Wednesday, 20 April 2011

The misguided quest for ultimate simplicity

There's a big tendency for politicians especially to say that the simpler something is, the better. Simplicity, above all else, should be the consideration. If they're looking at being elected, a promise to cut red tape - the red tape has now been cut so many times that the next government will be buying special nano-scissors to separate it atom by atom - is inevitable.

Now, I'm not saying that simplicity is a bad thing as such. Any added complexity has to be justified, because the more complex a system is to understand, the fewer people will understand it, and the more likely it is that someone whose life, health, family or livelihood depends on understanding it won't be able to.

Generally this added complexity on average benefits default people - who are more likely to have the spare time, energy and resources to look into it - and harms less privileged people. Most people spend most of their time not interacting with any particular system. The more complex and unintuitive1 it is, therefore, the harder it'll be to use when they have to.

The process for claiming benefits is designed to be complex to "catch fraud". What's not been acknowledged by government is that - almost by definition - real fraudsters are going to have more time to understand the forms, get the various bits of supporting paperwork together, and so on, than a person who actually needs the benefits and is trying to complete the form while also dealing with the situation that means they need the benefits in the first place.

At a commercial level, companies can offer a confusing range of choices to make it unlikely that their customers will pick the optimal one.

So, simplicity is good. But like a lot of things, not an absolute all-overriding good thing. Added complexity in the form of protective regulations can be helpful - equalities, health and safety, employment law, housing standards, etc. - and yes, it's more complex and costly2 to run a business or build a house to those standards, but the benefits received in terms of people being alive, more productive and happier are usually worth the costs.

When mainstream politicians talk about "cutting red tape", though, they don't generally mean "simplifying the benefits process". Instead, they go for the regulations which are probably doing more good than harm to keep. But it's a populist stance, and it's generally supported by the press - whose owners and editors also have a dislike for regulations that prevent them maximising their short-term profit - so it's one of those rarely-questioned truths: "simplicity good, complexity bad".

I was reminded of this recently when some publicity for the coming referendum dropped through the door, for the No campaign. One of the pages had a comparison of the FPTP and AV systems - the FPTP system with a short three sentences describing it, the AV system with a full column of long paragraphs: all accurate, but written in a very expanded form. Above both, a description of why one should vote No for simplicity.

In the spirit of that - and since the No campaign is opposed to spending money on voting:

Which is simpler?
You do what I say.

We divide the country into a number of "constituencies" of approximately equal population size, attempting to make the boundaries between the constituencies follow natural borders, historical divisions, group similar settlements together, and so forth. These boundaries are then reviewed and revised at great expense every few years to account for population migration, new buildings, and so on.

Within each constituency, all citizens aged 18 or over (almost all, anyway - see Appendix A for exceptions) will be placed on an electoral register to register their entitlement to vote in elections.

Every five years, or sooner if the government wishes or if the government is unable to maintain the confidence of parliament, or later but at the earliest possible opportunity in the event of a major national emergency, parliament will be dissolved. There will then be an election by the following process, with an expected cost of around £100 million.

Candidates will then be appointed on payment of a £500 deposit in each constituency, requiring at least ten signatures of residents of that constituency in support of their candidature. All eligible voters will be entitled to be a candidate if they can meet those requirements, and need not be registered within the constituency they are standing for election in.

[several pages about postal votes, proxy votes, campaign funding regulations, campaign regulations, vote secrecy, vote security, payment rates for counting staff, the largely honorary role of the Returning Officer, avenues of appeal to the courts, etc. omitted at this stage]

The winners of the elections then form the parliament.

[several more pages about by-elections, forming of governments, etc. omitted.]


1 A term that implies that things can be universally "intuitive" despite different people having different intuitions, based on their personal history, culture, neuro[a]typicality, learning, practice, and so on. Still, even if it's not possible for a system to be entirely intuitive, it can be designed so that it isn't needlessly obscure.

2 Some types of accessible design are actually either cheaper or no more expensive than the "traditional" inaccessible forms, provided that they're planned for from the start. Fitting them into the design as an afterthought, or on to existing technology or buildings, on the other hand, can be very expensive.