Quick Reaction To PG’s Essay, “Refragmentation”

As Paul Graham usually does with his essays, he makes a deep point with relative ease. Today’s essay, which many of you have read (and if you haven’t, please read it here) is even more intense. I had a sense he would be writing about the relationship between the proliferation of technology and its potential relationship to income disparity, and I know from some of his tweets over the course of the last month, he was likely extra careful in selecting his language for fear of being accused of insensitivity given the consequences if his theory is correct.

I’ve read the essay now twice and, as someone who grew up studying history and still tries to pay attention to how broader themes shape nations, societies, and cultures, I have to say that Graham’s essay is one of the most powerful I’ve ever read. He spends more time explaining “why” the world is the way it is today, and doesn’t wade into the territory of “OK, so what do we do about it?” Those are thornier issues to discuss, and perhaps the first step in addressing them is the acceptance that they’re happening to begin with.

If I had to boil down the essay (which is unfair, because it is very economical in its language already), my big takeaway is that as the 21st century unfolds with the power of the Internet at scale through mobile devices and the proliferation of technology and computing power into every industry, it calls into question The Coase Theorem which, for decades, defined why larger corporations existed — to manage complex transactional costs. Today, based on this essay, one could argue The Graham Theorem is now that networks of smaller companies render Coase obsolete. A step further, these smaller networks of companies (like that come out of YC) leverage technologies (through instruments like APIs) and can be run by smaller numbers of people, yet will likely accrue financial leverage as a result of the ratio of human:computer labor needed to manage and execute those transaction costs. Computers now make it cheaper, faster, and more efficient.

(As an aside, I’m sure there are other factors to consider in the argument that were not discussed. It is his essay and point of view. I am sure land issues could be another factor that drives inequality, or complex issues around socioeconomic status, race, gender, as well as access to educational resources.)

I do mean to imply there aren’t other factor at play, but to me, the more interesting question posed by the essay is — “Let’s assume this is happening. Then, what?”

Graham stops short of suggesting “what.” I understand why. It is too much for this post. The first step is acceptance. Income inequality is happening, and it is (perhaps not entirely?) driven by the accelerating rate and power of technological proliferation.

If society does “accept” The Graham Theorem as the prevailing OS of society (replacing Coase), what should our collective response be? Here are some ideas that are cited, and likely we will need all of these and more:

1/ Education: “We need more access to better education.” Yet, many believe our higher education institutes have been raising tuition while failing to keep students on pace for changes in the workforce. Perhaps YC is one of the first examples of that — forgoing graduate school, for instance, to join an accelerator.

2/ Guaranteed Income: This has been an argument — to proactively set basic income levels for citizens — advanced most notably by USV’s Albert Wenger, and most recently by YC’s current President, Sam Altman. Albert has been writing about the intersection of these issues for many years now, and I believe is writing a book that will touch on the topic. You can read one of his posts (which links to a video talk) here.

3/ Taxation: Graham points out we could theoretically tax economically accretive behavior to slow the effects of The Graham Theorem and redistribute wealth, but if done too bluntly, those creators could move to another nation that has more favorable laws and would want to compete for that talent, just like companies are competing for talent today. (This also is connected to global immigration, see below.)

4/ Immigration Reform: This is a sad topic in America today for so many reasons, especially given what is happening in other parts of the world today and how so many of America’s great entrepreneurial stories are carried through by first-generation immigrants. Many have clearly argued for immediate reform, but that seems politically infeasible in a country where xenophobia feels on the rise and where more and more people are beginning to feel excluded from the pistons which drive today’s global economy: technology. (Earlier, Graham has argued on his blog for U.S. immigration reform, citing global competition for talent as a potentially zero-sum game.)

Graham ends his essay today with the following warning (italics and emphasis added):

I worry that if we don’t acknowledge this, we’re headed for trouble. If we think 20th century cohesion disappeared because of few policy tweaks, we’ll be deluded into thinking we can get it back (minus the bad parts, somehow) with a few countertweaks. And then we’ll waste our time trying to eliminate fragmentation, when we’d be better off thinking about how to mitigate its consequences.

It’s a strong statement by someone who is careful with their words: “we’re headed for trouble.” As a reader of Albert’s blog for many years, I’ve come to admire his academic-style and now policy work to advance the issue. He and I have disagreed on what will actually happen. I admire that he is using his power to advance the issue and writing a book which touches on the subject. I am not as optimistic (sadly), where I see the “trouble” being more pronounced, leading to real conflict, causing many people to feel excluded from the economy and from the means of production, and while I want essays like this and Albert’s book to have a positive impact on policy and legislation, the country’s politics and willingness to make hard choices upfront seems to be running on limited reserves. That is my own personal takeaway from reflecting on Graham’s excellent post.

Haystack is written by Semil Shah, and is published under a Creative Commons BY-NC-SA license. Copyright © 2018 Semil Shah.

“I write this not for the many, but for you; each of us is enough of an audience for the other.”— Epicurus