news article background image
2018-03-22Macro `n Cheese

Investing the time

In this week’s post Roger is trying to chart out what the data says about the modern productivity paradox or, to (para)phrase it in a more familiar way: “you can see the robots everywhere, but in the productivity statistics”.

The debate on productivity or, rather, the lack of productivity has been one of the most hotly debated issues over the last 10-15 years. And no matter if you belong to the techno-optimists or the techno-pessimists it is a question we should indeed follow up on. Below, due to time constraints, I have decided to mainly chart the question out and spare you my usual tirades (don’t get used to it though).

First, this is what investments and productivity looks like (I will mainly use US data, but some of the charts should be replicable for other countries as well).

Macrobond Document Macrobond Document
Chart 1a & b: Hard to come by

Studying investment developments at large, it should be clear that the investment trends of yesteryear are broken. What is more, and problematic from a productivity perspective, is that the gap between actual- and trend- developments are increasing in vital areas such as information processing equipment, R&D and software investments. These developments are also mirrored in productivity statistics.

Macrobond Document
Chart 2:Productivity falling short of trends

The disappointing developments seem to be wide spread over the private sector, but perhaps particularly acute in the durable goods sector, despite the advent of robots. Speaking of which…

Macrobond Document
Chart 3: Developments that are hard to square

The graph above is, if nothing else, an illustration of the paradoxical developments. The number of industrial robots is moving up, pronouncedly so, but the shipment stats, which are a good indicator of investments (in national accounts), is showing only minor improvements.

To gain some more insight we would of course like to study productivity developments on a more detailed basis and, if possible, find out what production factors are driving the deterioration in growth, investments and productivity. There are a couple of datasets (available in MB) that are extremely well adapted for such endeavors.

In particular I want to highlight the Conference Board’s Total Economy Database (TED) and the EU’s KLEMS database, with the latter having both intermediate inputs and a detailed breakdown, for sector level data as well.

Macrobond Document
Chart 4:TED talks

Looking at TED-data we can see that the labor contribution to output growth is more or less the same size as before the crisis. Worryingly, what seems to be going on is that the contribution to GDP-growth from ICT-assets and, even worse, total factor productivity (i.e., from innovation) is retreating.

That said, from the above graph we can also read that the contribution from ICT-assets is reverting to levels pre-dating the Goldilocks-era, which fits well with the now common perception that the (90’s) ICT-boom was more a shift of production techniques (implying a level shift) than utilization (which would imply higher growth).

Now, and conveniently barring any Cambridge critique, the pronounced deceleration in total factor productivity (TFP) after the GFC only serves to deepen the productivity paradox. How can it be, that after discounting the contribution from labor and capital deepening (more capital per employee) that the truest form of productivity, TFP – innovation – has more or less completely stagnated? It suggests there are none of the expected positive effects from robotization, artificial intelligence etc.

Let’s dig deeper. Looking at the EU-KLEMS database we can perform a similar analysis to TED, but with some more detail.

Macrobond Document Macrobond Document
Chart 5: Is the US-growth spurt only cyclical?

Here, for instance, I have chosen to look at a sector that should benefit from an increased robotization of the economy; ‘Installation of machinery and equipment’. At the least, we see a cyclical improvement as hours worked is improving. At best, we can hope that these are not just cyclical developments but also a precursor to stronger investments and innovation. If you have the Macrobond application, you can download the charts above and open the Macrobond document. That way you can easily substitute other sectors and countries.

If we look at the different growth contributions and see how their means have developed between the period before and after the financial crisis, we have another neat way of slicing and dicing the data.

Macrobond Document
Macrobond Document Macrobond Document
Macrobond Document Macrobond Document
Note: It is possible to delve even deeper into sectoral details in the EU-KLEMS database. Here I have chosen to study only the main aggregates. Also, it is very easy to substitute the country.
Chart 6a-e: Drivers of value added growth in the US, per sector

To be honest, most of the improvements in value-added growth seem to be purely cyclical, as it is almost only in hours worked that we see improvements throughout. To add insult to injury, improvements in contributions are becoming considerably more concentrated post-2011, which corroborates the proposition that productivity (and hence, profits and wages) are also becoming more concentrated (the superstar vs zombie-firm debate).

When studying main sectors such as manufacturing, it is quite hard to discern any sign of increased innovation as a slight uptick in non-ICT capital is counteracted by deterioration on almost all other accounts. In trade, we do see a slight improvement in TFP-contributions (innovation), but those developments are hard to square with other developments, e.g., that the labor composition (educational attainment etc.) has deteriorated. The only sector where the contributions seem coherent and not just the result of cyclical developments is within the mining and quarrying sector where innovation is strong and capital deepening and labor composition also indicate that a more fundamental improvement is taking place. That said, this is where we find that the shale industry is complex, as it as has had quite a (cyclical) ride during the post-GFC years.

To wrap up on an uncharacteristically and hopelessly optimistic note, I cannot help but find it intriguing to see TFP – innovation – developing quite well in wholesale and retail trade, real estate, professional and scientific as well as recreational services etc. These are all industries that stand to benefit not only from robotization, but from artificial intelligence, machine learning and other “hot” technologies.

If nothing else, these nascent signs are rays of light in an otherwise quite depressing dissection of growth, investments and productivity developments.

Get these posts delivered to your inbox

If you like what you read, why not subscribe to have these analyses delivered to your inbox every Friday?

Sign up


We don’t usually have views and opinions about economic and financial states of affairs, (not ones that we express publicly as a company, anyway). We do believe, however, that people can and do appreciate a variety of perspectives. What you’ve just read is the perspective of our resident chief economist. While we think he’s very smart, Macrobond Financial does not expressly endorse the views he presents here. And, as the old adage goes, you shouldn’t believe everything you read (not without finding the data, performing a few analyses and presenting it in a nice chart). We want to make it clear that we are not offering this information as investment advice. That being said, if you have the application you can easily check everything that’s mentioned here, and decide for yourself. If you don’t have the application, now you have a great reason to get it.