s when it experienced one of its highest ever rates of productivity growth (described by Alexander Field in A Great Leap Forward). This was driven by just three sections of the economy including manufacturing, which benefitted from improved factory layouts and increased scale, as well as innovations like new chemical processes. Rail and road transportation also saw increased growth, enabled by the building of the interstate highway network after 1926. This in turn benefited the wholesale and retail sectors. Hence, it was a combination of manufacturing, transportation and distribution that generated such high levels of productivity growth despite weak aggregate demand during that decade.
National productivity growth will therefore always reflect the unequal growth across different sections of the economy at different times. Yet it does not follow that the right policy to increase productivity growth is for the government to throw money at cool technologies along sectoral lines. Nor are other current UK government policies supporting productivity growth. Leaving the EU single market has unsurprisingly negatively impacted the productivity of manufacturing, given that manufacturers are part of global supply chains. Furthermore, if regulations lead to a divergence from the current EU technical requirements as part of the new UKCA mark they will have a negative impact on the productivity of affected businesses in potential high growth sectors such as medical technology. The UK also lacks enough skilled labour, particularly in areas like computer programming. Effective strategies therefore do not necessarily require lots of money to be thrown at specific technologies.
The evidence from countries like Taiwan indicates that as long as some firms in a handful of sectors experience faster rates of productivity growth, they can drive up demand for local services, thereby increasing wages too. The last time the UK relied on faster aggregate demand growth, it was the productivity laggard. Growth in some leading, competitive sectors is essential too.
[1] Long live the labour shortages | Financial Times (ft.com) Martin Sandbu and Productivity: firing on all cylinders | The Institute for Government Giles Wilkes
[2] Vergeer & Kleinknecht 2014 Do labour market reforms reduce labour productivity growth?
[3] Based on Tang & Wang, Canadian Journal of Economics: Sources of aggregate labour productivity growth in Canada and the United States 2004
The views and opinions expressed in this post are those of the author(s) and not necessarily those of the Bennett Institute for Public Policy.