Since somebody came with the idea to promote MinimalistLaw, it's time to explore measurements for law. If the law is to be minimal, we'd better have some metric to know if there's progress. One of the engineering mottos being: you don't know what you can't measure.

I for one would be very curious to find out how many megabytes of text do all the laws applicable to my life (US, California, nevermind municipal laws) contain, and how this measure varied over the years. Anybody can provide some info on such studies ? --CostinCozianu

Some measurements, searching the 109th US congress only (2005-2006), at: http://www.gpoaccess.gov/bills/index.html The statistics were:

 The database contains 8,142,044 words in 5,691 documents. 

I don't have a clue if it's good or bad, though.

On one hand a law professional said:

I'm not aware of any quantitative analysis of the amount of laws before and after various reform movements. I think that just counting laws would be a bit silly. Perhaps more than "a bit". One simple-sounding law can be very complex if the definitions are complex or if it interacts with a lot of different state powers. It's very possible to slash the number of laws in half while multiplying its complexity by thousands. I can say that, for example, prior to the code pleading reform of the 1930s, if someone made a piddling technical error in bringing or participating in a legal action, the legal rules in place at the time made that error disastrous in scope, necessitating an enormous amount of work on the part of lawyers (work the client would have to pay for) simply to be sure every comma was correct and every tiny technical detail was covered. This was replaced with "notice pleading" in which it was only required that there be fair notice to the other parties sufficient to allow them to respond, something that's so simple that nonlawyers can easily do it, and lawyers can do it in a tiny fraction of the time (and expense to the client.) It eliminated all the multifarious and technical "forms of action" and reduced it all to one type of action, you no longer had to decide which crazy latin or legal term defined your action, it was just one kind. Look at the rule listing requirements for Arizona's civil complaints: "1. A short and plain statement of the grounds upon which the court's jurisdiction depends, unless the court already has jurisdiction and the claim needs no new grounds of jurisdiction to support it. 2. A short and plain statement of the claim showing that the pleader is entitled to relief. 3. A demand for judgment for the relief the pleader seeks. Relief in the alternative or of several different types may be demanded." That's all! Before the reform movement of the thirties (and the further simplification in the seventies), the requirements were enormous, and enormously technical. (By the by, would you count this as three laws or just one, in your quantitative analysis?) Lawyers nearly always support making the law simpler, because that means they can take on more clients and serve them effectively. The financial interest runs the opposite of the way you believe it does.

I am not claiming that the law requires an enormous amount of insight or training, it doesn't. But to talk about the complexity of the law without knowing what that complexity is and where it comes from is a little presumptuous. If I looked at your engineering design and said "It's too complicated, make something simpler, you're just trying to pad your bill" when you had made it complicated for a reason, you would be quite justified in telling me that I didn't know what I was talking about.-- JasonCorley

On the other hand one engineer's point of view:

Quantitative measures could (and in my opinion should) be applied to laws. Laws are system of logic and in particular modal logic (if X do Y, A shall do B and shall not do C). There have been experiments with expert systems applied to legal problems that probably you have heard of in the course of your studies and if the complexity of programs can be measured so the complexity of laws can be. There are measures for the complexity of a logical system, of a language, of a computer programs. You can count the primitive clauses in it, you can count the megabytes of data required to store the text of the law. You can also quantify a lot of social aspects related to the functioning of laws: what is the average person's experience with the law on a scale from excellent (10) to disastruous (0), another interesting measure would be to find out how knowledgeable is the average person about the laws that directly impact them ). If the legal profession is not concerned with such measurements then they are disregarding their customer (which is the public at large) and are not concerned with the quality of the process.

To clarify even further I'm not saying quantitative measurements will give the conclusive evidence on whether laws are more or less complex, nor even if they are unnecessarily complex. However, they give enough evidence to warrant a look. If the body of laws grows visibly in some numericla dimensions (megabytes of data, number of primitive clauses), the claim that the law is going towards simplification loses credibility. Quantitative analysis can and often does signal significant qualitative problems, so more often than not it should be done. It's just good engineering, good management, good customer service. And speaking of qualifications, software engineers eat complexity on bread for breakfast: that's what we do (arguably) - manage complexity. --CostinCozianu

A few notes:

• I assume that (western industrial country) law grows by any measurement method you use.
• Measurement of a complex system is difficult. Measuring software is done, but not very successfully. Try "measure a forest" as a beginner's example.
• Measuring law may not be a worthwhile goal on its own because it seems a special case of measuring knowledge (I always wondered how schools can get along without measuring to optimize what they can provide in restricted time).
• Law isn't equal law. The budget law of Austria is approximitely 2000 pages every year. This has no long-term effect, but it may have effects for an (unknown) number of years.
• From looking into some laws I see that an important part of laws defines the words used. A law about cars would define what the word "car" means and probably a hundred additional terms. This is hardly what laymen expect when thinking about laws.

-- HelmutLeitner

The "success" of any measurement depends on what you're trying to accomplish with it. In software for example, if you use LOC to measure productivity you're likely to fail. If you use LOC and even more detailed complexity measurements with the goal to monitor for anomalies and detect problems, it can be quite successful.Now it is apparent that the law grows. Parliaments are happier to use their power to make laws rather than the power to erase laws. But how it grows ? Does it grow linearly ? Should we expect an atenuation of the curve any time soon ? Is this an indication of a potential problem ? I think the law professional, especially academics, should try to to answer these, as an engineer, I raise what it seems to me as obvious questions. --CostinCozianu

An interesting challenge, this is. -- HansWobbe

I'm not at all sure that an acceptable metric is achievable, but there may be a few that are indicative of the extent to which laws are important. It should also be relatively easy to get these metrics from existing sources. I think it would be interesting to speculate about some, such as:

• Number of law books publiched and added to the LOC classification system, annually.
• Number of judgements published by courts.
• Number of practicing lawyers, by type of practice. Number of graduating lawyers, Bar association Memberships
• Length of time taken to enact a new law of a specific type.
• Length of time taken to resolve a case.
• Budgets for a jurisdiction's judical operations.
• Annual revenues of the companies engaged in publishing law books.
• Customer satisfaction numbers. How is the legal system reflected in public opinion.
• The fraction of the GDP that is consumed by legal activities.
• Scope of the applicable jurisdiction.
  • Consider the number of people that fall within the scope of any law and the operational facts of a 'democracy'.
• ...

This discussion had no direct and obvious bearing on practical affairs of participants of the site. If it did, we would be forced to take all data seriously in order to succeed. Since we are not compelled by the fear of failure or the lure of success to open minded to the truth, it's necessary that we hold our own ears open. Try to answer the actual question, "How do you measure the complexity of laws?" in a fair, open-minded, and disciplined way. Let's not bring in Reagan and Republican anti-FDR ideology (e.g. re: tax codes). Note that none of us actually is in charge of the tax code of our respective countries, so let's please ground the discussion in something related to one of our projects ('barns'). -- SunirShah

• Thanks for dropping by ,Sunir. This discussion started from practical matters related to Wikis. Laws govern societies, but they also govern wikis, other online communities, corporations, and so on, so forth. Since BayleShanks proposed MinimalistLaw as a model or ideal to be followed, and Helmut intervened quite vehemently against it, so I thought it nice to intervene with strong arguments in favor of it. Then the fundamental problem to be decided is when is a law too complex ? In order to help in that decision, on a societal scale the system of laws and its evolution needs to be measured. Several interesting measurements have been proposed.

• As for the tax code example, I don't know how you could have gotten the idea that such an example brings in Reagan and Republican anti-FDR ideology. I brought the example as a clear case where the distinction between simple law and complex law can be made. The example is also good because the problems in complex laws are readily aparent and most of the people have been exposed to tax codes of some kind or the other. I'm not trying to get the US congress to read MeatBall and change the tax code, but relevant examples are helpful in any kind of discussion. It strikes me as odd for anyone to look at it as an example of "republican anti-FDR ideology", unless one holds the completely misguided idea that a complex tax code benefits the poor and the middle class at the expense of the rich. It is actually the rich who benefit from complexity in the tax code, as the poor cannot afford tax lawyers. And even if you want to play wealth redistribution through the tax code, you can still do it in one-line formula. --CostinCozianu

Here's an interesting (and quick) question, then: should inheritance be taxed? Does it count as income? -- ChrisPurcell

No, it doesn't count as income. Whether it should be taxed is I think out of the scope of Meatball. Whether it makes sense for a community to tax it, is an interesting question. In Austria there is a 2-6% tax depending on the family relationship and the total value of the inheritance. -- HelmutLeitner

I bring it up because it illustrates another thing we need to be able to measure in law: ambiguity. Minimizing ambiguity is in tension with brevity, but both are needed for simplicity. We could perhaps begin to include coverage of ambiguity by measuring definitions separately from laws, allowing for a very precise but perhaps long definition of a word whilst still keeping the laws themselves simple and short. Something would be needed to keep law from creeping into definition (e.g. something like "income is payment for services, unless the service is growing maize and wheat on an EU-sanctioned farm with properly-rotated crops, in which case income is 46% of the payment" would be dodgy). Perhaps a census of how many people would agree, after some thought, that a definition matches the (or a) common usage of the word?

Measuring the complexity of law is equivalent to measuring the complexity of any system, including code as has been cited frequently. As we know from measuring code, it is impossible to judge how 'simple' code is from internal measures, such as lines of code, number of modules, code paths, memory consumed, cycles consumed. There is a theoretical internal measure called 'function points', but that is not actually measurable.

Further, internal measures of a system may not actually determine the general complexity of the whole system ecologically speaking, as it may be fewer lines of code to embed a virtual machine in a system and then ask the user to input as 'data' the actual program they want to run (which is not an unusual solution! e.g. JavaScript).

One uses internal measures to identify potential places for improvement, but there are many tradeoffs involved in building a system. We choose between them based on cost/benefit tradeoffs that are only sensible from the external frame. For instance, we trade space for time, and both space and time against hardware costs, and all against maintainability. As the external frame changes, choices change too. When hardware was most expensive, space and time efficient coding was most important. Now labour is more expensive, so maintainable code is more important.

Thus, the best way to empirically measure the actual efficacy of systems is from external measures. How effective are they at achieving their objective? How many resources does it consume as input? How productive is its output? How effective is its output at feeding back at acquiring more input? How many resources does it take to build? And how many resources does it take on an ongoing basis to maintain?

When looking at the legal system, we could look at a law based on its stated objective, and determine whether it is having the desired impact and to a sufficient measure, and then use this feedback to improve the law.

However, the law does not act against the universe, but against actors who have control over the law. Since there will be winners and losers, we cannot simply evaluate the law based on the maintainability of the law--rather, we must use a measure that is not yet often employed in ComputerScience called fairness. Fairness is measured inversely against the amount of power being levied against the status quo. Until a law is fair, it will not stabilize.

Since our dummy example is the tax system, while it's a simplistic argument that a simple tax code would cost less, it's obvious that the people who are advocating the simple tax code are the ones who would most benefit. The progressive tax system has been built up and encrufted over sixty years as part of a huge negotiation between all interested stakeholders. Eliminating it would simply unravel the balance of power that has been struck; not necessarily bad, but not trivial to do in a stable way. The problem is that there is no one neutral enough to quickly do the unravelling in a fair way.

The problem with politics is that no one wants democracy. We all want a tyranny that reflects our views. However, the reason we have a democracy is to balance all the potential tyrants against each other so no one can dominate, and that all can be treated fairly.

If we were to move towards MinimalistLaw, the group that is acting will obviously bias how the final law is constructed in their favour. The progressive regime, starting with F.D.R., swung the power dynamic away from what now constitutes the base of the Republicans. It's only natural that, starting with Reagan, they have been articulating a "The government is not the solution, it's the problem" argument for MinimalistLaw. Such a system would benefit them the most as they (the majority, the powerful, the wealthy) are the ones being limited. The contrasting power, the Democrats, has seen the government as a protector. Their natural reaction is to be frightened that their nanny state is dying and they have to fend for themselves.

Neither position is the best, which is why there is such a thing as democracy to sort it all out. The final answer will not be the lowest costing law, but the fairest law.

But is this interesting for Meatball, since American politics always ends up at GodwinsLaw's front door, and has nothing to do with the MeatballMission? Can we reify some of these conclusions away from what is so blatantly the current George W. Bush agenda towards something meaningful to us? -- SunirShah

Sunir, before you imply that somebody's promoting George W. Bush agenda (but would it be such a crime, in principle ?) , make sure you get your facts straight. Both this page and the MinimalistLaw go quite the opposite the actual policies of George W. Bush's administration, who: (a) increased both government spending and government employment big time even when sibtracting the budget for the "war on terror", (b) wants to wield the blunt axe of the law to further shape the social life to a "christian right" mold, including but not limited to by promoting ammendments to the constitution. Mr. George W. Bush is very much big government and big law as long as it's "christian right" government and "christian right" law. If you want to argue the opposite of MinimalistLaw, you have to realize that you'll be closer to George W. Bush's views than you may feel confortable with.

To make a long story short, MinimalistLaw protects against abuses both on the left and on the right (which Mr. George W. Bush is a perfect example of). So this discussion is very much orthogonal to left versus right politics, just for the record I lean on the left. --CostinCozianu

My point is that if we are talking about State politics, not OnlineCommunities, this page is off-topic. See the MeatballMission. -- SunirShah

Of course, you are right. Time for a mea culpa. What attracted to me in this discussion was the idea that MinimalistLaw is good for wikis and other online communities, as long as technical means are there to empower both the community and the individual. Of course, it's easier to promote MinimalistLaw on wikis if "big law" is proven to have bad effects on society at large. --CostinCozianu.