This is not unlike how an opposing party will criticize a bill based on its number of typed pages. The thump of a couple of inches of paper on a podium makes for great and reductive theater. It's likewise naïve to consider the number of regulations to be a direct barometer of the goodness or badness of regulation in general. Next we'll see the refactoring of existing regulations into smaller or larger components to affect their numerological residue.
Wanton numerological regulation slashers should be wary of Chesterton's Fence[1].
On the scale of meaningfulness, this idea is somewhere between the Dow Jones Industrial Average and middle school kids playing with margins, spacing, and font size to satisfy a teacher's page-length requirement.
The actual Tacitus quote doesn't really serve your argument. It's "The more corrupt the government, the more numerous the laws." -- you've got it backwards.
Complexity is one thing (although it should be obvious that complex domains often require complex descriptions). But number is arbitrary.
Suppose there exists a set of three distinct regulations regarding ways to safely manufacture a drug, and a dangerous new process is invented for which there is wide consensus that a new regulation is required. Which two of those first three should be discarded? Must every new regulation be weighed in arbitrary relative value against every other possible combination of two regulations in that subject matter area? That department? The country? By reductio ad absurdum, the cut-two-to-add-one rule means that the ultimately correct number of regulations is one. And if it's not one, then there must be some higher "correct" number of regulations at which point the subtraction rule would no longer apply.
Your argument seems to be "whatever that number is, surely it's lower than what we have now". How is that number decided and by whom? Subject matter experts? Voters weighing ballot options written by government officials? Does the number somehow fall or rise to make room for new needs judged important enough? If so, by whom? Is one complex regulation better than ten simple ones? Are ten simple ones more "corrupt" than one complex one?
Unfortunately there's no obvious way to apply a tree-shaking algorithm to the full body of regulations, other than to have humans look at a rule and all agree that the "blue dress on Sundays during harvest" rule can be scrapped. And Chesterton's fence makes many of those judgments risky. So now we're back at politics.
The other thing is that presumably it's easily circumvented by dressing up multiple distinct regulations as a single new regulation.
If you make a habit of making regulations long and multifaceted because scope to pass new ones is limited, it probably becomes easier to bury terrible ideas in amongst things people actually really want to pass.
That I completely agree with, there's a lot of lobbying done to obfuscate regulations and tax code for that reason, and it should be dealt with. I was merely pointing out that it's not just a binary issue, like the proposals in the .pdf suggests.
> Sure, but the priority order should be: complexity,length,number.
I think that, of these three, reducing complexity is the only one to which we should directly aspire; reducing the length and number of regulations is, I think, helpful only to the extent that it reduces, or at least doesn't increase, complexity.
Completely agreed. That and reducing ambiguity. On the whole Increasing both length and number is perfectly acceptable if doing so reduces complexity and ambiguity.
Much like with code really. Your cleve 2 line function is really clever, but most of the time the easy to read 30 lines and 3 function version if probably better.
I agree. I hope that some version of this process is implemented and that it will change government from how it currently operates: adding layers of regulation upon layers, to a process with iteratively improves regulation--weeding out the out-dated, the irrelevant, the imprecise, the ineffective, and replacing it with improved, timely, precise, relevant, effective regulation. So we end up with a regulatory base which is more comprehendible, manageable, and relevant.
This scheme is vulnerable to the same sort of gerrymandering game we play with districts. It provides no definition or guideline on the size or shape of a (presumably atomic) regulation. It's a meaningless bullet point.
How do you guys deal with re-factoring code? Imagine a codebase that has been largely additive for 100 years.
While the "+1 minus two guideline" has plenty of shortcomings in the long term, there is lots of low hanging fruit now, and it's an important mindset shift.
Guys, don't just add LOC. Refactor, clean it up and make it better. Remove blocks we don't use anymore.
Government is strongly limited by time. There's a hard ceiling on the amount of things that can be proposed, debated, and voted on within a term. Consequently, if there are things in the statute that are no longer used it's often better to simply ignore them than to spend precious time in the house arguing about removing them. It might only take 5 minutes to call the house to order, have someone stand up and say "We don't need a law banning witchcraft any more" and then have a vote where the result is a foregone conclusion, that's 5 minutes that the government isn't doing something useful that will actually impact people's lives.
The repeal of pointless old laws comes up relatively often here in the UK. Some of our laws are really old - the government was talking about repealing some that were passed almost 750 years ago recently http://www.bbc.co.uk/news/uk-politics-30334812
>>>Government is strongly limited by time. There's a hard ceiling on the amount of things that can be proposed, debated, and voted on within a term. Consequently, if there are things in the statute that are no longer used it's often better to simply ignore them than to spend precious time in the house arguing about removing them.
That is exactly why all laws should have sunset rules built into them.
We should have listened to Thomas Jefferson who wanted all laws, even the constitution itself to expire every 19 years
I think we programmers are ahead here - the general public needs sometime to think this over. The "+1 - 2" rule looks like a good start for opening the discussion.
Then again there were 3,378 new fed regulations totalling 81,611 pages in 2015. It can be counter productive if the new regulations are so voluminous that no one has time to read them.
I ran into problems along those lines in the UK buying some stuff in Spain. Apparently "don't money launder" translates in to a 1 ft high stack of EU regulations that no one understands. I'm not convinced they are hugely better than the three word version.
Wanton numerological regulation slashers should be wary of Chesterton's Fence[1].
On the scale of meaningfulness, this idea is somewhere between the Dow Jones Industrial Average and middle school kids playing with margins, spacing, and font size to satisfy a teacher's page-length requirement.
[1] https://en.wikipedia.org/wiki/Wikipedia:Chesterton%27s_fence