At this point is there any disadvantage to legislating that all companies should share a transparency report where they list their levels + avg, min, max, median salaries for each level?
Lots or reporting requirements only take effect once a firm reaches a certain size. This makes a lot of sense given that tiny firms (say less than 10 employees) may not have formalized wage scales, in-house administrative staff, and have a harder time amortizing compliance costs. So yes, there are disadvantages to consider.
I’m opposed to disclosure laws that would be prone to expose the income of individual people.
If Alice is the only SWEPR5 at a company, that proposal publishes her salary. If Bob is clearly the best SWEPR4, Cathy the middle, and Dave just barely got promoted to that level, their salaries are also individually published.
I don’t object to good faith ranges being required in job listings, but I oppose anything that would tend to publish an individual’s data that they should be able to keep private.
(That’s before even getting to the fact that salary is nowhere near the full compensation picture.)
Salary is about 90% of the compensation picture for almost everyone in the US. If you add in things like 401(k) matching and employee-paid insurance premiums, you've got about 99% of the picture covered (in addition to having something far too complex to publish on any individual level). The vast majority of companies offer no stock, no or minimal/token bonuses, and no perks with any real financial value.
> If Bob is clearly the best SWEPR4, Cathy the middle, and Dave just barely got promoted to that level ...
I think you're making a big presumption that the "clearly the best" gets the most income!
(Although the more usual argument against transparency proposals is that everyone thinks they're "clealry the best" so all except the highest paid would be annoyed.)
Good money being: any company that tries to do things properly (or, relatively properly).
Bad money being: profit, profit, profit. Take risks and misreport. Don’t put anything in writing. Pay bonuses under the table. Etc etc etc.
Not only does the good money need to spend a lot more money to comply, they also wind up wasting a lot of money on “fake” compliance or “over-“ compliance. For example, insurance companies often find ways to make companies insure the same risk multiple times. Or consultants over-analyse and say “This whole thing is a risk for your company. It needs to be fixed. I can do it for you.”
Good money will pay consultants to prepare excruciatingly detailed salary transparency reports, and they’ll live with the cost of their honesty (“you pay women less than men!!!! it doesn’t matter that your state has very few qualified women in that field!!! pay them all the same!!!”).
Bad money will do just enough to be compliant, and they’ll fudge the numbers (with sneaky fringe-benefit bonuses to their employees so they don’t care).
The information is specific to a unit and published monthly (e.g. if 6 units were rented out, there are 6 entries in the database). It’s quite helpful when looking for a place to rent because you can look up the exact building see what people are paying.
The data is obscured a bit because units are described as “50 to 70 sq m”, so while you can guess if it’s a 1 bed or 2 bed, you cant tell if it has 1 or 2 bathrooms, what floor it on, whether it comes furnished or not.
But you can look at a group of transactions that are similar to the unit you’re looking at and get of what the market rate is.
If it is legislated that everyone has to do it, then people looking to sell their labor can do so with perfect knowledge of where the market is at, reducing the negotiating strength of these shitty companies.
People in general don't like to have the hard conversations where they tell Bob that Sue is better than him and she makes more money. It's easier to just normalize transparent salaries lower. If everyone is transparent it will change over time though.
I worked for several hedge funds with transparent comp that have been some of the most lucrative of my career. If we are just throwing around anecdotes.
There are already companies collecting such information and making them available for their customers (who supposedly provide some collated data set to them in exchange). That's what all the "paying top of market rate" talk in various companies is based on.
So: the potential for collusion already exists, but unless regular folks can get access to the analysts' data sets as well, it's a very one-sided deal.
So right now all the companies do share salary data with each other. They typically wash it through a 3rd party so it's not that blatant.
So, a change where the companies also posted their salary data publicly would only additionally inform employees not HR. But yes tech-HR has in the past gotten in trouble over salary fixing.
Exactly the contrary: if all wages are public, that wouldn't be possible, because only one company would have to violate that conspiracy to get all employees at other companies to start applying to them.
No, because they need to attract employees and employees would generally not be attracted to uncompetitively low compensation packages.
As a result, companies absolutely track how much employees of a type and level get paid elsewhere, so that they know where to aim compensation packages competitively.
Would be interesting to see people rough idea of YOE to Level mapping for MSFT.
Everything from Level 50 through 60 looks pretty junior, like grad hires and possibly outside HCOL areas. Surprised there is so much granularity.
I do also find the base to stock ratios interesting.
I guess base pay caps out in FAANG the same way it does in fintech / Wall Street tech. $360k base is a lot of money obviously, but its interesting you can have like a $1M signing bonus & $1.25M TC but never get about $360K base.
For software engineers, 59 is entry level (e.g. a college hire). Promotion to 61 (SWE 2) usually happens in 2-4 years of experience. Promotion to 63 (Senior) is usually 5-7 YoE. Promotion to 65 (Principal) is highly variable but a capable person who climbed the ladder all the way from 59 and didn’t get stuck with a crappy team/manager along the way would usually get there around 9-10 years, give or take.
Level 59 is entry-level for an SE (e.g. an undergraduate hire), Level 58 is for Interns. People at Level 59 are expected to reach Level 60 within 2 years or they'll basically get managed-out, but once you get past Level 61 people generally take things more slowly.
In West Coast circles (Blind.app, et al) Microsoft's reputation is that it's still "the country club" where you go if you want a stable career without too much excitement (or stress) - if you play your cards right (and dodge the occassional round of layoffs) you can still live the baby-boomer dream having a single-employer-for-life untill you retire.
The downside is you have to put up with Microsoft's ineptitude. I was on a Microsoft project for about 6 months and it was a terrible experience. Everyone was out for themselves, asking questions was treated as an annoyance. Condescension at every turn. It was a small team and perhaps not representative of all of Microsoft, but it was a bad enough experience that I'd never consider working there in any capacity.
The base pay is effectively capped because salaries are taken out of money that is taxed. That's why CEO's used to stop at $1M. (Maybe they still do.) Everything else is stock because it's "free" money for the company to give away.
I'm not sure about Microsoft specifically, but at other companies I've worked at (including the federal government) a level number is somewhat standardized across all roles, not just engineering.
So that means an entry level engineer would start at a higher level than a senior secretary, or whatever the case is.
I've noticed that a lot of US organizations love to use "levels", but coming from the outside it's borderline impossible to know what those levels mean. We have IC levels, and I'm sure I'm on some level, but have zero clue as to what the level may be or mean.
In cases like this where you just have payment guidelines for a level 52, you might as well not know. What's level 52? Is that an intern or senior management?
Context is key - for that level, they're pretty bad numbers compared to the rest of big tech. Equivalent at Meta/Google/Amazon would be making vastly more. Microsoft usually punches well below its weight for compensation.
They return the favor with a chiller work-life balance and Washington State's lower taxes (obv not applicable to the large campus in Sunnyvale/Santa Clara)
Microsoft uses a system of levels to denote seniority. Higher levels come with higher offers, according to the guidelines. Levels also vary by field and location, but generally, "senior" starts at level 63, "principal" starts at level 65, and "partner" starts at level 68, according to insiders.
Not all types of roles go up to level 70, but in engineering, an employee at this level would be considered a "distinguished engineer.'" There's also a level 80 for a "technical fellow," which is typically one of the highest-ranking executives at the company.
I started off at MSFT over a decade ago and though I haven't worked there in quite a while I'll say that these numbers look very old to me - there's no way Level 59's base salary range is 74-110k today (e.g. about 3 years ago, a new-hire Lvl 59 friend of mine said he was on $115k base and he said he felt he was in the middle)
The publicly disclosed numbers on Levels.fyi are much closer to reality, IMO.
74k is definitely too low. In 2019, 59 was 98k. Usually goes up 1-2% each year. 115k sounds high. Each level technically has a 20% range up and down from the compa mid-point, but most offers are made dead-center. Intern return offers tend to be the juiciest 59s. In some cases college grads returning from their internships make more than those with 15-20 years of experience at 64. It's wild.
Am a manager at Microsoft so I can see where the policy-based midpoints are once regional adjustments are made. The base numbers look low by a rough 20% compared to what a Redmond HQ dev would get paid at the given level number. Bonuses are based on a level-dependent target percentage which you can find if you search for it.
And yeah, the Levels.fyi base pay distributions are closer to the truth.
https://archive.ph/k58ID
At this point is there any disadvantage to legislating that all companies should share a transparency report where they list their levels + avg, min, max, median salaries for each level?
Lots or reporting requirements only take effect once a firm reaches a certain size. This makes a lot of sense given that tiny firms (say less than 10 employees) may not have formalized wage scales, in-house administrative staff, and have a harder time amortizing compliance costs. So yes, there are disadvantages to consider.
Smaller companies can just report whatever they're paying each employee that year, then?
12 replies →
I’m opposed to disclosure laws that would be prone to expose the income of individual people.
If Alice is the only SWEPR5 at a company, that proposal publishes her salary. If Bob is clearly the best SWEPR4, Cathy the middle, and Dave just barely got promoted to that level, their salaries are also individually published.
I don’t object to good faith ranges being required in job listings, but I oppose anything that would tend to publish an individual’s data that they should be able to keep private.
(That’s before even getting to the fact that salary is nowhere near the full compensation picture.)
Did you know there are some countries where you can look up anyone's income?
33 replies →
Management is the only one who benefits when labor doesn’t openly share salary information. Let the information flow. Knowledge is power.
19 replies →
Why does that matter? They should all be able to know each other's salaries.
7 replies →
Transparency also helps to find out that you the foreigner gets paid less
Why shouldn't Dave and Cathy know what Bob is earning? They're the same level, they should earn the same.
31 replies →
Salary is about 90% of the compensation picture for almost everyone in the US. If you add in things like 401(k) matching and employee-paid insurance premiums, you've got about 99% of the picture covered (in addition to having something far too complex to publish on any individual level). The vast majority of companies offer no stock, no or minimal/token bonuses, and no perks with any real financial value.
> If Bob is clearly the best SWEPR4, Cathy the middle, and Dave just barely got promoted to that level ...
I think you're making a big presumption that the "clearly the best" gets the most income!
(Although the more usual argument against transparency proposals is that everyone thinks they're "clealry the best" so all except the highest paid would be annoyed.)
It's already the case for H1Bs.
I get more accurate date about Alice but levels.fyi already gave me a good ballpark
[dead]
Why would it matter if Alice was the only one or not? Disclosure laws inherently expose the income of everyone
18 replies →
This significantly disadvantages good money.
Good money being: any company that tries to do things properly (or, relatively properly).
Bad money being: profit, profit, profit. Take risks and misreport. Don’t put anything in writing. Pay bonuses under the table. Etc etc etc.
Not only does the good money need to spend a lot more money to comply, they also wind up wasting a lot of money on “fake” compliance or “over-“ compliance. For example, insurance companies often find ways to make companies insure the same risk multiple times. Or consultants over-analyse and say “This whole thing is a risk for your company. It needs to be fixed. I can do it for you.”
Good money will pay consultants to prepare excruciatingly detailed salary transparency reports, and they’ll live with the cost of their honesty (“you pay women less than men!!!! it doesn’t matter that your state has very few qualified women in that field!!! pay them all the same!!!”).
Bad money will do just enough to be compliant, and they’ll fudge the numbers (with sneaky fringe-benefit bonuses to their employees so they don’t care).
Nice theory but this hasn’t been borne out in countries where wage transparency is already the law.
As an aside, rent is public information in Singapore, at least for private rental (not for rentals of public housing).
Since all tenancies need to pay a stamp tax to be legit contracts, the government lists every contracts rental price on a website.
https://www.ura.gov.sg/property-market-information/pmiReside...
The information is specific to a unit and published monthly (e.g. if 6 units were rented out, there are 6 entries in the database). It’s quite helpful when looking for a place to rent because you can look up the exact building see what people are paying.
The data is obscured a bit because units are described as “50 to 70 sq m”, so while you can guess if it’s a 1 bed or 2 bed, you cant tell if it has 1 or 2 bathrooms, what floor it on, whether it comes furnished or not.
But you can look at a group of transactions that are similar to the unit you’re looking at and get of what the market rate is.
Maybe the region would have more women in the field if the companies were transparent about pay and benefits?
Every company I've talked to with "transparent" comp has offered absolutely terrible comp, even for their highest possible offer.
In reality "transparent comp" becomes an excuse for companies to not negotiate.
If it is legislated that everyone has to do it, then people looking to sell their labor can do so with perfect knowledge of where the market is at, reducing the negotiating strength of these shitty companies.
This is how it already works in Norway. https://www.bbc.com/news/magazine-40669239
10 replies →
People in general don't like to have the hard conversations where they tell Bob that Sue is better than him and she makes more money. It's easier to just normalize transparent salaries lower. If everyone is transparent it will change over time though.
2 replies →
I worked for several hedge funds with transparent comp that have been some of the most lucrative of my career. If we are just throwing around anecdotes.
Wouldn’t all the companies then just conspire to pay everyone lower wages?
There are already companies collecting such information and making them available for their customers (who supposedly provide some collated data set to them in exchange). That's what all the "paying top of market rate" talk in various companies is based on.
So: the potential for collusion already exists, but unless regular folks can get access to the analysts' data sets as well, it's a very one-sided deal.
4 replies →
Wait till you find out what they do under the current system where wages are secret information
https://en.wikipedia.org/wiki/High-Tech_Employee_Antitrust_L...
So right now all the companies do share salary data with each other. They typically wash it through a 3rd party so it's not that blatant.
So, a change where the companies also posted their salary data publicly would only additionally inform employees not HR. But yes tech-HR has in the past gotten in trouble over salary fixing.
Exactly the contrary: if all wages are public, that wouldn't be possible, because only one company would have to violate that conspiracy to get all employees at other companies to start applying to them.
No, because they need to attract employees and employees would generally not be attracted to uncompetitively low compensation packages.
As a result, companies absolutely track how much employees of a type and level get paid elsewhere, so that they know where to aim compensation packages competitively.
1 reply →
They're way ahead of you! https://en.wikipedia.org/wiki/High-Tech_Employee_Antitrust_L...
Wouldn't all the employees then just conspire to unionize around the highest paid employee?
EDIT: forgot to mention I'm not advocating for it.
It would cause wage growth leading to a lower percent (but an increase) increase in other prices (housing, etc).
So things that are happening regardless?
5 replies →
No?
This is such a random comment I don't know how to respond to besides a "No?"
3 replies →
Would be interesting to see people rough idea of YOE to Level mapping for MSFT. Everything from Level 50 through 60 looks pretty junior, like grad hires and possibly outside HCOL areas. Surprised there is so much granularity.
I do also find the base to stock ratios interesting. I guess base pay caps out in FAANG the same way it does in fintech / Wall Street tech. $360k base is a lot of money obviously, but its interesting you can have like a $1M signing bonus & $1.25M TC but never get about $360K base.
For software engineers, 59 is entry level (e.g. a college hire). Promotion to 61 (SWE 2) usually happens in 2-4 years of experience. Promotion to 63 (Senior) is usually 5-7 YoE. Promotion to 65 (Principal) is highly variable but a capable person who climbed the ladder all the way from 59 and didn’t get stuck with a crappy team/manager along the way would usually get there around 9-10 years, give or take.
Level 59 is entry-level for an SE (e.g. an undergraduate hire), Level 58 is for Interns. People at Level 59 are expected to reach Level 60 within 2 years or they'll basically get managed-out, but once you get past Level 61 people generally take things more slowly.
In West Coast circles (Blind.app, et al) Microsoft's reputation is that it's still "the country club" where you go if you want a stable career without too much excitement (or stress) - if you play your cards right (and dodge the occassional round of layoffs) you can still live the baby-boomer dream having a single-employer-for-life untill you retire.
The downside is you have to put up with Microsoft's ineptitude. I was on a Microsoft project for about 6 months and it was a terrible experience. Everyone was out for themselves, asking questions was treated as an annoyance. Condescension at every turn. It was a small team and perhaps not representative of all of Microsoft, but it was a bad enough experience that I'd never consider working there in any capacity.
2 replies →
> are expected to reach Level 60 within 2 years or they're automatically out-the-door
That part isn't true. Some people stay at L59 for much longer and there isn't any "automatic out", at least in the US.
The base pay is effectively capped because salaries are taken out of money that is taxed. That's why CEO's used to stop at $1M. (Maybe they still do.) Everything else is stock because it's "free" money for the company to give away.
This isn't true if they are restricted stock units (RSUs) which are taxed as income yearly even if you do not sell
Anyone know why they start at level 52?
Is there some story behind 52-70 instead of 1-19? Or is it just arbitrary?
I'm not sure about Microsoft specifically, but at other companies I've worked at (including the federal government) a level number is somewhat standardized across all roles, not just engineering.
So that means an entry level engineer would start at a higher level than a senior secretary, or whatever the case is.
I've noticed that a lot of US organizations love to use "levels", but coming from the outside it's borderline impossible to know what those levels mean. We have IC levels, and I'm sure I'm on some level, but have zero clue as to what the level may be or mean.
In cases like this where you just have payment guidelines for a level 52, you might as well not know. What's level 52? Is that an intern or senior management?
https://www.levels.fyi does a good job of showing how levels compare across companies.
Not sure if there's an actual reason. But 59 is the actual starting level that matters for software developers.
>On-hire stock awards: $310,000 default to $1.2 million with approval
>Annual stock award range: $0 to $945,000
Wow. Those numbers are huge
Context is key - for that level, they're pretty bad numbers compared to the rest of big tech. Equivalent at Meta/Google/Amazon would be making vastly more. Microsoft usually punches well below its weight for compensation.
They return the favor with a chiller work-life balance and Washington State's lower taxes (obv not applicable to the large campus in Sunnyvale/Santa Clara)
There are only a few hundred people at this level. Maybe like 200-300 people.
From the article:
---
Level 70:
Level 69:
Level 68:
Level 67:
Level 66:
Level 65:
Level 64:
Level 63:
Level 62:
Level 61:
Level 60:
Level 59:
Level 58:
Level 57:
Level 56:
Level 55:
Level 54:
Level 53:
Level 52:
I'll add this from the article:
Microsoft uses a system of levels to denote seniority. Higher levels come with higher offers, according to the guidelines. Levels also vary by field and location, but generally, "senior" starts at level 63, "principal" starts at level 65, and "partner" starts at level 68, according to insiders. Not all types of roles go up to level 70, but in engineering, an employee at this level would be considered a "distinguished engineer.'" There's also a level 80 for a "technical fellow," which is typically one of the highest-ranking executives at the company.
Other than the really granular levels, this is typical for a large corporation.
I started off at MSFT over a decade ago and though I haven't worked there in quite a while I'll say that these numbers look very old to me - there's no way Level 59's base salary range is 74-110k today (e.g. about 3 years ago, a new-hire Lvl 59 friend of mine said he was on $115k base and he said he felt he was in the middle)
The publicly disclosed numbers on Levels.fyi are much closer to reality, IMO.
74k is definitely too low. In 2019, 59 was 98k. Usually goes up 1-2% each year. 115k sounds high. Each level technically has a 20% range up and down from the compa mid-point, but most offers are made dead-center. Intern return offers tend to be the juiciest 59s. In some cases college grads returning from their internships make more than those with 15-20 years of experience at 64. It's wild.
Yeah, I was thinking the same. I came in to level 59 in 2008 and my base was 80k-ish. These numbers seem really old.
Am a manager at Microsoft so I can see where the policy-based midpoints are once regional adjustments are made. The base numbers look low by a rough 20% compared to what a Redmond HQ dev would get paid at the given level number. Bonuses are based on a level-dependent target percentage which you can find if you search for it.
And yeah, the Levels.fyi base pay distributions are closer to the truth.
2 replies →
What is level 59 equivalent to in terms years of experience?
1 reply →
These are the old ranges before the bump in 2022.
These seem even older than that. These are what the ranges looked like when I joined in 2008...
Stock ranges look accurate. Base numbers I'm unsure about but the base change more frequently than the stock ranges.
Still peanuts but you don't work at MS to maximize $$$ but rather for WLB
Yeah, folks on Blind reporting that these are not accurate.
I want to know the salaries happening at level 80!
(I also want to know wtf is happening with their levelling system in general)
It's pay-walled, bruh.
Try https://dataconomy.com/2023/10/17/leaked-microsoft-pay-guide...
You can usually look for archive links in the comments that will get you past the paywall. Someone else posted this: https://archive.ph/k58ID
The archive post was after GP’s question.
That is a weird CEO picture. Is it AI generated?
That's a lot of levels...
Wasn’t this pretty much already known via https://levels.fyi ?