We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people.

Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house]

A short video about the Alpha Graphs

A few months ago I gave a talk to Libertarian Home, of the sort that happen regularly at the Rose and Crown in Southwark. (They have a speakerless social at the same venue which I intend to be at, tomorrow.) My talk was … well, to put it kindly, it was somewhat less than the sum of its parts. It had its moments, but it didn’t add up. Worse, the more I struggled to pull it together, the longer it went on and the more incoherent it got.

But something good may yet have emerged from this muddle, because Libertarian Home’s Simon Gibbs and I recently agreed that it might make sense to rescue (i.e. for Simon Gibbs to rescue) one of the somewhat better bits of this talk and make it into a video short. Simon has now done this, with added graphics.

The subject is something I have already blogged about here, namely the tendency of statist measures to start out quite good, only later going wrong and then ever more wrong, and on the other hand the tendency of a truly free market, when a particular bit of it starts, to be a mess, and only somewhat later to start getting seriously good and in the long run superb. Two intersecting graphs, in other words, one going up and then down and down, and the other going down and then up and up.

My first label for this phenomenon involved hockey sticks, but when it comes to graphs the hockey stick is well and truly taken, and now I’m calling my graphs “alpha” graphs, because that’s how they look when put together.

Alas, even this bit of my talk could have been a whole lot more eloquent. For starters, I should have waved my arms around in a way that fitted how the graphs would look to the audience. As it was, I got them the wrong way around, sideways I mean, and hence somewhat clashing with what Simon does with them in his superimposed graphics. Nevertheless, the basic idea survives, I think, and is usefully provocative of further thought, as Simon demonstrates with his own further thoughts.

My own main further thought about the Alpha Graphs (here’s hoping those capitals catch on) is that the Adam Smith Institute should be mentioned in connection with them. One of the ASI’s basic tactical insights from way back is that there are indeed often many advantages to be gained and gamed by politically well-connected individuals or organisations or companies, from statist policies rather than free market policies, but that with a bit of cunning these tendencies can be countered, for instance by making the arrival of a competitive market very much to the advantage of a few big early participants, or with right-to-buy, right-to-sell arrangements with regard to such things as public housing that goes back into the market. It’s a matter of how you sell the new market, and to whom. Instead of just using Public Choice Theory (the Alpha Graphs being a tiny part of all that) to excuse libertarian policy failure; use it to point you in better (because more politically effective) policy directions.

That isn’t the complete answer to the problems described by the Alpha Graphs, but it is certainly a part of it.

The other thing I want to repeat in this posting is that I think that short videos are an excellent way to go, when it comes to spreading libertarian ideas, provided only that you know how to produce them adequately. (The technique has recently been used with great effectiveness by the Adam Smith Institute’s own Madsen Pirie to explicate basic economics.) I hope Simon Gibbs produces many more such video quickies in the next few years, and helps and encourages others to do the same, both in the form of excerpts from other bigger performances (by no means only from performances that he himself has recorded), and in the form of original creations of his own. Such a program could be a great developer of future libertarian star performers, as well as a chance for older libertarians like me to add their pennyworths.

16 comments to A short video about the Alpha Graphs

  • It sounds to me that what you are trying to describe is the whole “top down” vs “bottom up” design thing. Something that software designers and the like are intimately familiar with.

    Top down is generally faster to market but almost always has a scalability problem because some critical detail got overlooked. Said detail turns out to impact performance once you get to be successful and generally speaking at this point fixes tend to be bandaid upon sticking plaster upon ducttape and you end up with a horrible mess that still doesn’t scale properly is hopelessly insecure etc.

    Bottom up tends to take longer to develop but tends to be inherently scalable because you start with the leaves, then you put them together inti twigs, then the twigs combine to branches etc. etc.

    One of the challenges of rapid software development is to take the top down developed hacky prototype / demo and work bottom up to produce the real product

  • A win-win outcome Brian. I get a nice bit of content (and more linkage) and you get your talk rescued in the form of an embeddable YouTube video.

    The arm waving does clash a little (worse if I had left the graph over the top of the actual arm waving), but it’s nice and clear and worth it I think, if only so that I get to practice how to put graphics over a bit of video.

    More substantially: how come, in your shampoo example, the net effect is negative despite the obvious advantages of cleaner and prettier hair. It is as if you are saying that momentarily confused ladies are a bigger deal than ladies having the clean pretty hair they deserve, which I don’t believe to be the case. Is it more the case with a different example? Which example?

  • Laird

    An interesting idea, worth developing further. I wonder if there’s any way to quantify it, at least partially. Not that that is absolutely necessary; there are plenty of curves used to illustrate concepts (the supply/demand curve, IS/LM graph and Laffer curve being a few examples), but if some concrete illustrations of either half of the Alpha Graph could be provided it would help with gaining acceptance of the concept.

    Incidentally, to me it looks less like the letter alpha and more like the Christian fish symbol (or ichthys). Not sure that’s a good thing, but there it is. http://christianity.about.com/od/symbolspictures/ig/Christian-Symbols-Glossary/Christian-Fish.htm

  • Brian Micklethwait (London)


    I agree about the fish thing. However I will stick with Alpha, because in my eyes these graphs go somewhat further into the future than Simon showed them going. He concentrates on the shorter run, and even at the end of his shorter run, the advantage of market arrangements is only just beginning to assert itself. But fast forward a few decades and the graphs stretch out, onward and upward for the market, onward and downward for state imposed arrangements. Alpha then makes more sense. If it’s still a fish, it’s a fish with an implausibly large tail. Not that there are no such fish, but you get my drift.

  • I think it would be an Alpha with a very large tail as well. The “alpha” tag seems to best describe the start of the phenomenon. Hocky stick describes the value proposition(s) a lot better and it’s a shame the phrase is associated so strongly with scientific fraud. Perhaps the phrase’s associations should not stop you from using it?

  • Rich Rostrom

    This essay by Clay Shirky seems highly relevant.

    Shirky notes that in IT areas, distributed, fragmented evolvable solutions always start out as inferior to centrally designed solutions – and always end up displacing and obsoleting them.

    “Centrally designed protocols start out strong and improve logarithmically. Evolvable protocols start out weak and improve exponentially. It’s dinosaurs vs. mammals, and the mammals win every time.”

  • Richard Thomas

    Are these graphs based on any actual data? Or is the “hand waving” a good description of what you’re doing with them? It seems to me that you may be guilty of trying to add more weight to an opinion by producing “impressive looking graphics” without real basis, something we should really be frowning on around here.

  • Brian Micklethwait (London)

    Richard Thomas

    These graphs illustrate my opinions about the contrasting benefits, over time, of state action compared to free markets. I invite agreement. I draw the graphs because I believe these opinions to be (a) true and (b) valuable to explain how the world works and does not work. Waving my arms about graphically draws attention to these opinions.

    Does my observation, to take just one example, that public housing in twentieth century Britain started off doing quite a lot of good and not a lot of harm, and ended up doing very little good and a lot of harm, amount to “data”? Perhaps not. But I think that’s what happened, based upon, you know, half a century of reading, observation, etc. (See also: NHS.) Similar stories about how markets start out crazy but later sort themselves out and then become ever more superb can also be told, in a way that is perhaps similarly lacking in “data”. I like the personal computer story because I have witnessed the entire story so far, and because I have been active in this market as a computer buyer and user. The graphs help me tell those stories, and make them more memorable and more likely to be remembered, because more likely to be noticed.

    I am indeed trying to add weight to my opinions in these matters, because I think they deserve it. In common parlance, I think I am onto something important here.

    I deny that I am “guilty” of anything.

  • Snorri Godhi

    It is an interesting coincidence that just a few days ago, after watching Gangster Squad, i reflected that anarcho-capitalism is unlikely to catch on: at the beginning it would probably be so bloody that most people would welcome a return to a monopoly of violence. Now i have a useful tag for the concept.

    Except that i am not sure alpha is the right word: there is a branch of the Greek alpha that goes down and then up, but the other branch goes up, then down … then up again! That’s not what government services usually do.

  • Richard Thomas

    Brian, since the graphs are based on your opinion, they do not represent reality yet that is the implied intention of placing something onto a graph. By doing so, your graph becomes the new reality yet may actually be very incorrect or even doing your cause a disservice. The tendency will be to build on the graph to create castles in the sky, perhaps with even more graphs. But what if they’re incorrect? What if those lines are not smooth hockey sticks but rather exponential at first with a discontinuity in the second part? What if they are actually straight but it is only perception that gives them that initial shape?

    I am not sure if you are familiar with this cartoon http://imgs.xkcd.com/comics/purity.png but use of graphs in this way is way, way <———-that way.

    Have an opinion, pick some information which informs that opinion and graph it.

  • But in my opinion they do represent reality.

  • Laird

    Richard, that is why, way back at the beginning of this thread, I asked if there is some way of quantifying it. But even without that I agree with Brian that his graph does represent reality. The key word being “represent”. He’s not claiming that it’s a mathematically accurate description of any actual governmental program or private action (that would require a different curve for every specific instance); it’s merely a stylized representation of general trends. In this respect it’s like other stylized illustrative graphs such as the Laffer curve: it doesn’t help you determine the precise point at which taxation is optimal (in terms of maximizing governmental revenues), it merely illustrates the concept that there is a point of diminishing returns after which tax rate increases decrease revenue. And Brian’s Alpha Graph illustrates the possibly counter-intuitive idea that government programs may start out well but eventually turn to sh*t, whereas private actions behave in the opposite way. A very useful intellectual concept, but not a mathematical model.

  • Tedd

    I think Richard has a point in that, to be really useful as an argument, there need to be real-world examples that can be graphed that way, from actual data. (And, ideally, a dearth of counterexamples.) Unlike the Laffer curve, which is more metaphor than math, I think that might actually be possible with this idea.

    On the other hand, with a bit of stylizing it would make a fetching logo for a libertarian party.

  • Richard Thomas

    The Laffer curve is actually a case in point. Whilst the concept is generally correct, zero taxation = no revenue, 100% taxation = no revenue with some kind of shape in between, that’s as far as it goes. Yet people (notably on our side) are quick to use it to back “lower taxes means more revenue”. Which is totally unfounded as we do not know the true shape of the curve or where we are on it. It also tends to lead to the implication that maximized revenues is some kind of desirable outcome and puts the focus on taxation when (in my opinion) the issue is spending

  • Laird

    No argument with that, Richard; in fact, it’s precisely the point I was making. The Laffer Curve is an idealized representation of a concept, not a predictive tool. And the same is true with Brian’s Alpha Graph.