Drug Baron

Please DO cut the science budget!

| 11 Comments

As President Obama signs the sequester cuts into the budget, amid dire predictions of global economic melt-down and howls of protest at the impending reduction of the science budget, there is at least one UK scientist who wishes the same were happening here.

DrugBaron would cut the science budget in the UK overnight by 10% – along with all other government spending

The US sequester, then, is the perfect policy.  Maybe its only fault is that it doesn’t go far enough.  It’s a once-only measure.  Spending has become so bloated over several decades that an annual reduction across the board would be justified for quite a number of years.

Not surprisingly, spending cuts get a bad press.  There is a mismatch in the coverage given to those who squeal because they lose out than to those who benefit from the reduced cost burden.  This mismatch arises due to gearing: cuts affect a smaller number of people to a bigger extent than the savings spread over the whole population.

Those who lose out then marshall all kinds of “but its bad for all of us” arguments to support their case.  In the case of the science budget, the superficially strong-sounding argument is usually one of growth – knowledge is the foundation of economic prosperity, so how can it make sense to cut the science budget?

Simply because the link between input cost and output value is not only imperfect, but completely illusory.  It is one of the most pervasive, and most dangerous, fallacies of the 21st Century world: that input equals output – that spending is a measure of production.

“There is no selective pressure in abundance” Kevin Johnson, Index Ventures

Trouble is, its simply not true – in science or anywhere else.  Recessions and cuts are not just beneficial, they are essential.  Only in the face of cuts are weaklings weeded out, and efficiency maintained.  The general public, investors and governments are addicted to an unachievable, and undesirable, fantasy of steady, monotonic economic growth.  Its time to learn to love cuts – like DrugBaron.

The argument for spending more on science goes something along these lines: economic growth stems from efficiency; efficiency depends on knowledge; science generates knowledge so spending more on science will automatically yield more knowledge, efficiency and hence economic growth.

The first two links in this chain of argument are solid enough.  Real economic growth (as opposed to what we measure using unhelpful tools like gross domestic product (GDP), which muddies the waters by double-counting a lot of activity) is entirely dependent on efficiency gains – ultimately in primary productivity (meaning food, energy and other essentials).  Its also easy to believe, if quite difficult to prove, that efficiency gains require knowledge.

It’s the last link that is highly suspect: spending more on science yielding more useful knowledge depends absolutely on availability of money being the limiting factor for science productivity.

In the complete absence of public-funded science this may well be the case.  Our brightest people would be unable to engage in knowledge-generation at all, instead being forced to earn a living closer to primary production (as they were in the middle ages).  More money to support them generating knowledge would translate efficiently into economic growth.  In this world (which is nothing like the world we live in today) money is the rate limiting step in a chain of chemical reactions.

But once you have a very substantial public (and indeed private) funded science community, the balance begins to shift.  Finding another really smart person to support becomes the challenge.

Assuming there is a degree of rationality in deciding who to support (that science funding isn’t handed out purely by lottery – which we fervently hope is not true, although at least in the UK the allocation system is deeply flawed if not yet a complete lottery), then the next person to be funded by any increase in the funding pool will be just a bit less talented than their predecessors.  Eventually, you are funding people who have little hope of contributing any knowledge useful for productivity gains at all.

Western countries passed this point decades ago

The reality is that academic science (and probably most global private R&D organizations like pharmaceutical companies) are stuffed to the gunnels with also-rans (in addition to a much smaller core of really talented individuals).

The fundamental problem with science is that those on the outside (the public who are funding it) understand – for the vast majority – less about science than those on the inside (the scientists).  They have no way to judge the quality of the people they employ, nor the usefulness of the knowledge they generate.  Its frankly impossible for the keeper of the purse-strings to know when the line is being crossed between a lack of money and a lack of smart enough people to support.

And faced with howls of protest from those on the inside whenever a cut in funding is mooted (driven entirely by self interest) and a vague belief among the electorate as a whole that more funding leads to more knowledge, more knowledge to greater efficiency and that in turn to economic growth, it takes a brave politician to wield the axe.

DrugBaron sees it differently.  Cutting fully half the science budget would have no discernable effect on the productivity of UK plc over the long term.  Of course, it would damage the ludicrous measure of GDP in the short term as thousands of scientists were thrown out of work, cutting their spending power – but since the C-grade scientists were producing little of value anyway, this short-term loss would be entirely illusory.  In the longer term, the productivity of the science sector would be massively boosted, since almost the same output would be generated at half the cost.

This contentious claim needs (at least) two caveats.  Firstly, DrugBaron’s measure of output differs from those typically used as metrics to assess the production of science – published papers and perhaps patents filed.  Neither of these are a measure of either quality or usefulness, being a measure more of heat than light.  In the same way that the number of scientists employed or the total spend on science are proxy measures of effort and aspiration rather than of utility, so too are publication metrics.  Clearly, then, harsh cuts to the science budget would lead to declines in these “effort-based” metrics even if the real loss of utility was negligible.

Secondly, and more trickily, it assumes that the axe fell on the C-graders

That is harder to engineer than it might be imagined – and ironically is harder to bring about with a gentle squeeze on spending than big cuts.  Human nature tends to spread the pain, and faced with a small cut in spending the natural reaction is to reduce the amount that everyone is getting.  That is highly inefficient, since the small fraction of the total pool who are doing the useful stuff are hampered as much as those who are spinning their wheels.

Only once the cuts become large enough that they are impossible to meet through tinkering at the edges, when radical action, such as losing whole departments, becomes necessary do the managers make a real effort to distinguish the winners from the losers.

This, then, is the wonderful healing power of recession – not just in science but across the economy as a whole.  Recession and cuts weed out the less productive elements in any industry – elements that survived perfectly fine during the growth phase.  Without this periodic cleansing, living in the modern fantasy of monotonically increasing growth, efficiency declines and declines, until the correction comes.  And the longer its held off, the more painful it eventually must be.

Kevin Johnson, Partner at Index Ventures, uses a neat gardening analogy: “Apply enough fertilizer and water to your garden, and everything will grow – weeds and all.  There is no selectivity in abundance.  Only when the water and nutrients are limiting do the strongest plants survive (even if only just) while the weaklings wilt and die.”

DrugBaron has spent long enough in academic science, sitting on grant awarding bodies for the government and medical charities, to know for a fact that there are a lot of C-grade scientists in the UK (just as there are, of course, very many excellent ones).  A period of drought may be the only way to weed them out with anything approaching precision.

Intellectually robust as this analysis may be, there is little possibility of it coming about

Well-rehearsed squeals of protest from the scientists themselves (the good ones, fearing the axe will be applied ‘equally’ as so often it has been in the past, as well as the C-graders that make up the majority) will be joined by the usual commentators who mistake quantity for quality.

This same “coalition of self-interest and ignorance” often seems to dominate the argument on cuts in the wider economy too.  Recession is not just good for science, its good for all industries – the permanent loss of capacity occurs predominantly among the least efficient, and as growth returns it is replaced by stronger, better players.

Of course, for those who find themselves in that ‘relegation zone’ vulnerable to the down-turn, it is hardly surprising they argue for continued spending to prop them up.  But we should not heed them – or else there is never any pressure to reform.

If that’s the “self interest” component, the “ignorance” comes mainly in the form of apostles to Keynsian economics – a cultured form of financial hari-kari that has remained popular for more than half a century.  You know the patter: “borrow more to invest in growth for the future”.  DrugBaron christened such debt as “chocolate-flavoured poison” because it tastes good, but it kills you in the end.

Keynsian stimulus works just fine, but only with two provisos – neither of which apply now (or have applied for decades).  The two provisos are that the extra spending has to really stimulate long-term growth (and not just provide an illusory jump in GDP, that dangerous metric, that results from the actual expenditure of the stimulus no matter on what); and that you don’t already have a large amount of debt in the first place.

The example of investing in science is the perfect example of why the first proviso isn’t met: lack of money is no longer the limiting factor for the production of useful knowledge.  The limiting factor is the availability of super-smart people.  So invest in education, plead the apostles of Keynes – indeed we should, but remember that there too its no certainty that more money results in better performance.  At some point, the inherent intelligence of the population becomes limiting on educational attainment rather than the quantity (or even quality) of the education available.

The second proviso is more subtle.  The problem here is that the more debt you are taking just to service “unproductive” future liabilities (such as paying pensions), the less effective will be “further borrowing” to invest in future growth.  There are no two separate pockets for the growth stimulus and the general expenditure – there is only one economy.

The only solution is to cut discretionary spending, and cut it hard.  The current policy of stagnation falls between both stools.  Whether the economy stagnates for a decade or declines sharply for a couple of years, the impact on living standards is the same – but the gentle squeeze, however long its applied, doesn’t have the cathartic effects of deep cuts.  The weeds are struggling, but they are hanging on in there!

Its time for every taxpayer to learn to love cuts – not support the “coalition of self-interest and ignorance” that deplores them

If deep, blanket cuts are actually the right medicine for a strong economy (and a strong and efficient science base as well), it is a shame that such a policy could only be enacted by default, through paralysis and bickering, in the form of the US sequester.  For those of us in the UK, its an even bigger shame that the right medicine is only going to be administered to the US economy.

Its time for the taxpayer majority in the UK to stand up to the “other coalition” (not the current governing coalition between the Conservatives and the Liberal Democrats, but the coalition of self-interest and ignorance) and demand the deep and painful cuts that are needed to see the UK economy back on its feet and fighting on the global stage.

If a life-long scientist can call for deep cuts in the science budget, maybe the rest of the turkeys can vote for Christmas?  Those that survive the cull will be plumper for it.

  • http://twitter.com/AWTaylor83 Alasdair Taylor

    Dear DrugBaron, you state that your analysis is “intellectually robust”. I apologise as I may have missed a link from this page to the economic evidence from which you base your argument. If you could provide said evidence then readers will be able to discern if its really true that:

    -”there are a lot of C-grade scientists in the UK”
    -”cutting fully half the science budget would have no discernable effect on the productivity of UK plc over the long term”
    -that cuts, “a period of drought”, are the best way to make the system more efficient and productive

    -that we’ve actually reached the point of abundance

    Finally, what is your opinion on the relationship between the private R&D investment and growth and productivity?

    Thank you

    • davidgrainger

      Thanks for your comment, Alasdair.

      To be clear, “intellectually robust” is different from “evidence-based”. I didnt say there was economic evidence for (or against) my analysis. Its remarkably hard to garner some of the evidence you request, which means policy (whether that currently followed, or that recommended by DrugBaron) has to be based on rational extrapolation from what is observed or otherwise known. That is what “intellectually robust” was intended to convey.

      Taking your points in turn:

      There is no objective scoring system for the quality of scientists. But having spend two decades reviewing papers for major journals and grants for major grant-giving bodies, I have accrued a lot of evidence that points to a large rump of low quality scientists. The best evidence is the number of internally self-inconsistent arguments and experimental datasets that compose those papers/grants.

      I do NOT claim that the UK is worse than elsewhere – it may well be better. But it is where I live, pay my taxes, and where my personal evidence base predominantly lies. I am not sure what better evidence could be accumulated to independently score scientist performance.

      In any case, wouldnt it be a surprise if every scientist were fantastic? Thats not true in any other industry – my arguments hit with full force just on the normal presumption that some people doing a particular job are a lot better than others. That always means that the N+1 recruit is a little worse than the Nth – no matter how good or bad the “average” is.

      The provocative statement that cutting fully half the science budget would have no discernible effect on growth over the long term has to remain a conjecture – no such experiment has ever been done anywhere, as far as I am aware. I based the statement on my experience of examining papers and grant proposals that were worse than valueless – losing them from the ecosystem can only be a good thing.

      It is also a caveat, clearly stated, that this would only be true if a mechanism could be found to cut the unproductive “C-graders” without damaging the performance of the ultra-useful core. That is by no means easy to achieve in practice.

      Hence the concept that a period of drought is the best way to force the system to self-select. There is no evidence that this is “the best way”. But it is an “intellectually robust” conjecture based on the impact of steep recessions on industrial performance during the industrial revolution. The last half of the 19th century was punctuated by years of extreme growth followed by extreme recession, a pattern that drove the increases in productivity we continue to benefit from today (see the excellent biography of Andrew Carnegie by Peter Krass to understand the extent and impact of these growth/contraction cycles in the 1880s and 1890s).

      Ultimately, we will never know for certain that we have reached a point of abundance. In reality, spending more will always produce a “bit” more output. So its not a “threshold effect”, burt rather a declining return for expenditure. At some point, the extra value of the output was no longer worth the linear increase in cost. The proposition explored in this article was that we have long since passed the cut-off point where cost exceeds value.

      Ultimately, its not possible to definitively support the contention that science spending should be protected (which is the current Government policy, the policy supported by most scientists, and probably the policy most taxpayers believe, rightly or wrongly, is in their best interests) any more than it can be PROVED that cuts would be harmless. The purpose of this article was to challenge the “coalition of self-interest and ignorance” that AUTOMATICALLY assumes that cuts are bad – and vociferously argues against them. I hope I have redressed the balance and at least made people question whether sustaining current spending levels (not just on science but more broadly) is the right thing to do.

      For your last question, Alasdair, I dont really differentiate between public and private expenditure. Cutting the size of pharma R&D would increase productivity by the same arguments I laid out for cutting the size of the science base more generally (although the economic stringency is stronger on private money, so there is probably less waste to root out than in the public-funded sector)

      • http://twitter.com/AWTaylor83 Alasdair Taylor

        Thank you David, I welcome the development of your argument.

        Science has made a case for greater funding based upon its economic benefits whilst alongside asking that scientific methods are used to obtain evidence for policy decisions. In my opinion, a consequence that will arise from this is that those holding the purse strings will ask the recipients of public money to demonstrate the economic and social benefits generated. Whether you agree that the government should interfere in such a way, and many in publicly-funded research will not (and for some valid reasons), it is already happening through the current REF process and pathways to impact statements, and will become more prevalent in the future. Hopefully, better methods to study, assess and report the impacts of research will be developed and used so we can obtain the evidence needed to spend public money more appropriately.

        I doubt that most of those asking for more research funding want it to be spent on poor science, although sometimes it is difficult to avoid it. Many scientists would also welcome better and more efficient methods of assigning public funds for research. What concerns me is that if cuts are made, would they be done in a productive manner? Would it actually weed-out C-grade scientists or would we see cutbacks to good scientists across the board or, perhaps, in specific areas not seen as desirable (rightly or wrongly)? Would we also discourage good scientists from pursuing a research career due to difficulties of finding money and positions?

        As you state there is a paucity of evidence currently about the benefits of publicly-funded research. What we have are anecdotes, single case studies and personal experience. Based upon my own, I will say that I have to disagree with your assertions in this blog, even if I do appreciate your reasoning. At present, I support the levels of science funding and would welcome an increase (and will continue to do so until I see evidence that strongly suggests otherwise). Of course, this doesn’t mean that research funding is simply a blank cheque handed out by government. Thank you again.

        • davidgrainger

          Despite the deliberately controversial rhetoric, our positions are actually not very far apart.

          I said that IN PRINCIPLE a cut of 50% would do little harm IF it could be targeted to remove the bottom 50% of our science base ranked by quality. Like you, I recognise that this is not possible in practice. As you say, no-one involved wants to see poor research, and we are all attempting to do the best we can to distribute the money to those who will use it most effectively. That we fail is not because there is a better way to assess quality than that which is currently in use (although I think there are examples where ‘we’ could do better than we do – such as the EU Framework programmes which are, for the most part, an egregious waste of money – in fact, anywhere where factors other than quality start to impact decision making, we see poor decisions: so the need to spread money ‘fairly’ across the EU prevents distributing it to where it would be most effective. Programmes to direct research in particular geographies or particular fields of study always suffer from this inefficiency).

          Ironically, and despite the attention-seeking headline, I do not favour cutting the science budget until we have enacted much bigger cuts on things where public money is much more obviously wasted. Public sector pay and pensions would be the most obvious target. And my article was intended to use the arguments about science (where I do have considerable experience and knowledge) to make a much wider point about the hysteria that surrounds any kind of cut to any kind of service (which is driven by the “coalition of self-interest and ignorance” I talked about). The point of focusing on science was to say (as the last sentence says, in a somewhat tongue in cheek way) that if a scientist can see the possible benefits of cuts to the science budget, maybe we should all have a more open mind about cuts that ostensible affect us.

          If I manage to make that point, I will be delighted. And I really appreciate your comments for helping to bring out the subtlety of the argument.

  • Martin Eglitis

    I agree in principle with you premise. But one problem with the existing system is the ability of the ‘haves’ to guard their resources. While too much money funds too much mediocrity, there is also a bias for those effective at linking to funding (often by proposing research that doesn’t disturb the apple cart) to get most funding. There is a trickle-down effect that if one is professionally linked to a ‘rainmaker,’ one gets easier access to funds. A consequence is that ‘out of the box’ innovation is harder to fund. Cutting funds must be done in a way to enable *more* innovative thinking to be funded. In essence, that would counter the argument that less innovation comes from less funding. Not if more novel thinking is acceptable with those decrease resources.

  • David Fox

    DrugBaron is not wrong. It may seem that indeed turkeys are voting for Christmas, but the current situation is not reasonable. It seems to me that nowadays much of academic research is more about career building than actually getting useful stuff done. The financial pressure that cuts produce can be very useful in refocussing the efforts of scientists. I am not particularly in favour of sacking them, just their “going nowhere” projects.

  • GCM

    Of course we should seek to drive efficiency in all levels of spending, and cuts/tough times certainly create an imperative to do so. But you fall into the fallacy of composition when you assume that what is valid applied to a specific area remains valid when applied to the economy in aggregate. Here I recommend you stick to the physical science and leave the macroeconomics to the economists. You don’t cut your way out of deep recessions where monetary policy is at the zero-bound, which it is right now. When all sectors (and countries) are seeking to deleverage and real interest rates are already at zero, fiscal policy must take up the slack if deflationary spirals are to be avoided. We learnt this way back in the 1930s. We should not need to repeat the same mistakes when we have 80 years more experience and knowledge in this area.

    • davidgrainger

      Of course, the main thrust of the piece was about science funding, but I defend the more general macroeconomic point as well.

      For sure, it is not the mainstream view (which your comment very well articulates) – any more than the proposed cut to science spending is mainstream within my own area. What I am presenting is a challenge to the status quo, being the position that you set out. Leaving the economics to the economists has, by any independent assessment, left us in an economic mess!

      To argue that we should “learn the lessons from 80 years ago” makes the fatal assumption that nothing has changed about the system. But two very important things HAVE changed – the accumulated debt burden is much higher now than it was then (making it harder to justify yet more borrowing to stimulate growth) and the infrastructure we have is much better than it was then, so further efficiencies are harder to gain through whatever investment is made. So rather than me being “naive” for ignoring history’s precedent, I contend that the mainstream economists are being complacent by assuming that what applied then, applies now.

      Even given that caveat about using “ancient” history as a guide, I agree with the principle that “you dont cut your way out of a recession” and that deflationary spirals are very real risks. But that begs the question “when do you cut?”. The mistake is not investing now, but failing to cut during the positive phase of the cycle. By kidding himself that he had “ridded the economy of boom and bust”, Gordon Brown set his spending based on economic productivity of the positive phase of the cycle (effectively leaving nothing left in the good years to pay down the debt that accumulates during the bad years). THAT was the critical error that leaves the current government with no room to follow the policy you recommend – the ill-fated Keynsian stimulus.

      Had Gordon Brown shown less hubris, and recognised that the long period of growth of the early naughties was just cyclical, he would have used the extra income in that period to pay down the debt, keeping a tight lid on expenditure (and certainly not ‘investing’ in infrastructure at that point). Then, as you suggest, we would be able to make that investment now, using it to stimulate growth.

      BUT that isnt what happened. We have to decide on our current economic policy based on the real situation we find ourselves in now, not on economic theory or on historical precedent. And that situation is one where the stimulus has been applied on full throttle throughout the economic cycle until the tank is left on empty just at the critical moment when you need the engine.

      Yes, cutting across the economy could lead to a deflationary spiral. Yes, it may take 10 years to get out of that. But that would leave the economy more nearly reflecting our REAL efficiencies in primary production, rather than riding on a super-ionflated plane built solely on debt. It will take a long and painful time for the correction to occur but we are better to take that pain that to continue to kid ourselves that “the only option” is more debt.

      I really appreciate your comment for giving me the opportunity to expand on the general macoreconomic argument, rather than ‘just’ the narrower topic of investment in science.

  • fromLondon

    Thanks for an interesting read. However, I am not
    convinced that too much money is really the reason that we have “C-grade”
    researchers. Rather, I believe that universities often save money by accepting
    very mediocre people into PhD programs. After all, a substantial amount of
    research does not require one to be “super-smart” but consists simply
    of manual labour. And while some of this could be done by technicians and
    research assistants it has turned out to be cheapest to get it done by
    students. In fact, working hard for next to no money is an accepted part
    of every science career. Hence, I would argue that the university research system
    today resembles something
    like a pyramid scheme: Ridiculously high numbers of students are lured into doing
    unpaid summer projects, then PhDs, then first Post-Docs with terrible salaries.
    Smart or not these individuals effectively run university research,
    investing huge amounts of time while earning small (or literally no) salaries -
    universities have come to rely on raising “C-grade scientists” exactly
    because they provide cheap labour.

    My worry is that as long as this system continues all that cuts will do is to make all PhDs a bit poorer, replace skilled technicians and research assistants with cheaper students (who, in contrast to the former, often want to become “C-grade” PIs later on) and potentially drive a few really talented people away to finance, etc.
    Maybe funding bodies (including the state) should rather think about the
    distribution of their money and offer fewer, better-paid PhDs thus
    raising the competition both between aspiring scientists and with
    non-research industry.

  • Integrity Lost?

    Your piece focuses on the cost, but ignores the two elements of time and quality. Assuming time as a constant, why would we not focus on quality and cut the estimated 50% or more of researchers who sustain careers on inappropriate or fraudulent research instead of cutting the C-graders? The damage done by misleading research is log orders higher than those who honestly experience multiple failures with the best of intentions.