Hacker News new | ask | show | jobs
by sofon 2856 days ago
Really interesting article!

The Peter principle states that "every employee tends to rise to his level of incompetence". The article cites an example from Sales teams. That it's the highest performing sales people that are promoted to sales team managers, and they don't perform well as managers (team performance drops).

They then talk about solutions, such are promoting people at random.

But it seems like at least part of the problem in endemic in our view of organisations. People often want management positions because they are paid more, are more secure, or have more power. Maybe if we shifted our view of management as "just being a different skill set" and not always being higher paid then we'd find people without the required skills wouldn't try and obtain those positions and would focus on what they are good at.

6 comments

> But it seems like at least part of the problem in endemic in our view of organisations. People often want management positions because they are paid more, are more secure, or have more power. Maybe if we shifted our view of management as "just being a different skill set" and not always being higher paid then we'd find people without the required skills wouldn't try and obtain those positions and would focus on what they are good at.

Some tech companies have started to do that with technical seniority tracks. I think one bit that's still missing remains the ability to move people back into previous positions without loss of face or income if they don't fit in the new one.

The problem, which I worry is unfixable, is that managers are the people who decide which jobs are most stable and paid the most. Unsurprisingly, they give themselves the majority of the spoils.
It's interesting too that democracy is lauded in the west but companies are never run that way.
Democracy isn't democratic either. Why else would concepts like seniority exist in the US Senate? Or sub committees... How is the US Senate in line with the concept of one person one vote? If we want our vote to count more, we should move to Wyoming (which gets two senators the same as California)?

If we have each employee one share each, I'm sure many would simply turn around and see if they can sell it. How can we retain one person one vote in a company? Make shares non transferable? Do we allow proxy voting?

That's because the US government is not organized as a democracy, but rather as a republic. The Senate is not intended to be democratic, it is intended to be a deliberative body of representatives who come to consensus on legislative questions. It's quite a large deliberative body, so its method of consensus is superficially akin to democracy in that many questions are answered through voting, but it isn't, and was never intended, to be democratic in the pure sense. Along the same lines, it is only recently that we began democratically electing Senators.

None of this is by accident, and was done this way for similar reasons as companies have for running themselves un-democratically.

> That's because the US government is not organized as a democracy, but rather as a republic. The Senate is not intended to be democratic, it is intended to be a deliberative body of representatives who come to consensus on legislative questions. It's quite a large deliberative body, so its method of consensus is superficially akin to democracy in that many questions are answered through voting, but it isn't, and was never intended, to be democratic in the pure sense. Along the same lines, it is only recently that we began democratically electing Senators.

How often is there a consensus in the Senate? It seems like almost everything happens along party lines. Who are senators supposed to represent?

I've heard this "a republic, not a democracy" so many times. Don't get me wrong. I am very grateful for a nation of laws. However, it doesn't seem that being a republic, not a democracy is what is protecting the rights of the minority or the rule of the law.

It just feels like the purpose was to prevent wild swings but I'm afraid what is supposed to protect us from abrupt, wild swings will make it near impossible to correct a slow swing.

It's a democracy and a republic. The UK is a democracy but not a republic.
I'd do it just the opposite - folks don't all want to be involved, but want somebody who they respect to do it for them. Freely allow proxies to be assigned, to anybody, not just to 'elected officials'. Those with proxies could reassign the whole wad to another person, and so on.
Is it ok to be reimbursed for giving someone proxy?
Companies in the west are democratic, but the voting rights are held by the owners of the company. If you are a citizen, it is your country and you get a vote. If you are an employee, it is not your company.
Not necessarily true. Many companies including Google have different classes of shares where they have more voting rights than thier share of ownership.
Not many public companies are structured like this, and for a while it was a requirement of some exchanges that one share be one vote. I know that people on Hacker News (high concentration of founders) like the idea of a public listing without one share one vote, but I think it is better for companies in the long run to have their board seats elected by the shareholders. It may seem good now to have Mark Zuckerberg to have total control of FB, but will it seem like a good idea for his children to have more control than the actual owners? This is the kind of trouble companies like Viacom have gotten into by giving their founders too much control.
That's not many companies actually but rather a tiny minority of firms that are exploiting their ability to make unusually huge profits to get investors despite poor and risky terms.
But... aren't they? It's one share one vote, not one employee one vote, but I think that still qualifies as democratic. Then, there are elected board members (the representatives) who hire the CEO (the prime minister?), etc.
That wouldn't fit most definitions of democracy (which would be variations of "every person gets a single equal vote"). Most companies are more of an oligarchy.
Neither is the US. The US is a democratic republic: we vote in congress, the senate, and the president who then make decisions for us on our behalf. Public companies are similar: shareholders vote in board members, who then elect a CEO for the company on our behalf.
That is assuming that there's only one share class. Companies often have different classes of membership/shareholder which have different voting powers (or none).
Try to start a democratic company (or run a thought experiment about doing so) and you will probably find the reasons why democracy works better as a way to manage a society than a single company.
It's not like it's unheard of, but it doesn't fill the pockets of the selected elite.

There are worker cooperatives: https://en.wikipedia.org/wiki/Worker_cooperative

Workers' self management: https://en.wikipedia.org/wiki/Workers%27_self-management

And serious economists who advocate for such a system: https://en.wikipedia.org/wiki/Branko_Horvat

We did. And though I am not sure it would work for every line of work, so far it has been extremely successful for us.

https://www.fountstudio.com/blog/yes-we-have-a-pirate-code-1...

It does happen. Both the largest supermarket chains in Switzerland are cooperatives.

However, small competing firms in a market where people vote with their wallets is already pretty democratic. The voting takes place at a different level - people choose companies rather than leaders or direct policies, but it works well enough.

Also capitalism ends once you’re in a company. It becomes a miniature command economy with all the same inefficiencies.
If you're curious to hear postulated explanations then a good search keyphrase is "The theory of the Firm"
I work at a company where at least on the technical side, this is not true. ICs and managers are on similar pay scales. And manager jobs are certainly less stable than IC positions. Additionally, "failure" as a manager often means returning to an IC role.
Can ICs fire managers?
Directly? No. By definition firing people isn't in an IC job description.

Can an IC get a manager fired? Yes. Skip-level reviews and 1:1s usually paint a clear enough picture for upper management that the correct course of action is to fire the manager. The more senior the IC, the more weight their review carries.

Can a manager fire an IC? Not unilaterally. Again, skip-level reviews and 1:1s are there to help upper management evaluate how well the manager is doing their job. Feedback from other ICs is also a big part of it. Senior ICs are generally involved not just in the decision but as part of ensuring the underperforming IC is getting sufficient support so they can be fairly evaluated.

I'm sure systems like this have broken down at other companies, but it seems to work well enough where I've worked.

That's not true where I am. HR decide pay scales based on market pay levels in our sector.
Okay, so your company delegates the decision to managers in other companies. The effect is the same.
I'm sure this is how it looks from the outside, but it's very much not the way things actually work. Except in very limited cases (like, owner of a small private company), your boss does not set his own salary without oversight. And even then, the best way to increase his own salary is for the company to grow, so he's incentivized to run a proper business and not a sweatshop.
The problem has multiple issues. Managers getting paid more is certainly an incentive for people who aren't qualified to want to become managers.

Managers having power is another issue. It's possible for a manager to have less power. Their job being to facilitate and coordinate rather than direct.

Managers have power on multiple levels as well. Some have power over people, who does what, but they also often have power of decisions, what gets done, what to make, where to focus.

In other words, to fix the incentives that pull non-qualified people into management positions it seems like all of those issues need to be separated from the manager position?

In other words someone might not be a people person so isn't good and facilitating and coordinating between people and or other teams but they could be good at one or more kinds of team level decision making and want that responsibility.

One partial solution to having a single power hierarchy is having multiple power hierarchies. Then no one person is the boss of me, and when someone wants something done, they can't just order me to do it. They can't be a jerk to me or they might find their work assigned to my lowest priority.

It can go horribly wrong, too ("I have eight different bosses, Bob!"). I don't think there's any power structure which can prevent selfish or malevolent people from harming the organization: "The machinery of government is always subordinate to the will of those who administer that machinery."

This works only if ICs get to evaluate performance of managers and have inputs into their remuneration. I haven't come across such an org yet.
I think the article mentions the core problem. It’s not necessarily the people can’t move down (although that would be good) it’s that we don’t take into account the skills of the job the person is GOING to when deciding if they should be promoted to it.

If that was the criteria used maybe this problem wouldn’t exist in the first place.

> I think one bit that's still missing remains the ability to move people back into previous positions without loss of face or income if they don't fit in the new one.

Don't know where you are at but my manager literally did just this by switching back to an individual contributor role and retained the same level. Also, managers and individual contributors make the same amount per level so he didn't lose any income.

At my employer, senior technical roles are bonus-eligible at rates similar to line management. Salaries are similar as well. And I know a few employees who have dabbled in management and later returned to senior technical roles without lasting negative impact.
At my company the levels for technical engineers (traditional engineering btw, not software) is: Engineer 1, 2, Senior, Lead, & Principal where Lead is supposedly equivalent to a Supervisor & Principle a manager. In reality we have only a few leads and no principals. So there is very little incentive to go technical as you'll never make more than a Supervisor and Supervisor has at least 4 more levels above it.

Part of the company is similar to the Peter principle, but where I am, all the smartest and most technical people go into management which creates a different problem. They're often sub par at managing people and the administrative side of things. It also creates brain drain where we constantly have a lack of know-how in the less senior ranks. On the good side is that things generally work better with highly technical management as they tend to almost always make superb decisions and don't have to constantly consult about things. My last manager could listen to work being done and know exactly which processes it impacted and why due to the specific SQL tables involved. They can also direct subordinates in the knowledge they took with them too. These managers tend to be more focused with getting the job done quickly and properly with minimum impacts rather than political infighting.

A technical manager without political skills can hurt the team. They can’t get raises for them, resources, or recognition. A non technical political manager can easily outmaneuver a technical manager. It’s good to have a technical manager to get things done and learn, but expect to have to change jobs to get ahead.
Agreed in some cases, but again, I've been pleasantly surprised.

Technical managers understand what needs to be done...etc and enable high performing teams and make it well known during bonus discussions. When your department is responsible for major success each year, the technical manager can push promotions...etc fairly well. At this rate I'm avoiding the other departments like the plague.

Ah, well failure to use positions that exist is a bit strange. My test engineer is a “senior principle”. Losing her to a management role would be a blow to both my team and the broader product-line (ERP, many dev teams). Of course, I was a principle dev and moved into management for a change of pace and new challenges, so title and salary are no guarantee that tech talent stays in tech roles.
> But it seems like at least part of the problem in endemic in our view of organisations. People often want management positions because they are paid more, are more secure, or have more power. Maybe if we shifted our view of management as "just being a different skill set" and not always being higher paid then we'd find people without the required skills wouldn't try and obtain those positions and would focus on what they are good at.

This is exactly it. Hierarchies are good for efficiently organizing and distributing work. I mean, they're trees which are used everywhere for load balancing and distribution in computer science. We just have to break the association of a work hierarchy with social status which we're hard-wired to see, and part of that is the higher salaries given to people higher in the hierarchy.

Really, salary should probably be tied to hours and seniority.

The organization that pays the market wage will usually out-compete the organization which pays based on hours and seniority. Seniority is a poor proxy for productivity (theer's definitely a correlation, but it's not strong and it ignores individual variation, which is large), while paying based on hours is rewarding poor productivity.
> The organization that pays the market wage will usually out-compete the organization which pays based on hours and seniority.

That's conjecture though isn't it? Because the organization that doesn't promote its star employees out of their range of competence will perform better and have less turnover because people won't be leaving for better opportunities elsewhere.

> Seniority is a poor proxy for productivity (theer's definitely a correlation, but it's not strong and it ignores individual variation, which is large),

Sure, obviously productivity is a consideration in whether to keep an employee, and possibly a consideration for raises. You don't want to keep any kind of non-productive employee. But while they're not productive in their current role, they might be productive in another role.

> while paying based on hours is rewarding poor productivity.

You're assuming a lot about what I meant by hours.

> The article cites an example from Sales teams.

Did they look at people other than at sales? Could it be that their results apply only to salespeople? Just wondering.

That was my first thought, too. It seems quite plausible that the exact skills suited to doing well in sales (competitiveness, focus on short-term gains, etc.) may be the exact opposite of skills suited for managing.

The article's generalization from a specific employee type to all employees is unfounded.

It's universal across organisations: you need people management skills to be a manager, while most other positions require skills specific to their domain such as being a good confidence artist to make sales, being a good logical abstract thinker to be a programmer, and so forth.

Generally speaking, promoting someone who has great technical competence into a soft skills position is a recipe for disaster: you lose both a good manger and a good technical specialist.

The fundamental flaw is perceiving management as a senior role, when management is actually a subservient role to the rest of the organisation.

To go this route you'd need to delegate substantial decision making power for technical questions. Most companies don't do that, and plan to have unicorn managers with both skill sets.

The ratio of tech to people skills depends on the difficulty of execution. In the 90s when every tech company was a "hard tech" company you needed technical representation at a high level. Bill Gates would only allow former programmers to manage his programmers for this reason.

> Maybe if we shifted our view of management as "just being a different skill set" and not always being higher paid then we'd find people without the required skills wouldn't try and obtain those positions and would focus on what they are good at.

Then how will we know who's better than everyone else and has "made it"? How am I supposed to validate myself other than by having lots of people "working under me"? The point of management is being able to tell people what to do, because they're beneath you. Also, being able to tell your friends and family you're Executive Chief Head Lead Architectural Director.

When you dont pay managers more, generally speaking people don't want to do it. Especially not well qualified people - and you don't want non qualified person there.
Managers get paid more because their benefit to the organization is that they amplify the productivity of the entire team (if they’re good). Of course there are many examples of bad managers where this doesn’t happen, but that’s the idea anyway.
You could say the same thing about Janitors or network engineers or devops, or many other positions in a company. If they're doing their jobs well, they increase the productivity of all other employees.

Managers might get paid more because they are a scare resource or for other reasons, but it's not just because they increase productivity.

Every job doesn’t get paid for the full value it provides to the company, otherwise the company wouldn’t be profitable, so saying that every job increases productivity is true to a point, but the scale of the increase is different. Only jobs that provide a significant increase in productivity or profits are the ones that benefit from higher salaries as a reward for that, like management and sales. Jobs like janitors, while they do provide some help in productivity, are also subject to market forces which keep the wages low.

And I didn’t say they get paid more just because of productivity increases.

On the other hand, that means bad managers are being paid more to do the opposite.