Bullsh*t jobs: Understanding Work and Value in the Information Age

“Where you work is going to inform your values, and that can be a good thing. But it can be a negative thing as well, when you’ve got too much bullshit in the ecosystem that actually isn’t worried about the outcome. It’s only worried about the appearance of those who are actors in the process.”

Intro

The difference between lying and bullshit is that liars want their audience to believe their message and bullshitters don’t care. Bullshit is about saying whatever it takes to convince another party of something about the speaker – usually that they are competent and useful. This concept has trickled into the workplace in an insidious way. Bullshit jobs make an organization look more robust, productive, and legitimate but actually provide very little return on investment or impact.

While bullshit jobs keep people employed, they can also have negative consequences for our economy and society. Holders of bullshit jobs make up the “bullshit army” backing a CEO or executive’s decision-making process. In this way, leaders may be able to get away with controversial decisions or ideas, through wave after wave of employees who nod their heads in agreement, confident in the knowledge that it’s easier to get ahead by sending signals up the chain that people want to hear—whether or not it’s what they need to hear.

When work is meaningful – and not “bullshit” – it can shape and appeal to our value systems, allowing us to feel fulfilled by a belief that we contribute to something bigger than ourselves. Even through non-meaningful work, people are often able to discover and embody their own value systems. In this way, work – which can take up half or more of our waking hours – becomes an important part of who we are, how we view the world, and what makes up our society’s values. When our jobs feel like bullshit, our social order suffers.

So, how can we bridge the gap between “bullshit jobs” and our aspiration for meaningful, value-informed work? One suggestion from our panelists – senior consultant Jayden Rae, research director Dr. Brooke Struck, and associate Nathan Collett – is defining our own success. Frequently, in North American workplaces, success comes down to meeting financial targets or completing a series of tasks delegated by executives. It is often misaligned with employees’ values and misses the mark on intrinsic motivation. Even when work does contribute to meaningful good, impact is often not communicated back to employees themselves, robbing them of the opportunity to feel fulfilled by their labor. For this reason, making impact more salient to employees and determining our own barometers for our success are crucial to living a meaningful, value-fulfilled life through work.

This conversation has been edited for clarity and concision.

Discussion

Nathan: In this latest instalment of TDL’s Perspectives project, I’m sitting down with Brooke Struck, our research director, and Jayden Rae, senior consultant here at The Decision Lab. Today, we’re going to be talking about bullshit jobs: what they are, where they come from, what defines meaningful work, and what role work plays in a meaningful life. So to get started, let’s think about this concept a little bit, and break it down. Brooke, can you tell me a little bit about what a bullshit job is and where the idea came from?

What Are “Bullshit Jobs?”

Brooke: Sure, let’s start with one component of this idea at a time. Harry Frankfurt is this American philosopher who wrote an excellent essay called “On Bullshit,” where he distinguishes bullshitting from telling the truth and lying. So, telling the truth and lying have one thing in common: they both have a respect for the truth. And when you’re telling the truth, you’re trying to convince someone that something is the case. When you’re telling a lie, you’re trying to convince someone that something is the case, it just so happens that that something isn’t true. That thing is false. But when you’re bullshitting, you’re not actually trying to convince anyone about anything out there in the world. You’re trying to convince someone about you. It’s more about the speaker than about the world that they’re speaking about. Bullshit, in that respect, is extremely performative.

So the way that that plays out in a professional context around business bullshit, for instance, is around jargon and acronym soup and all of these kinds of things that appear to be very meaningful and well-thought-out and cutting-edge and innovative, but in fact, they’re often extremely vacuous of any content. That’s what gives them their performative aspect. It’s a way that the elites in the business world do their little secret handshake to show each other that, “I’m an elite and you’re an elite and we agree that this system is effective in perpetuating our ability to exert power.”

Bullshit jobs are the downstream effect of that – in creating this whole economy of jargon and bullshit within an organization, it needs to have physical manifestations, as well. The bullshit meeting leads to the bullshit minutes, which leads to the bullshit action items, which leads to the bullshit jobs to go and chase all of those things down. That’s what a bullshit job really is: it’s someone whose role, majorly or entirely, is to do service to that bullshit economy, essentially just holding up the cape of the executive person who marches ahead of this army of bullshit jobbers behind them.

That’s what a bullshit job really is: it’s someone whose role, majoritarily or entirely, is to do service to that bullshit economy, essentially just holding up the cape of the executive person who marches ahead of this army of bullshit jobbers behind them.

Nathan: That’s great. So before we dig into each one of those concepts, Jayden, what do you think about the way this concept came about? Do you think there’s more bullshit out there today than there has been in the past?

Jayden: Yes, and that’s a really interesting question. I’ll draw from David Graeber, a social anthropologist who wrote the book which put this topic on center stage, not surprisingly called Bullshit Jobs. He examines how the division of labor has changed over time, and that this idea of bullshit jobs is actually a fairly recent phenomenon. He looks at the U.S. labor market over time. In the early 1900s, you saw that only one-fourth of jobs could be qualified as a professional job, something like corporate law or a consultant or a lobbyist, something or another job  that would fall into that category. But going into the early 2000s, over three-fourths of the jobs in the U.S. labor market would be considered professional jobs.

Graeber suggests that because of this general shift in developed countries towards outsourcing and the professionalization and industrialization of labor, bullshit jobs have necessarily been created to keep people employed. And a lot of these jobs are not actually productive jobs in the sense that previously blue collar or agricultural jobs used to be productive. And given a lot of the political and social incentives to keep people employed, and also these really deep-rooted Protestant work ethics that value over-work, he essentially argues that a lot of these jobs have been created despite the fact that they don’t produce the same types of material returns other, previously seen labor would.

Given a lot of the political and social incentives to keep people employed, and also these really deep-rooted Protestant work ethics that value over-work, [Graeber] essentially argues that a lot of these jobs have been created despite the fact that they don’t produce the same returns other, earlier forms of labor would.

Brooke: Yeah, that’s a really interesting point, that if we think about what the economy produces, when you’ve got 75% of your economy working in agricultural production or manufacturing production or these kinds of things, the outputs of those sections of the economy are quite tangible, and you can assess the quality of those outputs in some fairly concrete, standardized, and genuine ways. When it comes to certain modern professions, such as certain business or advisory groups, but also to some degree bureaucracies and governments, to a large extent, what they create is decisions and consensus. And assessing the quality of a decision, or even a decision-making process, is very hard to do. And often, how much organizational momentum has built up behind the decision subs in as a proxy for how good a decision is. This creates an incentive system for people to build up armies of people behind them, who trade in this economy of building up the appearance of organizational momentum behind a decision.

Often, how much organizational momentum has built up behind the decision often subs in as a proxy for how good a decision is. This creates an incentive system for people to build up armies of people behind them, who trade in this economy of building up the appearance of organizational momentum behind a decision.

Organizational momentum or “the bullshit army”

Nathan: Let’s talk more about the idea of organizational momentum and elites needing this support system behind them. Recently, there was some commentary on the way the Senate operates, especially when confirming judicial appointees. Masha Gessen writes in The New Yorker that in the confirmation of Amy Coney Barrett, senators made 25-minute speeches that had nothing to do with the topic – they made a farce of the entire process. This, Gessen suggests, is the epitome of a bullshit job. They’re facilitating a process and providing legitimacy to something that would otherwise necessitate critical thinking.

So, do you think this idea is related to a shift in the way people are conducting critical thinking? I mean, you’re suggesting that experts and elites have these support systems, or “bullshit armies.” Does not having to justify their ideas alone help them avoid potential controversy? Are they able to use this army of people working for them to make their arguments and ideas seem legitimate, even when they’re not?

Brooke: Yeah, I think that there’s definitely a strong argument to be made there, that one of the things that people really look to avoid is isolating themselves, especially around a decision. No one wants to feel like they’re standing on an island, left holding the fort on a single decision, because that creates great professional risk for them. It’s much easier to defuse that risk if you say, “Well, I may have made a bad decision, but there was a group of eight decision-makers around that table, and we all came to that decision together, supported by our armies of analysts, who each combed through the background material. So if a mistake was made – which of course I will never divulge or I will never acknowledge – then certainly, it can’t have been my fault. It must have been a totally reasonable error, such that all of us working on this thing together all came to the same conclusion.”

Nathan: To couch this process in terms that may be familiar to people in the decision-making or behavioral science field; we can make a distinction between process and outcome here, right? The outcome of the decision is one that is consequential and serious. It’s the process that’s being trivialized by these jobs that justify the outcomes without contributing to the facts of the case that might sway people by more legitimate means. Because these jobs don’t do anything, they should not actually affect the outcome, other than to support one that otherwise would not make it through an honest decision-vetting process. 

I want to pivot a bit to technology now and to explore how that feeds into this process. Let’s think about the evolution of bullshit jobs as a big part of the economy. There are certain technological developments that are responsible for or have contributed to the ease with which bullshit jobs take place. We talked about a shift from certain forms of employment that basically rely on either physical or manual labor. Now, we’re getting into the labor of decision-making, which is often facilitated by all sorts of different technology, whether it’s thousands of emails, Slack channels, or other softwares. It’s a little bit lighter and harder to pin down where the actual work is getting done and what work leads to measurable outcomes. 

Let’s think about what Jayden said earlier about the fact that the pursuit of full employment coupled with a Protestant work ethic makes for a dangerous combination, where self-image is tied in with the value of working hard. Is it bad that we see this shift to technological and more professional forms of work, and that we’re losing a sort of integrity to the way we work? Or is it a more complicated story than that?

Jayden: To turn your question on its head, again, going back to David Graeber, he has a really interesting thesis. He essentially says, “The main problem with bullshit jobs is that people derive their sense of self and self-worth from their work.” I guess it’s a complicated situation because people often enter jobs with underlying values and expectations that they will gain some self-development or some self-actualization through their work. Often, however, they enter a very bureaucratic system where they might not actually acknowledge through their work that they are fulfilling a pointless job. In these cases, we can see actual societal harm, when a lot of individuals fall into this category. So it’s actually problematic in the sense that the separation between process and outcome is actually harmful to individuals, as well.

People often enter jobs with underlying values and expectations that they will gain some self-development or some self-actualization through their work. Often, however, they enter a very bureaucratic system where they might not actually acknowledge through their work that they are acting out a pointless job.

Meaningful work

Nathan: That’s fascinating. In that case, what makes a good job? 

Jayden: I think one thing that has made me think about work a lot is the movement of effective altruism: thinking about how you can use your career in a way that aligns with other values or other aspirations. The effective altruism movement is in line with the philosophy of utilitarianism, which is basically the idea of people trying to create the most good. And so, they think about this in terms of trying to create a career that’s aligned with yourself, your own skill set, and to have the largest impact you can. I think for many people, if there’s an opportunity to create a career path that allows them to have an impact on the things they care about in the greatest way possible, but not only that, that those positive impacts or outcomes are really salient to them, that’s really important. This is something Nava Ashraf speaks and writes a lot about, centred around what she calls altruistic capital, which is essentially how much meaning and personal fulfillment a certain job or company can provide.

[Read here: Nava Ashraf on altruistic capital.]

I think for many people, if there’s an opportunity to create a career path that allows them to have an impact on the things they care about in the greatest way possible, [and also] that those impacts and those positive impacts or outcomes are really salient to them, that’s super powerful.

Nathan: And that’s basically the exact opposite of a bullshit job, right? All of a sudden, the work is not only productive, but it’s also meaningful to you because of the values that underlie it. We can expand and think about some other things that come into the picture when we think about valuable work and what it is to have a good job. At the risk of exposing my inadequate knowledge of his theory of labour, I can say that Marx talks about alienation and ownership being valuable in one’s work. Someone needs to have a connection to the product they’re spending eight hours a day laboring over, and that’s a source of value to them when carried out autonomously. Being connected to what comes out at the end of your day is important. Do we think that that is another source of value, or can we only really find value in social value, in our own values, as opposed to the monetary value of what we produce?

Brooke: Yeah, that’s one of the challenges with theories like utilitarianism, and even the economic theory we’ve held prominently in North America and Europe since the 1980s – this idea of fungibility, that a dollar is a dollar is a dollar. In the case of utilitarianism and the effective altruism movement, a quality is a quality is a quality. I don’t necessarily buy that in terms of work feeling meaningful. So certainly, Jayden, one of the things that you mentioned in what you were discussing before is about salience: feeling the tangible connection between the work you’re doing and the outcomes that you’re helping to promote. 

This other angle around fungibility is to acknowledge that there are different kinds of value. It’s  qualitatively different to promote environmental outcomes from promoting social equality or economic prosperity. It’s a different type of goal to promote community cohesion within a city versus within a province. We can’t make trade-offs between these goals, or at least those trade-offs are not as straightforward as just optimizing two things measured in the same units. So, I think, we need to unpack these kinds of questions to get at this notion of what constitutes a good job for an individual.

But at the structural level, what makes a good job, if we want to build on these two ideas, are first of all that you’re advancing a type of good that is important to you as an individual. And second of all, that the efforts that you put in to advance that good are done in a way that promotes a certain amount of salience as to how you are contributing to that process.

Bridging the gap

Nathan: Let’s take what we’ve been talking about, the last things Brooke was saying about having your values apply to your work versus recognizing your place in a process irrespective of its eventual output, and how someone who recognizes that they are in a non-ideal job might work to improve it. Or how could a non-ideal working environment be improved and made more meaningful to employees?

Jayden: Yeah, it’s a challenging question. I guess you can look at it from two perspectives. One is more at the organizational or institutional level of those people who are actually creating the structures of the organization, and the types of roles and activities that people who are employed should fill, and to be really critical about the purpose of all those positions. A lot of these more bullshit jobs exist because they’re creating non-lasting solutions to problems, or potentially, they’ve fallen into middle management where they’re creating extra work to look productive themselves, extra work that’s not necessarily productive. So I think really effective organizations are firstly really self-reflective and critical about the structure that they’ve adopted.

A lot of these more bullshit jobs exist because they’re creating non-lasting solutions to problems or they’re middle management. They’re creating extra work to look productive themselves, extra work that’s not necessarily productive. So I think really effective organizations are firstly really self-reflective and critical about the structure that they’ve adopted.

I guess for individuals, it’s quite a subjective experience. It’s possible someone is fulfilling what we may qualify as a pointless job, and they may have convinced themselves or genuinely believe that their job is not, in fact, pointless. So I think it’s important to preserve individuals’ agency and autonomy when we’re trying to classify different types of jobs, and not categorize a whole group of them as being pointless. That aside, I guess to the extent that you can, you should try to create relationships that give you leverage in trying to change what your job looks like. 

Brooke: Okay, so you talked about how it is that we can resuscitate jobs that might feel particularly bullshitty. One thing I think is really important here is assessment and evaluation. One of the challenges, as I mentioned, that drives the bullshit economy is how hard it is to assess a good decision-making process. So if indeed that is one of the key drivers of the bullshit economy, the solution then is to design effective alternative assessment processes to determine how well a decision-making iteration has gone. If you find something that you can put into the place of organizational momentum to assess the quality of a decision, then you remove the incentive to have a whole army of people lined up behind you to say, “This is the right thing to do.” Your incentive system shifts to whatever that new assessment mechanism is.

Nathan: What can that look like? I mean, do we have any strategies for assessing the quality of a decision? There’s a lot of talk about post-hoc versus in-the-moment assessments, of whether a decision is right before it gets made versus whether the result was the right one. But beyond that, what would assessing the quality of a decision look like?

Brooke: Now, the short answer is that getting rid of biases is what you need in order to have a good, clean decision-making process. That doesn’t mean that every decision you make is going to turn out well, but it does promote better decisions coming out the end of the pipeline. And in sufficiently large numbers, better decisions lead to better outcomes. But there’s this challenge around individuals being identified with individual decisions, which are identified with individual outcomes. To circle back to a point from earlier in our discussion, people do not want to be held accountable for big decisions until they pay off. So one of the uniquely punishing aspects of a bullshit job is that you’re there to diffuse accountability for a major decision-maker if things go badly. But that decision-maker will also be happy to take the credit themselves when things go well. And that creates a major problem that drives the existence of the bullshit economy.

But there’s this challenge around individual people being identified with individual decisions, which are identified with individual outcomes. And that creates a major problem that drives the existence of the bullshit economy.

I think that in terms of finding one’s own work fulfilling, there’s something to be said about how we evaluate our own contributions. Speaking from personal experience, I think one of the most important things in my career has been retaining ownership of the markers of what constitutes success. In my own career, all of the times I’ve done what I consider to be my best work and had the most impact has been in those moments where I was the arbiter of my own success. Whether things were going well or badly, it was something for me to figure out, rather than waiting for those around me to determine it on my behalf.

Nathan: That strikes me as something characteristic of more senior positions or smaller organizations, Junior people in bigger organizations would likely have to find their goals and approval from above.

Brooke: For sure it’s easier the more senior you are, though I started taking on this kind of ownership as early as university and graduate school. In professional contexts, I think the gap that I’d want to open up there is that what your manager tells you about your performance is not necessarily an accurate reflection of your value. We need to be able to internalize the idea that our value is something we need to determine. You can be inspired by others and you can pick up pieces from these external systems to validate your work, but ultimately, you are the one who needs to give validation about the work you do. It’s you who needs to be able to sleep at night doing the job that you’ve done. It’s you who needs to be able to motivate yourself to spring out of bed in the morning excited about your work. That’s a really valuable way to get away from this bullshit job, to go through the very, very difficult and wrenching process of self-examination to ask myself, “What do I think good work looks like? What constitutes a good job for me?”

What your manager tells you about your performance is not necessarily an accurate reflection of your value. We need to be able to internalize the idea that your value is something that you need to determine…you are the one who needs to give validation about the work that you do.

Those are very, very difficult questions to address. But ultimately, and again, speaking from personal experience, I’ve found that things have worked out most for me when I’ve stopped worrying too much what others thought. Acknowledging the enormous benefit and privilege that have allowed me to do so, I’ve essentially had the opportunity to define my own success criteria. There was obviously minimal stuff that I needed to feed to my committee during my PhD to keep them off my back, but more or less, as long as minimal boxes were being checked elsewhere, I could define my own success. If those boxes are not minimal, you don’t have as much freedom. (I say that in the sense of psychological freedom, freedom as an individual, to undertake those kinds of self-explorations and self-criticisms.) Because in fact, you can create danger for yourself by creating too much of a disconnect between what you define to be a good job for you and what the external ecosystem is demanding of you.

Defining our own success

Nathan: It’s interesting you say that. A lot of people struggle in grad school, precisely because of the struggle to define their own ideals, to work off their own timetable, but more importantly, what you’ve just been saying, to define what success looks like according to their own benchmarks, and to follow that. I wonder if we can find a distinction here, a really important one, between the hardship that comes with that sort of freedom, and the hardship that comes with adhering to someone else’s goals or timelines. In both cases, you’re facing a struggle and a challenge to articulate something about yourself.

And perhaps it’s just the struggle of self-fulfillment here that is a struggle worth undertaking. But it’s daunting to propose creating your own success criteria as a solution for most people, most of the time, because we don’t necessarily consider it a possibility. It’s a lot easier for you to follow someone else’s orders, often. That’s the norm in most cases, for most labor jobs. But is that a realistic framework? Is self-fulfillment through one’s own definition of value something achievable for most people in most jobs?

Brooke: A cynical side of me says, “No, not at all.” I’m not sure what the answer is to that question about how feasible that is at scale, or at what scale that is feasible. I know that the way we tell the success stories of others does not set us up well to tell positive stories and feel positively about our own experiences of success. In terms of these reflections of what constitutes the good life and all these kinds of things, there is a deep and rich history of that kind of inquiry in philosophy, in literature, in history, in art. I would say the vast majority of that tradition of discourse exists within the humanities, which is absolutely not the thing that we train people in.

The liberal arts refer literally to the art of liberation, specifically self-liberation, leading the life of a free citizen. That is what we denigrate when we denigrate liberal arts: the freedom of the individual. And so, we end up in this situation where not only does the economy not provide us many opportunities to define our own success, but we also are not very effectively tooled up to be able to entertain those conversations, even if we have the luxury of doing so.

Not only does the economy not provide us with very many opportunities to have the luxury of defining our own success, but we also are not very effectively tooled up from the start to be able to entertain those conversations, even if we have the luxury of doing so.

Nathan: Okay, so let’s complement that. To provide a middle ground that’s a little bit more positive on how we might achieve this for more people, more of the time, I’m going to point our attention to another New Yorker article. This article points to the fact that no one’s talking about one element of Biden’s economic plan, which is called the Civilian Climate Corps. Basically, it’s plagiarism of FDR’s creation of the Civilian Conservation Corps, which was a government relief program for unemployed people, allowing them the chance to do “voluntary” public work for a stipend. This program was a resounding success that ultimately employed over three million young men. Nowadays, for the Civilian Climate Corps, the jobs work toward climate change relief efforts. 

These jobs are manual labor, but they serve a greater purpose. Obviously, they’re in pursuit of increasing employment in a time of significant struggles to find work and rebuild a sort of national pride in the kind of labor they’re doing. They’re centering it around ideals that perhaps people can discover for themselves through the work. This is not necessarily because the workers decide what to do every day when they wake up, but, within a defined project, perhaps people can find fulfillment according to values that are somewhat removed from what they’re doing day-to-day.

So someone who’s working, whether it involves building a wind turbine or tallying numbers on spreadsheets, they are working toward a purpose that is solving the global issue of climate change. That’s a different job than someone who is doing the same thing towards facilitating fracking, for example, which is obviously harmful to the climate. So, these people may be doing similar jobs day-to-day – tallying numbers, looking at spreadsheets, working in factories – but their efforts are toward directly opposite goals. 

There is a difference there, I think, in the kind of values that are being pursued that you can sign up for. You can have your tasks delegated to you in this role, but you may understand something about yourself and what you believe in through the way these jobs shape your character or introduce you to a new environment and a new set of values. You may have even implicitly held these values before, but you might bootstrap that process of self-discovery through a more organized system. I think to some extent that connects to the idea of finding your values in your work, even if you don’t have to create each of those values yourself. You could find them through labor.

Brooke: Yeah, absolutely. I think about Marx and the alienation of the means of production from the value which is produced. There is a way for that to run in the opposite direction, as well. Marx’s point, I think, only has bite because when it’s done properly, reconnecting with the products of our labour can be a source of value. And so, we shouldn’t think of our values as one thing over here on this side of the firewall and our work as this kind of thing over here on the other side of the firewall, and that values need to come first and that’s going to inform the type of work that we do. Absolutely not. There is not a clean firewall between those things. Where you work is going to inform your values, and that can be a good thing. But it can be a negative thing as well, when you’ve got too much bullshit in the ecosystem that actually isn’t worried about the outcome. It’s only worried about the appearance of those who are actors in the process.

I think worrying about the outcome is essential to the general process of discovering your own values. If you already have a good idea of your values, work helps you execute them in the real world. If you’re still discovering them, which most of us are, then meaningful labor has to be a process for doing that, alongside other institutions. 

Nathan: I think worrying about the outcome is essential to the general process of discovering your own values. If you already have a good idea of your values, work helps you execute them in the real world. If you’re still discovering them, which most of us are, then meaningful labor has to be a process for doing that, alongside other institutions. Anthony Appiah has this idea that governments and families are traditional loci of people’s values, and also places of expression. When we debate the legitimacy of broadening the notion of family to more than two parents or non-heterosexual couples, for example, we’re redefining our values through the institutions of family. I think labor is also one of these institutions that we can use to execute and redefine our values.

Brooke: Yeah, I absolutely agree. In a way, we should ask ourselves how we ever believed otherwise. How could we have ever thought that we spend half of our waking hours—sometimes more—doing something and that that thing wouldn’t have a massive impact on our values and perceptions of the world?

Nathan: Alright – let’s leave it at that. Thank you both very much for sitting down with me and having this discussion. More to come soon. 

Why We Misjudge the Risks of New COVID Strains

How well-calibrated are we on the risk of the new COVID strains, coming from the U.K., South Africa, and elsewhere, and discovered across the U.S.?1,2,3

Health authorities, various experts, and the media have put the spotlight on concerns over vaccine effectiveness and claimed the strains are likely not going to escape the vaccine. They thus assert there’s no need to ring the alarm.4,5,6,7

For instance, Adm. Brett Giroir, the former White House coronavirus testing czar, gave an interview on December 27 to Fox News Sunday, where he said “It is not any more serious than the normal strains of COVID… we have no evidence to suggest, nor do we believe that the vaccine would not be effective.”8

Yet this proclamation may have come too soon: new research is showing that existing vaccines do indeed appear to be less effective against certain variants of the coronavirus.9 However, because scientists may quickly update the mRNA-based vaccines from Moderna and Pfizer/BioNTech to make them fully effective against these new COVID variants, most experts are confident that we will still be able to contain new variants.10 If the U.S. does indeed achieve herd immunity through mass vaccination by the end of 2021, these new strains may not make much of a difference.11

But there’s another element of these brand-new strains that should make you much more ready to update your risk assessment and change your plans: they’re a lot more infectious. However, the possible impact of their infectiousness has not received nearly enough attention. Such complacency is reminiscent of our “sleepwalking” response to the virus’s emergence last year, in spite of direct advance warning from myself and other experts in behavioral science and risk management.12,13,14 That’s why so many in the U.S. and elsewhere have not succeeded in the effort to plan and adapt successfully to this situation.15 Our brains don’t deal well with such threats, making it much more difficult to respond to slow-moving and high-impact train wrecks such as the pandemic itself, or to a much more infectious variant.

Are the New Strains Really More Infectious?

Researchers believe the U.K. strain to be anywhere from 56–70% more contagious than the previous dominant variant,16 meaning each individual person who gets the new strain infects 56-70% more people than the older COVID strain. The new variant quickly came to dominate the old strain of COVID in Southeast England, going from less than 1% of all tested samples at the start of November to over two-thirds by mid-December.17

Image courtesy of BBC

The South African strain appears even more infectious than the new U.K. strain.18 It came to dominate the country quickly: first found in October, it was responsible for over 80% of all new COVID cases by the end of December.19,20

To corroborate this research, let’s compare new daily COVID cases per million people over December.

Image courtesy of Our World In Data21

The U.K., U.S., Canada, Italy, and France all experienced a major rise in cases in September and October. That’s mainly because colder weather in the Northern Hemisphere drove people to interact indoors, where COVID spreads much more easily.

All these countries imposed various levels of lockdown in late October and early November. That decreased22 or stabilized23 their numbers by late November and early December, with the exception of a Thanksgiving-induced bump in the U.S. that stabilized by mid-December.

The U.K.’s numbers, however, look different. After going from under 250 new cases per day at the start of October, that number went to nearly 400 by mid-November. By early December, the country successfully bent the curve, with new cases dipping below 250 again. However—unlike any of the other countries—it then experienced a sharp uptick from 250 in early December to over 500 new cases daily by the end of December. Given that the U.K. didn’t experience any holiday bumps or changes in government policy, the new strain is the most likely culprit for this deadly surge.

By contrast, South Africa is in the Southern Hemisphere, and it’s summer there from December to February. Given the warm weather, COVID cases would be expected to decrease, not increase. However, South Africa experienced a major surge, from below 30 cases per day in early November to over 180 by the end of December.24 Given the absence of policy changes or other viable explanations, the new COVID variant is almost certainly to blame.

Let’s not make the same mistakes we made in the beginning of the pandemic, making sure to take seriously the real threat of these new strains.25 So what are the implications?

It took the U.K. about two weeks to double its numbers from December 10, at 240 new cases per day, to December 24, with 506 cases. In South Africa, we see a similar pattern of doubling, from 86 cases on December 10 to 182 on December 24. In both cases, it took about two months from the discovery of the variant to the start of the case surge as the new variant took over.26

Given that both variants had likely landed in the U.S. by late November, we can anticipate they have started taking over and could become predominant. Indeed, the CDC projects that the U.K. variant will be dominant in the U.S. by the end of March.27 Moreover, a COVID variant similar to the U.K. one became predominant in Central Ohio in an even shorter time frame, over three weeks from late December into early January, according to a study from Ohio State University study.28

As of March 17, only 21% of the U.S. population has been vaccinated. Given the already-overwhelmed situation within our hospitals, the potential threat of a caseload that doubles every two weeks—as in the U.K. and South Africa—is crystal clear.29,30 Even a 50% increase every two weeks would have a drastic impact.

It’s hard to imagine, but the numbers don’t lie. And a recent study31 suggests that in the U.S., the U.K. strain was responsible for over 3% of all cases by the end of January and, at that point, was doubling every 10 days. The alarm bells should be ringing very loudly.

Why we ignore slow-moving train wrecks 

How worried do you feel right now? Is your heartbeat pounding, do you feel a wave of heat, are your palms sweating? Are you thinking about how to change all your plans for the next six months?

Probably not, because our minds aren’t well-adapted to processing the implications of these seemingly abstract numbers. We fall into dangerous judgment errors that scholars in cognitive neuroscience, psychology, and behavioral economics call cognitive biases.32 They result from a combination of our evolutionary makeup and also the particular ways that our brains are wired.33

Such mental blindspots impact all areas of our lives, from health to politics and even our shopping habits.15,34,35,36 Fortunately, recent research has shown effective and pragmatic strategies to defeat these dangerous judgment errors. This starts with recognizing the biases that are most likely to harm us in the pandemic, by causing us to react poorly to slow-moving train wrecks.37

Normalcy Bias

The normalcy bias refers to the fact that our intuitions drive us to feel that the future, at least in the short and medium terms, will function in roughly the same way as the past: normally. As a result, we drastically underestimate both the likelihood of a serious disruption occurring and the impact of one if it does occur.38,39 Indeed, while evaluating the future based on past experience often works, it’s not a good approach for new situations like a novel coronavirus variant.

This bias leads individuals, businesses, and governments to fail to prepare even nearly as well as they should forc potential catastrophes, especially slow-moving disasters such as pandemics. Moreover, in the midst of the event itself, people react much more slowly than they ideally should, getting stuck in the mode of gathering information instead of deciding and acting.

It’s not surprising that going with our gut reactions leads us astray in response to catastrophes that slowly gather steam as they approach. Our threat response causes us, based on our personalities and predispositions, to go into fight-or-flight mode.

The fight response causes us to take immediate action in response to the problem, such as buying toilet paper and guns, as so many did in response to COVID-19. Others fall into the flight response, essentially freezing and ignoring important information—as was the case in many places where social distancing measures were slow to be implemented. It might also mean actual flight, where people try to leave their area for another one that’s perceived as safer, an option that proved popular for those with means.40

Of course, neither of these is the right response for the situation at hand. Fleeing an area or stocking up is fine for the typical disasters that may strike a region, such as a hurricane that might cause major flooding (as happened in Houston in September 2019). It’s not a good fit for the COVID-19 pandemic itself, nor the threat of the new strains.

Hyperbolic Discounting

In the ancient savanna environment, our ancestors had to live in and for the moment, since they couldn’t effectively invest resources to improve their future states (it’s not like they could freeze the meat of the mammoths they killed). Right now, we have many ways of investing in our future lives, such as saving money in banks. Yet our instincts orient us toward short-term rewards at the expense of our long-term futures, a mental blindspot called hyperbolic discounting.41 This cognitive bias causes us to underestimate the eventual impacts of clear trends, such as a more infectious strain of COVID.

That’s why leaders across Europe and the U.S.—in politics, media, and business alike—largely failed to act in a timely manner to address COVID-19 in the first place. Despite extensive evidence showing that outbreaks were growing exponentially, many leaders rejected calls in the early stages of outbreaks to take the course of action shown to be effective in China and other East Asian countries. Officials did not impose shutdowns and social distancing, along with thorough testing, contact tracing, and isolation. The media did not, by and large, ring alarm bells. Following their examples, business leaders largely proceeded with business as usual. Only a few large companies, such as Twitter, Square, and Google, led the way in encouraging employees to work from home, while the English Premier League led the way in canceling sports matches.

The political, media, and business leaders who failed to take action quickly must have felt in their gut that the short-term sacrifice of a shutdown outweighed the long-term benefits of decreasing the impact of the pandemic. It took enormous efforts to convince them otherwise. And despite the clear trend lines from the example of the U.K. and South Africa, we see that in the U.S. and parts of Canada, leaders more or less continue business as usual.

Planning Fallacy

We tend to feel optimistic about our plans: we made them, and therefore the plans must be good, right? We intuitively feel that the future will play out according to these plans. That mental blindspot, the planning fallacy, threatens our ability to prepare effectively for potential problems, and to pivot quickly when they happen.42,43,44 That includes neglecting both major unknown threats, also known as black swans, as well as, more surprisingly, big and obvious threats, known as “gray rhinos.”45 The new strains fall into the latter category, as it was never doubted that the novel coronavirus would mutate (as all viruses do).

Consider how the U.S. and many European countries delayed imposing restrictions in the initial stages of the pandemic, despite clear evidence they were needed. Similarly, in the fall, despite extensive warnings from experts of an upcoming surge, European countries and U.S. states waited until a sharp spike in hospitalizations and deaths to reimpose restrictions.46,47,48

We’re seeing the same happen right now with the new COVID strains. Despite the clear need to act, there’s no action, at least in the U.S., where many states are reducing coronavirus public health restrictions, despite rising cases and deaths.49,50

Attentional Bias

Attentional bias refers to our tendency to pay attention to information that we find most emotionally engaging, and ignore information that we don’t.51,52,53 Given the intense, in-the-moment nature of threats and opportunities in the ancestral savanna, this bias is understandable. Yet in the modern environment, sometimes information that doesn’t feel emotionally salient is actually really important.

In recent days, the headlines in the U.S. focused almost exclusively on the inauguration of the new administration and impeachment trial of Donald Trump, with brief mentions of the 400,000 death toll and slow vaccine rollout. Health officials and scientific experts have focused on discussing the vaccine rollout and how to advance it. There’s almost no discussion of the clear trends of the new strains, and the need to act before it’s too late.

The same problem plagued us early in the pandemic. For example, the fact that the novel coronavirus originated in Wuhan, China, and caused massive sickness and deaths there, didn’t draw much attention as a salient potential threat among Europeans and Americans. It proved too easy to dismiss the importance of the outbreak in Wuhan, due to stereotypical and inaccurate visions of the Chinese heartland as full of backwoods peasants.

In reality, Wuhan is a global metropolis. The largest city in central China, it has over 11 million people and produced over $22.5 billion (USD) in 2018. It has a good healthcare system, strengthened substantially after the SARS pandemic. A major travel hub, Wuhan’s nickname is “the Chicago of China”; it had over 500 international flights per day before the outbreak. If we assume an average of 250 people per plane, that’s 10,000 people a day flying out of Wuhan to the world.

Europeans and Americans, with the exception of a small number of experts, failed to perceive the threat to themselves from the breakdown of Wuhan’s solid healthcare system as it became overwhelmed by COVID-19. They arrogantly assumed this breakdown pointed to the backwardness of China, rather than accurately concluding that any modern medical system could become overwhelmed in the face of the novel coronavirus.

They also failed to recognize the thorough interconnectedness of Wuhan to the rest of the globe. Case in point: the first case of COVID-19 in the U.S. was diagnosed in a traveler from Wuhan in the state of Washington. The first epicenter of COVID-19 in Europe, northern Italy, has particularly close ties to Wuhan due to the robust fashion and garment industry in Milan and other northern Italian cities, which outsources manufacturing to Wuhan.54

Europeans only began paying serious attention to COVID-19 when it began to take root in Italy. Americans, in turn, started to pay attention when COVID-19 surfaced in Washington state.

Do you know who didn’t ignore Wuhan, besides a small number of experts? Those to whom this city felt emotionally salient. Those who understood that Wuhan could be fairly compared to Chicago in the U.S., Manchester in the U.K., or Frankfurt in Germany.

That emotional salience helps explain why many Chinese communities in Europe and America acted quickly and effectively to minimize the impact of COVID-19. For instance, the 50,000 Chinese in Pareto, Italy—a quarter of the city’s residents—went into voluntary lockdown at the end of January. That’s three weeks before the first recorded infection in Italy.55

With their connections to China, the members of these communities could envision what was to come, and spread the word to close their businesses, stay home as much as possible, and wear masks in the rare event that they ventured outside. That helps explain why, according to Renzo Berti, the top state health official for the region, none of the Chinese residents in Pareto got COVID-19 and the town’s infection rate was half of the Italian average, 62 cases per 100,000 people (compared to 115 for the entire country).

Imagine what would have happened if everyone behaved like these communities? Businesses, individuals, and governments acting together could have prevented the enormous death toll and economic devastation from the novel coronavirus. Yet our attentional bias led us astray. The same thing is happening right now with the new strains.

Conclusion

While this evidence may feel unreal, keep in mind that’s simply our cognitive biases pushing us to feel like that, just like they did early in the pandemic. We’re miscalibrated on the risks of these new strains, and only by taking into account our mental blindspots can we correct for them and take the steps needed to prepare effectively for the high likelihood of a major surge in spring 2021.

TDL Brief: COVID and the Climate

As the global climate situation becomes increasingly urgent, the movement against climate change has gained more and more traction. Despite the protests, the conferences, and the new policies, many are frustrated by the slow progress that is being made. 

In early 2020, the world was struck by a pandemic that brought daily life to a screeching halt. Among all the tragedy and the trauma brought on by COVID-19, many people found themselves searching for a silver lining. 

It is probably safe to say that we can all agree that the coronavirus pandemic has impacted nearly every facet of our lives. It has left nothing untouched – including the environment. The silver lining that so many were looking for was the realization that one of the many consequences of the pandemic might just be a reversal of some of the effects of climate change. The hope was that, with far fewer people travelling and less factories up and running, we could prove that the damage we caused could be undone. Then, that evidence could be used to push for lasting change. Perhaps it would not be so bad if life after the pandemic did not look exactly as it did before.

Although this bright spot in the pandemic certainly has value, there are many who look at it from a less optimistic perspective. They raise questions like “is the change we are seeing significant” and, if so, “is there any feasible way to make it last?”. These questions are too big to have just one easy answer; even experts are in disagreement about them. In spite of the uncertainty, perhaps this could still be used as a way to bring about change in a post-COVID world.

1. Present bias

By: Peter Masone, “Why did we mobilise for COVID-19 and not climate change?”, London School of Economics and Political Science, May 2020

It is undeniable that COVID-19 and climate change are both global crises that call for swift action. That being said, the way the world responded to the pandemic is strikingly different from our response to climate change. We adapted quickly to COVID-19. Policies were put in place to slow the spread, and we began wearing masks and limiting our contact with others. The pandemic called upon us to make unexpected sacrifices and completely turned our lives upside down. For the most part, people have complied with the public health measures and are doing their part to “flatten the curve”. Climate change, on the other hand, has not been received with nearly the same level of urgency and the actions taken to mitigate the effects of global warming have been minimal, at best. 

COVID-19 and climate change are both international crises that have serious implications for our future. While lifestyle changes at an individual level can be beneficial, they both require political action be taken against them, in order for real progress to be observed. The situations are similar, so why are our governments’ responses to them so drastically different? 

This discrepancy can be explained, in part, by a cognitive bias known as “present bias”. Present bias describes how we tend to value immediate payoff over future rewards, even if that future reward is substantially greater than the immediate one. The COVID-19 pandemic feels real and scary. It has created an atmosphere of fear and uncertainty, which has motivated us to act quickly to quell those anxieties. Furthermore, with the pandemic there is the feeling that if we all do our part to slow the spread there is a reward waiting for us in the near future. Namely, the end of the pandemic and the chance to get back into our old routines. Climate change, on the other hand, is a distant, abstract concept. Its effects can be observed but they do not always have a significant effect on people’s daily lives. We do not feel its effects immediately, so it does not incite the same level of fear within the population. Beyond that, we would have to make a lot of changes and sacrifices in order to undo the damage we have inflicted upon the planet. The payoff to ending the climate crisis would be extraordinary but, since we feel so disconnected from the issue and achieving that end goal would take many years, we do not feel the same incentive to act against climate change like we do for the COVID-19 pandemic.

2. Going green in global COVID recovery

By: “‘Green recovery’ from COVID-19 can slow climate change: UN environment report”, UN News, December 2020

Although we saw a small dip in global carbon emissions during the early stages of the pandemic, when the world was put on hold, we are still on track to boost global temperatures by 3 degrees Celsius in this century. That does not sound like much, but if it were to happen, the consequences would be devastating. However, the UN urges us not to lose hope, nor to give up the fight for a greener tomorrow. The UN Environment Programme executive director, Inger Anderson, announced that a “green recovery” from the COVID-19 pandemic might be enough to slow the rate of climate change, and pushed governments worldwide to allocate money to this initiative, while also making climate change a priority for 2021.

The report offers suggestions for changes on the large-scale, such as dedicating resources to reforestation of exploited lands, prioritizing zero-emissions technologies, and decreasing fossil fuel subsidies. Furthermore, due to the high demand from consumers, it has become clear that shipping and aviation technologies can no longer operate as they are. They must move away from fossil fuels and towards greener sources of energy, in order to decrease greenhouse gas emissions.

Recommendations are also made for changes we can all make in our daily lives. Take the train instead of booking a seat on a short domestic flight. Carpool to work or, even better, take up cycling. Make a concerted effort to limit food waste and to make your home more energy efficient. These small efforts add up to big changes. 

As we come out of this pandemic, we cannot simply go back to the same lives we were living before. While we recover from the toll COVID took on us, let us help the planet recover from the toll we have taken on it.

3. Public health

By: “Climate and COVID-19: Converging crises”, The Lancet, December 2020

COVID-19 has taken news headlines by storm and at the forefront of almost everyone’s minds. This is understandable, considering the scope of the issue, however we cannot let our attention be redirected away from other global issues entirely, especially not ones like climate change, which is not only a pressing issue, but one that is inextricably linked to the pandemic.

Research has shown that health and the climate go hand-in-hand. For example, air pollution is linked to asthma, food insecurity can result in poor diets, and the heat can pose a risk to the elderly. There are commonalities between the pandemic and the climate crises. Both have led to countless preventable deaths and both are spreading resources in healthcare thin. Furthermore, the same factors that drive climate change contribute to the propagation of diseases like coronavirus. Two major examples are international travel and urbanization, which has resulted in high density populations in cities. Another striking similarity between coronavirus and climate change is that they both disproportionately affect marginalized communities. 

The emergence of COVID-19 vaccines has been the light at the end of this very, very long tunnel. That being said, the public health crisis will not end with this pandemic. Governments are focusing on policies to bolster economic recovery from COVID-19, but they also need to be pushing policies to encourage a green recovery, because climate change is a public health crisis in its own right.

4. Nice try, but no cigar

By: Matt McGrath, “UN report: Covid crisis does little to slow climate change”, BBC News, September 2020

In the early days of the pandemic, the bright spot in the darkness for many was the possibility that carbon emissions might be dropping enough to slow climate change. Unfortunately, a report from the United Nations suggests that the reductions were too minimal to have any real impact.

There is no arguing the fact that lockdown measures had an immediate effect on greenhouse gas emissions. In April 2020, the daily levels were 17% lower than they were in April 2019. It is possible that this level of reduction may have been enough to make a difference. Unfortunately, it was not maintained for very long. By June, daily greenhouse gas emissions were only 5% lower than they had been in the previous year. Even with the drop in greenhouse gas emissions on the ground, levels of carbon dioxide in the atmosphere continue to increase and we are still on track for 2016-2020 to be the hottest five-year period on record. That is not the kind of record-breaking feat that garners much celebration.

One poignant statement from this report is that, in order to effectively combat global warming, a “pandemic-sized” reduction in greenhouse gas emissions would be required every year until 2030. Recall that this drop in greenhouse gas emissions resulted from a global lockdown. The level of change needed to reverse the effects of climate change is quite significant. In order to make the changes necessary and to heal the damage we have inflicted on the planet, the UN Secretary General, Antonio Guterres, announced that we must emerge from the pandemic ready to take on the climate crisis and achieve sustainable development, which can only be accomplished through “science, solidarity, and solutions”. 

Dr. Mitesh Patel on Nudging, Tech, and Health Care

Background

Health care is filled with choices. For providers, this can be choosing the right prescription, deciding whether to (de)escalate treatment, offering the influenza vaccination, or referring a patient to a specialist. For patients, it is reflected in the decisions underlying their willingness to accept and adhere to the doctor’s orders.

High-quality health care depends heavily on high-quality decision-making. But far too often, this isn’t the reality of medical practice. How choices are framed in the context of healthcare delivery has received very little attention until very recently. This is why I wanted to sit down and talk with Mitesh Patel, MD, MBA.

Dr. Patel is the Director of the Penn Medicine Nudge Unit, the world’s first behavioral design team embedded within a health system. He is also the Associate Director at the Center for Health Incentives and Behavioral Economics, on faculty at the Penn Medicine Center for Health Care Innovation and the Center for Health Incentives and Behavioral Economics, a Staff Physician at the Crescenz VA Medical Center in Philadelphia, and a Senior Fellow at the Leonard Davis Institute of Health Economics.

His research combines behavioral science with scalable technology platforms such as electronic health records and wearable devices to improve health and health care. He has led more than 25 clinical trials in partnership with health systems, insurers, employers, and community organizations, that testing interventions such as nudges, incentives, and gamification to change clinician and patient behavior. This work includes digital health interventions that use wearable devices and smartphones, and health system interventions using the electronic health record.

As an aspiring physician interested in the leveraging behavioral science to design and test approaches to steer medical decision-making toward higher value and improved patient outcomes, I was thrilled to interview Dr. Patel.

Interview

Sanketh: You describe that human behavior is the final common pathway for nearly every advance in medicine. Could you elaborate on this?

Dr. Patel: In order for medical therapies, tests, and treatments to benefit patients, two things need to obviously happen. The clinician needs to recognize that a patient meets criteria for that test, therapy, or treatment, and then prescribe or order it. Then, the patient has to understand how they will benefit, and then adhere by either taking the test or medication. Unfortunately, one or the other often does not happen. This creates opportunities for ways to improve the delivery of health care. 

About a third of health care starts to become unnecessary or redundant because we are doing too many things that we have already done to address issues that would actually be better solved through addressing these gaps in decision-making. 

Sanketh: Why do you think health is one of the last frontiers to embrace behavioral design as a creative, human-centered approach to defining and solving problems?

Dr. Patel: This line of work has gained much more traction in fields like finance, retail, and even entertainment. Health care, however, is much slower because it is highly regulated, and rightfully so. Nudges are really influential and if they are designed in the wrong manner, it can lead to adverse events. 

We can’t simply apply typical marketing techniques to health care. Health care has many logistical challenges because there are multiple stakeholders involved. Someone can’t simply advertise colon cancer screening to a patient. It has to be ordered by a clinician or health system. Then, on the patient level, patients often do not know what test or treatment is best for them, creating a communication gap that must be filled. 

Sanketh: In this discussion of workflow in health care, the topic of electronic health records (EHRs) is inevitable. How can we use EHRs as a vehicle to nudge providers and modify behaviors?

Dr. Patel: EHRs are now widely adopted by clinicians and health systems by more than 90% of them. It used to be that all of our health care decisions were made either by verbal communication or on pen and paper. Now, everything is funneled through the EHR, which gives us visibility into the decisions that are being made. It also introduces an environment in which those decisions can be influenced. 

A lot of effort goes into implementing EHRs and optimizing them, but there is still a lot of low-hanging fruit in terms of how we can change perhaps the framing of information, the ordering of options, the default option, when we prompt someone with an alert versus not, and so on. There are a lot of different tools in the choice environment within EHRs that help guide clinicians towards the right decision at the right time, and I think that’s where the opportunity really is—a way to do that systematically, to learn, and then to spread what works. In parallel, stopping what is not working.

Sanketh: There’s a lot of controversy around nudges because many view them as coercion or manipulation. However, in many cases, there is always an inherent default nudge that is already influencing behavior in some way. So implementing a nudge is more about adjusting the environment to redirect the nudge towards improved outcomes in a more structured way. Otherwise, the nudge is still present, but in an unstructured manner that could actually be promoting poor outcomes. 

Dr. Patel: That’s a great point. Many clinicians express how they have been an expert in their field for X decades and trained for Y years. They become very skeptical when they hear that someone else can influence their decisions regarding their patients’ outcomes in the right direction. We often have to make people aware that the current design of the electronic health record is already nudging them in one direction or another. Actually, in many cases, it causes extra work to do the things you already want to do. In other cases, it’s actually misaligned with what’s best for patients. So the question is not whether we should nudge, but rather how do we strategically align nudges in the appropriate direction while keeping patient outcomes and the burden on clinicians in mind.

Sanketh: It’s also widely recognized that EHRs are a leading contributor to provider burnout, and burnout contributes to medical errors, increased susceptibility to biases, and poor medical decision-making. Could you speak a little bit to how nudges can potentially mitigate burnout to reduce its downstream adverse effects on patient care?

Dr. Patel: When a lot of people hear the term nudges, they think that they’re going to receive an alert or extra information that they have to process. The most successful nudges are actually the ones that reduce burden on clinicians. 

There’s all kinds of alerts and information in the electronic health record which are not useful. However, there is a lack of research that has looked at this in a systematic way. By comparing alerts in one setting to another setting with no alerts, we can actually determine the effectiveness of these alerts and potentially provide strong evidence for turning off any ineffective and burdensome alerts, which really is just noise at that point. 

Then, we can design nudges that, instead of prompting clinicians, might default clinicians towards the easiest pathway. For example, we had a nudge for flu vaccination where instead of alerting physicians, we had an alert that went to medical assistants when they triage the patient for the vaccine. The alert simply told the patient, “Hey, it’s flu season you should talk to your doctor about this.” If the patient wanted the vaccine, the clinician would sign and it was easy. If they didn’t, they would then spend time talking to the patient about their concerns and try to address those. In this situation, the clinician is spending more time talking with patients, rather than dealing with those alerts, so I think there’s a lot of opportunities where nudges can reduce the burden on clinicians.

Another example is a study we published last month on cardiac rehab. Cardiac rehab is an evidence-based program consisting of 3 sessions a week, for 12 weeks, of structured exercise with a cardiologist on site who gives you advice on exercise, diet, and medications. It is found to reduce mortality and hospital readmissions by about 30%. In Penn’s health system and many others, however, less than 20% of patients were being referred for cardiac rehab. So let’s say 100 patients come in this month with heart attack or stroke. 85 would go out the door and never be referred for cardiac rehab. When you ask clinicians whether they would offer cardiac rehab to someone who just had a heart attack or stroke, they say that everyone should get it. It’s effective and paid for by insurance.

So we then started digging into why clinicians aren’t actually referring patients to this, and we learned that the process for making a referral was very burdensome for clinicians. They had to fill out a paper form consisting of about 20 fields amidst a busy day of rounds. Essentially, it was their responsibility to identify who is and isn’t eligible. 

We then spent three months going back and forth on how we could automate this. What we came up with was linking the form to our secure text messaging system so that care managers on rounds got the names and locations of the two or three patients every day who should be referred. The form was templated, so the clinician no longer had to fill it out. They still had to sign, but now it was electronic. And it was also designed so that it was opt-out, meaning that the patient would be automatically referred unless he or she presented any concerns. Then, the care manager went to the patient with a vetted list of cardiac rehab centers, near their home, covered by their insurance, and just made it easy for the patient to be able to do that.

We found that the cardiac rehab referral rate went from 15% to 85% and [was] sustained that way for two and a half years. We had two control hospitals which, once they found out about these great results at the hospital across the street, their rates went from 15% to 25%. Noticeably a smaller increase, [which is] a demonstration of how awareness alone is not enough. And if you think about it, all we did was we made it easier for clinicians to do something that they already wanted to do by leveraging technology.

Sanketh: Wow, these results are beyond astonishing, and what’s beautiful is that the intervention was so simple. It reminds me of the classic go-to example for defaults, which is the case of organ donation. In the United States, people have to opt-in to be an organ donor. In other countries, the default is being an organ donor with the option to opt-out. This is likely why we see a smaller percentage of the U.S. population being an organ donor.

Dr. Patel: Yes, exactly. Defaults are one of the most powerful measures you can implement. 

Sanketh: Now shifting gears, in a recent article of yours, you discuss how behavioral phenotyping can guide digital health design and innovation. Is there a trade-off between the personalization of a solution, and the scalability of that solution?

Dr. Patel: Most behavioral interventions currently designed are one-size-fits-all, which assumes that a single intervention will work for people in very different ways. We are currently figuring out the best methods to personalize these interventions. Newer forms of technology make it easier to personalize, and it’s oftentimes much simpler than people think. For example, delivering a personalized nudge to a million patients does not really mean sending out a million different messages. Instead, it means stratifying the population into five or six groups or however many that makes sense, and then sending out five or six messages. Oftentimes, it’s nothing more than changing a couple words in a text message. In some settings, however, it can be much more difficult to do that. We need to be careful in picking the situations where phenotyping and personalizing nudges can be implemented at scale.

Facebook and Google do this every day—you get different search results based on your location. Your feed will be different tomorrow based on the actions you took today. We need to find opportunities for such algorithms in the personalization of digital medicine.

Sanketh: We talked about how scalable technologies and digital platforms can democratize nudges. This applies to patient decision-making. For example, medication adherence is a huge problem in health care, and nudges have been proven to be effective. However, in the context of some patient populations, there are communities that are not as well-versed with technology. I actually currently work at a low-income clinic that serves uninsured patients, and many patients only use their phones to make phone calls. Even those with more advanced smartphones are not very aware of the phone’s full capabilities. How can we nudge such patient populations through means rather than technology?

Dr. Patel: This is another area where we have to think about phenotyping because a lot of tech solutions may not make as much sense for people that are less tech-savvy or don’t have access to technology, especially when they are expensive products like wearable devices. There are many ways we can deliver the same behavioral insights from such interventions in low-tech forms. For example, rather than fancy apps, we can simply send text messages or emails. 

We can also think about the way that we frame information that they’re getting already if they’re coming in to see their primary care doctor and they’re having a conversation. Many patients will still get access to different types of materials either electronically or on paper, and so I think the same insights apply. It’s just the delivery channel that might be different.

Interestingly, we even find for people that are tech-savvy, the low-tech solutions often work better. Sending someone a simple text message often gets higher engagement than asking them to communicate like a chatbot. There are some cases where the high-tech and machine-learning algorithms really are useful, but in most cases you can use simpler approaches.

Sanketh: Could you elaborate a bit more about how different patient populations respond differently to different interventions?

Dr. Patel: Current COVID-19 vaccination efforts are a perfect example. Many health systems are sending information regarding vaccine registration through emails. Sometimes, people who work at these health systems don’t access their email as often because it’s not part of their workflow. With these people, you might be able to reach them better through text messaging than other channels. Another example is lower-income countries where phones are used more than any other platform. In such countries, they don’t use fancy patient portals through the EHR, and texting is really the primary approach.

Sanketh: Speaking of COVID-19, I saw that you wrote an article about how we can use nudges to improve vaccine uptake. After the announcement of the three vaccine candidates, I also wrote an article about the behavioral hesitations to vaccine uptake. Now that we’re about three months in, how well do you think current public health communication and policies are doing on this front from a behavioral science perspective?

Dr. Patel: Right now, there’s more demand for the vaccine than there are vaccines. At some point, that will change to a situation where we have ample supply but half the population is hesitant or hasn’t gotten the vaccine. This is when behavioral science approaches will be most important. We’ve also seen early on that racial minority groups are getting the vaccine at much lower rates than other groups. There is still a lot of opportunity to do much better, not only in incorporating behavioral science but then testing those approaches. At the moment, there’s no data on what health systems or public health groups are doing, what their messaging is, and whether or not it’s working. Each group is doing it independently.

This morning, we actually posted online results from a mega-trial I tweeted about earlier today, where we tested 20 different text messages for flu vaccination. Our results could inform the COVID-19 vaccination efforts. We found the best message was one that said the flu shot is “reserved for you,” instead of saying “available.” This simple change correlated to a 11% relative increase in flu vaccination at a Walmart, where almost a million people go into the pharmacies.

However, the word “reserved” created some worry for pharmacies because it falsely implies that pharmacies might be out of the vaccine. Instead, we used the term “waiting for you” for available vaccines. This language also resulted in a 10% increase in uptake. Such simple interventions could be incorporated into the COVID-19 vaccination effort to improve vaccination rates.

Sanketh: I love that. At the onset of the pandemic, the government came up with the term social distancing, which focuses on the negative rather than the positive. A much better alternative would probably have been “physical distancing” or perhaps even something that focused on the benefits, such as “safety distancing.” Hopefully we see some of those changes adopted on a wide scale in the context of COVID-19 vaccine uptake. 

Let’s talk about the future of the field. In 2019, the nudge unit hosted a “Nudges in Health Care” symposium. And after the two-day event, many leaders in health care were inspired and started to take steps to launch their own behavioral design team for their own hospitals. How quickly are hospitals currently adopting the concept? Will there be a point where every large academic hospital will eventually have one in the future? 

Dr. Patel: The idea is spreading rapidly. We have our next Symposium on May 20, 2021, and we already have over 400 attendees from 200 organizations around the world sign up for that. We’ve had multiple units start in the last couple years, and we have helped them through the process. There’s now one at Geisinger, UCLA, and the University of Washington. I think many more will come over the next few years, but in others it might be just incorporating some of these principles into existing teams. 

There are a lot of teams that are already working on health informatics and how we think about decisions in the electronic health record. Every house’s health system is different. Some of them have really strong behavioral science expertise, others don’t. Some of them really have a strong team in terms of implementing these kinds of perspectives [and] interventions; others, not so much. Our goal is to help accelerate that movement and try to help interested systems have the resources and materials they need.

Sanketh: So how difficult is it for a hospital without a strong behavioral science backbone to implement the findings from your research studies at the nudge unit?

Dr. Patel: It’s not that difficult to get started. We do have ways that they can interact with our team and other teams to communicate and get insights. But that’s something that we’re working on further facilitating.

Our goal is to make it less difficult. We do this by publishing all of our research for people to learn from and leverage. We’ve got examples of more than 50 projects, many of which have had really great successes and others which haven’t worked. We’d like people to learn from even our failures so that they avoid making the same mistakes. We hold the symposiums to disseminate the work. 

Sanketh: Everything revolves around economics. What’s the ROI for hospitals looking to implement a nudge unit?

Dr. Patel: The cost varies, and it’s highly dependent on whether institutions are hiring new people for their teams, for example. Our group is about 90% funded by grants. We get a little bit of funding from our health system foundation, which we then leverage to get more grants that accelerate the work.

The ROI can be incredible. The first project looked at prescription rates of generic versus brand name medications. Our intervention made generic medications the default options, and clinicians had to actively opt out in order to prescribe a brand name medication. Over the course of two years, this intervention saved about $32 million. And what’s amazing is that the intervention only took one hour to implement. So the potential for ROI is huge in that just one or two projects can overcome the cost of implementing these types of teams.

Sanketh: That’s amazing, I’m looking forward to the wide adoption of behavioral science in hospital systems, and I am excited to attend the 2021 Nudges in Health Care Symposium in May!

Wrapping it up

Nudges and reframing choice architectures is quickly gaining attention in the healthcare industry, and is quickly proving its value for the multiple stakeholders involved in health care. Especially with the rise of digital health innovation, leveraging insights from behavioral science offers a great opportunity to increase the personalization of health care, and in turn, steer medical decision-making towards improved outcomes.

The Value Perception Dilemma

A few years ago, Rory Sutherland, Vice Chairman at Ogilvy and an enthusiastic practitioner of behavioral science, commented on a peculiar aspect of human value perception. He used the example of the British post office to illustrate his point: people often neglect the value and benefits of the highly-efficient mail service that is capable of delivering lovely postcards across the city for £1.85. Imagine you had to do this yourself: it would deprive you of the best hours in the day and cost you about £25 in transportation. Nonetheless,

“A service for which I might willingly have paid £10, were no cheaper alternative to exist, is sold to me for £1.85 – and yet I do not walk out of the post office punching the air with the feeling that I just saved £8.15 on a £10 good. Instead, I just think, ‘Hey, £1.85 – that’s what a package costs to send, so I guess that’s what it’s worth, meh.’” 1

On the other hand, a lot of us have become devoted customers of name-brand coffee franchises, willing to defend their elaborate drinks with fancy names—even though we could just as easily make coffee at home for as little as a tenth of the price.

From a strictly rational standpoint, i.e. an objective cost-benefit analysis, the price we are willing to pay for each of these two goods does not make any sense. Yet, as we will see in the examples ahead, cognitive bias can arise from the particular context associated with artisanal coffees—and lacking at the post office—that pushes us to different perceptions.

Our brain’s dual systems

After years of behavioral science experiments, Daniel Kahneman and Amos Tversky identified two modes of processing that govern our decision-making processes. Simply named “System 1 and System 2,” they represent cognitive responses triggered by different contexts and the individual’s awareness of their own decision-making processes.

System 1 symbolizes an “autopilot mode,” where decisions are made below the level of consciousness: when this system is dominant, we tend to make choices based on availability, representativeness, and other mental shortcuts, which in turn lead us to gravitate towards fast, intuitive, and effortless options. System 1 is also associated with emotional responses, on which we often rely to reach quick conclusions. System 2, on the contrary, is a deliberative mode that looks for facts, logic, and consistency. It plays an investigative role, running cost-benefit analyses and seeking to make choices based on reason.

Although we generally see ourselves as essentially logical beings, it is when System 1 is active that the “magic,” as Rory Sutherland defines it, happens.2 In his latest book, Alchemy: The Surprising Power of Ideas That Don’t Make Sense, he evokes a paradigm shift towards psychological and behavioral approaches for everyday problems, as states his 4th rule of alchemy:

“The nature of our attention affects the nature of our experience.”

With this in mind, he emphasizes how we underestimate System 1’s share in our decision-making processes, thus leading us to design solutions and responses tailored to the wrong kind of brain. Our “fast” mode is powerful, and has control over our behavior most of the time. In fact, it cannot be shut off, and we have to exert a deliberate (and somtimes great) effort to activate System 2 and delegate choices to it.3 Still, some companies are narrowly focused on System 1, focusing all of their branding, user experience, and advertising efforts instead on appeals to plain rationality.

Context and value perception

In addition to our dual cognitive systems, there is a fundamental element of influence that can distort decisions in many different ways: context. In his best-seller Predictably Irrational, Dan Ariely demonstrates how we rely on inconsistent principles to make decisions.4

In a field experiment conducted with Kristina Shampanier and Nina Mazar, Ariely set up a table at a public building and offered two kinds of chocolate: Lindt truffles for 15 cents and Hershey Kisses for one cent. Each “customer” could only choose one of them.

The results were not surprising. As people compared price and quality, about 73% opted for the truffles and 27% for the Kiss. However, the experimenters then decided to test if a “free” condition could change those preferences by reducing the price of both chocolates by one cent—Lindt truffles were then sold for 14 cents and the Hershey Kiss was given for free. Even though the price of each option had decreased by the same amount, the change had a major impact on people’s preferences: Now only 31% opted for the truffles.5Voilà!

Another experiment by Princeton psychologists John Darley and Daniel Batson illustrates the power of another contextual element: time pressure.6 In this study, 40 trainee Catholic priests completed a questionnaire about their motivation to join religious services. Right after that, they were asked to record a five-minute talk. Priests were then split into three different groups, where they were put under different levels of time pressure. Individuals from the first group, the “high-hurry” condition, were told “Oh, you’re late. They were expecting you a few minutes ago. The assistant should be waiting for you so you’d better hurry.” In the second group, the “intermediate-hurry” condition, arrivals were told, “The assistant is ready for you, so please go right over.” And finally, in the “low-hurry” condition, the message was, “It’ll be a few minutes before they’re ready for you, but you might as well head on over. If you wait over there, it shouldn’t be long.”

As the participants made their way to their destination, they passed a confederate (somebody who was secretly in on the experiment) pretending to be in distress. Darley and Batson wanted to investigate how many trainees would stop, and if time pressure would influence helping behavior.

The results confirmed their hypothesis: in the high-hurry condition, only 10% of the participants stopped, compared to 45% in the intermediate-hurry condition and 63% in the low-hurry condition. As Richard Shotton explains, “The situation, not the person, determined the behavior.”7

Customer experience and value perception

Back to our first example. Both the postal service and the gourmet coffee franchise work their messages towards the customers, but in doing so, they appeal to different systems.

The first one, the postal service, is focused on a more rational approach, emphasizing efficiency and speed. Even though this approach is valid, it depends, as we have seen above, on a specific level of processing. If the message recipient is not in the right state to parse this information deeply (System 2), those benefits could potentially be bypassed—customers will not even think about them.

Remember your last experience with the postal service: did you pay attention to any details about the postal service and the journey your mail would soon be taking, or were you simply eager to leave as soon as your package was posted? I bet you were staring at your phone for most—if not all—of the time you spent there. When we’re in autopilot mode like this, it is hard for us to perceive any value in the services we use.

On the other side, the gourmet coffee place provides its customers with an aura of luxury, relaxation, and detachment from “the outside world.” As you walk into the shop, you are greeted by a combination of aromas, sensations, and memories. Buying a drink is an immersive process, embedded in a specific culture. We follow a particular set of social norms that guide our movements, perceptions, and even our vocabulary—coffee comes with its own fancy lingo.8,9 You are not just buying a beverage; you are part of a tribe of coffee connoisseurs, relishing a Venti-sized, extra-dry latte. It does not even matter how much it costs, does it?

Hidden features, hidden perks, hidden… gorillas?

Back in 1999, psychologists Daniel Simons and Christopher Chabris ran an experiment to test the extent to which we are unaware of environmental details.10 They showed participants a video where two groups of people, one in white shirts and one in black shirts, are moving around and throwing basketballs back and forth. They asked volunteers to count the number of times the people in white shirts passed the ball. The task demands a high level of attention, in order to avoid being distracted by the players in black shirts. Try it for yourself:

Did you notice the trick? In the middle of the experiment, a person in a gorilla suit walks across the scene. If you missed it, you’re not alone: although most people could provide the right number of ball passes, almost half of the participants did not notice the gorilla.

This paper demonstrates that we often do not detect large changes to objects and scenes—what Simons & Chabris called “change blindness.” Furthermore, and as the authors stated, the “likelihood of noticing an unexpected object depends on the similarity of that object to other objects in the display and on how difficult the priming monitoring task is. Interestingly, the spatial proximity of the critical unattended object to attended locations does not appear to affect detection, suggesting that observers attend to objects and events, not spatial positions.”

In other words: even if something noteworthy is taking place around us, we may not process it if our attention is on something else. This is relevant as a lot of critical characteristics—and benefits—from the postal service can easily go unnoticed, even when they are clearly mentioned to us. As Daniel Kahneman once said, “What you see is all there is.”11

Other cognitive biases can further serve to distract us from important (but uninteresting) information. For example, the halo effect can often prevent us from thinking critically about other companies, brands, products, services, and other people. We form impressions based on a single aspect that is often unrelated to the main part of it—for example, when a product is aesthetically pleasing, we might also perceive it to be of higher quality, even if it isn’t really. Consequently, in places like the post office, spartan product design and an overall unengaging customer experience may lead us to a mistaken impression of this service as trivial and low-value.

Takeaways

Companies and brands can organize information and highlight aspects that will activate responses from one of our cognitive systems. Knowing which one of them to play to is a fundamental, though commonly unobserved, strategy to create value and guarantee it is perceived by customers. After all, this is not a self-evident aspect, especially if you are focusing your message on rational events alone.

Our example of the mail service could be improved by incorporating elements that grab our attention. In September 2020, for example, the Brazilian Post Service celebrated the day one of the country’s most famous writers, Clarice Lispector, would reach her centenary with a special stamp.12 Another interesting idea, this one associated with donation campaigns during Christmas time, was the creation of a mascot to promote a social service they have been running for the past thirty years, where thousands of kids from lower-income communities send letters to Santa Claus are converted in gifts that Santa himself takes to them.

One last example comes from Rory Sutherland’s popular TED Talk, when he recalls the British Post Office’s tremendous effort to improve their first-class mail success rate from 98% to 99%, an initiative that “almost broke the organization.”13 Sutherland points out that they could have achieved a better outcome by focusing on the perception average citizens had of the service: if you asked people on the street what percentage of first-class mail arrived the next day, most of them would likely give some number much lower than 98%. Raising the actual rate by 1% was probably much less effective than drawing people’s attention to the Post Office’s existing success.

In the age of countless stimuli vying for our attention, finding the right one to direct your customer’s focus towards represents one of the most important competitive advantages you can create.

Now, let me guess: you might be ready for a coffee.