This book was one of The Economist‘s books of the year in 2020 – and I bought in time for Christmas. It has taken me until now to read it. It tackles the question of why economic growth in the US, along with the rest of the developed world, slowed in the 21st century compared to the last half of the 20th. It is a topic that has troubled many, but the usual response is to attribute it to whatever the commentator’s pet theory happens to be, throwing in a few pieces of circumstantial evidence in support. I am as guilty of this as anybody: using my theory that it is mainly the Baumol effect – a switch in the economy from manufacturing to services. The virtue of Dietrich Vollrath’s book is that he is led by the evidence – though restricted to the United States. He concludes that, mostly, the slowdown arises from the problems of success, rather than being any failure of economic management.
Interestingly, Professor Vollrath admits to starting his search with his own pet theory: that the slowdown resulted from the growing concentration of big business, their consequent market power, and the resultant higher profit margins. That is doubtless why he devotes three whole chapters to it compared to one on Baumol – but his conclusion is that it did not have a major impact on growth – and its effect is even ambiguous. His overall conclusion is that the growth rate (in GDP per capita) slowed by 1.25% per annum, from 2.25% to 1.00% on average, and that 0.8% of this is due to demographics (i.e., a smaller proportion of people working), 0.2% is due to the Baumol effect, 0.15% due to a slower turnover of staff and firms, and 0.1% is due to a decline in geographic mobility. Changes to tax and regulation, rising inequality and trade with China were all examined but had an overall effect of nil. Both the demographic effects (smaller families, retirement, etc.) and the Baumol effect arise from choices to be expected as societies grow more prosperous – so he calls these the results of success. This, of course, fits in with what I have been saying for some time – though in my commentary the Baumol effect figured larger than demographics. Professor Vollmer finds this effect to be real but slow-acting, and dwarfed by the effect of demographic changes.
How much does this apply to Britain? According to some statistics I have found on an internet search, the growth of GDP per head in the UK was 2.4% per annum from 1950 to 2000, and 1.1% from 2000 to 2016 (actually a bit more than the US – but this may an issue with the statistical series) – a very similar level of decline. I don’t have comparable figures on demographics, but the same sort of thing was going on in the UK, with a clear baby boom in the 1940s and 1950s, followed by shrinking family sizes. If anything family sizes were shrinking further – and we had a lower rate of immigration. This week The Economist has started a series on Britain’s productivity problem – but it is hard to tie in their data with this book. It is looking at a shorter time period, and comparing output per hour worked with other countries. And Britain does seem to be less productive than many other countries – and productivity has fallen since the great financial crisis in 2007-09, though it is not so different from other countries in that (it has a similar decline to the US). But productivity figures are notoriously difficult to calculate, and they are not necessarily fully comparable from country to country, even in the same OECD data set. The most important conclusion arising from Professor Vollmer’s calculations is that, so far as the overall economy is concerned, productivity is liable to be trumped by demographics. And here the position in Britain post covid is grim. A lot of people have dropped out of the work force, though it is unclear why (early retirement looks the most likely). Immigration remains high, but with extra bureaucratic hurdles since Brexit, imported workers aren’t necessarily going to the places of highest stress.
One conclusion of this book is that it is hard to detect any growth effects from deregulation or tax changes – so it is hard to see that the government’s hopes for increased productivity will have much effect. The partial exception, according to Prof. Vollmer, is changes that allow greater housing mobility. Allowing more homes to be built in areas of high productivity is the leading aspect of this, and the government shows no sign of wanting to deal with this. One policy not talked of in the book is the idea of “levelling up” – of improving the productivity of lagging regions to be closer to that o the leading ones. According to The Economist Britain has a particularly big regional gap – so there is the possibility of growth through catch up. This probably entails substantial devolution of tax-raising and spending powers, which Westminster is fond of talking of, but never actually does much about. So the outlook for GDP growth in the UK looks weak – even if, as the book suggests, a lot of this is actually a problem of success.
This book does have a couple of major defects, which arise from the way the problem has been framed. The first is that there is no investigation of the impact of status goods (or services). These are significant because their economic characteristics differ from normal goods – the point of them is that they are expensive and so not generally accessible. Low productivity is often a feature, not a problem. It is to be expected that as an economy becomes more successful, the demand for status goods rises. This would have a very similar effect to the one Baumol identified (and, indeed it doubtless contributes to the Baumol effect, as a lot of services are status symbols). But I have seen no attempt to prove this statistically – and Prof. Vollmer does not even mention it. It may also be a factor in the rise of market power he discusses. Apple’s high margins, for example, arise in part from the fact that their products have become status symbols. Modelling the effects of status goods on the economy at large is complex, however – and it is not well supported by established statistical categories. So it is not surprising that the book ignores them. But a reduction in economic growth as a result of a move to status goods would support its “problems of success” thesis.
The second disappointment is the narrow way it deals with the effects of globalisation. It solely looks at the effect of workers displaced by the move of production in America abroad, and in particular to China. It finds a mild negative effect on productivity. Prof.Vollmer keeps any benefit of lower prices to American consumers from increased trade out of the scope of his analysis. This is disappointing, because I believe that it could be significant. I’m sure it was in the UK, whose economy is more exposed to world trade, as price reductions on imported manufacturing goods had a significant impact on overall prices in the first part of the 21st century. I’m not sure how this would work through into the statistics though (the D in GDP stands for “domestic” after all). This should manifest as a positive effect on growth in the late 1990s and early 2000s, going into reverse afterwards. But it would make the story significantly less tidy, so it is easy to see why Prof. Vollmer left it out.
This remains an important book, however. Economic growth attracts an awful lot of commentary from economists and politicians. Given that, it is surprising how weakly understood the phenomenon is, even amongst people who should know better. Alas that includes the writers at The Economist, whose article on Britain’s economic growth only mentions the effect of demographics in passing, and superficially at that. And that from a journal that made the book recommended reading. This is yet another sign of just how stilted most public discussion of economics remains.
Very good piece.
Thank you David
Interesting. The Economist article seemed to me to have a fair point when it laments that recently the UK’s output per hour has not been behaving well, and that the immediate prospect for this metric is poor – where-as in Thatcher’s time this metric performed well, it pointedly remarks. But what was it about the Thatcher policies which achieved this, and can we reproduce this today? The Economists remarks about the failings of housing policy correspond to the point you, note, Matthew, that allowing houses to be built in areas of high productivity does assist growth, where-as according to Vollrath , Thathertite changes to tax and regulation would not.
The strength of the Economist’s argument is that it focuses on the output per hour metric. That means it sidesteps the demographics argument. It is also looking at shorter timespans then Vollrath’s book.It would good to get behind the rise in productivity in the 1980s though. As ever there are lots of ideas, but pinning them down to hard evidence is tricky. There was a lot of business closure and start-up, and quite a bit of population movement, which Vollrath would recognise as positive factors. And his analysis of tax and regulation is very focused on the US. The didn’t have such low-hanging fruit as BT and British Gas to reform!
I haven’t read the book so I’m not sure about this. However, my initial thought is that a better sub-title for the book might be “Why a stagnant economy should be, but isn’t, a sign of success”.
I can’t see why you think economic growth is only “weakly understood.” Human beings are naturally inventive and innovative. Give anyone a task and they’ll find a way to do it quicker the second time, then quicker still the third time etc. There is a natural incentive within the capitalist system to make use of this process and produce the same thing with fewer resources from one year to the next. Fewer resources means fewer people.
This is good for the owners of the process but not so good for anyone who is displaced and loses their livelihood.
Alternatively, there can be a general expansion which usually involves government involvement to run the correct fiscal and monetary policies to expand the economy, keeping both the workers happy with extra jobs and the employers happy with extra production and profits.
So what would the economy have to look like for a zero growth policy to be generally beneficial? The ownership of the means of production can’t be in private hands for starters. They’ll always want to use any improvements in working methods, and there is no preventing new and better inventions, for their own benefit. They are unlikely to agree that there is going to be any benefit at all for them.
For everyone else, it will mean that an increased level of productivity will translate into fewer working hours. I can just about see how all this might be possible but we’re a long way off making it a reality.
Actually the the book title is misleading – surely the publisher’s idea and not the author’s. Vollrath does not say that the US economy is stagnant. He goes out of his way to point out that it was still growing in the 21st century – just at a slower percentage rate. He even points out that the arithmetic scale of the later growth is about the same as before – it is just that the denominator is larger. Indeed he suggests that there is little danger of China catching up wit the US in per capita income. This is indeed the direction that Baumol’s insights point to.
I suggest that growth is weakly understood because of all the nonsense I see written about it. There seems to be little understanding of how its components (human capital, physical capital, productivity, etc, work together). There is almost no awareness of the Baumol effect, even though it is standard BSc curriculum material. There is often a fallacy of composition similar to the way people confuse managing national finance with household finance. People seem to think that the production side of the economy works like a single factory scaled up. Explanationsof the “productivity puzzle” come and go without people showing a clear grasp of the actual dynamics.
I agree that the distribution of wealth, and the benefits of growth, is critical. The capitalist system has conspicuously failed here since the neoliberal consensus got going. There is a widespread assumption that growth means that everybody gets richer – and that questions of distribution are secondary. But as lower proportional growth is here to stay, distribution is critical. And if workers truly get the power to choose, they may well choose to work less…
An excellent piece, Matthew.
I notice there is no mention of the prediction from the Robert Solow’s classic growth model that I learnt about in my formal study of economics twenty-odd years ago. In a short piece I am drafting at the moment, I have said (without mentioning Solow):
“Then there is the problem of maintaining our capital stock. A big house costs more to heat and maintain than a small one. The more physical capital we have to maintain, the less spare income we have for growth-enhancing investment. For this reason, rich countries often grow more slowly than poor ones.”
Perhaps the Solow effect isn’t as important as I thought it was.
Vollrath does discuss the effect of capital stock on growth – but he doesn’t identify a major effect. His explanation for poorer countries growing faster is mainly around technological catch-up – an old story. Doubtless demographics have role too, though shrinkage to family size has been dramatic.
Thank you, Matthew, for this article.
One issue that often seems to be missing from the debate on productivity is that more than a decade of suppression of public sector pay has had a significant impact on productivity. As public pay has, in real terms, fallen, so has the sector’s productivity, thanks to the way it is measured. Increase public sector pay, and it will automatically improve overall productivity.
As you say, productivity in the service sector is difficult to measure. If a shop assistant sells one dress priced at £1000, is she or he really twice as efficient as the assistant selling 5 £100 dresses?
On the Economist, one of it’s current weakness (in my view) is its obsession with house building. Yes, it could make a difference in theory but loosening planning laws is unlike to achieve the housing mobility they want, not least because of the complexity of the housing market and its relationship to interest rates rather than demand. Is moving the jobs away from the South East really less effective than moving people to the South East?
Thanks Richard. The picture on public sector pay can be ambiguous. Recently (obviously not now), the growth of “flexible” working in the private sector meant that public sector pay was pulling ahead. But there are clearly big issues in the public sector. One issue has been the outsourcing of work formerly done by professional staff to private suppliers (including not-for-profits) using unqualified staff instead. This arrangement has proved unable to deal with the complex problems that are such a feature of public sector work (just think about social work). There is no good way of measuring effective output.
I agree about the Economist’s fixation with housebuilding – though funnily enough it is in line with Prof. Vollrath’s thinking on the point. Another problem is that it assumption that bigger companies are more productive, and that the European failure to produce a Google or Apple is sign of major shortcomings. I think I will look into this in a future blog post!