Update on the Budget Proposal: NSF

Back in March, I wrote about OMB Director Mick Mulvaney’s budget blueprint. I discussed, among other things, what the budget and his commentary appeared to imply about the OMB director’s views on how science advances and what its role is in our economy.

Because of the fallout of the events of the last few weeks, less attention will likely be paid to the white house’s new budget proposal: A New Foundation for American Greatness (note that this is only a proposal and will likely not get passed in its current form). Unlike the proposal released in March, this budget proposal mentions the National Science Foundations (NSF).

The proposed cut to the NSF is approximately $800 million or 10.7% of the current budget, and when asked about this, Mick Mulvaney joked about the fact NSF funds had gone to a musical about climate change “last year” , before mentioning that the this white house intends to cut back on climate change research. I should mention that it is true that the NSF awarded a grant to develop this play, although it is worth noting that its goals went beyond merely getting the word out about climate change and the fund ended in 2014.  More importantly, this $700k, 4-year expenditure is being used to cut over 1000 times that much from the annual NSF budget. Of course, the overall budget for any environmental/earth science research is a little over twice the proposed cuts, so it is likely impossible that all of the cuts will come from climate change research.

It is more likely that these cuts will be dispersed across the board, leading the NSF to prioritize existing projects over new ones, leading to a sharp decline in funds for new researchers many of whom would be forced out of science altogether. If this prediction is accurate, and these cuts are maintained for a few years, fewer students will be able to get advanced science degrees at universities across the country, leading to a significant reduction in the size of the scientific community. Undoing the effects of these cuts would likely cost many times what will be saved and take significantly longer. To be fair, Director Mulvaney only intends to harm climate science: harm to the other fields of study is merely a price he is evidently willing to pay.

This was all covered in my first post, when I discussed cuts to the Department of Energy and National Institutes of Health. NSF funding differs from funding from these and other agencies in one important respect: it comes with an outreach requirement. The outreach requirement not only ensures that recipients of NSF funding participate in events intended to increase public interest in and awareness of the sciences, but helps facilitate a dialogue between the scientific community and the rest of society.  As I mentioned in my first blog post, this dialogue becomes increasingly important as fields of study advance, becoming more specialized and insular.

The administration’s cavalier attitude towards science education represents the closest any administration has come in recent history to being openly anti-science. To be fair, these cuts are not localized to science: they part of a larger set of cuts meant to help fund increases in defense spending, nuclear security, and border security, as well as tax cuts. The problem with making these decisions about scientific funding in comparative ignorance is that the usefulness of basic research is difficult to gauge when you know a field well and extremely difficult to gauge when you don’t. This means that when it comes time to either increase or decrease spending on research, it would be best to consult with scientists and engineers in a wide variety of fields to discuss the ramifications of each change if a politician wants to approach this decision with prudence.

This brings me to the other possibility for funding cuts: that they are NOT spread out, and are localized to climate science, as the White House claims. While scientists outside of atmospheric sciences would now only have to contend with cuts made to other th agencies, the atmospheric science community would have to contend with a particularly brutal decimation since the NSF accounts for 59% of federal funding for environmental sciences. This contraction would arrive alongside severe cuts to other sources of climate science funding (discussed at greater length in the budget blueprint released in March), leading to a drought for climate science research in the United States. This drought would likely immediately end the careers of climate scientists across the country, and make it nearly impossible for new climate scientists to enter the field for years to come, since any newly-minted PhD’s would now have to compete for scraps with an over-crowded field of veteran researchers.  Some might rejoice at the diminished funding into studying anthropogenic climate change. I would like to remind these people that the damage would spread to all climate science research, partly because no one studies “pure” climate change and partly because the cuts don’t leave much money for any other type of climate science.


Grad Students and the Proposed Tax Plan

I recently came across this article in The Washington Post by David Walsh, a grad student studying history at Princeton. The article discussed the ramifications of removing Section 117 the US tax code on the taxes paid by grad students.

Section 117 makes certain school expenses (i.e. tuition and supplies) tax exempt. According to Walsh’s article, the tax plan proposed by the House would eliminate these tax exemptions.

Before proceeding, I would like to state that I have not had the time to read the plan in detail, so I cannot comment on the particulars. If I were to speculate, I would say that eliminating Section 117 is likely part of a larger attempt to go after tax deduction and close loopholes that allow people to avoid paying taxes like. Eliminating Section 117 would be consistent with, though not equivalent to, taxing all non-monetary income in the private sector. To be clear, I am NOT saying that these two types of tax exemption are equivalent, to say that would be a major fallacy.

Now, to the consequences of the proposed plan…

For middle class families of undergraduate students, the elimination of Section 117 would make college significantly more expensive, likely furthering the opportunity gap (by which I mean the difference in professional opportunities) between families of different means. I mention this primarily because, while this article is about graduate students, I would be remiss if I did not mention a more pressing human consequence of the plan.

The consequences would be particularly stark for graduate students. Masters students do not typically get financial aid, perhaps because a graduate degree is not considered as vital to a person’s prospects as an undergraduate degree. For students paying out of pocket, this means a significant amount of debt and student loans, which typically consumes a sizeable portion of their disposable income for years after graudation. Masters and other professional degrees (i.e. M.D. or J.D.) offer an important (and sometimes a vital) leg up in many sectors of the professional world. Students already go into considerable debt when they get an advanced degree, so increasing this burden will  drive down the number of students seeking admission to these programs.

One major issue is the cost of university degrees, which has sky-rocketed in recent years. It may well be argued that universities really don’t need to charge so much for degrees and that this proposed overhaul of the tax system would force them to lower their costs. However, this argument does not take the elasticity of demand for advanced degrees into account. Students already go into considerable debt when pursuing graduate degrees in the hopes that pursuing a Masters or Doctorate will advance their careers. Increasing the burden of debt on students will likely result in students simply taking on additional debt, with no real costs born by universities or lending agencies.

Even if the number of students seeking graduate degrees were to diminish, universities would still be able to continue filling their graduate programs. Graduate programs are already highly exclusive: universities would likely still be able to fill their programs with students they would have otherwise rejected if fewer students were to apply.  In other words, doors will open for students who can afford to pay taxes on the income they devote to tuition.

The starkest consequence will likely be for Ph.D. programs, where students’ incomes are largely derived from their stipends. Before I go any further, I would like to clear the air about what Ph.D. students actually do.

When I was a graduate student, the majority of people with whom I spoke assumed I was spending most of my time in lectures and occasionally doing teaching work. This was true during the first year of my program (the masters portion of the Ph.D.), as it would be for most graduate students. However, during the second or third year of graduate school, Ph.D. students diminish their class load and eventually stop taking classes altogether because the purpose of graduate-level coursework is to prepare students to work for the university, typically as researchers. When students cannot convince someone to pay them a full-time salary to do research, they support themselves by teaching. Teaching does not advance a student within a program: it is merely how they support themselves while they work.

A Ph.D. is granted after the a student’s body of original research becomes sufficient in breadth to convince the faculty of a department that they are going to be able to be productive academics or researchers. The time required varies between departments and universities, but what does NOT vary is the fact that students spend most of their time doing research. In other words, they are working full time for the university. This, along with the fact that Ph.D. students are typically paid a stipend, made it seem odd to me that I as “paying” a sort of invisible tuition.

This is where the repeal of Section 117 comes in. As Walsh points out, instead of paying between taxes on a $18k-40k a year stipend, students would pay also  pay income taxes on the $30k-60k tuition,  meaning that students’ would be paying taxes on a total income that is over twice their actual income, because they are being spared the cost of tens of thousands of dollars they should theoretically be charged in order to work full time and almost never take classes at a university. I will follow the example of the article and use Princeton as the example.

An elephant in the room here is that it seems stupid to charge Ph.D. students this tuition in the first place. There are a few counter-intuitive and administrative reasons for students to “pay” tuition (which is really paid for by their department, advisor, or external grant/fellowship) which I will not go into at the moment for lack of time. For now, the most important fact is that this practice of charging (then waving) tuition for Ph.D. students is here to stay.

If this tax bill goes through, then a student at Princeton earning approximately $32k a year would have a net income of approximately $18k a year as opposed to $29k or $30k, coming to about $1500 a month for rent, utilities, food, and other daily expenses. This is not viable in that area, and the university would not be lowering rents around there because of the revenue generated by  the legion of students and faculty paying rent. This leads to two possible outcomes:

(i) Princeton increases the graduate student stipend to adjust for an after-tax income that would be livable in the area

(ii) Princeton does not change the stipend, so students are all forced to find extra sources of income.

Let’s look at the scenario (i) in the case of engineering and science departments (sticking to what I know). Lab groups in these department are usually responsible for paying for their own graduate students, and the new, much more exorbitant costs of graduate students will make them an undesirable expense for most groups. Already, postdoctoral fellows, who come in knowing more than graduate students are less expensive at most universities, so many groups CURRENTLY need to be incentivized to spend resources on costly graduate students who require significant training. Therefore, most departments will cut back significantly on the number of Ph.D. students they admit.

Furthermore, because the funds for these departments come from external grants which are primarily governmental since academic research tends to produce public goods, the large taxes effectively paid by the research groups themselves would mean that the taxpayers would be getting less “bang for their buck” in research. Let me say that again: in scenario (i), by trying to save taxpayers money, the new tax plan will ensure that taxpayers get less for the money they pay.

So, let’s look at scenario (ii). Here, departments would not have to pay more for Ph.D. students, but the students themselves would have to find extra sources of income to make up the lost income which, in many cases, could be as much as $10 k. The first problem that comes to mind is the difficulties students might have finding part-time jobs, particularly when universities are located in college towns. For now, let’s assume that there are no such difficulties and that somehow, part-time work abounds and students are all able to find reliable sources of extra income to provide them with a living wage. Even making $5k a year from these jobs will require significant amounts of time for students to spend NOT working for their lab groups. Their attentions will be split, and the student’s group’s productivity will suffer. This will  ALSO be a blow to the bang for the taxpayer’s buck spent on research. It doesn’t stop there though: Ph.D. students already tend to find their workload to require 100% of their attention so adding the implicit requirement of an extra part-time job will prevent Ph.D. students from focusing their full attention on their studies. This will likely lead to many more Ph.D. students burning out than currently do, thus reducing the labor pool of prospective scientists, engineers, and academics.

There is a silver lining here: the job market for people like me who got their Ph.D.’s before this tax plan. We would face reduced competition from a younger generation of scientists for grants. Given our overcrowded job market, this is a very non-trivial gift for us. However, I don’t think I need to make a case for the fact that improving the job prospects of current Ph.D.’s by reducing the number of future Ph.D.’s is not beneficial to the country as a whole.

One common argument I have heard throughout this debate is that students used to pay for their education by also working part time therefore students today should be able to do so. While there is truth in this premise, there are a few problems with drawing any sort of conclusion from this fact. Firstly, the costs needed to make up for short-falls in income were significantly smaller than those needed for today. Even adjusted for inflation, tuition has increased sharply over time. Secondly, most people who had to “work their way through school” were undergraduates. Thirdly, the competition for post-Ph.D. jobs has become RADICALLY more fierce, limiting the amount of time students can spend outside of a Ph.D. program and still land a decent job.

This last consideration is particularly important. Most of these stories also come from a time of almost unnatural prosperity in the US most of our potential economic rivals were recovering from World War II. During this time of prosperity, the government was able to fund research in a comparatively more lavish manner than they do today AND they were willing to due to a competition with the USSR. The labor market for researchers was also significantly smaller than it is today, so that finding a position was easier than it is today (this is not to say that it was easy). This meant that putting yourself at a slight disadvantage by partially diverting your attention was not as fatal as it would be today to your prospects for future employment.

Also, as I have mentioned earlier, graduate students are basically employees for the university and having to support your ‘indulgence’ of working does not really make much sense.

There is something to be said about the argument that graduate students are essentially going through an indentured servitude with the hope or expectation that it will pay off later on in their careers. I should mention here that while this may be true of masters/professional degrees, almost no PhD.’s elevate one’s earning potential above that of a masters degree, and instead take large amounts of time and effort that could be spent earning more elsewhere.

Most of the payment of a PhD is getting to work on an interesting problem, which is an extremely generous payout. The problem is that this is the reason for academics having lower salaries than their counterparts in the private sector. It should also be noted that while there seems to be a stigma that academics get to research any topic they choose, no matter how obscure or impractical, academics do not generally have complete control over what they work on: they need to get grants from someone who would like to see their proposed research carried out.

One of my inspirations for writing this post was the comments section in the Washington Post article. I realize that internet trolls are a common phenomenon, but the comments here seemed to be much more sincerely believed, and they echoed conversations I had had in real life. I hope that I have helped to clear the air.


The Budget Blueprint: A Physicist’s Perspective

This blog focuses on my take on current events as a young condensed matter physicist. Welcome!

Last Thursday, the Trump administration put out their budget blueprint. Before delving into the details, I feel compelled to remind the reader that this budget blueprint is not final. This budget proposal still needs to go to the budget committees in the House and Senate, which will come up with resolutions and submit them for a vote: the power of the purse still belongs to Congress. It’s also worth remembering that President Trump has a habit of treating most confrontations as negotiations: so perhaps we should think of this budget proposal as an opening bid and not a wish list.

Many have attacked the budget for its draconian cuts to programs like Meals on Wheels, FEMA, afterschool programs, NPR, Americorps, and PBS, and that certainly deserves focus, though I will not discuss them in this post as I will be focusing on the portion of the budget dedicated to science and engineering research. I will also refrain from discussing the  proposed 30% cut to the EPA, the cuts to NASA’s earth science research, and cuts within NOAA, since these appear to be the result of the increasingly toxic political climate, a topic too complex for me to summarize within a subsection of a single blog post.

What I want to discuss is the cuts to funding for basic science and engineering research.

At the bottom of page 19, the proposal justifies eliminating the Advanced Research Projects Agency-Energy (ARPA-E) by saying that “the private sector is better positioned to finance disruptive energy research and development and to commercialize innovative technologies”. On page 20, Mulvaney writes that despite a cut of $900 Million [(approximately 16%)] to the Office of Science, this budget “Ensures the Office of Science continues to invest in the highest priority basic science and energy research and development as well as operation and maintenance of existing scientific facilities for the community The budget also boasts savings of approximately $2 billion (approximately 42%) by cutting the Office of Energy Efficiency and Renewable Energy, the Office of Nuclear Energy, the Office of Electricity Delivery and Energy Reliability, and the Fossil Energy Research and Development program in addition to eliminating the combined the Weatherization Assistance Program (which has a total budget of $230 million), and the State Energy program (which has a total budget of $70,000). The proposal also cuts $5.8 billion (approximately 18%) from the National Institute of Health. The blueprint makes no mention of the National Science Foundation (NSF) , which may mean the foundation will see no cuts, though it is likely that though the 10% across-the-board cuts in non-defense spending may cut into it.

While there may well be administrative inefficiency in these agencies (as stated by office of management and budget director Mick Mulvaney pointed out in a press conference), these cuts are far larger than that. These cuts reflect the spirit of the following facebook post by Mulvaney:  “Do we really need government-funded research at all?”

I think that this is a fair question, especially if you haven’t worked in scientific research.  I will answer it in three parts:

  1. How is science useful?
  2. How do fields in science and engineering advance today?
  3. How does increased specialization affect science and engineering?

As is my custom, I will focus primarily on physics and pretend the other fields don’t exist. Let’s begin…

How is science useful?

Shortly after his discovery of electromagnetic induction in 1831, Michael Faraday was famously asked “what is the use of your new discovery”, to which he replied “what is the use of a newborn baby?”

In addition to inductors helping advance AC circuitry to the point that it became widespread in the 1880’s , Faraday’s “newborn baby” grew up to join the intellectual offspring of Ampère, Poisson, Gauss, and Lagrange in becoming a critical to James Clark Maxwell’s formulation of electromagnetism in the early 1860’s (Maxwell’s equations), which reshaped our understanding of phenomena like electrodynamics and electromagnetic radiation, and essentially created the field of electrical engineering.

This story is meant to highlight two facts:

  • The implications of science can be unexpected and far-reaching
  • These implications can take a long time to materialize

In other words, basic research can realize important applications, though what t hose applications will be or when they will materialize is highly uncertain. This is because, from a technological perspective, scientific discoveries are tools which inventors and engineers can use to create new devices. A scientist does not know what the results of an experiment will be before carrying it out, so it becomes very difficult to control exactly what kind of applications it will have.

How do fields in science and engineering advanced today?

When most people imagine scientific research, they think of what they see in movies. A scientist discovers some fundamental effect, and within a year has put it to use to create some invention which either alters the course of humanity, makes someone fantastically wealthy, or turns someone into a superhero.  Those who’ve taken a science class or two might be forgiven for thinking that it consists of coming up with groundbreaking equations like, well, Faraday!

If you go to the homepage of cutting edge scientific journal like nature, physical review letters, or applied physics letters and look at the flashiest new findings, you will find that all of these advances apply to what seems like very specific scientific subfields. The sorts of papers we learn about in introductory science classes would be hard to find today: as would someone with the surprisingly common 18th and 19th century job titles like “chemist, physicist, inventor, philosopher, and farmer”.

The loss of the scientific “renaissance man” is hardly surprising, given that research has become increasingly specialized as our scientific knowledge advanced, leading to a greater need for collaboration in research: especially when it comes to making a field more applied.

How does increased specialization affect science and engineering?

The increasing specialization of study in an advancing scientific field can be likened to the increasing greater specialization of labor in a developing economy.

Basic scientific research could be likened to the acquisition of raw materials, falling squarely into the primary sector of the economy (i.e. mining, fishing, or farming). One crucial difference between research and the primary sector is that resources are limited, resulting in zero-sum competition, whereas scientific knowledge and principles are clearly not: it’s not possible to “use up” science. Basic science research, unlike engineering research, cannot (generally) be patented, making it difficult to capitalize on basic research directly. For this reason, while basic research helps advance entire industries, it is very rare that an individual company finds it in its best interest to pursue basic research itself.

More basic applied science and engineering research is similarly equivalent to the part of the secondary sector devoted to processing raw materials like metal fabrication or the construction of copper pipes for more specific use. An example of this sort of research would be the use of advances in scientific knowledge to construct and study of devices or systems with extraordinary properties such as photonic devices or metamaterials with negative indices of refraction. While these systems promise manifold applications further down the road, technology or cost prevent us from being able to implement at this time.

A good example of a project in which a company would engage in this research is the endeavor by tech companies (including Intel, IBM, Google, and Microsoft) to build quantum computers. These companies research a wide variety of topics including: how to the construct qubits, what quantum circuitry would be effective, how to overcome problems of decoherence, and even quantum information theory.

This sort of R&D project can be thought of as a huge, long-term investment as well as (intellectual) vertical integration. Not surprisingly, more basic applied science and engineering research in the private sector has significant barriers to entry (most often, an oligopoly). Because the companies come out with patents relating to many discoveries and inventions made while working on this project. This means that industries are highly competitive, where companies have small or nonexistent economic profit, this sort of research would not be conducted without funding from the government or foundations. This is not to say that more basic research does not benefit these companies: they are able to devote R&D to more applied engineering research (i.e. they can make smaller, shorter-term investments). This is the premise of ARPA-E Funding!

This is not to say that an industry will remain this way forever: so let’s take a look at industries in which companies are able to fund more basic research themselves. So…

How does research conducted in the public sector differ from research conducted in the private sector?

Before proceeding, it is crucial to remember that research in the private sector is advanced by the profit motive, a formidable force for fostering innovation. When comparing research conducted in the public and private sectors, we should remember one thing. Whatever the shortcomings of research in the private sector, it has produced spectacular results industries ranging from semiconductors to pharmaceuticals.

The first difference between research conducted in the public and private sector is the goal. Research conducted by a private company has focused goals: it aims to have profitable applications in a specific field. Public research, on the other hand, is more general: it can aim to create a specific advice, but it can also aim to simply explore general properties of a certain class of devices or it may be entirely devoted to answer a question. Therefore, even in industries where companies routinely conduct fundamental research of their own, it is possible to publicly fund useful engineering research that would otherwise not be carried out! Another implication: when an experiment in the public sector turns out to have completely unexpected results, it can be more exciting than when everything goes right. This gives scientists and engineers working in the public sector more flexibility in following the results of an experiment instead of being forced to say “that result is interesting, but it’s not going to lead me anywhere I want to go”.

Public and private research also differs in the way in findings are shared. Unlike universities or national labs, where researchers are incentivized to publish, private companies are incentivized to make most of their findings confidential or proprietary. This means that the company will delay sharing their findings with the larger research community and when they do, they would do well to limit others’ ability to capitalize on their discoveries, limiting the way in which somewhat fundamental findings can be applied by others.

This is NOT to say, however, that the government should not reevaluate what research merits investment with public funds. It should and it does! Research grants are not easy to get from the government: in fact, much of the time the success or failure of a person’s career has little to do with their skill or even past productivity. I have heard countless stories of people who were good scientists, but entered the job market at the wrong time and found their careers at an end. In fact, a major reason many physicists leave physics for other technical fields like quantitative finance.

Many will say, however, that times are tough and the government should cut the budget so as to limit the ballooning debt which will drive up interest rates and crowd out investment. This is not a good time for long-term investments: we should restrict funding to only the most promising research.  Here’s the issue with that: evaluating the usefulness of funding for science and engineering is tricky because it is almost impossible to calculate an expected ROI for a future experiment.  Even calculating an ROI for PAST DISCOVERIES can be very difficult. What is the work of Faraday worth? He invented the inductor, but circuits also use the previous inventions of capacitors and resistors, in addition to later inventions like diodes and transistors: almost every discovery has been made by combining myriad previous discoveries and innovations with an researcher or inventor’s ingenuity. In short: while research is clearly beneficial to society, even the value of past research is extremely difficult to quantify. Also note that this hasn’t even touched upon the fact that the implications of research often won’t materialize for years or even decades!
So what? This is just an opening bid!

Finally, I get to the meat. The apparent cavalier attitude with which these programs were cut and the fact research only seemed to get funding if it appeared to promise immediate use like the National Nuclear Security Agency (NNSA) belies a short-sightedness that worries me. Either the blueprint was written without an understanding of how science and engineering research works (unlikely, though not inconceivable) or it was written with that knowledge and it is difficult to tell which is worse.

If the budget was written without knowledge of how research in science in engineering works, then Mulvaney didn’t try to familiarize himself with the topic (not difficult to do), and decided, instead, to make these cuts in ignorance. If this is true, then the budget director does not feel the need to properly educate himself before making important decisions. Scary but, as I said earlier, not likely.

If the budget was written with knowledge of how science and engineering works, then the government just decided t0 actively forego future growth for a slight reduction to the deficit in the present. The main exceptions to this rule are the increased money allocated towards the NNSA and NASA’s space exploration budget. The NNSA makes sense in the immediate future, but expanding interplanetary travel taking priority while cutting fundamental scientific and engineering research? There are three unsettling (comparatively likely) possibilities:

  1. Mulvaney does not believe something is beneficial or promising unless he can understand every step himself (“if it had value, I would understand it”)
  2. Mulvaney sees the importance of funding research (even those he doesn’t understand), but believes that cutting the budget is such an immediate concern that it is worth scaling back scientific research
  3. Mulvaney sees the importance of funding research (even those he doesn’t understand), but does not care because many of the effects will not be felt in the next 4-8 years.

I believe the third possibility to be the most likely. It’s also highly troubling because I have never been able to come up with a convincing answer to the question: what keeps people from borrowing from the future to benefit the present?