This week’s leader in the Economist is titled, “A Boom Like No Other”. It is a critical review of the promise and progress of Donald Trump’s administration and its efforts to unleash an economic engine. The article explores the progress across three dimension:
- The state of investment
- The state of employees
- The state of competition
The newspaper argues that these three factors need to change in a certain way for a worthwhile productivity jump to be realized. I won’t, of course, focus on the political side of his analysis for there are no winners or losers in doing so. I will however comment on the economic data used in the article and what they mean.
Firstly the Economist shows that investment is up. Economy-wide investment is up 7% in the first quarter of 2018 compared with the same period the year before. Average investment growth was about 10% during the surge of the mid-2000s. so 7% isn’t too shabby. Of this, big listed firm investment (capital and R&D) is up 19%, according to analysis of the S&P 500. That is impressive, even scary. But then the newspaper pours scorn on the data when looking deeper.
Apparently the make-up of the firms leading the investment charge has changed. Where once it used to be AT&T, Chevron, ExxonMobil, and GE, now it is Alphabet, Amazon, Apple, Intel and Microsoft. The IT sector is therefore driving the lion’s share of investment. Taking this data in hand, the article argues therefore that investment growth is not broad-based but skewed to a lucky few sectors. I do not regard this is bad; indeed it is needed and natural. Let me explain why.
The cycle we are in today is little different to what has happened in the past but too few people are aware of it or its importance. The IT sector is going through its own re-generation. Expensive server-based, on-premises and data center-based technology is being re-engineered for the cloud. Over a period of time this will save the IT industry a lot of money and as such will drive the productivity growth in the IT sector. This will create two follow-on benefits in subsequent time periods.
First, IT costs for current work will fall. As an imputed cost to doing business in the private sector and serving citizens in the public, IT costs will fall and so productivity related to work and business processes using that IT will the increase in downstream consumers of IT services. This cycle has been seen before but it takes years to work its way through.
Secondly, and this is the scary part, beyond ‘work today’ we will see new work and innovation. This is another hard cycle to understand and predict. IT is also innovating over and above “cloud computing” which is, after all, little more than a sourcing factor. General purpose technologies (GPT) like AI are going through a renaissance and being applied to specific business problems and opportunities. These will yield special purpose technologies (SPT), or what some will call product or service innovation, and many of these will be disruptive to established work.
This will both increase business value and cut costs at the same time. Thus a second wave of change will follow the first, only the second wave will be larger and more broad-based beyond IT then the first. Look here for an explanation for the various cycles and waves of innovation that may evolve: Where You Spend Your Firms’ Capital Matters. The question is, which firms, downstream of IT, will move first?
There are other investment growth sectors beyond IT too. The energy sector is “jumping off” as well as foreign firms investing in America, though focusing on intellectual-property intensive industries such as tech and biotech. This last note should not be news to readers of this blog: we have talked of this factor before and it’s role in productivity dynamics. See Book Review: Capitalism without Capital – The Rise of the Intangible Economy.
On the competition front it is generally bleak news. Industry concentration remains at near all-time highs; new start-up figures are up slightly, though mostly at the same levels as noted under Obama toward the end of his presidency. The problem here is more political: huge sums of money tied up in lobbying firms who reinforce the position of the largest firms, and the investment cycle has been turned upside down with Dodd-Frank, near-zero interest rates and quantitative easing.
There are small signs things may soon change. The recent revision to Dodd-Frank removes many cost-prohibitive burden from small and midsize banks. This will free up those banks to start lending again to small firms; neither group- small, mid-size banks and small start-ups, were per of then financial crisis that Dodd-Frank was meant to resolve. So Trump’s political action may in fact simply undo an unintended consequence, for the good. But the investment side of the economy needs, and quickly, interest rates to recover (to help the market return to its pseudo rational expectations) and the Fed to dump its balance sheet of bond assets (to help the investment/margin side of the market to shift from money to investment).
On the labor front things are also staring to look interesting. “The absolute pay bill for all firms rose by 5%” in Q1 2018. This is above the headline rate of inflation. More interestingly is the fact that the share of gross corporate profits that is paid to workers has risen to 78% from a low of 76% in 2014. This is huge and won’t be liked by Mr. Piketty. The choice is to accrue profits to labor or capital so labor is now growing its share. As interest rates and the Fed balance sheet come down, this rate should increase further. It will do so until the long term capital investments being made now start to payoff in future productivity growth.
This will alter the share in two ways. As automation improves so costs of labor go down and capital will accrue more of the profits. But beyond automation, process and newer product/service innovation that will be a mix of capital (information and/or technology) and labor, will fight for a bugger share of profits.
In conclusion the Economist suggests the current tech boom is like no other. It turns out that is more like the others than not. It just depends on how you interpret the data.