The Magic of Science

A group of billionaire stargazers say that the American answer to our great environmental crisis is $11bn more research.

As an afterthought, on page 8 – not as one of their main five recommendations – they add:

A strong market signal will increase the intensity of energy research, add large private-sector commitments, reduce barriers between the lab and market, and ensure technologies perform better and cost less over time. Those policies may include some combination of a price or cap on CO2, a clean energy or renewable energy portfolio requirement, and technology performance standards.

Really, that’s all there is to it. The necessary and sufficient condition for a transition to green energy is creating demand. (Note “green energy”, not overall emissions reduction; land-use changes like deforestation and cattle-raising are a distinct problem.)

Consider two alternate histories of aviation.

  • In track A the US government decided in 1910 that it would fund hundreds of millions of dollars in research to produce working prototypes of a combat aeroplane flying at 1,000 mph and a civilian airliner of 300 seats flying at 600mph. It did not offer meanwhile to buy planes with lower specifications, nor did it subsidise others to do so.
  • In track B the US government did not fund any research at all, but regularly issued calls for incrementally improved combat and transport aircraft and bought large numbers of the winners of competitions. It also gave tax breaks to operators of civilian aircraft, and subsidised airports, pilot training and air traffic control.

Track B is close to the actual history of rapid innovation. Does anybody believe that track A would have yielded anything much at all?

Very little long-range, fundamental, or blue-sky research will get done if the government doesn’t pay for it. The great days of Bell Labs are gone the way of the rentier science of Charles Darwin. At the other end of the innovation chain, firms are glad to develop a product if they believe the technological risk is low enough and the market prospects good enough, and governments are bad at replacing them.

There’s some point in between where the balance should shift from public to private-sector research. I don’t know where this is exactly. But is there any doubt on which side sustainable energy falls today?

There are a whole raft of well-known technologies out there in power generation (photovoltaics, solar concentrating, wind, geothermal), distribution (hydro storage, smart grid, hydrogen for fuel cells) and consumption (battery transport, light and heavy rail, smart energy management). Flue gas sequestration and tidal power are almost there. We are definitely at the point where the need is not for fundamental understanding but cost-effective practical engineering. You mobilise your technological resources, and expand them through economies of scale, networking, and experience, by creating effective demand. This is why and how Chinese firms are cracking the problem of polysilicon supply for photovoltaics (refuting old gloom from me, current prices have been falling for a year, and see this gripe by a speculator). Spain and India have weaker science bases but bigger wind turbine firms than Britain or France. Substituting public for private-sector funding for close-to-market development is irrelevant and likely even harmful.

In theory publicly-funded research could get prototypes to the point of eliminating both technological risk and excess costs. Can anyone point to cases where this sort of corporate welfare has ever actually midwifed a major innovation? Leaving out the bit where businessmen, engineers, and craftsmen fiddle with the design and the process to make them cheaper and more reliable (learning by doing, or economies of experience), talk to each other, don’t talk to each other, emulate and vie with each other (economies of networking, competition), and lower costs by mass production (economies of scale)? Count me a market fundamentalist on this one.

Let me anticipate some possible objections. Soviet and US rockets were developed entirely by public funding. John Harrison built the first marine chronometer over 40 years’ work in response to a very large prize of £20,000 created by the British Parliament in 1714 after the Navy had lost an entire fleet in 1707 to a navigation error. His final version, H4 of 1759, was a complete and reproducible solution. The episode is instructive: prizes can replace patents for long-shot problems. NASA is now trying them for launch rockets.

These innovations were all insensitive to price. The NASA of the moon programme and its Soviet counterpart had no hard budget constraint. Early marine chronometers were still worthwhile buying [update from comments: if only for the Royal Navy initially] at the cost of 30% of the whole ship; the cost came down later. The entire problem of sustainable energy is price.

Nuclear power: it doesn’t look as if private sector development will by itself make this energy competitive, so shouldn’t the government pay? Answer: governments – quite a few of them – have been throwing public money at fission power research for half a century, and it’s still not economic. I suspect the engineering is about as good now as it can be, and the problem lies with inherent features of risk and waste disposal. The public may be irrational in holding nuclear to a higher standard of safety than any other technology, but there it is. So plants are slow to design and get approval (I doubt whether the French could build their nuclear park today), and slow to build, so you can’t get economies of scale or rapid innovation cycles. It’s just a second-best technology, like steam or fuel cells in cars. We can either clone existing designs or give up, more research won’t help.

A more serious objection is that while we can get massive reductions in carbon emissions just refining existing technologies, it will be very hard to get all the way to carbon neutrality. Think of aluminium, steel, cement, aviation, shipping … So we still need long-range research on carbon sequestration, fusion, genetically modified trees and algae, solar cement kilns – and geoengineering if it still goes pear-shaped. I entirely agree. But I don’t think the Innovation Council were thinking of spending $16bn on scientists mad enough for me.

Comments

  1. DCA says

    A bit of pedantic historical correction, which does however reinforce your argument: Harrison's final chronometer, H4, was just barely reproducible, and only at a price that was in fact well beyond what the market would pay: the only copies were made for the Royal Navy. However, once it was shown that clockmaking was a solution to the longitude problem, commercial makers got involved (Arnold and Earnshaw) and invented the quite different mechanisms that became the standard for the next 150 years.

  2. Brett says

    Answer: governments – quite a few of them – have been throwing public money at fission power research for half a century, and it’s still not economic.

    In the US, maybe – but that's because of over-the-top regulations that drive the time and cost of setting up a nuclear plant through the roof. Once you get the plants actually set up, though, they tend to be profitable to run, which is why there are a number of firms pushing for expansion.

    I suspect the engineering is about as good now as it can be, and the problem lies with inherent features of risk and waste disposal.

    You'd be wrong, then. There have been a number of major innovations in plant design (Pebble Bed reactors come to mind). As for risk and waste disposal, that stuff is vastly over-stated (the waste can be mostly re-processed into fuel, and the risk is minimal as long as you don't do what the Chernobyl guys did, and deliberately run your reactor roughshod with the safety constraints turned off and no containment dome), and largely solved decades ago.

    The public may be irrational in holding nuclear to a higher standard of safety than any other technology, but there it is.

    It's a terrible waste, and an unnecessary one (France gets most of its power from nuclear, so it's not like this is some universal thing). Nuclear Power is perfect for a situation in which you need to keep things as compact as possible and use rail instead of oil-driven transport.

    It’s just a second-best technology, like steam or fuel cells in cars. We can either clone existing designs or give up, more research won’t help.

    No, there's plenty of room left for innovation. There's an entire field of possible reactors (Thorium reactors) that we've only somewhat tapped into, for example.

  3. says

    DCA: thanks.

    Brett: France is pretty much the outlier, not the US. Britain, Germany, and Japan have fewer and unpopular reactors. On your last paragraph: I'm sure the nuclear lobby is quite strong enough to see that lots of taxpayers' money is still spent on these exciting possibilities. Isn't it funny how the private sector is curiously reluctant to take them on? Who knows, we might buy ourselves a demonstration thorium reactor by 2020!

  4. Brett Bellmore says

    The private sector is reluctant, because you never know when the public sector will decide, halfway through your expensive investment, to change the rules.

  5. yoyo says

    lol. pretty much only countries with very good governtments (ie government is a high-status job) can do nuclear.

    i love how engineers don't think things can ever go wrong.

  6. Brett says

    Isn’t it funny how the private sector is curiously reluctant to take them on? Who knows, we might buy ourselves a demonstration thorium reactor by 2020!

    Why bother when there's plenty of room for experimentation with uranium reactors, which they are very keen on presently?

    I love the snarky tone, though. As if solar panels and wind – still in the single digits of US power generation – are paragons by comparison in terms of the market (*cough* renewable tax credit!), or somehow immune to the same type of idiotic NIMBYism that has harmed nuclear power in the US.

    Or better in terms of environmental friendliness – have fun driving your gas-guzzling trucks all over the plains and desert to operate and repair wind farms, while nuke plants can be built and supplied by rail. Or your dependency on imported lithium, or trying to make a smart grid where your power sources are out in the middle of nowhere work without cheap super-conductors (major power losses over long transmission + intermittent power source = bummer, man).

    The private sector is reluctant, because you never know when the public sector will decide, halfway through your expensive investment, to change the rules.

    For once, I actually agree with you, other Brett.