At least in Germany a lot of research isn't being done at universities but at dedicated research institutes. Quite a few PhD students are working part time at uni and part time at one of the Fraunhofer Institutes or at Forschungszentrum Jülich. Another European example is CERN in Switzerland.

This model just doesn't conform with all those research university rankings.

I wonder how much it contributes.

American Visiting Researchers at CERN, since they remain attached to their Universities, probably count toward their institution's ranking. But a full time researcher at CERN doesn't.

The JPL at Caltech and the Lincoln Laboratory at MIT are the opposite of that. They do, what could be considered mostly government research but I'm sure 100% of their citation count goes toward their respective institutions.

Or even more importantly the Max Planck Institutes and Helmholz. Max Planck society is a heavy weight for sure and the co-operation with universities pretty unique as they need each other and are fully public (other than e.g. American private entities like JPL).

I agree. Also, some universities in Europe, like ENS in Paris, produced more field medal recipients than all top US universities combined. But way more people heard about Harvard than ENS. English helps too.

Citation needed—I see Harvard and Princeton as having 7 alumni each, which combined is more than ENS' 12.

From the list in Wikipedia it seems a lot of researchers move institutions after they were awarded, so maybe that is it. But yes, the score might change with time.

Not only English, but also the marketing through Hollywood.

mathematics have historically been a focus of scientific training in France for a very long time. I don't think it is representative of a more general trend TBH.

I've noticed EU CS grads are on average much stronger scientifically than US ones. With notable exceptions, the typical US undergraduate curriculum is more of a trade school. That’s fine of course, but definitely not the skillset needed for novel research or even properly applying recent research.

Edit: Apparently I’ve struck a nerve. But this has been my experience working and hiring at two of the FAANGs. The teams doing the most scientifically challenging work are predominantly European educated and the team members that aren’t are from the most elite US institutions.

I would not say that EU grads are better than American grads or the opposite. They both have different set of skills and strengths.

My experience is the following: I’m an European grad but did my PhD in the USA . In Europe tests were brutal and you had no help whatsoever from the professors. Your grade in a course was 80%-100% a single final test and lots of people would fail. Also, professors would not restraint criticism toward students.

In USA students seldom fail, grades are inflated in most universities and they not only depend on tests but also on homework and a variety of activities. They also spend lots of time doing extra activities and “leadership” stuff and learning how to self promote.

That being said, I feel that American students end up being much more driven and optimistic and many of them end up trying and succeeding at doing new things.

I am a grad student in a top european university in CS, and tough love has been my experience as well. We have access to the grade distributions and they are steep but not multimodal, with many students failing the first mandatory and notorious courses.

The usually grade depends on a single test, often oral which puts additional burden, plus, to even qualify for the test, students need to achieve well above 50% average on homework assignments.

I agree a lot with this, the European educational system is too much "learn by tough love" sometimes.

I think a less pot-stirring way of saying this is to just point out that most US colleges/universities are closer to Fachhochschulen than Universitäten.

The difference between the two regions is IMO mostly overstated. Especially in CS. Both the US and EU have institutions with trade school-style CS bachelors/masters as well as truly top research institutions with real Computer Science bachelors/masters degrees. Saying much more than that isn't particularly helpful.

NB: I'm willing to bet that most of the "hard core science" teams in Europe have Americans over-represented relative to other teams in the company. This says more about how immigration systems work than anything else.

Also NB: being on "hard core science" teams isn't the best way to maximize income/career trajectory at FAANGs, and isn't even the most secure... remember a buncha years back when Microsoft killed off its bay area research lab?

How do we always go from university rankings based on research results straight to comparing what in the US are predominantly Bachelor graduates?

I think it is abundantly clear that research prowess really has very little bearing on how competent you are in teaching say Linear Algebra. Which is how we got the whole MOOC thing in the first place, there is no magic Harvard Linear Algebra but there sure are tons of social mobility barriers and they don't really care for the curriculum one bit as much as having that upper strata origin Ivory league stamp of approval.

> The United States trains the best computer graduates ― by far: We find that CS seniors in the United States substantially outperform seniors in China, India, and Russia. The average computer science student in the United States ranks higher than about 80 percent of students tested in China, India, and Russia. Seniors in elite institutions in the United States similarly outperform seniors in elite institutions in China, India, and Russia by approximately 0.85 Standard Deviations (SDs). Importantly, the skills advantage of the United States is not because it has a large proportion of high-scoring international students.

https://blogs.worldbank.org/education/who-trains-best-comput...

Original paper

https://www.pnas.org/content/early/2019/03/12/1814646116

This is comparison between US and (China, India, and Russia). The original poster was comparing between US and Europe, which this paper does not cover at all.

For reference, the title of the paper is "Computer science skills across China, India, Russia, and the United States"

The EU CS grads that make it to the US are stronger than the average US ones. They are also stronger than the average EU ones.

I think that some elite US institutions provide solid CS foundations but the classes that supply this don’t also supply good scientific thinking. I think that the institutions thought that science and engineering requirements would have supplied this. However, that is very hit and miss based on the quality of the professors you happened to take for introductory science courses.

The second tier US schools are definitely trade schools. The professors do at best second tier research. The average student is weaker. Canada appears even more explicitly trade school like than the US with significant portions of schooling devoted to getting industry experience.

Also, the recent rise in compensation has relatively recently tilted even the elite institutions towards being trade schools. A greater and greater percentage of undergraduates at these elite institutions are majoring in CS.

Maybe survivor's bias? I assume you live in the USA. The europeans that FANG bothers to move across the pond might be only the really good ones attracted by the top payment. The mediocre europeans don't make it and thus remain in europe, outside of your view. Different culture, different success measures, red tape around visas, etc. all put larger cost onto international hires vs just hiring domestically.

Exactly.

I can think of a few reasons:

In many European countries, the five-year master's is still seen as the primary undergraduate degree, at least in universities. (There may be other more applied institutes with shorter degrees.) When most students are expected to graduate with a master's instead of bachelor's, the program is likely to be more theoretical.

Elite schools are a foreign concept in many European countries. In such countries, ambitious students motivated by status and success are more likely to pursue a professional degree such as law, medicine, or business. The ones that choose an academic program such as CS are more likely to be motivated by the subject matter itself.

You don't see a representative sample of European graduates. Moving to another country requires huge sacrifices. Most people from wealthy countries are unwilling to make those sacrifices simply to get a better job. Those that do emigrate for work tend to be more career-oriented than the average.

> Elite schools are a foreign concept in many European countries

The most well known universities in England, France, Germany, Switzerland are what you can call Elite Schools. Yes, even in technical fields like ETH, École Polytechnique in Paris, etc

By elite schools, I mean universities where a substantial part of the value to the students is the brand value and networking with people who can be expected to be particularly successful in the future. The UK has that at Oxbridge and to a lesser degree at other top universities. In France that means some grandes écoles. Germany traditionally didn't have elite schools, with the assumption that all universities are equivalent, but things may be changing. I'm not familiar enough with the Swiss system.

I am not impressed with european masters alumni and I have worked in europe and the usa, it definitely seemed like it was the bachelors of europe, everyone had it.. I believe it is more difficult to get into a masters program in the usa.

> But this has been my experience

Not sure why you're being downvoted, that has been my experience as well.

I had the complete opposite impression with compspci grads I met in Germany. For undergrad they had a lot of diverse job skills, opposite for US

Could this be due to EU universities studying specifically one subject while US universities require credits in other topics like Languages?

I think that is mostly compensated by bachelors taking an additional year.

I don't know if what parent comment says is true at all, but if it is I think it's because historically universities didn't have a Bachelor's/Masters system in Europe. At least in Germany and the Netherlands you would study for 5 years and that is still the expectation. For less academically inclined students there are other options such as trade schools.

I don't think so, or at least it's not necessary. A lot of my CS classes (US state university) were just kind of watered down so weaker students could pass. In short, grade inflation.

Do US engineering programs actually require foreign languages and other liberal arts coursework? That wasn't the case at UVA or VT in the 90s.

Yes, most U.S. universities require a broad range of subjects outside of one's chosen major. This is a good thing in my book. You have your whole life to develop an ever-narrowing level of specialization.

No disagreement here. But then my degree is in economics, not engineering.

IIRC, UVA's engineering curriculum didn't require foreign language, had it's own writing course, most of the remaining cross-over outside engineering was hard science, with a small number of gen-ed electives (maybe 1/year or 1/semester?).

Compared to my Econ degree, which required 2 years of language, 2 semesters of writing, a mix of hard science and math, and somewhere around 4 gen-ed courses per year.

I wasn't required to take foreign language classes but my engineering degree did have a significant liberal arts core. This is pretty standard in the US from what I've seen.

Lol at my state school a cs course counted towards the foreign language credit for engineering majors... that said i did feel there was enough liberal arts requirements overall.

Yeah, many of them do.

In my medphys program we had a unit on how to get research funding after grad school. The instructor explained that the most likely source by far is industry, followed by government, and last academia/other. It's been this way for a while as I understand it.

I believe this is almost 100% wrong - reversed, until fairly recently. European basic research was in the toilet for decades; they often did the second experiment in everything.

Most government agencies like NIH and NSF are funding mostly basic research, because as you note private companies do not invest in this.

> application-focused (something that EU is not very good at)

Probably because of the investor climate, and applied research is a good fit for companies.

Research in the U.S. has tended towards applications due to promotion of agricultural productivity research and agricultural extension services by various U.S. founders in the 1700s and due to the land-grant university system established in the 1800s.

https://en.wikipedia.org/wiki/Land-grant_university

Citation needed? Never heard or noticed any bias like this before.

Look at university research labs, and see where they get their funding from.

Companies tend to fund applied research.

Funding in Europe, whether academic or industry motivated, is simply lacking across all scientific, engineering, and social science disciplines.

You seriously think that university research in the US is largely funded or driven by companies? This is utterly wrong.

*Edit*. The question is why the united states has the "best research universities". That is a very general-level question and the Parent made a very general, and very wrong, claim in response. Regardless of whether some specific fields might have more industrial funding than others, it's easy to google the overall portion: and industry funding is quite minor.

https://www.aaas.org/sites/default/files/2018-11/UniSource1....

Depending on the field, it can be. Lots of University-affiliated labs in the medical field get significant industry funding.

> is largely funded or driven by companies

GP never implied it was the largest contributor, and yes, depending on your field a significant amount of funding does come from the private sector.

This seems on topic: https://www8.gsb.columbia.edu/newsroom/newsn/10721/the-color...

> You seriously think that university research in the US is largely funded or driven by companies? This is utterly wrong.

Slow down there bud. Where did I say that?

Unless you’re talking about AI/ML, then yes, university labs receive more from companies, than government funding agencies.

> Slow down there bud. Where did I say that?

If I miserepresented you, then I am sorry.

> Unless you’re talking about AI/ML, then yes, university labs receive more from companies, than government funding agencies.

This is simply untrue. Companies comprise less than 10% of overall university research funding, and AI/ML is far from dominant. It's easy to verify this.

Have you heard of Bell Labs?

> Have you heard of Bell Labs?

While I would tend to agree that private funding for research is way better in the US, Bell Labs is not necessarily a good example.

True, the tech. output of the outfit was amazing, but they had massive govt support (among other things, they were given an official monopoly by the USG).

It wasn't just the monopoly itself but their revenue and profit structure was driven by the research (cap ex and op ex were treated differently by the consent decree, and they had performance metrics to meet as demand grew). Research also had valuable P.R. value.

I don’t know why anyone should believe you. Can you source some concrete facts if any?

This is a provocative comment ready to start flame wars.

Please tell specifics and not paint broad brushstrokes to flame the fires. I’m sorry to flag your comment.

Edit: One source we can look at the awards given to fundamental research. They’re overwhelmingly US-based. Fields, Nobel, Turing, etc. This contradicts the popular generalization.

A bit of an overreaction here? I’m pretty sure it’s ok for people to share their personal observations.

I don’t think so. I’ve seen time and again that it leads to completely unpleasant nationalism flame war.

Did anyone read the article at all?

I'm surprised the trajectory of the forum would make you think that way. Also meta-conversation is very dangerous to focus.

Meta-meta-conversation is always fine though.

> They’re overwhelmingly US-based. Fields, Nobel, Turing, etc

This isn't really clear. For example, EU has generated about twice the fields medal recipients as the US. That's reasonable in terms of roughly equivalent populations, but does beg the question that if someone grows up in Europe, trains and spends early career in Europe, but ends up a professor in USA before being given a Nobel prize, where do you count the result?

Someone having positive opinion of non American is not starting flamewar. Overreaction to it is. Threatening to flag it just because you disagree, or because someone having positive opinion of something in eu hurts you feeling is absurd.

So far, this thread has been the most uninteresting. These types of opinions lead to completely unsubstantiated back and forth. It’s not even remotely close to truth unless we have some facts to disclose. Threatening to flag!? I apologized to flag and provided reasons for it. Far cry from a threat. Obviously, you’ve all agreed that it was an overreaction.

I’d say, the entire article is flamewar prone. How can you say something is best?

Why is this provocative? It was certainly not meant as a black and white contrasty "it is always true"-law of nature, bit rather someone describing vague undercurrents based onpersonal experience?

Weird overreaction on your side tho.

The commenter said no references, it seemed like an expression of broad opinion/experience, not like a strong declaration of truth.

"European and American universities both have strengths and weaknesses" is not that provocative.

> US-based. Fields, Nobel, Turing,

How many of these were actually educated in the US, as in from primary school, and how many came to the US to complete a PhD after having received an education outside of the US?