When More Is Not Better: Book Review Part One
When More is Not Better, the new book by Roger L. Martin, will be published on September 29th by Harvard Business Review Press (Pre-order). Roger is professor emeritus at the Rotman School of Management at the University of Toronto and was formally the school’s Dean. In 2017 he was also named the world’s leading management thinker by Thinkers50. He has written 11 previous books and is a strategy advisor to the CEOs of Procter & Gamble, Lego, and Ford.
Overcoming an Obsession with Economic Efficiency
The new book is focused on “overcoming America’s Obsession with Economic Efficiency”. He notes, “our obsession with economic efficiency has featured too much pressure, too much connectedness , and too much pursuit of perfection, all of which has produced a dangerously unbalanced economy lacking resilience”. He points to the fact, “our dominant model of the economy is that of a machine” and suggests the blame for that can be directed to economics generally, and to Professor Wassily Leontief, a former Nobel Prize for Economics winner, and a former professor of economics at Harvard Business School from 1931–75.
Leontief left his mark on Harvard, where Martin himself started studying economics just two months after Leontief’s departure. Martin says he graduated in 1979 thinking he knew how the US economy worked, but soon learned that he had been taught only one model of economics and it was wrong. He admits he was “naïve to the power of models to shape action and the power of metaphors to drive the adoption of models”. He believed the models he was taught were descriptions of how the economy actually works, not how they might work. His professors “oversold the veracity of their model”, and that still irks him to this day.
All Models Are Wrong
Martin notes that he later learned about the “foibles of models” from John Sterman, MIT professor in systems-dynamics who pointed out “all models are wrong” in an article with that title. But Martin also notes we cannot operate without models, whether we realise it or not, and most of the time we do not. Models simplify the complexity of the real world to help us make decisions and take actions.
Martin goes on to note that we use analogies and metaphors to help us make sense of the new. “We select our metaphors, and afterwards, our metaphors shape us”. They do so because different metaphor drive adoption of different models that produce very different outcomes, says Martin.
Metaphors are the second core component of four in the core components of models, argues Martin. The first is a desired outcome. The third is a cause-and-effect sequence to get to a desired outcome. And the fourth is a desired proxy, or proxies, to measure progress towards the desired outcome.
He also explains the above in a slightly different way, based on the insights of Sterman. “When we have a goal in mind, we create or choose a model to pursue that goal , and the model will indicate proxies that we can use to measure progress. We do so whether we are conscious or not of building and deploying a model. And the model we create or chose is almost always grounded in a metaphor we can easily relate to.” All of which we often do unconsciously.
Machine is a Bad Metaphor for the Economy
We did not always view the economy with the metaphor of a machine and, whilst Leontief did not use the machine metaphor himself, his thinking produced that result. The analogy provides a sense of reassurance, that the economy can be steered or controlled, even though the evidence suggests otherwise. The idea evolved to influence job and organisational design, and the teaching and operation of business by functional siloes — too often overlooking the fact the whole is greater than the sum of the parts.
None of the above represents fresh thinking. Martin acknowledges this by suggesting Peter Drucker, among others, warned of the problems the machine analogy and the misuse of models would create. But, despite repeated warnings, the machine analogy did lead to management thinking obsessed with the efficient performance of the machine. It has dominated the whole history of management theory and practice.
Taking friction or slack out of the system, maximising the return on assets, minimising costs — they have been the focus of management since the idea of the division of labour and economics of scale, espoused by Adam Smith. They were the focus of Frederick Winslow Taylors scientific management, and of later management theories, tools, and models such as Business Process Reengineering, Total Quality Management and Just-In-Time, for example. Since they are all focused on the bottom-line of the profit equation, I have suggested the outcome has been what I call Denominator Capitalism, rather than Numerator Capitalism.
The Metaphor Models and Measures are All Wrong
Martin observes that the result is a lack of resilience. Businesses become fragile. Shocks such as the global banking crisis of 2007/8 and the current Covid-19 induced crisis expose the lack of resilience, the fragility, of firms. More worryingly they also expose the lack of resilience of the economy as a whole. But the main argument made by Martin is that the mechanistic analogy is wrong, and our models are therefore wrong.
The associated problem is this, “the proxies [measures] that we have adopted for measuring and driving efficiency are turning our pursuit of efficiency into a destructive force”, says Martin.
He states that proxies, measures of progress towards goals, are important in this context. He notes “they have become indistinguishable in the minds of their users from the efficiency they are supposed to measure”, distorting outcomes as a result. Or, as I have put it previously, “for many the ‘means’ have become the ‘ends’, and we then lose sight of what the ends should be”. Gross Domestic Product, as a very poor proxy of the prosperity of a nation, is perhaps the best example of this problem at a national level.
The point is worth emphasising, “In addition to assuming that the model is the reality, we compound the problem by mistaking the proxies for the desired outcome of our model” and the name for this is surrogation, a process whereby a measure for a desired outcome becomes the surrogate for that outcome”. Another example that Martin gives is the price of a companies stock being considered a true and complete manifestation of the value of a company, resulting in it becoming the goal of management to increase the stock price, and incentivising this with stock-based compensation to generate perverse consequences and bad behaviours, which Martin illustrates.
The Economy Is Unbalanced and Distorted as a Consequence
Not only are businesses and the economy not resilient in the face of shocks, they have also become very unbalanced even in normal times, and they are failing to deliver the kind of outcomes “American democratic capitalism” has produced in the past, and that people expect of it. Our, to put it another way, for many the American Dream is dead.
Martin shares the findings of the Persona Project, part of a six-year Martin Prosperity Institute Project on the future of democratic Capitalism in America. It sought to understand how the average American was experiencing it. The two key findings were that most no longer feel the economy works for them, and most are “decidedly disengaged from politics”.
Two Centuries of Progress Has Stalled
The reasons are obvious from the data that Martin goes on to present. He notes that over its first two hundred years America sustained long-term growth, so the American Systems of democratic capitalism generated prosperity that was widely shared, and the median family in America saw their incomes increase 100% between 1947 and 1976. Since then, in a period almost twice as long, the median income rose only 31%.
Martin also notes that in two ways the current situation is worse than in the period following the great depression. First, “After the worst depression in history, it took only twelve years for average American income to grow 29% above the pre-depression peak, and fifteen years after it had grown by 100%. He adds, “In stark contrast, on the current trajectory, it would take the median 1976 American family one hundred years to double its real family income — a span covering three generations”.
Second, and perhaps more important in Martin’s opinion, “the Great Depression hit the incomes of the top-earning Americans more severely than it did those of average Americans”. “Nothing could be further from the truth in the current economy”. He adds, “While the median family is stagnating as never before, the top 1 percent (and 0.1 percent and 0.01 percent) are doing better than they ever have in American history — and there is no sign of that stopping.”
Martin goes on to say that until as late as the 1980’s, “the rule in the American economy was that the poorer you were, the more you benefited from growth in the American economy”, but, “around the same time that the income of the top earners began accelerating in the late 1970s, economic growth slowed”, and by 2014 the situation had “flipped entirely”, the richer you were the more you benefitted.
To stress the point Martin goes on to say, “there is no better illustration of how the average worker is losing out than the dramatic shift in the relationship between productivity growth and wage growth. Up to the mid 1970’s there was a tight relationship, but between 1973 and 2018 a 77% increase in productivity resulted in a much lower, 12%, increase in wage compensation. As a result, economic mobility, the basis of the American Dream, ground to a halt. He concludes, “we have to ask — given all that the data shows — whether there may be a fundamental structural problem with democratic capitalism. If so, can we fix it?”
Martin believes there is a structural problem, for the reasons I have already outlined above: the wrong analogy (thinking of the economy as a machine) linked to the wrong models and the wrong measures (proxies) of progress. In addition, the problem is compounded by our mistaking the proxies for the desired outcome of our model.
From Gussian to Pareto Distributions
To illustrate the point Martin notes that in the past the desired economic outcome was predicated on a Gussian distribution of incomes, a reference to bell-curve shaped chart of distribution by the German mathematician and physicist Carl Freidrich Gauss. It is called the ‘normal distribution’ curve because it is such a frequently found phenomenon. It represents our expectation, that the vast majority of the population earn a median income and are ‘middle class’. And, as the economy grows, the distribution remains proportionate, the curve retains its shape, whilst everyone benefits.
“This Gussian narrative was a reasonably good description of the American political economy, at least during the first two hundred years of the country’s existence.” Incomes for all families improved steadily. But over the past four decades the pattern has become increasingly distorted. He suggests we have moved “toward a Pareto economy”, a reference to the work of Italian economist Vilfredo Pareto who’s observations resulted in a different kind of distribution, producing a different shaped curve. The core feature is the absence of any meaningful mean or median. This is because of pressures on the system and connection among the participants in the system.
Martin’s key insight is that, “increases in pressure, fuelled by an obsession with efficiency, measured through surrogated proxies, in a context of increased connectedness, are making the distributions of outcomes in almost all spheres of economic activities increasingly Pareto, crowding out the historically Gaussian patterns that we have always assumed.”
He illustrates this in relation to income distribution, the jobs market, and the distribution of companies in an industry — wealth becomes increasingly concentrated, jobs become extremely highly paid for a very small minority and poorly paid for many, mergers and acquisitions lead to extreme industry consolidation.
Time for a New Analogy
The good news is that we can fix the system, but doing so will not be easy. First, we must re-consider the desired outcome or, “specify what we want the economy to deliver”, as Martin puts it. Then, recognising “hidden in the answer is an assumption about how “we” benefit from economic growth”, “we need to re-examine our theories about what improves the economy as well as the proxies we employ to measure ‘progress’ and guide further action”, warns Martin.
He proposes we drop the machine analogy of the economy in favour of one that “recognizes it as a natural system that needs constant tweaking”. More specifically he suggests it is a “complex adaptive system”, which is what natural systems are. He uses ‘natural’ for the sake of simplicity.
In such systems dynamic interactions take place making it too hard to determine in advance what the outcomes may be. The analogies he uses are rainforests, our bodies as natural systems.
The point is, “If the economy really were like a machine, in which any given subsystem operates independently and does not influence any other subsystem, we would probably have been protected” from the downside of the global banking crisis, for example. The fact we saw so much ‘contagion’, as it was called, indicates the degree of interconnection and interdependence even at the level of subsystems.
The implication is important. It is not possible to forecast outcomes using modelling approaches that assume linear, or at least moderately discernible input-output relationships, says Martin. And complex adaptive systems continuously adapt. They are dynamic.
Additionally, he cautions, “not all natural systems are alike” and behavioural adaptation is especially difficult to predict in social systems such as the economy. Actors in social systems can attempt to game the system for their personal advantage, leading to behaviours that cause crises, such as the actions that produced the global banking crisis.
All that I have covered above deals with the first Part of the book. The second part offers solutions, and particular solutions for business executives, political leaders, educators, and citizens. I will cover this second part in a second article.
Conclusion
To conclude this article let me say that I agree with the observations Martin makes, and that will be no surprise to readers of my previous articles. But given how long the machine analogy has been used to think about the economy, and the way it is managed, we should not underestimate the challenge of disregarding it for a more accurate alternative. Particularly the complex and adaptive system alternative, made-up of many sub-systems.
Under such dynamic and inherently uncertain systems, that can be influenced but not controlled, decisions still need to be made. It seems to be human nature to seek certainty. We analyse problems in search of the answer, assuming there is only one, and that it will remain the right one. This is a very wrong and dangerous assumption in complex adaptive systems, but we are all too willing to go with the false sense of certainty we seek, and to choose to be wilfully blind to realities that result in crises — predictable surprises.
On March 9th this year, just before the impact of Covid-19 was starting to really hit us, I ran a conference in London. Undaunted: How Successful Leaders Face-Up to Wicked Problems and Avoid Predictable Surprises. We covered much of the ground that Martin covers and produced similar insights including uncertainty, complexity and systems thinking, for example.
We have much more to do before we fully understand these issues. Part of the challenge is that the problems we seek to solve are different in nature, the systems we rely on are also very different in nature. We fail to recognise that this is the case. We are also very bad at understanding how systems interrelate and are interdependent, although each crisis exposes our ignorance and our lack of a holistic perspective.
Martin serves us well by adding his voice to the calls for change, consolidates earlier insights and introduces new ones. But I think readers will need to establish for themselves a much deeper understand of systems thinking than Martin is able to provide in this book.
Our lives are governed by a complex constellation of systems — economics, finance, health, education, transport, communication — how well they are designed and managed is critically important. Understanding their purpose, how they will create value in the form of a contribution to human flourishing and wellbeing, must determine their design and how we measure their performance, to avoid the ‘means’ being viewed as ‘ends’ in themselves. Or as Martin says, to avoid surrogation, the process whereby a measure for a desired outcome becomes the surrogate for that outcome.
Previously I reviewed the book Value Sensitive Design. The discipline offers very relevant insights. Read the Review
