[FD3SA]

Directy democracy fails for exactly the reasons you have stated, yet those reasons all have one source: complexity.

The system of law was designed not by engineers, but by law makers who had no concept of system architecture and runaway complexity. They employed an imprecise language to construct an unstable structure which requires continuous patching and propping, resulting in a quagmire of legislation, open to manipulation due to its inherently contradictory and subjective foundation.

This problem can only be solved by proper system design, but it requires a hard reset of extant legislation. Once the law is written such that every literate citizen can understand it, we will have arrived at direct democracy.

[JimmyL]

I'm not even sure that clear, literate laws would do it. They'd help a huge degree, for certain, but I think the breadth of issues that Congress has to consider would still overwhelm pretty much anyone who wants to deal with it, save those who can afford a professional staff to help them out.

Consider an issue like water-use rights on the Colorado River. Even if you codified all the regulations relating to it into a well-organized and clear format, there's still the difficulty of figuring out what the right answer is. How would an average literate and interested citizen consult with farmers, environmental groups, conservationists, power companies, land-use stakeholders, first-nations groups, agricultural companies, municipalities, individuals who have been on the river for decades, bureaucrats, etc. and then link all those consultations to what's going on with other affiliated riversheds, general environmental and developmental polities, fish stocks, etc.

Even if the law is super-easy to understand, I think it would still be difficult for an average unsupported person to understand the subtitles and implications of something like a water-use bill, let alone one that regulates complex financial and regulatory transactions.

[nitrogen]

Use interactive infographics to visualize the effects of various legal algorithms. Let people drag allocation around, tweak parameters, etc. Combine all the data, including expected satisfaction of the various interested parties, with expert statements from all perspectives. Display the results of regression tests (e.g. display how this change in the algorithmic law would affect reality now if it had been made 5, 10, or 20 years ago). Thus, you allow non-experts to make informed decisions (exactly how congresspeople are supposed to work now).

In the resource allocation example, each interested party would state the amount of the resource they currently receive, the minimum amount they need to survive, and the maximum amount they could use efficiently, and their expected revenue in all cases (if they are a commercial resource user). They could also indicate their relative satisfaction with the various allocation algorithms proposed by engineers and others. Independent audits could determine how essential a given resource or resource user's product is to the economy, and how efficient the various resource users are.

I'd expect this would require less supporting staff than our current legislative system. Right now, each representative has his or her own complete advisory staff. If the same data is shared by everyone, you'd only need enough people to audit the data, algorithms, and visualization systems to avoid manipulation of outcomes, freeing up current staffers to contribute directly to the economy.

[FD3SA]

This method would have the immense benefit of turning a subjective "policy" into an objective algorithm, whereupon its direct impact could be measured numerically via simulations. As it stands, policy submissions are vague tomes of legal jargon with no measurable effects, coupled with a grandiose vision of implementation which is designed to elicit a positive emotional response from the electorate.

[davekinkead]

Representational democracy also fails for the same reason.