July 18, 2017
There’s something unsettling about finding yourself at home in bed, under the duvet, Pringles in one hand, remote in the other, binge-watching ‘House of Cards’.
As you try to come to terms with where you have ended up, can you take some comfort in the fact that a significant chunk of the reasonably well-educated bourgeoisie is also addicted? Are we, at all hours of the night, gorging on Netflix, checking emails, responding to Whatsapp groups, booking an Uber or buying something on Amazon?
Is technology taking over our world and what does it mean for the global economy and for our way of thinking about the world?
In this update, let’s tease out some of the connections – not always apparent – between technological innovation, low productivity, wage deflation and the conundrum faced by central banks everywhere. The central banks’ dilemma is that they are itching to raise rates but finding no reason to do so.
Economics is constantly evolving
At the beginning of every year, when my students sit down in front of me at Trinity College in Dublin, I can’t stress enough the importance of not only what they’ll learn in textbooks, but what they see happening in the world around them. As the world changes, our understanding of the economy should change too. But sometimes it doesn’t.
Too regularly, economists – and in particular policymakers – display a form of ‘groupthink’ and seem wedded to old models that appear impervious to the changes that are taking place in the real world.
Groupthink, when wrong, can lead to huge policy errors and have enormous unnecessary repercussions. As the global economy is growing and unemployment is falling, the default position of central bankers, from the Fed to the BoE to the ECB, is that inflation has to be around the corner.
But, what if it’s not?
What if policymakers in their ivory towers wedded to old ideas do not appreciate that technological change and the likes of Whatsapp, Uber, Amazon and Netflix, are driving deflation not inflation?
Both issues, (1) the deflationary impact of technological disruption and (2) the productivity paradox, are interrelated and have huge policy implications. Let’s explore them and more, but for the sake of clarity, take each one in turn.
When is a curve not a curve?
Now it might be helpful to remind ourselves that central bankers believe in the Phillips Curve. The Phillips Curve is the traditional relationship between inflation and unemployment. Every religion has a creed – a set of core beliefs. For central bankers, the creed is the Phillips Curve.
The mantra goes like this: as unemployment falls, more and more firms want to employ more and more workers, but there are fewer and fewer workers available. So wages rise. The worker knows that if his boss doesn’t give him a rise, he can go across the road to the guy with the ‘vacancies’ sign in the window and get paid more. This process triggers wage inflation.
At some point the economy runs out of workers and wages rise. Traditionally, this threshold was an unemployment rate of between 5% and 6%.
While higher wages give more income to workers, they constitute higher costs to employers and ultimately less profit. Because companies traditionally work on a fixed profit margin basis, as the employer’s costs go up, the company simply increases retail prices to sustain profit margin. And, because everyone is getting a pay rise, everyone has more income. This increased income allows the market to absorb the increases in prices as everyone has more money.
Therefore, the traditional mantra of central banking is based on two crucial beliefs: first, the ability of the worker to push up wages when unemployment is falling (or at least falls to a certain level) and second, the subsequent ability of the employer to pass on the increased costs through higher prices.
Years ago, as a rookie economist at the Central Bank of Ireland, my first job was to assess these inflationary relationships. Sure we did lots of research, but my favourite was ‘crane counting’. This highly sophisticated approach involved standing on the top floor of the bank, one of the tallest buildings in Dublin (a low-rise Georgian city) and counting cranes on the horizon. The more cranes, the more activity, the higher future wages in construction and somewhere down the line, we could be sure of an uptick in inflation.
Common sense works – or at least it did.
However, the relationship between unemployment and inflation appears to be breaking down or at least shifting. Could technology be the reason?
Unemployment is falling everywhere but inflation isn’t rising. In the US and the UK unemployment has been falling and falling but wages have remained depressed. Even in the eurozone, which has created more than five million new jobs in the past three years, wages remain depressed.
This creates problems for the major central banks because after years of QE they are trying to ‘normalise’ before the next downturn. By ‘normalising’ the bankers mean they want to unravel QE, shrink their balance sheet and raise interest rates. They want to get back to the pre-crisis status quo.
The last thing the central banks want is to enter a new recession with rates close to zero because if rates are already at zero, how can the central bank fight a recession?
But in order to raise rates they need an inflation trigger. The dilemma for the central banks is that there is no inflation, even though unemployment is below the level (5-6%) where we traditionally see rising wages.
This defies economic wisdom, particularly in the US, which is about to enter its ninth year of economic expansion and unemployment is headed towards 4%.
So what is happening?
Why are wages and inflation in general not rising? And what might watching Netflix have to do with all this?
To explain this, let’s go back to theory. The Austrian economist Joseph Schumpeter came up with the term ‘creative destruction’ to explain the natural process whereby recessions and human ingenuity create innovations and these innovations produce ‘disruptive technologies’. Such disruptive technologies like Uber, Amazon, Whatsapp, Google maps or Netflix, destroy old industries, slash prices and change the ground rules. Schumpeter argued that these forces are so strong they drive down inflation structurally.
In terms of Uber, Whatsapp and Amazon, these new technologies are driving down prices undermining the old fixed margin businesses. Taxis, retailers, telecom companies and all sorts of leisure companies can no longer add margin to their price to make up for increased costs.
So the issue is what if Schumpeter is right and Phillips is wrong?
In a sense, Schumpeter argues that we should embrace change because it is invigorating, while the central banks using Phillips are constantly trying to deploy policy to get us back to where we were before.
Could we be in a Schumpeterian world of technologically-inspired deflation, where the old relationship between wages, employment and inflation has broken down?
For this to be the case, something profound also has to be happening to wages.
Amazon undermines the fixed profit margin model of retail, but what if some other deflationary force is going on in the labour market even before the company worries about margin?
This brings us to robots and the productivity paradox.
The productivity paradox
Part of the ‘technology is everywhere’ story is the notion that robots are taking our jobs. This sounds plausible, very 21st century and is part of the general giddiness about Silicon Valley, tech stocks and general market hype. But the economist in me is concerned; if robots are really taking our jobs, why is productivity so low?
Robots replacing people should drive up productivity, but productivity clearly hasn’t read the script. Furthermore, because productivity increases determine wage rises, depressed productivity means depressed wages.
But, you might ask, are we not surrounded by all this amazing technology that is changing our lives? Well, yes and no. Clearly in entertainment, lifestyle and media, new technology is everywhere. And because we are all using it, like my Netflix addiction, we feel that it is more ubiquitous than it actually is. The hard data shows that technology is taking over not so much our lives, as much as taking over our spare time.
If we were actually living through enormous technological change four things would be happening. First productivity would be rising as we use this fabulous new technology to create more things. Second, wages would be rising too. Third, employment would be falling because companies would be investing in cheap robots to take over from more expensive humans. And fourth, all this investment in new technology would be captured in a surge in business investment.
But none of these four things are happening. Actually, the opposite is occurring.
More and more, rather than less and less, jobs are being created. Productivity is not exploding but stagnating. Wages are depressed and business investment is not surging but is, in fact, tepid.
So, it’s not that we are embracing technology but we aren’t embracing it enough. It’s not that there are too many robots out there – there aren’t enough of them!
Incidentally, when you think about it, the robot problem – employed workers getting paid more to do less – would be a ‘nice’ problem to have. Instead we are experiencing the opposite: workers are getting paid less and working more.
So why is this?
It has to do with three potential factors.
The first is structural. In the past thirty years, each recession has been taken as an opportunity to wring more inflation out of the system. In plain English this means that the worker pays. In a recession unemployment rises and wages fall and when the economy finally turns and the worker gets up off the canvas, he does so at a lower real wage than before. This outcome is policy-inspired wage compression.
The second factor is that technological disruption has made workers cheap all over the world. Technology allows companies to scour the world for cheaper workers and there has therefore been a convergence of the cost of labour around the world. Granted certain professions have done well as a result of technological change but, on average, wages have been kept in check due to competition. The global supply chain means that the global supply of workers, rather than the national pool of workers, sets the national price of labour and thus wages.
You could say the labour force has been Uberized!
You could also look at the productivity paradox from another angle. Many believe that productivity is low because investment has been weak, so workers have not had access to enough capital investment. But what if it’s actually the other way around? What if investment in technology, which drives productivity, has been low precisely because wages are low?
We are seeing low levels of productive investment because there’s no real need for companies to save money as workers are already cheap. The financial imperative for a company to invest in order to lower costs has gone away.
Therefore the causal relationship is from wages to technology, not from technology to wages.
It’s not that there is no productive investment going on, but it is not as much as the tech evangelists of Silicon Valley might have us believe. We are seeing more of the same. There’s nothing new about humans being replaced by machines. Ask the Luddites. It is the story of the global economy and will continue to be.
When those workers lose their jobs they don’t disappear, they try to get other jobs. Typically, they slip backwards into jobs that are less well-paid because they still have to put bread on the table. This competition also pushes down wages. And, the lower wages are relative to profits, the less investment there will be because companies won’t bother to invest in expensive machines when people are cheap. This same type of rationale explains the industrial under-development of the Southern slave-holding states relative to their Northern counterparts in the Antebellum US.
As a result, we get low productivity being driven by low wages, not the other way around.
Societally, this leads to higher levels of inequality between citizens who depend on wages for their income and those who depend on rents, dividends and assets. Politically, this drives populism as the disenfranchised worker might not have a good wage, but he has a legitimate vote.
Now let’s go back to the central bankers’ immediate dilemma.
To raise, or not to raise…
Despite central banks’ desire to tighten monetary policy, persistent deflationary pressures appear to be the problem, rather than incipient inflationary stresses. This is being driven by disruptive technology, which is undermining the ability of firms to pass on increased costs. This is keeping retail prices low.
Then there is the productivity paradox. How in a world of mass technology and robots, can wages be so low? Rather than taking our jobs as everyone is suggesting, robots aren’t taking enough. Wages are not low because productivity is low but rather the opposite is the case, productivity is low because wages are low.
Low wages mean cheap workers will do the job just fine, while the split between profits and wages will continue to go increasingly to profits. This split towards profits rather than wages has profound ramifications for the general economy because if wage growth remains weak, demand remains weak because rich people don’t create demand – or jobs for that matter.
Poor people spend, rich people hoard.
The average guy with a good wage, buying little bits all the time creates demand. Rich people in general hoard money rather than spend it. Therefore, if more of the return from the economy is going to dividends than wages, this money doesn’t filter down through the economy in the same way.
In Keynesian terms the poor guy’s multiplier is much larger than the rich guy’s. You get a much bigger bang for your buck from the poor guy than the rich guy.
Therefore, aggregate demand in this cycle has remained fragile.
So what’s the rush to normalise monetary policy? Why the rush to get back to the pre-crisis status quo?
It is because of ‘groupthink’ deep in the central banks. They want to get back to the status quo because that’s the way it’s always been. It’s an intellectual form of mean reversion!
There is no need to rush. In fact, the immediate issue isn’t inflation. It’s deflation, implying that the rate of unemployment can go lower and lower before wages begin to rise. The unemployment rate at which wage inflation rises could be 4%, it could be 3% or even lower!
If central bankers rush to raise rates mistaking a structural issue for a cyclical one, they may end up doing untold damage.