If it can't be expressed in figures, it is not science; it is opinion.
This is neither science nor opinion. It's disputing definitions.
Some people define freedom as being able to do things without consequences. Some people define it as being able to do things, even if you must suffer the consequences. Some people define it is being allowed to do things, even if you don't have the resources to actually do it. Some people define it as things you could theoretically do if you tried hard enough. Some people only count if it's something that you could be reasonably expected to be able to accomplish.
The question to ask is: why does any of this matter? Once you know what you actually care about, measure that. Also, to avoid confusion, don't call it liberty, assuming it even resembles liberty. I don't know about you, but I just care about happiness.
disputing definitions
An interesting point, that I'll have to think harder about.
I just care about happiness
Another definition of freedom/liberty I've heard is, roughly, being able to explore the solution space of what might be able to bring you happiness.
The question to ask is: why does any of this matter? Once you know what you actually care about, measure that. Also, to avoid confusion, don't call it liberty, assuming it even resembles liberty. I don't know about you, but I just care about happiness.
If you just care about happiness, would you approve of an orgasmium tiled universe?
I'm not even sure I disapprove of that, but I'm pretty confident that I don't really just care about happiness.
The trouble here (aside from dragging politics in - you could have started at paragraph 4, for example) is that you took a route that wasn't very helpful to the reader. Halfway through, after some quotes, you basically went "and ignoring that, here's what some other people came up with," and then presented the straight numbers without talking about their methodology for quantifying "liberty," your original topic.
And of course correlation/causation mixups are bad, more in-depth statistical analysis is good, but that has to be solved by reading a book or two, while giving us the details of peoples' methodology can be solved by just reading a few papers.
you basically went "and ignoring that, here's what some other people came up with,"
Freely admitted - though I tried to explain why; simply, that I wasn't able to find the numbers I went looking for.
but that has to be solved by reading a book or two
At the time of writing, I was mostly still at the 'sniff test' stage, to try and do a simple test to figure out if it was worth spending more effort on this line, or if I should look for some other approach. The initial response seems relatively promising - the next stage would seem to be finding /which/ "book or two" to read, or at least find some in-depth data source to get better numbers.
Does anyone have a good 'in' with, say, a university sociology department, who might know where to find "median discretionary income" stats, or even have them already handy in a spreadsheet?
You asked what you did wrong here so I'm going to point out one little thing.
To calculate discretionary income, both taxes and health care are subtracted from gross income.
Why health care? Why not, say, food or rent? There are tons of unmentioned premises here.
Food and rent /are/ also subtracted from gross income to calculate discretionary income; I was highlighting those two items, not defining 'discretionary income', which already has a fairly well-established meaning.
Discretionary income is a complex concept that can hide many premises. I order to fully define it, you need to know how much/what quality of food, housing, and health care are discretionary. (If you would find such a classification useful, don't try to specify each case separately; determine the rule that you use to make such choices.)
Wow, this topic got downvoted into oblivion quickly. I already knew I'm just a beginner at proper Bayesian rationality, and I can take this as a piece of evidence supporting that, plus I got a relevant book recommendation out of it, so it wasn't a complete loss. (I'd have been happier with a few more posts explaining, not just why the downvotes, but what I could do to write on this topic in a better way in the future, but I'll take what I can get.)
I'm trying to apply some LessWrong lessons to certain aspects of politics, in order to encourage some people to adopt LW techniques. Yes, I know it's the mind-killer, but it still consists of problems to try to solve... and one of the core difficulties is that for this particular audience, their main focus, 'liberty', is generally treated qualitatively, resulting in much discussion that is more opinion than evidence-based. So I asked myself - is it possible to treat this thing quantitatively, and if so, what would that mean? - resulting in the following exploration:
Heinlein wrote:
> What are the facts? Again and again and again - what are the facts? Shun
> wishful thinking, ignore divine revelation, forget what "the stars foretell," avoid
> opinion, care not what the neighbors think, never mind the unguessable
> "verdict of history" - what are the facts, and to how many decimal places?
> You pilot always into an unknown future; facts are your single clue. Get the
> facts!
Just as the basic scientific method is the bare minimum necessary to advance real knowledge, but there are specific methods within science (such as Bayesian reasoning) that come closer to the ideal of Solomonoff Induction and thus increase knowledge that much faster, it seems plausible that the Zero-Aggression Principle itself is merely the bare minimum necessary to support individuals' freedoms, and that there could be some particular methods within the overall aegis of libertarianism that are better than others at that task.
But in order to find out what those methods /are/, we need some way to measure which methods do better - some way to quantify how much freedom an individual has. I don't recall having read anything on the matter in libertarian circles - if work on this has already been done, then I'd be happy to hear about it. But until then, I'll just have to try to muddle through on my own.
Yudkowsky wrote:
> When there's a confusing problem and you're just starting out and you have a falsifiable
> hypothesis, go test it. Find some simple, easy way of doing a basic check and do it
> right away. Don't worry about designing an elaborate course of experiments that would
> make a grant proposal look impressive to a funding agency. Just check as fast as
> possible whether your ideas are false before you start investing huge amounts of effort
> in them.
So - /is/ there some reasonably simple way of measuring peoples' freedom to do stuff? Some groups have released studies measuring quality of life in different countries, amalgamating various factors into a 'happiness index' or 'quality of life' measurement... and somewhere in there may be something useful.
But we /might/ have a set of numbers that are easier to find than that. Looked at one way, the general method used these days to determine how to allocate resources is through 'price' and 'money'; and the more money one has to distribute, the more freedom one has to distribute it to various things. So, in at least one sense, we could try comparing the 'discretionary income' of various groups, which is income minus the costs of the necessities of life (eg, food, shelter, health care, paying taxes to keep out of jail, etc). And, if possible, instead of the mean discretionary income, we'd be better served by the median, that of the average person in that society. For similar reasons, we'd also be better served by calculating the total discretionary income acquired throughout someone's lifetime, than simply looking at the annual figure. If this idea works out, then those groups which tend to be qualitatively higher in respecting individual rights (such as one or more of the sortings listed at https://secure.wikimedia.org/wikipedia/en/wiki/List_of_freedom_indices ) should at least roughly correspond to having higher median discretionary income. And, if not, then this is yet another wrong idea to replace with a better one.
Unfortunately, just because a given metric is chosen doesn't mean that the numbers can be immediately found through Google, in which case you just have to work with whatever closely-related numbers /can/ be found. In this case, the closest I've been able to find so far is http://www.worldsalaries.org/total-personal-income.shtml , which lists about 30 countries by disposable income rather than discretionary. But since discretionary income is at least related to disposable income, we can try checking those 30 countries to their listing on freedom indices to see, as a minimal standard, if the top half of the list has more countries listed as 'free' than the bottom half.
(Note - at this point, I /don't know/ what the answer will be. I have a /guess/ - but that guess could turn out to be wrong just as easily as it turns out to be right.)
Looking at the data gives us:
1st column: Freedom House 2011
2nd column: Economic Freedom 2011
3rd column: Press Freedom 2010
4th column: Democracy Index 2010
Good to bad: blue, green, yellow, orange, red
US: blue, green, green, blue
Switzerland: blue, blue, blue, blue
Germany: blue, green, blue, blue
UK: blue, green, green, blue
Austria: blue, green, blue, blue
France: blue, yellow, green, green
Netherlands: blue, green, blue, blue
Australia: blue, blue, green, blue
Taiwan: blue, green, green, green
Sweden: blue, green, blue, blue
New Zealand: blue, blue, blue, blue
Japan: blue, green, green, blue
Spain: blue, green, green, blue
Korea: blue, yellow, green, blue
Hong Kong: yellow, blue, green, orange
Singapore: yellow, blue, orange, orange
Israel: blue, yellow, yellow, green
Russia: red, orange, orange, orange
Czech Republic: blue, green, green, blue
Latvia: blue, yellow, green, green
Slovakia: blue, orange, green, green
Brazil: blue, orange, yellow, green
Lithuania: blue, green, blue, green
Poland: blue, yellow, green, green
Thailand: yellow, yellow, orange, green
Mexico: yellow, yellow, orange, green
China: red, orange, orange, red
Romania: blue, yellow, yellow, green
Philippines: yellow, orange, orange, green
And counting up the columns:
1st half: Column 1: 14 blue, 1 yellow, 0 red.
2nd half: Column 1: 8 blue, 4 yellow, 2 red
1st half: Column 2: 4 blue, 9 green, 2 yellow, 0 orange
2nd half: Column 2: 1 blue, 2 green, 7 yellow, 4 orange
1st half: Column 3: 6 blue, 9 green, 0 yellow, 0 orange
2nd half: Column 3: 1 blue, 4 green, 3 yellow, 6 orange
1st half: Column 4: 12 blue, 2 green, 1 orange, 0 red
2nd half: Column 4: 1 blue, 10 green, 2 orange, 1 red
... so by all four measures, the countries with lower disposable incomes tend to be less free. This means that this idea doesn't /immediately/ fail the sniff test... and so it just might be worth following up on, to see if this continues to be a useful quantification, and if so, what predictions can be made from it that can help guide us as we make plans to increase our freedom. Or perhaps some other measurement will be discovered to be more useful in helping us make such predictions - in which case it would be worthwhile to find /it/.
So, what useful ideas can the simple existence of "lifetime median discretionary income" as a metric give us?
The most obvious is that it when we're faced with a question of what we can do to increase liberty, we know that there are particular actions we can take which will have the /greatest/ benefit, and we can concentrate on identifying them.
We can also get at least a rough feel of what some of those actions may involve, which we might not have otherwise thought of.
For an obvious example: When someone dies, they no longer receive any income at all; thus, working to prevent deaths, and generally extend lives, helps to increase the total discretionary income those people can accumulate over their lives, and is thus an action promoted by use of this metric - which meshes very well with the standard libertarian view of the Right to Life being rather important. So far, so good.
For a counterintuitive example: To calculate discretionary income, both taxes and health care are subtracted from gross income. Reducing taxes means less is taken out of gross, and so is one of the most obvious ways to increase discretionary income. However, /if/ reducing taxes /increases/ the cost of health care by more than the taxes are reduced, then discretionary income will go down - which the use of this metric advises against. What it /does/ advise is to use whatever system results in the /minimum/ cost for health care, whether that cost is paid directly or via taxes (along with some further suggestions about maximizing the benefit-to-cost ratio for health care practices). This perspective goes against the usual grain of libertarian thought, in which taxes are generally viewed as an unalloyed evil... but it may be worth considering what your true goal actually is: increasing liberty /even if/ doing so requires the use of government, or getting rid of government /even if/ doing so reduces individuals' liberty.
Perhaps this counterintuitive piece of advice means you think that the metric being used is utterly useless. You are entirely free to think that, and say so however you wish. I am also entirely free to ignore what you say if you simply disparage one metric without suggesting a better one. After all, as Heinlein wrote:
> If it can't be expressed in figures, it is not science; it is opinion.