Are we missing the point?
In recent times, I have heard from academics and those inside industry who worry that accelerated automation will contribute significantly to economic problems that already exist and which are squeezing the American middle class. I'm not convinced it is that simple.
When I critiqued Silicon Collar I felt that the author ducked an important issue around the social impact and the economics in play. I said:
The problem with the arguments that Mirchandani puts forward are that they are rooted in variations of the economic status quo, regardless of whether you are ‘left’ or ‘right’ leaning in your assessment of economic theory.
In that sense, Mirchandani doesn’t consider alternatives to those theories, recusing himself on the grounds that he doesn’t have a good answer to that or, the emerging question of ethics in technology.
For his part, Mirchandani is very clear that he views the world from a much more optimistic framework, even going so far as to suggest that the pessimists are little more than whiners. There is nothing inherently wrong with that, provided one accepts that it is a single vision that may be as flawed as it is attractive.
With those thoughts in mind, I was interested in hearing Martin Ford, author of Rise of the Robots: Technology and the Threat of a Jobless Future, who had a session at the recent BoxWorks event and which took its title from the book. The blurbs for it said:
We might imagine and hope that that today's industrial revolution will unfold like the last: even as some jobs are eliminated, more will be created to deal with the new innovations of a new era...As progress continues, blue and white collar jobs alike will evaporate, squeezing working and middle class families even further.
It is extraordinarily brave of a technology company like Box to put up a session of this kind. Why? Box and BoxRelay are all about automation. It's a technology advance I like and can see us using at some point. Does it threaten jobs? Possibly.
Crush that overhead
From my standpoint, it might mean that we don't have to employ someone to undertake important but repeatable tasks. That's an opportunity cost saving that has a hidden impact on the economy.
Indeed, when I think about the decisions diginomica has taken over its short life span and the fact that we have not added a penny in headcount overhead to run the business in that time then it is safe to say that we eat our own automation dog food as rabidly as any other business.
Are we denying the potential for those who might fulfill a useful function for us? I don't think so when we collectively view our prime requirement as top quality analyst/writers and, are prepared to pay well to achieve that end.
In short, our focus is not on job elimination but on expansion of the right kind of work for our business model. But then we are a service business. We don't produce 'stuff' as such that might require people to manage machines.
In both his presentation and his book, Ford is very clear in his argument that automation has made a major contribution to the growing gap between productivity growth and the growth in incomes. By way of opening illustration, Ford showed a graphic from the Economic Policy Institute paper of a year ago titled: Understanding the Historic Divergence Between Productivity and a Typical Worker’s Pay - Why It Matters and Why It’s Real, (see above) which certainly makes the jaws drop. Remember this is an all encompassing set of metrics, it is not restricted to mass production industries.
The end of accounting
Ford then went on to show a succession of graphs from which you can conclude that many sectors within the economy will continue to see an adverse impact on employment. One example that stands out is book-keepers and routine accounting. Ford cited a WSJ article from 2015 titled The New Bookkeeper Is a Robot which demonstrated the decline in the number of finance staff required per billion dollars worth of sales. (See image right.)
That decline is set to continue as we see more automation enter the finance department, improved business processes and the relentless pursuit of cheap offshoring. As Phil Fersht, CEO Horse for Sources would likely say: Robotistan is real.
In 2006, I predicted that the accounting profession - as opposed to those accountants employed by industry - would see a dramatic decline in the requirement for book-keeping as a service. This followed the emergence of services like Xero and others. It is a trend that is continuing although not at the pace I imagined. Yet.
Counterbalancing that decline, I also predicted that we would see the rise of a new type of accounting firm that seeks to add value to the basic accounts and tax work that has to be undertaken for compliance and regulatory purposes. There has certainly been some genuine innovation in that regard but by and large, the profession is tenaciously hanging on to the old models for as long as possible. It cannot last forever.
Brian Sommer gets it right when he says that our ability to analyze, understand and respond to changing conditions will absolutely require a reliance upon automated systems that do far more than crunch the numbers. In Mirchandani's book, he goes further:
But, as an industry, accountants don’t handle change well or quickly. They like to dissect every potential change, be it a new FASB [Financial Accounting Standards Board] accounting rule proposal or a new technology. The automatic reaction to any potential change is to not change but instead go into ‘study’ mode. This pace, often glacial in nature, was a luxury that could be humored when business, technology, and global economies also changed at a similar, slow pace. That’s just not realistic anymore.
In ongoing conversations with finance chiefs in every industry, the conversation is always the same - get rid of routine functions so that we can concentrate upon higher value work that helps the business. See the tension?
The inexorable decline in labor's shareFord's book makes the plausible case that all corners of industry will be impacted by automation where there are predictable processes and that labor will continue to be displaced. During his discussion on the topic at BoxWorks, he noted that the general decline in labor's share of the economy is not US specific but that you can find clear evidence in many major economies. (See image left)
We can readily imagine the thesis to be correct because as Ford correctly points out, there has been a steady decline in the number of people needed to make the same amount of 'stuff,' a trend that can be clearly found in the narrative around productivity improvements from the 1950s through to the present day. But as someone said to me yesterday, "It is rare for there to be a single cause for any condition."
Ford doesn't quite fall into that trap completely, reviewing factors like the erosion of labor rights and globalization, but I question the prominence of automation given the timeline shown by EPI in my first illustration. Is that an unreasonable position given what we are seeing in the marketplace?
To put this into perspective, I remember the first electronic calculators coming into the accounting offices I worked in during the 1970s but it wasn't at least until the mid-1980s that I saw the start of the dramatic rise in the use of desktop computers and the replacement of autoworkers with robots.
But then I equally remember the political climate of the day which saw labor unions in both the UK and the US crushed under political pressure. I also remember the wild swings in interest rates as successive governments tried to first accelerate and then put the brakes on the economy. Those old enough to remember may recall the derision with which politicians talked about the 'stop-start' economy.
To my disappointment, Ford didn't explore those other things in his session. That was a lost opportunity in my mind, but then I fully understand why he would choose not to step into that particular minefield. When I spoke with him afterwards, I asked about economic and policy topics but he wouldn't be drawn. My sense is that while happy to explore at a superficial level, the complexities of how economies works weighs heavily, something to which the book alludes. In that, I agree.
Economic theory isn't helping
Where Ford gets it right is in his characterization of the current state of economic thinking. Here it is in a simplified nutshell.
On the one side we have those who think that the cure for all ills is a combination of low taxes, light regulation and allowing the market to do its work. Some might recognize it as Reaganomics and Thatcherism.
On the other side are those who say that the consequences of what happens with the first group theory is the creation of a polarized society where massive inequality is the norm. They cite studies like the PI one to account for the massive increase in wealth concentration and the stagnation of compensation among the American middle class. These same thinkers believe that there comes a point where improved productivity doesn't matter any more because demand is too weak to support consumption snd that ultimately, the economy fails. In that scenario, it doesn't matter how much automation is in place. Demand has no way to keep up with supply.
The difficulty in my mind is that these economic positions are so far apart from one another that it is impossible to see how they can be resolved without some kind of cataclysmic event. In the book, Ford says that:
...professional economists - all of whom have access to the same objective data - are completely unable to agree on what I would characterize as an extraordinarily fundamental economic question: Is a demand shortfall holding back economic growth and if so, is income inequality an important contributor to the problem?...in the field of economics, the opinions all too often break cleanly along predefined political ones...In other words, if you're waiting for the economists to deliver some sort of definitive verdict on the impact that advancing technology is having on the economy, you may have a very long wait.
Ford considers a variety of possible scenarios, including the possibility of the dystopian future illustrated in Elysium. While possible, it's not an image I want to entertain with any degree of enthusiasm.
Help a brother?
Does Ford have a solution? Sadly, he doesn't offer much beyond the sop of a Universal Basic Income as a way of at least alleviating rising inequality. You can read plenty of debates around this hotly debated topic. Here, I believe the WSJ's Greg Ip sets the scene nicely but comes to some nutty conclusions alongside an economic doomsday scenario.
I have no position on this other than to postulate that however disagreeable to many, UBI may be the only short term measure that stimulates demand while partially redressing the inequality issue.
Ford goes on to talk about topics like wealth redistribution, the need for retraining, access to education and the requirements of a burgeoning healthcare industry. All of these are laudable and frequently debated topics. But I wonder whether all of these proposed solutions are just as rooted in the polarized economics of today and therefore flawed. Regardless, I didn't detect a great deal of enthusiasm for any particular measure and a preponderance of ifs, buts and maybes suggest that while happy to surface ideas, Ford is on very uncertain ground when it comes to solutions that are sustainable.
Pay up, suck it up?
Others are looking for a fix but not necessarily offering a complete solution. For example. raising the minimum wage is another ongoing and hotly debated topic. Here, I am kind of persuaded by Tim O'Reilly's argument where he says:
...a $15 minimum wage might cost Walmart on the order of $5 billion/year. This is no small number. It represents about a quarter of Walmart’s annual profits, and about 1.25% of its annual US revenues. But it might save taxpayers $6 billion per year (and that’s just the amount used to subsidize Walmart; including all the other low-wage employers in America, the number is far larger.)
If Walmart weren’t able to pass off part of its true costs onto taxpayers, the company would have to accept lower profits or raise its prices. But is that really such a bad thing? Let’s do some back of the napkin math. If Walmart’s profits were reduced by $5 billion (approximately 20%), its market cap might fall, a loss to shareholders. But leaving aside the shock of a sudden drop in earnings due to a change in the rules, would the owners of Walmart really not have wanted to own it if it generated $20 billion a year in profit instead of $25 billion?
If Walmart were to pass along the additional costs to consumers, prices would have to go up by 1.25% (or $1.25 for every $100 spent at Walmart.) If the costs were split between capital and consumers, that would require only a 10% drop in Walmart profits and an additional 62 cents per $100 spent by consumers. Would people really stop shopping at Walmart if they had to spend little more than an additional half cent for every dollar?
Those higher prices might discourage some customers, but the higher incomes of workers might encourage them to spend more. So it’s not inconceivable that Walmart and its shareholders would come out whole.
The problem is that any minimum wage cuts across everyone. Most recently we saw this in San Diego where a rise in the minimum wage led to an almost immediate hike in beer prices at the local bars. First world problem? Sure. But it didn't go un-noticed.
I'm a huge fan of automation. I'd hate to have to go back to the world of accounting 'tick and bash' or the mindless keying in of bank statement entries. Other than when a Tesla goes on fire, we almost never talk about vehicle reliability these days. That came about because automated machines can make things with a precision that's impossible for humans to accomplish at scale. Automation in many industries has certainly created an environment where it is possible to make 'stuff' at unprecedented speed.
On the other hand, there are plenty of questions abut how the algorithms that pre-select job candidates contain unintentional and unconscious bias but I have to be optimistic in believing those problems will be worked out. And then there is the emerging question of defining what artificial intelligence really means.
Whether you take Mirchandani's optimistic vision of a world of plentiful choice or Ford's statements of at times depressing inevitability around the impact of automation, it's not enough to have a fleeting understanding of the economic impact in a defined or narrow sector. As an industry, we have a responsibility to widen the debate to include questions about exclusion and inequality. The odd philanthropic gesture won't cut it.
We have so far, been perfectly prepared to champion the cause of disadvantaged LGBT persons and the disgraceful discrimination against women. But we have not begun to explore the outcomes that technology brings beyond shrugging at the obvious displacement of jobs or tacitly accepting the assumption that everything will somehow work out OK. Evidence is mounting that large parts of the workforce are deeply unhappy about this laissez-faire attitude.
I believe we are at a point where we need to open up the debate that are inclusive of broader factors that play into the complexity that is real world economics. How do we conduct that debate without becoming mired with pre-conceived notions of political bias and the all too emotional responses that go with one or other position? That is a question for an other day.
As a final thought, a paradox is emerging for which there is no simple answer. Much of the past efforts in automation have been about cost reduction and efficiency. As the evidence suggests, this is something that, along with other factors, has worked out very well for those who own capital. That trend will continue because it makes so much sense to those who are makers. In recent times, the information technology emphasis has been on harnessing data and improving processes aimed at getting us to consume more. Think omni-channel. The current evidence suggests that demand is stalling. What happens then?