This is cool. I was less interested in the ratio and more interested in the median pay at places. Some places I’ve never heard of have a significantly higher median pay than my company! And I’m pretty surprised at how high median pay is at FAANG
Seems like a bad way to measure. You're heavily penalizing CEOs who lead companies with low median wages.
If I wanted to see if CEO wages were unjustifiably low or high, I'd look at things like total revenue, profit, stock price growth, etc. Then you could at least tell if pay is linked to performance.
> You're heavily penalizing CEOs who lead companies with low median wages.
I have a feeling that this is the entire point.
edit: to be clear, I disagree with it because it's a terrible way to measure what I think OP is trying to measure. As another commenter mentioned, this simple ratio doesn't tell the whole story.
It doesn't really prove anything. No one expects Walmart to pay the average employee as much as Facebook. Makes for some political rhetoric but nothing insightful
The minimum wage hasn't increased since the 80s (when you take the inflation adjusted numbers). The lower the median wage, the higher likelihood of tax payers subsidizes for the lower "half" at the direct benefit of upper "half."
Not saying this is the only point just that outsourcing isn't the only issue here.
I would go a step further and say this is an actively harmful way to measure CEO pay since it further incentivizes reducing the number of low paid full time employees in favor of contract workers or complete outsourcing of that work. Both those options would likely lead to worse overall compensation for low paid employees.
So whether or not a CEO is "overpaid" depends on their business model?
Hire a part-time janitor at $15k and you look like Mr Burns, but out-source it so your next lowest paid employee is making $150k and you look like a saint.
But nothing's change with how much the CEO is paid.
I think what you're missing is that this particular metric doesn't care about performance of the company or if the CEO "deserves" that much money or not, and that's by design. When people compare the pay of a CEO to the pay of their employees it's typically because the entire point of the comparison is to to draw judgement on the CEO's business model.
For example, a common example where this metric is used is to cast judgement on the owners of Walmart for being multi-billionaires while their employees make near minimum wage. The point is not to say anything about the specific performance of those owners or of the employees (and often intentionally ignores it as a factor), but rather to specifically call attention to the disparity in distribution of wealth within the company and to cast judgement on the business model where one person gets rich while depending on the work of people who remain poor.
Again, I don't agree with the use of this tactic to measure anything. I'm just explaining why its done.
I would say that a CEOs salary should be based on: 1) the level of responsibility (how big is the company), 2) the CEOs performance (is the CEO improving performance).
Not a bad metric. Of course there's alot of ways to slice it, but if you compare the millionaire landlord of say low income housing vs the owner of high income housing - same income - people do look down on the 'slum lord'.
Someone making $10m per year paying poverty wages (infer they are getting rich by paying poverty level wages) is going to be judged differently by people then the CEO paying $200k median.
It would be great if there were a clear definition of what "pay" means. Sundar Pichai's pay includes his $240m stock package granted in 2019. That's a truckload of stock but is that to be considered pay?
https://www.theverge.com/2019/12/20/21031629/google-ceo-sund...
Why wouldn't it be? Would you consider Steve Jobs' "pay" back when he was running Apple for the second time to be $1? I consider Jack Dorsey's entry to be an error, or a case of incomplete data.
Apples to oranges. You should be considering salary to salary.
If shareholders want to pay their CEO more directly from their pockets in the form of dilution, it doesn't hurt or affect the standard employee in any way.
And how does the CEO salary hurt the standard employee? Most of the S&P 500 companies employ around 100k of people. dividing $20M salary equally between them would give each of them $16.6 per month.
You are correct. It doesn't. The entire CEO pay controversy is a wedge issue. The only people who should be complaining about CEO pay are shareholders.
A CEO gets paid in stock, which the company later buys back with cash. Ergo, the CEO getting stock compensation reduced the amount of money available to pay workers.
GAAP requires companies to report stock compensation as an expense because it does actually move wealth around in a zero-sum fashion. It’s not a magic free money tree.
No, the buybacks do. Youre making it seem as if the only way for a CEO to sell stock is if the company buys it back.
And it's specifically called a non-cash expense, because it doesn't affect cash (which salary does). You are correct in that it isn't a free money tree. It's paid for by shareholders. If the shareholders wish to pay their CEO $X, the government has no reason to intervene. It doesn't make the employees any better or worse off.
The $1.7million reflects not the increase in Bezos' net worth, but the cost of Bezos' employment to the company (salary + security and transportation). This is what actually shows up in expenses in Amazon's financials.
AFAIK, Jeff hasn't gotten a stock grant since the IPO in 1997, so the vast majority of his net worth gains are purely capitol gains from holding amazon stock for over 20 years. Since there have not been additional stocks granted, this increase in Bezos' wealth actually costs Amazon nothing, not even dilution, as it's just an increase of value of existing shares.
Get this, I just found out Google hires for a role called "Application Engineer" which is a role where you program but are lower class and lowered paid than regular "software engineers."
Google might not agree with the "lower class" part but the fact that for an application engineer to switch roles to software engineering within google requires them to go through a harder interview process indicates a deliberate caste based ranking structure that limits growth based off of your ability to pass an algorithm interview.
I actually believe it's 100% within google's rights to do something like this, but I think it's very bad that they disguise this ranking with the role name. I know people who didn't realize what they were getting into until they joined google.
Seriously, just name the role for what it is: High IQ Software Engineers vs. Average IQ Software Engineers because that's essentially what they're interviewing for.
There is nothing wrong with this. I am saying what is wrong is that people aren't made aware that this is what the position entails. You are being hired on as a lower class engineer.
I've had friends hired without being aware of this fact. Literally great software engineers at other companies with google fully aware of their previous title than suddenly unknowingly delegated to a lower class position with google not elucidating the difference.
It's like a software engineer applying for a software technician job. And you as the hiring manager not mentioning anything because you know the person applying isn't yet fully aware of the difference. These roles are literally targetting people who have software engineer in their titles but can't make the cut of the normal google interview.
It should be made clear that there is a difference between a technician and an engineer... just like it needs to be made clear that an application engineer is not a software engineer at google.
Also I would argue that the expectations are generally the same. You are spending your days programming there's really no difference.
Additionally it's an elitist attitude on googles part. Basically their saying that everyone who is normally considered a software engineer at other places aren't qualified to be software engineers at google. The only people considered to be software engineers are the people who pass that interview which is less than 1% of all people who apply. So they funnel these people into a lower caste that can never move on to be a software engineer EVEN when performance proves otherwise.
A lot of companies with <$15k median pay as well. Seeing as US federal minimum wage is ~$15k (higher than that in more than half of states) and there are companies listed with even <$10k median pay, either part-time or outsourced workers are being included, neither of which seems like a fair comparison to make.
Don't know about outsourced, but part-time isn't a fair comparison since their hourly wage could be comparatively high but their total wage low. It would be fair if salaries were given per-hour, but I'm pretty sure that top executives don't have the same concept of working hours as low level employees
If it’s not normalized for time worked, then it’s not really pay ratio that represents the value a company sees in an employee, or the economic comfort of that employee. So, what’s it really trying to show?
Some jobs don’t require full time work, like consultation or janitorial services, and these jobs are structured around this, by having multiple clients. Trying to take any direct meaning from these ratios isn’t really possible since another dimension is required for these numbers to have any real use.
Because it must include labor in other countries where the pay scale is very different. A barista at Starbucks in NYC is going to have a different income than one in Vietnam. By making these all US companies, you are comparing the CEO pay of an American against an average of global norms which are far lower.
It would be useful (to me) to include another column that's just the median pay of US-based employees.
Another interesting tact would be to show what a Vietnamese CEO makes versus their company's median pay, or what this looked like in the US in 1950, for example.
Does Netflix have full-time employees making content? Or do they just outsource to other studios? It takes a lot of people to make shows/movies, and not all of them are paid especially well. Could easily outnumber the tech employees.
Is Jack Dorsey paid any more than any other billionaire who owns a few Twitter shares? He could choose to not work for Twitter (some people might say he already has...) and still earn the same from his Twitter holdings, so it seems reasonable to say his compensation for being Twitter CEO is just the $1.40/year. And he "only" owns 2% of Twitter, so it's definitely reasonable that some other rich person could have the same size shareholding as a passive investment.
if you're being pedantic? no, but it's obvious that if you massively frontload your effective compensation in the form of stock in your company and employees coming in will never make as much on the stock as you do, that is still a gap in compensation. it doesn't pass the smell test for most people. WeWork is a perfect example of this
I would say that the overwhelming thing that separates Jack Dorsey from Twitter employees financially is that he is massively wealthy, not that he earns more for working there. If he founded a new small startup today, valued at a nominal amount, and paid all employees including himself $75k/year, wouldn't there be the same inequality? You'd hardly argue that he should need to light on fire $100mm/year for the privilege of working at the new startup, in order to make his compensation somehow the same as the other employees.
Edit: Seeing a lot of people complaining about the data quality. A lot of the more interesting breakdown on salary vs stock etc was only available by clicking on each row. Maybe a more advanced scrape could reveal that stuff.
This is a great idea and the Google Sheet share is a great format, but there are many flaws in the representation of the data. A great, immediate change would be to add a row at the top indicating some of the flaws, and then start making incremental changes to fix them (e.g. create a column for stock based compensation).
Cause overall this is a great topic, just needs some work so people can focus on the topic at hand rather then get bogged down in the data quality issue.
Appreciate it. Meant it mostly as a "Wow didn't know X company had 1000 to 1 CEO to median pay ratio". I'll see if I can get some more data / detail on what goes into the pay numbers.
Some of these numbers seem to be off. For example, Harley Davidson has a median pay of $187k? Where is that coming from? They've got 5000 employees... I assure you that 2500 of them aren't making more than $187k.
Meanwhile at the other end of the spectrum is Gap, with a reported median pay of ~$6k. This seems very low. At first I thought that maybe the overwhelming of their employees are oversees working in factories, but then I looked up their SEC filing and they don't manufacture their clothes. I'm assuming they have a bunch of part-time employees, but even then, $6000/year assuming $10/hour means employees are working an average of 11 hours per week, which still seems low.
This is a bad way to do this. There's executives with multiple c-suite roles, compensation for multi-year performance targets, founders who elect to take low salaries and company size disparities.
Contrarian viewpoint: Having a CEO make ~1,000x the median wage doesn't seem that practically wrong to me, seen from a lens of "At most of these places, if the CEO messes up many, many thousands of people will lose their jobs." Ie, if one thinks of "right" compensation as being proportional to impact, not to skill or job difficulty.
Another way to think about this is how much the CEO is actually taking vs others. If the Aptiv CEO worked for free, giving his $15 million pear year to the 147k employees that would only be $100/year per employee. Meaningless.
Aptiv spends $1 billion per year on salaries.
So the very "worst CEO" (by this metric) receives 1.40% of all salary paid by the company. Google's CEO receives is 1%.
I think this explains why CEOs make so much relative to an individual. Paying them 1% of all salaries feels, and is, very cheap given how important their role is.
If Steve Jobs had been paid 50% of all salary it would have been worth it to shareholders. Of course lots of bad CEOs get paid huge amounts as well but why is this bad? The salary they received is not what hurts the company, it only cost around 1%. What really hurts is the damage a bad CEO does to the company. That can cost the entire resources of the company.
I like how most of the 6 figure employee salaries are clustered around the lower ratios - evidencing that top employee pay isn't at all dependent on top CEO salaries.
79 comments
[ 5.4 ms ] story [ 177 ms ] thread"according to a Wall Street Journal analysis of data from MyLogIQ and Institutional Shareholder Services"
https://www.sec.gov/news/pressrelease/2015-160.html
If I wanted to see if CEO wages were unjustifiably low or high, I'd look at things like total revenue, profit, stock price growth, etc. Then you could at least tell if pay is linked to performance.
I have a feeling that this is the entire point.
edit: to be clear, I disagree with it because it's a terrible way to measure what I think OP is trying to measure. As another commenter mentioned, this simple ratio doesn't tell the whole story.
Why is it more ethical to outsource low paying work to vendors to get it off payroll?
Not saying this is the only point just that outsourcing isn't the only issue here.
Hire a part-time janitor at $15k and you look like Mr Burns, but out-source it so your next lowest paid employee is making $150k and you look like a saint.
But nothing's change with how much the CEO is paid.
I think what you're missing is that this particular metric doesn't care about performance of the company or if the CEO "deserves" that much money or not, and that's by design. When people compare the pay of a CEO to the pay of their employees it's typically because the entire point of the comparison is to to draw judgement on the CEO's business model.
For example, a common example where this metric is used is to cast judgement on the owners of Walmart for being multi-billionaires while their employees make near minimum wage. The point is not to say anything about the specific performance of those owners or of the employees (and often intentionally ignores it as a factor), but rather to specifically call attention to the disparity in distribution of wealth within the company and to cast judgement on the business model where one person gets rich while depending on the work of people who remain poor.
Again, I don't agree with the use of this tactic to measure anything. I'm just explaining why its done.
Which would make CEOs who outsource lower wage work to vendors and contractors look better.
Someone making $10m per year paying poverty wages (infer they are getting rich by paying poverty level wages) is going to be judged differently by people then the CEO paying $200k median.
If shareholders want to pay their CEO more directly from their pockets in the form of dilution, it doesn't hurt or affect the standard employee in any way.
(Many do in fact. Carl Icahn for instance.)
GAAP requires companies to report stock compensation as an expense because it does actually move wealth around in a zero-sum fashion. It’s not a magic free money tree.
And it's specifically called a non-cash expense, because it doesn't affect cash (which salary does). You are correct in that it isn't a free money tree. It's paid for by shareholders. If the shareholders wish to pay their CEO $X, the government has no reason to intervene. It doesn't make the employees any better or worse off.
https://www.marketwatch.com/story/amazon-ceo-jeff-bezos-has-...
AFAIK, Jeff hasn't gotten a stock grant since the IPO in 1997, so the vast majority of his net worth gains are purely capitol gains from holding amazon stock for over 20 years. Since there have not been additional stocks granted, this increase in Bezos' wealth actually costs Amazon nothing, not even dilution, as it's just an increase of value of existing shares.
Google might not agree with the "lower class" part but the fact that for an application engineer to switch roles to software engineering within google requires them to go through a harder interview process indicates a deliberate caste based ranking structure that limits growth based off of your ability to pass an algorithm interview.
I actually believe it's 100% within google's rights to do something like this, but I think it's very bad that they disguise this ranking with the role name. I know people who didn't realize what they were getting into until they joined google.
Seriously, just name the role for what it is: High IQ Software Engineers vs. Average IQ Software Engineers because that's essentially what they're interviewing for.
I've had friends hired without being aware of this fact. Literally great software engineers at other companies with google fully aware of their previous title than suddenly unknowingly delegated to a lower class position with google not elucidating the difference.
It's like a software engineer applying for a software technician job. And you as the hiring manager not mentioning anything because you know the person applying isn't yet fully aware of the difference. These roles are literally targetting people who have software engineer in their titles but can't make the cut of the normal google interview.
It should be made clear that there is a difference between a technician and an engineer... just like it needs to be made clear that an application engineer is not a software engineer at google.
Also I would argue that the expectations are generally the same. You are spending your days programming there's really no difference.
Additionally it's an elitist attitude on googles part. Basically their saying that everyone who is normally considered a software engineer at other places aren't qualified to be software engineers at google. The only people considered to be software engineers are the people who pass that interview which is less than 1% of all people who apply. So they funnel these people into a lower caste that can never move on to be a software engineer EVEN when performance proves otherwise.
Never is it mentioned that this is a position that is not a software engineer and is paid less than a software engineer.
Like do I need to ask whether a backend engineer is a software engineer? If not why do I need to ask this about an Application Engineer?
This role is where google puts most of their contractors.
Warren is paid in growth of his net worth.
What is the validity of this data set? Netflix Median Pay being lower than Google's doesn't sit right with me.
Broadcom's Median Pay is $354k? Impossible.
Again, I'm not knocking the idea as it's a good thought exercise but it's important to be realistic and factual when coming up with conclusions.
Some jobs don’t require full time work, like consultation or janitorial services, and these jobs are structured around this, by having multiple clients. Trying to take any direct meaning from these ratios isn’t really possible since another dimension is required for these numbers to have any real use.
It would be useful (to me) to include another column that's just the median pay of US-based employees.
Another interesting tact would be to show what a Vietnamese CEO makes versus their company's median pay, or what this looked like in the US in 1950, for example.
Edit: Seeing a lot of people complaining about the data quality. A lot of the more interesting breakdown on salary vs stock etc was only available by clicking on each row. Maybe a more advanced scrape could reveal that stuff.
Cause overall this is a great topic, just needs some work so people can focus on the topic at hand rather then get bogged down in the data quality issue.
Second biggest surprise is Valero. I imagine they contract out most of the work, so have few employees. (Correction welcome.)
Alphabet at $259k seems high. I wonder if the median includes any actual Google/Youtube/etc. employees, or only Alphabet management.
Meanwhile at the other end of the spectrum is Gap, with a reported median pay of ~$6k. This seems very low. At first I thought that maybe the overwhelming of their employees are oversees working in factories, but then I looked up their SEC filing and they don't manufacture their clothes. I'm assuming they have a bunch of part-time employees, but even then, $6000/year assuming $10/hour means employees are working an average of 11 hours per week, which still seems low.
Aptiv spends $1 billion per year on salaries.
So the very "worst CEO" (by this metric) receives 1.40% of all salary paid by the company. Google's CEO receives is 1%.
I think this explains why CEOs make so much relative to an individual. Paying them 1% of all salaries feels, and is, very cheap given how important their role is.
If Steve Jobs had been paid 50% of all salary it would have been worth it to shareholders. Of course lots of bad CEOs get paid huge amounts as well but why is this bad? The salary they received is not what hurts the company, it only cost around 1%. What really hurts is the damage a bad CEO does to the company. That can cost the entire resources of the company.