Why does research cost so much money?

Question

I have been dragged into an argument with someone who can't understand why millions are being raised to fund ALS research (that's the "ice bucket challenge", love it or hate it). He doesn't get why research costs so much money because - and I genuinely quote - "it only takes time and effort".

My first instinct was to say "Are you being serious or just trying to wind me up?". But then I realized that maybe from an outsider's perspective, this might actually be difficult to understand.

So I am asking for points to make when explaining the cost of research to lay people, and how to articulate these points in a way they can identify with them.

Answer 1

it only takes time and effort

I think the answer to the question is already there. It takes time and effort. So it means it takes at least the money to pay the people for a long time. And effort means you need a lot of people.

More precisely, for medical research, reagents, animal model, clinical test are really expensive. Many different drugs need to be developed to have only one working in the end. Finally when a drug seems promising, you have to do year-long clinical tests, just to ensure patients safety.

So time and effort == lot of money

Answer 2

People's salaries cost money. Given the overheads of running a business or university, plus the cost of fringe benefits like retirement and health insurance, you can basically double someone's salary to get the full cost of employing them (whether the number is directly charged or comes in through a national healthcare/retirement scheme). If their talent's are in demand, then companies may bid up salaries in order to attract them away from universities in order to try to make a profit on their labors.

STEM talents are in demand.

Therfore, each PhDed person who works on a project costs somewhere between $150,000 and $400,000 (salary is half that, recall) in a broad range of fields relevant to solving problems like ALS or cancer or other diseases.

Research is hard. Most drugs don't pan out, so you need lots of people trying lots of different ones in order to develop some that do work.

Mathematics is cheaper to pursue. That can literally be one person reading and scribbling and talking to colleagues until she has a breakthrough. These folks often get paid primarily to teach, but if they're willing to work on algorithms for companies or the government, then they are worth a lot, so salaries also get bid up.

Answer 3

I'm an experimenter, so in addition to the cost of some highly trained people's salaries and benefits and the cost of keeping the lights on and the white boards cleaned there is equipment and consumables.

Some of the equipment is precision kit. Lots of it are produced in small runs because only a few hundred sites in the whole world need that kind of stuff, so there is no economies of scale. Some of the consumables are pretty exotic and cost a whole heck of a lot.

All of this applies perfectly well to the kind of research mentioned in the question, plus they have to deal with human subjects concerns and that doesn't come cheap either.

---

And just because I like to talk about my work...

Because I'm in big science, when it come time to stop tinkering around with prototypes and testbeds and get really serious we usually build a one-of-a-kind multi-ton detector with thousands or tens of thousands of instrumented channels. Giga-bytes per day data streams, massive computing infrastructure, hundreds of salaries and travel money.

The question isn't "Why does it cost so much?" but "How do you expect us to get it done with so little?". Then we go ahead and try because we're happy to have the chance to do it, even on a "shoestring".

Answer 4

I think it depends totally on the area of study.

Like everybody mentioned already, only the human costs are already relatively high, and equipment for bio sciences is crazy expensive as well.

Let's just make some estimates.

An average neuroscience lab has at least one EEG machine, which for those levels cost between 30 and 70 thousand bucks. You need good computers, which usually range from 1500 to 2500 USD, at least one for each member, in a 5 person group. That is already ~100,000 USD only in basic equipment.

You sometimes need to pay for your space in some universities, like a rent, having access to live animals to do experiments also costs about 2000 bucks per experiment for small animals, larger animals have all sorts of extra costs, I do not know, but having a chimpanzee should be costing a lab well over half a million bucks/ year to maintain and do experiments.

Want an MRA to do functional scanning of the brain, first get a University that has one, also crazy expensive, then pay the rent.

Conferences plus travel expenses is another sizeable chunk.

Thanks to the ripoff that Academic publishing is, Universities pay hefty sums of money to get access to research journals.

All of these costs are not even taking into account salaries for Professors, PostDocs, Grad Students, etc

Answer 5

Beyond the cost of lab spaces, researchers, guinea pigs, machinery, and raw materials, I would like to add a few costs caused or worsened by the lack of open science (a great TED video about open science if you aren't familiar with it):

• Paid access to research articles. In addition to preventing the general public from having easy access to and contributing, it has two main consequences:

  • Universities must pay some insane amount of money to access publications, typically a few millions USD yearly.
  • Even the most highly-ranked universities don't have access to all articles (far from it), which mean researchers sometimes waste time to try to get access to some papers.

• Data sets unavailability or price:

  • Unavailability: researchers sometime have to create their own data sets or abandon project ideas due to the unwillingness of others to share their data sets, because sharing data set might mean wasting some publication opportunities.
  • Price: some researchers do share their data set, but but not free. For example, in the natural language processing community some of the key data sets are only available on the Linguistic Data Consortium website, which charges download.

• Unreleased source code. Many publications don't share their source code that was used for the article. A couple of reasons might explain the behavior: just like for data set, access to source code gives an edge over other researchers, who will have a harder time improving, amending, etc, the article. Also, the source code might be badly written and researchers can be embarrassed about it. It might be a way to avoid other people finding bugs in your code that invalidates some of your results. I asked Are there any journals or conferences that take into account the availability and the quality of the source code when selecting the papers to publish? one day, even in 2014 it is hard to find... Also, see reference on availability of source code used in computer science research articles: "in computer science systems, out of 410 papers that were analyzed, only 85 has a link to source in the paper".

• Too many articles due to overreliance on bibliometrics to assess researchers (grants, promotions, etc.), which push researchers to over-publish. I sometimes feel that I am a documentalist, trying to navigate my way through myriads of papers that are written in some unnecessarily complicated way with barely any contribution.

• Lack of good, widely-used platforms to publicly comments on existing articles. E.g. if something is unclear in an article and a researcher spent his morning to understand, there is no good platform for him to leave a comment to help other future readers who will have the same understanding issue.

• Publication bias, which results from the fact that positive results have a much higher chance to be published than negative findings. This slows down research too: see the article Why Most Published Research Findings Are False or the TED video: Ben Goldacre: What doctors don't know about the drugs they prescribe (really worth watching if you don't know the extent of the problem).

• Page limit of the articles inherited from the Middle age where articles were printed and distributed through paper form (and charged $30 per article, without any dime going to the authors' pockets). This forces authors to chop off some of the information they would have liked to convey. How many times did you wonder how the authors performed some mathematical derivation? How many times did you wonder which parameters did the experimentalists used?

All these inefficiencies cost money too. E.g. How to Make More Published Research True:

Currently, many published research findings are false or exaggerated, and an estimated 85% of research resources are wasted. (in biomedical research)

---

Some key points from the TED video Ben Goldacre: What doctors don't know about the drugs they prescribe:

Former drug company researcher Glenn Begley looked at 53 papers in the world's top journals, and found that he and a team of scientists could NOT replicate 47 of the 53 published studies — all of which were considered important and valuable for the future of cancer treatments.

Half of all clinical trials ever completed on the medical treatments currently in use have never been published in the medical literature. Trials with positive results for the test treatment are about twice as likely to be published, and this applies to both academic research and industry studies.>

In 2010, three researchers from Harvard and Toronto identified all the published trials for five major classes of drugs, and then measured two key features: Were they positive, and were they funded by industry? Out of a total of 500 trials, 85 percent of the industry-funded studies were positive, compared to 50 percent of the government-funded trials.

According to a 2011 study in the Journal of Medical Ethics,4 nearly 32 percent of retracted papers were not noted as having been retracted by the journal in question, leaving the readers completely in the dark about the inaccuracies in those studies.

Answer 6

Even in purely theoretical fields or fields without expensive lab costs, research is expensive.

I think the root of the problem is that your friend seems to be measuring 'cost per result (e.g. cost to cure ALS)'. Most lay people probably don't understand what it means to learn something truly new about the world. Discovering something new is really, really, really hard—even for very intelligent people (remember Einstein's 1% inspiration, 99% perspiration). Thus, whilst it may be true that "it only takes time and effort", the amount of those things needed to achieve even minor progress is huge. To achieve a major breakthrough, like curing a disease or proving a major theorem, might take thousands of scholars working for years or even a century (see, e.g., the Poincaré conjecture).

The cost to achieve one of these major 'results' is indeed large, but that is because they stand among the crowning achievements of mankind.

Answer 7

To deal with this answer more generally, I would insist on using their own argument: nearly everything on earth is just a matter of time and effort.

This includes everything from The Great Pyramids to battleships to cars, books to research, farming to mining. They are all products of the application of effort and time.

The reason things cost money is not mostly because of the raw materials. With most things in life, the raw material is the easy part!

I like farming/gardening as an example. Much of a garden or even a farm can be done with little to no raw materials. You need dirt, sure, but that stuff is everywhere! It isn't always good for growing, but you can fix that with time and effort to till, cultivate, amend with more cheap stuff (waste products and so on). Bugs? Squish them - time and effort. Seed? Harvest it wild, make your own selective cultivation to raise yield - time and effort.

Unlike gardening, though, research isn't growing the same crops over and over again every season. With growing food, it often gets easier each year if you do it right - you learn what works and what doesn't, the soil can be made more fertile. But with research, once you have your product you don't get to just repeat the same process again and call it good. That wouldn't bring about a publishable result - no one cares if you "discover electricity for the 40th time."

When research is easy and quick, it gets done easily and quickly, and then there is no reason to do it again. The low hanging fruit gets picked, and it doesn't grow back!

Therefore, we can expect research to get harder as we as a society learn more - more time, more effort per discovery is required. This means research should only get more and more expensive, while making and growing stuff should get cheaper and cheaper. Economics (research!) is more complicated than this of course.

So let's talk about what else costs money. 'Stuff', raw materials, do cost money. Research needs labs and offices, electricity/fuel, and most need all sorts of special equipment and apparatus. Medical labs need stupid amounts of these, and everything costs so much money because research is picky - poisoning someone with an injection because you didn't spring for that stupidly expensive "approved" beaker instead of using a $1 drinking glass from the second-hand store that wasn't resistant to the chemical regents you used is generally frowned upon.

And if a person thinks human time and effort should be free and people should not ask anything in exchange for their life dedication and work - why in blue blazes are you still talking with this person? Suddenly, a relevant XKCD:

And if the Senator seriously thinks medical research - or any research - should be cheap or free...well, good luck to you in your noble fight, friend.

Answer 8

The other answers address the cost of actual research extensively enough, but there are, sadly, other aspects to consider.

Academic research is, for the majority, funded by the government. That means that, although I have a high opinion of some government funding agencies, the decisions on how to spend the money is often political. As any other government organization, members of academia work in a loop to protect their interests and privileges.

As a result, in addition to the natural amount of effort and money spent on endeavors and projects that turn out to be dead ends (these are actually useful to research as they rule out possibilities), there is a tremendous amount of money lost in generating heat.

It's frequent to see faculty being hired because they are the spouse of another faculty, or because they have social traits that contribute to the 'diversity', grants being attributed hoping that the awardee will return the favor, funding being attributed to people on the basis of the number of publications instead of the quality of them (even more so now with the open-access movement and the subsequent logarithmic raise in paper count), etc. Not to mention cases (e.g. in medicine, socio-economics, history) where funding goes to promote the political agenda of the governing majority.

At the end I think at least a good third of academic research is either marginally incremental, frankly redundant or even completely bogus.

It should be noted that while these sums ('millions!') seem huge to us, the amount of money spent in academic research is typically a fraction of the government's budget.

Answer 9

You could put in perspective by comparing it to other ventures.

For instance, running Walmart "cost" around $450bn last year. Running IBM cost $115bn last year. Running Pfizer cost around $20bn last year.

Now, compare this to the cost of finding the Higgs boson: around ~$13.25bn over the total time taken. Hubble: around $10bn up to 2010.

Scientific research is not cheap, but it's way cheaper than running a business and has much wider reaching impacts. What sticks in people's minds is that the money comes from "funding" and "charity" and that strikes them as "waste".

Answer 10

Research does not necessarily cost a lot of money. Even today, some areas such as literature, philosophy and theoretical mathematics (non-computational) literally cost only as much as a desk, pen and paper. In theory. In practice, as many have already pointed out, even in these fields it can make sense to do a little cost-time trade-off.

But what about the other fields? Since the question is about ALS, I will focus on more technology-heavy fields. I will also try to explain more the barriers to making the same research cost less, rather than try to guess what that particular researcher's budget was.

• You need to be associated with a university or institute, and they take a big cut out of your funding. A vast bulk of research is done by university professors. While on the face, you might think that such institute-affiliated researchers take a monthly salary, and then get money from the government on top to spend on their research, that isn't quite true. When you receive a grant, the university or institute takes a large amount of this, eg. 50% or even 80%. Part of this is probably just taken because the president wants a nice house. But it also pays for the electricity in your office, the electron microscope everyone gets to use, the salary of the janitor, the salary of the health and safety people whose approval is required by law for you to obtain restricted research chemicals, etc. You could save a lot of money by not being a university affiliated researcher, but you can't just up and go and do research from your garage - you must deal with a lot of bureaucracy unless you want to get fined or jailed, and you lose the very important benefit of easily being able to eat lunch with leading scientists and talk to them about science.
• Materials are expensive. Others have explained in detail why consumables (chemicals, enzymes, single use sterile tools such as petri dishes) are expensive. A lot of these you could in theory make yourself. I know many biology labs who eschew modern kits and still use DIY methods from decades ago to save money - but it's a lot of work and introduces a lot of risk for error. Even then, some crucial reagents are simply impossible to manufacture if you don't have a large chemical plant. Think by analogy to computers: You can do a lot with DIY electronics, but nobody is going to be building an i7 out of scrap metal in their garage.
• Equipment is very expensive. Even the simplest biological research equipment tends to run from thousands of dollars to hundred of thousands or even millions. Even something as basic as a centrifuge can run you two grand, and you cannot do any molecular biology without one. If you want to do sequence based research (absolutely necessary for ALS), either you must buy a very expensive sequencer or you must pay someone to run samples on theirs.
• Scientists need to eat. Research isn't a hobby, it's a full-time job. Perhaps the professor's salary gets paid by the university - but often grad students and postdocs are paid from the grant money. All these people must be paid a salary, otherwise their landlord will kick them out and they will starve. It would be ridiculous to expect someone with a 40-hour job to do a few hours of research every weekend and get somewhere.

With ALS and biology in particular, there are also some "soft" factors contributing to research costs: Firstly "big data" and high-throughput studies are currently in vogue, and both require expensive, specialized systems. Second, it is getting harder and harder to find problems that aren't expensive to solve.

There are many studies which can be done for a $10.000 in reagents (which isn't a lot, other costs notwithstanding) over 1-3 years. Many labs already do this. But not every lab is this lucky.

What happens if your disease has hundreds of variants which must be characterized by spending thousands on genotyping (it costs a few hundred per person, to cover things like cost of manufacturing the genotyping chip, cost of chemically preparing the sample, salary of people who do the specialized genotyping, and the profit margin of the genotyping company)? What happens if the protein that causes the disease must be purified using an extremely expensive chemical? What happens if the aberration that causes the disease is so microscopic, that you need a million dollar microscope to see it? What happens when the thing you study turns out to be so complex, that only a supercomputer (which are very expensive to have or use) can hope to make sense of the data? What happens when you are studying a very dangerous pathogen that will kill you unless you have a $100 million BSL-4 lab that takes $2 million every year to maintain?

What's worse is that biology has been around for quite some time, and a lot of the questions that are "cheap" to answer have already been answered by someone else. Problems like ALS, which have stayed unsolved for decades, are unsolved for a reason: Sometimes it's just because nobody happened to be clever enough to come up with the right idea, but a lot of the time it's because the scale of research needed to attack them was prohibitively expensive. Now technology has advanced, and it is no longer prohibitive, but still expensive (meanwhile the technology affording that discount is itself not cheap).

---

it only takes time and effort

Certainly not. There is some research to which this applies in a trivial way, but a lot of important research is practically impossible without using the technological infrastructure built over the centuries of human history, and using that infrastructure is rarely free.

Why does research cost so much money?

How much does he think it should cost? Based on the quote, I suspect it is $0. That's not gonna happen.

It costs what it costs because that's what it costs. If you can come up with your own budget for a given study that is much smaller, yet still makes it feasible, go ahead. But just writing numbers on paper and asking why it doesn't cost this much is wishful thinking. You don't go to the car store and tell the guy you think his Ferrari "ought" to cost $32.27 because that's how much cash you happen to have in your pocket.

points to make when explaining the cost of research to lay people

The condensed version of the above is: Even smart people sometimes cannot solve a problem without buying expensive things. You cannot study moon rocks without buying an expensive rocket to go to the moon. You cannot study what's inside the atom without buying an expensive atom smasher. You cannot test a cancer drug if you don't have expensive cancer cell cultures to test it on.

Science is not just sitting around in a room and philosophizing. You must also do experiments. The experiment's set up can sometimes get involved and complicated. Observing the outcome of the experiment can require specialized sensors or measurement tools. Processing the data gathered can require state of the art supercomputers.

how to articulate these points in a way that they can identify with them.

"You get what you pay for."

Answer 11

One part of the answer that was not mentioned so far is that highly skilled people want to get paid good. Or to phrase it differently: You will not find people with the skills you need to do research without paying them accordingly. Especially for pharmacy / medical studies you need a lot of man-hours of people who probably made a PhD at university. They spend a lot of their money to get to this point and they want to get paid accordingly.

As those studies require - as already mentioned - a lot of time, you need to pay a lot of expensive man-hours.

Some numbers

• The median education debt for indebted medical school graduates in 2012 was $170,000, and 86 percent of graduates report having education debt. Specifically the 2012 the median debt at graduation was $160,000 at public institutions and $190,000 at private institutions. (Source: How Much Does Medical School Cost?)

Answer 12

The research group that I've been fortunate enough to work with as part of my Masters dissertation (Microelectronics) have recently received funding of £400,000. To me that sounded like an immense amount of money; but then I saw some invoices for how much it costs to repair and maintain the antiquated machinery they're forced to use.

Plus, does your friend really not understand that "time and effort" cost money in and of themselves before you even start thinking about equipment and whatnot? Does he think that researchers don't need a salary because they are immune to starvation and the elements? US$1million would pay the salaries of 5 research assistants for 4 years; and those salaries are not exactly generous.

Answer 13

I would emphasize who's time and effort is involved, and how much of it is involved. It isn't like we have a couple of undergrad research assistants working on the problem so we can pay them next to nothing. ALS research occupies the time of teams of doctors.

Second they seem to be underestimating the type and length of the work involved. It isn't like they can go over to CVS and say I have this prescription for this never before seen drug whip some up for me while I browse the magazine rack, or go to some medical supply store and order a never seen before medical device.

Third, safety protocols, it isn't like they fabricate some new concoction in the lab then go jab it in some guy's arm like they do on TV. It takes a long time because they are actively trying to not kill people, and exercise an abundance of caution.

Answer 14

One point to make is to realize that what is "so much" is a very relative term. If you see value in what is being produced then the cost may seem small and definitely "worth it". If you do not see the value, then of course the result is the opposite. It would be possible to go through, for example, public expenditures in government budgets and argue for military costs, for social benefit schemes, public health, public schools etc. and find similar disagreements.

It is becoming popular in certain political circles to want research to be directly translated into profits within a short time span. Although there of course is nothing wrong with gaining profitable results many discoveries have only yielded such results long after the timing of the discovery. The spin-offs from discoveries are also often realized once the basic research has been done, for example as a result from unexpected discoveries. As the saying goes, "If I knew what I was doing, it wouldn't be science".

Hence, I think the view of "cost so much" is either a lack of perspective, or insight, or simply not finding the progress of any value. Having this view point is certainly valid but as many responses have already discussed, one must then argue what costs should or could be reduced.

To call research inefficient is not sufficient. We can see that certain problems can be solved if sufficient finances are supplied over a sufficient period, take the US space program to put a person on the moon as an example. But, was that a cheap project? Obviously for some but not to others.

If no time or effort was spent on anything, life would remain simple but hardly bearable, provided our intellectual development would occur spontaneously.

Answer 15

First think about all of the other challenging and expensive endeavors we do. Running a business, building a bridge (or writing a computer program)... these were all done before by someone else. And if you are looking to build a bridge you are probably reading a book titled something like "how to build a bridge" and written by the last person to attempt what you are wanting to do.

Research involves solving problems that are open, there is no book to read because it has not been done before. Now, you can gather a lot of materials from this person over here that had one idea that helps you, and that person over there. But these scattered resources are harder to find (I.E. in academia) and at the end of the day you are the one who needs to put it all together and build the very first bridge in existence.

Now do us all a favor and put a whole bunch more money into safety testing.

Answer 16

This is well explained by an American expression: "Time is money."

You have to pay "people" for their time. And these are not random people, but highly trained and paid researchers making salaries, typically in the top 20-30 per cent of the population.

Also, such researchers normally have to be supported on most projects by "advanced" scientific equipment, which also costs a lot of money.

Answer 17

Apparently your friend who says "it only takes time and effort" has never seen the proof that time=money. Since effort=time, it's clear that time=effort=money.

In my field (computer science), grant support for a middling-size research project might last five years and provide support for two PI's, a postdoc, and two to four graduate students. Let's look at an example personnel budget. Note- you can find lots of NSF or NIH sample budgets online that illuminate this further.

Assume both PI's ask for two months of summer support per year. If their base salary is $90,000 for a nine-month contract, then their base monthly salary is $10,000. Thus, if they both ask for two months, then that's $40,000 per year in base salary for the PI's.

Then you add on PI fringe benefits costs. This money provides benefits like health care. Different universities have different fringe rates. But, let's say that our rate is 15%. Then we have an additional $40,000*.15 = $6,000 per year.

Postdocs are similar, but they're paid full time to do research. They might have a base salary of $55,000 per year. Then their fringe benefits are $8250

Now let's talk grad students. Let's say we have three graduate students. Each graduate student has a monthly stipend of $2,500- or $30,000 each / $90,000 combined per year. Then you charge fringe on those students, which is going to be slightly less, say 10%. Thus, the fringe for all three per year is $9,000.

These are all the "direct salary costs". To recap, we have:

         Salary  + Fringe:
PI 1:    $20,000 + $3,000
PI 2:    $20,000 + $3,000
Postdoc: $55,000 + $8,250
Grad 1:  $30,000 + $3,000
Grad 2:  $30,000 + $3,000
Grad 3:  $30,000 + $3,000
--------------------------
Total:   $185,000 + $23,250 = $208,250 per year

But, these are only the direct costs. All grants also include what are called "indirect costs", which are paid to the university and go towards things like building maintenance, utilities, and non-research-staff salaries. A pretty normal indirect cost rate is 50%. So then we also have:

Indirect Costs = 0.5 * Direct Costs
$104,125       = 0.5 * $208,250

Finally, our total yearly cost:

Total Cost = Direct Costs + Indirect Costs
$312,375   = $208,250     + $104,125

Just like when you go to the grocery and just throw stuff in the basket, it adds up a lot quicker than you think. Notice that all of this is just to support six people per year (two PhD PI's, one PhD postdoc, and three grad students). And this is just counting salaries, it doesn't include other costs like equipment, publication fees, or travel.

The salary-only cost for a three year project under the above would be $937,125. The salary-only cost for a five year project would be $1,561,875. Note that the actual costs would be slightly higher in real life because of things like pay raises and cost of living adjustments.

If you consider published papers to be the basic unit of academic research, then suppose that this project turns out 8 papers per year (which is reasonable but optimistic). The cost per published paper is then a little over $39,000 per paper.

Answer 18

Well, in academia most of the funds go into salaries and I feel like people who never employed anyone just don't know how expensive this is. Here in Austria 3 years of PhD cost the university around 110 000€ and a Postdoc (the more experienced researchers) is around 200 000€. So if you got 2 PhD Students and a PostDoc working for 3 years, and that's not a very large Project, it's around 420 000€ just salaries, but nothing happened yet.

Now the other costs depend very much on the research topic and field, but in Chemistry you need to buy chemicals and also other supplies. They might be expensive or cheap, but for such a project you might have another 100 000€ of budget. keep in mind, that's only around 900€ a month for every researcher. Now I don't know if this would be a lot or not for something not familiar with Chemistry and I have to agree, that's not a bad budget but also not extremly large.

Let's say that's it for this project and we don't need any new equipment or something like that. But now, at least in my country, there's also what they call "overhead", another 20% of that what's in the budget right now goes to the university. Why? Because they provide the infrastructure (HR people, non-scientific personell, electricity, equipment, labspace,...) which of course also produces cost. So this rather small project would be around 624 000€ with nearly 70% (420 000) just salaries.