A popular explanation for the UK's comparatively low productivity growth is its low level of investment in research and development (R&D). In 2014, I authored a report for the UK government on "Insights from international benchmarking of the UK science and innovation system" that goes into detail on these arguments. The upshot is that just increasing R&D expenditure alone is unlikely to be an effective policy for boosting growth.
In this short blog, I want to bring some nuance to the facts about the UK's "low" investment in R&D. Putting aside any definitional issues (which are significant), there are two important adjustments that need to be made to make any meaningful comparisons: the inclusion of financial services product development in R&D investment; and an adjustment for sector mix (manufacturing firms typically invest 30 times as much in R&D as those in the retail and wholesale sector).
As the picture above shows, including financial services product development in the calculation increases the UK's R&D investment as a share of GDP from 1.7% to 1.9%. Not a huge boost; and, of course, this adjustment also increases other countries' figures. However, even if small, I would argue that qualitatively, thinking of the UK's financial services sector as a key source of innovation does matter for the country's economic narrative.
The second adjustment makes more of a difference, in relative terms. While the sector mix differences between different developed nations are surprisingly small, the UK does have a larger share of, for example, construction, and a lower share of manufacturing in its GDP than other comparative countries. And average levels of investment in R&D in construction are much lower than in manufacturing.
There are many ways to make these comparisons, but perhaps the least confusing is this. On traditional metrics, the UK's R&D investment as share of GDP is 1.7%, or 0.8%-points below the average of the comparator countries (2.5%). By including financial services, and simulating what countries' figures would look like if they all had an "average" industrial structure, increases the UK's figure to 1.9% and reduces the gap to comparators to 0.5%-points (which now average 2.4%). The reduced gap is primarily because many countries' R&D intensity in the original metrics is boosted by a large share of manufacturing (and sometimes defence) in their sector mix.
Why does any of this matter? After all, it doesn't fundamentally alter the broad conclusion that UK is not a heavy investor in R&D. At one level, it is only interesting to data geeks like me. I am happier when I feel that I'm comparing at least different apple varieties rather than apples and oranges.
But it also matters for policy. If a country is targeting a certain amount of R&D, it is unlikely to succeed if it doesn't pay close attention to the characteristics of its existing business population and the overall sector mix of its economy. So any set of policies aiming at increasing growth or productivity through higher levels of R&D investment will need to look well beyond just increasing support for R&D.