Phil Rosenzweig criticize pseudo scientific tendencies in the explanation of business performance in specific books (those that offer secrets of guaranteed business success)
The Halo Effect is a 2007 book by business academic Phil Rosenzweig that criticize pseudo scientific tendencies in the explanation of business performance. As well as many business magazines and newspapers, it targets specific books (those that offer secrets of guaranteed business success) and academic research published by business schools. It outlines nine "delusions": mistakes of reasoning that undermine these recipes for business success. In light of these mistakes, Rosenzweig argues, much of business writing is what Richard Feynman called "cargo cult science", having the superficial trappings of science but operating at the level of story-telling. The book also considers some more scientific business research, whose conclusions are more rigorous but do not promise a simple recipe for success. The subtitle of the 2007 US edition is "and the eight other business delusions that deceive managers" while that of the 2008 UK edition is "How Managers Let Themselves Be Deceived".
The book was named "Business Book of the Year" 2007 at the Frankfurt Book Fair It has been described as part of a trend for books that encourage evidence-based practice in business research.
The author told reporters the book had been written over the course of 25 years of experience in business consultancy and academia. Rosenzweig earned his PhD at the University of Pennsylvania, before serving on the faculty at Harvard Business School and later at the International Institute for Management Development in Switzerland.His corporate career included seven years at Hewlett-Packard.
The book is critical of a genre of business books including In Search of Excellence, Good to Great,What Really Works and Built to Last. It finds similar faults with a swathe of business journalism.
- The Halo Effect of the book's title refers to the cognitive bias in which the perception of one quality is contaminated by a more readily available quality (for example good-looking people being rated as more intelligent). In the context of business, observers think they are making judgments of a company's customer-focus, quality of leadership or other virtues, but their judgement is contaminated by indicators of company performance such as share price or profitability. Correlations of, for example, customer-focus with business success then become meaningless, because success was the basis for the measure of customer focus.
- The Delusion of Correlation and Causality: mistakenly thinking that correlation is causation.
- The Delusion of Single Explanations: arguments that factor X improves performance by 40% and factor Y improves by another 40%, so both at once will result in an 80% improvement. The fallacy is that X and Y might be very strongly correlated. E.g. X might improve performance by causing Y.
- The Delusion of Connecting the Winning Dots: looking only at successful companies and finding their common features, without comparing them against unsuccessful companies.
- The Delusion of Rigorous Research: Some authors boast of the amount of data that they have collected, as though that in itself made the conclusions of the research valid.
- The Delusion of Lasting Success: the "secrets of success" books imply that lasting success is achievable, if only managers will follow their recommended approach. Rosenzweig argues that truly lasting success (outperforming the market for more than a generation) never happens in business.
- The Delusion of Absolute Performance: market performance is down to what competitors do as well as what the company itself does. A company can do everything right and yet still fall behind.
- The Delusion of the Wrong End of the Stick: getting cause the wrong way round. E.g. successful companies have a Corporate Social Responsibility policy. Should we infer that CSR contributes to success, or that profitable companies have money to spend on CSR?
- The Delusion of Organisational Physics: the idea that business performance is non-chaotically determined by discoverable factors, so that there are rules for success out there if only we can find them
Interview with Phil Rosenzweig, Author of The Halo Effect by steve millar
QUESTION: Can we start with a little personal background? We know from The Halo Effectjacket cover that you have a doctorate from Wharton, have taught at Harvard Business School, and are currently a professor at IMD in Switzerland.
ROSENZWEIG: I’m originally from California, and worked at Hewlett-Packard from 1979 until 1986 – at which point I shifted to a career as a business school professor. I was at Harvard from 1990 to 1996, and since 1996 I’ve have been on the faculty at IMD. We work mainly in executive education, so I work closely with managers from a variety of companies.
QUESTION: What are your areas of expertise in business? Do you consider yourself a strategist? How did you get started on the research that led to The Halo Effect?
ROSENZWEIG: My work with executives has led me to conclude that most of them are smart and hard-working, but have a limited ability to think rigorously and critically, especially about business research. As a result, they tend to be gullible and willing to believe research that claims to give the secrets to high performance. Unfortunately, for all the appearance of solid research, much of this work – which includes some of the biggest best sellers in recent years – is badly flawed and nowhere near as rigorous as it claims to be. My book tries to give managers the tools they need to be more discerning, more appropriately skeptical about what they read, and to be more intelligent consumers of management research. I hope to raise the level of discussion – and to set a higher standard for those who produce business research, including consultants, gurus, journalists, and professors like me!
QUESTION: Could you give us a short synopsis of the main themes of The Halo Effect?
ROSENZWEIG: A great deal of thinking the business world is shaped by mistakes in logic and errors in judgment – which I call “delusions.” In my book, I identify nine delusions, but they are not equally important. The most important one is the halo effect, which speaks of data independence and bias. Once researchers make that mistake, others tend to follow. As a result, researchers or consultants may claim to have found the keys to high performance, but often they have the causality backwards – rather than finding what drives high performance, they have found how high-performing companies tend to be described, which is a very different matter!
QUESTION: Do you think that some of the delusions noted in the book – the halo effect, single explanations, correlation and causality, rigorous research and lasting success – derive from our psychological need for simplicity and tidiness of cause-and-effect thinking? Is some just due to the desire of authors to sell books?
ROSENZWEIG: Probably some of both. No question, the halo effect is a very common mental shortcut that is born out of a natural desire for simplicity. But it is probably also true that some researchers don’t want to face up to the problem of data independence because they are content to sell simple stories.
QUESTION: The Chicago Cubs have had a down and now (currently) up year. When they were awful, the sports media was positing a broken organization, an inept farm system, an over-the-hill manager, and incompetent, prima donna players. Now that they’ve been on a winning streak, it’s a shrewd management team, with a productive farm system, a rejuvenated manager, and unselfish players. Is this the halo effect in practice? Is the press primarily responsible for this type of thinking due to their short-term focus?
ROSENZWEIG: This is a great example of the halo effect – and we see it frequently in sports. Exactly right! When a team is doing well, we tend to describe it in very different terms from when it was doing poorly.
QUESTION: Nassim Nicholas Taleb, author of Fooled by Randomness and The Black Swan, has extolled The Halo Effect as “one of the most important management books of all time.” Like you, Taleb rails against what he considers the delusions of those who think only in black and white with no grey, who deal only in certainty with no randomness, who seek only simple explanations, who think that life is linear, symmetric, and bell-shaped. Could Taleb’s diatribes fit into The Halo Effect?
ROSENZWEIG: Nassim is concerned with misconceptions and errors in thinking that pervade the world of financial markets. My concern is with the world of management and business performance. So we are looking at somewhat different fields, but our approaches are complementary and some of our ideas are very similar. I like his work a great deal, and am glad that he has also liked my book.
QUESTION: If business success is much less tidy and predictable than bestselling business book authors would have us believe, what are the implications for research and intelligence? If we cannot identify a single smoking pistol, is it all about increasing the odds of success?
ROSENZWEIG: Yes, trying to improve the probabilities of success is the way we should look at things. Business performance is, after all, relative and not absolute – we are in competition with other companies, and our success is also affected by what they do. There are not, therefore, formulas that can reliably and predictably lead to high performance. Our task as managers is to make judgments, under conditions of uncertainty, that stand the best chance of improving our likelihood of success in a competitive market setting.
QUESTION: You note methodological flaws in several of the designs that drive bestselling business book research. Do you agree with those who think most such designs are too weak to support any definitive conclusions? I have espoused randomized experiments and panel designs (time series with control groups) as exemplary for business and intelligence. Do you often see these designs with bestselling business book research?
ROSENZWEIG: Some well-known best sellers, such as In Search of Excellence and Blue Ocean Strategy, do indeed have problems of research design – they only examined successful companies, and therefore cannot say what makes successful companies different from less successful ones. The problem with Built to Last and Good to Great is different: the design, which used matched pairs, isn’t bad, but much of the data they relied on are not independent of performance, which is the very thing they are trying to explain. The problem here is not one of design, but of the validity of data.
More generally, some things in business do lend themselves to randomized experiments – such as retailing, pricing, merchandising, customer service, warranties, and more. Any time we have many discrete transactions, we can attempt randomized experiments – we can assign different treatments and compare results. But that approach doesn’t work for the big questions like mergers and acquisitions, or new market entry, or reorganizations. These things are few in number, infrequent, and hugely consequential – and therefore not the stuff of experiments. So we have to rely on other designs. That’s inevitable – but since we have to gather data from different sources, we also have to guard against certain kinds of bias, notably those that are based on performance.
QUESTION: In The Halo Effect, you mention the thinking of Michael Porter from Harvard Business School that business performance is driven by strategy and execution. Robert Kaplan and David Norton of Balanced Scorecard fame speak of strategy as hypothesis, where companies posit “if we do this, then results will be that,” or “the more we do of this, the less will happen of that.” Do you think business strategy can be adequately represented by hypotheses similar to those from science? Do you see more bottom-up strategy driven by randomized experiments? If so, what implications does that have for business performance evaluation?
ROSENZWEIG: I like much of what Porter, and Kaplan and Norton have written. That said, it is important to remember that a hypothesis in scientific experimentation – say, chemistry or physics – is very different from company performance in business, in that a chemical reaction or some physical phenomenon is generally absolute, not relative. In business, the key dependent variable – company performance – is better understood as relative, not absolute. And that changes everything. We simply cannot expect the predictability and replicability of the hard sciences when it comes to studying business performance.
QUESTION: Jeffrey Pfeffer and Robert I. Sutton from Stanford have written a best seller,Hard Facts, Dangerous Half-Truths & Total Nonsense, Profiting from Evidence-Based Management, which attempts to do for business what evidence-based medicine has done for health care. Among the tenets of evidence-based management are experiments and other powerful designs for evaluating strategy, along with a prototype management mentality and the feedback of strong intelligence. What are your thoughts on this development for business?
ROSENZWEIG: I like Hard Facts by Pfeffer and Sutton very much. I would only add two comments. First, everyone will quickly agree that “evidence” is important, but that only begs the question – we also have to agree on what constitutes good evidence! Unfortunately, the halo effect and other examples of bias have the effect of undermining the quality of evidence. So it’s not enough to emphasize “evidence” – we have to be clear about what good evidence really means. Second, the origins of evidence-based management come from medicine, and here again we have to be careful. In medicine, things such as patient recovery are absolute, not relative – I can have a ward full of patients and they can all recover thanks to a good treatment. The recovery of one patient is not dependent on the lack of recovery of someone else. In business, however, performance is relative more than absolute; therefore, the way we think about our dependent variable – our desired performance outcome – will be somewhat different. I like the basic tenets of evidence-based management, but we have to be careful when we draw simple parallels from medicine.
QUESTION: What has been the feedback/fallout from business academics and consultants over The Halo Effect? Has Jim Collins, author of Good to Great, called?
ROSENZWEIG: The response from colleagues and managers has been excellent – I think a lot of people are tired of the sort of nonsense that is being peddled and are ready for a clearer way of thinking. I have not heard directly from the gurus that I criticize, and don’t expect to – they wouldn’t see the benefit of giving me any more attention than I am already getting!
QUESTION: Where are you heading currently with your teaching, research, and consulting? Has The Halo Effect reshaped your work?
ROSENZWEIG: I continue to work with companies at IMD on questions of strategy and organization, but am now doing a bit more about rigorous thinking and understanding business performance.
QUESTION: Do you consult with business? If so, what are the main foci of that consulting derived from The Halo Effect? Do you see positive practical applications of your work?
ROSENZWEIG: At least one company has approached me to try to put ideas from my book into application. This company is asking: Do the metrics we have been using really capture the appropriate phenomena, and are they valid measures, or do we have problems of data validity and lack of independence? These are good questions that all companies should be asking.
QUESTION: The focus of the Business Intelligence Network (BeyeNETWORK.com) isbusiness intelligence (BI). BI can be viewed as the use of data, technology, and analytics to help business evaluate, optimize, and manage its strategies and business processes. The takeaways from The Halo Effect are certainly pertinent for BI. What advice would you have for management that wants to use intelligence as a differentiator?
ROSENZWEIG: My advice is for managers not to be impressed with the quantity of data in a given study, but to insist on the quality of data. Be sure that your independent variables are truly independent of the things you are seeking to explain. Also, watch out for the survivor bias: choose your sample population and time period in such a way as to capture the full phenomenon, not just the survivors.
QUESTION: Be an optimist and a realist. Over time, where do you see business headed with the research and intelligence it will increasingly accumulate?
ROSENZWEIG: Business management and leadership is a mix of hard and soft, of rigorous analytics and of intangibles like confidence and inspiration. I would never suggest that business is only about intelligence and rigor – it is art as well as science. The best managers, I think, have the ability to inspire and motivate others, but also are clear-eyed about analysis, about probabilities, and about the fundamentally uncertain nature of performance in a competitive market economy. There will never be predictable formulas that can ensure high performance – the nature of competition makes that impossible – but there are ways that business intelligence can improve our abilities to execute and be more efficient.
source :wikipedia/haloeffect/book & e bay network