Hacker News Reader

Comment by elvinyung | Hacker News Reader

I have a pet theory that goes one step further.

In Discipline and Punish, Foucault talks about how in the early modern era, the actions of subordinate individuals (workers, students, soldiers, etc.) became increasingly regimented through more and more precise specifications of procedures as sequences of elementary actions. Then in the 19th and 20th centuries, we started to see "rationalist" economic efficiency projects like scientific management/Taylorism, which essentially tried to deskill factory workers by fragmenting their work based on time and motion studies.

The gig economy is the neoliberal revival of these scientific management principles, plastered over with the neoliberal ideals of marketization and entrepreneurialism. Just like before, it aims to improve efficiency by quantifying and compensating work at finer-grained scales than ever before.

It has the potential to drastically increase economic efficiency, but at the expense of financial security and the psychological wellbeing of entire generations.

View on HN

What do you mean by economic efficiency? It seems like the model of a lot of gig jobs encourages the employer to waste the worker's time.

This happens because the employer pays per job. For example, Bird and Lime waste workers time by letting them all chase after the same scooters to be charged so a worker might spend time going to pick-up a scooter only to be beat to it by another by a few moments. It seems highly suspect that this is the most efficient way to get this done unless your idea of efficiency only means lowest cost to the company.

I meant something like "fraction of the employee's time spent on what the employer considers to be productive/useful work".

Contrast it with a typical salaried job. In such a position, people basically get paid to take bathroom breaks, chat with coworkers, check social media, eat lunch, etc. and they might not even be in the office for the full 40 hours a week. They get to eat up even more resources by participating in benefits like health insurance, paid time off, and pensions. Isn't that such a waste of resources?

Optimize that ratio, ignore productivity, achieve glory.

That salary work mostly pays a lot more than gig work tells you something about the economic impact of both. Gig work might be more efficient than hourly pay for the same work, but it isn't maximizing overall economic efficiency, it's a middle man capturing more of the value.

But salaried work also needs things like a university degree (mostly). What I'm worried about isn't so much that there will always (or for the foreseeable future) be "skilled" and "unskilled" buckets of labor, but that there's this self-reinforcing loop that creates a permanent underclass of gig workers.

Lots of skilled jobs don't need college degrees, employers just require them because it's convenient filter that doesn't have much of a downside for them.

You did qualify your assertion with "mostly", but IME (I switched from trad salaried employment to FT consulting 2 years ago), there's a lot more money to be made in the latter. You have to deal with more risk and more uncertainty, but in terms of income, the upside is way, way higher.

Consulting is not a "gig economy" type job

I get that there is linguistic overlap and some similar conditions, but business consulting is better described as a professional service (like lawyering or accounting) than it is as gig labor.

Marx discussed this thoroughly. The same thing is happening to salaried jobs, just to a lesser extent. This phenomenon was one of his bigger objections to capitalism.

All productive labour is being broken down into pipelined work streams where any individual worker performs as simple and as well defined a task as you can give them, and performs that task all day every day. This increases the economic efficiency of the worker, at the expense of depriving their work life of any meaning and turning them into an unremarkable, disposable, exchangeable cog in a machine.

Pretty much every industry in the developed world follows this trend, including tech.

Yes, exactly.

It almost feels like secure salaried positions were actually the abnormal deviation in history, a small island of stability caused by a temporary surge of social democratic politics in the post-Great Depression era.

> democratic socialist politics

Do you mean social democratic?

Gah, sorry, fixed -- thanks!

Quibble - this comes from Smith. It's not specifically a Marxist complaint - just one popular with them.

Mind referencing where Smith highlights this as a drawback of capitalism? I'd like to learn more.

Book V of the Wealth of Nations has a long passage about the ill effects of the division of labor (or more specifically, of the constant repetition of simple tasks) on the workers.

Here: https://www.marxists.org/reference/archive/smith-adam/works/... (search for "In the progress of the division of labour")

Cheers, that was a good read. I'd still attribute the argument to Marx, because Smith doesn't seem to link this effect with the inherent values of capitalism.

Turns out there's also a wikipedia page outlining the history of this line of reasoning:

https://en.wikipedia.org/wiki/Division_of_labour#Henri-Louis...

But the word 'Capitalism' did not exist in Smith's day!

I can say this absolutely exists in Corp tech shops.

I worked in an environment where the CIO would frequently refer to the tech staff as “the factory”. I often encouraged this CIO and others in the management team not to because it was demoralizing and not true - the work was highly skilled, highly dynamic and not repetitive.

Oh and if you treated tech work like a factory it looked like lots of management overhead, poor execution from low skilled workers, predictable timelines (long timelines), and last but not least higher cost.

Well it has existed throughout history.

We haven't discovered great ways of winning wars or building empire without being inefficient and treating people like pawns.

Our history books don't spend time making us think about the morale of the front line soldiers in Alexander/Genghis Khan/Napolean/Stalin's armies. Instead we have entire shelves of books devoted to the mindlessly ambitious guys at the top of the food chain.

We make Edison an icon and give Marconi a Nobel, but if you know your history, these guys ended up at the top of the food chain because they were ruthless. It hardly mattered what Oppenheimer/Einstein/Feynman thought about the nuclear bomb. Because they were the "factory".

Much like the "factory" of people at the pentagon/wall st/google/facebook etc.

We still haven't figured out a good way to keep "factories" running or expand/defend empires without propping up mindlessly ambitious people. The inequality rates suggest our faith in them is at an all time high.

Our current default method much like the cold war or a chimp troupe, is to keep mindlessly ambitious people at the top of the food chain constantly paranoid about each other. Ofcourse we can do better. But the agency isn't there yet. For example, Facebook could turn into a Wikipedia type entity by tomorrow if the "factory" of workers revolt.

I’m not sure Oppenheimer/Einstein/Feynman would have been referred to at the time as “the factory” nor would I imagine anyone consider their work needing to be highly efficient nor would I imagine anyone would dream the appropriate management style to be to eek out all the efficiency in what they were building.

I don’t disagree with your thoughts on charismatic leadership but the response seems misplaced or slightly off topic.

I think factory management and motive is slightly different than dictatorial/power leadership.

In my company they have started to refer to people as "resources". It feels really dehumanizing.

This has been the case for years there is usually a department with it in the name - human resources.

D.E. Shaw apparently uses the term "human capital".

the CIO would frequently refer to the tech staff as “the factory”. I often encouraged this CIO and others in the management team not to

At least he was being honest in revealing management thinking. Worse would be a guy who said the company was like a big family, while planning which kids to lay-off next quarter.

We have some of those too but they are usually junior managers who have a single level team (and likely haven’t had to fire or lay someone off). Org leaders usually have more perspective than to call it a family.

And these cogs are more and more actual machines, computer or mechanical. I read somewhere that if your job does not consist of some Rube Goldberg like process where you never are doing the same thing more than a few times, then your job is likely to be automated as the machines continue to improve.

A successful path forward for humanity in such a world might involve somehow getting most people to have ownership of the robots that produce what they need/want. A difficult proposition because, if one has the freedom to manage capital one also has the freedom to be conned out of that capital or take on too much risk. It also seems very unlikely at the moment as at least the US's educational system seems to be run by people that think thinking about capital and investing is not valuable or a worthy use of time. Wouldn't it be great if we were all capitalists with an income stream we controlled that could support a minimum lifestyle. If it was supported by mostly robots/machines and not other people, it could be a great outcome for everyone.

Efficiency, to me, means getting the same result for less inputs. This could take the form of a less expensive method for performing some work or a higher utilization of existing time / equipment. A lot of these gig employers aren't increasing efficiency they are just shifting costs to unsophisticated workers who don't realize it.

> A lot of these gig employers aren't increasing efficiency

Regarding ridesharing, I disagree. App-based hailing is objectively more efficient and scalable (to lower population densities) than street hailing, and more efficient than privately-owned cars requiring massive, distributed investments in parking infrastructure.

Agreed about app-based hailing being more efficient, but I think it’s not germane to the issue at hand – those same organizational improvements would be seen if Uber classified its employees as employees, or if a city’s licensed taxi providers could get their shit together for a good app.

You realize you have always been able to call a cab in low density areas right? Ride-sharing eliminates the dispatcher. It doesn't change the economics of a taxi service in lower density areas. Those higher costs are shifted to the gig employee. Last I checked a driver doesn't get any extra fee if she has to drive 5 miles vs 500 feet to get to her next fair.

At least until recently these services didn't even seem to help with the utilization problem where drivers would be sitting for periods w/o a fair. At least Lyft now will schedule a drive with their next rider while the current one is in the car.

They do. It's called a Long Pickup Fee or a Long Pickup Premium and it depends on time as well.

Yeah, they just instituted that in 2017 because drivers were getting wise.

By hiring a salaried worker, you are essentially locking all of their work to your company's benefit. A quality gig worker can finish an assignment at your company and go work for your biggest competitor. There is value in monopolizing their value to your organization.

In practice this switching ability isn't true or easy. Sandwich workers in Seattle had to sign non-competes and couldn't go work at other restaurants. There was a lawsuit from the city to stop it.

Same in Massachusetts with sandwich workers, they finally had to make a law limiting no competes even though there had been political opposition to that idea for years.

Companies are attempting to monopolize the employees output, but only pay for the seconds the employee is producing for them

Sandwich workers are hourly not salaried.

salary vs contractor is pretty similar to wholesale vs retail. if you know up front that you need a large amount of work from someone over a long period of time, you hire them as an FTE with a salary. the wholesale rate for developer time is much lower than what a comparable contractor will charge you per hour for the convenience of only paying for the hours where they do useful work.

a possible reason why it is uneconomic to give gig economy workers salaries is because the necessary discount would put them below minimum wage.

> unless your idea of efficiency only means lowest cost to the company

I believe this is exactly what economosts mean when they talk about efficiency. It’s professional jargon. And yes it differs frim the common sense definition of efficiency.

"What do you mean by economic efficiency? It seems like the model of a lot of gig jobs encourages the employer to waste the worker's time."

But the employers aren't paying for that time.

Your comment brings to mind a fascinating short story that follows this idea: http://marshallbrain.com/manna1.htm

I am pretty sure I found it on HN a few years ago.

Oh wow, thanks so much for linking this story! I was actually thinking of it as I was writing the post, but I couldn't remember the title.

That sounds like a fancy way of saying advances in technology displace the need for certain (and increasing) types of work, increasing the supply of labor, but not the demand, and therefore decreasing the price of labor.

Plus globalization and adding women to labor supply.

It specifically isn't about that.

The idea is that these projects aim to make more and more kinds of manual work fungible and exchangeable by breaking them down into small enough subtasks. Each of these elementary tasks get commoditized into a market of near-perfect competition when they become deskilled this way. It's not automation, it's a hyper-division of labor, like a worldwide assembly line connected by ethernet and optical fiber.

It is specifically that it isn't automation that makes it a problem. If you were totally replaced by a machine, then at least you wouldn't have to work. But here, you aren't being replaced by machines -- you are part of the machine, like a component that's completely replaceable and interchangeable.

I think it still qualifies as “automation”. The automation doesn’t completely replace people, but it does replace portions of their tasks, and when you start adding it up it does replace people.

Two people needed to do 5 tasks now done by 1 person who can handle all 5 tasks. Full time secretaries for everyone are now down to 1 administrative assistant for the whole team or office. Taxi dispatchers not needed anymore.

That’s what technology enables people to do. Measure, target inefficiencies, and commoditize. But I don’t think it’s anything like a grand conspiracy, it’s pretty much the only way to move forward.

Problem is the humans aren’t so interested in distributing the fruits of the resulting efficiencies in an equitable manner.

Well, sure, but I think there's something to be cautious about here. The legend goes that the Romans never found an economically useful application for the steam engine (the aeolipile invented by Heron of Alexandria) because they already relied on a massive labor force of slaves.

But it's precisely this barrier that we've begun to cross that indicates there are some very old questions that we should be asking ourselves again.

I think you and I agree that at the end of the day, the only thing that matters is that this increased economic efficiency has the potential to improve everyone's wellbeing, not just concentrate wealth in the hands of a few.

Yes, hopefully we can find a balance between a system that rewards effort and ingenuity at the same time as providing opportunity for everyone.

In thd gig economh, the machine isn’t doing any of the work. When I order something on instacart, it’s not functionally different from calling in an order to a human shopper. The machine isn’t doing any part of the shopping or driving. Indeed, instead of displacing human labor, like you suggest, it’s increasing the amount of work available by bringing previously non-market work into the market. The machine does that by reducing the transaction costs to make outsourcing that kind kind of labor feasible. Relevant reading: Ronald Coase’s Theory of the Firm.

The machine (internet/instacart) is doing work, it's reducing the transacation cost so that you and the delivery person can make a deal. It’s just like an exchange making trading easier, it increases liquidity and facilitates transactions.

“Non market” work seems like another word for labor you can’t afford. Rich people with drivers and butlers or whatever have had instacart forever.

Having a central dispatch to relay what you want to a driver, have them go pick it up and deliver, and then re route them to the next one all while handling incoming calls is probably too expensive to offset the convenience of an individual not driving to Target and picking it up. But with a computer and some programming, you don’t need any dispatch, and you don’t waste time communicating what you need, which drives down the cost of getting something delivered to you so that it’s now affordable. The machine is doing the work of people that would have had to coordinate and pass all that information between two parties. Not to mention the risk of not being paid (taken care of by the payment card networks).

Also, I’m not convinced these many of these “gig” companies are going to be around forever. Instacart might work in a few urban locales with a few rich young people with no families, but it’s far from proven that it’s valuable enough and cheap enough for most people.

> The machine (internet/instacart) is doing work, it's reducing the transacation cost so that you and the delivery person can make a deal. It’s just like an exchange making trading easier, it increases liquidity and facilitates transactions.

“Work” is different than transaction costs. The amount of human labor needed hasn’t changed in any meaningful way compared to what was needed before, so it’s not the reduction of labor demand through automation that’s at issue. What’s changed is, as the OP explained, that technology now permits more fluid allocation of that labor so as to reduce transaction costs. This is a different phenomenon from automation eliminating the need for actual work.

> “Non market” work seems like another word for labor you can’t afford.

Non-market work is a well understood term in economics for work households perform themselves, such as child care, etc. Technology that allows moving non-market work into the market doesn’t decrease labor opportunities, it increases them. That’s the opposite of your thesis.

Again, two different phenomenon. A roomba is automation—it replaces the need for a human to vacuum and decreases the amount of human labor needed. An app that makes it easy to hire someone to clean dishes after a big party is not automation and doesn’t decrease the amount of human labor needed.

> Rich people with drivers and butlers or whatever have had instacart forever.

Yes, and the amount of human labor involved in that scenario was about the same as today. What changed was the transaction costs, enabling a shift from an employee relationship to a contractor relationship Read the Theory of the Firm. It addresses this issue directly.

All labor opportunities are not equal. Driving (especially with GPS and navigation), cleaning, preparing food, moving things might be labor opportunities, but they are things that many more people can easily do with minimal training. Therefore more people end up in the labor supply for these types of positions, and therefore that type of work pays less.

The fact that someone would even be willing to accept "gig" work is because the better opportunities they might have had such as manufacturing, customer service, secretarial, travel agent, etc is automated or outsourced, and/or they need to supplement their income because it's not rising in real terms.

>The amount of human labor needed hasn’t changed in any meaningful way compared to what was needed before

I disagree with this statement because I think that is all that has changed. Before and after instacart or uber or whatever, the cost of gas, vehicles, maintenance did not change in any meaningful way. So why did people not have things delivered for them? Because the equivalent amount of human labor required to do what internet+app+GPS would have been too costly.

The nature of rich people and drivers and butlers hasn't changed, I'm sure those that can afford to have them full time prefer that than some random person on Instacart. And while transaction costs have changed, what has also changed is a demand for labor that needs to be permanently employed for 40+ years. So you have the perfect situation where labor has nowhere to turn but these apps, and these apps can reduce the cost of utilizing this temporary labor such that the upper middle class can afford to use it.

We had taxis and home delivery services for years.

What is the difference today? They are much more affordable due to technological advances in thes upply chain.

Folks in this discussion are missing the point. Just because it is no longer human labor, does not mean it is no longer work.

When apps connect customers with providers instead of people, that is automation.

Then there is good news for you. This Foucaultian critique of the neoliberal notion of entrepreneurialism is precisely what Italian philosopher Maurizio Lazzarato has articulated in his book, The Making of the Indebted Man.

He argues that (to quote Zizek), "in daily ideology, our servitude is presented to us as our freedom. He demonstrates how we are all treated as capitalists who invest in our lives. Indebtedness implies a functional discipline. It is today a new way of maintaining control over individuals, all whilst promoting the illusion of free choice. Even the fragility of our career path and chronic insecurity is presented to us a chance to reinvent ourselves every two or three years. And it works very well."

Interesting. I've only vaguely interacted with Lazzarato's work through Hardt and Negri. I'm guessing the book builds on what Deleuze says about the society of control, e.g. "Man is no longer man enclosed, but man in debt."

Now that's an excellent way to put it.

It's so much more cost-effective than slavery. Slaves have to be fed, clothed, and housed. Gig workers don't.

Not only that, gig workers have freedom and opportunity. Not much of it, but they can fantasize about escaping their trap. There's a whole ideological system telling them it's all their fault. Capitalism, Christianity, and Judaism all line up on this.

Rebellion is very difficult. The debt system doesn't oppress. It's merely indifferent to your suffering. It simply ignores you until you die.

i'm late to the discussion, but i think you buried the lede here (so to speak).

it's not so much that we're reaching for some neoliberal efficiency (whatever that means) at the expense of humanity, but that such kinds of industrial regimentalism allows the managerial class to place themselves above and apart from the proletariat in the corporation.

regimentalism marginalizes workers such that managers dehumanize and treat workers indiscriminately. it allows managers to make rational yet ruthless decisions without considering worker impact. it sets them apart; it others the workers, and thereby gives managers all-important status over workers, even where otherwise undeserved.

at the same time, the proletariat worker self-disciplines themselves (as foucault likes to put it) around the ever-intricate rules and ever-tighter tolerances, so much so that they cannot pay attention to the man behind the curtain pulling the strings (and realizing that he is pitiful and small).

this is the most important aspect of regimentalism, and the most insidious. the proletariat gig-economy worker loses money, power, and esteem in the transaction, but they're too busy jumping through hoops to do anything about it.

I share that point of view. (My pet theory is that the ruling classes are rolling back the "educated peasants' intermezzo".)

Reply on news.ycombinator.com