On Interest Rates, Technological Job Automation and Inflation
Recent developments have meant it’s time to take a look at the wider economic picture again. Yes that’s right, we’re talking about the US Federal Reserve hiking interest rates last week as concerns of an overheating economy come to the fore, the World Economic Forum’s report that artificial intelligence and robotics will overtake humans at performing tasks in the workplace by 2025 and Jim Cramer, famed host of CNBC’s Mad Money show claiming that the Fed is wrong to hike rates given technological disruption of the workforce by Silicon Valley.
So, are we today facing a Fed constrained by conventional thinking and overlooking the real issues? It’s important to note that as we predicted in 2016, US monetary policy has continued to tighten apace with President Trump even going as far as recently saying that the Fed was his “biggest threat” due to the fact that it keeps raising rates. The problem is that one of the fundamental priorities of most central banks (and of course the Fed is no exception here) is to balance the pursuit of full employment with the stability of prices. The US system for governing interest rates is an applauded model. Other countries for much of their existence (such as the UK until 1997) had politicians firmly in control of the levers of monetary policy which is viewed by many as undesirable given the tendency to use it as a tool for political gain rather than economic wellbeing which is obviously what it should be.
The overall balance should be such that full employment should be sought, but in the event that it’s actually achieved it would probably be detrimental to the overall economy as at that point employees would become so valuable that they would demand unreasonable pay increases and inflation would go up as a result which is in direct conflict with the other part of the Fed’s remit to maintain stable prices. At a macro level, some degree of unemployment is a good thing as it means there is slack in the economy to move in a new direction as and when needed, the objective is that those unemployed should not be long term unemployed, but those rotating between jobs and skillsets.
Setting monetary policy then is a balance between conflicting requirements which is as it should be. It means that those weighing the decision should be regularly analysing the state of the economy in relation to those requirements. The trouble is, it also has another effect which is demand for the currency since if interest rates are higher, generally that means a higher rate of return can be achieved on bonds and the like bought in that currency among other factors. It’s also worth noting that while President Trump decries the Fed, he’s also likely pleased with some of what they’re doing in other areas such as rolling back financial regulation which restricts banks.
To Job Automation
We’ve written about this several times in the past and it’s an undeniable fact. Jobs becoming redundant due to technological advancement has been taking place for a long time. The question really becomes one of whether job automation leads us to a post-scarcity future where everyone can just sit around (unlikely) or whether the automation will lead to other jobs. The WEF surveyed some companies and found that approximately 29% of tasks in the workplace are performed by technology but that the expectation is this will rise to over 40% by 2022 which is a staggering number to think about.
That’s not necessarily the end of the world. Technology is often more consistent than humans, particularly for mundane tasks and tasks where extreme levels of precision are required. It’s clear in my daily life when it seems that in any given month, I can walk into yet another shop or post office and see less people working a counter and more machines where I serve myself. Many of these jobs which are being automated out of existence then are relatively low paid and menial, technology has aimed at these as the low hanging fruit for while they don’t pay much in terms of the work they are replacing, they are a significant cost burden on many companies since the sheer numbers of staff required for these jobs mean that they’re huge drags on the bottom line.
Professions however should also beware. As I’ve written about in the past, automation is coming for them too and AI stock analysts, lawyers and doctors are already being refined on a daily basis. The job market then is being attacked at both ends and across the spectrum. In the short term, it’s difficult to see the roles of company officers and other senior managers being given up to machines as the overall process for corporate governance still has a degree of punishment built into the system in the event that senior managers fail in their duties and it would prove difficult to fine a computer or send it to jail in the event of a catastrophic failure of oversight.
So the question really becomes, what happens to the people who lose their jobs due to automation? The theoretical objective here is that the automation itself will create more jobs and indeed the WEF report notes that although almost half of the firms surveyed think they will reduce headcount due to automation, 38% of them think they will increase their employment numbers in new roles which enhance productivity and therefore margins and bottom line.
The WEF surveyed companies representing approximately 15 million workers and of those surveyed, the trend indicated a reduction in existing jobs of just under a million with a gain of new jobs at about 1.75 million. Unfortunately the WEF has then taken an overly simplistic view and simply scaled this to the global economy (excluding agriculture) and concluded that we would look at 75 million job displacements and 133 million new jobs. Realistically of course this is unlikely to scale exactly but theoretically, the net effect is positive.
The difficulty when it comes to this is whether these new jobs are the kinds of jobs which menial workers (many of whom will be displaced by technological automation) are capable of retraining into. I’ve talked many times in the past about how human intelligence likely approximates a normal distribution and if indeed that’s broadly accurate and the menial jobs being displaced are people at the lower end of the distribution, the likelihood of them all being able to retrain into productivity enhancing roles seems slim so this does lead to the likelihood that technological automation may lead to a longer term unemployed or excess of supply at the lower end of the skill spectrum, further depressing wages of this job profile.
Jim Cramer then ultimately argues that inflation isn’t a problem and the Fed shouldn’t be hiking interest rates further (which the market is obviously expecting, particularly given the hawkish comments which were revealed in the FOMC minutes revealed last week) due to the technological innovation which is coming out of Silicon Valley and will give more slack to the economy. The basic premise then is that even if we run out of people to do jobs, don’t worry, technology will step in and do stuff which means we won’t have inflation. The difficulty this thought process presents are unfortunately numerous as although there is a degree of truth in the fact that automation is putting people out of work, if we take the WEF research at face value, even more jobs will be created leading to an overheating job market, at least potentially in the higher skilled arena.
Couple this with the fact that inflation is what’s known as a lagging indicator and we have a problem. The Fed can’t be seen to sit idly by while the economy overheats and then inflation subsequently runs rampant. Lagging indicators are so called because they tend to only surface after the economic trend has already been determined. So in this case, the inflation will only be evident once it is too late to do anything about. This makes sense if you think about it at the micro level. Suppose you run a small business and all of a sudden you can’t hire people anymore at the rate you used to hire them for. There must be a lot of demand for staff. Assuming that your staff have a reasonable understanding of how much their skills are worth in the marketplace, they may demand a pay rise to stay, your costs are going up. How do you still run a business with costs going up? You’ll likely need to eventually increase prices, if this happens across enough companies and industries, inflation goes up.
So what do we have, a strong economy, prices already rising due to the potential currency manipulations of others and the tariffs we are imposing on imports but not too significantly (yet), strong job numbers and (relatively to global) high interest rates leading to higher demand for the dollar meaning that you have more purchasing power from other countries, but also that your exports are not as competitively priced now so your balance of payments for international trade situation worsens which is a problem that Trump is trying to solve with tariffs… you can see where this is going.
Ultimately, the Fed is doing what is probably the right thing. The difficulty of course is that it won’t really be visible if it is doing the right thing or not until it’s too late to change to hit that particular economic cycle. Job automation may solve some of the problem but is unlikely to fix it all.
There are longer term affects of interest rates going up, if US government bonds are paying higher rates, it is of course inevitable that corporations and individuals will also have to pay more to borrow money. For companies like NVIDIA (NASDAQ:NVDA) with a decent cash pile and margin in its business this isn’t the end of the world but for others like Tesla (NASDAQ:TSLA) who are still loss making and running on empty with a possibility of needing to raise cash and bonds coming due, this may be a problem. We’ll keep an eye on the situation as it develops.