Merit, Value, Luddites, and AI

Don't equate your income, or anyone else's, with moral worth.

Back in 2004, before many of today’s college students were born, I wrote an article about Friedrich Hayek for The Boston Globe. It came out well, but I couldn’t work in one of the most interesting things I heard in my reporting. Political theorist Glyn Morgan (now at Syracuse) had been instrumental in adding Hayek to the syllabus for Social Studies 10, a year-long survey of major thinkers in social theory. In our interview, he talked an aspect of Hayek’s thought, building on Adam Smith, that troubled students:

Smith encourages people to be stoical about whether they succeed or fail in the market. This is a bit different from the Protestant that Weber writes about. The Protestants want to persuade us that whether we succeed or fail in the market is a sign of our merit or worth.

And Hayek’s like Smith in the sense of disagreeing with the Protestants. He realizes that to a certain extent it’s a game of chance and whether you succeed or fail there’s going to be a lot of luck involved. I think that hard-working, successful, achievement-oriented Harvard students don’t like that idea. They’re troubled with the idea that there’s a lot of luck. I tell them it’s the same about grades. Whether you get an A or a C, there’s an element of luck in it. You can’t always come up with a great idea that day. We all won’t write brilliant papers every time. There’s a lot of element of luck in life. They don’t like that. They think there’s destiny and merit and if you work hard you will succeed.

Value is not the same as merit. Harvard students may not like it, but the distinction is a critically important one—morally, economically, socially, and politically.

In Hayek’s terminology, value is simply the outcome of market processes at the moment—of the supply and demand that reflect current circumstances. Those circumstances and and likely will change. We shouldn’t confuse economic success with just deserts. Talent and hard work may sometimes be necessary to success, but they definitely aren’t sufficient. What constitutes talent in one era may be a low-value attribute in another and, besides, talent isn’t something you earn.

In an August 3 post, Harvard grad Matt Yglesias explained the difference:

The fact of the matter is that whatever privileges I may have benefitted from in life, a big part of the reason I’m successful is that I am, in fact, much better than the average person at writing high volumes of cogent articles. But this is a bit of a time-bound and arbitrary talent. I’ve worked alongside a lot of people who are better than I am at taking 2–3 weeks to write a single long-form article. I’m more successful than most of those people, because it happens to be the case that at our particular point in history, taking 2–3 weeks to write a single long-form article is a less valuable skill than writing 4–5 columns per week. But 25 years ago it was exactly the opposite; in the print age, column inches were scarce and the winning strategy was to be someone who, given enough time and effort, could generate something really, really good. In the digital age, space is plentiful and the winning strategy is to fill the space with content that’s pretty interesting or informative or entertaining.

These are related skills — the people who are really good at one are usually at least above average at the other — but they are distinct. And it’s not just a question of practice. I have tried over the years to do feature writing, and I just don’t reach the heights that many other people in the field are capable of.

I’m fortunate not only to be alive at a time when the kind of content production I’m good at is valuable but to have come along in the industry right when things were turning. I graduated from college in 2003, and for the first four or five years of my career, digital work was still considered low status. That meant it was relatively easy to get my foot in the door, and I was already a well-established voice when the 2008 financial crisis blew up print journalism in a way it’s never recovered from. That’s tremendous good luck in life. But it’s not luck as opposed to merit, it’s that questions of what skills happen to be valuable and who has them and at what time are all highly contingent. And these market judgments about what skills are valuable not only shift over time, they aren’t ethical judgments about people’s worth.

Yglesias doesn’t cite Hayek but, given the timing, I have to wonder whether he read him in Social Science 10.

Hayek’s original discussion tends to equate “merit” with effort, but it could mean anything that is considered a virtue.¹

The inborn as well as the acquired gifts of a person clearly have a value to his fellows which does not depend on any credit due to him for possessing them. There is little a man can do to alter the fact that his special talents are very common or exceedingly rare. A good mind or a fine voice, a beautiful face or a skilful hand, and a ready wit or an attractive personality are in a large measure as independent of a person's efforts as the opportunities or the ex­periences he has had. In all these instances the value which a per­son's capacities or services have for us and for which he is recom­pensed has little relation to anything that we can call moral merit or deserts. Our problem is whether it is desirable that people should enjoy advantages in proportion to the benefits which their fellows derive from their activities or whether the distribution of these advantages should be based on other men's views of their merits.

Reward according to merit must in practice mean reward according to assessable merit, merit that other people can recog­nize and agree upon and not merit merely in the sight of some high­er power. Assessable merit in this sense presupposes that we can ascertain that a man has done what some accepted rule of conduct demanded of him and that this has cost him some pain and effort. Whether this has been the case cannot be judged by the result: merit is not a matter of the objective outcome but of subjective effort. The attempt to achieve a valuable result may be highly meritorious but a complete failure, and full success may be en­tirely the result of accident and thus without merit. If we know that a man has done his best, we will often wish to see him re­warded irrespective of the result; and if we know that a most valuable achievement is almost entirely due to luck or favorable circumstances, we will give little credit to the author….

To decide on merit presupposes that we can judge whether people have made such use of their opportunities as they ought to have made and how much effort of will or self-denial this has cost them; it pre­supposes also that we can distinguish between that part of their achievement which is due to circumstances within their control and that part which is not….

If a man's ability in a given field is more valuable after thirty years' work than it was earlier, this is independent of whether these thirty years were most profitable and enjoyable or whether they were a time of unceasing sacrifice and worry. If the pursuit of a hobby produces a special skill or an accidental invention turns out to be extremely useful to others, the fact that there is little merit in it does not make it any less valuable than if the result had been pro­duced by painful effort….

A society in which it was generally assumed that a high income was proof of merit and a low income of the lack of it, in which it was universally believed that position and remuneration corresponded to merit, in which there was no other road to success than the approval of one's conduct by the majority of one's fellows, would probably be much more unbearable to the unsuccessful ones than one in which it was frankly recognized that there was no necessary connection between merit and success.

I can’t help thinking that America would be less riven by conflict if everyone “frankly recognized that there [is] no necessary connection between merit and success.” On both sides of the red-blue divide are economic winners who resent challenges to their sense of superiority—especially from other affluent people with different backgrounds and politics—and economic losers who want both more money and more respect. Recognizing the distinction between economic success and moral worth wouldn’t eliminate conflict, but it might ameliorate it.

Hayek saw the distinction between merit and value as an argument against economic redistribution in the name of equality. If merit and value are different, he believed, there is no moral justification for overriding what markets reward and there are many good arguments against the intervention that would be required to force a “meritorious” outcome. At the same time, I suspect his clear distinction between merit and value also helps to explain why, unlike more absolutist libertarians, Hayek supported safety-net transfers. He didn’t see them as intrinsically unjust. Hayek’s primary concern was allowing people the freedom to act on their own knowledge (including knowledge of their own subjective preferences), without being overridden from on high. I don’t know how he would have assessed the Danish welfare state but, like Yglesias, he certainly wouldn’t have been disappointed that the country’s market economy still produces unequal results.²

My husband Steven Postrel argues that “merit tracks value with a lag,” meaning that whatever happens to be economically valuable will eventually be seen as meritorious. Deirdre McCloskey argues more or less the opposite—that a change in the assessment of the merits of commercial, bourgeois virtues led to an economic takeoff.

One thing is sure, however: People who are doing well almost always think that’s the way the world should be. If you equate value and merit, a fall in your economic fortunes threatens your moral assessment of society (and perhaps yourself) as well.

Luddites and “Email Jobs”

One of the little recognized facts about the historical Luddites is that these displaced handweavers were the beneficiaries of an earlier wave of disruptive technologies: the original spinning machines of the Industrial Revolution, which made once-scarce weaving yarn abundant. Like the power looms that the Luddites destroyed, machines for preparing and spinning fiber also sparked protests. As I wrote in The Fabric of Civilization:

Protesters smashed machinery and demanded government relief. Pending parliamentary action, the town of Wigan halted the “use of all Machines and Engines worked by Water or Horse, for carding, roving, or spinning of Cotton.” A petition to Parliament explained that the “Evil in Question is the Introduction of Patent Machines and Engines, of various Descriptions, which have superseded Manual Labour to such a fatal and alarming Degree, that...many Thousands...with their Families, [are] pining for want of Employment.”

Parliament commissioned a report but decided against action. “A very valuable Manufacture of Callicoes has been established in the said Country by the Use of the Patent Machines,” it concluded. For all the disruption it caused, the new technology was creating new kinds of jobs and benefiting the nation as a whole.

A pamphlet with the unwieldy but clear-cut title Thoughts on the Use of Machines in the Cotton Manufacture. Addressed to the Working People in that Manufacture and the Poor in General laid out a case that could be made about music streaming, self-driving cars, drone deliveries, or any other fear that the robots are taking our jobs.

Those who were thrown out of their old employments, will find, or learn new ones. Those, who now get less by their labour, will be aiming at the more profitable branches. Those who, by striking early into new inventions, get a disproportionate gain, will soon find so many rivals, that they must sink their terms, and reduce their profits….In fact, the cotton manufacture is almost a new trade. The fabrick, the quality of the goods we make, is amazingly changed. How many new kinds of cloth are made, in very great quantities, which could not possibly, have been made, at least in any quantity, or so cheap to sell, without our machines?

Though perhaps too sanguine about the immediate fates of individuals, the tract writer was correct about the big picture. By making thread plentiful, the “patent machines” changed the world. From clothing to sails, bed linens to flour sacks, essential items were suddenly much cheaper, more varied, and more easily obtained. Women were liberated from their spindles and distaffs. It was the beginning of what economic historian Deirdre McCloskey calls “the Great Enrichment,” the centuries-long economic takeoff that lifted global living standards. Just as string allowed early humans to conquer the world, abundant yarn created ripple effects within nearly every aspect of life.

In the decades following the introduction of mechanical spinning, handloom weavers “enjoyed a golden heyday,” writes historian Beverly Lemire, “with as much work as they could want and at high rates of pay. The technological imbalance between weaving and spinning worked to the advantage of that generation of skilled and semi-skilled weavers, men and women.” A generation later, power looms arrived and, their own livelihood threatened, once-prosperous weavers became Luddite rioters. Their case against new technologies wasn’t principled, in other words. It was entirely self-interested.

I thought of the Luddites while reading Freddie DeBoer’s recent post on what he calls “email jobs.” He gives a list of characteristics.³ These jobs:

Depend, naturally, on email and other digital communicative tools like video conferencing, online calendars, and networked workspaces for the large majority of their actual productive capability

Are staffed almost entirely by people with college degrees, but while they do take advantage of time management and organizational skills that can be developed in college, almost never call on domain-specific knowledge related to a particular major

Dedicate a considerable amount of time not to the named productive goals of the job themselves but to meta-tasks that are meant to facilitate those goals (scheduling, coordinating, assigning responsibility, “touching base,” enhancing productivity, ensuring compliance with various HR-mediated job requirements and odd whims of the boss)

Have no immediate observable impact on the material world; an email job might involve coordinating or supporting or assessing a project that will eventually move some atoms around, but the email job itself results only in the manipulation of bits

Cannot be considered creative in any meaningful sense - they do not entail the production of new stories, scripts, code, images, video, blueprints, patents, research papers, etc - but may involve the creation of materials that are subsidiary to larger administrative goals, such as PowerPoint presentations, reports, postmortems, or white papers

May or may not be partially or fully remote but could likely be performed fully remotely/on a “work from home” basis without issue

Can involve supervising lower-level workers, even teams, but these positions are not themselves fundamentally supervisory and the holder of an email job is rarely the only “report” for anyone; these positions, in other words, are not executive or executive-track, though some may escape the email job track and gain entry to the executive track

Tend to top out at middle management, and often have a salary range (with a great deal of wiggle) between $50,000 and $200,000/year….

He offers an example:

Take someone who works in accreditation at a college in a large public university system. He or she didn’t get a major in accreditation (there is no such major) and is unlikely to have majored in education, and even if they did they would have learned about pedagogy and “theory” and assessment rather than anything having to do with their daily work lives. Essentially everything they do for work takes place within the confines of their laptop screen, and the exception is various in-person meetings that accomplish nothing beyond delegating various tasks, defining roles, critiquing past performance, and otherwise reflecting on how to do a better job of supporting the tasks that other people do. A person in this job might have a secretary or lower-level administrative functionary that reports to them, but they are not on a track that makes advancement likely - becoming a VP somewhere will likely require many years of service and going on the job market to get a job at another school. A person in this position will never interact with students in any real capacity, demonstrating the psychic distance between email jobs and the actual function of their institutions. Though they have a clear and defined set of responsibilities written into their job description, their overall impact on the day-to-day functioning of their college is nebulous, and far more time is spent on administrivia than their “real” duties. They live between the 50th and 75th percentile for individual income in their state.

Although rarely the focus of public discussions about employment, these pixel-pushing jobs soak up a huge number college graduates who earn solid incomes. DeBoer’s point is that we’re thinking about college the wrong way: “I’m willing to argue that many millions of Americans have email jobs, that their share of the workforce is growing, and that the constant tendency to think about college as a route from a particular major (prelaw, premed, computer science) to a particular educated position (lawyer, doctor, programmer) is therefore flawed.”

He’s probably right. But that’s not what makes email jobs interesting in 2023 as opposed to 1993—especially if, like DeBoer, you think that “email jobs are desirable jobs, and placing people in steady email jobs is a kind of success for our economic model.” [Emphasis in the original.]

These are the Luddite jobs of the 21st century—the jobs where a college education brings a “golden heyday” up until AI replaces the pixel-pushing, or at least allows one person to do the work that formerly occupied 10. For society as a whole, that productivity boost could be fantastic, but for the pixel-pushers it will be bad news—all the more so if their sense of worth is invested in their status as a well-educated email worker.

Value is not merit. Merit is not value. Please remember that.

Fun with Subtle Relevance

1. Published as chapter 6 in The Constitution of Liberty and available in a terribly formatted, copyright-violating version here.

2. An unrelated tidbit about the Danish welfare state: While researching The Fabric of Civilization I spent several weeks at the Centre for Textile Research at the University of Copenhagen. During the time I was there, the center enjoyed the services of an all-purpose research assistant/administrator—we called her the director’s “entourage"—who had an academic background in archaeology. Having been laid off from an unrelated position, she was working as a volunteer. Volunteering at a nonprofit was a requirement for collecting unemployment compensation while she looked for a paid position. Given the funds, I’m sure CTR would have hired her as permanent staff.

3. Substack won’t let me use the bullets and an indentation for a block quote so I omitted them.