Sameer Srivastava 00:01
In many cases, the managers have assumed, you know, we've been able to maintain high levels of productivity, things are still getting done. It's all fine. And I think we don't really understand the ways in which the cultures in many, many organizations these days are fraying.
Amir Goldberg 00:17
So, if you're a leader, if you're a manager, if you have a vision to build a company and you think algorithms are going to solve that problem for you, you're in for a big surprise.
Alison Taylor 00:25
HR can't own culture and compliance can't own culture and senior leadership can't own culture. Each has a slice of culture, as do employees as a whole. So, if we still think in terms of delegating to specific functions to solve these issues, I don't believe they will ever be solved.
Per Hugander 00:43
When you want to influence culture, when you want to make that a stronger part of your culture, what you need, if you want to change the culture is to change people's assumptions.
JB 00:52
This is Bank Notes: Banking Culture Reform. The views expressed in this podcast do not represent those of the New York Fed, or the Federal Reserve System.
Toni Dechario 01:01
Hello, and welcome to Season Two of the Banking Culture Reform podcast part of the New York Fed’s initiative to drive awareness and change in financial services culture. This season, we're exploring themes of trust, technology, and the new workplace. We introduced these topics in our 2022 webinars, which you can watch at NewYorkFed.org. In this season's podcast, we’ve brought back panelists from each webinar to take a more in-depth look at their work around these three themes. My name is
Toni Dechario, and I'm with the New York Fed’s culture team.
Toni Dechario 01:33
Our guest today is Alison Taylor, Executive Director of Ethical Systems, and an adjunct professor at NYU Stern School of Business. Alison's also writing a book, which is focused, in her words, on how business can do the right thing in a turbulent world.
Toni Dechario 01:48
Hi, Alison. Welcome.
Alison Taylor 01:51
Hi, Toni, thank you so much for having me. It's so wonderful to be here with you.
Toni Dechario 01:54
Same here. Let's start out just by outlining what it is that you do. You wear many hats. Maybe you could describe a few of them.
Alison Taylor 02:03
Sure. So, the biggest hat I wear is that I'm Executive Director of Ethical Systems. And that's a research collaboration based at NYU Stern School of Business. It was founded by a prominent social psychologist called Jonathan Haidt, in 2011, and I took over at the end of 2019. And the idea behind it is to bring the best ideas from academia on behavioral science and social psychology into corporate practice. So, one of the things that Jon observed and that I also observe, is that there is a lot of focus in the private sector on consulting solutions, and less focus on the academic research that might actually help us build more ethical and sustainable cultures. So, in terms of how I think about what I do in the world, I think about myself primarily as a translator, as someone that acts as a bridge between real-life practice and real-life experience, and what's going on in universities and the big learnings we can have there. You mentioned also, I'm writing a book. So, I've been interviewing lots of leaders and lots of practitioners for the last year and a half on all these big-picture questions. And then I have the immense privilege of teaching at Stern on a lot of core ethics and sustainability classes, which I think, again, really gives me this hands-on view of what's going on out there and how today's business students are thinking about these questions.
Toni Dechario 03:33
Thank you for that overview. Why did you move into an organization focused on organizational culture? Why do you think organizational culture matters? And what kind of brought you to that realization?
Alison Taylor 03:45
Oh, such a good question. The short answer is I've had a variety of roles, mainly in consultancies, working with, on the one hand, ethics and compliance teams, legal teams, investigating fraud, doing due diligence on anti-corruption, looking at corporate integrity, looking at the background of founders and companies prior to transactions, and that kind of thing. More recently, I worked in sustainability, so materiality assessments, helping companies identify social and environmental priorities, and trying to help drive change to more responsible business. In both sets of experience, it has been very, very difficult to avoid the conclusion that culture is what really matters. So, I first started to notice this, really in the run up to the 2008 financial crisis, where I was doing a lot of anti-corruption and anti-money laundering diligence, very often for compliance teams, and very often prior to a big investment banking transaction. And it was very, very hard to avoid the conclusion that the sales lead, or the banker had the ability, in practice, to override a lot of Compliance and Risk checks, not by saying that they thought those were stupid, but just by undermining them and pushing the way that the system worked to get the result that they needed. Or I would do a fraud investigation and I'd watch the senior leadership team find a convenient scapegoat to blame even though it was quite obvious to me that there was greater knowledge of this fraud in the organization then the organization was prepared to admit to. And so, I got very, very interested as a result of these experiences in questions of culture, because I started to see the same things happen over and over again, and started to ask myself questions about what I was seeing, what I was hearing, what I was experiencing, but I didn't have the concepts to understand. So that took me back to do a second Master's in Organizational Psychology. After that Master's, I moved into the sustainability role. And many of the same questions about power, about culture, about leadership, about incentives cropped up there as well. So, I think in order to understand where we are with responsible business, and where we might end up, you can't get away from culture. And certainly studying culture and social psychology has been more useful to me than anything else I've ever studied. It is real-world tools and real-world insights that I still use every day.
Toni Dechario 06:21
Is there anything in particular that you think is important for financial services firms to understand about organizational culture and kind of how and why it matters to them specifically?
Alison Taylor 06:33
I mean, there's a lot of academic evidence about the kind of people that go into financial services and the specific sorts of risks that manifest in financial services. I take some of the research about the kind of personalities of people that become bankers with a bit of a pinch of salt. But this is an industry with an enormous amount of power over the rest of the economy. It's an industry that is valorized. It's an industry whose high performers are arguably excessively rewarded for their role in society. And so, I think that leads to a lot of ethical pressures. There's also in financial services, particularly since the turn of the millennium, a lot of regulation, so a lot of requirements. So, I think this is both a strength in financial services, you tend to see more rigor and more oversight in banking, for example, than you do in many other industries. But I think it can also lead to a focus on process, at the expense of culture, at the expense of wider organizational dynamics. And so there tends to be an idea that if we put in place the right process, and the right system and the right oversight, and the right carrots and sticks, our wider culture and trust problems will be addressed. And unfortunately, I think, oversight processes while they're entirely necessary, I would never say they're unimportant. They're foundational. I would describe them as necessary, but not sufficient.
Toni Dechario 07:58
Right. You can have the best controls in the world. But at the end of the day, if the culture encourages people to find ways around them, they're going to find ways to do so.
Alison Taylor 08:05
Exactly. I mean, it's been very, very useful to me. And this is part of what I'm trying to write about in this book. It's very useful, I think, to think about a lot of these efforts as a sort of defense mechanisms to deflect either the regulator or to deflect reputational risk. And so there tends to be more of a focus on protecting the value of the organization from these perceived negative threats from the outside, than actually doing what it takes to build an ethical, sustainable culture where you perhaps wouldn't need so many defenses because you'd have more societal trust.
Toni Dechario 08:41
That's a great segue, actually. So, the three topics that we at the New York Fed have been looking at this year, with reference to culture, are trust, technology, and the new workplace. But you moderated our most recent web panel on culture in the new workplace. So, I'd love first to get your perspective on a couple of learnings for you. Is there anything that you talked about in that session or that you thought about in preparing for that session related to how kind of our post-pandemic world is impacting culture and the way that we interact with one another? What are you kind of still thinking about?
Alison Taylor 09:24
I mean, the first thing I want to say is it was such a wonderful panel, and anyone that is listening to this right now and hasn't listened to that panel, I highly recommend that you do so. Three amazing people, amazing academics, amazing experience. And each of them did a lot to push me on my prior assumptions. And I certainly walked away with very different conclusions than I walked in with. So, I think the first thing I would say is I had the assumption that a remote culture might be a culture with weaker signals and a culture where it is less easy to make employees aware of what your priorities are and what your values are. But what the panelists showed me is, actually culture can be stronger, culture can even – or technology rather – can even come further into your home. It can make the organization feel more present. It can create new kinds of dynamics and new kinds of ways of behaving. So, it's not just a case of remote culture is weaker, and in-person culture is stronger, but there are entirely new considerations if you move to a remote or a hybrid way of thinking. The second piece of insight from the panel, I think was about the need to experiment, and this is very much in line with Ethical Systems’ own mission, but the panelists repeatedly emphasized that we're in this unprecedented time where we don't really know what's going to happen next. We're dealing with a lot of things that have never happened before. And we all, I think, collectively focused on the tendency of a lot of CEOs – perhaps particularly in financial services – to say, “This is the way it's going to be. This is the way my culture is. I'm going to make this decision about remote work or incentives or hiring, and it will have this impact.” And I think what we all agreed was, we don't actually know. So, there is a need to do proper behavioral experimental research in real organizations. There is far less of that around than there should be. We do a lot of A-B testing and so on to try and make people buy more, but much less to try and make our cultures, to try and make our ethics, or to try and make our leadership more effective. So really, I think it really made a strong case for experimental research and working with academic partners. And then thirdly, and perhaps most significantly, I have this view of technology and surveillance as very dystopian, and very disturbing. And the panelists convinced me that really, this is a set of tools that could be used to drive far more positive culture. It could be used to empower employees. It could be used to solve a lot of the problems from the past. We're not necessarily doomed to this dystopian future of all being plugged into a matrix and surveilled to death. So, where I really landed, was feeling much, much more optimistic that if we could get these insights and tools into the hands of leaders with the appropriate thoughts, and the appropriate priorities, we would be in a much, much better position to weather some of the storms out there at the moment.
Toni Dechario 12:22
I think maybe first, I'll pick up on your comment that now we can bring the office into the home more. I know that you have some views on that. There's been kind of a trend in the last few years of bringing your whole self to work. But I think that you've identified maybe some pitfalls associated with that, can you talk a little bit about your perspective on kind of the strengths and weaknesses, perhaps, of bringing your whole self to work, and particularly as it relates to ethical and unethical behavior and how it might impact people's motivations?
Alison Taylor 12:57
Sure. So, the first thing I want to say about this topic is that I think the notion of bringing your whole self to work comes from a really, really, really positive place. It's very much tied to diversity and inclusion considerations. It's very much tied to the idea that if you don't feel you can be your whole self in the office, if you are hiding some aspect of yourself of your social identity – let's take an obvious example, your sexuality – then you will not be able to participate fully in the culture. You are also less likely to raise your voice when you have ethical concerns. You are less likely to feel included. So, I think all of this comes from an incredibly positive place of trying to make organizational culture more relevant and more appropriate and more comfortable for every employee and not just a particular subset of employees. And I see this in the classroom every day: to see that young people expect more from their careers than I did at their age, to see that they expect to have a contribution, they expect to work somewhere that has a positive contribution to society. And I think all of that is really awesome, and in strong contrast to how I felt, which was that it was all about getting told what to do until you had enough power to tell other people what to do. So, having said all of that, I see a lot of surveys and a lot of discussion out there about how employees today expect corporations to align with their personal values and expect them to speak up on questions that they feel passionate about. And unfortunately, I think that is going nowhere good, because it is somewhere between difficult and impossible, I would say, for any corporation today to accurately reflect the personal values of all its employees. In fact, it's an – it's an obviously absurd aspiration. Our values are personal, they are highly varied. There is probably less of a global consensus on core values than there has been certainly in my lifetime. And so how we would expect any global corporation to reflect the personal values of all its employees I have, I have no idea. And so I think it puts companies in a bind. It exposes them to risks of hypocrisy and unrealistic expectations. And I think it's also, in some cases – and certainly in the US context – helping to increase polarization and helping to further confuse us on the role of business in society. So, what I would rather see corporations do is have maybe less of an appetite to impose a set of values on all their employees, which is, by definition, going to make some group of employees feel silenced and overridden and resentful and as if there's a dominant culture. I think it's far more important to encourage pluralism, to encourage mutual respect, to encourage conflict resolution, critical thinking. And so rather than maybe taking a stand on the controversial social or political issue of the day, you provide employees time off to vote, you allow them to use their own salaries, to speak up and find whatever causes they believe. And so, I think we've gone too far, and expecting corporations to fulfill a role we used to expect from governments. And I think we are in some cases causing further conflict and fragmentation inside our organizations than we need to. What I would personally do if I was a CEO, today is say, “if you want to thrive and be senior in this organization, I need to see you work successfully with somebody with very, very different values from you and who you may even personally dislike.” I also would kind of close by saying, I would like to shut my front door at night and know that my employer is not going to come into my house and know what I'm going to do. So maybe where we need to go is don't bring your whole self to work and we won't bring HR into your home.
Toni Dechario 16:47
So, picking up on kind of another of the learnings that you talked about, which is the idea of experimentation, that we don't actually know what the right mix is between virtual and in-person work. We don't actually know who should be working with whom and where and why and how that impacts culture. It's different for every organization, I think, is one of the big takeaways I had from that session, that you moderated brilliantly, by the way. But I think that there's a real resistance to experimentation. I think this is particularly true, actually, in the highly regulated industries like banking, but true anywhere where it is a careful trapeze act to be seen as experimenting with your employees, right. But you have a lot of experience with this at Ethical Systems, and what is it that kind of resonates with people and gets them over that hump and willing to try new things and maybe make a mistake?
Alison Taylor 17:45
I mean, I don't think it's super easy to encourage any private sector organization to do this kind of research. When I'm trying to persuade an organization that this is a good idea, I emphasize that they will have the benefit of academic rigor, they will have the benefit of independence, anything they do will be documented. We would try something with a control group. We would be able to measure before and after. So, one big selling point is it will stop you wasting maybe millions of dollars on, for example, DEI training that doesn't work, and have some hope of designing something that might work better and might actually improve inclusion. But you also talk about how difficult it is to experiment in areas like compliance or like ESG, where there's so much reputational and regulatory pressure, and perhaps a nervousness with, you know, we're experimenting with our compliance program that already sounds really risky. There is an idea in sociology called isomorphism. And I think you see this very, very strongly in compliance and in sustainability is that firms are essentially copying each other and herding together to try and prevent regulatory scrutiny. And I think it's really stopping us progressing. So, there are ways to design experimental research that are ethical. They're not creepy. They don't involve violating employees’ privacy, I certainly think it would be important to have rigorous academic partners to do that. But I think that we've reached a limit in terms of how we can progress on any of these issues if we just keep doing the same thing over and over again and expecting a different result.
Toni Dechario 19:21
Isomorphism, I like that concept. I have to admit to not knowing it before you mentioned it. It makes me think about kind of the challenge that many industries face, I'm sure this isn't limited to financial services, but I know that in financial services, there is a kind of a collective action problem when it comes to culture and behavior, because there's so much movement among and between organizations, that you kind of have an industry culture as much as you have an individual firm culture. So how can firms assert a different approach to culture kind of within that collective-action problem, if you will, and make it like a strategic strength?
Alison Taylor 20:06
I think it's true in all industries to some degree, but I certainly had a client from banking a few years ago, who said, “We like to walk across the street together holding hands.” And I think because of the regulatory pressure, there's certainly a very, very big driver to kind of herd together and not put your head above the parapet. And that's incredibly understandable. One solution would be collective action. You see that also on – things like climate with GFANZ (Glasgow Financial Alliance for Net Zero), so getting companies to sign up to the same approach to do the same kind of culture assessment to share findings. Obviously, there are antitrust considerations, but they are not insurmountable, and so on. So, I think that's really, really positive. I mean, something I certainly see is that there is more emphasis on culture and conduct risk thanks partly to the Fed, but more emphasis on these questions than there is in many, many other industries. And so I think you are starting from a really, really good base, I would say it's financial services and pharmaceuticals that have gone the furthest here. That have gone furthest in hiring behavioral scientists, that have gone furthest in bringing in these insights. So certainly not starting from scratch, and certainly in a much, much stronger position than other industries. But I think differentiation and trying to do anything truly innovative, it's going to continue to be a struggle.
Toni Dechario 21:21
You talked about folks in banking, having brought in behavioral scientists to help them with this work. You sitting in your seat at Ethical Systems brings a super wide range of perspectives to bear on this issue. You have members that are sociologists, organizational psychologists, economists, accountants, who are kind of all bringing a different perspective and indeed, bring a necessary perspective, you kind of need to be able to see it from many different angles. There are many ways to skin a cat, if you will. Obviously, a single organization can bring in folks with all of those different perspectives to kind of help them solve their unique problems. So, what do you see as effective ways for an organization trying to attack culture as something that can be formed and influenced how they bring all this to bear when facing this huge amount of literature?
Alison Taylor 22:20
So, I think this is – you've kind of really with your question expressed some of the problems with culture, which is that it's perceived to be almost impossible to measure. And it's perceived to be not just complex, but sort of fluffy and soft, compared to other things we might do around structure or process or things that are much easier to understand, easier for the average business to implement. So, I think that's all a shame. Because if you get culture, right, all these other things would likely fix themselves. So, it's not easy to answer your question with a kind of pat, couple of pieces of recommendations. But I think a very good place to start is just by trying to measure your culture. And not measure it in the sense of, I want to have a score, and I want to understand how I compare to my peer banks, and where I'm higher or lower, and how I can improve my score. But understand kind of some key points about how the culture works, how employees feel about their roles, how employees feel about their bosses, how comfortable employees would feel speaking up, and so on. And so I would, Toni, just point any listener to, there's a page on the Ethical Systems website, where we have summarized what we know of as being the best available culture, ethical culture assessments out there. We've done a link to all of them, we've done a review of them. And so that would be a really, really good place to start. If you could get a baseline measure of where you are, then you will be able to understand and track change and progress over time. And from there, maybe start to think about what themes are occurring, or different themes are rising in different regions and different groups and different product lines in different teams? And then how might we go about improving some of the issues that have been identified? So, the short answer is we're not going to be able to change or improve anything if we don't know what the current situation is and can't measure it. So, the first step is always to measure culture and have some curiosity about what's going on in your organization. And then it's not as mysterious and not as fluffy as people think.
Toni Dechario 24:19
So going back to new ways of working, and how culture or the transmission of culture has maybe changed in the last couple of years. Can you give us some perspective on whether there are different expectations now, around acceptable behaviors? And I asked this specifically in the context of, you know, news articles that we've seen about firms that have ended up paying pretty hefty fines for the use of unauthorized channels to conduct business. Is there something about our new approach to work that's perhaps normalized some behaviors that weren't considered okay in the past?
Alison Taylor 25:03
I think certainly the shift to remote work has created some dilemmas. It's created some additional opportunities for unethical behavior. It has also created additional opportunities for intrusive surveillance. So, I think setting and conveying culture and being much more explicit about your expectations, and really focusing on what it takes to be a good team manager are real imperatives in a way that I guess they've always been, but there are things you can do in a physical environment as a manager, there are things you can do in terms of oversight that are more difficult in a remote or hybrid environment. The other thing that's very, very much on my mind is changing expectations about work from younger generations. So, millennials, first of all, and now Gen Z, and we talked a little bit about this on the webinar, but we have at this point, anybody under 26, does not have that much experience, if any, of working, you know, in a normal – what we used to consider a normal – pre-COVID office environment. And I think that combined with these possibly unrealistic expectations about what work and leadership and jobs are for, and the increased ability and willingness of young people to raise concerns, and possibly unrealistic demands is all coming to maybe mean that management has never been so difficult. And I'm hearing this over and over again, in daily conversations that today, your 25-year-old intern, or your new hire, will be quite comfortable challenging the CEO in a town hall meeting about, you know, the mix, the diversity mix on the senior leadership team. I've heard that across countries. And I think it's – it’s pretty jaw-dropping for anyone my age, to think of challenging power in that way. So, on the one hand, there's this huge opportunity with this rise of voice to kind of treat this as an advantage to identify problems you're going to necessarily miss, just because you have a lot of power at the top. On the other hand, I'm also frequently hearing that there's a lot of entitlement, there's a lot of unwillingness to adapt to the culture, there's a lot of for example, feeling like, I would like to be able to have the schedule and work from where I like, but I would like to also have all the benefits of the in-person culture and the kind of mentorship and encouragement and so not maybe a very realistic view of the trade-off. So, I conclude really, by feeling quite grateful that I'm not managing a big team these days. But I think we need to do a lot more to support managers and middle managers in particular. I also think we've emphasized over the last few years, sort of strategic inspirational leadership at the expense of just good, hands-on competent management. And I really think we need to go back there and provide today's managers with better tools, and better approaches to navigate some of these challenges.
Toni Dechario 27:54
That's really interesting. I have been thinking myself lately about how our experiences, particularly in the last couple of years, even non-Gen Z folks like us, are so bespoke. And you know that the – there's been research done by Microsoft, which is cited by lots of people, showing that our world has become smaller, and our ties with the people that we’re close to, that we talked to a lot, have grown stronger, but our kind of what they call “weak ties,” you know – we have fewer of them. And we've become much more kind of siloed. And so I'm curious about how that's impacting organizations’ ability to kind of influence via large social groups, because we are so bespoke. And you talked about how kind of the next generation comes in and they want their specific values to be reflected. Well, they all have different values. And so I'm curious about whether the way that we influence behaviors, the way that we influence kind of acceptable and unacceptable behaviors has maybe shifted.
Alison Taylor 28:58
I think this is sort of like the ultimate philosophical question. But I think I'm back to this sort of notion of how difficult it is now to set values from the top down. My impression is this more CEO town halls and kind of company-wide meetings, and then everybody goes back into their tiny teams. And what we're not doing as well as we used to maybe – not that I don't think we've ever done it particularly well – is work across silos and take a kind of whole organization perspective. That's unfortunate because most of the most pressing and difficult problems facing organizations today aren't anyone's job to deal with. So, whether it's deciding what social or political issues to speak up on, whether it's thinking about ethics, beyond compliance, whether it's thinking about how you can contribute more positively to society and set environmental and social goals, whether it is about detecting wrongdoing – all of these things – require cross-functional collaboration and someone thinking about risks in a much, much more holistic way. I think we'd need more of a focus on holistic integrity. I've worked on this for a long time, but closer connections between marketing, public affairs, HR, compliance, ethics, ESG, more strategic thinking about integrity and ethics. And now I think a lot of private sector organizations would really benefit from academic help and support and measurement and understanding what is going on out there and also understanding these aren't issues confined to one organization, pretty much everybody is struggling with them. And we all need to collectively come up with better approaches, which I don't think will be achieved if we just use the old tools and tired old assumptions we've been using from before the pandemic.
Toni Dechario 30:45
What is holistic integrity?
Alison Taylor 30:46
I think it is about thinking about the connections between the voluntary and the regulated and thinking coherently about, for example, it is not sufficient to have even a leading sustainability program with a great materiality assessment and all the disclosures out there, if you haven't, for example, given thought to your political contributions, and given thought to what you're going to speak on publicly. And so the default, I think, is to give, you know, compliance one set of issues, sustainability another set of issues, risk another set of issues, senior leadership another set of issues, strategy another set of issues, HR another set of issues, but we can only make sense of some of these questions if we can sync across those functions. And so, a very obvious thing we've been talking about so far on this podcast is organizational culture. Well, HR can't own culture, and compliance can't own culture, and senior leadership can't own culture. Each has a slice of culture, as do employees as a whole. So, if we still think in terms of delegating to specific functions to solve these issues, I don't believe they will ever be solved. That's what I mean by holistic integrity is thinking more about working across the organization and making some of these questions the responsibility of the senior leadership team, so we can put in place the right governance mechanisms to at least think through and make our best effort on answering these questions.
Toni Dechario 32:15
And that way, you also make it kind of more of an institutional thing, like it's more kind of embedded throughout the institution, which is relevant, because I'm now thinking about the Great Resignation, which maybe is a myth. People are saying now, I don't know – I'm sure you, you know, better than I. Is it real? And if so, how is it impacting firms’ ability to influence behavior as you have as much churn and turnover as people have these days?
Alison Taylor 32:44
There's been a ton of churn and turnover, and in some cases, also exponential growth. So, I also know a lot of firms that have grown substantially during the pandemic. People don't know each other. People aren't used to the culture and so on. On the Great Resignation, I mean, there's a lot of data from surveys and so on out there. There's all sorts of evidence that things like long COVID are really affecting the workforce. So, I think it's a pretty murky picture. But I think perhaps this is another example of a failure or a weakness and thinking holistically about culture and holistically about employee motivation. What I tend to see is a lot of CEO rhetoric around purpose and mission, and I have the impression that all these CEOs have read a lot of surveys showing that more dedicated passionate employees are more productive. But you've seen all this rhetoric that's very, very grandiose, in a lot of cases, advancing along with employees feeling apparently less motivated, apparently less committed, apparently more exploited, apparently on email at 11 p.m. at night. And so, we are possibly trying to use a lot of rhetoric about purpose and mission that we feel will attract investors or deflect scrutiny in some way, without first concentrating on the basics of dignity and respect and employee satisfaction. And so, I think we need to focus on these more basic issues. We need to walk before we run, maybe it's one way to put it. I think it is less important that you have some kind of noble-sounding societal goal and more that you have superordinate goals, and that employees feel some level of pride in what they're doing every day. And so, I think we've a little bit lost sight of some of those basics. The good news being if any company can get all of that right, I think it would provide a significant source of strategic advantage today.
Toni Dechario 34:40
What are your thoughts on kind of whether there have been any opportunities raised either through the introduction and acceptance of more and more technology between us, or the kind of global experience of all of us going home and now coming back to work? Are there any opportunities that kind of came up that you think could still be harnessed? And I guess I'll add to that question, just an observation that I think there are many people generally, and especially in financial services, who want to go back to kind of the way things were, you know, who are going back to be in the office full time. And my observation is, even if we're all in the office, the way that we interact with one another has really changed and that's probably not going anywhere. Even if we're all in the office, we're still using technology much more than we were before. We're much more likely to use Slack or Teams to communicate with one another than we are to pick up the phone, which kind of pre-pandemic, we were probably more likely to pick up the phone – or at least those of us in older generations who hadn't embraced technology yet. And so, where do you see kind of the opportunities that exist for everyone, even those that are kind of more traditional organizations.
Alison Taylor 36:01
I mean, I think the big opportunity is to redesign work so that we are making the best use of the time that we need to spend together, whether it's to generate ideas or get to know each other, or, you know, work on something strategic together, versus what we are best placed to having the freedom to do at home on our own. And so there's a huge opportunity. But to take advantage of that opportunity means really looking at every single person's job role, understanding what they do all day, understanding what part of that can only be achieved through in-person collaboration, what part of it is best to leave the employee to work on their own behalf. And so it's not a case that you – just my impression, is a lot of organizations have defaulted to two, three days a week in person whenever you feel like it. And then people end up sitting in the office with lots of Zoom feedback on other people's computers. You could design a better workplace where you have physical spaces where you are bringing employees together for the activities that are really, really important for them to do together. And otherwise, allowing them to live in other places, visit other places, work more freely, take care of family obligations, is obviously a huge opportunity to advance on DEI goals. It could give people with family obligations, more flexibility, or ability to achieve. There's some evidence that more kind of toxic diversity dynamics, or leadership dynamics show up less in remote environments. So, we get more of a focus on competence in which employees can actually deliver rather than which just sort of talk a good game in meetings. So, the tools are all there; the thinking is all there. In fact, from technology, the insights are potentially all there. But those insights need to be put in someone's hands that know what to do with them and know how to use them for the collective good. So, it's an all opportunity, I think, for an organization that does that, if you don't think deliberately about this, if you just kind of default, and align with what everybody else is doing and, again, kind of try and apply ideas from before, then I don't think you will take advantage of these great developments. And I think probably the net result will be worse than it was before the pandemic. So, there's bigger upside and bigger downside, I would say.
Toni Dechario 38:23
Okay, so you are placed at Ethical Systems in a seat, where you see so much work being done on questions related to ethics, related to organizational behavior, to incentives. So, I want to kind of dig into some resources with you because you are an excellent source of information about resources. So, if I, if I were a bank executive, who has just entered a new organization, and wanted to get my head around what is organizational culture, I'd love to hear your take on the best resources to bone up on in order to approach that. And then moving into – once I've been able to fully understand the culture that I'm dealing with, or the set of cultures that I'm dealing with, what are the best resources for me to consult in thinking about how to shape and influence organizational culture? Particularly when it comes to ethical behavior and ethical decision-making.
Alison Taylor 39:19
I mean, there's a lot of resources out there, there is a really good reading list that we created maybe a year or so ago with really good behavioral science resources aimed at ethics leaders that I will share, and we can have in the in the show notes. That's books, it’s articles, it’s podcasts, it's all stuff that I think is clear and compelling, and you don't need all the jargon to understand. So, one of the really great things about this field is there's so many amazing vivid writers, it's not like trying to get to grips with regulatory compliance or accounting, or something like that. These are, I think, concepts and ways of thinking that are familiar to everybody. They're often in popular discourse. These ideas shouldn't be viewed as impenetrable as they tend to be. I'd certainly be paying attention to the New York Fed’s culture and behavior sessions. I think I would be looking to get some advice and some strategic partnerships to try and kind of put some of these ideas into practice. So, you know, I think my key takeaway is really, that there's a lot out there that is, is accessible and is helpful, and there's more and more being developed every day and so no need to feel intimidated. And no need to feel that this is outside your normal skill set. I had a professor at Columbia that said, “You don't change culture by trying to change culture.” So, very often what you need to do to change your culture is not kind of more happy hours and free ping pong and yoga apps, you need to look at things like incentives and oversight and power and rewards. And then the culture will change. Microsoft is a very, very good example recently of a new leader who really came in and very deliberately changed the culture, and specifically moved the company from a focus on individual results to team performance and behavior. So, it can be done. It can be done quite quickly. It just requires a bit of thought and a bit of planning. And again, not to sound like a broken record, but experimentation and measurement are not just throwing spaghetti against the wall and seeing what sticks.
Toni Dechario 41:33
I think that sums it up perfectly. I will end it there. Thank you so much, Alison, for being here today.
Alison Taylor 41:38
Thank you so much for having me.
Toni Dechario 41:40
If you want to hear more from Alison, watch the New York Fed’s webinar on culture in the new workplace, which you'll find at NewYorkFed.org/governance-and-culture-reform.
Sameer Srivastava 00:01
In many cases, the managers have assumed, you know, we've been able to maintain high levels of productivity, things are still getting done. It's all fine. And I think we don't really understand the ways in which the cultures in many, many organizations these days are fraying.
Amir Goldberg 00:17
So, if you're a leader, if you're a manager, if you have a vision to build a company and you think algorithms are going to solve that problem for you, you're in for a big surprise.
Alison Taylor 00:25
HR can't own culture and compliance can't own culture and senior leadership can't own culture. Each has a slice of culture, as do employees as a whole. So, if we still think in terms of delegating to specific functions to solve these issues, I don't believe they will ever be solved.
Per Hugander 00:43
When you want to influence culture, when you want to make that a stronger part of your culture, what you need, if you want to change the culture is to change people's assumptions.
JB 00:52
This is Bank Notes: Banking Culture Reform. The views expressed in this podcast do not represent those of the New York Fed, or the Federal Reserve System.
Toni Dechario 01:01
Hello, and welcome to Season Two of the Banking Culture Reform podcast part of the New York Fed’s initiative to drive awareness and change in financial services culture. This season, we're exploring themes of trust, technology, and the new workplace. We introduced these topics in our 2022 webinars, which you can watch at NewYorkFed.org. In this season's podcast, we’ve brought back panelists from each webinar to take a more in-depth look at their work around these three themes. My name is
Toni Dechario, and I'm with the New York Fed’s culture team.
Toni Dechario 01:32
Our guest today is Per Hugander. Per is an engineer-turned-strategy consultant who writes extensively about psychological safety and trust. He's a strategic advisor at the Nordic Bank, SEB, and he teaches at the Hult/Ashridge International Business School. Welcome, Per.
Per Hugander 01:48
Thank you, Toni.
Toni Dechario 01:49
So, Per, maybe to get started, you could describe to me what it is that you do.
Per Hugander 01:54
Absolutely, I will try to do that. So, what I do, I help teams to make progress on complex challenges. And I help them to make better decisions. And then I try to help them to infuse the capacity to make those decisions into their organizational culture.
Toni Dechario 02:13
So, you started off as an engineer, what you just described doesn't sound like what an engineer does. Maybe you could describe to us how you ended up in this role, and how you ended up deciding that this was a really important thing for organizations to be able to do.
Per Hugander 02:28
Absolutely. So, before I became an engineer, I was a basketball player. And I was always really, really struck by how the best coaches that I had, could do these small things that would make our team a little bit better, and it could be stuff in the preseason; it could be a smart play that they drew up at the end of a game; could be, you know, how they mentally prepared us before the game. And it's those small things that would give us an edge on the competition. And I think when I stopped playing basketball, and I decided to go into business, I brought that with me, and I wanted to help organizations with that same thing, to get that edge. And to start with I was a strategy consultant. I thought strategy was the thing, you know, the best way to help organizations to succeed with what they wanted to do. But after seeing a few organizations trying to go through or make it – old organizations – trying to make it in a transforming industry, I realized that probably wasn't strategy that I was supposed to help them with. I was probably doing the wrong thing. It's very much like, I don't know if you read the book “Strategy and the Fat Smoker.” It's a Harvard professor named David Maister who wrote it. And he basically says that it's not, you know, coming up with a great strategy or the strategic plan that's going to make you successful. It's actually how you make that happen. And it was pretty clear to me that it wasn't, you know, lack of good strategies, or lack of resources that was getting in the way for these giants to succeed. It was actually that the way they led the organization, and the culture was really getting in the way of everything that they were trying to do. So, after having that insight, and after seeing more and more organizations trying to make it through transformation, I realized I should probably head back into leadership that I'd been looking at, but when I was a basketball player, right, and culture, so that's how I started venturing this way. That's a long answer to a short question.
Toni Dechario 04:42
I love that. The book's title is so illustrative. It really helps.
Per Hugander 04:46
I love that book and there’s a short article on it as well, that I greatly recommend everyone to read.
Toni Dechario 04:52
So, you can have the best ideas and strategies, but really what matters is how you put it into practice, and whether you can really implement them. So alright, you figured out that this is what matters. How did you figure out what to focus on?
Per Hugander 05:03
I realized that we're asking a lot of our leaders. It's a pretty daunting task to be a leader these days. You’ve – first, you have your team that you're supposed to lead, then you got to do the one-on-ones and manage your employees’ development and then there's the administrative burden of being a manager. And then you also have to stay on top of your area of expertise. And that alone, in a changing industry or a transforming industry is a quite big ask. And then on top of that, there's all these things that organizations are struggling with, or were struggling with, when I started getting into it. You need to be an innovative leader, and you need to be a collaborative leader, and you need to be a transformative leader and an inclusive leader. There’re all these things that we need to succeed with. And it was just hard for the leaders that I met to use the stuff that was available. There wasn't a shortage of theories or models. But all the great stuff that was available, wasn't really being used, because it was, it wasn't delivered in a way that, you know, the leadership industry wasn't doing a good job of getting what we know about leading in complexity, to be used by leaders. So, many leaders were just overwhelmed by all these books and theories, and they ended up not implementing very many of them. So, what I tried to do was to go through many of these fantastic theories and models, and see are there commonalities? What is it in these theories, that really makes a difference? Because they were all good. It's just that there were so many of them that they didn't really help. So, I was trying to find those common threads and the skills that you needed to use them. And I started to see a pattern of a few skills, important skills that kept coming back, and that kept being important. It didn't matter if it was to drive innovation, or if it was inclusion, or if it was, you know, creating a better team, there's a few things that come back. And some of those things, two of them that I thought we'd talk about later is the ability to take another person's perspective, to really go into perspective-taking. So, that was something that came through, and also the ability to create a psychologically safe environment. So, an environment where people dare to take interpersonal risks, you know, ask risky stuff, challenge each other, throw out new ideas that might sound a little bit crazy at first or admit when things aren't going well. Or if there's stuff you don't know, those things mattered. So, I decided to focus on those skills and try to find ways to help leaders to improve them and put them to use. And that was quite helpful. So, I kept doing that. And I found a few things to focus on, I thought it was a good idea to focus on the skills, not so much on the tools and on the theories and methods, but to focus on a few skills and help people to get good at those. So, that's how I found the focus. Then later on, I ventured into trying to infuse those skills into organizational culture.
Toni Dechario 08:23
So, the skills were the ability to take perspective, and the ability to create and establish a psychologically safe environment, and kind of build psychological safety. There's actually a Harvard Business School case study written about how you went about doing that. The case study for those who are listening was written by Amy Edmondson and Elena Corsi. And it kind of describes how your approach brought really tangible and positive business results. We'll link to it in the show notes. And I recommend people review it. It's a fascinating case. It indicates that, you know, the firm has been able to solve strategic problems that it had been working on for years, that it's helped the firm make decisions more quickly, more efficiently, and it even links your approach to improved market share. Can you describe kind of broadly, what that approach looks like? How you actually kind of created a skills-building approach within the organization?
Per Hugander 09:23
So, to change culture, you need to change people's assumptions. So, people need to assume that psychologically, or psychological safety will create good outcomes, and not theoretically buy into that. You know, Amy Edmondson, who's a fantastic professor from Harvard, who's done all kinds of research on psychological safety, proved that it really makes a difference for organizations in complexity. But that theoretical insight doesn't really do it when you want to create culture change. Or when you want to, I don't usually talk about culture change, but when you want to influence culture, when you want to make that a stronger part of your culture. What you need, if you want to change the culture is to change people's assumptions. So, have people assume that this makes a difference for us, here, on our topics, in our market, with our people, if we create the psychologically safe atmosphere, and if we use perspective-taking to explore these tough challenges that we're facing, we will create positive outcomes, successful outcomes. If I can help people to make that connection, then I can start changing the culture or influence the culture. That's my approach, to really go deep and try to make people experience having success with these skills. So, instead of the normal way, or the more practiced way of leading culture change, which would be lots of workshops with, you know, hundreds, maybe thousands of people, if you have a large organization, I would try to find one team that were willing to invest the time needed to learn these skills, and then help them to put the skills to use on the challenges or the opportunities that mattered the most to them. That's what I did. So, in the case study, you'll, you'll read about one team, it was the chief risk officer’s team, so, the senior risk team of the organization. And we practiced this, we had several full day workshops and off-sites and weekly meetings to really get these skills and then put them to use on the really important topics. And we created some really cool results that you can read about – you mentioned some of them – read about them in the case. When we did it with other teams and had similar or better results. And when you have those stories, then you can start doing lots of big workshops and rollouts. But my approach is really about going deep. So, you give people proof that this works in our context.
Toni Dechario 12:04
So, when you say, ‘this works,’ what is it exactly that you're doing with these teams?
Per Hugander 12:08
So first, you practice the skills. So, you get to practice perspective-taking first and in like a facilitated environment. So, we do it together, and we use it on a few topics that the team shares, but then I have them going out and testing these skills, because it's not hard skills, it’s stuff that that we can all do. It's just that we don't do it when it matters the most. We usually don't go into perspective-taking when there's conflict, or you know, when there's tension. But if we do, you usually create much better, you know, both insights and outcomes. So, I would have them go out and try and test this in their home environment, first on low impact stuff, but then gradually using them on more and more important business challenges. And then they'd come back every other week. And we'd share stories and talk about what worked and we’d visualize going forward, what would it look like if I use this skill, the coming week on a specific challenge or conversation. And gradually, they got more and more experiences of perspective-taking leading to more insights, more ideas, better decisions. So, that's how I got them to link this perspective-taking to successful outcomes, you know that it works. Then that it works is different for everyone depends on what you work with. But you know, it works to get progress on stuff, where we might be stuck.
Toni Dechario 13:36
So, by taking someone else's perspective, by kind of putting yourself in their shoes and thinking, “how would I approach this decision that they're facing, or this kind of conflict that they're needing to deal with?” helps you to identify solutions, kind of build on that particular problem, as well as on other problems? Is that what you're saying?
Per Hugander 13:58
Not really, I take it one step further. So, you said you know, “I put myself in their shoes and think about how I would approach that.” The key to perspective-taking is not to put yourself in their shoes – it’s to put yourself in their head. “What is it like to be Toni having this conversation?” With Toni's past and motivation and skills and whatever you bring with you and me trying to understand what is it like to be this other person? So, it's goes beyond putting yourself in their shoes, and really trying to understand what is it like for this person to be in this conversation or to deal with this challenge or struggle?
Toni Dechario 14:37
How do people do that?
Per Hugander 14:39
It’s not very hard. I mean, right now, you could probably try to, “I wonder what it's like to be that weird Swedish guy who's on my podcast,” and think about what would that be, from what you know about me, and just try to do that. And then I would ask you to, you know, go do that with your daughter about something that you might not get along around, or you know, where you might have some difficulties. And once you've done that, “Okay, you know, I get it. I know her pretty well.” So, it's not so hard to try to see it from her perspective. And then I’d ask you to do it with one of your colleagues and it's, it's just something that you can get into a habit of thinking about what's it like to be this other person talking about this challenge. And then, you know, gradually, I would help them do it with situations where there might be conflict, you know, someone that you really have different opinions or different motives or agendas. And you've been struggling and maybe fighting for a long time. To just switch out of, “How do I see it?” and just switch into, “How does she see it, or he see it, or they see it?” And it makes a big difference. Not just because you get more information, because in that process, you also want to ask lots of questions and really interview them about their perspective. So, it's not just imagining what it's like, but in the conversation, you let them know that “I really want to understand your perspective on this thing, you know, so we can move forward. So, that's why I'm asking all these questions.” So, you ask questions, you imagine it. And what happens then as you get more information, you learn more. But also, what's really interesting that I learned through a collaboration that we have with the Wharton Neuroscience Initiative, is that when you switch into perspective-taking, and you switch away from arguing or trying to solve a problem, you actually start using other parts of your brain. So, what they call the default network, which is another part of the brain compared to when you're trying to solve problems. And this part is more connected to creativity. So, that – so it’s – you're more likely to have new ideas about stuff that might work. You get more information from asking all the questions. And, and however you tried to get into that perspective, but also just from the process of switching out of your perspective and switching into someone else's to go into that exploration, you switch on another part of the brain, so you get access to more parts of your brain? I'm not sure. I don't think they let me say that you get smarter, you get more intelligent, but I think you access more parts of your brain. So, maybe that's why we come up with new ideas.
Toni Dechario 17:32
Fascinating. So, this is unusual. This is an unusual exercise to do in a bank. It's kind of different, I think, from how many organizations – particularly more traditional organizations – operate. So, can you talk a little bit more about how you got this organization to start practicing this and how you kind of helped people to overcome any skepticism about this approach, and kind of start building this as a habit?
Per Hugander 18:03
So, that's the tricky part, you need to find someone who really believes in this and believes that it's worthwhile spending, you know, a day or two and then doing these weekly follow-ups. That's a big ask, that's a lot of time. And for lots of people – it doesn't matter if it's in a traditional organization, or the bank, or if it was when I was helping scale-ups and startups – is taking that time. People all want the results, right? But taking that time, it's really hard. So, you need to find someone who's willing to do that, and who's willing to protect that time when his or her team starts, you know, complaining about blocking so much time in the calendar. So, I was lucky, I – before joining the bank, it wasn't that hard, because then I was a consultant. So then, you know, people who contacted me, they already had bought in. They heard about me, they've taken my programs or, or knew someone who's been a part of my program. So, they were already into it. So it wasn't that hard. But being internal, trying to find someone who want to go beyond, you know, doing a half-day workshop or a one-day workshop, and really get into the process of becoming good at this and then putting it to use on their strategic challenges, that was hard. So, I almost gave up before I met this first person, that's in the case study as well, who was willing to do it. And then you know, we get the results, and then another one tagged along. And then we talked about this at a big conference, and all of a sudden you could get more and more people because they wanted the results. But it's really hard, you know, those big asks, of days and hours, it’s not easy, especially when you're asking senior managers to take that time. So that's the hard part. You’ve got to find that first person who's willing to invest.
Toni Dechario 19:54
And in some of your writing, you write about how the proof is in the pudding, right? And when people start to see that practicing this approach is actually helping them make better decisions. It's helping them ultimately become more efficient, and so they're using less of their time not more of their time once they kind of get into the habit. There's kind of a snowball effect, right? Can you talk a little bit about assumptions, organizational assumptions, and why you want to kind of go for those in order to influence culture and influence the way that people do things?
Per Hugander 20:28
Yes, I think everyone who wants to work with influencing culture needs to take a look at Edgar Schein’s organizational culture model. It's a classic and it really stands the test of time. And the foundation of culture according to Schein is the shared underlying assumptions that exist in an organization. So, it's the assumptions about what creates successful outcomes in our specific context. And those assumptions once they are shared, and they become almost automatic, so people act and behave in accordance with those assumptions. And that's how culture forms. So, that's why I focus so much on assumptions. Because if you can influence people's assumptions, you can start to create new assumptions. And after a while, maybe they will get both shared, and hopefully underlying, you know, become underlying assumptions that people just use and act by, without even thinking about it. You know, that takes a long time. But to get started, if you want to influence culture, that's where you want to start, a lot of people start with the espoused values, which is, you know, the middle layer in Schein’s model. It’s this iceberg, right? So, in the bottom is the shared assumptions, and then it's the espoused values, and then on top of the of you know, that what you can see above the surface of the water is the behavior. So it – Schein says is that it's these shared assumptions that it makes the difference. Espoused values is not a very effective way of working if you want to influence culture. And if you want to change culture – I don't really like the word “change” culture. I think you influence culture, not change it. But if that's your aim, that's where you want to work: espoused values. The values that are enacted in the organization will only change if assumptions change. So, it's, I find it a lot more fruitful to focus on the assumptions. If you want to infuse something new into the culture that has to do with, you know, leadership or culture work is about culture. So, influencing leadership or culture, you need to break that assumption. So, the way I try to do with that is try to give people proof quite quickly, of you know, that this will help on your most important challenges. Like you said before, once they understand that they're more willing to invest more, but if you don't give them that proof early on, it's a good chance that they will start, you know, to try to negotiate and use less time. So, what you want to do is give them proof quite early. So, I usually have like a two-day off-site to kick these initiatives off, like the one in the case study. And at the end of the first day, the full first day we just try to build a more and more psychologically safe environment, and we train perspective-taking. At the end of the day, I ask them to identify one strategic important challenge that they all want to make progress on. And then we use two hours to do a structured dialogue around that challenge. This is something that's important to them. And during those two hours, you know, 95% of the time, we'll make some really good, substantial progress. And so they go to bed, realizing that ‘wow, this really makes a difference.’ And then the next day, we have a morning reflection, and they talk about how they're surprised about how this really helped. And with that you're beginning to both influence the individual assumption, but also from having them share those stories, that makes it easier, you know, to kind of start changing your assumptions about these, you know, soft-skill workshops or culture initiatives, then also about psychological safety perspective-taking. So, it's all about giving them proof that this works early on. There's a great theory. – it’s by Nelson Repenning at MIT – and it's called the “capability trap,” and it's all about, if the investment is up front and the payoff is in the future, it's really hard to get people to invest.
Toni Dechario 24:53
Well and also, we're social creatures. And so, coming to this realization jointly, I suspect, is effective in that where we are reinforcing one another's responses and reactions and, and kind of building into the into shared assumptions, not, just personal assumptions, the fact that this was effective and in helping us solve this strategic problem that we all care about.
Per Hugander 25:16
Yeah, absolutely. And that's also why we do these weekly check-ins, about the perspective-taking, because that's when the only thing we do with the – it's, it's an hour, and we talk about stuff that worked the two weeks before, and then we talk about where we might want to use this, the coming two weeks. And all of that is stories about how this works, how it could work, how they believe you know, where they believe it will work. So, it's all about creating shared beliefs and proof that it does work. So, putting it into social context, sharing stories, sharing experiences, that's a big part of influencing and helping people create new assumptions.
Toni Dechario 26:01
So, one of the things you said I want to pick up on, which is before you start teaching perspective-taking and before you have people start practicing perspective-taking, you create a psychologically safe environment. That sounds wonderful. How do you do that?
Per Hugander 26:15
What I do to establish a psychologically safe environment early on is that I have people practicing being vulnerable. And this is nothing new, people have done this, being vulnerable about, you know, sharing life stories, sharing things, it's nothing new, we've done that where we’ve practiced trust for decades. But what I do after that is that I have them be vulnerable about work related stuff, you know, have them share and talk about situations, challenges, topics that they're facing, where they might not know the way forward or they might have, have, you know, made a mistake, or they really want to maybe challenge the way that things are done, but they haven't done it. And have them push those boundaries, given all of those kinds of topics, there's a risk, right, there's this interpersonal risk, which is psychological safety. Psychological safety is a willingness to take interpersonal risk. So, you want to help them to get over that uncomfortable feeling that that risk creates. And all of a sudden, they realize that that uncomfortable feeling, that might not be something I should listen to that much. Because when I do lift these things in, I can get help from my colleagues. And we actually make progress. So gradually, by sharing topics and themes that they're unsure of, we build an atmosphere where that's okay. That's the game, it's not something that I do before we do the perspective-taking. It’s something that I do – these go hand-in-hand. So, I do a little bit of the vulnerability practice, I do a little bit of perspective-taking. And actually, perspective-taking is a really good way of building psychological safety, especially in a team where there might be some tension. Or you know, when there's two or three people that don't get along, when you have them practice perspective-taking on each other, and really getting into asking questions to the point where they understand each other. That's a pretty powerful session right there. And it's, you know, you can do that in twenty or thirty minutes. And that actually builds psychological safety, when you realize that that person that I've been struggling with for a long time really understands me now, you know, even if we always had – she actually seemed to care about, of what I thought about the topic or how I saw it. And from that, you start building that psychological sense. So, they work in tandem, these two things, psychological safety, makes it easier to do perspective-taking, but perspective-taking also increases psychological safety. So that's a nice thing about working with these two, at the same time, and I would never try to work with one of them and not the other. If I'd work with a team, they go hand in hand, pretty much.
Toni Dechario 29:05
So, you've been at this for a number of years, now. You've kind of broken through the obstacle maybe that you first faced, which was finding a group to start the ball rolling, right? To be willing to give you the time and the trust to work on perspective-taking and psychological safety. And to get them to a point where they started to see results. I'm hoping that maybe you can provide some stories about what your results have been, you know, where you've kind of seen this tangibly provide outcomes for the organization. And any lessons learned from the years you've spent doing this.
Per Hugander 29:47
So, this stretches, you know, long before I joined this bank, I've done this with lots of organizations and lots of teams, and it's always a bit different, you know, but it always relates to what's the most important thing for these people right now, because that's what I tried to focus in on in order to create that proof that this works here. So, you know, sometimes they're going through a merger, and you know, there's always tensions and misunderstandings in mergers. So, in those situations, a lot of the times the stories will be about how they could find a way through. Sometimes it's roadblocks, you know, you have this strategy, you know where you want to go, but there's just stuff internally that we can't get through, can't get past a lot of the time, it's a conflict. And a lot of time, it might be between one or two, either individuals or groups and organizations that for some reason, might have different interests. So those are things that we have lots of proof and lots of organizations that you can over them, so you can get over them quite quickly, stuff that people have been struggling with, and not really talking about in a productive way for years, you can get past it in two, three hours workshop if you really put these skills to use. So those are a few things, you know that like you said, a few situations, it's been about creating innovation and trying and testing new things in the market and that led to more market share. And it's different for every team and every organization, but where it makes the difference is where they choose to use it. And for me, the stuff that you see in that case study is just the proof points, right? That this works. The goal is that they internalize these skills and start using them more and more and get progress on all kinds of topics that I don't know about, because I trained them years ago, and hopefully, they're still, you know, I get stories here and there from teams that I worked with, months and years ago, and there's new things popping up, and I can't really keep track. But you know, making progress on those complex challenges, something that most teams need. And, and it's something that isn't that hard. If you just get into exploring, instead of trying to solve you know, get into that perspective-taking and start sharing the most important information, despite that interpersonal risk, you can make progress on most stuff.
Toni Dechario 32:10
It's really just building a habit and practice.
Per Hugander 32:23
Yeah, and I think also shared language. What really helps is when you have that shared language around this and say, “Hey, guys, this looks like a complex challenge that we're facing, maybe we should stop talking about the way we're talking about it and do an hour perspective-taking.” If you don't have the language that can sound really weird, especially if it's a conflict, right? But now when we have that language, and in organizations and teams, where you have that language, it's really easy to just put it there and say, “Let's do perspective-taking for an hour or two and see where we go.” And that's really powerful to have that shared language. So, if you have the language, you have the proof that it works, so you dare to use the language, then, you know, it's habit, you get better every time you do it. But you don't have to practice it for months and months and months to get results. Usually, people do pretty well the first day, if they just really try to get into perspective-taking then it gets easier, the longer you go through the process, but it's not this very complicated thing. It's just, you know, you got to put your mind to doing it.
Toni Dechario 33:26
Okay, so one of the topics that we're looking at this year is new ways of work. And a big part of thinking about new ways of work is thinking about hybrid work, and not always being in person. Does perspective-taking work, when you're doing it either remotely or in a hybrid way? Is there something to being in person that's important to the sessions that you're holding? What's your take on how the change in the way that we work is affecting your ability to bring perspective-taking and psychological safety to your clients?
Per Hugander 34:01
So, it definitely has an impact on how we can reach out and make this happen in the organization. So, it's good and it's bad. So, a few of the things that you would get from meeting in person you don't get from a digital conversation. And when Corona hit and all of a sudden everyone had to work from home, we were stuck with the question, “Are we going to keep doing this?” Because we were just getting to a point where this was kind of, you know, you got a pretty strong internal trademark on these – on this concept. So, do we want to try to transfer that to digital or do we just want to wait? And so, we decided we don't want to do the whole thing because we don't want to burn the trademark but what we wanted to do was keep going with the cohorts that we were already in progress with, and it turned out to work really, really well. You lose a little bit. So, what happens when you sit in a room together, I usually have people sit close, you know, no tables, knees almost touching, making eye contact, something builds that’s called synchronicity, which is when they – we start synchronizing brainwaves and heartbeats. And it sounds really weird, but it happens when people sit close and talk. So, you don't get that as much when you're meeting through a screen. But lots of other stuff works really well, you know, asking the questions, trying to mentally understand what it's like to be the other person, most of this stuff works. So I – my estimation this is, you know, not scientifically proven, or anything – is that it's 90%, as good to do these sessions, digital. There's something that you lose, but it's not a lot, it's definitely worth doing it virtually. But then we get to the good points, which is when you meet virtually to do these sessions – especially when it's when it's senior leaders – who are hard to get into one space, and there's always going to be someone who can't be there, when you have an hourly, person-to-person meeting every week. But when it's virtual, it's easier to make it. You know, there's no traveling time. You don't have to go in and out of different meetings, walking from one end of the office to the next and all that stuff. So, more people attend, and more people are there more consistently through the process. And I think that has a really positive impact. So, I would say that they equal each other out. In the best of worlds, you know, it would be in-person, everyone there all the time, but I think digital has some great benefits. And when it comes to the dialogues, you know, that's also core to this process, you know, to have a dialogue on a really important complex challenge. I think that works almost better digitally, because it's easier to manage the, you know, who speaks and who's listening and to spread the word evenly and move in and out of perspective-taking and sharing, which is a key to those dialogues, that works really well, digitally. It's just lots of people have concluded that digital isn't good for creative work. I think it's just because we don't use it for creative work, we haven't really tried it. But when you do try to facilitate creative conversation, it can be great. So, I think it works really well digitally. But I do admit that I want to start out person-to-person, face-to-face. That's my preference. I'm sure that it works the other way around as well. But there's something in that synchronicity. So, I do like the hybrid versions of doing it. There is something about having that person-to-person start when you want to build psychological safety that's hard to get over. You can do it virtually, but why chance it when you're investing all of this time.
Toni Dechario 37:59
In person, you can use the neural synchrony that takes place when you're in person to almost speed up perspective-taking is that right? It kind of facilitates our ability to take someone else's perspective, because we are like almost biologically synchronizing our heartbeats and our brainwaves are starting to synchronize.
Per Hugander 38:23
So, let's actually think about impacts the psychological safety even more, I would say, you know, it's a really good way of creating that psychologically safe space, when you're able to get people who don't agree to sit in those closer circles and have conversations with each other. It really helps the psychological safety and that makes perspective-taking easier. So yeah, they feed off each other. But I wouldn't say that it's only the perspective-taking, I think it's just as much the psychological safety, but you can really leverage it. You can virtually as well, but then you want to really make sure that people aren't sitting in a, you know, an open landscape environment and being disturbed and stuff flying around them and emails coming in. If you're in your own room and you shut off your emails and you know, messengers and Teams notifications and stuff, and you're really focused on what's going on on this screen, you can get almost the same quality, absolutely.
Toni Dechario 39:27
Fascinating. One of the things I wanted to talk to you about today, Per, is trust. That's another topic that the New York Fed has been focused on this year within the culture initiative. I know that you have some strong feelings about trust versus psychological safety. The listeners will notice the word “trust” hasn't really come up yet, in our conversation. Can you talk about why you focus so much on psychological safety over trust? And maybe some of the pitfalls that you see with trust?
Per Hugander 39:56
Yes, so trust. First of all, I want to make sure that everyone understands I love trust. I think trust is great. I would like to see more trust in the world. So, a lot of the time people point me out as the guy who doesn't like trust, the anti-trust guy. And I'm not saying trust is bad. I'm just saying that it doesn't do it for me when I try to help organizations succeed. So, there's a few problems with trust. And especially today, you know, there's all these sayings, right? “Teamwork is all about trust, collaboration is all about trust.” There – It seems like everything that we do is all about trust. So that could be true, because trust has so many different definitions. When you look at – and I've explored trust quite rigorously. Firstly, trust has lots of different definitions. And that's a problem. If I want to help an organization to move forward. If I say teamwork is all about trust, people can ask me, “Okay, so what kind of trust? Is it that I trust in your competence? Or is it that I trust that you won't hurt me? Or is that instinctive feeling that I have when we meet about trust or distrust?” There's so many kinds of trust. And when you look at the research, there was someone who – but I don't remember who – but there's a really good meta study on trust, where they concluded that there's so many definitions that the research on trust combines to create meta nonsense – no, macro nonsense, sorry – because there's too many definitions. So that's one problem. But when we look at the way, why I don’t use trust, there's so many ways I'll start with, with trust, if you focus on trust, you automatically also focus on trustworthiness. And people managing their trustworthiness, or their appearance to be trustworthy, is really problematic for me, in the kind of conversations, in the dialogues that I need. Because in those dialogues, if I'm going to help people to make progress on their most complex strategic challenges, I need the best possible information on the table. So, I need people to dare to put unfiltered information and a lot of the time, that is risky, because it could show that I don't know everything, I've made a mistake, or I need to challenge someone that you know, maybe someone above me in the organization, and all of those things could harm how people evaluate me as trustworthy.
Toni Dechario 42:31
That makes a lot of sense because one of the components of trustworthiness, according to Onora O'Neill – who I think of is kind of the primary thinker on trustworthiness – is competence. And so, if you're acknowledging that something is difficult for you, or that you maybe don't have full 100% competence in some area where you think people might expect you to, if you're focused on being trustworthy, then that's going to hurt psychological safety.
Per Hugander 42:58
Yes, exactly. So that's the problem. So, with trust comes that proving that I'm trustworthy, and that a lot of the time helps or stops people from voicing the information that I need them to voice to make progress on the most complex challenges. So that's one problem. But another problem. So, like you said, most definitions in organizational context about trust has something to do with competence, a lot of the times something to do with character and benevolence, and that through socializing with someone or interacting with someone for a long period of time, you can start evaluating how competent, trustworthy and benevolent that person is. So, it takes time to build that trust as part of the definition. So the problematic part here is that lots of organizations, definitely the one I work with, but lots of organizations that I've worked with before, they need to bring in new people, right, they need to bring in lots of tech people, if we're looking at the financial industry, you know, tech people, these key people, lots of new people that needs to come in, and help out in the, in the different decision-making processes and in the teams. So, when you add lots of people in a short period of time, those people who have the least chance of having to. – having established trust, right, we're running a risk of not evaluating the information they bring in as much because we don't feel trust for them yet.
Toni Dechario 44:28
Are you saying you know; trust is kind of personal? And psychological safety is more cultural?
Per Hugander 44:34
Yes, I would say trust is between two people. So, I can trust you, but I can't trust a team. I can trust all the individuals in a team. And then we have trust, but I can't trust the team. There's nothing there to really trust. I can trust the individuals. But you can create a psychologically safe atmosphere or environment or if you want to call it culture. That's the that's the atmosphere that you have. I don't have to trust everyone. I'm sure you've been to these, that could be a party, or it could be a training or somewhere where you haven't established trust, you hardly know people, but the atmosphere – they’re really open and great. And you can both share information and challenge each other. You can have great conversation without trusting people, but the atmosphere is there. So that would be the way I would think of the difference. So, there's lots of research, if you look at research on teaming, if you look at research on collective intelligence, that says that there's other things that comes in, that's really important that we can do, that doesn't have to do with trust. And it has to do with you know, sharing, contributing equally to creating a psychological safe environment to take each other's perspectives. If we do those things, we can do really well, before we've had time to, to establish trust. I’ve thought a lot about trust, and it's just not that helpful. It's a lot more helpful to focus on psychological safety, you know, that willingness to take interpersonal risk, to dare to put the most important information on the table, because most of the time, that is what we need in order to make progress on the stuff that we're, we're challenged by now. And it goes for, you know, big financial organizations, but it also goes for really small startups. A lot of the time, it's those complex challenges are getting in the way. And when your focus is to get progress on those, trust isn't going to help that much, it might actually make things worse.
Toni Dechario 46:33
That's interesting. It reminds me of one of the panelists in one of the webinars we did this year, a guy at NYU called Jay Van Bavel has done some writing about cohesion. And how, indeed, cohesion in a group helps encourage challenge, right, because you already have that cohesion, you already, people feel safe to kind of come up with wacky ideas, because they in your lexicon now feel that they've built that trust in that they've established trustworthiness already. But he says – and I think this reflects what you were talking about – that the process of trying to create cohesion hurts challenge. Sounds similar to your description of the process of trying to build trustworthiness or trying to build credibility can hurt challenge because you don't want to appear unsure, or, you know, kind of not competent.
Per Hugander 47:31
Absolutely. And this is, this is very problematic, right? Because in the situations where we need to make good decisions, this goes for companies, society, everywhere. Now, when people come together, if we have a period of time, where we're working to create trust, or cohesion, or whatever you want to call it, where people are holding back, not sharing good information, now trying maybe even to challenge each other to see if we're trustworthy, which can create conflict, if we have that, we're probably going to make really bad decisions. So, what I'm trying to do, what I always try to do is to find ways to get over that how can we perform during that trust-building phase or cohesion phase? How can we perform that as well? Because the way I see most of the important decisions we take, are going to be taking in that phase? So how do we do that? That’s the challenge. Then psychological safety to quickly get that up, but also to be trained to do perspective-taking, even if I don't like you, or if we looked at whatever it is, if I can quickly realize that this is a complex challenge, we need to make progress on it. And I can go into perspective-taking. And that also will create psychological safety, we can get to a space where we can make good decisions during this trust-building phase or, or cohesion phase and, and actually what I think and what I’ve – think I've seen proof of – is that when you stop emphasizing trust, and you start emphasizing psychological safety and perspective-taking, people can start making progress. And without trusting each other, or at least with the definition of trust that trust builds over time. That is actually trust-building in itself, to make progress together to see that, you know, even if we might not like each other, or we don't know each other, or I can't evaluate if you're competent or not. We have this shared experience of working together and making progress. That's actually a really good way of building trust. So maybe we can build trust, not by focusing on trust and saying that it all comes down to trust, but by not focusing on trust. Maybe it's a quick way to get that cohesion.
Toni Dechario 49:49
And ultimately, in the end, it's all about making progress on a decision, on a challenge, on the outcome and you're building confidence in your ability to jointly do that. And I'll bring it back actually to the case study and it shows that you have been able to get really tangible, clear, kind of business outcomes from all of this, which sounds soft. And, you know, a little like trust and psychological safety, some of the vocabulary can kind of create a resistance. You know, I think especially in like a field like finance, you know, which is wants to see the numbers and kind of the quantitative outcomes. And all of this psychological safety stuff is really soft. I think what's really important is you’ve shown hard outcomes. And so, it's a fascinating study. If you can make one recommendation to the people running financial firms, when it comes to influencing culture, what would that be?
Per Hugander 50:49
I can only do one? Okay, so one recommendation, go deep before you go wide. You know, really give one group of people, preferably someone who works with something important, that's quite senior in the organization, give them the chance, you know, a few days and a few hours to really learn the skills so that they can get the experience of this helps to make progress on our challenges. If you can do that, before you ask them to get behind something like psychological safety, or perspective taking or these soft words, it helps a lot. There's lots of people that are now getting behind this, that are promoting this, you know, both the skills and the vocabulary, that originally were quite skeptic to that soft language. And we even call it empathic listening to start with, but I switched to perspective-taking, but you want to give people a chance to really experience the benefits so that they can get behind it. That's where you need to start. Don’t start rolling out culture change based on theories about something that worked some other time, some other place and some other industry, you know. Make sure that people have proof that it works here, for us, in our context, and then try to spread that. So that would be, you know, one thing, if you're going to work on culture.
Toni Dechario 52:19
I love it that you changed what you called perspective-taking it – that you had previously called it empathic listening. What kind of a difference did it make when you made that change? Because I think language really matters, and how people respond and react to language is real. I'm curious about why you made that switch, and what results you saw when you made it?
Per Hugander 52:40
So, it's really hard to choose, you know, what are you going to call it, because we you could call these skills, both psychological safety and perspective-taking, you could call them other things. So, starting out, especially when you're in, you know, if you're going into a financial firm, or an engineering company, I feel that I want to have something with, you know, really credible research behind it. So, when you choose something, choose something that you can point to, you know, this is from Harvard, or MIT or INSEAD or Ashridge, or some really credible school, and some really credible researcher, because a lot of the time, some of the skeptics need that to come along. So, what we started out with empathic listening, that's because Otto Scharmer of MIT, you know, he’s a pretty credible resource. So that's why we use that language. We realized that it was quite tough for some people, empathy and listening, you know, it sounded maybe a little bit too soft, passive. So, when I started finding really credible research on perspective-taking, then it was easier to make that switch. I could have changed it to, to lots of different words, but I don't want to change it into something that doesn't have, you know, that isn't backed by research. Because if you want to spread this to lots of people and you want to spread it in an organization, you want to choose wisely, so that you can always back it up. But since we found lots of research on perspective-taking and the definition of perspective-taking in that research was very, very close to the way Otto describes empathic listening. So, it's not the same, you know, empathy isn't the same as, perspective-taking but empathic listening, the way he describes it to take the perspective of that other person is basically the same as perspective-taking. So, then I felt that it was fairly safe because I could back it up with research, and it was easier for people to buy into it. So that's why we changed that one. I've been thinking about, you know, psychological safety is also a hard word to get behind for lots of people. But Amy Edmondson has done such fantastic research and there's so much proof that this really makes a difference on complex challenges. So, I've stuck with that. And it works. It doesn't work with everyone. Language is important.
Toni Dechario 55:19
Wonderful. Thank you so much. This has been really fun. Excited to do this and to get the word out. You're doing really interesting work.
Toni Dechario 55:26
if you want to hear more from Per, watch the New York Fed’s webinar on “Trust in Banks: Where are We Now?” which you'll find at NewYorkFed.org/governance-and-culture-reform.
Sameer Srivastava 00:01
In many cases, the managers have assumed, you know, we've been able to maintain high levels of productivity, things are still getting done. It's all fine. And I think we don't really understand the ways in which the cultures in many, many organizations these days are fraying.
Amir Goldberg 00:17
So, if you're a leader, if you're a manager, if you have a vision to build a company and you think algorithms are going to solve that problem for you, you're in for a big surprise.
Alison Taylor 00:25
HR can't own culture and compliance can't own culture and senior leadership can't own culture. Each has a slice of culture, as do employees as a whole. So, if we still think in terms of delegating to specific functions to solve these issues, I don't believe they will ever be solved.
Per Hugander 00:43
When you want to influence culture, when you want to make that a stronger part of your culture, what you need, if you want to change the culture is to change people's assumptions.
Jeremy Brisiel 00:52
This is Bank Notes: Banking Culture Reform. The views expressed in this podcast do not represent those of the New York Fed, or the Federal Reserve System.
Toni Dechario 01:01
Hello, and welcome to Season Two of the Banking Culture Reform podcast part of the New York Fed’s initiative to drive awareness and change in financial services culture. This season, we're exploring themes of trust, technology, and the new workplace. We introduced these topics in our 2022 webinars, which you can watch at NewYorkFed.org. In this season's podcast, we’ve brought back panelists from each webinar to take a more in-depth look at their work around these three themes. My name is
Toni Dechario, and I'm with the New York Fed’s culture team.
Toni Dechario 01:33
Our guests today are Amir Goldberg and Sameer Srivastava. Amir is an associate professor of organizational behavior at Stanford's Graduate School of Business. Sameer is the Ewald T. Grether Professor of Business Administration and Public Policy at the UC Berkeley Haas School. He also co-directs the UC Berkeley Culture Center, Amir and Sameer co-direct The Computational Culture Lab, a joint initiative between Stanford and Berkeley. Welcome, Sameer and Amir. You're both doing really fascinating work on the intersections between technology and culture. How did each of you get to this point? How did you come to believe that culture is important? Sameer, how about you?
Sameer Srivastava 02:12
Great. Thanks so much, Toni. Thanks for having us for this conversation. So, the answer to your question for me goes back to my undergraduate days when I was an economics major, and I wrote a game theory senior thesis, which was a mathematical model of the behavior of firms and the decisions that managers inside those firms made. And the model worked out really nicely. But then I was looking for empirical examples of actual managers and actual firms behaving this way, and I couldn't find any. And it made me feel really disillusioned about this model and made me think I wanted to really understand how real managers, how real people in organizations made decisions. So rather than pursuing a PhD at that time, which had been my plan, I went into industry, and I worked for a management consulting firm. I ended up staying for about 14 years and spent a lot of time really helping organizations think about aligning their organizations with their competitive strategy, and very quickly came to see the role that social networks and culture both play in the way decisions get made, right? So, we might have an idea that we think is right, but then decisions are socially produced through interactions with other people and who you come into contact with. And the nature of those interactions, importantly, shapes what the outcome is. And so I wanted to really understand that better. And that led me to go back kind of mid-career to do a PhD. This time, I decided to go back and do a PhD in sociology rather than economics, because it felt like the discipline that better understood and took seriously the role of culture and social networks. And then as I started doing research, which was initially focused more on the role of social networks in shaping individual and organizational outcomes, I began to see that that literature was disconnected from the cultural context in which networks operate. And so a lot of the work I've been doing together with Amir, and I'll let him say more about his background, has been really trying to take culture seriously. And find new ways of measuring it, new ways of tracking it. And really understanding how people come to create group culture and also how they are influenced by the culture in which they are operating, whether they realize it or not.
Toni Dechario 04:36
Thanks, Sameer. Amir, do you want to tell us your origin story a little bit?
Amir Goldberg 04:40
Sure. Well, first of all, thank you for having us. This is really exciting to be here. For me, I think the foundational moment was when I moved to the United States. So, I have the advantage and the curse of being an immigrant, but I think it gives you it gives you a lot of perspective. I grew – I grew up in Israel, which is a very Mediterranean Middle Eastern culture, with a lot of like, history and fraught history to it. When you arrive in a different country you just suddenly become aware of the extent to which things that seem so natural to you are, in fact, very unique to that particular context. And it's very frustrating because I would say things – and I knew that I understood the literal meanings of what I was saying or what other people were telling me – but I didn't understand the implied meanings. One of the most common things that I've noticed in American kind of work culture is people say, “We should really have coffee sometime.” But it's not a literal invitation to having coffee, it's more of a nice way to end the conversation and suggest that maybe in the future, there will be an opportunity for us to socialize. And it takes a long time to understand these. Or the – or the baseball references and what they imply. And that was a moment where I understood that this is such a fascinating, and such an important aspect of our social lives, but one that we are so deeply and surrounded by that we don't even see that. You need to travel over the Atlantic to be able to observe it. And that was the moment where I decided that that's what I wanted to start doing. And it's very difficult to take that distance, and to look at yourself and others from the outside and ask yourself, what are the beliefs and interpretations that are shared across a group of people that they believe are just the natural order of things, but in fact, are a result of, of a very rich history that they're mostly unaware of. And that's what's so fascinating about culture. And now, with respect to the question of why financial institutions should care about culture, I think there are two main reasons. First of all is just because it's around us. You can't escape culture. It's kind of like the fable of the fish that is taken out of the water and asked, “How's the water today?” and the fish says, “What water?” So, we're swimming in the water of culture, whether we like it or not. And I think financial decision leader – makers, financial leaders, that are better aware of how their decisions and interpretations or bias through cultural contexts are gonna – are going to make eventually better decisions. And I think it's particularly acute in the finance industry, because we know from the study of culture that culture is one of the most unique things that humans have that has enabled us to become such a successful species on this planet. And one of the reasons is because culture is a way for us to share knowledge in a way that no individual could discover by him or herself. And the more the – the external environment is complex, the more culture becomes an important way for us to resolve some of these uncertainties. And financial markets, by their nature, are complex. And they're just becoming even more and more and more complex. And this was very much clear during the 2008 financial crisis, which was happening while I was doing my PhD studies, where a lot of the decisions of investors that seemed to be rational, profit-maximization types of calculations were, in fact, driven a lot by cultural processes, by shared interpretations that had nothing to do with the underlying fundamentals of the instruments that they were using. So, I think every person should think about culture, but in finance, in particular, in light of the complexity of the reality that investors are dealing with.
Toni Dechario 08:44
So Amir, if culture is thought of as kind of just the way things are and the water that we're swimming in. Does that mean it can't be changed?
Amir Goldberg 08:53
I think what it means is that, obviously, it can be changed, because it's not the natural order of things. But the reason that it is perceived as if it is natural, is because it is so deeply shared by people that - you know that the technical concept that sociologists and social scientists would use here is something known as social construction. And social construction is a reality that we have all come to believe is natural, but isn't necessarily the – the natural order of things. But the reason that it is held so strongly is because everybody shares it. So that means that changing that perception is going to be very difficult. So, if you think of one of the most dramatic achievements in that respect of the 20th Century is changing people's perceptions about gender and gender roles. And, you know, in the early 20th Century, many people held the position that women are simply incapable of leadership because they're biologically innately not disposed towards certain types of qualities and behaviors. And it took many, many, many thousands of people to provide evidence, to make kind of strong arguments, to protest. But we saw that change happen. And usually when it happens, it happens in a landslide, because suddenly, it requires a large contingent of people to change their opinions en masse. And then a lot of things happen. As a consequence, we saw – and also with, with the gay rights movement, in the latter part of the 20th Century – doesn't mean that things can’t change. But it means that their opinions are strongly entrenched. And in changing people's opinions, it is an uphill struggle.
Toni Dechario 10:53
So I want to get into the guts of what you both do. One of the themes that we're thinking about this year and that we've been focused on this year is technology and that relationship between technology and culture. This is why we brought you two in – you co-lead the Computational Culture Lab. I'm wondering, Sameer, maybe you can give me an introduction to what the Computational Culture Lab is, and what inspired its creation.
Sameer Srivastava 11:23
Happy to. So, I think it's been about seven years since Amir and I started the lab. And it came about the way a lot of academic collaborations come out, which is, I invited him here to come give a talk at Berkeley. And we had a nice chat afterwards. And he invited me to come give a talk at Stanford, and we had a nice chat afterwards. And then we decided to meet in the middle in San Francisco. Over lunch, we started to think about research collaboration. And as we started thinking about the ideas that we were both interested in, we also began to realize that the data – we would need to really get at these questions just weren't accessible and wouldn't be accessible using the typical methods one has available to study culture, things like surveys. Surveys are incredibly useful. They're very widespread, but they have important limitations, right? They tell you about what people think and what they report in an instrument. They're subjected to various forms of reporting bias. They're costly to implement in the sense that it takes time for people to fill them out. In many organizations, response rates to surveys have been on the decline for years. And so we really were starting to think about, you know, are there other ways to get at facets of culture? Again, this is about seven years ago, when we were still relatively early in what I think of now as, as the digital revolution as – as it relates to the study of culture. So, we started to think about, well, what are data sources that one could use to get at more behavioral indicators of culture, and ones that could be mined. So that rather than having to just get data, every three months, or once a year, you could get real time continuous measures. So, we started thinking about, for example, interactional language use: the idea that people in organizations are routinely communicating with each other. And increasingly, there are digital traces of that – whether it's in emails, or whether it's in Slack posts, whether it's on Glassdoor reviews – there are lots of different ways in which we can collect these digital traces. And around the same time, there was a proliferation of tools and methods for structuring that data, for analyzing it at scale, computational costs were declining dramatically. And the tools from computer science and computational linguistics were becoming more readily available. And so I think, a big idea, the Lab, the inspiration for the Lab was, can we harness these new data sources and these new methods to help us tackle good old-fashioned social science questions, but in new ways? And that, I think, is really, in some sense, an encapsulation of what we've been trying to do. Amir, what would you add?
Amir Goldberg 14:06
I think you’ve provided a fantastic answer. You know, I will just add that related to our – to your first question about our backgrounds, I think also Sameer, and I come from slightly different training backgrounds. Sameer started in economics, and then went into consulting kind of experience, the limitations of – of the theories and the practices firsthand. I did my undergraduate degree in computer science. And I also kind of understood that this was a moment where data digitization, new algorithmic capabilities are going to revolutionize how we think about, quote, unquote, “softest” aspects of human interaction and – and sociability. And we were very excited because we knew that these technologies are out there. But then our friends in computer science and statistics weren't thinking about culture as – as their immediate candidate. And that was an opportunity to be seized, but also a way to move the needle.
Toni Dechario 15:09
So maybe, Amir, you can go into as layman terms as possible, what are some of these computational methods that you're using? And then you can talk a little bit, perhaps, about exactly how you're applying them to aspects of culture.
Amir Goldberg 15:25
Sure. So, you know, one of the biggest advancements in computation, and in quote, unquote, “artificial intelligence” – which is, as a sidenote, I would say, a term I dislike, because we definitely haven't reached the moment of an intelligent, artificial algorithm in the sense that humans intuitively understand intelligence. But one of the biggest advances over the last ten, fifteen decades has been in natural language processing. The pace at which these algorithms are developing is just staggering. Just over – since Sameer and I started the Lab, the pace has been unmanageable. If you don't stay up to speed with what's happening in engineering and computer science for a few months, you really lose out. If you use methods that were developed in 2018, and it’s 2022, you're like, a century behind. So, it's exciting. And what – what these advances have kind of implied are several things. First of all, they systematically are able to analyze text at scale, in a way that no human can. A human can code, maybe hundreds of thousands of documents. These algorithms can scan millions, billions of documents, and learn something from them. And recently, there's really been a lot of advancement in semantics basically, in understanding the underlying meanings. And not just the statistical patterns of text. So, you know, in the late 1990s, the way by which natural language processing happened was that you would count frequencies, and you would look at, okay, how many negative words how many positive words? Now there are what's known as “deep learning” methods. And we won't get into the details of what exactly differentiates deep learning from other forms of learning. But these are machine-learning algorithms that appear to be able to represent semantics in a way that's quite consistent with how humans understand the meanings of words and the meanings of sentences. And some of them, novel algorithms, are just amazing. I encourage everybody who's listening to this podcast to go to Open AI’s GPT-3 model and play with it a little bit. You ask it questions and it answers quite astonishingly well. It's quite easy to break it. So, you will very quickly realize that it's not intelligent in the way that humans are. But for certain tasks, its ability to answer very open-ended questions is astonishing. And we use that as a way to try and infer what do people mean when they say, “Let's have coffee sometime.” And do these set of people mean different things than this set of people? And can we learn something about what's driving their behavior and why they're making the decisions that they're making, by bringing in the insights that these algorithms are giving us? So, in one of our current studies that I think is going to be very relevant to the topic of our conversation, we analyze the text in quarterly earnings calls, where executives talk to analysts and – and describe their firm performance and their vision and their strategy, etc. So, one of the things that we've done was to apply one family of algorithms known as “word embeddings.” These are deep learning algorithms are quite amazing in inferring semantics. So, what they are really good at doing is – they study a large body of text, and they can locate each word in a multi-dimensional space. And it turns out that the distances between these words in that space strongly correspond to how humans understand the meanings of words. So, the word ‘dog’ and ‘cat’ will end up being far closer in space, then ‘dog’ and ‘podcast’ which are semantically unrelated terms. So, what we've done is we use that attribute of word-embedding models to analyze the text of executives and quarterly earning calls. And when we – we observe an executive speaking in such a call, we ask, is the way that this person is speaking, semantically similar or different from how her competitors are speaking contemporaneously? We call this “performative atypicality.” So how atypical they are in their performance – in their linguistic performance – in these calls. And we find something quite interesting. We find that firms whose executives speak atypically in these types of calls, analysts tend to overestimate the future performance of these firms, which, ironically, leads to a negative earnings surprise. They are kind of over-buoyant about – about how this firm is going to perform. And we find that this is especially the case, if these executives atypicality, it kind of emulates the way by which celebrated innovators speak. That tends to influence analysts, and to bias them to believe that you have something special. They believe it so much, that they are overly optimistic about how well your firm is going to perform in the future. So that's just one example of how you can use these algorithms at scale to discover these patterns that are quite subtle, and would have been extremely difficult for a human coder to – to understand at scale.
Toni Dechario 21:10
That's so interesting, because my intuition would be the opposite. You know, I've been, you know, part of lots of different cultures in my life, right? It's not just your work culture, it's the culture of your family, and the culture of any – any sort of kind of identity that you might have, has, has a culture associated with it. And I find that people often end up also using the same words. You know, we do have shared meanings, and we even have shared ways of saying things, even intonation, sometimes you kind of start to just at – people start to really sound alike. And so my intuition would be that it gives people kind of comfort to hear someone who sounds like the others that are doing something similar. And in fact, maybe that is what you're saying. It's just like, the visionaries of that group, are who do well, but what about a few speak differently, but not like a visionary, just like yourself?
Amir Goldberg 22:11
You're absolutely right. This is like a perennial tension, not just in quarterly earnings calls, but in every social situation. There's these kinds of cross pressures of fitting in, but also being different when you're dating. You don't want to be too weird, but you don't want to look like everybody else. Right? When you're making a presentation, when you're talking at the family dinner, you don't just want to parrot what everybody else said, but you also don't want to be entirely unintelligible. There are different ways for people to look different than to be similar. What's really interesting is that if you are idiosyncratically different, you just look weird. If you are, however, kind of, are very good at picking on the subtle ways by which innovation and creativity are conventionally understood, and you are able to differentiate yourself from others in that way, that's when that has the biggest effect. I kind of think about it as like being cool, right? What does it mean to be cool? Just being different doesn't make you cool. There are the right and the wrong ways of being cool. And I think there are also right and wrong ways of being cool in quarterly earning calls in a way that very kind of subconsciously affects the interpretations of analysts. And what Sameer and I enjoy is that you can use algorithms to identify that in a way that none of us could have done it individually. When we read those transcripts, we just drown in the data. We don't understand what's going on, we need the help of the algorithm to do that for us. Sameer, do you want to add to that?
Sameer Srivastava 23:51
I was gonna say, first of all, that Amir and I figured a lot of things out over the last seven years. But one thing we haven't figured out is the right way to be cool. Still working on that.
Toni Dechario 24:01
I think you're cool. I think this is cool.
Sameer Srivastava 24:02
Very kind of you to say. So, the other example I wanted to highlight which is from the financial-services sector is the use of such tools that Amir was just describing, to actually measure post-merger cultural integration. The specific manifestation here is just a machine-learning classifier, right? It's a – it's a way of trying to take a given form of communication – in this case, we take email messages – and we try to classify them. Is this a message that was written in the style of the legacy organization pre-merger and to what extent can we actually distinguish between the language being used by people post-merger? And what this allows us to do – the kind of academic research side of it – is trying to develop a language-based measure of what we refer to as “symbolic boundaries.” But which meant, you know, people in organizations will think about as the various subtle “Us-Them” distinctions that we maintain between groups. And we do this all the time. And we do it often automatically. But it turns out, we can actually assess this in language, we can see people communicating in ways that reify their old legacy organization. And we see people coming up even as culture shifts and language changes, finding new ways to draw cultural distinctions between their group and the other group. And so, in this particular setting, where we have one bank that buys two other banks and integrates with them, we can actually assess where people are in terms of their linguistic style. And, 18 months after the fact, still figure out what their legacy organization was. And then from a managerial perspective, this type of approach helped us give the leadership team some advice about which bank branches, for example, are better- or less-well culturally integrated, and therefore where do they need to spend their time, right? That's an example of how these tools can be used for managerial purposes, but in ways that are not about revealing any individual's score, or doing so in ways that are – that can be used in nefarious ways. So, you can use it as a responsible tool. But – we can talk more about this – we need to put in some safeguards to make sure they're used responsibly.
Toni Dechario 26:17
Yeah, and I definitely want to spend some time on that. But I want to spend a little more time on the research first. Sameer, it would seem like if you can use this post-acquisition to figure out kind of maybe where the holdouts are, you could apply it to any organization to see subcultures. And the existence of subcultures isn't necessarily a bad thing, right? I mean, there are going to be subcultures in any big organization, you can't have one behemoth. That's if we've learned anything, we've learned that. You can see – my understanding of semantics similarities – is that what you can see is kind of how tight that subculture is. If I'm able to describe what Amir explained about semantic similarities, you can kind of map – physically map – how people can associate different concepts, how closely they associate different concepts to one another. And the more similar those maps are to one another – that each individual's map is to one another – kind of the more semantic similarity they have. I wonder, what does that reveal about a group? Is there something either kind of good or bad that can be deduced from how close they are?
Sameer Srivastava 27:28
So, a couple of things, which you said that I want to pick up on, Toni. One is: you're absolutely right; you're talking about similarities and differences. And that reflects a way that Amir and I have really approached this question of culture, which is, broadly speaking, you could think about content-based approaches to culture. These are ones that emphasize certain terms or certain values were innovation-focused or adaptability-focused or integrity-focused. Or you can think about this distributional approach to culture, which is, to what extent are individuals and groups and the organization as a whole similar to or different from other entities, culturally independent of what that content is? And these methods we've been describing are especially good at the latter. Of course, they can you can get a former too, but they're especially good at the latter. And so one of the things that you can start to see going back to subcultures is that managers and leaders have assumptions they make – mental models about where the subcultures exist. And those assumptions typically follow the contours of formal structure. We're going to assume, “Oh, those risk management people have this kind of culture, and the lawyers have that kind of culture and,” and so on, right? But with data like this, one can actually find subcultures that sometimes cut across groups. So, the assumptions we have that – the stereotypes, almost, we have about groups may or may not hold. We may find that there are subcultures that exist that are totally independent of these groups. And then we can start to investigate. And this is a harder question of trying to pin down when a subculture is harmful or hurtful, because sometimes subcultures form in response to that performance, as opposed to causing the bad performance. So, the causality can be difficult, but you could then begin, once you've identified those subcultures, then you can start examining the consequences of those cultures.
Toni Dechario 29:21
We've spoken with other guests about network analysis, and kind of the understanding you can have about subcultures and where you're kind of nodes of influence are using network analysis. So, this sounds to me like you can actually come up with some similar information using NLP. Have you done any work combining the two?
Sameer Srivastava 29:42
Amir, you want to take this one?
Amir Goldberg 29:43
Yes, we have. Some of our earliest work was precisely focused on the relationship between networks and, and culture. I would say, you know, to begin with, is that there is the assumption that culture, and the structure of the network, necessarily are coextensive with one another. So, you would imagine: if we have an organization where one group of people only interact with themselves and the other group of people interact with themselves too, and there's very little bridging between them, that there should be different cultures emerging in these two subgroups. That's often the case, but it's not necessarily the case. And good organizations can actually cultivate shared cultures, even if there are pockets of dense interaction that don't have a lot of bridges across them. Our early investigations were to try and explore this type of duality. And one of the things that we found quite interestingly, was that there's a trade-off, especially if we look at individuals, between the tendency to align and become very similar to your colleagues and your peers that you interact with or be different with them. And it's contingent on the type of network that you are part of. So, on average, a general finding in the study of culture is that there are higher returns to conformism and to thinking like other people around you. Most people who express different opinions are just shocked by their peers, and we never hear about them. So, for every Steve Jobs there are probably thousands of Steve Jobs wannabes. So, but we found that, actually, when you are embedded in a very dense network, where your colleagues have strong friendships with one another, you're kind of part of a quasi-family in the organization. That's when you have more latitude to be a little bit different. I like to think about it as kind of your weird uncle during a family dinner. The uncle is part of the family and he knows that he has kind of the security of the trust of the family members are not ever going to kick him out. And he can occasionally say weird things. But nobody's gonna suspect him for trying to sabotage you know, the family. So, there's an interesting trade-off there. And I think they're also insights for managers. Some organizations need to cultivate more heterodox cultures, need to encourage people thinking outside of the box, especially if their competitive advantage is around innovation and creativity. And the insight would be that if you want to encourage people to think outside the box, you need to create networks that provide the kind of social safety net for people to express unconventional opinions and not risk the penalty of being shunned as a consequence. So the two are not necessarily coextensive with one another.
Sameer Srivastava 32:51
Could I just add, also, that network analysis, of course, is a mature tool and technology. And, broadly speaking, there are kind of two approaches to it. One is there are a lot of firms out there that are providing services around meta-analysis of internal communications, right? So we can get through just the metadata on who's sending email messages to whom or who's sending Slack communications to whom, the structure of the communication network and how it changes over time. And that's great for understanding the network at scale. And then there's another approach, which is to run a network survey, where people are asked to nominate those who are in their network, or they've gone to for different kinds of resources. That can be really useful for getting at the content of the tie. The metadata doesn't tell you anything about who's mentoring whom, or who's seeking information from whom, or who's sabotaging whom. All you know is that they're in touch with each other. The survey data gives you much better insight into what's flowing through the tie. But it's very hard to collect those data at scale, especially if you're a larger organization, you get all kinds of again, recall bias. These are well documented biases, and whom we tend to list when we're asked a network question. So, to me, one of the opportunities is really to bring language analysis into better dialogue with network analysis, just as you were suggesting, Toni, because through language, we can get a much richer insight into the content of what's flowing through a network. And through the digital trace metadata, we can get much more comprehensive insights into who's communicating with whom. And there's some work that we have underway right now that's trying to do just that, together with a validated survey.
Toni Dechario 34:36
I actually have the benefit of having read some of your research, that I don't think has actually come out yet, into semantic similarities. Which is why I was able to get it on the first try, because – because I had the benefit of reading a little bit about it in advance – and one of the stories that you talk about in your research that hopefully we'll be able to link to in the notes to this show – is levels of what you call organizational identification. We've started to get into that a little bit with your identification of subgroups. But this is kind of a larger organizational question. But I'd be interested in in having – having you describe kind of what organizational identification is, how you're able to measure it. And why this approach is – is potentially superior to surveys.
Amir Goldberg 35:21
Generally speaking, there is a huge limitation with survey analysis. At the outset, I just want to say, I don't think that algorithmic approaches are necessarily superior to surveys. There are certain things that surveys can get that you cannot get by using our approaches. But the limitations are, aside from the obvious limitations, which are scale and granularity, you can’t ask the same person to answer a survey day in, day out, and do it like a huge scale. People will get pissed off. And it's also – it's just simply not a viable option. But aside from that, you know, when you ask people a question, in a survey, you activate their conscious cognition, and they become self-aware. And often they give you unreliable answers, not because they're lying, but because their self-presentational self is very different from their authentic self, and they sometimes are unaware of it. We all are. We have been socialized heavily not to be authentic, right? When we go up the elevator, and there's somebody that looks weird to us, we're not going to say to them, “Hey, you look weird.” We have been very kind of aggressively socialized to keep our opinions to ourselves. And that happens when we answer surveys. What this means is, especially when your employer is asking you questions, you're going to put on your presentational self, whether you do it consciously or not. And that's going to skew the answer. You're probably going to answer your ideal answer: what you would have liked as opposed to what you truly experienced at that moment. So that's the advantage of using our methods relative to surveys, especially when you ask these kinds of very abstract, but fundamental questions. “How strongly do I identify? Do you identify with an organization?” And the other thing is that when you think about the problem of identification – by problem, I mean, as a social scientific question – why do people identify more than others? Is it something about the difference between different organizations or simply, certain managers or leaders better at cultivating identification? Is it about the difference between different individuals, or maybe there are actually a lot of circumstantial situational differences that lead to the fact that there's actually a lot of variance within people over time. And we all experience that. You know, when we were hired, our level of identification with our employer kind of fluctuates. It's a function of a lot of things that are happening in our lives and in our professional and work lives. And when you bring our methodology at scale, you can suddenly look at those trends with a magnifying glass in a way that a survey could simply never afford. And we were able to discover that, yes, a lot of the difference in why some people identify more than others has to do with stable differences between individuals. Some people are just like Groucho Marx, they're cynics. They don't want to be a member of a club that would accept them as a member. Some people are very kind of gung-ho and would want to be a member of any club, whether it accepts them or not. A lot of the difference is explained by that, but we also discover that the type of networks that you're embedded in to build on the – the research that, that I described earlier. Whether you are a member of a tight-knit group really appears to affect your level of identification. I want to take Sameer’s caution, we don't know exactly what is the causal direction here. Is it because you identify that you create an – a certain network around you? Or is it because you have a network that your identification, intensifies, or weakens? But we definitely see a relationship between the two. The more you're part of a tight-knit group, the stronger your identification with the organization, but also, the more your relationships somehow span across the organization, where you have access – whether directly or indirectly through your immediate colleagues – to people who are scattered across the organization and are not only exclusively members of your particular subculture. Do you also identify more with the group? There are two kinds of trends here: you want to be part of a dense clique, but at the same time, you don't want your clique to be an island that's entirely disconnected from the mothership. We believe that this isn't actually only true about organizations, maybe this is a way for us to think about some of the problems that the United States – and actually all industrialized countries – are experiencing right now about political polarization, ethnic animosity between subgroups. There's something about being able to be simultaneously part of a strong community, but also feeling that your community has ties to other communities that intensifies your identification with this higher order group. Sameer, you want to add to that?
Sameer Srivastava 40:33
Yeah, I was gonna add just a couple of things. So, one is just in terms of the specifics of the measurement strategy. So again, this is a paper in which we use these word-embedding models that Amir previously described. And we focus in particular on pronoun use. So there have been prior attempts to try to measure facets of identification based on just the counts again, of “we” language or “I” language, right? To what extent am I thinking about the collective when I'm communicating? But here we're trying to measure something a little bit more subtle, which is the degree to which my conception of the self – that is, “I” – overlaps with my conception of the collective – that is the “we.” And this is something that you can't measure through just sheer counts of words, right? We have to actually look at the semantic meanings, as you were saying, Toni, between these words and how it shifts over time. And you can actually see that for those who identify more strongly as measured, for example, in survey data, that they, over time come to have greater overlap. Like, they literally are thinking about the self and the and the group in ways that are more aligned, that are more in unison, and that we think is the novelty of that measurement strategy. And also very quickly, wanted to name some of the collaborators on the various projects we've just mentioned, because we have not mentioned them. So the project that Amir mentioned with analyst calls, that's with Paul Gouvard, who's an alumnus of the Lab now at the University of Lugano. The project I described using the post-merger integration and financial-services cultural integration is with Anjali Bhatt who is an alumna of our Lab now a professor at Harvard Business School. Govind Manian is an alumnus of the Lab who was involved in the project that Amir mentioned regarding fitting in or standing out. This last one is with a current PhD student, Lara Yang.
Amir Goldberg 42:24
Forgot to mention our [Stanford] colleague, Chris Potts, at computer science and linguistics, was involved in some of our earlier research.
Sameer Srivastava 42:30
Absolutely.
Toni Dechario 42:31
So, you're spreading your network. You've got – you've got them moving out, building their own networks. Okay, so I can actually get into the – the nitty gritty of your studies all day long. But I do want to move into a discussion of the ethics of all of this, because I think that's really important. And it's a pretty big debate, what is the line between kind of using these powers for good and using these powers for bad?
Sameer Srivastava 43:00
I'm happy to go first on this one. I'll give you an example of the use of these tools and what I think are really misguided ways. So, there's an organization – I won't name it – that we came across, that uses some of these same tools, and they get access to internal communication data from a subset of firms. And they mined these data. And they can predict, because it's not that hard to predict when someone is at risk of exiting the firm. And it's one thing to make these predictions across large groups, or look at these aggregate patterns, but it's quite another thing when you get down to the individual level, and – we're trying to predict, Toni, when you might leave the Fed. You know, that's a very, very sensitive thing, which, if that's known to you, or known to leaders, can have all kinds of unintended consequences, how your manager then treats you or doesn't treat you or what opportunities become available to you or don't become available to you. And this particular organization was producing reports – weekly reports – for all people managers, with the predicted probabilities of each of their people leaving at the end of the month. And not only were these subject to false precision, to the Nth degree. You get predictions of 10.56% right – which we just know is ridiculous, given the accuracy of these models. But there was no attempt to think about unintended consequences. And so I think among the things that we encourage firms to do, of course, we can collaborate with them on research, and we can put in place our own protocols and our universities require us to put in place protocols to protect human subjects and, of course, we follow those. But after we leave, firms can do what they wish and our advice to them is to really think about using these tools in ways that are about self-empowerment of employees, rather than that are about surveillance and monitoring and using them in punitive kinds of ways. Insofar as these are opt-in tools, we think they're much, much better. So, you know, imagine a set of tools, and these are out there where people can opt-in and get data reports that are just for them and how they're communicating, and maybe some suggestions for what they could be doing differently or better. That seems like a really useful tool, a force for good. But I think the other important point here is transparency. So, if it becomes known that such data are being used, and they're being used in ways that are opaque, and mysterious, it will very quickly erode trust in your organization, very quickly erode trust in leadership and in very subtle-but-predictable ways actually destroy the culture of the organization. So – and whenever we do this kind of work, we will often go to a town hall meeting and explain exactly what data we're working with and what the procedure is for de-identifying the data and show some examples of ‘here's the raw text. And here's what it looks like after it's been processed,’ and answer questions because we think that kind of approach is really what people need to feel like these tools can be a force for good as opposed to evil.
Amir Goldberg 46:09
If I may add to that, I think technology is morally neutral. You know, atomic energy was used to reduce the cost of producing energy, and this had huge environmental positive effects. But it also killed hundreds of thousands of people in Hiroshima and Nagasaki. So really, the technology in and of itself affords all kinds of things. And the question is, how do people use it? So, I think it's really important that researchers like us are being held accountable by the public. It's really important that organizations are being held accountable by the public. I want to say very clearly, I don't trust organizations are going to do the right thing. I think it's true what Sameer said that doing the right thing ethically will probably be the right thing also from a profit-maximizing perspective, because the downside of ruining your culture is huge. So, you want to self-empower, and you want to make sure that you're not alienating your employees. But there are – I think there are two particular risks with the use of technology of this type. The first is that when you use data, quote, unquote, you have the false confidence that you have some “objective representation” of reality. We researchers understand that this is not the case. Figuring out what causes what or whether the measurement actually truly measures what we think it measures in the world is so complicated, and we argue about it all the time. But when a manager is presented by some, like, you know, this person is 2.3% more likely to leave than that person, we understand that that number is probably meaningless. And it might only mean something when you look at the aggregate, but not at one individual. So that's very dangerous. I don't think that managers are numerically kind of educated in a way that they understand statistics to make the right inferences. And that's our responsibility as educators. The second thing is that a huge danger here is that the same algorithms might be deployed across multiple systems. And if there's algorithms are biased, and they are by definition biased, because they are trained on data that was produced by humans who are biased, the same person might be disadvantaged across different contexts. So, they will be seen as a misfit culturally, in that firm, and in another firm, and they might not be hired, and they might not get credit. The use of algorithms introduces all kinds of new problems, even though the general problem which is how do we evaluate other people is problematic whether an algorithm does it or whether another human does it. And I think it's our responsibility as a society to have that conversation. I'm always disappointed – I hear, you know, Biden, and Trump and others talk about a lot of things. They rarely talk about these issues. This is like one of the most pressing problems of our time, period. And we have to talk about it.
Toni Dechario 49:10
And actually, one of the other topics that we've been really focused on this year – there have been three – technology is one of them and trust is another one. And trust isn't just an issue with financial-services firms, although that was a huge issue coming out of the financial crisis. But trust is an issue everywhere with everyone these days. Are there ways that technology can be harnessed to actually improve trust or more specifically within organizations to enhance psychological safety? I mean, is – is there a way to fight this…using technology?
Sameer Srivastava 49:45
I mean, so let's just take an example of the self-empowerment tools we talked about. So, in many cases, you know, what leads a workplace or a social group to feel less psychologically safe is because people communicate, sometimes in unintended ways that signal that they are being judgey or evaluative or that someone else's ideas aren't worthy. And again, it's sometimes you're engaging difference, you're not necessarily denigrating when you're interacting with somebody, but that's how they perceive it, right? And so an example of a way technology can help with the building of trust is by giving people some tools. You're about to communicate in a way, whether you intended to or not, that the other person is likely to perceive as denigrating. And if that's your intention, go ahead. But if it wasn't, maybe you want to self-correct, you know, and the person can choose to do it or not do it. It's not shared with the manager; it's not used for performance management; it's just a way to help the person become aware that they may be stepping into something that's going to undermine trust, right? And so that's just one simple example of how you could use it.
Toni Dechario 50:53
I'm curious about how you actually reassure people that their data isn't being looked at, though, that it's that it's genuinely just for them. You know, don't you run into kind of given the state of trust in our society, don’t you kind of run into the problem of people thinking, “No they're – they're using this for something else. I just, I just don't believe that.”
Amir Goldberg 51:17
Of course you do. And I think the fundamental challenge of leadership is to build trustful relationships. So, if you're a leader, if you're a manager, if you have a vision to build a company, and you think algorithms are going to solve that problem for you, you're in for a big surprise. This is sometimes the wrong way of thinking about, about this whole discussion of using algorithms in work. And basically, in other contexts where there's human productivity is to think about them as complete substitutes. “Okay, we're going to just outsource this job to the algorithm.” I don't think that's what's going to happen. What is going to happen is good, thoughtful leaders are going to ask themselves, what are the various tasks that I can use the algorithm, either to fully substitute me on or to give me good predictive analytics on what are tasks that I've never done that the algorithm now enables me to do, but there are certain things that I need to continue doing and building trust with your employees is one of the most important things that that a manager does. So, it's not something about the design of the algorithm in my view. It is something about the culture of the firm. And my knowledge, that when my manager tells me, “We're implementing this algorithm as a self-empowerment tool,” that I can trust them to live by their promises. And we know from many, many years of research and organizational behavior, that the way to do it is not just through exercising cheap talk, is when managers show that they're willing to make sacrifices when these types of dilemmas emerge; that they're willing to pay a cost – that the company is willing to pay a cost – in order to protect its values. That's the only way to communicate to employees that there's true kind of action behind those words that are otherwise vacuous. Nobody in Enron said, “Our goal is to cheat our customers.” They surely had, like, very beautiful statements about how important trust is for them. But in practice, they never implemented that. So, I don't think algorithms change any of that. They just introduce new types of challenges and problems. And good leaders are going to need to think hard about how to solve these problems. Sorry, we can’t – we're not going to solve that problem for you.
Toni Dechario 53:43
I think that sums it up. Alright, I have just a few more questions, kind of rapid fire. What's on the horizon? Kind of what directions? Do you plan to take your research going forward? Can you give us kind of a sneak peek?
Sameer Srivastava 53:57
I'm happy to go first on this one. So, I think one of the things that we've alluded to, is the need to validate a lot of the digital trace measures we've talked about, right? And so we have just started on a few projects where we're bringing together traditional surveys with digital trace data and measures. And I think a really important question is to understand the conditions under which the two are aligned and tell us something similar; the conditions under which they tell us about different facets of culture. And when some of the new measures we're developing actually aren't all that informative at all, which I think is going to be the case for many of these. The sort of metaphor that comes to mind and thinking about where we are, with the explosion of some of these tools is, you know, the drunk searching for keys under the lamppost. In the sense that right now, there's a huge amount of effort on mining existing and readily available data. So, Amir and I had one of the first papers that tried to measure organizational culture using Glassdoor reviews. Since then, there's been an explosion of papers and now a firm that's even providing culture scores based on Glassdoor data. And it's great because the data are widely accessible, and you can work with them. And you can do lots of things. But if you really think about it, there's a – there are a lot of problems with those data too, right? We have certain people selecting onto the platform and writing a review or not writing a review. What leads them to show up on the platform or not. We know that some HR organizations game Glassdoor reviews, particularly right before the best places to work lists get generated. We know that there are structured prompts on such platforms that get you only a window into certain facets of question. So, if we think about it, to me, another big question is, there's a whole range of data available, some of which is easy to get, some of which is less easy to get. So, this interaction that we're having right now over Zoom has a lot of cultural signal in it that has yet to be mined. We have facial expressions we're making; nonverbal layers of communication. We have paraverbal layers of communication, in terms of how we say what we say, the tone, and the pauses, and so on. We have various acts of subtle deference and lack of deference happening amongst us. And there's culture in there. So, getting at some of these other facets of culture beyond what you can just measure through text and these structured forms, I think, is a big area. And then the last part, which we can talk about more is really understanding what works or doesn't work in terms of changing culture. And I think to do that, well, we're going to have to think about running randomized controlled trials of the kind that get run in pharmaceuticals and that firms actually do all the time with their websites, A/B testing, but we don't really do good A/B testing as it relates to cultural practices. And we should.
Toni Dechario 56:55
So, Sameer, maybe you could give some examples of ways that actually academia and industry could work together to do some of that testing.
Sameer Srivastava 57:06
Yeah. So, I think there's a few different ideas that come to mind. One is, you know, Amir and I – and not just us, a whole range of people within and outside of our Lab – are developing new ways of measuring culture. We're publishing the stuff mostly in academic journals. And then there are providers out there that purport to provide different ways of measuring culture through platforms and tools. The two worlds have generally not been that well connected. And so, I think one simple approach would be, how do you take some of the research findings and begin to responsibly productize them in ways that will provide insight without the negative consequences we talked about? Second is this idea of a randomized control trial at scale, which we refer to as a mega study. And one of the things I know Toni and I have talked about is the idea of a financial services-focused mega study that would be about trying to shift the culture across a set of banks and institutions. And that's something that, you know, we're very excited to explore. But I think a third area that comes to mind is a lot of the focus in organizations is about what's happening inside what's happening to the employees of the organization, or maybe just to the direct customer. But I think a really important, untapped opportunity is for the academic world and the world of practice to come together to understand the broader social impact of what happens in organizations. And here I'm thinking about in the banking world, what happens in underserved communities. So, we know for example, from the work of Fred Wherry, who’s a sociologist at Princeton, about the importance of dignity and respect. And a big factor in why people in those communities are hesitant to go to their banks, and instead often choose to go to the payday lender, you know, that's nearby, is because they actually get treated with more dignity and respect. So, here's an example of how the internal culture inside of banks has a consequence that spills over and is having a negative influence on society in terms of inequalities in the world. There's another very nice recent paper by Mario Small, a sociologist at Columbia, who's looked at literally the travel time: how long does it take you to walk to the nearest payday lender versus the nearest bank branch, if you live in a low-income community? And, it turns out, it actually takes more time to get to the bank branch, right? The locational choices we're making are also having negative consequences. So, I think, more work to really understand like the spillovers and how the financial-services sector is contributing to or – and hopefully over time helping to address – these bigger social problems is a real important next step too.
Toni Dechario 59:54
That's fascinating. Amir, do you have anything to add?
Amir Goldberg 59:58
Sameer and I are both positioned in a place that allows us to interface with industry more than in most academic positions, because we operate in business schools. So, I interact a lot with my MBA students, and with alumni and with executives who come here to do executive-education sessions. And I often find that the conversation is very fruitful. There's also sometimes a danger when firms and academics collaborate, because then the financial for-profit incentives, and then knowledge-generation incentives can get kind of intertwined and might contaminate one another. So, it's often very important for us to very clearly create that boundary and say, “The research is intended to advance knowledge on the topic, not to advance your revenue in particular.” And I found there some firms that are amazing and willing to actually make that sacrifice for the common good. Some are less so. I also think, pursuant to our conversation earlier on, on the ethics of it, that I think we should – we should have a tighter relationship with nonprofit organizations, with government, and to encourage the conversation around these topics. As policymakers are trying to address policy through a changing world. Here at Stanford, there's the HAI, the Human-Centered AI Center, which is really leading the world in terms of discussions around the implications of AI, and its societal impact, its economic impact. And I would like government, business, academics, to be involved in those conversations. And I hope that, you know, voters will, will demand those conversations to be had. Because everybody's going to be influenced by an algorithmic – whether you want it or not. Algorithmic decision makers are going to be there when you get hired, are going to be there when your salary is being negotiated, when your credit is going to be negotiated. When you buy a ticket for an airline ticket, it's already being driven by algorithms. And we need to have a very kind of robust conversation about what's the role of government and what is the responsibility of organizations in making sure that the kinds of inequalities that Sameer spoke about earlier on not reinforced at scale.
Sameer Srivastava 1:02:35
One last thing I wanted to say on this topic is: for those who are interested in getting better connected to the world of academic research, one vehicle for doing that is the Berkeley Culture Center, which I co-direct with my colleague, Jenny Chatman. We have an annual conference that brings together industry leaders who are thinking about culture, as well as academics around the world who are studying culture, from different disciplinary vantage points, and a series of ongoing initiatives that people can use to engage with academics throughout the year, including on joint research. So, if you're interested, feel free to check that out: Berkeley Cultural Center.
Amir Goldberg 1:03:10
It's a great program and exactly – and embodies this academia, business relationship, but in a way where the different objectives of these different parties are still maintained. Researchers are coming in to talk about the researchers and research and business interests are coming in to, to voice their opinions and to learn new things. So it's a really amazing initiative that Sameer built.
Toni Dechario 1:03:35
That's a great resource. Thank you. Final question: if you could make one recommendation, each of you, to the people running financial firms, when it comes to influencing culture, what would it be? Amir, I'll start with you.
Amir Goldberg 1:03:47
I would say get out of your comfort zone and speak to people. You can’t learn about culture by looking at aggregate statistics or dashboards or reading Harvard Business Review articles. You need to go and talk to people, and you need to go to talk to people across all ranks in your organization, but also across all of your constituents outside your organization. And not just your immediate stakeholders, like your investors or your board members, but your customers – especially the customers that are culturally most distant from you. And if you don't talk to them, you just – you won't know – the only way to discover things you don't know is to have conversation. You can't imagine what you don't know unless you actually engage in interaction.
Toni Dechario 1:04:39
The famous “unknown unknowns.”
Sameer Srivastava 1:04:42
Yeah, and I was, I was... it's funny, Amir and I did not coordinate on this, but I wrote down one word on this question, as you were asking it Toni, which is “ethnography.” You know, this was the way in which people did culture work a long time ago because they would go as Amir said and observe and interact and ask questions and learn. And I think this whole idea of the old management, by walking around, has in many ways been lost. And in the post-pandemic world, where we're in hybrid modes of operating, and there's much more remote work, I think it's even harder to do. And I think, in many cases, managers have assumed, you know, we've been able to maintain high levels of productivity, things are still getting done, it's all fine. And I think we don't really understand the ways in which the cultures in many, many organizations these days are fraying. And it's partly through the walking around equivalent of walking around in a hybrid world that managers will get some intuition for it. And then if you pair that intuition with good, rigorous data, I think that's what's going to help managers figure out what to do about this.
Toni Dechario 1:05:47
I was struck by the fact that after the whole conversation we've had about technology, both of your recommendations were very analog, they were very human. Also, something that your research reminded me of is a Microsoft study that looked at how culture has changed over the course of the pandemic, and found that actually, silos became much stronger, and the links between them became weaker. Is that a permanent change? Are we seeing maybe, in Amir's words, “a return to closer community?” Or are we likely to go back to our more global ways?
Amir Goldberg 1:06:19
First of all, the thing that researchers hate doing the most is making predictions about the future. We're good at studying the past, rather than making predictions about the future. But I'm going to go on a limb and I'm going to say, we're definitely not going back to February 2020. That's evident already. And then we need to ask ourselves, what has changed? Actually, a lot of the trends that we were seeing before the pandemic, were simply exacerbated by the pandemic, because we suddenly discovered that we can interact over Zoom, and we can be productive. But also, the communities that we’re part of are starting to change. And there's greater sense of isolation and a mental-health kind of epidemic that was catalyzed as a consequence. My personal view is that the important thing that has happened through the Zoom-ification of work, catalyzed by the pandemic is that the boundary between the home and work has really, really been blurred. And that is a boundary that it took an industrial revolution to create. So, it was a huge accomplishment of the 19th Century to create this idea that, you know, there's a private self that you are at home, where you were maybe a parent or a spouse or a child, but then you go to the office and you occupy a role. And there's a lot of symbolic movement there, you send, you wear different clothes, and you also physically move to one space from one space to the other. And that has been vehemently and very dramatically disrupted during that pandemic. And I think the repercussions are going to be huge. And one of the biggest questions we need to ask ourselves is, if you think about it, a lot of the contemporary American’s life is defined by the organization that occupies – that employs them, I'm sorry – when you ask people, “What do you do?,” the first thing they say is who they work for. That's a huge part of who they are. That's a huge part of how they create relationships, especially amongst the upper middle classes, the educated knowledge kind of workers who tend to work for a firm that's thousands of miles away from where they grew up. So, I think all of that is going to change. And one of the biggest challenges that leaders are going to have is, what does it mean to have people whose sense of self is diminishingly related to who they are employed by? And we have also, the technologies now to intermediate work – kind of Uber-ification of work. That might mean that you could work for multiple organizations, and you don't have to be exclusively employed by any of them. I think those are huge, huge questions and huge changes that are going to emerge. So, if we need to make a prediction in the long run, I think that 100 or 200 years from now, we will look at 2020 as an inflection point in redefining the way that people understand themselves socially, insofar as the organization that employs them plays a role in that definition.
Toni Dechario 1:09:25
That's so interesting, because at the same time that you have this strengthening of subcultures and this kind of strengthening of small communities, small groups. You also have, you touched on this Amir, kind of hyper-individualization of our experience, and those would seem to be working at cross purposes.
Amir Goldberg 1:09:49
Yeah. But yeah, you know, our – we talked about human nature earlier, human-nature changes. If we think about human nature as something that's embedded in our DNA, that changes on a far longer scale and time horizon than our beliefs and our perceptions – and there's clearly some incongruity between the technological environments that we've created in our evolutionarily produced need for being co-present with other people – there's something that happens through sense, through motion, that we don't fully understand, that technology doesn't intermediate.
Sameer Srivastava 1:10:27
Going back to this idea of, of experimentation, and, of like fully understanding what works, what doesn't work, we don't know – as Amir was saying – what the future ought to look like, right? We know it's not February 2020. But it's probably also not just having Zoom calls with your immediate work team and being productive because people lose a sense of meaning and purpose if all they're doing is producing outputs, right? A big part of work life in most of the 20th Century was this was your social world. We would go and you might meet your spouse there, you might, you know, have close friends that you stay with, for the rest of your life that form there. And so, we have to figure out a way to reproduce some elements of the organization as a social world, but in this new reality. And what does that mean, in terms of how often are people coming back physically versus not? What are the ways in which they're connecting through a technology, when it's not just about having a meeting to make a decision or to get work done? When it's just socializing? What does that look like? And how do you make it meaningful? No one, as far as I know, has any good answers to that. Managers have, in some cases, very strong points of view – should be three days a week, should be two days a week – that are not informed by any data as far as I can tell. And this is another opportunity where we could actually learn by running some structured experiments. Let's see, what actually is the minimum level of in-person interaction needed to have meaningful experiences at work? And that's knowable.
Amir Goldberg 1:11:56
But wait, you said reproduce. And I think that's a big question, do we want to reproduce? Or do we want to imagine a new type of social world, and I think parts would be reproduced and parts are going to be new. And experimentation is happening all around us already. It's not systematized in the way that academics would want it to be necessarily, but firms are experimenting.
Sameer Srivastava 1:12:18
I would say what they're doing is piloting random ideas on strong priors they have. I wouldn't call it experimentation in the sense of a controlled experiment where you're learning, you know, A versus B. And there's, in principle – it's a, there's not much extra effort involved in going to that versus just trying this versus trying that.
Amir Goldberg 1:12:35
I think another way of saying what Sameer was saying is that they're doing it anyway, on the basis of their strong convictions that were formed probably in the 80s and the 90s, when they were coming of age, about what an organization ought to be like. And some of those are misperceptions. And some of those are socially constructed to use a term we introduced earlier beliefs about the nature of humans which isn't actually the nature of humans, it is the nature of the world that you were socialized into. And so, one of our biggest challenges is to persuade these leaders, that maybe their preconceptions are incorrect. And if they experiment carefully, that might lead them to find better ways of recreating congruence between the social experiences and the technologies that we have.
Toni Dechario 1:13:34
I think there are lots of reasons for resistance. But I think one of them actually goes back to an earlier conversation that we had, which is about the ethicality, right of surveilling, for instance. Don't you kind of run into some of the same questions with running experiments? I mean, I think that there's a discomfort and running experiments because of the implications to your staff to telling your staff that you're experimenting with their career. This is a serious question, right?
Amir Goldberg 1:14:03
You know, it came out – recently, there was a big article in The New York Times about research that our colleagues here at Stanford and MIT did through A/B experimentation on LinkedIn. You know, there was a big uproar. What, LinkedIn experiments with its users? Well, yes. So does the New York Times, by the way. We all accept, for example, at this stage, that if we are given a COVID shot, there was a randomized controlled trial that was done earlier, such that you are not just given a shot without having had it tested thoroughly beforehand through an experiment. And I think it's a cultural shift that we need to go through. Of course, managers don't need to just haphazardly experiment with their customers or with their employees. But there's a strong case to be made that ethical experimentation leads to better outcomes for everybody. Because employees will benefit. The organization will benefit. Society will benefit. It requires us to kind of reboot that conversation and to make people less afraid about the idea of experiments, but also to put in mechanisms to make sure that managers don't exploit that in a nefarious way. Universities have institutional review boards, because researchers in the 20s and 30s misused experimentation in really nefarious ways. Maybe that's something we should think about for-profit organizations as well.
Toni Dechario 1:15:43
There's also a kind of a vulnerability on the part of senior management to acknowledge that they have to experiment, that they don't have the answer.
Sameer Srivastava 1:15:50
Is it a vulnerability? Or is it a strength? Someone tells me – a senior leader – to find that they know the right mix of in-person versus remote. I don't trust them very much as a leader, right? If instead they say, “This is not knowable, and here's the way we're going to get to a better answer,” I had – my trust in that person actually just went up.
Amir Goldberg 1:16:07
And that's the cultural shift that we need. We need to encourage leaders to talk in that way and to show that successful leaders are actually aware of when are they making decisions that are based on strong evidence? And when are they just operating on the basis of gut? And when I have CEOs coming into my, my classroom, they often say, you know, students ask them, “How did you make this decision?” And they said, you know, “It was my intuition.” That's a very dissatisfying answer. What does it mean, “your intuition”? Can I learn something from your intuition? As an investor, I would be taken aback if a CEO will tell me I'm pursuing this strategy because it is my intuition.
Toni Dechario 1:16:51
Thank you both so much for doing this. Appreciate it.
Sameer Srivastava 1:16:55
Thanks so much for having us.
Toni Dechario 1:16:57
If you want to hear more from Sameer, watch the New York Fed's webinar on Shifting Norms: The Intersection between Technology and Culture in Financial Services, which you'll find at NewYorkFed.org/governance-and-culture-reform.
