The model skips an important part of the reasoning - some people might deserve to be paid more because they're more valuable (due to being more productive or whatever the equivalent is in a given industry). Pay transparency means that if you want to pay someone more they have to be able to justify it. In a well managed company you need to have some form of employee assessment system to judge how well they're doing. This is already true so that good people get rewarded, but becomes even more important with transparent pay because you have to be able to point to it and say this is why that person gets 50% more than you. You have to be able to have real conversations like "they brought on 2 new clients and delivered all their projects on time, whereas your last project was on budget but two weeks late". Being good and useful (and comparing salaries at other companies) provides the employee upward leverage on price (the inverse version of my role play above).
That said - I don't disagree with the conclusions of the paper (showing correlation between pay transparency and reduced pay), but I don't think it's an important conclusion. Pay is a frequency distribution. Talking about the mean/median of that distribution is less important than making there there are good reasons for the people at either ends of it (e.g. "I'm at the lower end because I'm female). To put it another way - I'm more worried about inequality than overall pay level because the people in the middle are fine, it's people at the bottom who need help.
The problem with specific justifications is that many roles do not have objective KPIs like this. How do you compare the worst teammate on a successful project with a superstar on a project that was doomed from the start?
For that matter, there are also the office intangibles that don’t show up nicely on an evaluation but do cause problems when nobody does them.
These are all problems that exist already and are just highlighted by pay transparency. I agree that they're difficult problems, but they're important ones.
You definitely can't do a proper employee evaluation by a completely objective set of numerical KPIs (even in places it seems possible like sales numbers, it results in gaming the system and knock on negative effects). On the other hand, you can definitely have a consistent system for looking at the different areas of people's performance and judging how much they've added. It takes some effort to get right, and there will always be an element of subjectivity but it is possible to do a decent job at it.
I dont think it's possible tbh, it will always have to be subject to market forces.
It's not up only to one company to reward optimally, it's the entire network which provides employment alternatives to underpaid employees that will eventually lead to an optimum.
For instance, I joined my first local company as an immigrant from a vastly different culture and was the only foreigner. It caused issue as I knew little about the way people interacted and they had trouble accepting my particularities. I was blocked out of suspicion, under constant worried monitoring, for fear of rocking a status quo that had lasted for a decade, and was paid well below market rate for what I could have done.
Well, gladly a few months in that environment let me learn the ropes of my new culture, I was able to quickly change job while selecting a more balanced company, and eventually tripled that initial salary in 4 years, by switching position or job.
I dont want just performance assessment by humans, it's also good to have an open employment network you can adapt to and that can adapt to you. This is something that doesnt exist in my original socialist country and create intense misery, with a 10% unemployment rate.
In comparison, my new ultra capitalist country has 3% unemployment and provides new job opportunity every week, at the cost of a small effort on my side to rectify a position of weakness. Employee performance is tied to a social environment that is sometimes nonsensical to an individual judged as a failure, it is important to balance this reward mechanism with outside competition too.
If you are an employer in a jurisdiction with pay transparency, and you need to get a superstar employee, you just create a titled position for them. Then you can justify paying them 350% the average wage.
The paper claims that there will be a wage equilibrium. Yes, for average employees there will be a wage equilibrium, and average employees will see their wages increase because of this.
The paper tries to claim that this wage equilibrium will negatively affect superstar employees. This is complete nonsense. It bases this on the assumption that employers cannot justify discriminating between normal employees and superstar employees. But this is an unwarranted assumption. Literally just give the guy a title. Secondly it assumes that even for these superstar employees their wage would reach an equilibrium. This is another misguided assumption, because these superstar employees are not a fungible resource like the rest of the employees in the same occupation. They very likely possess unique experience and skillsets. So the pay of other superstar employees should not have any bearing on how much they themselves get paid.
Basically the model seems to think that somehow all the employees are equal and entirely fungible, but at the same time some employees somehow manage to negotiate much higher salaries.
Now of course I am not an economist so it is ridiculous to believe that I have found the fatal flaw in a published paper, but because of the political nature of such articles I find it hard to trust the "hard numbers". I remain skeptical.
>Now of course I am not an economist so it is ridiculous to believe that I have found the fatal flaw in a published paper, but because of the political nature of such articles I find it hard to trust the "hard numbers". I remain skeptical.
I admittedly haven't read the paper properly, but did study economics at undergrad and read many similar papers. This is how a lot of economics works, you end up with published papers claiming completely different results because of small differences model assumptions. Multiple models can be consistent with observed data while providing vastly different policy implications. If you want to look at a well publicized example (though I think it might have to do more with different econometric assumptions) look at the research published on the Mariel boatlift and its impact on labor markets. See here - https://www.bruegel.org/2017/06/the-mariel-boatlift-controve...
If you are hiring someone to do a specific job which is different to other roles then sure you can create such titled positions. But if they are basically doing similar work to others (Just 10X better or whatever), then a fancy title isn't really going to fool anyone.
Coworkers can actually see how productive someone is, assuming they are actually 10X that’s not a problem. I have worked with people that I thought should be making bank.
The only issue is people outside your organization, but they have no idea what a given title means. Scrum Lord’s make 500k, wonder what they do?
Some types of high productivity are easily visible, some aren't.
The dev who can whip out tons of stable code is easily measured. The dev who fights to stop an unneeded re-architecure or veto feature requests that would have maintenance overhead unjustified by their value is harder to judge. And yet that veto may be the equivalent of 2-3 FTEs savings.
Haha, when I had been a software dev for a year, I interviewed at MS and they asked what my biggest success was. I said it was convincing an architect and 3 FTEs that they didn't need to spend 2 months on a 100% solution, given that we had all the pieces of an 90% solution ready at hand.
They did not seem impressed. I did get hired though...
More seriously, I've met several people who really struggled to get recognized officially for their work. One was someone who was just amazing at cross team collaboration. Specifically, we were a small group, doing an important strategic project, but we needed some work from other teams that were serving the companies larger, single product. It was hard to get them to free up a couple of dev-days, even when we had the explicit support up and down the management chain. She was amazing at getting the ICs on the team to get on board. But really struggled during review cycles because a fair amount of her work wasn't traditional lines-of-code/jiras.
Another example was a developer who was a king at driving customer solutions through the dev team. He was a developer, and even though he's probably responsible for millions in not-lost-sales, he struggled to get the recognition he deserved.
Anyway, just agreeing with some examples from my professional life.
Some coworkers can see high performance. If the quantitative performance is only 30% higher, it may not be apparent. If there is a qualitative outperformance, the dullard coworkers may not realize the impact.
Right, what is the spread of salaries among star and journeymen players in professional sports leagues?
The Indian cricket league IPL has a public auction every year. Salaries range from 2 million at the top to 20/30K at the bottom. And most players are towards the bottom.
I think this is conflating two separate effects. One is pay transparency, and the other is legislated pay equity.
Legislated pay equity will always result in lower wages simply because the employer is legally unable to raise their bid for talent. No surprises there.
Why ask others to guess what your motivations are if you can instead tell us?
Reducing disparity within a company is surely achievable and almost inevitable.
I don’t personally believe it’s desirable in endeavors where value contributions can be heavily unevenly distributed, because it likely leads to the company being unable to employ the very top value contributors as another company can [and likely will choose to] pay them more, which is consistent with the main claims in the article.
yeah, as best I can tell you absolutely can pay someone more if they are better at their job under equal pay laws.
In California for example, the equal pay act requires 'equal pay for employees who perform “substantially similar work,” when viewed as a composite of skill, effort, and responsibility.'
Someone who contributes more inherently must have greater skills and/or put in greater effort, and likely greater responsibility, given they are accomplishing more tasks.
You can pay differently based on a seniority system, a merit system, a system that measures earnings by quantity or quality of production, or another bonafide factor such as education, training or experience.
MA law is fairly similar. My concerns around transparency are mostly individual privacy concerns rather than believing that employees (or even regulators) can’t understand generally why two people could justly be paid different quantities of the local currency.
No place is (yet) mandating perfect sameness of the latter and I don’t expect that to be policy in my lifetime or while we have the concept of money as a society.
> Why ask others to guess what your motivations are if you can instead tell us?
Because directly stating "the authors of this paper misunderstood the purpose of pay transparency: it doesn't matter what the average pay becomes, as the goal was to reduce pay disparity" is a more forceful statement that makes more assumptions and simply isn't as friendly as asking it in the form of a question (which might be taken as a socratic dialog, but that also feels more friendly to me than a lecture).
How does decreasing the average imply we are "all" poorer? Literally 99% of everyone could be making more money, but if one person who was only making more money because it could be done in secret is now making considerably less, then the average would still go down. I don't know anyone who actually cared about "average wage"...
A higher average, even if I'm not participating in it, will benefit me.
Looking at a list of countries sorted by GDP per capita reveals that the ones nearest the top (with the highest average) are, with few exceptions, the best places to live even if one is earning less than that average:
You don't really need to make up a theoretical model for this. Just analyze Norway or Sweden where pay transparancy exists to discover how it impacts pay. I remember reading that it reduces overall pay as the model here predicts, but I don't remember the source.
A non theoretical model's results wouldnt be able to eliminate confounding variables in Finland/Norway. Did pay go down because of pay transparency or because of 17 other variables that were changing at the same time? No way to tell.
A theoretical model is likely trash - you can pretty much reverse the outcome by tweaking a couple of assumptions.
The only way I can think of that might work would be to measure the effect on pay for similar jobs across a state border - where one side implemented pay transparency and the other didn't (similar to how the Dube, Lester, Reich study was constructed).
This could also have to do with Scandinavian social tendencies, where doing excessively better than your peers financially is seen as being in poor taste.
The pay transparency leading to wage equilibrium is just enforcement of the societal expectation. There is also very low class mobility there compared to, say, the US.
The idea that Scandinavian countries have lower class mobility than the U.S. goes against almost every study and observation on the issue. Here is the World Economic Forum's findings [1], which places Scandinavian countries at the top across a host of factors including income, education, and health.
I'd also be careful about placing much emphasis on things like culture, not because countries are homogeneous culturally, but because often when people talk about how culture in Europe is so drastically different than culture in the U.S., they are referring to stereotypes that are mostly at the margins and don't often have much of an impact on socioeconomic factors. As humans, we have a tendency to focus on superficial differences and exaggerate those differences to explain away phenomenon.
I highly doubt you'd find a credible study indicating that Scandinavians will forgo attaining a higher income and a better quality of life just so they can fit into a mold and satisfy some kind of cultural tendency, but if you do have any such studies for review I'd be very happy to see them.
This report is almost entirely studying the ability of people to move up from poor to middle class. That's one kind of mobility sure and those countries focus social programs heavily on evening out everyone's income. That's exactly what I was saying.
What I'm talking about wrt low mobility is how often people in those countries go from poor or average to wealthy or from wealthy to average or to poor. Generally wealthy people stay wealthy there and average income people stay average income.
That study's data supports this -- it's very easy to obtain a mean income if you had less; after that movement is very low. This should be obvious but if you have very low income inequality then there's very little potential to be either rich or poor.
This isn't even a criticism! It's great that people can generally earn a living there. I'm just stating what everyone seems to know there -- that it's very hard to become wealthy and generally looked down upon as a goal. Scandinavians don't like to boast. Startup costs for new businesses are high and you can't just go buying up tons of real estate to rent out to people because land ownership is heavily regulated. I can't even buy an apartment as a non-resident foreigner (in Denmark) without a very difficult to obtain government approval. Danish monetary policy is having its banks charge customers penalty rates for having a meager cash savings.
In the US we have much greater swings in fortune, good and bad. Heck, I've been both a millionaire and homeless within my lifetime. That kind of outcome is absurdly unlikely in Scandinavian countries.
The reason this doesn’t work as well as you’d hope is that there are confounders that make it difficult to apply Norway’s experiences to the US. The design of using a “natural experiment” where a localized change lets you measure before/after at a discontinuity is a good way of solving this. That’s what this paper looked at to test their model.
I’m not familiar with Sweden, but this is how it works in Finland. The data are not published online, but can be queried by phone or in person from the tax office. National newspapers publish lists of the top earners each year when the data are published; here is one example (in Finnish), including everyone with income over €100,000: https://www.iltalehti.fi/verokone
> So, it's in employers' interest to make the salaries transparent, isn't it? I bet the model is incomplete.
Not necessarily - without transparency, an employer might decide to hire Dr Super Engineer, and might be happy to pay her 2× the mean wage of their existing engineering team.
With transparency, they might choose not to hire her, because they are only happy to pay her 2× if the existing engineering team are not aware, and she might refuse to take the job if they won't give her 2×.
>With transparency, they might choose not to hire her, because they are only happy to pay her 2× if the existing engineering team are not aware, and she might refuse to take the job if they won't give her 2×.
What's stopping them from creating a new role/position/title that pays 2x?
Nothing, but there's no need for that, because there's nothing stopping them from paying her 2x for the same position, based on the actual skills and experience someone has.
You can totally pay people different wages under these laws for the same position, if they are better at their job, you just have to show that that's why.
Employers are already intrinsically motivated to pay as little as they can get away with. Isn't the fact that they are willing to pay more the best kind of proof?
That would only work as proof if you believe that salary is a perfect metric for job performance, and that nobody is paid less than they could get elsewhere (IE that the labor market operates with perfect efficiency).
I imagine we both know neither of those is the case.
Since those are not the case, you can ask yourself, are bills requiring equal pay for equivalent work and pay transparency each likely to improve salary as a metric for job performance, and labor market efficiency?
I suspect they do both, but if you believe otherwise, feel free to explain.
> That would only work as proof if you believe that salary is a perfect metric for job performance
There exists no perfect metric for job performance, so what is your point? No alternative system you propose can do it perfectly.
> and that nobody is paid less than they could get elsewhere (IE that the labor market operates with perfect efficiency)
Pay is not the only measure of a good job. I could easily find a job that pays me 50% more than I currently make.
> Since those are not the case, you can ask yourself, are bills requiring equal pay for equivalent work and pay transparency each likely to improve salary as a metric for job performance, and labor market efficiency?
No, they absolutely aren't. Who decides what is equivalent work? Having the same job title doesn't mean you add the same value to an organization, and trying to use law to force that will exacerbate the problem, not make it better.
Some people work more, practice their craft outside work, or are available after hours as needed. Those people are more valuable and should be compensated as such, not lumped in with the lowest common denominator.
All of the people I have talked to who supported these "equal pay" measures were people who were bad (or at best mediocre) at their job.
Businesses in the US are driven almost entirely by money. The people making less at their jobs are really, truly providing value, they can use that as leverage to negotiate a higher salary. Or just take matters into their own hands and give themselves a raise by getting a new job.
I don't get how the end of the political spectrum that complains about pay gaps is the same one that constantly complains about how corporations are driven by greed and money.
No employer wants to pay any person any more than they have to. I, a white man, have had to do a lot of work to justify my own raises. And when I got screwed by my employer, you bet your ass I left and found something better.
If people think they really can get more for what they do, they should go out there and do it. At the end of the day, money talks. If you truly provide value, you have leverage.
I’m in favor of a light version of salary transparency. (“Here’s the anonymized spread of compensation and you can find where you land in the spread.”)
I’m not in favor of publishing the individual salary information of people who joined my company without that publication being part of the bargain. (I view it as a serious privacy violation whether I publish “here’s what XYZ makes” or whether I publish enough data that lets a data scientist figure out which point in the dataset is XYZ with greater than 80% certainty.)
>I’m in favor of a light version of salary transparency.
In countries with salary transparency, (the ones I've worked in anyway), there is usually a published grid of roles, and salary ranges. Similar to how the government does things[0]. Companies then make use of "benefits" to reward better performing employees, since those benefits are not part of salary.
Bargaining power is mainly dependant on offer and demand; it can be low, or it can be high, depending on how valuable you are to the company. Pay transparency will simply increase the number of employees who will try to bargain. I guess that if you're not valuable enough that could backfire for everyone. If you do are valuable the company will just need to pay.
Pay transparency might result in lower wages at the transparent company, as underpaid employees jump ship to another company - leaving only the lower-value companies being perhaps more fairly paid.
That doesn't mean lower wages overall for all workers involved, though. The ones that jumped ship may be getting much more somewhere else, raising the average in a way that wasn't looked at in this model.
That said - I don't disagree with the conclusions of the paper (showing correlation between pay transparency and reduced pay), but I don't think it's an important conclusion. Pay is a frequency distribution. Talking about the mean/median of that distribution is less important than making there there are good reasons for the people at either ends of it (e.g. "I'm at the lower end because I'm female). To put it another way - I'm more worried about inequality than overall pay level because the people in the middle are fine, it's people at the bottom who need help.