Hacker News Reader

Comment by cperciva | Hacker News Reader

If you formed a union at your workplace, you and your colleagues would get to decide the rules like that.

Yes, but I am more skilled than my colleagues.

Unions inherently negotiate on behalf of the median worker who was employed at the time that the contract negotiations took place. As other people have pointed out, the fact the unions negotiate for current employees rather than future employees contributes to the "screw the new guy" phenomenon; the fact that unions negotiate for the median worker can result in them undervaluing their high-productivity workers.

This is why unions are effective primarily in low-standard-deviation-productivity fields: If the best secretary in the office is 10% more productive than the median, but the union manages to raise the median secretary's income by 20% compared to the market clearing rate, then every secretary benefits. However, if the top software developer is 10x as productive as the median...

View on HN

10x developer is a myth, but a fun one that makes you feel warm and fuzzy if you believe it.

You can negotiate a higher pay with your employer if you excel, the union provides the bottom from which you can negotiate.

Unfortunately, right now, that bottom is below our what our fair market-rate earnings would be.

> You can negotiate a higher pay with your employer if you excel, the union provides the bottom from which you can negotiate.

I've never heard of a unionized industry that worked this way. Unions negotiate pay for every worker that belongs to the union. The only way to negotiate for higher pay is to quit your unionized job and take another non-unionized job.

Depends on the union. Some allow for negotiation, some have a scale which you can work with.

Bonuses and supplementary pay are usually not determined by unions and are up to you to negotiate in this situation.

Evidently this is highly industry-dependent. My experience is in the US auto industry, where all union workers' pay and benefits are fixed by what the union negotiates with the company. (At least, they were when I worked in that industry.)

I don't know much about how most unions work, but at least TV & movie actors and writers, classical musicians, and professional athletes have unions that work as task_queue describes. It's hard for me to imagine a programmer's union that didn't work this way getting off the ground.

Here's an example: The Screen Actor's Guild.

Matt Damon is a SAG member but he does not get paid scale. Scale is a minimum wage set by the union but if you're a big enough star, you negotiate your own wages.

Pretty much all the unions in industries where there are recognized hyper-elite workers work this way (SAG s mentioned in a sibling comment, most of the major professional sports players unions, etc.)

Most other unions don't work that way.

It's not a myth, some developers really do suck that much.

This is exactly what I am thinking. Its not that I believe I am amazing, I just see so much incompetence around me.

10x developer is very rare, but I have worked with guys like that - though of course, this depends on tasks. If task is well defined and bounded by a given tool set and rigid rulesets, etc, 10x cannot really be achieved. On the other hand, if there is just a vision of achieving goal X, and the median developer doesn't even know what tools to start with, an outstanding developer is much more than 10x productive than median.

However, in regular development jobs, I have also seen 0.1x developers, and worse. Even guys who definitely do damage if they are given access to the tools. The variation is huge.

I do believe that top developers are 10x as productive as the median. I don't see this as an indication that top developers are particularly good, though; rather, it's a commentary on how awful the median developer is.

If they do their job, they negotiate the minimum wage and benefits. Nothing prevents you from getting more than that if you're skilled, but it does in fact prevent the new guy from being screwed over if done properly.

You say that now, just you wait until supply of programmers catch up.

Yep. It has always amazed me how my peers can assume that their pay should be proportionate to their value in one breath while ranting about unions in the next.

I mean, it's one thing to realize that the supply/demand situation for programmers currently makes unions unnecessary, but it's quite another to slam people in other fields for doing what they have to in order to secure the same damn privilege (getting paid according to value).

The opportunity to make this extremely modest dream a reality is not some kind of right inherent to the free market! It's something we're going to have to fight for once supply/demand equilibrium has been reached.

When the supply of programmers catch up, the weak go to slaughter. 2 people digging a hole for $5 an hour are more effective than 1 person digging for $10 an hour, almost completely irrelevant to digging ability. 2 dummies don't make 1 smartie in programming. Unions exist in fields where replaceability is high. That happens much less in educated professions. The main other reason they exist is for professions with a high risk of injuries. Carpal tunnels only takes so many programmers every year.

Ok, so how do you explain the situation for chemists? The variation in their productivity due to competence is at least comparable to that of programmers if not more -- yet their pay is uniformly low and the job security is uniformly poor. The only escape route is having the flair (!=competence) required to go into administration or attract investors (which is basically the same thing). OTOH if you assume that supply/demand is king, it all makes sense.

To assume that your competence will save you, you need both to assume that your evaluation of your own skills is correct and that others will confidently agree. Both of those propositions sound shaky. I've met a lot of overconfident programmers and I've met a lot of disinterested HR people.

I don't know anything about chemists. However, I struggle to believe an industry with a career path starting freshman year has more variance in ability than a technical field pulling poli-sci majors off the curb to fill seats.

Supply/demand of a chair takes a distance second to the value of the product of that chair. Medicine/law/finance are all oversupplied with high salaries. Try showing up to a teaching position interview and convincing them to pay you more because of econ 101.

> secure the same damn privilege (getting paid according to value).

I think this is the hard part: what is the value according to which people should be paid? Who should decide it? Not easy.

It gets particularly hairy when technological change has an impact on value. This is happening faster and faster.

Consider the classical example of Wapping dispute [0] where old hot-metal methods of printing, which unions sought to protect, were replaced by electronic publishing tools and printing machines. Today, it would look quite funny if papers still had to employ Linotype professionals to be able to print things in newspapers. The more acute concern is whether publishers print anything at all on paper, or if everything is in the net. If the unions could decide, of course you couldn't publish in the net, because it is destructive to jobs in the printing industry.

At some time, the commercial "value" of Linotype operators was very high. Now it is only historical; as a commercial operation, the machines have gone to obscurity. Who should have decided the value of those workers? I think the only one who really can decide is the one who owns and runs the operation. The unions mostly have an incentive to preserve the status quo - while happily accepting a change in other businesses, if it benefits them. No doubt that the modern-day Linotype operator would see it fit that he uses his union pay to buy an iPhone.

[0] http://en.wikipedia.org/wiki/Wapping_dispute

> what is the value according to which people should be paid? Who should decide it? Not easy.

On the contrary, it's economics 101. I have a feeling our disagreement is due to definitions so let me be more specific about my intended meanings.

There is a minimum price at which you are willing to not switch vocations and there is a maximum price at which you can be profitably employed. The latter is what I was calling value, as in "value to the company." Relative bargaining power determines where on that spectrum your salary will be. It's arguable that the final price is "value" but that's not the meaning I intended. What should happen is a question for the philosophers. What does happen is up to us.

The Wapping dispute is an example of a union failing to recognize that the value (same definition I was using above) of their employees had fallen below their asking price. Sure, it happens -- I have no doubt that unions increase friction, which is bad and hurts the economy. But underpaying workers "because you can" is also bad, also introduces friction (increased need for loans, decreased ability of workers to bootstrap), and also hurts the economy. So I don't buy the argument that one is better or worse. As far as I'm concerned it's a zero-sum power struggle, philosophically and morally speaking. I'd rather be on the winning side and I don't begrudge others for wanting to be on the winning side.

> maximum price at which you can be profitably employed.

I think such a figure is extremely difficult to calculate in many if not most jobs.

> just you wait until supply of programmers catch up

Been hearing this for decades and the only things happening are:

1. The multiplier effect of programmers keeps going up.

2. Programming work is growing faster than the supply of programmers.

I don't plan on seeing these trends reverse during my lifetime. It's not a zero-sum game.

Is there evidence of this?

In Europe all I have seen is wages going down over the last 10 - 15 years. Admittedly I only started taking notice during the dot com boom, but I get the impression they were better before that.

The other thing I notice is that I personally manage a database that would have taken three of four people ten years ago. We have loads more frameworks and libraries that make the task manageable. I use Django, whereas ten years ago, the stuff done by Django would likely be done with in house libraries, so needed more people. Also I think many of the problems are "solved", we have ideas about how to do things that were not as clear when the internet was staring out.

A small subset of the population even has the potential to program and that's probably not going to change. If a lot more people are encouraged to try it out then a small amount more of the people who had an aptitude for it but didn't realize it will enter the workforce. The leftover majority will still not be able to. No dramatic increase in programmers is going to happen.

http://www.eis.mdx.ac.uk/research/PhDArea/saeed/

http://blog.codinghorror.com/separating-programming-sheep-fr...

Unions exist to increase the bargaining strength of the labor supply when it is very weak. The bargaining strength of developers is rather strong, and good developers can demonstrate their ability.

Good developers would not want a union, as it would likely drag down their compensation while forcing them to work with hiring mistakes.

Those who tilt at these giants mean well, but they will unhelpfully collide with windmills.

>>However, if the top software developer is 10x as productive as the median...

I wonder what's point in having a fast super car and complaining you are stuck in traffic often.

If you are good by a factor of 10, it hardly makes sense for you to ever work with a group of people.

Reply on news.ycombinator.com