The author forgets to mention that using genetic information for hiring decisions is illegal. It has been since 2008 when Bush signed the Genetic Information Nondiscrimination Act.
Gattaca[1] is a fun sci-fi movie depicting a world where gene selection is predominant and workforce selection is based on genetics. It's pretty entertaining, and not a half bad look at the problems of that society.
It is ominously-possible, in limited, but plausible, degrees. Our genomes will no doubt end up being used for indirect discrimination in dating, friendships, child care, school acceptance, certain military/clandestine work, frat/sororities, social clubs, exclusive event invitations, [insert activity here]. It's already used in healthcare and gamete donation.
Quite, and think how for some people their criminal record already serves this purpose. There was a nauseating story on my local news last night about a 12 or 13 year old autistic kid who was placed under arrest by police officers who contract with the school district for the hideous crime of scraping his initials on the sidewalk concrete of his middle school campus with a handy stone: http://www.nbcbayarea.com/investigations/Thousands-of-Bay-Ar...
(sorry about the autoplay video, but there's a full text transcript at the link)
The notion that a moment of juvenile whimsy is now a permanent part of a person's digital footprint and could turn up in background checks for the remainder of their life is hideously Orwellian; we are busily crafting the sociopolitical infrastructure of a totalitarian state at the same time as we have 'serious' candidates for the highest national political office expressing unbounded enthusiasm for running such an entity.
Evil is not some spiritual or abstract concept, it's the banality and legitimization of common-place BS injustices, spanning the gamut of great to small.
Great article. I don't think we will ever stop in finding ways of measuring productivity. In a sense, measurement is fundamental as it helps us determine the truth. Look at how important measurement is in the sciences and in engineering.
But the prospect of an employer using previously collected data to make an employment decisions seems rather scary. People do a lot of things when they're young and immature and it seems paranoid to rely on that information when they would probably grow out of it. How accurately can past information predict future productivity? I feel like we can't (yet) measure the effects of external factors on human nature, such as new ideas, new habits, new relationships and so on.
Yup. But as a competing future employer--- I'm okay with an employer making that kind of mistake. I'm tolerant of a variety of life circumstances that the curmudgeon employers might not be, and I'd be happy to have an easier time hiring those people.
It would be a problem is 99.9% of employers used the same metrics for hiring, because it would just exclude whole groups of people.
> I don't think we will ever stop in finding ways of measuring productivity. In a sense, measurement is fundamental as it helps us determine the truth. Look at how important measurement is in the sciences and in engineering.
Beware of cargo-cult :). I'm not saying it to you personally - but I can imagine some manager or marketer saying those words and meaning that "if measuring is so important in science and engineering then it must be good, therefore let's use measuring!". Measurements are only as good as the questions you want them to answer; ask about wrong things, and even the best data will lead you astray.
If we accept Cowen's premise regarding what constitutes value, Kant and Hegel were crappy authors because they had a small number of pageviews. We probably should have valued their contributions less; it's great we have such objective measures now, and we probably won't make that mistake again.
From the article:
Today a digital media company knows exactly how many people are reading which articles for how long, and also whether they click through to other links. The exactness and the transparency offered by information technology allow us to measure value fairly precisely.
There are two different logics at play here, with incommensurate notions of value. The firm's primary goal is to make money; as such a story's value is largely a function of page views (or click-through on ads, etc.) However, we can also assess the value of some produce (article, etc.) at a longer-term societal level. In this case a good story is one that helps policymakers reconceptualize an issue or foment some sort of social change (e.g. a New Yorker article like this one [1]). Within the language of free-market economics these longer-term positive effects might be called "externalities" or "intangibles"; intuitively they are much harder to measure, especially with capital flows. But think about it— do we really want to incentivize firms, and by proxy employees, to maximize value as measured by immediate capital flows?
More generally, there isn't a natural law that says that compensation should be proportional to output—quality of life may well be better even at the top if this relationship is somewhat dampened. The cognitive cost may be lower, for example, of not continually switching jobs based on market conditions (this is certainly true in the limit) or of reducing anxiety about losing the ability to produce (e.g. health and psychological wellbeing). These are issues we need to think about and discuss at length, not take as premises.
More generally, there isn't a natural law that says that compensation should be proportional to output.
There's just a natural tendency of a company to want it that way since this maximizes their profits. Once regulations and conventions held-in-check the tendency to profit from the exact measurement and control of workers. Those hindrances seem to have broken down and the advance of fluid financial capital and technology seems like a freight train with enough momentum to batter down any and all limits to their advancement regardless of our opinions on the subject.
Kant and Hegel were not journalists. In the context of philosophy, value (qua influence) is more easily quantified by the breadth and depth of citation networks, which are already the object of some study in academic and legal circles.
I agree with your overall argument that we should be careful about which premises we begin with, as they can easily become axiomatic and stifle thinking (a perennial problem in politics). But I think Cowan in choosing his premises here to illustrate how technology that liberates capital flows has the potential to undermine capitalism by limiting people's ability to freely participate in it.
Of course such observations summon the ghost of Karl Marx to whisper 'I told you so.' I wonder how many thoughtful observations of that philosopher are overlooked because of a misplaced reification of his labor theory of value.
I forgot to mention that I think the labor theory of value is a complete crock and I can't understand why the radical left keeps punching itself in the face with it.
I'm not sure I see much compelling here. First off, all he's saying is that technology provides more information about the market. But technology has been doing that forever. Second, he's chosen a very simple example that has a measurement that maps directly to a single person and their economic value (journalists and how often their material is read). Try to apply that to teams (like software teams) in which a productivity is created by a diversity of personalities and specialty skill sets that are focused collectively on a problem - things get muddy real fast when you're trying to measure individual productivity in that environment. And that environment is where to the truly high value work is being done the modern global economy.
You won't believe these ten tips for being a more effective journalist, #6 will blow your mind. Doctors hate him! Learn this one weird metric busting trick.
Yeah, i don't really know how people get stuff done. Metrics can be nice bumpers, to go see if someone needs help. But actual productivity is pretty nebulous. (to me anyway)
This is an interesting article, yet I believe it largely misses the issues people have with income inequality. The article seems to focus on pay discrepancies between people with the same role, missing a lot of the big picture.
Completely missing from the article are any references to differences between jobs with easily measurable output and those with difficult to measure value.
Software engineers and management are two great examples of professions where value is very difficult to measure. In the opinion of many, software engineers tend to be underpaid compared to the value they produce. Managers on the other hand, tend to be viewed as overpaid or at least better compensated compared to the value the produce. I believe that managers do a better job of capturing their value primarily because they (or someone just like them a step up the chain) hold the purse strings.
Back to what the article focuses on for a moment. Sure a superior producer might be able to capture more of their value than their low-producing counterpart. That's a drop in the bucket compared to C-level compensation vs salaryman pay.
In summation, people aren't upset that the guy in the next cubicle makes 20% more money because he's more productive. They're upset that the CEO makes 100x more than they do even in bad times for the company.
>software engineers tend to be underpaid compared to the value they produce
I think comparing salary to the value produced is a slippery slope. Software engineers are certainly not paid below the average or median salary. Their "value" is just a function of the job, not a measure of its worth.
I understand your point, but there's a glaring error in what you said. People need to get paid less than the value they produce, otherwise their employer would just be a passive participant gaining no advantage from their employment.
If the employers produce no value, then we should get rid of them. If they are producing value, then the overall value produced should be greater than the sum of the value produced by their employees.
That's only true if you believe the you can add up the value every employee produces to get the value the company produces. But the company produces more than the sum of its parts. Or should, anyway.
That's why efficient markets are only considered to be an ideal.
Note in edit: I suspect what I remember as the Econ 101 explanation is probably accurate enough for most purposes: Both the employer and the worker have temporary, local monopolies: The employer has plenty of money, but a shortage of time, and the worker has plenty of time but no money. So each is willing to trade with the other. The employer could make money in excess of the value added by the employees, by exploiting other kinds of effective monopolies such as patents, trade secrets, even things like brand recognition and goodwill.
> The employer could make money in excess of the value added by the employees, by exploiting other kinds of effective monopolies such as patents, trade secrets, even things like brand recognition and goodwill.
Exactly. That's why the ideal case is for the employee to be paid for the value he provides, and the employer also paid for the value he provides (i.e. he provides patents, trade secrets, brand recognition, ...).
> Their "value" is just a function of the job, not a measure of its worth.
There are two kinds of value. The first is economic, and is measured in money. In a market economy, where scarce resources are allocated to the one who values them most (and therefore will pay the most for them), that is the definition of value.
The second kind of value is social or moral. People have value or worth by virtue of being human, completely apart from their economic value. (For example, this is why it's illegal to murder poor people.) And jobs have value in this sense, too, that is different from their economic value. (Perhaps one sees the difference most starkly in the social value vs. economic value of elementary school teachers.)
I think you are comparing salary to social or moral value, which is not just a slippery slope, but is a meaningless comparison - a category error. Comparing salary to economic value, on the other hand, is completely reasonable.
>People have value or worth by virtue of being human, completely apart from their economic value. (For example, this is why it's illegal to murder poor people.)
Ohhh... thats why we dont murder poor people? Never knew that.
"Their "value" is just a function of the job, not a measure of its worth."
I think we need to be careful with terminology here. I'm defining value as return on investment (and sometimes I'll be using the expected return on investment interchangeably).
An employer hires an engineer because they want product x. If the employer thought they could receive $100k selling product x, they'd be silly to pay >= $100k to the engineer. So at the very least, we can agree that the upper bound for salary depends on the value the employee provides.
Salespeople in particular are great at capturing the value they produce. It's extremely easy to measure the production of someone in sales. This is why it's common in many industries for the compensation of a salesperson to be an extremely high percentage of the margin on a sale. By virtue of their value being so easily measurable, salespeople are able to capture a much greater amount of their value than many other professions.
All I'm saying is that comparatively, software engineers do a poor job of capturing their value.
Yeah, agreed - the main point of income inequality is missed.
Beyond the employee who makes 100x less than the CEO, it's also the people who have little to no hope of even getting a job at such a company, because they weren't lucky enough to be born into an environment that allowed them to develop to a certain degree of their potential.
Are you saying that there's no way of for the poors to dig them out of their poor-holes? How do you explain all the penniless immigrants and third-world inventors?
The problem with your rationale is that this type of thinking allows lazy people to blame their situation for their lot in life.
They're saying that the odds are stacked against people born into less well-off backgrounds. A lucky few might have the natural ability or freak opportunity to increase their livelihood; that doesn't justify environmentally influenced wage inequality.
No, I'm not saying there's no way to achieve social mobility. I'm saying that it's just very unlikely. There's always exceptions you can hold up as examples. Also, the idea that everyone living in the "third world" is penniless is ignorant. In every developing country, there is an elite whose children will have the same benefits during their upbringing as in any other country. The idea that because an inventor came from a developing country means that everyone has the same chance at social mobility is laughable.
Lazy people, indeed people in general, are experts at rationalizing away their problems. They don't need the help of my rationale. Of course, everyone should believe that they are capable of achieving anything they set their minds to - research shows that those who believe they are in control of their own destiny tend to be more successful in life. But from a statistical standpoint, there's no denying that people from a tough upbringing have a much harder time than otherwise.
Dang, can we do something about the sexist title? I've already complained about it on their site; I have no idea if they'll do anything about it, but seems to me we could change it here to "The Measured Worker" with no loss of intelligibility.
You jest, but it's worth considering. I mean, if productivity is uneven among contributors and contributions ate now measurable and poor contributors are numerous and can be identified and taken out of the workforce, we could end up with perennially a jobless class of people.
What do we do with them, do we distribute their load across employers (as we do now, for the most part) or do we allow discrimination but make up for this by allowing them income vouchers, knowing under this scenario they'd be hard pressed to hold a job, or do we fabricate work suitable for them?
This guy is a professor of economics. Says it all really. He presents his utopian vision absent of any connection to reality.
He completely ignores data protection legislation. Most of the KPIs he describe measure presentation, not productivity.
There are so many instances here of selecting evidence (anecdotal, fanciful) to suit your hypothesis here he would make a first year social studies student blush. What was the hypothesis anyway? That inequality is linked to value production and that will be borne out by measurements gathered at some future time?
Just goes to show, Economics is a field of study (not a science) where learning the "right" answers and courting of research grants trump cold hard empirical analysis.
44 comments
[ 3.6 ms ] story [ 108 ms ] thread[1] http://www.imdb.com/title/tt0119177/
(sorry about the autoplay video, but there's a full text transcript at the link)
The notion that a moment of juvenile whimsy is now a permanent part of a person's digital footprint and could turn up in background checks for the remainder of their life is hideously Orwellian; we are busily crafting the sociopolitical infrastructure of a totalitarian state at the same time as we have 'serious' candidates for the highest national political office expressing unbounded enthusiasm for running such an entity.
But the prospect of an employer using previously collected data to make an employment decisions seems rather scary. People do a lot of things when they're young and immature and it seems paranoid to rely on that information when they would probably grow out of it. How accurately can past information predict future productivity? I feel like we can't (yet) measure the effects of external factors on human nature, such as new ideas, new habits, new relationships and so on.
It would be a problem is 99.9% of employers used the same metrics for hiring, because it would just exclude whole groups of people.
Beware of cargo-cult :). I'm not saying it to you personally - but I can imagine some manager or marketer saying those words and meaning that "if measuring is so important in science and engineering then it must be good, therefore let's use measuring!". Measurements are only as good as the questions you want them to answer; ask about wrong things, and even the best data will lead you astray.
Which is my way of saying, "Doing good science is actually pretty hard."
From the article:
Today a digital media company knows exactly how many people are reading which articles for how long, and also whether they click through to other links. The exactness and the transparency offered by information technology allow us to measure value fairly precisely.
There are two different logics at play here, with incommensurate notions of value. The firm's primary goal is to make money; as such a story's value is largely a function of page views (or click-through on ads, etc.) However, we can also assess the value of some produce (article, etc.) at a longer-term societal level. In this case a good story is one that helps policymakers reconceptualize an issue or foment some sort of social change (e.g. a New Yorker article like this one [1]). Within the language of free-market economics these longer-term positive effects might be called "externalities" or "intangibles"; intuitively they are much harder to measure, especially with capital flows. But think about it— do we really want to incentivize firms, and by proxy employees, to maximize value as measured by immediate capital flows?
More generally, there isn't a natural law that says that compensation should be proportional to output—quality of life may well be better even at the top if this relationship is somewhat dampened. The cognitive cost may be lower, for example, of not continually switching jobs based on market conditions (this is certainly true in the limit) or of reducing anxiety about losing the ability to produce (e.g. health and psychological wellbeing). These are issues we need to think about and discuss at length, not take as premises.
[1] http://www.newyorker.com/magazine/2015/07/06/revenge-killing
Edited to fix italics and to correct the spelling of Cowen's name.
There's just a natural tendency of a company to want it that way since this maximizes their profits. Once regulations and conventions held-in-check the tendency to profit from the exact measurement and control of workers. Those hindrances seem to have broken down and the advance of fluid financial capital and technology seems like a freight train with enough momentum to batter down any and all limits to their advancement regardless of our opinions on the subject.
I don't see visible force standing in their way.
I agree with your overall argument that we should be careful about which premises we begin with, as they can easily become axiomatic and stifle thinking (a perennial problem in politics). But I think Cowan in choosing his premises here to illustrate how technology that liberates capital flows has the potential to undermine capitalism by limiting people's ability to freely participate in it.
Of course such observations summon the ghost of Karl Marx to whisper 'I told you so.' I wonder how many thoughtful observations of that philosopher are overlooked because of a misplaced reification of his labor theory of value.
Yeah, i don't really know how people get stuff done. Metrics can be nice bumpers, to go see if someone needs help. But actual productivity is pretty nebulous. (to me anyway)
Completely missing from the article are any references to differences between jobs with easily measurable output and those with difficult to measure value.
Software engineers and management are two great examples of professions where value is very difficult to measure. In the opinion of many, software engineers tend to be underpaid compared to the value they produce. Managers on the other hand, tend to be viewed as overpaid or at least better compensated compared to the value the produce. I believe that managers do a better job of capturing their value primarily because they (or someone just like them a step up the chain) hold the purse strings.
Back to what the article focuses on for a moment. Sure a superior producer might be able to capture more of their value than their low-producing counterpart. That's a drop in the bucket compared to C-level compensation vs salaryman pay.
In summation, people aren't upset that the guy in the next cubicle makes 20% more money because he's more productive. They're upset that the CEO makes 100x more than they do even in bad times for the company.
I think comparing salary to the value produced is a slippery slope. Software engineers are certainly not paid below the average or median salary. Their "value" is just a function of the job, not a measure of its worth.
Can you elaborate? Surely the efficient-market ideal is that everybody gets paid an amount equal to exactly the amount of value they produce.
Note in edit: I suspect what I remember as the Econ 101 explanation is probably accurate enough for most purposes: Both the employer and the worker have temporary, local monopolies: The employer has plenty of money, but a shortage of time, and the worker has plenty of time but no money. So each is willing to trade with the other. The employer could make money in excess of the value added by the employees, by exploiting other kinds of effective monopolies such as patents, trade secrets, even things like brand recognition and goodwill.
Exactly. That's why the ideal case is for the employee to be paid for the value he provides, and the employer also paid for the value he provides (i.e. he provides patents, trade secrets, brand recognition, ...).
There are two kinds of value. The first is economic, and is measured in money. In a market economy, where scarce resources are allocated to the one who values them most (and therefore will pay the most for them), that is the definition of value.
The second kind of value is social or moral. People have value or worth by virtue of being human, completely apart from their economic value. (For example, this is why it's illegal to murder poor people.) And jobs have value in this sense, too, that is different from their economic value. (Perhaps one sees the difference most starkly in the social value vs. economic value of elementary school teachers.)
I think you are comparing salary to social or moral value, which is not just a slippery slope, but is a meaningless comparison - a category error. Comparing salary to economic value, on the other hand, is completely reasonable.
Ohhh... thats why we dont murder poor people? Never knew that.
I think we need to be careful with terminology here. I'm defining value as return on investment (and sometimes I'll be using the expected return on investment interchangeably).
An employer hires an engineer because they want product x. If the employer thought they could receive $100k selling product x, they'd be silly to pay >= $100k to the engineer. So at the very least, we can agree that the upper bound for salary depends on the value the employee provides.
Salespeople in particular are great at capturing the value they produce. It's extremely easy to measure the production of someone in sales. This is why it's common in many industries for the compensation of a salesperson to be an extremely high percentage of the margin on a sale. By virtue of their value being so easily measurable, salespeople are able to capture a much greater amount of their value than many other professions.
All I'm saying is that comparatively, software engineers do a poor job of capturing their value.
Beyond the employee who makes 100x less than the CEO, it's also the people who have little to no hope of even getting a job at such a company, because they weren't lucky enough to be born into an environment that allowed them to develop to a certain degree of their potential.
The problem with your rationale is that this type of thinking allows lazy people to blame their situation for their lot in life.
Lazy people, indeed people in general, are experts at rationalizing away their problems. They don't need the help of my rationale. Of course, everyone should believe that they are capable of achieving anything they set their minds to - research shows that those who believe they are in control of their own destiny tend to be more successful in life. But from a statistical standpoint, there's no denying that people from a tough upbringing have a much harder time than otherwise.
Btw it's better to email questions like this to hn@ycombinator.com. We only saw this one by accident.
I will email in future. Thanks!
In other words: the incompetent are increasingly discriminated against. Maybe they should be protected by law? ;)
What do we do with them, do we distribute their load across employers (as we do now, for the most part) or do we allow discrimination but make up for this by allowing them income vouchers, knowing under this scenario they'd be hard pressed to hold a job, or do we fabricate work suitable for them?
He completely ignores data protection legislation. Most of the KPIs he describe measure presentation, not productivity.
There are so many instances here of selecting evidence (anecdotal, fanciful) to suit your hypothesis here he would make a first year social studies student blush. What was the hypothesis anyway? That inequality is linked to value production and that will be borne out by measurements gathered at some future time?
Just goes to show, Economics is a field of study (not a science) where learning the "right" answers and courting of research grants trump cold hard empirical analysis.