I will answer this question quantitatively, and revise as I collect more data.

Definition: I am defining 'demand' as the number of tenure-track faculty searches at research universities.

Methodology:

• I perused past issues of Physics Today, where most of the faculty searches are posted ~August-November. So far, I have only looked at October and November 2014, but I will update with other months/years if I track down the old issues. During the fall, this is available online to anyone (Jobs | Physics Today Jobs), but these postings are removed after some time.
• I counted the job postings (note: this just indicates that the university planned to do a search, not that they hired someone) by field. Some of the postings were rather specific, so I fit them into a category as best I could. If I am not familiar with the field, I might have done this wrong.
• I only counted research universities in the US. Since research and teaching universities fall on a spectrum, some might disagree with what I counted or excluded. Also, national labs were excluded, because there were a shit ton of postings to sift through just considering universities.
• Searches in departments other than physics (EE, MechE, etc) were excluded. Open searches (no subfield specified) are excluded from this summary, as are fields with only 1 or 2 positions in the time period of study. The latter might be added if I collect more data.

Results:

• Condensed matter physics: 30 Experiment/14 Theory
• Astrophysics and Cosmology: 14 Experiment or Observation/11 Theory
• High Energy physics, Nuclear Physics, and Particle Physics: 9 Experiment/7 Theory
• AMO: 6 Experiment/4 Theory
• Biophysics: 5 Experiment/4 Theory
• Soft Matter: 3 Experiment/3 Theory

Discussion:

Almost every field has more experimental positions (demand) than theoretical ones. This makes sense because experiments require more manpower (which requires more graduate students and postdocs, which bring in money to the university via tuition and/or overhead) and materials/supplies (which brings in money to the university via overhead). Condensed matter is particularly skewed towards experimentalists, but there are more theory positions in that field than any of the others (at least in 2014). In astro/cosmology and HEP/NuPa, the demand for experimentalists and theorists appears to be more balanced. As I collect more data, I will add to this summary.

It should be noted that this coarse-grained approach doesn't capture the granular demand which exists within each subfield. Someone in a less "in-demand" subfield might be the hottest thing on the job market one year because their niche is very trendy/fundable and the converse is also true.

Edit: Here are results from sept/oct 2012:

• Condensed matter physics: 13 Experiment/5 Theory
• Astrophysics and Cosmology: 5 Experiment or Observation/5 Theory
• High Energy physics, Nuclear Physics, and Particle Physics: 7 Experiment/3 Theory
• AMO: 1 Experiment/1 Theory
• Biophysics: 4 Experiment/0 Theory
• Soft Matter: 1 Experiment/0 Theory

Edit 2: here are results from this past year (jobs posted in Physics Today Aug-Nov 2017). Notably higher relative number of astro jobs than years past, presumably because of recent Nobel prize. Also lower number of jobs overall, perhaps because of political/economic uncertainty.

• Condensed matter physics: 14 Experiment/5 Theory
• Astrophysics and Cosmology: 9 Experiment or Observation/11 Theory
• High Energy physics, Nuclear Physics, and Particle Physics: 7 Experiment/5 Theory
• AMO: 9 Experiment/1 Theory
• Biophysics: 4 Experiment/1 Theory
• Soft Matter: 3 Experiment/0 Theory
• Quantum optics/optics: 3
• Quantum information: 2

all years available here: Phys Today Job Posting Field summary

These are excellent questions, although I wonder on the distinction made here between the two. Theory and experiment are inextricably connected, especially in modern times when serious science is done in the form of collaborations.

Particle physics and Condensed Matter physics seem to be two major disciplines of contemporary physics where we need both of these. Since there is a large base of theoretical knowledge already available, I would say that we need a bit more experimental concentration here. But, having said that, since we have more accelerators and laboratories available now, perhaps it would not harm to have more theorists coming to explain what we measure or observe.

A relatively newer and very important field is Biophysics where once again we need expertise in both the theory and experiment.

There is one field where perhaps theorists excel more and mostly, that is computational physics. This is a powerful tool which is used in virtually all physics, and even in economics and biology. This is one area where we need more theorists to come and contribute to the fields computational physics addresses.

Experimental physics is not for the curious. Curious people can learn more in a few hours in the library than an experimental physicist can learn in a year.

Experimental physics is for the adventurer. Think of the early explorers, working under difficult conditions, in foreign lands, short of money, short of food, not knowing if what they were pursuing even existed. They lived with discomfort. They were seeking discovery, and they did it by surrounding themselves with things they didn't understand but struggled to figure out. And they were constantly attacked by the natives.

Experimental physics takes patience, lots of it, endurance, and good judgement to know what is worth pursuing. Even more important, it takes a large degree of self-skepticism. You have to know when to give up. A former graduate student of mine, when talking about his recent Nobel Prize, told of the many project I had going in my lab at the time he was working there. Most of those projects went nowhere, and had been quickly abandoned. And yet it also takes self confidence, a willingness to continue even though the referees have rejected your application for funds.

When money is short, or I am tired, I do theory. That is much less stressful, but not nearly as rewarding.

Many people dislike experimental physics because of the uncertainty, the long time it takes to get anywhere, the difficulty of analyzing data with high self-skepticism. But for the few, the proud, the experimentalists, it feels like living the life of the ancient explorers, of Captain James Cook, of Admiral Perry, of Burton and Speke.

Interesting question!

I was a practicing experimental nuclear physicist for nearly 50- years but, deep down, I was a wannabe theorist but, although my math was pretty good, I felt that it didn’t quite make the grade. More importantly, I felt I never really had any great ideas that I wanted to pursue with the analytical tools that I did have. I kind of fell into doing experiments; I wrote or was co-author to almost 100 technical publications and, early on, I realized that I liked the feeling of being part of some of the minor discoveries that I and my colleagues made. I have worked closely with theorists who helped me better understand some of the discoveries that we made and I did envy their skill and ability to do that.

However, as I got older, I appreciated more the experimental experience that I did have because it allowed me to be rather mobile in my career. After 20 years as simply studying the structure of nuclei for the simple reason that “they were there”, I was asked to help with a problem in applied physics. Through that experience, I found I really enjoyed doing and producing something that somebody was essentially able to “use the next day” to make what they needed to do a bit easier. And so, for the rest of my career, I did these kinds of problems - problems that had direct applications to real-world problems and which were more like engineering problems.

Reflecting on that career and the question at hand, I think the following story might explain it. As a young teenager, I was enrolled in a private high school and was all excited that I could sign up for a large variety of shop courses that the school offered because I had decided that I wanted to be a carpenter or a machinist. I wanted to work with my hands and build things. My parents, on the other hand, essentially told me that they were not going to pay the high tuition cost at the school so that I could attend to “play with wood”. I was told in no uncertain terms, that at this school “… you will attend so that you will get well educated so that you can go to college; after that, if you still want to be a carpenter, that career path will still be open to you.” Upon entering the school I was greeted with upper-level courses in Latin, algebra, and other required college-prep studies which transform the high-school experience into real and often painful work. I howled and complained but all my protestations fell on deaf ears! Somehow, I endured that - and much more - and did it successfully, So, perhaps, my becoming an experimental physicist was my compromise career choice over being that great carpenter or machinist that I once dreamt of being!!

Yes, you can’t design an experiment unless you understand what you are doing. People don’t understand how difficult measurement is. For example most people think when you stick a thermometer in the air you are measuring air temperature, but the instrument also absorbs thermal infrared radiated by the surroundings. How much of the registered temperature is due to air temperature and how much is due to the radiant environment depends on the thermometer geometry and airfow, and can’t be easily estimated. So something as simple as measuring air temperature involves radiation transfer and fluid dynamics, and if far, far more complex than you would think. (It turns out the propagation speed of sound is the best way to measure pure air temperature, but you still have to get the pressure and water vapor content in order to calculate air temperature. Oh, and measure simultaneously in different directions to cancel out wind speed).

All of this helps explain why experimentalist take almost twice as long to get their PhD as theorists. But theorists get all the glory.

Before I studied physics in college I had mentioned to a college professor that I didn’t think I was “good” enough to be a theorist. He gave me a look and said “experiment is the higher calling”. It took me over a decade to fully understand the truth of that statement.

It’s a lot easier. Theoretical physics requires that you not only get a Ph.D. in Physics but also that you get the equivalent of a degree in advanced math. I managed to be an experimentalist for over a decade with differential equations as my highest mathematical training. That wouldn’t cut it for a theoretical physicist. Nowadays they have to learn some esoteric stuff like group theory or non-Euclidian geometry depending on their speciality.

I once read a science fiction story set in the not-so-distant future. The 50-year-old protagonist was a theoretical physicist who was still in graduate school. He wanted to be a particle physicist, but by this point it took decades to “learn the basics.” I wouldn’t be surprised if it comes to that.

There’s also the pragmatic issue of getting a job. Industry doesn’t have much patience for employees who spend most of their days just thinking about the universe. So most theorists are driven to academic positions which are limited in number and, assuming you are a conscientious person, very labor intensive.

While I went the route of becoming an experimentalist, I was in awe of theoreticians. It’s a difficult profession.

A2A by John Mandlbaur.

(Awww, did you miss me, John? I can’t say that I actually missed you… and thoroughly enjoyed the all too brief respite from your insanity).

If physicists are so clever and I am wrong,

There’s no “if” involved at all here… Physicists are “so clever”, and you are clearly wrong.

then why can't physicists predict how much a ball on a string spins faster with any accuracy,

They clearly can, John. Your apparent inability to understand this, or how they can actually do so, only serves to highlight your fundamental ignorance. Perhaps if you actually bothered to learn all of the actual relevant physics and maths, you’d have a better perspective on this.

and why does my theory make surprisingly accurate predictions?

Well, to put it simply, John, it doesn’t. In fact, the only “predictions that it does seem to be able to make inherently involve you cherry picking only the selected data points that actually agree with your claims, while fudging and blustering nonsense about all the data points that outright falsify your ideas… which isn’t too surprising, really, because as I’ve proved, mathematically, in several previous answers to your questions, your alleged “theory” is completely inconsistent with everything we actually know about reality.

What you do just isn’t science, John, and no one is under the slightest obligation to take any of it seriously.

And, no, John, this answer isn’t a “gish gallop” simply because I’ve broken your question up into sections… it’s point by point ridicule and mockery.

Theoretical physicists should never be allowed in a lab. They break things. (Joke!)

Generally there is a disconnect between theoreticians and experimentalists. Theoreticians think about physics from a more abstract perspective, whereas experimentalists need to relate a theoretical prediction to an actual experiment and a potential result that they can relate back to the theory.

Experimentalists are nuts and bolts people who think about what's possible based on the operating principles of various widgets.

Most theoreticians never think about the widgets.

Experimentalists love widgets. When they're not building experiments, they're pouring over catalogues of new widgets thinking about what they could build with them. Every physics conference is filled with widget makers trying to sell their latest high-tech widgets to experimentalists. The theoreticians are usually elsewhere chatting about all the wonderful things that are just waiting to be discovered just over the rainbow.

I've been a little facetious, but not too much. I've worked with a number of theoreticians and there really is a skill involved in bridging the disconnect between what they'd like us to do and what we actually can do. That said, collaborations between experimentalists and theoreticians can be extremely fruitful and rewarding.

Excellent question. Thank you for asking me.

Yes, indeed, if you have a good life-long preparation in mathematics, you enjoy computations, and you are a born experimentalist (ie experimental physics is your forte’, you enjoy doing experiments and have a talent for it, your fingers have that peculiar dexterity etc. etc.), you are a true physicist and can be an asset.

A lot of good physicists do both, and in fact, part of a good physicist’s training involves both. For instance, in my training as a particle physicist, I had to do both things. We did cross section calculations, made theoretical models and did extensive and long calculations, at times with complex numbers and once with large 36 x 36 matrices etc. Even my dissertation’s three long chapters were exclusively devoted to theoretical modelling, experiment development and simulations of my problem.

Even beside the work I do at the moment, in developing experiments, in addition to doing calculations involved with it, I try to read stuff and think about theoretical bases of those. I am not good at advanced theoretical stuff (such as Riemannian geometry and manifolds, etc) involved with GR etc. but I love doing analytical work and long calculations, often off my head without using a calculator.

I have replied similar questions about theoretical and experimental physics here on Quora, you might like to read them.

Well if you are asking this question after watching "the big bang theory" then I guess you would favour theoretical physicist superior to experimental physicist as Sheldon is one.
But when it comes to real life then there is no comparison between them,because they both are like shadow and object.As the shadow is invisible until light is thrown over an object so are the both physicist.Consider a new discovery as the light,which falls on the body I.e theoretical physicist, now it will reflect the shadow until it is proved experimentally,so here the shadow acts as experimental physicist. So the purpose of light is only fulfilled until unless the shadow is visible,so as picture is completed.
Now,in todays world people only believe whose proof is available.Just consider if NASA wouldn't have published the images of Pluto then how would we believe their spaceship reached pluto,same stands for theoretical and experimental physicist, the theoretical physicist give up theory but experimental physicist prove it.So no comparison.:)

There are two types of experimental physicists: researchers and innovators. Researchers have a successful career in a vertical sector, mostly in an academic environment, and will push the frontiers of understanding by a disciplined, long term systematic effort of canonical experimental knowledge. Then, there are innovators. They will probably jump among projects and disciplines, move between academy and industry, move across applied and fundamental research by building an heterogeneous understanding. In both cases it is an exciting life adventure, and it is mostly a matter of personality and instinct to push you in one direction or another.

An experimental particle physicist would have a great deal of exposure to things like radiation (health physics), radiation detection, electrical engineering and signals analysis. If you want to put aside the experimental particle physics (and give up the search for the God particle, for instance), you could go into a career in one of those fields.

Personally I'd find medical imaging to be highly interesting. It really exercises math skills, computer skills, medical knowledge, etc.

Health Physicists jobs:
http://www.indeed.com/q-Health-Physicist-jobs.html

Medical Physics (nuclear medicine, designing radiological imaging systems):
http://en.wikipedia.org/wiki/Medical_physics

Signal processing is an area of electrical engineering and applied mathematics

that deals with operations on or analysis of signals, in either

discrete or continuous time, to perform useful operations on those

signals. Signals of interest can include sound, images, time-varying measurement values and sensor data, for example biological data such as electrocardiograms, control system signals, telecommunication transmission signals such as radio signals, and many others. Signals

are analog or digital electrical representations of time-varying or

spatial-varying physical quantities. In the context of signal

processing, arbitrary binary data streams and on-off signalling are not considered as signals, but only analog and digital signals that are representations of analog physical quantities.

In the United States, the person within an organization responsible for the safe use of radiation and radioactive materials as well as regulatory compliance. An organization licensed by the Nuclear Regulatory Commission to use radioactive materials must designate a Radiation Safety Officer in writing.

http://en.wikipedia.org/wiki/Radiation_Safety_Officer

[I speak here concerning high energy (particle) physics and astrophysics; the reality might be slightly different for solid state physics or other sub-fields.]

Both positions are insanely competitive, at least as far as getting a tenure-track professorship or permanent position at a national lab.

Although one might consider the theorists as being at the “top of the pyramid,” and it is indeed true that somebody who is not at the absolute top of his graduate class will be heavily discouraged from pursuing a theory career, successful experimentalists need to be incredibly sharp, and will often have a deep and comprehensive grasp of theoretical physics as well.

The reason there are fewer theory positions is frankly that progress in the field *needs* a larger number of experimenters, and that an experimenter who is merely very good can make important contributions, whereas the “second-level” theorist is much less vital to progress. So there are much fewer theory positions open at any given institution.

Physicists are in high demand in many different industries.

They're wildly popular in finance. In fact, they're partly to blame for the recent economic crises. The "quants" (quantitative analysts) put together models that were deeply flawed but had the imprimatur of PhD physicists and mathematicians. That led them to put too much money into things that were too risky, and we're still dealing with that as it unwinds.

Any field that requires applied mathematics can be done by a physicist. Experimental physicists have a particular advantage in that they've been involved in construction and management of large projects, especially experimental particle physics, since particle physics these days is done almost exclusively in large, expensive works. Only a few are actually involved in the management and design of the equipment, but they're used to working in and managing teams.

So you end up with former physicists in all sorts of fields. Just among my friends, I've seen them in intelligence (the government kind), computer software, civil engineering, web site troubleshooting, aviation, weapons inspection, and even law.

A2A.: No. NO! Hell, NO!!!

But —in a different sense, kind-of/sort-of, somewhat, guardedly… y-yes —but not necessarily (nor preferably) in that order.

First of all, as written, the only sensible answer to the question is the first sentence in Mark John Fernee's answer (see also Simon Bridge's and Jonathan Hardis's answers), and without his disclaimer. If the Pauli effect does not convince you, I can provide first-hand (and quite disastrous) experimental data on the topic.

To be clear: theoretical physicists do not test their theories using the machinery that experimentalist make. That is one of the things taht experimentalist do: Theorists theorize, experimentalists experiment. The two approaches to understanding Nature are complementary, and require very different skills, and even mindsets. I know no one who has both.

Now, with this “subtlety” clarified, there is a “yet, somehow…” By the nature of its logistics, cheap tabletop experimenting can be done haphazardly and free if any theoretical preconceptions. That is how Alessandro_Volta discovered batteries! That is how Alexander_Fleming discovered penicillin! That is how Michael Faraday discovered electromagnetic induction! None of these epoch-making discoveries (to name but an extremely very few!!) have been performed with intentions of testing a theory—as some theorists might have developed it! It is true, of course, that those experimentalists do know the established theoretical background which pertains to the experimenting that they perform; they don’t just step off the proverbial ledge “to see what would happen.” The emphasis here is however on the motivations and intentions —and therefore the breadth and width of experimenting, when not focused on testing a specific theoretical prediction or chiseling a measurement to one more decimal place. (Not that there is anything wrong with that either! We do need all kinds of experimenting!!)

Such experimenting (and discovery!!) “unbridled” by intentions to test an existing theory is of course still happening today.
(••• Experimentalists: do chime in, please! •••)

But no longer in elementary particle physics.

By the nature of its logistics, multi-billion-dollar installations such as CERN exist and operate by the graces of extremely complex ekono-socio-political syzygy of some 23 partnering member-states and 50 observer-nations… Such experiments do in fact get designed to test a known theory. (Can you imagine convincing hundreds of politicians and bankers to approve the use of land and other multi-billion dollar resources to “see what happens if…”???) It can, of course, happen that an experiment designed to “X” ends up indicating “Y” —wholly unexpectedly, surprising both the experimentalists and the theorists. However, in carefully tuned multi-billion-dollar installations, chances of that happening are essentially tuned out.

…and therein lies at least one of the difficulties of the predicament in which elementary particle physics (a.k.a. fundamental physics) finds itself in the past several decades.

It’s incredibly context-dependent. In my world, it’s mechanical, electrical, optical, and software engineers that work with physicists (like me) who have gotten into biology to make medical diagnostic devices. They tell me what machines they make and how they make them, and I tell them how well their designs are working and what implications that has for infectious disease screening.

The guy who sits next to me is a PhD physicist and works on heavily theoretical stuff - his engineering colleagues are EEs, mostly, who do freedom-to-operate and patent search kind of things.

A dear friend of mine (PhD in physics; dissertation on string theory) is a Big Data guy nowadays. He really only works with software engineers. Another dear friend of mine (PhD in physics; dissertation on physics education) works with very few engineers - he’s setting up lab equipment for demonstrations and teaching lectures for classes.

If you’re looking for a very general answer, anyone who does experimentation is going to need a good engineer to design, set up, calibrate, and maintain their equipment. This includes (but is not limited to) astrophysicists, condensed matter physicists, biophysicists, and particle physicists. A large number of experimentalists simply buy their equipment from engineering companies.

Forgive me for answering a different question, but that is not really quite the right question. It doesn’t work that way!

Rather “Classical Physics” is a less accurate representation of reality.

Reality *is* quantum. Science in general, and physics in particular, is an exercise in continually deepening and improving our understanding. In the early stages of understanding, our theories for how things work are simpler, but also less accurate. Physics proceeds by searching for new evidence and checking our existing theories against it all the time.

Classical physics is a good approximation, and allowed us to do some great things like build telescopes and steam engines and so on. However, as we learned more, and collected more data, there were things that we saw happening in the real world that classical physics couldn’t explain. This really came down to the care and precision with which we observed things. As we observed things more carefully, classical explanations didn’t make sense.

A good example of this is the “orbits” of electrons around the nucleus of an atom. At the end of the 19th century, an atom was thought to be like a currant bun, with electrons being the currants stuck to the surface. Then people started firing particles at matter and the particles would often pass straight through. They worked out that atoms were mostly empty space. So a, classical, model was thought up to explain it. This imagined a nucleus like the Sun and the electrons like planets orbiting the Sun.

This idea didn’t last very long, because electrons moving in a field generate electromagnetic radiation. That means they emit energy. Classical physics (Newton’s 2nd Law) says you can’t create energy from nothing so the electron, in orbit, would have to loose energy. That means its orbit must decay. When physicists did the maths for this, they calculated that all of the electrons in all of the atoms in the universe would collapse into the nucleus in tiny fractions of a second. Our universe could not exist based on Classical Physics!

To explain what really happens, confirmed by every experiment that has been tried so far, you need to treat the atom, and electrons, and their “orbits” as quantum things. Electrons don’t “orbit” they exist as a probability cloud, in a superposition of locations around the nucleus.

So Quantum Physics is the best description of reality that we currently have, though sometimes hard to understand. Classical Physics is a rough approximation that is a very useful tool in some circumstances, but does not really describe reality.

That is a delicious question. Let me try to answer it off the top of my head.

Let me consider just faculty, and those who do research actively.

When I look at rankings for universities, typically there are about 400-500 listed. So lets say the total number of universities/institutes in the world is about 1000.

IISER Pune is smallish, and has about 30 faculty, of which roughly 15-20 are experimental. Larger departments can have 100 faculty, smaller ones are typically larger than 15 (smaller institutes are dominated by theorist-heavy departments). Based on this, I'd say the average number of physics faculty at these 1000 institutes is about 60, of which a reasonable estimate would be half are experimental physicists.

So thats 1000 * 60 * 0.5 = 30,000 experimental physicist faculty.

I'll bet an ice-cream (which you must collect in person) that I'm within say a factor of 3 of the right answer.

In current physics research, it's almost impossible to conduct either an experimental or a theoretical PhD without a significant computing component. Computers are the tools of the trade in all areas of research; experimental physicists will typically write codes to acquire, process and analyse the data from their experiments, as well as develop and work with codes which use first-principles models to derive expected experimental signals. Theoretical physicists rely heavily on computers too; there are very few areas in which work is done solely with analytic theory (paper and pencil approach). All theoretical physicists will at some point find themselves translating their theoretical models into computer code to study the behaviour of a given system numerically.

That being said, there is little doubt in my mind that the post-PhD career prospects are a strong function of the amount of computing evident in the student's work. The reason for this is pretty simple; modern, high-tech companies want employees who code, and there are not many better ways to demonstrate coding ability than to have developed one's own programs for use in cutting-edge research. Essentially, there will be a 'sweet spot' between the full-on experimental researcher who spent most of their time fixing vacuum flanges and aligning laser optics, and the abstract theorist who spent years immersed in advanced tensor calculus, filling pages upon pages with advanced mathematics. Companies outside of academia are looking for someone between the two extremes, who has shown the ability to grasp the theoretical fundamentals and translate them to working, efficient and novel computer code, as well as to see how their numerical calculations align with the results coming out of experiments.

The foregoing is also likely to be true of most positions within academic research; the demand for researchers who have demonstrated the ability to develop their own code and apply it in a way which bridges the theoretical/experimental divide is almost certainly higher than for applied mathematicians or dedicated, full-time experimentalists. I should add that this situation has obviously developed over the half-century since the advent of electronic computers; prior to this, there was more room in research for the 'pure' experimental or 'pure' theoretical physicist.

That's a great question. My point of view is that it 1) really is an historical matter and 2) related to what is the measure of smartness.

1) Few hundred years ago, people were able to do both, and they actually had to do both, I mean experiments and theory, to progress. But now is a time where it is extremely difficult to do both, because both domains have made tremendous progress. Therefore, this dichotomy is nowadays meaningful, while it was not 'back then'. You can also see that in the average fields change physicists are involved nowadays. While, before, they were able to cover many fields, jump from one to another, it is now much more difficult. All the more for experimentalists, who very often have to deal with very complex, time consuming, and expensive setups. I would say that it is easier for theorist to do such fields change and therefore to look like smarter.

2) Celebration or fame also depend what is easier to 'sell', 'present', etc; it also depends on the historical context, the personality of the one you choose. For instance, Marie Curie got two Nobel prize, but she was a woman, not playing drums, and no picture of her fooling around with a nice, pink, tongue. So, although she is the only person winning two such distinctions in science in two different scientific fields, she isn't that famous. Also, I can speak only for the country I live in, there is a strong feeling in society that abstraction is sexier than pragmatic considerations. So this societal considerations may induce bias as 'smarter' people will, in average, choose theory if they can. In few years, you could end up asking: who is smarter, physicist or quantitative finance analyst? In many of the best French schools, best student choose quantitative finance, because of money and because it is highly math based.

And fame, I guess, is also related to the simplest way to remember things. It's easier to remember that the famous Boson that has recently been identified "belongs" to M. Higgs rather than to a scientific conglomerate. Note also that it sometimes is a question of timing, or else. For instance, who knows or remembers the name Englert or Brout or Hagen or Guralnik? (I myself had to google that)

I am not sure this is a very structured answer, but my point is that I don't think one can rate each "species" of physicist. There are too many parameters that influence the result. What I am sure of is that when you see a bright one, you will recognize him/her. And what I am sure of, is that for Feynman, Landau and Heisenberg, they have been discussing hours, days, all their life, with experimentalists, and that cross fertilizing is most often where the richness comes from.

Theoretical physicists often understand the extrapolations of a theory to regions far beyond technology. A theoretical physicist like Hawking will extrapolate quantum mechanics back 14.2 BY to when the universe was at ‘maximum’ density. Or he could extrapolate the same theory 100 BY into the future, in the that death of the universe.

Hawking may know in a vague sort of way how general relativity applies to an atomic clock on an airplane. However, he is not interested in sorting out the different processes that occur in the atomic clock. His career is making extrapolations into an unforseeable future.

Experimental physicists often know how the same theory would apply to laboratory equipment on a bread board scale. Physicists Hefele and Keating applied general relativity and special relativity to atomic clocks, some strapped to couches in airplanes. They analyzed the data from actual experiment, not hypothetical twins traveling in imaginary rocket ships.

The calculations performed by Hefele and Keating were not really extrapolations of relativity. They were right on applied to a set of clocks on real platforms and real airplanes.

These were the same type of calculations used to make an accurate GPS system. So Hefele and Keating were not extrapolating into an unforeseen future. Their techniques were used by their children, maybe even by them. Their great grandchildren are driving cars with navigational devices programmed to alonmg the lines of the calculations done by Hefele and Keating.

To my knowledge, neither Hefele or Keating said anything about the Big Bang or the Heat Death. I don’t even know if they were qualified to make such an extrapolation. However, they really understood how general relativity exists on a bench top level. So they were experimental physicists.

Finance:

ex-Brother in law:

Ronald Kahn, PhD in physics from Harvard

Managing Director, Global Head of Scientific Equity Research

Ronald N. Kahn, PhD, Managing Director, is Global Head of Systematic Equity Research at BlackRock. He has overall responsibility for the research underpinning the Systematic Active Equity (SAE) products.

His service with the firm dates back to 1998, including his years with Barclays Global Investors (BGI), which merged with BlackRock in 2009. Prior to joining BGI, he worked as Director of Research at Barra, where his research covered equity and fixed income markets.

Ronald Kahn is a well-known expert on portfolio management and quantitative investing. He has published numerous articles on investment management, and, with Richard Grinold, authored the influential book Active Portfolio Management: Quantitative Theory and Applications. The two of them are the 2013 winners of James R. Vertin award, presented periodically by the CFA Institute to recognize individuals who have produced a body of research notable for its relevance and enduring value to investment professionals. He is a 2007 winner of the Bernstein Fabozzi/Jacobs Levy award for best article in the Journal of Portfolio Management. He serves on the editorial advisory boards of the Financial Analysts Journal, the Journal of Portfolio Management and the Journal of Investment Consulting. The 2007 book How I Became a Quant includes his essay describing his transition from physics to finance.

He teaches the equities half of the course, "International Equity and Currency Markets" in UC Berkeley's Master of Financial Engineering Program.

He earned an AB degree in physics, summa cum laude, from Princeton University, and a PhD in physics from Harvard University. He was a post-doctoral fellow in physics at University of California, Berkeley.

(This reminds me of the riddle, How is a raven like a writing desk?)

You're right, each kind of physicist requires knowledge from the other. An experimentalist benefits from additional knowledge of engineering in order to design and construct experiments. But their jobs are quite different.

When reporting their findings, experimentalists often interpret their own data, and there is even some cultural pressure to do so. But ideally the two activities are kept separate to lessen the effect of personal bias on the data.

The same division exists as well in other sciences. A theoretician's laboratory might be a desk and a computer or stationery. An experimental psychologist needs that, plus a room with volunteers or a zoo with its animals. Thus a possible solution to Lewis Carrol's unanswered riddle How is a raven like a writing desk? is that each could occupy, or be, the laboratory of a psychologist.

It wasn’t always like that but these days, the two roles are pretty much separated to two distinct occupations. Almost every physicist may be pretty much clearly classified as either an experimental physicist or a theoretical physicist. The detailed thickness and character of the separator depends on the subfield of physics, however.

There is a class of physicists in between. Enrico Fermi was surely an excellent experimental physicist and an excellent theoretical physicist. Most experimental physicists are mediocre theorists and vice versa.

A physicist should be at least very good in at least one of these activities.

No, I would say that most EEs are a far cry from exptl. physicists. I’ve worked with both (matter of fact, I’ve BEEN both!) and, in general, a good way to tell an engineer from a physicist is the engineer knows the formula or equation to use to solve the problem (or design the widget), but cannot derive the formula from first principles, because it was never part of their training, whereas a decent physicist CAN derive the formula. So physicists go much more in depth in regards to the mathematical development of what they work with. Not to be negative about engineers, there’s plenty of smart and competent ones out there, but I can state this having been on both sides. Just one little “for example”: when Stanford U. was designing the components for SLAC, their linear accelerator, they needed very high power klystrons, higher than any previous designs. So they sent their requirements to several hi-tech corporations, such as RCA and GE, and the replies were that you couldn’t build them to that level of output. So the Stanford Physics and EE departments got together and built a prototype, then sent their design specs to the companies that initially said it couldn’t be done to have the rest built.

The subfields in physics aren’t that important. There is a significant difference between theorists and experimentalists. Even there, the differences aren’t fundamental.

Theorists and experimentalists both have to work with computers. So employers may be looking at the computation skills. They may not care about theorists and experimentalists.

If you major in nuclear physics, you will probably find that there are few jobs in this particular field. However, a research group may want someone knowing about imaging. The mathematics of imaging and nuclear scattering aren’tthat different. Or maybe they want someone skilled in spectral sorting.

The mathematics isn’t that different. Most employers know that. Most aren’t so dumb as to want a narrow specialist. They will give you about two to five years to catch up to whatever field they hire you for.

They first hire you to be a problem solver, not a intellectual know it all. Your ‘specialty’ develops AFTER your first hire. You should not be asked to repeat your PhD thesis the first weeks out of school. You should not expect to repeat your PhD thesis.

A beginning physicist is expected to have a broad understanding of physics. He can often get a beginning job in a multidisciplinary field that merely includes physics.

An employer may not even care about the difference between a chemist and a physicists. If you really want to be a physicists, you can not ignore chemistry. Most down to earth applications of physics will involve chemistry.

So don’t worry too much about the subfield in physics. Instead, worry about having a broad based knowledge of all physical sciences AND computers.

A broad knowledge base is better than a deep knowledge base. There are few jobs for narrow specialists.