More gruel
The Interest-Convergence Dilemma: How IBM Got Woke

The Interest-Convergence Dilemma: How IBM Got Woke

While putting together this series on race and mythology I stumbled across “Policy Letter #4,” written by IBM President Tom Watson, Jr in 1953:

Each of the citizens of this country has an equal right to live and work in America. It is the policy of this organization to hire people who have the personality, talent and background necessary to fill a given job, regardless of race, color or creed. If everyone in IBM who hires a new employee will observe this rule, the corporation will obtain the type of people it requires, and at the same time we will be affording an equal opportunity to all in accordance with American tradition.

I am fascinated.

This is before Brown v. Board of Education. Before the Montgomery Bus Boycott. Before anyone had heard of Martin Luther King, Jr.

Watson’s policy directive was aimed less at his own managers than at the segregationist governors of Kentucky and North Carolina. IBM wanted to expand its manufacturing operations into the South, but it could not allow Jim Crow to undermine the company’s priorities. Those facilities would open in the late 50’s with white and black employees side by side, almost a decade ahead of any federal requirement.

In the 1870’s railroad barons teamed up with Southern planters to pull the plug on Reconstruction. Now, corporations compete to offer the friendliest environments for employees from marginalized groups, hosting gay pride events and Black History Month lectures. Getting caught wearing a MAGA hat is a good way to get fired from almost any good-paying job. Being a Republican is something you’d want to keep very quiet if you have serious career ambitions.

What happened? Are corporations nice now?

IBM’s stance on racial equity wasn’t just posturing. In 1958 they abolished the hourly wage in an effort to bridge the divide between their white collar and manufacturing workforces. When he became CEO, Tom Watson Jr. would look for ways to bridge a growing divide in compensation between management and front-line workers with innovations in profit sharing and employee stock ownership. He wrote:

I thought that the model corporation of the future should be largely owned by the people who work for it, not by banks or mutual funds or shareholders who have inherited the stock from their parents and done nothing to earn it.

In the 60’s the company would step up its investments in racial equity, working with Robert Kennedy on a project to open IBM facilities in the Bed-Stuy neighborhood of Brooklyn and backing Kennedy’s redevelopment agenda. While Dixiecrat Senators were battling to halt the Civil Rights Act in 1964, its author, Assistant Attorney General and Head of the Civil Rights Division, Burke Marshall, was fighting behind the scenes to win votes. After he won, IBM would make Marshall its general counsel.

We have an instinct to place halos on our heroes and horns on our villains. Do not make this mistake. Neither Thomas Watson, Sr nor Jr. were ideological crusaders, but relentless advocates for their company’s interests. Don’t look for saints. Welcome allies, while understanding the landscape of their interests.

When Tom Jr wrote Policy Letter #4, he was still serving under the leadership of his dad, CEO Tom Watson, Sr. Two decades earlier, when IBM’s goals required collaboration with the Nazis, Tom Sr. didn’t hesitate. IBM’s punch-card tabulators, developed first for the US Census, then for the Social Security Administration, became crucial to the Nazi effort to round up Jews. IBM’s German subsidiary was its second-most profitable unit during the Depression.

When the Nazis insisted they could only do business with IBM if it removed Jews from senior management, the company complied, at least in Europe. Tom Watson Sr. met Hitler in 1937, where he was presented with The Order of the German Eagle. He returned the medal in 1940. There’s no evidence that CEO Tom Sr. wrung his hands over Nazi anti-Semitism in the 1930’s, or that he resisted Tom Jr.’s move toward desegregation in the 1950’s. Neither move was about ideology. This is a very important factor to understand

Illinois Republican Senator Everett Dirksen took the rostrum in the final day of the great 1964 Dixiecrat filibuster and quoted Victor Hugo, “Stronger than all the armies is an idea whose time has come.” He could have put it another way: Mythologies follow power.

Why did white supremacy fail? A tide “stronger than all the armies” was shifting against its foundations. Professor Derrick Bell in 1980 described what he called the “interest-convergence dilemma” in civil rights law: “The interest of blacks in achieving racial equality will be accommodated only when it converges with the interests of whites.” This may not be an iron law, but it is wise in politics to recognize when their interests align with powerful forces and use that alignment to their advantage.

Myths follow power. The rise of knowledge capitalism shifted the landscape of power away from capital investments in extractive businesses toward capital investments in brainpower. Have you noticed that we’re seeing a striking pivot in public life toward expressions of racial and gender equity without much movement on questions of class equity? We have same-sex marriage, a Black former President, and a disastrous health care system that no one is close to fixing. The interest convergence dilemma provides an explanation. In the Woke Era, progressive ideals can advance right up to the point where they begin to inconvenience educated, affluent people. We have relative progress on racial equity because that progress benefits powerful people. Recognize this pattern of interests and it can become a lever toward real progress.

Look closely at the language of Tom Jr’s policy statement. What goal does it identify? “The company will obtain the type of people it requires.” What does that mean?

How many cotton planters ever used the word “talent” to describe their business needs? What about coal barons? Steel companies in the 19th or early 20th century weren’t looking for workers with particular “personality, talent and background,” to borrow Watson’s phrasing.

The same way a gold miner searches for “the mother lode,” a new brand of company in the 20th century was starting to search for “talent.” Talent is not a finite resource. You don’t capture talent in a war. You can’t make it produce under the lash. Talent requires cultivation. A poverty-striken, war-torn country has just as many smart people as a prosperous, peaceful nation, but it doesn’t have as much talent ready for knowledge work. It is impossible to develop the human potential of a population, making it ready to power a knowledge economy, without significant, long-term investments in that population’s well-being. By the second half of the 20th century, nations were in competition to create the best conditions for developing talent.

By the post-WW2 era, a new generation of entrepreneurs was accumulating wealth, which breeds power, by cultivating human ingenuity. What they wanted from a government was not just different from the needs of planters and industrialists, but starkly at odds with these older power centers. White supremacy was becoming an obstacle to money and power.

IBM and similarly placed companies needed a mass of healthy, comfortably raised, highly educated and creative people prepared to devote their minds to problem-solving. These new knowledge workers were not fungible. This labor pool did not fit into classical Marxist categories. Knowledge workers’ talents and skills were not sold in the marketplace by the hour but could become a platform for capital investment. Those skills could be improved by investment, usually in the form of education or training, raising their yields, like capital. And these workers were beginning to be compensated directly in the form of capital ownership in their employers’ companies. For knowledge capital investors, development of a population’s health, welfare and talent made everyone richer, even as the workers themselves earned more. Knowledge capitalists needed a government unlike anything humans had ever seen. The government they want still hasn’t fully taken shape.

What IBM, other knowledge capitalists, and the national defense infrastructure needed from a political system was not what cotton planters and factory-owners needed from a political system. White supremacy, as developed, refined and perfected to harmonize the interests of rural and industrial capitalists would be at war, almost literally, with the needs of knowledge capitalists, for decades. Knowledge capitalists won this war because they have vastly more wealth, and wealth-creation power, at their disposal and because a powerful vein of activist egalitarian politics survived in the US, ready to seize this opening.

This is bad news for America’s remaining white supremacists. Whether it becomes good news for humanity is a question that remains unanswered.


  1. One thing I wonder is if we’re running out of gas via forms of organizational regulation, once you start thinking about quite a few staff members who are reflexively suspicious of (most often) Black people in particular.

    For example, back in 1960, you could probably get a few people who never thought about the problem of poor treatment of various kinds before to start thinking about it, and make some gains. In 2020, it seems to me that many processes are rendered unfair by but a few, but the few causing the problems are hard to locate efficiently and definitively, and have proven numerous enough that it’s difficult to take decisive action to remove them.

    This hinges on the observation that many processes are not a “majority rules” type of thing. It’s almost always preferred to have more agreement rather than less: one overly suspicious person on a hiring board of five or even ten is not going to be so well overruled by by the other four or nine. Building a case for dismissal is tough, especially if the industry has already deterred applicants that suffer discrimination and there is little data to build such a case. This is a difficult bit of asymmetric conflict between more numerous egalitarians and versus a few anti-egalitarians.

    In this, I have my own field of familiarity, technology, in mind. While perhaps its practitioners are, I think, more inclined towards egalitarianism than other fields by at least a little, the workforce is so lopsided against Black people and women that the 10th percentile of most knuckle-dragging types is going to be serious sand in the gears. They do not need a numerical majority or anything close to it to uphold the status quo.

    1. In tech, there’s a meritocratic libertarianism that has its pluses and minuses. It’s rare to find people who are noisy bigots. It’s also rare to find people who actually give a damn. The general attitude is that they’ll embrace anyone who can perform, but there’s very little interest in understanding what it takes for people to get in a position to perform.

      We’re seeing a lot of this kind of resistance to our efforts to change our hiring process. People who refuse to change their hiring filters insist that there are no minority candidates who are “qualified.” Efforts to open hiring to a broader range of candidates are getting “colorblind” pushback.

      What might break matters open is the dramatic success of people who are making an effort to hire people from more diverse backgrounds. Teams that are succeeding in these efforts are producing some absolute stars.

      1. Noisy bigots are nearly extinct, yes. Though there’s an ideological character, I think that’s the lagging indicator of something more fundamental: some people are just suspicious of, say, Black people, or women, with certain job descriptions. These are pretty old chauvinisms, to reduce their more damaging adherents below 10% would be very difficult, especially if they just clam up — which is still better than nothing.

        I think this type of character would be overly difficult to persuade as to the merit in even dramatic success unless they were already intimately familiar with the case. But the odds of someone of this mold being so closely positioned as to ascribe merit fairly is tiny.

        My more general point is that we need to evaluate to what extent a uniformity assumption (i.e. harm agents are evenly distributed) is being used in certain theories of why discrimination persists, and ask: can that really be? I strongly suspect that harm distribution is power-distributed (in the statistical sense, not the sociological sense), and our processes are sensitive to that margin.

        For example, the interview process:

        Firstly, without calibration and standardization, you’re already cooked for obvious reasons…and these practices are not yet the mainstream.

        But let’s suppose even with a more standardized metric that one of five panelists is about 50% harder to give correct scoring for a certain appearance of interviewee. Well, hiring is not a majority rules thing so much: people will move on to candidates that have maximum agreement. What if it were standard practice to drop one outlier, for example? The odds of a double fault (two such interviewers) is far smaller.

        Various asymmetric processes are everywhere once I stop to think about it. The rate and responsiveness of patch review, of receptivity to good ideas suggested by the discriminated party, and so on. Perhaps it is possible to signpost more egalitarian pockets in an organization? Is there any way to do that when the company is very lopsided, i.e. there’s almost no data being generated/emitted about how well various pockets are working for people? And so on.

  2. The other question is also, how do we deal with modern Luddites, people hwo see the future and openly reject it.

    One of my favorite online streamers recently lamented “beware the Nazbol vortex.” and this is something I think is going to happen in the coming years. The Republicans are going to fall further and further down this hole. Basically they are going to start “advocating” for workers rights and class issues, but use that as a way to directly attack BLM, trans-rights and other social issues, the way the Nazi’s did in Germany against their minorities.

    The thing I’ve been saying is that the Democrats’ and republicans are essentially switching places to what the parties were before WWII, where the republicans were the ones who pushed for most of the racial integration legislation pre-WWII and the democrats opposed it due to the inclusion of southern democrats that have now joined the republicans. Trump is basically the culmination of that a Neo Confederate president.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.