The Goldman Career Path

When I started in investment banking in 1992, Goldman typically hired new college graduates as financial analysts. Financial analysts, who made a two-year commitment to the firm, did very junior work, usually supporting more-senior people. In their third year, excellent performers were sometimes offered an opportunity to stay for a year, and top performers might be invited to work in another group or geographic region of interest. However, to become an associate, the next step on the typical career path at Goldman, one had to leave, attend business school or law school, and then apply to return as an associate. Occasionally, financial analysts were directly promoted to associate, meaning that no graduate school or reapplication was necessary (that’s the path I was invited to take).

Associates were typically hired right out of top business schools or law schools, and most of them spent their summer vacations working at Goldman, thereby allowing the firm to make sure there was a cultural as well as a performance fit. Associates did financial analysis but also spent time with clients and thinking more broadly about what was required for the client. After four years (if they hadn’t been transitioned out of the firm or to another group), most associates were automatically promoted to the title of vice president (“executive director” outside the United States). Vice presidents spent most of their time with clients and reviewed work done by analysts and associates. Typically, VPs worked with partners on larger deals or with larger clients. One way Goldman maintained its culture was through an elaborate employee review process begun in the early 1990s.³ The anonymous 360-degree reviews were so extensive, even for analysts completing their first year, that they sometimes filled a binder the size of a phone book. As many as twenty people (managers and peers) contributed to the meticulous preparation, which included quantitative rankings, areas for comments, and also a self-review and statement of goals. The review meeting, conducted annually, was scheduled far enough in advance of decisions about compensation that it was not a discussion about compensation, but rather about performance and improvement.

When I gave reviews as a business unit manager, people typically asked where they ranked in their class. By sitting in on reviews and in training sessions about giving reviews, I had learned to maneuver the discussion to methods for improvement and not relative class rankings or compensation. My understanding was that employees were supposed to leave the review feeling positive but, at the same time, with some degree of insecurity, knowing that they had things to work on, though no one ever formally discussed with me the rationale for (and the consequences of) delivering this mix of positive and negative messages. In my experience, most people heard only the negative message, thus socializing the idea that they needed the firm more than the firm needed them. We also had less-formal six-month reviews, especially with more junior employees, to give timely feedback for improvement.

Reviews became more sophisticated, quantitative, and professionalized during my years at Goldman. When I started in 1992, I never heard of a policy or ritual to cull the bottom 5 to 10 percent of a class. The classes were very small, and there seemed to be little differentiation among the members. However, as classes became larger and some people were viewed as stronger than others, Goldman seemed to informally adopt the 5 to 10 percent policy (a policy often credited to Jack Welch, then CEO of General Electric). People who did not perform well enough were told in the annual review that they were in “the bottom quartile,” which at Goldman was interpreted as a professional way of saying one should start looking for a job elsewhere. At the time, Goldman tried not to fire people outright but instead to give messages tactfully through reviews and compensation so that people could find another job. The trick in many ways was to convince the people not in the bottom quartile, but in the middle of their class, that they were on track for partner so that they were motivated, stayed focused, and did not leave. Though nearly all professionals at Goldman thought they were on the partnership track, in the back of their minds, typically they knew the chances were slim. Nonetheless they stayed.⁴

Typically, it took four to six years as a vice president in investment banking to be considered seriously for partnership election. The election process served as a sort of “up or out” mechanism, because being up for partnership a few times and not making it was essentially a public message about one’s future with the firm. Being denied partnership too many times could damage one’s external market value, so getting turned down for partnership the first time presented a difficult decision. One rejection could be rationalized, but failure to make partner two or three times was far more difficult to explain away to a potential employer, and by then one’s compensation and responsibilities probably reflected diminished prospects at Goldman.

After 1986, when Goldman’s primary competitor, Morgan Stanley, went public, Goldman was the only comparable firm that was a partnership and still used the title “partner.” Partnership election happened every two years. Until 1996, when the firm changed to a limited liability corporation (LLC) and adopted the title managing director, the progression was from analyst to associate to vice president to partner. Unlike now—when titles have proliferated at Goldman and on the rest of Wall Street—the head of a department or specialty area in the early 1990s typically was a partner. That was it—only four titles (analyst, associate, vice president, and partner)—making Goldman a very flat organization.⁵ The cubicles where analysts and associates sat were right next to the partners’ offices, and they worked directly with the partners and shared assistants. In fact, some partners tried to sit in cubicles to make themselves more accessible and approachable. They abandoned this approach because the partners had so many meetings with so many people it distracted the junior people around them. Also, the junior people who sat next to the partners felt their personal lives and comings and goings were becoming too transparent. One banking partner instituted trading desks instead of cubicles in his department to facilitate communication. This approach didn’t work, either, for much the same reasons.

Being a partner at Goldman was one of the most sought-after positions on Wall Street in part because of partners’ prestige and because Goldman partners generally were the highest-paid people on Wall Street. A Goldman partnership was “about as close to a sure thing as there is anywhere in the business world—a ticket to wealth and prestige.”⁶ Partners tended to manage departments or work with the most important clients and had responsibilities other than working on deals. For example, in the M&A department several partners would have to agree to sign something called a fairness opinion, a document issued by Goldman to a public company’s board of directors stating that the M&A transaction in question was “fair from a financial point of view” to shareholders. This “opinion” gave the board members confidence in the deal and was thought to protect them if they were to be sued or their judgment was questioned by shareholders.

Issuing an opinion was risky, because it could be challenged in court and Goldman could be subjected to legal and financial liability. In fact, Goldman’s IPO prospectus stated, “Our exposure to legal liability is significant,” citing “potential liability for the ‘fairness opinions’ and other advice we provide to participants in corporate transactions.” Before the IPO, the partners personally shared this liability. (After converting to a LLC and the IPO, partners did not have personal liability in the event of a successful lawsuit; any financial consequences would be borne by the firm and not the private partnership.) So partners in the trading business had to trust that the M&A partners were being very careful and thoughtful in issuing fairness opinions.

The Partnership Prize

Compensation can certainly send a message, but typically it is a private message to the employee from the company. Partnership, however, is a public recognition of the value of one’s contribution and ability to uphold the principles. It worked together with the culture to motivate people to accept lower compensation and work grueling hours. In addition, it served as a mechanism related to financial interdependence.

In a 2009 op-ed letter about the debate over bonuses and Wall Street pay, Peter Weinberg, son of Jimmy Weinberg and nephew of John L. Weinberg, described Goldman partnership and financial interdependence this way:

The only private partnership I can talk about authoritatively is the one in which I was a partner from 1992 to 1999, when the firm went public: Goldman Sachs. Partners there owned the equity of the firm. When elected a partner, you were required to make a cash investment into the firm that was large enough to be material to your net worth. Each partner had a percentage ownership of the earnings every year, but the earnings would remain in the firm. A partner’s annual cash compensation amounted only to a small salary and a modest cash return on his or her capital account. A partner was not allowed to withdraw any capital from the firm until retirement, at which time typically 75%–80% of one’s net worth was still in the firm. Even then, a retired (“limited”) partner could only withdraw his or her capital over a three-year period. Finally, and perhaps most importantly, all partners had personal liability for the exposure of the firm, right down to their homes and cars. The focus on risk was intense, and wealth creation was more like a career bonus rather than a series of annual bonuses.⁷

The partners knew an elected partner’s actions and performance would reflect on them, and if they made an unwise choice, that could have serious financial and reputational consequences for them all. Partners also had to believe that transferring a percentage of profits from themselves to incoming partners would be advantageous—that the potential dilution of their own wealth and the added risk were worth the potential gain.

Consequently, the Goldman partnership process was carried out in painstaking detail, and that was generally understood throughout the firm. Partnership election was taken seriously, as shown by the personal involvement of senior executives. The partners judging a candidate’s merit and suitability were the very people with whom the candidate would be financially interdependent if granted partnership. Written nominations were solicited from the partners; from those nominations, the partnership committee created a preliminary list of potential candidates.⁸

During the months preceding partnership elections, it was difficult to get time with the partners because they had so many phone calls and meetings about candidates. Goldman held its partnership elections every two years rather than annually, in part because the process was so arduous and time-consuming.

Although the process was shrouded in secrecy outside Goldman and discussed in hushed voices or behind closed doors, some people knew what was going on from off-the-record conversations with partners with whom they were close. The conversations were supposed to be confidential, but rumors were quietly passed on. People knew who was a “lay-up” (a basketball term for an easy score), who was “on the bubble” (not sure to be elected), and who did not have “a prayer.”

Written nominations were solicited from Goldman’s partners, and from those nominations the partnership committee created a preliminary list of potential candidates, kicking off a paper chase. Endorsement letters were filed, and internal former FBI and CIA employees conducted background checks on all serious contenders. An internal investigation (called “cross-ruffing,” a term borrowed from the card game bridge) took place, led by a partner not in the candidate’s division. Then the list was narrowed during the meetings to discuss the would-be partners, a photo of the candidate was displayed on a screen, and from this list, a committee selected names to be placed before the partnership for a vote. The partners generally wanted to keep the percentage of partners to total employees about the same, so the number elected would be carefully chosen depending on retirements and growth.

The process might seem coldhearted, but it was viewed as essential to maintaining the culture, and everyone knew it.⁹

The behavior you needed to exhibit to make partner was clear: make money for the firm while embodying its principles. Everyone knew that becoming a partner required the support of people outside your division and region. Someone in banking needed support from asset management or trading. Someone in North America needed support from people in Asia. This was a key tenet of teamwork; to make partner, you had better help people across the firm. This meant that if you were asked to be on a 4:00 a.m. call with a Japanese client, even if it had nothing to do with your clients, you would set the alarm and do it without question, knowing that if you did not rise to the occasion, it could hurt your chances of becoming a partner (or staying a partner). If you were asked to bring someone from another area into a project because it could help improve the firm’s advice, then of course he or she was invited. And the new person participated even though the revenues of the project might not be attributed to you or your group. Candidates’ behavior was being vetted as much as their ability to contribute financially to the partnership.¹⁰

I was surprised when very senior partners asked—in a roundabout, casual way—my opinion of vice presidents who were widely known as being considered for partnership, and as I got more senior they privately told me why someone had or had not been elected—with the implied messages such information carried. Once a partner showed me the standard form used to evaluate candidates. The blank form included numerical rankings of listed qualities related to the Goldman principles.

Former Goldman vice chairman Rob Kaplan says that many of his most stressful and difficult moments at Goldman occurred during partner promotions, particularly when he had to deliver the news to a highly qualified person that he or she would have to wait another two years for reconsideration. It was little wonder that Kaplan writes in What to Ask the Person in the Mirror that this message of failure to make partner was often met with “expressions of betrayal.”¹¹

One person I interviewed—a managing director (now at another firm) who was passed over for partnership at Goldman in the 1990s before the IPO—said that of course he was furious. He had done everything he could—he had given up weekends, birthdays, anniversaries, and vacation time with his family—and then had not been elected in spite of what he thought were positive indications (assurances that he felt were essentially promises) that he would be. Not making partner was devastating to him, he said. He told me that, for people at Goldman, who you are, and what you think about yourself, was ultimately decided at that moment. The thing he feared most was the public embarrassment and humiliation, even more so than the loss of potential riches. When he failed to receive the customary, congratulatory expected call from the senior partner of the firm on the appointed morning, he explained, he felt betrayed. He did not want to hear that he might very well become a partner the next time, in two years. He felt that people wanted him to stick around so that they could get two more years of relatively cheap labor from him and that they would say anything, just as they had assured him before, to keep him. But he was determined not to stick around while his peers from business school were called managing directors at other firms and he was called a vice president for another two years. His staying would be “a cheap option” for Goldman, so he swallowed his pride and successfully cashed in his pedigree by leaping to another firm.

In a 1993 interview, Steve Friedman, then a senior partner, explained that the partnership election process was used as a way to convey that people would be rewarded for doing what was best for Goldman and would be denied the ultimate reward—partnership—if they paid more attention to their own agenda than to the firm’s. Friedman described delivering that message to one disappointed banker: “I have looked people in the eyes and said, ‘You did not become a partner this time despite your basic abilities, your candle power, your energy. You had all the goods to have achieved it, but you did not become a partner because you were perceived as having too damned much of your own agenda, and you were ignoring what we were telling you was in the broader interests of the firm.’”¹²

The flip side of the partner election was that, to add partners, Goldman had to discreetly ask others to retire, for the simple reason that maintaining a greater number of partners would mean each would own a smaller percentage of the firm.¹³ If someone was departnered, it was only after careful deliberation by the most senior partners and usually resulted from making too small a contribution relative to ownership (Goldman could get someone else to do it for less) or from doing something that jeopardized the firm’s reputation and put the partners at financial risk. Thus, the pressure was on for performance and proper behavior, even for partners, and the pressure was intense.

Generally, there have consistently been more partner additions than departures, but the ratio of partners to the total number of employees has stayed relatively constant because of continued growth. For example, the total number of partners noticeably increased from 1984 to 2011, which corresponded to the increase of total employees: both nearly doubled from 1998 to 2011. However, the percentage of partners to employees (the partner ratio) remained steady at 1.5 percent to 2.0 percent. Growth was important to provide more opportunities for partnership and for the partnership to be financially attractive.

A retired partner I interviewed said that he never worked harder at Goldman than when he was a partner. He had partner responsibilities and obligations on top of his regular job. He explained that he had done everything by the book and was relieved when he made partner.

“Why were you relieved—was it the money?” I asked.

He explained, when he got to that point, it was less about the money than it was because everyone knew he was being considered for partnership—his wife, his clients, his business school classmates, everyone at the firm. He retired after the IPO, despite having been a partner for only a short time, because he was burned out, even though he knew that each year he could “hold on” represented millions of dollars. The pressure was taking a toll on his health and his family, he told me. Today he still values the social status of having been a partner, and that is how he is usually introduced in social contexts: “a retired pre-IPO partner of Goldman Sachs.”

No one was exempt from performing or from upholding the firm’s values. The departnering conversations were held discreetly and privately with the senior partner of the firm, but when the internal memo came out, there were almost always rumors, typically related to performance. The former partner hadn’t pulled his weight or wasn’t doing the expected culture-carrying tasks, such as recruiting. Or the outcast had done something harmful to the Goldman image, such as having an extramarital affair with someone at the firm or saying something inappropriate, or had subjected the partnership to unnecessary financial risk. Even if a partner left voluntarily for primarily personal reasons, there was almost always speculation about the “real” reasons.

A current partner told me that the organization looks for an explanation beyond the desire to leave because he or she may simply feel he or she has made enough money. That’s not acceptable, because it might send a message that money is the primary driver. If the reason for leaving is that the partner no longer enjoys the work, then people would wonder what there is about Goldman not to like, this partner told me. The firm convinces people that being a Goldman partner is something one would never want to surrender; it gives one social identification, prestige, money, and access, and it is perceived as serving a higher good. That is what is sold to potential and current employees. So if a partner leaves, something has to be wrong with him, and the firm perpetuates that belief through whispers. The partner explained that the only acceptable answer is that the partner is retiring to serve a higher purpose, which is to go into community and public service—and many do. He said it wasn’t the money, it was that their work ultimately needed a higher meaning.

Historically, Goldman’s process of partner election and departnering are exemplary of what sociologists term closure, the tight coordination within a group, which ensures that people comply with the organization’s norms.¹⁴ According to sociologist Ronald Burt, “closure increases the odds of a person being caught and punished for displaying belief or behavior inconsistent with the preferences in the closed network.”¹⁵ In his view, closure strengthens organizations by ensuring that people not adhering to expected norms can be removed.¹⁶ Making partner was so lucrative and the identity meant so much to people that they modified their behavior to enhance their chances of being elected, and, once they were partners, it became an integral part of their social identity.¹⁷ Rewarded behavior helped the firm as a whole. Partners worked hard to make more money but also were pressured to promote teamwork, the culture and the principles, and to stay within the firm’s rules and values.

Meanwhile, the close scrutiny of each partner’s contributions and adherence to the mandate during the partnership election provided closure by removing partners whose continued tenure was not to the advantage of the firm, thus ensuring trust among those who remained. In this way, partnership election and departnering reinforced the distinction between insiders and outsiders. The effectiveness of this practice is best seen, according to Charles Ellis, the author of The Partnership—The Making of Goldman Sachs, in the “speed and clarity with which long-serving partners who left went from being insiders to outsiders and were soon forgotten.”¹⁸

A Social Network of Trust

While the partnership election at Goldman was grueling, partnership offered camaraderie to those who made it.¹⁹ A retired partner explained to me that there were regularly heated disagreements among partners over business decisions. He felt the partners were in fact like a family or club. As in many families, the partnership was “dysfunctional” and had “black sheep”; there might be questions of motives or agenda; and sometimes there were even sharp elbows or personality mismatches, but there was a good deal of trust and familiarity among the members.

Most partners had spent their entire careers at the firm, and many senior partners had mentored the newer partners over years. They had gone through the same election process and knew how tough it was to make partner and how demanding the job was. Partners had outings together as well as annual dinners. Many lived in the same suburbs. Generally, they knew each other well. Some ties were stronger than others because of business school friendships, experience in working in the same department or the same part of the world, or location in the same neighborhoods. However, all of them were financially interdependent and needed to trust one another. Several partners agreed when I interviewed them that the phrase “social network of trust” was a fitting description.²⁰ In 1994, a partner, addressing a gathering of new partners, highlighted the relationship between the partnership structure and the cultural view it promoted: “We own this business … We are partners—emotionally, psychologically, and financially. There can be no borders between us, no secrets.”²¹

Partners also created stronger networks because of the trust. Building a network involves connecting the dots, or rather connecting Goldman people to each other and to important people outside the firm. Almost a decade before he crafted the firm’s business principles, John Whitehead wrote a set of guidelines for the investment banking services area, one of which was, “Important people like to deal with important people. Are you one?”²² Gaining access to important people requires introductions from people willing to make the call. Partners were more willing to do that for the partners who were in their social network of trust and with whom they had financial interdependency.

For example, an investment banking partner explained to me that he was at ease connecting fellow partners to his contacts outside the firm, more so than connecting a vice president or regular managing director. Partners, he said, wouldn’t leave the company and take the relationship to a competitor. He trusted his partners in private banking to use good judgment and not to put his investment banking relationship at risk. This kind of trust made it easy to offer multiple perspectives to help a client. It also aided the firm in cross-selling: providing a full solution to clients. For example, Goldman might provide the merger advice to sell a company, and then the M&A partner would introduce a Goldman private banking partner to help the client manage the proceeds from the sale.²³ “Cross-selling” significantly improves a firm’s financial performance, maximizing revenue opportunities.

A virtuous reinforcement loop was erected; partnership was enhanced by the trust fostered by the social network and financial interdependence, and the social network was enhanced by the trust.²⁴

Productive Dissonance

The emphasis on shared values in the Goldman partnership model might lead to the impression that the Goldman culture was rigid, monolithic, and intolerant of diverse opinions. Generally this was not the case. Although the partners held common values and many of them came from similar backgrounds, Goldman recognized that different people had different ideas and perspectives and that a diversity of people and ideas was important to the firm’s vitality and productivity.²⁵ Diversity was also important in its making the right decisions and giving clients the best service and judgment. Such diversity of experiences and expertise gave Goldman the flexibility to deal with constant change.

Goldman promoted cross-function communication and organizational cohesion through rotational programs for employees into other departments and regions and firmwide committees consisting of people from different departments.²⁶ This “small-world network” of people who built ties, had financial interdependence, and trusted one another led to innovation and high performance. Committees drew together partners or potential partners from different areas to work together on partnership election, capital commitments, risk, culture, lifestyle, brand, and more. These committees, together with partnership meetings, served as places of exchange like those Ronald Burt describes as essential for optimizing the number of brokerage opportunities for networks.²⁷ This practice created valuable networks, as well as a systematic, structured way to bring people together to discuss ideas, challenge each other, and seek solutions.²⁸ Goldman’s flat organizational structure also encouraged people to interact, bringing their diverse opinions to the table.²⁹ The biennial change in partnership, with a balance of new partners joining and old partners leaving, kept the ideas fresh, but generally it did not introduce so many differences that cohesion was lost.³⁰

Goldman’s partnership culture allowed for disagreement in part because the partners have a stake in more than only their own areas of responsibility. They were financially interdependent. Banking partners had nothing to do with trading and vice versa, but trading partners were affected if a banking partner hurt the reputation of the firm, and banking partners were affected if trading partners didn’t properly manage risk, because they are risking the capital of all the partners. In a typical big bank, by contrast—one without a partnership ownership structure—compensation is based primarily on departmental and individual performance (and peer group compensation averages); what others do or think in another department or divison is typically of little concern to anyone else, because one’s own bonus is not materially threatened. The lack of financial interdependence typically limits the amount of productive disagreement you find at most Wall Street firms. Even with compensation in stock vesting over years, the attitude is much different because what one does typically has limited impact on the earnings or stock price of the entire firm.

The disagreement at Goldman was described by several Goldman partners as valuable.³¹ Rob Kaplan indicated that “irritating, distracting, and uncomfortable” discussions were “extremely good medicine for a healthy organization.”³² For example, one seemingly contentious relationship was that between senior executive John F. W. Rogers, considered by many to be one of the firm’s most powerful people, and Lucas van Praag, Goldman’s public relations spokesman from 2000 to 2012. Jack Martin, CEO of the PR firm Hill and Knowlton, said that he had seen the two “disagree aggressively, but that at the end of the day they [came] together as one.”³³ Although others may see friction between Rogers and van Praag as unhealthy, Blankfein expressed confidence in them and asserted that “dissent and disagreement is healthy.”³⁴

Whitehead described how he and his co-leader, John L. Weinberg, strove to maintain an environment conducive to productive disagreement within the management committee: “We met every Monday morning. We had an agenda, we went down the agenda, we made decisions as a group, and John and I tried not to dominate the committee because we wanted their input. It was very important to have the input of everybody on the management committee. There were seven of us, then nine of us, and then eleven. It got bigger as the firm’s diversity grew.”³⁵

A Harvard Business School case study about Goldman’s training found that Goldman used the Socratic method to explore questions through discussion and debate. One senior manager described a typical discussion: “That was a good argument. But next time, it would actually be really interesting if you also added these three things.”³⁶ One partner expressed his delight in this learning environment: “Goldman Sachs when I started was a fantastic place to be planted and grow. They treated me the right way, encouraged me the right way … It’s a Socratic, collaborative style. Bouncing things off of each other is fun, and you encourage that at every turn.”³⁷

According to a former management committee member, the dialogue was predicated on collaboration: “We were always taught that the odds are high you’ll have better outcomes with a shared work effort than with that of a single individual. At Goldman Sachs it’s pretty rare that an additional perspective doesn’t give you a better outcome. As problems become more complex, the ability for a single individual to have the best perspective declines dramatically.”³⁸

David Stark, a Columbia University sociologist, argues that dissonance, or friction, over competing values promotes an organizational reflexivity that makes it easier for a company to change and deal with market uncertainty.³⁹ Dissonance prevents groupthink: what happens psychologically when a group is so concerned with maintaining unanimity in the face of opposition that alternatives and options fail to be identified or properly evaluated.⁴⁰ Stark notes that dissonance can become resonance, a “dangerous form of cognitive interdependence,” when too many people overlook a key issue, giving “misplaced assurance” to those who think similarly.⁴¹ A partnership structure seems to mitigate against this phenomenon in part because of the financial interdependence and social network of trust among partners. One researcher explained this effect well in describing the era of the private investment banking partnerships as one in which “no one could take excessive risk with the firm’s capital, because of the vigilance of the partners. If Partner A wasn’t comfortable with a new business or a new client, Partner B would have to convince him of the merits of the business—and to do so, partner B would have to go well beyond the argument that it would generate a lot of short-term fees.”⁴²

Stark points out that the “ability to keep multiple evaluative principles in play and to exploit the resulting friction of their interplay” is a competitive advantage. Organizations typically try to avoid the perplexing situations that arise “when there is principled disagreement about what counts,” when, in fact, they should embrace such situations and “recognize that it is legitimate to articulate alternative conceptions of what is valuable, what is worthy, what counts.”⁴³ Partners I interviewed generally agreed that this is a good description of the productive dissonance that characterized Goldman’s partnership culture model.

Goldman excelled at adapting to change and dealing with market uncertainty not so much because of specific individuals but because the organization’s partnership structure and culture sustained ongoing and productive discussion and disagreement. Stark notes that when faced with uncertainty, “instead of concentrating their resources for strategic planning among a narrow set of senior executives or delegating that function to a specialized department, heterarchical firms [flat organizations or those with a limited hierarchy] embark on radical decentralization in which every unit becomes engaged in innovation.”⁴⁴ Therefore, hierarchical firms—those with a long chain of command—are less supportive of innovation and entrepreneurship. Dissonance—supported by the flat organizational structure that facilitated interaction and information transfer, and the financial interdependence of partnership that fostered trust and aligned interests—gave Goldman a significant competitive advantage. Based on my interviews, dissonance of this degree did not exist at most Wall Street firms, because there was no financial interdependence, earnings were distributed, or the partnership was dominated by a few individuals, decreasing the financial interdependence. Goldman’s culture of debate took on special significance during the credit crisis, as discussed in a later chapter.

If dissonance encourages entrepreneurship, innovation, or flexibility in meeting competitive pressure, why was Goldman in the 1980s not considered innovative? For the most part, Goldman had a long history of watching and waiting while others did the innovating; then it moved in with an improved or enhanced version. Whitehead said, “[As] far as new products were concerned, I never felt that we had to be first when we did something. We had a reputation for being absolutely first-class in everything we did, but we didn’t have to be first with every idea. In fact, I enjoyed it when our competitors had the new idea and tried it out first.”⁴⁵

Whitehead pointed out to me another reason Goldman was not the first to innovate: constrained capital limited what the firm was willing and able to do, something that changed as it first began raising more outside capital and then accelerated when it went public. It took bigger financial risks and grew more quickly. When I discussed innovation with partners and competitors, most said they did not see Goldman as an innovator in the 1980s and early 1990s, but they thought that dissonance added to the firm’s superior financial and competitive performance and allowed it to get to the best answers for the firm and for clients.

When I asked Whitehead about the perception of Goldman in the 1980s of not being an innovator, he pointed out a nuance that many people were missing: Goldman’s strength was not a matter of investing in and potentially taking new risks by developing new products per se. Instead, its strength lay in the ability for people to come together as an organization to figure out how to change and adapt and market products that were better for clients, no matter who created them. To Whitehead, that practice was as innovative as developing products or investing in innovations that might not work or benefit clients—but much harder to execute because of the barriers to getting people to work together, and a lot less risky.

Another significant innovation in the late 1980s and early 1990s was the heavy use of the emerging technology of voicemail, adopted much earlier and used more effectively at Goldman than at other firms. Voicemail helped improve coordination and teamwork, because multiple people could be given information simultaneously and they could hear the tone of a message. The technology resulted in better execution for clients and gave Goldman a competitive advantage. Goldman continued to rely on voicemail heavily even after it had e-mail capability, because e-mail lacked the inflection and expressive capacity of the human voice and was less effective in maintaining the firm’s social network.

Goldman people responded quickly to voicemail (and later e-mail messages) because the culture demanded that they do so. A quick and comprehensive response was and still is the norm; it would be culturally unacceptable not to respond as soon as possible. No matter the hour, you can get a partner or junior person on the phone to help you think about a problem or speak to a client. That was true when I was at Goldman, and my interviews confirm it is still the expectation.

Many people from other firms were (and are) shocked at the speed and amount of information shared, often mentioning their need to adjust to Goldman’s urgent voicemail (e-mail) culture.

Financial Interdependence and Risk Management

As a private partnership, if Goldman suffered large trading losses, lost a lawsuit, or received fines for criminal or illegal practices, then the partners were personally liable. Their equity was at risk every day. Even with the protection afforded personal assets by the LLC structure, which Goldman changed to in 1996, partners still faced the risk of losing their invested capital, which for many represented the bulk of their wealth.⁴⁶

When one’s own money is at stake, management of risk, both financial and reputational, is a key concern.⁴⁷ According to the partners I interviewed, managing risk for one’s entire net worth, and that of all the partners, brings a higher level of intensity than managing risk for shareholders and a longer-term orientation than today’s typical three-to-five-year equity vesting period. That is probably why, in Goldman’s own accounts of its culture, risk management often is not included: it was so obviously a no-brainer that it was assumed.⁴⁸ As mentioned earlier, Peter Weinberg stressed this issue in his Wall Street Journal op-ed, “right down to their homes and cars,” noting that “[t]he focus on risk was intense.”⁴⁹

The firm’s culture and principles, combined with the partners’ financial interdependence and the risk to their own capital, ensured that collective risk management and reputational risk management would be a priority. The structure of Goldman’s pre-IPO partnership resulted in financial interdependence and a social network of trust, virtually ensuring teamwork and dissonance; in turn, the business practices, policies, and values supported it.