Watch Steve Mandis, author and Goldman Sachs insider, discuss his latest book, What Happened at Goldman Sachs, as well as other great videos from our complimentary Hot Off The Press Series!

William C. Dudley, the president of the Federal Reserve Bank of New York, gave a speech Monday in which he used the word “culture” 45 times. Here’s how he defined it:

Culture relates to the implicit norms that guide behavior in the absence of regulations or compliance rules—and sometimes despite those explicit restraints. … Culture reflects the prevailing attitudes and behaviors within a firm.  It is how people react not only to black and white, but to all of the shades of grey. Like a gentle breeze, culture may be hard to see, but you can feel it. Culture relates to what “should” I do, and not to what “can” I do.

Dudley has a doctorate in economics, and spent a decade as chief economist at Goldman Sachs. But in his remarks he sounded more like a sociologist than an economist. His many mentions of “culture” could be significant. I’m hoping they mark the beginning of a change in how regulators think about reining in law-breaking and excessive risk-taking at banks. I’m also hoping that I had something to do with them.

I studied economics too, as an undergraduate. Then I went to work at Goldman Sachs, in mergers and acquisitions first, proprietary trading later. I spent a dozen years there — a tenure that overlapped with Dudley’s — went on to work at several other firms, and then, after the financial crisis, decided to go back to school. I now have a Ph.D. in sociology from Columbia University and teach at Columbia Business School.

Many people find it peculiar that a former proprietary trader with a background in economics would go back to school and study sociology. As I reflected upon my career at Goldman Sachs, though, what stood out was the importance of its organizational structure. That’s something sociologists pay a lot of attention to, while economists generally don’t.

So I studied sociology, and for my doctoral dissertation focused on the organizational culture of Goldman Sachs. The dissertation became a book, titled What Happened to Goldman Sachs: An Insider’s Story of Organizational Drift and Its Unintended Consequences (HBR Press, 2013). One of the changes I document in the book is how Goldman drifted from a focus on ethical standards of behavior to legal ones — from what one “should” do to what one “can” do.

After the book was published, Dudley got in touch. I met with him and his people, and discussed what I had learned in my study of sociology and, in particular, my in-depth study of Goldman. I made recommendations on how to improve regulation. Also, I sent him two pieces I wrote for HBR.org, one on the importance of focusing on organizational behavior and not just individuals, the other asserting that culture had more to do with the financial crisis than leverage ratios did.

One of the key conclusions I drew from my study was that to achieve sustained success and avoid firm-endangering risks, a firm like Goldman has to cultivate financial interdependence among its top employees. I wrote:

Financial interdependence is important as a self-regulator … leaders should disproportionately and jointly share in fines, settlements, and other negative consequences out of their compensation plan or their stock … meaningful restrictions on leaders’ ability to sell or hedge shares should be imposed, which can lead to better self-regulating and longer-term thinking.

Consider what actually happened at JP Morgan Chase after the gigantic “London whale” trading loss. The company’s board docked the pay of CEO Jamie Dimon by more than half, to $11.5 million from $23 million. The bank also went after the bonuses of the individuals involved. That’s something, but the bulk of the loss was of course borne by shareholders. And what happened to the compensation of the typical JP Morgan managing director? According to people that I interviewed, not much (other than losses on their JP Morgan stock, which in most cases represents only a fraction of their overall net worth). The main reason, I was told, is that JP Morgan must pay competitively or lose its most talented employees. The second reason was that most managing directors had nothing directly to do with the losses.

But these people were important parts of an organization that messed up. Plus, they’d gotten big bonuses in previous years based in part on profits they had nothing to do with. One banker that I interviewed suggested that if JP Morgan’s managing directors collectively had to pay a large portion of settlements or losses related to the misbehavior out of their bonus pool, perhaps they as a group would take stronger internal actions to prevent such behavior, reward those who acted responsibly, and kick out those who did not. Maybe they would hold their leaders to higher standards and question each other’s activities.

This in fact is how things generally worked at Goldman Sachs and other Wall Street firms back when they were partnerships instead of publicly traded corporations. Each managing director was financially interdependent with every other. Typically, each received a fixed percentage of the overall annual bonus pool and was personally liable for other managing directors’ actions. At Goldman there was the added restriction that partners could not pull out their capital until after they retired (a far cry from the three-to-five-year vesting that bankers complain about today). The organizational regulation created by this structure was key to managing risk, and we should be thinking about ways to bring it back.

I am not suggesting the banks return to being private partnerships. But they should move away from today’s norm of discretionary annual bonuses for managing directors to, at least for a select group of top employees (at Goldman the elected “partner-managing directors” represent around 1.5 to 2% of total employees), a shared bonus pool with fixed percentages that would pay a large portion of settlements or losses related to misbehavior and have greater restrictions on selling stock. Managing directors would share in the firm’s successes, but also feel it when others incurred losses or when the firm got hit with fines. Giving bankers reason to hold each other accountable would cause them to pay much more attention to asking questions and managing risk and misbehavior. Restricting stock sales could push their thinking and actions in a more long-term direction. 

In his speech Monday, New York Fed President Dudley urged some moves in this direction. Much of the compensation for high-level bank employees should be deferred for years, he said: “This would create a strong incentive for individuals to monitor the actions of their colleagues, and to call attention to any issues.” And when a bank is hit with a big fine, he argued — and this is something I don’t recall hearing before from a U.S. financial regulator — that some of the money be taken out of that deferred compensation pool:

Today, when a financial firm is assessed a large fine it is paid by the shareholders of the firm.  Although senior management may own equity in the firm, their combined ownership share is likely small, and so management bears only a small fraction of the fine. … Assume instead that a sizeable portion of the fine is now paid for out of the firm’s deferred … compensation, with only the remaining balance paid for by shareholders.  In other words, in the case of a large fine, the senior management and the material risk-takers would forfeit its performance bond.

This kind of interdependence has the potential to move the focus back to ethical standards of behavior instead of just legal ones. It might also drive away some talented employees. But if these people can’t take a longer-term approach and trust one another, should we trust them — and should they really be working at systemically important banks?

This blog first appeared on Harvard Business Review on 10/24/2014.

View our complete listing of Strategic HR and Compensation and Benefits blogs.