Sam Roggeveen at the Lowy Interpreter links to yet another economic growth is coming to a end post.
As well as an argument that median income growth will be even slower than economic growth (has it occurred to the author that the distribution of income is a societal choice?) the author of the original NBER paper, and the FT Alphaville blog post, seem to think that our lack of supersonic passenger planes and flying cars somehow illustrates that progress has dramatically slowed.
I should note that their argument is considerably more complicated than that (and there’s a good point that the rise of Asia may have slowed the development of industrial automation), but I’d just like to address a couple of points.
The first is the whole notion that “progress” has slowed to a crawl. While it’s probably true that consumer goods that manipulate the physical world haven’t changed as much in the last few decades than they did in the decades immediately preceding, that is an almost completely useless measure of progress.
Consider this: we may not be traveling any faster than we were back in the 1960s (if you were able to afford air travel back then, which 99% of the population couldn’t), but we have ten more years of life in which to travel.
And it’s not like our life expectancy back then was that bad back in the 1960s either – it was over 70 years:
But that’s not really what this post is about. Flying cars (as well as vacations on the Moon) are the favourite punching bag for so many writers with their bee in a bonnet about something or other. But very, very few of these writers have more than the most superficial understanding of why we’re not zipping to work in them, despite the grand predictions of science fiction authors of the 1950s.
Let’s first note that we’ve had a pretty close approximation to a flying car for many, many years. It’s called the helicopter, and you can have a brand-new four-seat Robinson R44 for the knockdown price of 434,000 American dollars.
So why so expensive? Why can you buy 30-odd Hyundais for the price of one low-end helicopter?
It’s not that the power plant is inherently complicated. While larger helicopters use gas turbine engines (it’s actually debatable that they are more complicated, but we’ll leave that for now) the R44 uses a bog-standard Lycoming petrol engine. A big one, mind you – it displaces nearly nine litres, bigger than even a Dodge Viper – but not particularly sophisticated. Its design dates from 1957, with a primitive fuel-injection system that dates from the 1970s. Heck, aviation petrol engines still run on leaded fuel!
A bigger issue is the cost of the materials of the body and rotors of a helicopter. I’m not sure exactly what the R44 is made out of, but aircraft aren’t made out of the cheap, easy to work with, but heavy steel that is used in automobiles. It’s some combination of aluminium and carbon fibre composites, perhaps with a bit of titanium thrown in. But even that doesn’t convince me – the McLaren MP4-12C road car is built around a carbon fibre “tub” and is yours for $229,000 (in the USA).
Electronics? Pricey as hell, but not because of any inherent complexity to them. A GPS unit/radio/moving map with ILS (another radio navigation system used by aircraft) is a $26,000 upgrade. In other words, a GPS navigation system with several different kinds of radio and a dinky little LCD screen. I’ll take an iPhone with more functionality for a fiftieth of the cost, thanks!
So, we come back to the question – why is a helicopter so much more damned expensive than the sum of its parts in more familiar context would suggest?
The R44 has been in production for almost 20 years, but as of last year only 5,324 of them have left the factory. Roughly one per working day, in other words. Only once in human history – the latter days of the Second World War – were aircraft truly mass-produced; Henry Ford himself was able to churn out 650 B-24 Liberators per month from one gargantuan aircraft factory, and Russian industry churned out even greater numbers of the Il-2 Sturmovik. For the rest of the history of aviation, aircraft have been built in the same way cars were built before Henry Ford came along – small, highly skilled workforces hand-assembling in (relatively) simple, low-cost factories. Eclipse Aviation, a startup founded by a former Microsoft executive, had a crack at mass-producing a low-cost four-place corporate jet. Unfortunately, their engine of choice failed to deliver on its promises, and by the time they got that sorted out they ran slap-bang into the GFC.
The other problem, of course, is the R&D costs. Like cars, aircraft have to meet fairly strict regulations to become “certified” production models – and any change requires re-certification. This process is very costly, which is incidentally one of the main reasons why most of the innovation in small fixed-wing aircraft happens in the “homebuilt” and ultralight sector which bypasses much of this regulation. But, in any case, the costs of certifying the R44 and all the associated equipment have to be distributed across hundreds, rather than hundreds of thousands, of units.
The short version? While I don’t have hard numbers, it’s my strong suspicion that if you put in an order for a lazy half a million R44s (keep in mind that there are 60 million cars produced every year), the unit cost would be well under $100,000, and probably closer to $50,000 than $100,000.
So why aren’t dentists and solicitors commuting to work in their own personal R44?
In a nutshell, because they’re too dangerous and difficult for Joe Average to fly, even with the comparatively low levels of traffic we see at the moment.
It’s very difficult to get a pilot’s licence; harder still to get one that allows you to fly over built-up areas, and harder again to get an instrument rating that allows you to fly in bad weather (to simplify a bit). It takes thousands of dollars and many hours of practice. And that’s for relatively simple (and relatively safe) fixed-wing aircraft. Getting a helicopter pilot’s licence is far harder.
Part of that is that it is, by all reports, genuinely difficult to fly a helicopter – much harder than a car, or fixed-wing aircraft. Nothing you do in a helicopter is straightforward; they require constant correction just to maintain a constant altitude and direction. Any change isn’t just a matter of pushing one control; everything affects everything else.
But it’s also, at least implicitly, a deliberate strategy on the part of government authorities to discourage all but the most committed from piloting their own helo.
Despite the training, the certification of the aircraft, and the maintenance regimes you’re required to follow to keep them in the air, you’re roughly 25 times more likely to die in a privately-piloted helicopter than if you did the same journey in that well-known death trap, the car. They are as dangerous to their occupants as motorcycles. 1,500 Australians a year die in car crashes. Could you imagine an Australia where nearly 40,000 Australians were dying every year in chopper crashes?
And, of course, if they were as ubiquitous in our urban skies as Hyundais are on our roads, a fair number of those crashes would result in R44s plummeting into people’s houses.
Building a safer helicopter is possible; in large part by taking the human pilot out of the equation to stop them crashing in to overhead wiring, flying the wrong side of the dead man’s curve, and so on.
We’re now reaching the point where building software to pilot a helicopter safely, from takeoff to landing, is feasible. But actually building it to the demonstrated reliability level required would be very costly. Then you’d have the people most interested in the development – existing helicopter pilots – fighting the development all the way, because it would potentially make their hard-won and lucrative skills redundant.
At some point, our increasing incomes, continued improvements in technology, and increasing social acceptance of automated vehicles will make the self-piloting personal helicopter inevitable, and somebody will make the very large investments required to overcome some of the other problems associated with them (pollution and noise). But that point is still probably some time away.
So why didn’t fifties futurists anticipate any of these issues? Look at the social context. The environmental and health consquences of technology were barely even beginning to be considered – public health in developed countries was still struggling with things like polio. For those few sci-fi authors who appreciated that personal flying machines would require automatic control, the difficulty of building reliable software wasn’t seriously appreciated until the late 1960s.