Seeking Firm Info Hypocrisy Examples

Arnold Kling:

[James] Manzi is a fan of randomized controlled experiments in business and public policy (in the latter, examples include the Rand health care study and the Wisconsin income-maintenance studies). I believe that decision-makers will resist this approach, for the same reason that they resist Robin Hanson’s suggestion to use prediction markets. That is, decisions are not necessarily about achieving results. They are often about establishing the status of the decision-maker. For a decision-maker to conduct experiments or to employ prediction markets is to admit ignorance and doubt, which lowers the decision-maker’s status. (more)

The examples of disinterest in random trials and prediction markets both help convince me that management is often less interested in information that it pretends to be. But since I’m giving a talk on the subject soon, I’d like other examples. So I ask you, dear readers: what common patterns of manager behavior suggests they are less (or more) interested in info than they let on?

GD Star Rating
Tagged as: ,
Trackback URL:
  • Joe

    Mangers will often make the decision on highly technical matters without real practical grasp of the technology, complexity, or implementation knowledge needed when the direct reports who have all of these things present options to them.

  • Matt

    I would argue that the premise is wrong. Most decisions that executives make aren’t necessarily good candidates for the type of randomization and trials that you are talking about.

    In the tech industry we use random trails all the time, A/B testing of website designs and features. Facebook and Google both have this as a big part of their company cultures.

    However there is a limit to this type of work, often you have time constraints, competitive pressures, and limited information. Status is lost by making bad decisions, not at first and not always, but eventually.

    • nazgulnarsil

      I would argue that business decisions are more fault tolerant than generally assumed. There is a strong incentive to act as if they are not because that raises the status of any decision maker.

      • xxd

        Agreed. There is a significant difference between operational and strategic decision making. Strategic decision making by it’s nature is much fluffier than operational. Is the Strategy even correct? Who is to say. My opinion is that any strategy that is different from provide the customer with what they want for the minimum cost for the maximum profit is the best strategy but this is notoriously difficult to enunciate into strategic vision with concrete divisible operational chunks and often the operational decisions are divorced from the strategic decisions.

        An excellent book on this is “The Halo Effect”.

        From a managerial incompetence perspective it’s much easier to hide bad decisions at a higher level because of the aforesaid disconnect between operations and strategy.

        A more interesting topic is to ask the question: should high level executives even be required to make strategic decisions?

        I’d be bold and stick my neck out and say that since the optimal strategy is almost always the same that there is no need whatsoever for higher level executives to even make strategic decisions at all and the operational decisions should be up to the operational level managers. Further: my position is that since higher level executives provably make little to no difference other than motivationally “the vision thing” to the company bottom line and in fact only serve to be good speakers in front of VC guys or stock market analysts or groups of investors that we are in fact asking too much of senior executives if we ask them to make decisions and instead should limit them to doing what they’re good at which is giving presentations and doing PR for the above mentioned groups of external stakeholders.

      • Mark M

        nazgulnarsil –

        I agree about the fault tolerance. This is why so many incompetent managers and executives are able to survive and thrive in the corporate world. I’ve also seen a surprising number of small business owners make surprisingly stupid mistakes without any lasting impact.

      • Doug S.

        I suspect that the function of the CEO and other high executives is primarily to arbitrate disputes between lower ranking members of the firm. Two upper-mid-level managers can argue forever about what to do, but the CEO can simply tell them who’s “right” and they have to shut up and listen.

  • Rob

    Kahneman discusses this resistance, even including some personal anecdotes, in his book.

    “…The moment you have a system that is a more structured system – then that system can be used to second guess the decisions of people. And people don’t like to be second guessed…”

  • The way that many companies structure the questions they ask in polls, and even then abandon the results if they don’t like them.

    The way that many corporations and politicians simply deny science (and even arithmetic) if it doesn’t suit their business and political interests, for example tobacco causing health problems, pollution causing health problems, green house gases causing global warming, austerity through reduced government spending contracting the economy, turning Medicare into a fixed voucher system is not “ending Medicare as we know it”.

    If someone is not going to change their action based on data or on a better understanding reality, there is no point in even going through the motions.

    I think a lot of this is deliberate, that is those in charge know what they want to accomplish, but need the “cover” of a more benign or neutral policy to justify it.

    A good example of this was when Eric Cantor said that pollution controls cost too much, but then objected to Obama characterizing Cantor as putting making money ahead of having a cleaner environment which is exactly what Cantor was doing. People may disagree what the trade-off between profits and pollution should be, but if you reject pollution controls because they cost too much, you are putting making money ahead of having a cleaner environment.

    The reality is that the costs of pollution control equipment fall on the owners of factories and their customers, where the costs of pollution fall on whoever lives down wind. Libertarians don’t have a factual basis for asserting that living in a polluted environment has no costs, so they simply deny it via hypocrisy.

    Another good example is AGW denialism. The denialists say there isn’t enough data, but then they reject all the data that there is and any new data that doesn’t fit what they want the data to say.

    I think ignoring evidence and denying reality is mostly a Conservative tactic. Reality has a liberal bias, so liberals don’t need to do that. 😉

    • I think ignoring evidence and denying reality is mostly a Conservative tactic. Reality has a liberal bias, so liberals don’t need to do that.

      I know that I should not but I had to answer this. Both side have their biases and if you do not see bias among liberals you are not looking. I think they are equally good at ignoring science that does not go their way.

      • For example any data about gender or race differences.

      • Or smoking being not so bad actually.

      • Poelmo

        Liberals (in the American sense of the word) do seem to be on the side of the facts and reason more often, they don’t always understand why, but they do seem to be better at figuring out who to listen to if they don’t know much about the subject themselves. So no, they’re not two sides of the same coin, they don’t just have opposite dogma’s. The difference between having an open mind and feverishly defending (caricatures, I know, but like any good caricature they preserve the essence) traditions because you’re afraid of the unknown, goes much deeper. Liberals and conservatives really think differently.

        This deevaluation of reason and facts is a disease. No, when you believe black people are inherently dumb you don’t have merely the same bias as someone who believes humans of all skin colors have the same average intelligence, they are not just two opinions of equal value. One has only prejudiced softcore research (questionable psychology and statistics research), while the other is rooted in cold hard genetics research and is logically the default stance. It’s the same with climate change: people contrive all sorts of ways to deny something they don’t remotely understand, purely because they’re afraid of carbon taxes. (American) politics are scary in this way and the media are complicit because of what they call “balanced reporting”.

      • I disagree. Liberals don’t try to deny what the facts are, they simply have a different explanation for what the facts imply.

        To a liberal, racial, ethnic and gender differences in achievement are real, but they are essentially all due to differences in socioeconomic status, poverty, abuse, poor at home environment, poor prenatal care, unequal opportunity and stereotype threat.

        Conservatives make up “facts” to suit their beliefs. Conservatives believe that ethnic differences in achievement are due to differential genetics, despite the fact that no genes have been found that are consistent with such beliefs.

        In other words, Liberals explain differential achievement by looking at the real differences in socioeconomic status which are known to mediate differences in achievement and which have been demonstrated to do so across many cultures, and then seek to mitigate those adverse effects.

        Conservatives ignore the known adverse effects and attribute differential achievement to unknown hypothetical genetics and bad choices, such as choosing to be born in poverty. They do this because they don’t want to do anything that might mitigate what is really causing differential achievement.

      • Be aware that we’ve already had the IQ argument here before, so I’d encourage anyone participating in it now to read the previous discussion so we don’t retread old ground.

        “differences in socioeconomic status, poverty, abuse, poor at home environment, poor prenatal care, unequal opportunity and stereotype threat.”
        Most of those can be examined through adoption studies, prenatal care notably excluded. We can even examine the effect that the perception of others have through studies that show whether parents mistakenly believe their children are or are not identical vs fraternal, or even biracial. A lot of that can be skipped with a GWAS like the recent Ian Deary study on IQ, following up on the Dolan GWAS on height I referenced in the previous OB link.

  • “When I started writing I thought if I proved X was a stupid thing to do people would stop doing X. I was wrong. ”
    —Bill James in his 1984 Baseball Abstract

    A/B testing is used in web design but it could be much more widely applied

  • Poelmo


    What you are describing is a religious belief: I believe A and if someone presents me with evidence of B I will dismiss, I will never investigate any weaknesses A might have but I will actively look for evidence that supports A. Besides religion this attitude is common conspiracy theorists and the American conservative movement. In fact it was the default basis for human thinking and philosophy for thousands of years.

    • Just because we are not what we call “religious” does not mean we are immune from religious thinking. Religious thinking is social thinking. Many beliefs common to liberals, some of which I share, are protected by an aura of emotion and taboo. We reject hypotheses that seem nasty in some way (essentially a religious feeling) without giving them the benefit of rational consideration. We only imagine that our emotional rejection of them is the same as rational consideration.

      • Poelmo

        Yes, you are right, but that’s what I said: you don’t have to be religious to think that way, being a conspiracy theorist also works.

  • JR

    Often managers take and run with the first approximate answer. Further refinements/improvements/corrections must support (and not contradict) this first answer, particularly if it has already been communicated up the ladder.

    Be very careful when communicating preliminary results…

  • Rikk Hill

    Most of this behaviour I’ve seen tends to focus around risk management.

    An archetypal example: a manager will read of/dream up/hallucinate a mode of failure for a business-critical system, then insist a failsafe be put in place to deal with that circumstance. The mode of failure will be a legitimate concern, but the actual likelihood of that failure mode occurring isn’t taken into consideration. Meanwhile, adding the failsafe introduces extra modes of failure that are more abstract and harder to visualise, and the actual process of implementing it increases the scope for human error.

    On balance, the risks from implementing the failsafe are greater than the risk of the failure mode. Still, implementing the failsafe makes the manager look astute and cautious, and any problems that arise from the implementation are seen as the failure of his subordinates, not him.

    This problem is exacerbated by widespread poor understanding of risk and statistics, but the argument is pretty straightforward. “We’re more likely to make a horrible mistake when safeguarding against X than we are to simply experience X”. They don’t care.

    • This type of behavior is a recurrent theme in literature, particularly in Greek myths, where the Oracle would predict some future disaster, and in trying to avoid that disaster the disaster would be caused.

      An example in current events is preventive war. Trying to prevent war by going to war. Trying to prevent Iran from acquiring nuclear weapons by bombing them would actually ensure that Iran would develop nuclear weapons.

  • Sid

    Personal interviews. They are much less effective than the interviewers make them out to be. They are more about establishing status and getting people into the firm who would respect existing status hierarchies. But I don’t know if there are documented studies describing this.

  • Kevin

    great interview by Frank Zappa on this exact phenomenon in the music industry:

  • DW

    Firms often have many employees but are usually only effective at achieving whatever the CEO is currently focused on.

    Good CEOs care about info. Everyone else just spends their time trying to figure out what the CEO wants to hear. This may be related to info.

  • Nathan Whitehead

    I see this one all the time. The exact same facts are presented to management through two different channels. The response varies wildly depending on the source of the facts. The responses are not about dealing with the new information but about responding to the politics of the developing situation.

  • Roger

    Change changes the power distribution within an organization. Managers and the divisions they run compete for resources and status. When new ideas or products are presented, much of the evaluation of the idea is based upon whether it will hurt or help the power, status and relative size of a manager’s division. The important factors become who suggested the idea and whether it increases or decreases the budget of my department relative to my colleagues and competitors.

    If you think the evaluation is based upon how effective it will be to large scale organizational success, then you are missing much of the dynamic. They don’t want randomized trials or prediction markets because these things threaten their power and status. Both decentralize decision making and make it empirical rather than political.

    This is my experience from leading product innovation at a major financial institution.

  • Robin – How about the ubiquitous “hockey-stick” curves for market size produced by consultants and industry analysts.

    No matter what the technology or market, the projection is always hockey-stick growth, with the knee of the curve just a few years off. (I could give a long list of examples.)

    I think this is because that’s what managers want to see – confirmation of the correctness of their existing plans.

    Consultants who don’t deliver the confirmatory goods go out of business, so selection leaves only those with the hockey-stick graphs.

    • Someone from the other side

      Speaking as consultant with one of the top firms, I more than once (in fact, I am going to do it starting next week) wheeled out scarily big guns to disprove pre-existing hockey sticks as the bullshit they are (it usually takes all of 15 minutes to figure that out and then weeks of nasty analyses to prove it beyond doubt).

  • Mark M

    Robin –

    Isn’t this mainly the age-old principal-agent problem? Perhaps compounded with confirmation bias?

    I don’t think managers are necessarily disinterested – I think they’re solving different problems than what appears on the surface. The trick is to understand what problems they are solving beyond what is readily apparent. Signalling status is definitely part of that equation. Another factor is inertia: resistance to change, fear of the unknown, or inability to understand.

    As with the recent example of using high-status consultants, signalling may not work if the underlying reason for the decision is known, so managers have an incentive to hide the fact that signalling is part of their objective.

    For a specific example, when a recommendation is sent back for further study. I saw this when the President of a publicly traded company where I was employed was a former IBMer. A study of products resulted in a recommendation to go with an IBM competitor. When asked to re-evaluate, we had the same results. Then again. After many delays and studies, we finally bought the IBM product. The President, who was newly hired and didn’t stay long, had other objectives besides the health of the company. In retrospect, those objectives were maintaining a solid network for future job opportunities. Although we knew our president was being influenced by this relationship, we didn’t understand that he didn’t really care what the study said, and he couldn’t tell us his true motivation.

  • kirk

    The ubiquity of Program Management institutionalizes management ignorance of individual projects. Simple questions like “does it work” and “can we sell it” become meaningless reports about schedule and design budget tracking and reporting. PM won the war on metrics by replacing performance requirements and market expectations (which are largely invented out of whole cloth before the project gets rolling) with ginormous heaps of meaningless info-tainment. Project teams are trained to project specious outcomes which become embalmed in layers and layers of pointy headed manager heirarchies. If we give a PM a team of individuals that are invested in building a car that floats and flys FROM DAY ONE we get project plans that demonstrate… winning plans for cars that float and fly. On time and under budget.

  • Someone had a great post on “The costs of savoring”. The same logic applies to managers: opening one’s mind to more information, alternate approaches, new ideas, is costly, requiring ,requiring time and energy. And the returns for the additional investment in time and energy are highly uncertain with no proven track record. And almost by definition getting more information means embarking on a path which one’s boss has never trod, and hence will require an additional investment in selling the boss.

  • Danny

    Our company had to choose between two options, let’s call them ‘plan A’ and ‘plan B’. I presented my manager with an analysis that showed ‘plan A’ was overwhelmingly superior, given corporate objectives. BUT ‘plan A’ needed the approval of management two levels above my manager, and ‘plan B’ only needed the approval of management one level above my manager. So my manager made an attempt to sell ‘plan A’ to his manager, then my manager’s manager made a half-hearted attempt to sell ‘plan A’ to his manager. A little bit of push back led him to conclude it wasn’t worth the bother of trying to change his boss’s mind (and potentially annoying him in the process), and so he gave the instruction to go ahead with ‘plan B’ instead.

    • Robert Koslover

      Can you be entirely confident that your understanding of the overall needs of the organization (aka, the “corporate objectives)” was superior to that of the understanding of the manager three levels above you?

  • NotJohn

    Don’t underestimate the importance of a manager being prepared to make a decision. Very often there isn’t a huge difference between the likely outcomes of different answers. In such cases being an effective manager just means being prepared to make a decision and then back it. Of course that will sometimes backfire, but most good managers I’ve known have had that characteristic.

  • Eric Hammer

    I used to work for a business unit on the planning side that had to make a decision between two sets of demand forecasts. One was pretty labor intensive and infrequently updated based on past history (something like once a quarter), and had rather bad accuracy. The other was almost entirely automated and didn’t require much in the way of human time, updated every month, and had better accuracy.

    The business, after being shown a great deal of data demonstrating these facts, chose the more expensive and less accurate forecast. On purpose.