Testimony

of

Jon Roland

before the

Oversight Hearing on Year 2000 Preparedness

of the California Legislature

Wednesday, February 24, 1999

For further information see the WWW page at

http://www.crl.com/~jdr/y2k/y2k.htm


The Year 2000 problem ... is going to have implications in the world ... that we can't even comprehend. ... If we built houses the way we build software, the first woodpecker to come along would destroy civilization. -- U.S. Deputy Defense Secretary John J. Hamre, in testimony before the U.S. Senate Armed Services Committee, June, 1998.

1. Introduction

In this testimony I will be focused on Plan C. Plan A is remediation of the Y2K problem. Plan B is replacement of non-compliant systems. Plan C is about what we do if Plan A and Plan B don't work. It is also about what we do if public anticipation of the failure of Plans A and B results in disaster, something that now seems likely to occur even if Plans A and B do work.

While Y2K is in many ways unprecedented in human history, the potential scenarios stemming from public loss of confidence in the monetary system and in the delivery of vital products and services do have precedent in history. We need to study those precedents.

First, we need to plan for the collapse of the world monetary system. It is essentially fragile, a gigantic Ponzi scheme to which we have become addicted. The Y2K problem doesn't have to be very severe to trigger a world depression worse than that of the 1930s, and at this point it seems likely that Y2K will be severe enough. We have started too late, put too much emphasis on remediation instead of on replacement, and there is no longer enough time to either remedy or replace everything. People know that, and knowing that, they will do what they have to do to protect themselves, and the result will be a cascade failure of the division of labor.

The onset of economic collapse is likely to occur before the worst of the computer problems do, and make it more difficult to deal with them when they do. But when they do, the failure of many systems is likely to compound the depression with effects that most nearly resemble those that have come with civil wars and natural disasters. But whereas most natural disasters have been confined to small areas, this will be global, depriving us of the relief that would be available to disaster areas from unaffected areas.

At its worst, Y2K could be very bad indeed. As bad as a civil war following a depression. Everything depends on how long essential supplies and services are disrupted. If it is only a few weeks, we could avoid major public disorders, but if it goes into months we could have a major loss of life.

We need to keep in mind, however, that while Y2K may provide the occasion for this systemic failure, there are many other things that could trigger it, and probably will during the years to come. Therefore, the Plan C preparations we contemplate today are likely to be needed even if Y2K effects turn out not to be as great as they could be. They are like preparations for an earthquake. We may not know when it is coming, but we know it will come eventually, and almost anything we do to prepare for it is likely to ultimately be a good investment.

We'll reach the end of the century still using software that will crash when it has to handle years starting with "20", and people won't get excited until it's too late to do anything about it. -- Jon Roland, in a talk before a chapter of the Association for Computing Machinery, 1978

2. Preparing the public

The dilemma for public officials has been to try to prepare the public without causing them to panic and thereby precipitate an economic collapse. They have therefore been underplaying the severity of the Y2K problem, while at the same time advising people to take precautions. This approach won't work. The people are too smart for that. And if they feel that they are being deceived, the result will be worse than it would have been if officials play it straight.

At this point I am recommending the public be advised to stock up on a month's worth of supplies, in anticipation of a failure of most services and supplies, including electric power, water, fuel, food, and medicine. More than that would create a demand that can't be met anyway, and any less is likely to leave them short. Most suppliers don't have storage facilities that can handle more than that, so we need to get people to store what they will need in their homes and places of work.

The people also need to be told, bluntly, that they cannot expect normal police, fire, and emergency medical services, that they will be on their own. The lesson of South Central Los Angeles is clear. If there are riots, the police won't be there, and there will not be enough National Guard or military personnel to cover ever city, much less the countryside, and involving the military, besides being a violation of law, would be a fundamental mistake, leading to a police state and civil war. The people are going to have to provide their own police, fire, and emergency services in their neighborhoods. That means they need to be told to get together with their neighbors to organize, train, and equip themselves to do so. The main job for regular police, fire, and emergency personnel will be to assist in the organizing and training of citizens, who will need to supply their own equipment.

The people also need to be told that evacuation to the countryside is not a viable option. There is no way that hoards of refugees from the cities can sustain themselves there, without housing or other facilities, and they are likely to meet armed resistance if they try to seek refuge on private property. People might have been able to feed themselves by hunting during the Great Depression, but that is no longer possible. There are too many people now, and not enough game.

They also need to be told that families and households are not viable under such circumstances, even in a remote area. Only neighborhood groups of at least several hundred persons can effectively meet one another's needs. Each such group needs to have a mix of skills not unlike those needed by a military unit in the field. Everyone needs to do his or her part. There must be no free riders. And groups having a surplus of supplies, equipment, or persons with some skills need to be prepared to trade with those lacking those supplies, equipment, or skills.

3. Community organizing

The first step in community organizing is to divide the state into neighborhoods. These will vary in size, and the boundaries will need to be somewhat flexible, but in cities people need to be organized down to the level of the city block. A meeting site needs to be designated for each such neighborhood, such as a school, church, office, or warehouse, and the people asked to meet together and elect a neighborhood commander and vice-commander. Existing political divisions, such as voting precincts, may serve as the basis for this division, and existing community leaders, such as party chairpersons, neighborhood watch coordinators, or ministers, might be asked to lead the organizing effort, going door to door to urge everyone to participate.

The key to making this work is to make it socially acceptable for organizers to approach their neighbors for this purpose. One of the main impediments to such neighborhood organizing is the anonymity of most neighbors today. Too many people live next to one another without knowing one another, and most people feel awkward approaching strangers. If this effort accomplishes nothing more than getting neighbors to know one another, and become friends, it will be worth a great deal to the future of this society.

While we must try to get the people organized voluntarily, we must not leave it entirely to volunteer action. In case volunteerism fails, we need to have a backup.

The legislature needs to adopt legislation authorizing local magistrates to impose fines and jail time on non-exempt persons who fail to respond to a community call-up, in the same way they may be imposed for failure to respond to a notice for jury duty, and penalties for failure to obey lawful orders of the neighborhood commander or his elected superiors, who shall be designated constables. The sheriff shall be the commander of each county. Exempt persons would be those with disabilities or who have official or professional duties that take precedence over their community duties.

Such penalties should be imposed only as a last resort, after voluntary methods fail, and after repeated and willful resistance. The people in this country have lost the tradition of such community participation and discipline, and it will take some time to get them used to it again. If it is done right, they will embrace it rather than resent it, and develop a new community spirit that will sustain community participation and service voluntarily.

People should be asked to meet on a regular schedule, at a regular location, for preparation and training, more frequently as the end of the year approaches. They should be urged to follow a regular program that covers every aspect of the preparations that need to be made. The program should also include preparations to deal with such eventualities as nuclear and biological attacks, toxic chemical releases, and other disasters that might require evacuation, and each neighborhood should have an evacuation plan to a designated evacuation site, with at least two backup sites. They should rehearse their responses to a variety of possible threats, including injuries to community members.

4. Communications

One of the critical services threatened by Y2K is electronic communications. We have seen how during disasters either phone lines go down or are tied up as everyone tries to call one another at the same time. The advent of digital communications like the Internet has brought a more efficient use of available bandwidth, as messages are broken up into digital packets, which are routed from one node to the next, perhaps along multiple paths, until they are reassembled at their destination. The Internet was designed to survive a nuclear attack on the United States, but the reality of the Internet today is that, while local subnets might survive calamity, the Internet as a whole depends on long-distance, high-bandwidth backbones to carry the major part of the message traffic. Therefore, it is dependent on "wires", that is, fixed channels that require more power than batteries can supply, including fiberoptic cables and microwave satellite links. These are subject to disruption.

The traditional answer to the problem of emergency communications has been the amateur radio community. The current system it offers is the National Traffic System, which supports a cooperating network of radio operators using a variety of equipment and transmission protocols. For more details see http://www.arrl.org/field/pscm/sec2-ch1.html Unfortunately, the number of participants in it are few and scattered, and while this system may be helpful for localized disasters of short duration, it would almost certainly be overwhelmed by a worst-case Y2K scenario.

The ideal system would be a wireless Internet, consisting of a nationwide network of independent portable, battery-operated handheld or wearable computers, linked together by high-bandwidth transceivers, that could relay digital message packets along a route to their destination, without ever having to go through a fixed station or backbone. I have been advocating such a system since 1994, and my latest short version of the proposal, which I call the WIPnet, for Wireless Internet of Portable Nodes, is at http://www.constitution.org/wipnet/wipnetprop.htm

The good news is that the hardware components needed to build such a network at a low unit cost are now becoming available, just in time for Y2K, and the software needed to make them work could be developed during the next few months, if a focused effort to do so were being made. The bad news is that the hardware manufacturers, many of which are based in California, are mainly focused on niche commercial markets in which they can tie their transceivers to base stations and charge usage fees. Research on something like a WIPnet is being left to poorly-funded amateurs.

Perhaps the single most important thing members of the California legislature could do would be to put personal pressure on the manufacturers of the components needed to build a WIPnet to focus their companies on developing the essential hardware and software, and on marketing the units at a low cost. The legislature might budget for the purchase of some number of such units, at a fixed cost, as an incentive to the companies to do so.

Such state-purchased computer-transceivers would then be distributed to key agencies, perhaps even down to the level of local neighborhood commanders or their designated communications officers.

However, if we could establish a statewide or nationwide network of computer-transceivers we would still require a way to communicate in broadcast mode with the general population. For that purpose, since power may not be available for commercial broadcasting, I recommend that we encourage the establishment of a network of AM and FM microbroadcast stations, radiating at 1 to 10 W, to serve as a link between the WIPnet and the population. The FCC is currently considering granting licenses to station operators of this kind, but it is being opposed by the commercial stations, and it seems unlikely that they will act in time to meet the emergency that Y2K presents unless pressure is put on them. See http://www.fcc.gov/Bureaus/Mass_Media/Notices/1999/fcc99006.txt

The California legislature should address a resolution to the FCC to immediately grant free AM and FM operator licenses up to 10 W for non-commercial, public-service microbroadcasting, subject to noninterference with established commercial stations. It should also ask them to allocate more bandwidth for spread-spectrum transmission to be used for the WIPnet.

A plan should then be made to get such microbroadcast stations set up in strategic locations, covering the entire state, and operated by volunteer operators as part of the statewide emergency communications system. The frequencies of these stations should be widely publicized, and persons encouraged to listen to them at regular intervals for important public information. There should be regular reports on Y2K and other preparedness subjects, and news of any early difficulties that may arise and responses to them.

As developments unfold, a regular news service should emerge to keep everyone informed, provide necessary guidance, and allay public alarm. If disruptions should occur, people can be warned to avoid the areas of disturbance, or respond to contain it.

The legislature should budget funds to convert all printed materials relevant to Y2K and other preparedness matters into digital form and put on Web sites or otherwise distributed over the Internet.

There is also a need to free up the time of state and local personnel having special skills that need to be imparted to citizens in training sessions.

The legislature should authorize and direct diversion of state and local personnel with preparedness skills from their regular duties, and be allocated travel funds, so that they can travel and provide training to local citizens preparedness groups all across the state.

5. Other recommendations

Enact greater discretion to department heads to reallocate resources to solve computer problems. Present legislation tends to micromanage departments to the extent that they often cannot even evaluate their needs, much less do anything about them, without an act of the legislature.

Authorize increased pay for programmers, with discretion to managers to pay what it takes to get qualified personnel. A top programmer may be 100 times as productive as an ordinary programmer. The state is likely to lose many of the programmers it needs to solve its Y2K problems to the private sector if it does not match what the private sector will offer them.

Reorganize departments to have technical decisions made, and technical personnel hired, by technically trained persons.

Set up a network of investigators, outside the State Auditor's office, to inquire into how well departments are operating generally, especially on computer matters, and how well critical private sector providers are dealing with their Y2K problems. The legislature should not have to rely only on reports by top officials who may or may not understand what is going on in their own organizations. Someone needs to talk to the people actually doing the work. And if some sector like the railroads refuses to testify, someone needs to be able to find out why and what is really going on.

Put the priority on replacement, not remediation. Most old software and the machines it runs on are probably long overdue for replacement anyway. And replacement is the only way anyone can be fairly sure there won't be Y2K failures, especially in the time available. Mainframes should be phased out wherever possible, and replaced by networks of supermicros. A regular schedule of replacement should be adopted when systems are acquired, with hardware replaced at least every five years and software replaced at least every seven. It if ain't broke, it will be soon. When in doubt, replace it anyway.

Break up the 58 existing California counties into at least 200 smaller ones. Present counties are too large. They are like small states, and have too much political clout to manage them effectively. Most authority should be moved down to such counties. They should be further subdivided into wards and precincts with their own elected officials who enforce the laws at the lowest possible level.

Forget gun control. What we need to do is arm the good guys and disarm the bad guys, but everything proposed on the subject so far will do the exact opposite. The last thing we need during a worst-case Y2K catastrophe will be for innocent citizens to be unable to protect themselves and their neighbors, when the police won't be able to do it. Without deterrence from a well-armed citizenry, we can expect to see the emergence of gangs of looters seeking easy targets. And the only constitutional way to disable the right to keep and bear arms, or any other right, is either within the sentence upon conviction in a criminal trial, or by order of the court in a competency hearing. The state should maintain a database of such disablements, rather than just records of convictions or commitments, disablement on the basis of which are constitutionally prohibited bills of attainder.


WIPnet Proposal | "Counterintuitive Behavior of Social Systems" | "Three Futures for Earth"