xaa

Hmm, I can see how you got there, but I don't think it has worked out that way in my case.

The amount of time you have to spend writing grants is a function of: 1) how much money you are burning, 2) how much help do you have writing, and 3) how good is your science (better science will be more successful).

No doubt I hurt on #1. But I think I help enough on #2 and #3 to compensate, and therefore I don't think my PI spends a net larger percentage of time writing with me than he would without me. Anecdotally, over my time with him, I think has been a slow decrease in the fraction of his time he spends writing.

If we're talking about unicorns, a competent "staff grant writer" has to be the rarest of them all. To write a grant well, you have to deeply understand the specific science at hand and be a good writer. Someone who would voluntarily choose such a job would also be someone who enjoys the process of writing.

So, in other words, someone perfectly suited to be a PI, a more prestigious and better-paying position with more freedom.

Our institution has one you can ask to help write/review/correct grants. They are pretty useless because they cannot possibly understand the science each individual lab is engaged in at an adequate level. Usually they just offer superficial suggestions/corrections.

Sure. The most mature part of this is analysis of expression data from GEO. We had to take the raw expression data and textual metadata, extract attributes from the semi-structured metadata into something statistically useful, and devise methods to do automated meta-analysis between experiments and platforms on the expression data. We are talking about >1M samples here by 20K-50K transcripts.

Now we are basically extending the same idea to new data types and sources, like methylation and sequencing-based data. When we ask "what should we do next", we look for basically what archives are biggest, and if they conceivably have something useful to say about aging, we put it on the queue.

The big-picture goal is to take this approach and apply it to all these different levels of organization (DNA, epigenetics, RNA, and ultimately hopefully protein) and put it all together in a cogent picture, using way more data than anyone has done before. We ask questions like "how does the aging process differ between tissues?". We also want to go "wide" by comparing species, as the whole "intervention X works in mice but not humans" is an important thing to address.

Then on the smaller scale, the system is also of use and interest to individual investigators who want to know "what's happening with my gene X with age in tissue Y?". They like to include figures from us for example if they have some findings from mice and want to give evidence that the pattern may also hold in humans.

Finally, we built the platform to be a pretty generic meta-analysis tool, so we have occasionally used it to answer questions completely unrelated to aging. Once you do all the background legwork, asking many different questions is largely a matter of changing the meta-analytic model.

I thought it would be this way too. But I do actually try to help out with funding as much as I can.

Ultimately, though, I'm afraid I can't summon too much sympathy for a PI who thinks this way, because they knew exactly what they were getting into when they accepted that promotion. It is common knowledge that the tradeoff you make upon becoming PI is that you get the authority to direct the research aims while sacrificing the ability to do them yourself.

Exaggerating somewhat, a PI gets 100% of the intellectual freedom and responsibility, and has 0% of the fun. A grad student or postdoc gets 0% and 100%. I was looking for something like 50%/50%.

The reason this was easily possible in my case is because of the peculiarities of bioinformatics. Namely, equipment costs are low, so there is more money for staff per grant dollar.

We also have had much more luck with center and collaborative grants than R01-style individual grants. On collaborative grants, our sections are almost always the highest scored, on top of the fact that the pay lines are inherently higher. On individual grants, it is hit-or-miss to say the least, because reviewers seem to think, somewhat rightly, "why do you need $1M to play on computers?".

People come to us and say, "hey, I want to put you in as a small section for $75K, $100K, or whatever". Over the course of time it has meant we have a diverse and relatively stable revenue stream and low overhead, and there is less risk (because less time spent) on each individual grant. Although amusingly, my PI is still associate despite pulling in way more money overall than many fulls, because of this idea that only PI grants count for promotion.

> I say that because that means I now have to pay for both of us while you go do science.

Yeah, although I said in my OP that "I just want to do science", in reality I write parts of grants, mostly the technical parts, all the time, and I go to grant strategy meetings, where I feel my input is valued as I can often give a better idea of what's feasible than the PIs can. I guess more precisely, I don't want the responsibility and pressure of being in charge of the whole package, and I don't like dealing with all the paperwork. So in our case the funding is a team effort as well.

> way things like training grants are structured

I have been with him since grad school. In the early days, the way we got around this was titling me as a postdoc and paying me like a staff scientist, which was fine by me.

Yeah. Although it is annoying that we both have to clarify "well, I could be a PI if I wanted to", because of this social norm that if you aren't a PI by the appropriate age, it means you weren't good enough. Because, obviously, everyone wants to be a middle manager. And we aren't talking to an academic crowd here, but a lot of academics would be skeptical when we say that.

If enough people in our generation take this approach, I think it will be very good for science. It is a happy medium between the old system of "everyone gets tenure eventually" and "no-job-security eternal postdoc" of the last few decades.

Science is inherently risky enough without also forcing everyone to essentially become an entrepreneur. That phenomenon is likely behind so much of the exaggerated findings that are getting published.

My interest is in applying bioinformatics methods to understand aging. His is bioinformatics in general. So most of the time there is actually no conflict between "my" projects and "his" projects, my interests just being a subset of his.

It doesn't hurt that my interests have introduced him to new people and funding mechanisms that he would not have otherwise had access to.

I'm not sure how much detail you want. That was the short version. The slightly longer version is that there are huge datasets out there, and I want to understand aging by building tools and workflows for the many datasets which were not necessarily originally gathered to understand aging, but which still contain information about it. And also to work with and complement wet-lab aging researchers by giving them context or leads for experiments.

My career goal is to be a staff scientist (in biology/bioinformatics). I hate writing grants and the related political bullshit. All I want is to do actual science, even if it means a marginal pay cut and some degree of decrease in autonomy -- although in practice, I'm not sure how much autonomy there actually in being a PI, where you are "autonomous" in that small portion of your time that doesn't involve raising money.

I made a simple proposition to my PI: I'm skilled. I'm at least 3x as productive as a grad student (and I could make 10x in industry). You pay me 3x, and I'll work on your projects some of the time and mine some of the time. Give me (a realistic and workable level of) freedom, a living wage, and treat me right, and I'll stay here as long as I can. He agreed.

It takes an experienced PI to understand just how much more value a staff scientist gives over a early-stage postdoc or grad student, which are cheaper. You get stability, expertise, and experience for what is ultimately not that high a price tag. I myself have watched my PI undergo this evolution. I have been with him 10 years, and in the early days, he was all about summer students etc. Now he knows that you simply cannot get real work done with cheap, temporary labor.

I could get higher-paying or more prestigious jobs elsewhere. And I do believe I could cut it as a PI if I wanted. But I really, really value this arrangement.

> the challenge is finding a "job" that you love and are excited to tackle every day

This is great and very important, but it doesn't prevent burnout necessarily. I have a job and career that I really enjoy and consider to be very meaningful (aging research), but I have been severely burned out several times.

In some perverse ways, having a "meaningful" job can increase stress and propensity to burnout exactly because you know it's important. I imagine this is why doctors and lawyers have a lot of burnout, stress, and substance abuse problems. OTOH, you don't really hear about corporate accountants getting burned out.

My personal theory of burnout is that it happens when you start to feel you are on a treadmill -- running hard to achieve an important goal, but getting nowhere. Then the stress and burnout symptoms increasingly make this a self-fulfilling prophecy and feedback loop.

> Have you been using it in an enterprise scenario?

Yeah, sort of, an academic/hospital setting. Actually, I combined my academic and personal e-mail addresses into one at a custom domain, and forward other accounts to it, which annoys some colleagues.

Most people here use Outlook. Sometimes people might have a mailto: link in sig (which gives me a chuckle as it's pointless), but I don't ever remember encountering embedded images.

The biggest human offenders for sending HTML-heavy e-mails are university-wide announcements. In my mind, these fall closer to spam than ham, and what useful information is in them can usually be gained from the title.

> Roughly 90% of the direct email I receive in my inbox comes through in HTML form.

Same, because I haven't bothered to set up spam filtering. But of the e-mails I actually want to read, well below 5% do.

> Plain text is more often coming from mailing lists and automated systems that email me.

Fascinating. For me it is just the opposite. People don't usually seem to bother using HTML, consciously at least. But automated systems use it quite liberally.

> unfortunately HTML email is the way of the world

Is it really? I still use mutt, and I have found 99% of the time, heavy HTML in e-mail means the mail is spam.

Occasionally an important e-mail (bank, etc) might contain a few HTML links, easy enough to visually parse. Real individual people IME send plain text, even non-techies.

Although this article never mentions the word, it's really about abstractions, not "implicitness" per se.

No one would say that if you have an almost-perfect abstraction, like a physics equation for fluid dynamics, that you should go to the effort of dropping down to the lower level (specifying what each molecule should do, in this case).

The problem is that this ORM example is an extremely leaky abstraction. Several of the inferences it is making are inferences that are not predictable if you don't already know how it will work. And they are inferences that matter.

Anyone who has ever used an ORM knows that it seems like magic, right up until you try to run some large query and it takes an eternity, and then you have to unpack it all and add all the explicit stuff to fix the problem that you "avoided" in the first place.

The real art of programming is in knowing what needs to be explicit and what can be implicit. I'd agree that if something can be made implicit without mystifying callers, it should be.

> I think dmix hit on a very important point when he mentioned that its hard for politicians to sell a future thats not all rosy and prosperous for everyone.

I couldn't find where he made that exact point, if he did, I would reply to it by saying that what politicians sell and what they actually do are two completely different things. Both sides of the aisle have been selling rosy futures and working towards the exact opposite for decades. I'd say that Bernie & Trump were evidence that the message "your future isn't all that rosy unless we change substantially" resonated this election.

dmix and I agree globalization is here to stay. I was just a little irritated by his implication that the way we need to respond is by "developing new industries" and "adapting to the global marketplace" and all that jargon. I'd say the S&P has adapted just fine to the global marketplace. The problem is that the median American is worse off anyway. Your second paragraph is apt about some potential solutions.

> So the dichotomy between historical fluke and market trends doesn't really exist. Historical flukes are market trends.

On further thought, you're right about this. I was rather inarticulately trying to remark on the fact that the supposed "inevitability" of market trends is always used as an excuse to avoid pro-labor legislation, but never pro-business legislation.

> I hardly see globalism as something that's detrimental to America as a whole

Well, we're increasingly seeing that there's no such thing as America "as a whole". There's capital, who it's good for, and labor, who it may be good and bad for in various ways, but a net bad, I'd say. Good for your 401K and the cost of your iPhone, not so much for your salary or employment prospects.

If you take "good for GDP" to mean "good for America as a whole", then yeah, it's probably good. Problem is where that GDP growth is going.

> you think the US government has been highly influential in developing 'globalism'

Yes and no. I think the gov't has been highly influential in ensuring the gains from it go to capital and that its negative effects on labor are not reined in. Globalization would have happened anyway but the govt's approach to it has resulted in more inequality and dislocation than was necessary.

As for whether the government could have slowed or did hasten the process, hard to say, but I think so.

> market trends and phenomenons can be controlled somehow, that they are simply choices that people make, and those choices can be controlled

This double standard galls me. If a trend is detrimental to American workers, it's market trends, like the hand of God, irresistible. If it's beneficial, it's a historical fluke (post-WWII boom).

I might buy this if the U.S. government had been a neutral bystander in this process of globalization and the hollowing out of the middle class. In that case, we could say that it was just economics. But, in reality, the government:

- Incentivizes and fails to do anything about offshore tax havens

- Doesn't bust trusts

- Has low tax rates on capital

- Sets immigration and trade laws favorable to corporations

- Spends tons of money on pork, including the porkiest pork of them all, the military

and so on, ad infinitum. Would this process have happened anyway? Maybe. But our government, and the companies running it, actively accelerated it in every way they could. If these trends can't be controlled to some extent, why have so many lobbying dollars been spent to accelerate them?

Thanks so much. Fascinating. I've made many of these types of app but never to "web scale". Bioinformatics apps are a bit niche, and we take the view that our non-paying academic "customers" can wait however long it takes to finish the query, in the unlikely event that there is high load.

Just to make sure I understand correctly, that's about $1-2K of one time cost per 10qps (w/o SSD, and not counting power and maintenance, etc)? When I first saw "cloud server", I thought that was a per-month rental cost, but the link is for actual in-house hardware. If this is even close to correct, my suspicions seem confirmed.

Except for one thing. I have no idea how many qps a site like Academia would have. 100qps was completely out of my ass, but it seemed hard to imagine it being any more than 1-2 orders of magnitude higher, at most. Any guess on that?

Bioinformatician here. I appreciate that you need to make money somehow. I do find this a little hard to believe, though:

> full-text search of 20MM PDFs at our traffic levels is expensive to operate

Assuming you convert them to text once, index them, and put them in a standard FTS engine, I'd guess it is on the order of 100GB-1T of text (max), plus some more for the index (basing these estimates with my experience text mining PubMed Central and MEDLINE). So it can all fit on a pretty standard server. Maybe at 100 req/s it would take a few. Yes, you'd want replication.

The number of servers required to get good latency FTS is the part of this that I'm least familiar with. Anyone have a ballpark, given these estimates, on what kind of hardware would be required? (I could easily be wrong, and indeed this is very expensive. If so, I'd be curious about ballpark numbers)

> exurbs with their cheap sprawling subdivisions and giant SUVs benefit from it

I guess I'm with you to the extent this means they "recoup a larger proportion of their losses on these pointless endeavors".

> Your not being in San Francisco costs the government $30k/year. Any subsidy below $30k/year that gets you to that more productive job is profitable for the taxpayer.

With this I agree, although it seems at odds with what I think you were saying earlier, that tax should be COL-adjusted. If this were the case, it would no longer be true.

> People are most needed where they would create the most value, regardless of cost of living.

Nope! True, this would be best from the government's perspective. From the individual's perspective, it is most rational to live where they have the biggest surplus of wages over costs. Doesn't do you much good to make $1M if it's $1.5M for a crappy apartment and basic food.

Unless you are separating the concept of "people are needed most at X" from the concept of "it is most economically rational to live at X".

> imagine someone who costs taxpayers $20k/year in Appalachia, but could generate $30k/year ($5k in taxes) as a barista at the Googleplex

OK, yes, but this sets up a mostly false dichotomy between "employed in SV (etc)" or "unemployed in Appalachia". There are a lot of places in the US that are less dysfunctional than either one of those two places.

> I simply argue that subsidies to get people to high-productivity cities are much less costly (they can even be profitable!) than subsidies to help people stay in low-productivity situations.

It just seems self-contradictory. You say that people in SV pay more tax than people not in SV. That's right. But then you say people in SV should be subsidized, which removes the logical basis you're using to argue they should be there in the first place.

It is helpful to remember that subsidizing someone's COL in a low-cost area is by definition going to cost less than someone in a high-cost area. In short, if we HAVE to subsidize people a lot, then we should subsidize them to live in Appalachia, not SV. But if only a little (i.e., they still pay tax in net), then it could make economic sense to subsidize people in SV. Ironically, by this logic, it makes more sense to subsidize the $90K person in SV than the $30K person.

> income taxes are not cost-of-living adjusted

Well, this is another issue. It's incoherent, honestly, that someone can be considered both rich and poor from the federal perspective. But in the end, a COL adjustment on income tax would be equivalent to the type of subsidy in question.

- The diplomatic, covert, and military apparatus which dominates the world to the extent necessary to secure the flow of cheap oil for your car.

lel. If that was ever true, which I doubt, it isn't any more. If there was ever a point that our military paid for itself (even on the margin) in terms of cheaper oil, I'll eat my hat.

> Welfare so that unemployed people can stay in dead-end towns, instead of moving to where they are needed.

This is really the core of it. If people are really needed, economically and relatively to elsewhere, in supercities, then supercities will pay enough for them to live there. Ergo, you are overstating how much they are needed. This is exactly analogous to employers claiming there is a "shortage" of programmers, but they aren't willing to raise wages.

Besides, am I to believe that there are no unemployed people on welfare in cities? I don't know what the relative rate is. I do agree people should be incentivized to move where they are needed. We may have different views on where that is, exactly. Any kind of welfare distorts economic "need" signals, which is a big downside of it.

Imagining welfare or wealth transfers didn't exist, I'd say that the place where people are most "needed" is a place where benefits of living - COL is highest. I have no data on where that might be, but I doubt SV ranks high on the list unless you are a programmer. By that metric I would imagine that there would be a dynamic equilibrium between urbanization and countryside, as there is now.

In some ways, SV programmers defending subsidies for service workers in SV sounds like "please subsidize my $10 latte and Whole Foods bill".

> the low-density regions they enable

This is pretty creative. If cities had to grow their own food and extract their own natural resources, I think they'd have quite a problem, yes? It may be fair to say cities & countryside are codependent, but the idea that cities are one-sidedly enabling everyone else is not credible.

It is quite likely that there is a net transfer of wealth from cities to the countryside. The reason I would argue this is "fair" is because people in the cities are also the primary beneficiaries of GDP growth.

BTW, I don't live in the country, I live in a medium-sized city. I am limiting my criticisms to cities that have so poorly planned expansion, or are so crowded, that the COL is sky-high, and they want a bailout from everyone else.

No, I'm not. I'm suggesting that they be given the choice of moving somewhere else, of their choice, with subsidies for the move, or trying to go it alone in place.

And this only for those people who fall in that gap between (approximately) the national low-income line and the local "low-income" line. People who fall below the national low-income line should be given subsidies, regardless of where they live.

I believe in helping people, but not subsidizing irrational behavior that exacerbates an existing problem.

Just think about how this looks to someone like me: the low-income threshold in CA is over twice my household income. And I am not classified as low-income. So to some degree my tax dollars will be transferred to subsidize some terrible decision-making. If this looks bad to me, a liberal, think how it looks to moderates and conservatives!

Let's take this to the logical extreme: suppose I decide to live literally in Hollywood hills, where one of my relatives lives. I make nowhere near enough to survive there. Should I be subsidized anyway? Does it make a difference if I was born in Hollywood and this situation arose subsequently? I think not, when it comes to the question of what should be done about it now, although reasonable minds may differ.

I must say that as someone who does NOT live in a crazy high-rent state or town, I am not well pleased by the idea of my federal tax dollars going to people who choose to live there but can't make ends meet, and this even though I'm generally pro-spending on social welfare.

You may say, "but the poor can't afford to move!". Well, if that's the case, the government should subsidize their move instead. It would be way cheaper and would incidentally drive real estate prices closer to the national median.

I'm not trying to cast any blame on people in this situation, as they are likely there through no fault of their own, but this can hardly be the most efficient way to solve the problem.

I feel like economics is screaming a message to people in the Bay Area: "There are too many people here! There's tons of land in the U.S.!", but no one is listening, just casting blame on the techies, gentrification, land policy, or whatever.

Coming from a state where even $50K/hh is quite comfortable (maybe slightly less so with a family of 4, but still beyond "rice and beans" level), I have to ask: how is this even possible? It strains credulity.

OK, sure, rents/real estate are higher. A LOT higher. But other than that, you're not going to pay substantially more for a car, computer, or any other consumer good in CA than anywhere else. Gas and food are somewhat higher, although some of that, based on my visits to and relatives in CA, seems to be based on more picky eating preferences than inherent logistical problems (but you say you don't do that).

It seems outlandish on its face: you know that a $400K household has to be way above the median, even in Oakland. So how is everyone else surviving?

> Now, I'd like to be ethical. I want to choose to be. And I'd like us all to be ethical to each other. But you should make the case that it's beneficial for people to be ethical to each other

Well, you just implicitly made your own case for ethics, one of the strongest ones, IMO: any situation where there is a tradeoff between individual gain and communal loss, ethical thinking is required. If that tradeoff doesn't exist, like the win-win situations that can be encountered in many economic situations, ethics aren't required.

Therefore it's never "beneficial" to you as an individual to be ethical in any individual situation. It's only as the ethical decisions begin to pile up across society that we see clearly it is more beneficial in the aggregate if everyone co-operates than if everyone defects.

anigbrowl is just noting some common heuristics that generate similar ethical outcomes but using simpler logic.

Anyway, I think the reason people are going beyond legality to morality here is that Uber has established a pattern of behavior over time as an actor that doesn't follow the rules and does whatever it wants. It isn't accidental; they've exhausted their reserves of good faith.

> It's sad how HN is here demanding meticulous adherence to a clause buried deep in the incumbent's ToS and pretending that violating it is some basic moral misdeed.

Probably because the context for a "TOS violation" is very different between a large company and an individual, and because the motives for this violation are venal. And because Uber is a serial violator of pretty much every rule it comes into contact with. I doubt you'd find many here who would see a problem with, say, a researcher violating the same clause to get data for a research project.

Large companies can afford lawyers to read TOS documents. Normal people don't have the time or expertise to read them, usually. And who writes TOS documents in the first place? Large companies. It is a double standard if a company wants their TOS followed, but breaks other companies'. So it is coherent to argue that large companies have a higher burden to follow a TOS more "meticulously" than individuals do.

> I enjoy discussing and debating this on HN, but large powerful corporations are likely to get their way.

Haha, so true. I think it's fun and worthwhile to at least figure out what's going on. At some point I have transitioned my interest in politics from "I want to change the world for the better!" to a more detached, curiosity-based position.

WRT STEM MS, my understanding is that in "hard" engineering (electrical, mechanical, etc), MS is economically worthwhile. In science and probably math, it is useless, you need a Ph.D. Not sure the breakdown for CS. In general, there is always a salary increase for each level of education, but the difference b/t fields is in how many years it will take to "pay back" the investment.

But more generally, I'm not completely sure what you're proposing as the ideal solution here. When you say:

> This is why I oppose efforts to allow corporations to control the conditions under which immigrants are allowed to come to the US

Do you mean you oppose "chaining" to a particular company, selection of particular skillsets, or what? Are you saying you support an immigration quota with no particular selection criteria (minus specific "Einstein" exceptions)?

Much of your argument seems to boil down to the idea that it's unfair to would-be immigrants to make them sink costs in education to have an uncertain shot at coming to the U.S. Maybe, but that has no bearing on what is in the U.S.'s (i.e., median US voter's) best interests.

> Ironically, in other words, if you hear corporations talking about a "shortage" in a field, it may be rational to avoid that field if you're a citizen with freedom to choose your job and educational path.

Yes!!! This was my takeaway as well. "Shortage" talk is a giveaway that a sector is trying to apply some kind of pressure to reduce salaries in that sector. Even in a field like nursing, where there is supposedly a "shortage", but there is no prospect of getting immigration to fill the gap, what is happening is they are trying to sucker young people into this field in which the salary is simply too low after you account for all the headaches like weird hours, stress, etc.

Such an "odd" correlation that the fields with so-called "shortages" are always ones with higher than median salary-to-education ratio.

But perhaps it is not so clear. Because there would be no "shortage" talk in the first place unless the job were already economically attractive. The rational decision probably depends on how successful you expect the sector to be in bringing down the sector's wage towards the national median.

Another thing you've helped me realize in the course of this conversation is that all the talk of "job creation", as if the number of jobs is a static thing independent of wage, is totally ridiculous. "We need policy change X to create jobs in sector Y" is code for "we want to reduce salaries in sector Y". In this regard "job creation" may be regarded as equivalent in goals and deceptiveness to "shortage".

> One area where I may disagree with you a bit is on the Einstein thing. I consider this a bit of a red herring.

I meant "Einstein" to be a stand-in for "person with high potential economic/cultural/scientific value". The implicit question was, should we maximize immigration for highest value-per-immigrant (however that is defined), or should we select random immigrants?

> ...leave everyone better off.

Oh, my. If you had said "average GDP per capita" or "total real GDP" or even "many people" better off, OK. Free immigration is very much like free trade: it makes the system as a whole more efficient, but there definitely are winners and losers in the process. In the limiting case of 1 immigrant, employees in that immigrant's field lose (slightly), and everyone else wins, including that immigrant's employer.

So, in my view, these things should be controlled in such a way to try to make sure the winners aren't people already winning and vice-versa. But not to the point of cutting off our nose to spite our face; there is this tradeoff between equity and efficiency.

> If immigration forces people to be programmers, then homegrown programmers will be disproportionally affected by immigration...I don't think this would have happened had immigration been "free"

Well, I agree, and this is pretty close to a tautology. As the "high-skill natives" in an H1B field, we're the primary losers from the policy, whereas most people benefit.

I'm coming at this from the assumption that there is some number of immigrants per year, X, that society is willing to tolerate. Would we prefer X Albert Einsteins or X random people? You are arguing that the former would cause a market distortion; maybe true, but still preferable, I'd say.

And I think that it doesn't really matter whether immigrants are enslaved to a particular company, or whether they're simply selected for a particular set of skills. If you allow only programmer immigrants, most of them will program, whether or not they're required to, because their alternative is unskilled labor.

> I think the H1B visa is too narrowly defined

I agree, but that's the power of lobbying for you. Actually I'm not in programming per se, but science (with computing aspects). Tons of H1Bs here for science as well.

It would make more sense if it were defined by salary rather than field, although may be hard to implement.

Very interesting argument. While the "serfdom" aspect of this is definitely morally repugnant, the most optimistic interpretation I can see for the H1B program is that it is supposedly a way to encourage or select for immigrants with high economic value.

It absolutely does distort the market, but if we are under a political constraint that we can only have X immigrants per year, it does make sense to take the ones that will be working in high-wage sectors, if we are out to maximize GDP.

> It will also ensure that the price of lemons will never reflect the opportunity cost, so the market signals that would bring prices back into equilibrium

Hm, I'm not sure. It seems like restricted immigration would basically move the supply curve for lemons to the right by decreasing cost of wages for lemons (only). Then because avocados are competing for the same land, avocado supply curve will move left.

The result will be more lemons at a lower price and less avocados at a higher price, but they will both still be in "equilibrium". In either case, the rate of return per unit land should be equal for both avocados and lemons, yes?

It seems the main effect of interest to us would be that wages for avocado-growing employees would be higher and lemon-growing employees lower under a restricted immigration policy vs an unlimited one. So, back to reality, I have never thought about the fact that H1Bs are actually good for everyone else's wages (outside the affected sectors) compared to unrestricted immigration.

If we concerned about "fairness", we could note that it might therefore also be more morally sound to restrict immigration to high-wage jobs because it would actually compress wage inequality and the low-wage natives would benefit. The losers in this scenario are high-skill natives and low-skill foreigners.

> The claim is the supply is inelastic in the short term.

That is correct, and it may be true. The reason I'm very skeptical is that we have been hearing this same spiel for 5 years at least. If companies are unwilling to raise salaries short-term, it will not incentivize long-term migration into this sector.

If you substitute "programmers" with "widgets", it is obvious what is going on here. Companies want "trade" (immigration) barriers lowered so they can get their widgets more easily and at a lower price. They don't care whether their widgets come from endogeneous US production or from abroad, so long as it is at a minimal price. Best of all for them if there are tons of widgets available both domestically and internationally, so that the price will be minimal.

> What does this even mean? Doesn't basic economics tell us about supply and demand. At a low salary, demand for developers will be high, and supply will be low. At a high salary, demand will be lower, and supply will be higher.

It means, "we're trying to convince people to get into this field because then supply will increase and we will have to pay less". Obviously, any individual company could fill as many positions as it wanted if the price were right.

You are quite right that "half a million open jobs" is a meaningless statement missing a variable, like quantifying speed by saying my sailboat moved 500 meters.

They're "high" relative to some (most) jobs and not others. It sounds like you may be implying something like "programmers make enough already". Perhaps true on a moral plane, has nothing to do with economics though. Why is gold so expensive? Because it's demanded. There is no inherent correct price to anything absent a market.

> The fact is that they could offer $1,000,000 starting salaries tomorrow and the pool of 'qualified' applicants would not change because salary is not the determining factor here.

OK. I think this is the core of the matter. Let's assume the number of "qualified" programmers is fixed over the short term.

The market is made of actors/companies. If a company wants more of the qualified programmers, and they pay $1M starting salary, that company will get them. And then the market price will rise. The fact that it hasn't risen to $1M implies the companies don't really want them that badly.

And in the long term, a price raise would encourage more programmers to become qualified and the problem would be rectified. But if what you are saying is true, and they really only want Stanford grads or equivalents, why would they care about H1Bs? It's not. They want to keep the price down.

Even if we assumed the number of qualified programmers were constant, there would be no more of a "shortage" than there is a shortage of gold. We don't say there's a gold shortage even though its quantity is "fixed" and presumably everyone would like more of it if it were cheaper.

I think people make a stronger argument for a shortage in the labor market WRT doctors. In that case there is a cartel (the AMA) keeping the quantity artificially low. There is no such phenomenon in programming. There's nothing stopping a university hiring all top-notch CS profs and becoming the next CMU/Stanford, and charging students a bundle...if the price were right.