[Week 38 of 2026] Partnerships

Welcome back to Price and Prejudice with a few musings from Week 38 of 2026. Today we talk about the partnership model in law firms and expert services more broadly.

Law firms are famously organized as partnerships. Why? Levin and Tadelis argue that when clients cannot judge advice, shared profits make partners selective about admitting colleagues who might lower average quality. Therefore, their willingness to share earnings helps reassure clients. A related argument is from Rebitzer and Taylor, who argue that knowledge of clients resides in individual lawyers, so giving those lawyers ownership makes staying more attractive than leaving with their clients.

Well, this Friday's WSJ article reports that law firms are hiring retirement coaches to persuade senior partners to leave. Equity partners own the firm and split what is left after salaries and expenses. For example, Wachtell's average profits per equity partner is about $12 million last year, which includes payment for the partner's work and the income from ownership. A partner also has a capital account recording her investment in the business, so the arrangement looks much like a part-owner of a restaurant who also cooks.

Other professions face versions of this problem. For example, in 2014, Deloitte defended retirement at 62 before Congress as succession planning, noting that partners remained well paid afterward. McKinsey asks consultants to advance or leave and limits its managing partner to two three-year terms. From what I can tell, tenured professors at universities are also basically like partners, although the "return" on their partnerships may not necessarily be in fungible money.

The article discusses a few ways that firms are trying to make the partners leave, but what else can they do? One version could be to have a voluntary buyout largest at age 62 and shrink each year until it expired at 68. That makes waiting costly, but requires cash upfront and may encourage productive partners to leave too soon. Another version could be to have a partner exchange her equity and her vote for a fixed income, paid for a set period by the practice group that inherits her clients. This would allow her to work on matters her successor assigns, although clients might continue treating her as the person in charge.

From the law firm's perspective, it's a bit tricky because in a world with AI that automates much of the tasks that employees can do, the partners will likely become more valuable. A law firm's revenue has two parts: the clients that partners bring in, and the hours that associates bill working on those clients' matters. AI eats into the second part first, i.e. document review, first drafts, and research that can now be done with fewer people. So what's left is the client's trust in a particular person and someone to hold accountable when the advice turns out to be wrong. And both of these sit with the partner.

One partial solution to the law firm's conundrum is the IPO market. It avoids the potentially costly process of negotiating each retirement one at a time, funding it out of the remaining partners' income, and arguing about what the claim is worth. Instead, a listed firm's departing partner simply sells shares to an outside investor at a price that already capitalizes the firm's future profits, and the firm's capital does not leave with her. This is roughly what happened to investment banks.

Generally, problems like these are fertile ground for financial innovation. The employee stock ownership plan was invented in 1956 so that the retiring owners of a California newspaper could sell the business to its employees rather than to an outside buyer. Wall Street partnerships went public one by one from 1970 onward, partly so that partners could take their capital out without weakening the firm. Unfortunately, not every version has worked – the physician practice-management companies of the 1990s bought thousands of medical practices from retiring doctors and mostly went bankrupt. But at least the pattern is the same every time: someone holds a valuable claim that cannot be sold, and eventually a structure appears that lets them sell part of it.

By tying a professional's income, her capital, and her clients to the same person, partnerships allowed the clients to trust the firm and the firm to trust the partner to stay. The cost of that solution is that leaving has no price. The pressure will likely continue as AI raises the value of the people who hold client relationships. Other professions eventually found a buyer for that claim, but law firms have not yet, and the retirement coaches are what fills the gap in the meantime.