Today’s chart is a reproduction of one that featured in a recent blog on the Bank Underground website, which is the very informative blog-platform of the Bank of England, the central bank of the UK. It shows the Great Britain’s productivity growth since 1761 (the thin grey line) and the ten-year moving average (the black line), as well as the 100-year average growth rates of the 1800s and the 1900s. It doesn’t take a genius to get the main takeaway of this chart: that productivity has been trending down for decades. In fact, we have now reached the point that growth as measured as a ten-year average has turned negative. To give you an idea of how serious this is, a quote from the author’s blog will probably suffice: “How poor has the past decade of productivity growth been by historical standards? Exceptionally.”
It’s an impressive chart for sure, which leaves one with the slightly depressing feeling that a new world order, in which productivity growth is a thing of the past, is upon us. So is it? My ability to predict the future is as limited as the next person’s, so I cannot claim to know that things will turn out just fine. Still, let me offer a bit of pushback on the chart and the gloomy feeling it gives us.
For starters, I cannot help but be reminded of what people were ‘blogging’ about at the beginning of the 1900s. Productivity growth had peaked as high as 2% in the 1870s, after which things started to deteriorate. Productivity trended down for decades and by the turn of the century dropped to the point that growth as measured as a ten-year average had turned negative. By 1908 the ten-year average growth rate was -0.22%, even lower than the figure recorded in 2016 (-0.17%). Things must have seemed pretty bleak, right? In hindsight, we can see that it was a weak start of what would become a record-breaking century. With a bit of imagination, a similar pattern can be seen around the 1800s. The century started with two weak decades, after which things picked up. To show this, I have plotted the productivity growth at the start of the three centuries in one chart. Maybe the dip is not so exceptional after all…
Now, I am not a big fan of wave cycles, so I am certainly not suggesting that this is a century cycle. The main point I am trying to make is that we have been here before and on previous occasions the harm appears to have been temporary.
So what were people talking about at the beginning of the 1900s? I haven’t looked it up, but one thing I am pretty sure of, and that is that they were not writing about these productivity numbers. The reason for this is simple: the data wasn’t available back then. The productivity numbers presented here are based on GDP equivalent data, but as the whole concept of GDP wasn’t even introduced until the mid-1930s in the US, by far the biggest chunk of the graph is based on guesstimates of what the UK economy was doing. Whereas we currently have statistics offices that follow strict guidelines to come up with exact numbers for the GDP, inflation corrections and the like, they did not have the same ability to collect data back in the day. With the wars, bank crises, bubbles and deflation scares, it is clear that the data gathered before 1945 should be taken with a large pinch of salt.
This can be said of normal GDP growth data, but is even more true for the productivity measure presented in the chart. This chart looks at Total Factor Productivity data (TFP), a concept that – even in modern times – has been criticized. Without getting too much into the weeds here, broadly speaking, there are two types of productivity data. The concept that most people know and intuitively understand is the concept of labor productivity. This idea is simple: how much does one ‘unit’ of labor produce? This measure is calculated by taking a country’s total production (GDP) and dividing it by the hours worked. From this perspective, growth can either come from more labor, or from labor becoming more productive.
Total Factor Productivity takes productivity a step further. Not only do you adjust for the number of hours worked, but also for the amount of capital used. This measure is calculated by taking a country’s total production (GDP) and dividing it by the hours worked and capital employed. From this point of view, growth can either come from more labor, more capital or from productivity gains. Conceptually, it is a much more interesting concept to work with, but you can imagine that this higher level of sophistication can only work if you have an equally sophisticated method for gathering the data. We may have reached that level by now, but we certainly weren’t there yet in 1936, 1880 or 1795… Even nowadays, with the Amazons of the world offering more and more cloud-based services, it would be difficult to assess ‘the’ amount of capital used in each country.
So where does this leave us? Given the lack of sophistication in the past, I would propose looking at the simpler of the two measures, as is done in this final graph. The pattern is still pretty much the same as in the first chart, the good news being that we still are in positive territory with respect to ten-year productivity growth. Admittedly, the drop in productivity is still significant and worrying, but the term ‘exceptionally’ seems to be an overstatement in this case.