Maurice Williamson - Report of the Y2K Task Force

Appendices

INTRODUCTION

The nature of the problem
The international perspective
The investment for readiness and the costs of failure
The Task Force's brief
The Task Force's response
The Task Force's report

2.0 Introduction

The nature of the problem 2.1 The essence of the Year 2000 problem is that the computers on which we have come to rely so heavily may not work as intended.

2.2 This is not about one computer or piece of equipment containing computer technology¹. It is about any product or system that works using computer technology, and the resulting impact of any widespread failure on essential goods and services. It could involve any machine or system from the family PC to a city's sewerage system.

2.3 The Year 2000 problem - also described as the "millennium bug" - has arisen because many organisations throughout the world have information technology (IT) systems which were programmed using a two-digit representation² for the year.

2.4 This problem does not affect all computers, all systems and all machines or only the one date - 1 January, 2000. The Year 2000 problem involves mainly those machines and systems regulated by embedded computers³ whose programming does not correctly deal with dates in the next century. Errors may occur where two digit year dates have been used (so that 2001 cannot be distinguished from 1901), by the failure of the chip to recognise 2000 as a leap year or specific dates have been used to convey indefinite information, such as 9/9/99 to denote the end of a computer file. These errors - unless identified and corrected - could have significant consequences for the organisations concerned and the wider communities and countries in which they operate Affected systems and machines will behave unpredictably or simply fail.

2.5 Some embedded systems do not incorporate a date function. These will not present problems nor will those systems already correctly programmed with a four-digit year date. Some systems have been programmed incorrectly and these will present problems.

2.6 The biggest difficulty of all is identifying in advance those systems that will present problems and those systems that will not.

2.7 The difficulties created by those systems that do not correctly deal with the change in century will range from the mildly inconvenient (not being able to access a building or use a lift), to critical situations threatening life or personal safety (failure of hospital life support or water systems) or the national economy (failure of the financial system).

2.8 It is not a problem confined to each organisation nor is it capable of being remedied entirely on an individual basis. Maintaining the supply of critical goods and services could present major problems for many organisations. An organisation's own systems may be fault-free, but those of its suppliers may not be. To paraphrase John Donne, no organisation is an island.
2.9 The incidence of failure may not be high. Much safety-critical⁴ software has been developed according to stringent engineering design and implementation criteria. The Gartner Group estimates that there will be 50 billion embedded systems in use worldwide and that around 1 to 3 percent of these will have Year 2000-related failures leading to shutdowns, erroneous results or chaotic behaviour. Of this, a fraction are in mission- critical systems, leaving in the order of 25 million systems which must be repaired worldwide.⁵

2.10 Uncertainty over the extent of possible failure is a key element of the Year 2000 problem, getting in the way of a "perfect" solution. We believe it is important to recognise that there will be problems. As a country, we need to do the best we can, in the time available, to manage a difficult situation. That means identifying the critical areas and being assured that they will not present problems or, if they do, that there are robust contingency plans⁶ to manage the situation effectively.

2.11 Testing to see that everything will work according to plan is not easy. Testing is both time and resource intensive. Sometimes it is difficult to find the embedded chips and, once found, the mechanism for changing the date on an embedded device may not be obvious. Even devices with the same model number may differ simply because the chips used in their construction vary. Finally, in interlinked systems⁷ problems may only occur when all elements of the system are moved forward simultaneously. Effective testing of these systems may be particularly difficult when not all segments of a system are under the control of a single organisation, such as New Zealand's electricity supply network.

2.12 It is important to recognise that the longer the delay in addressing Year 2000 problems, the fewer the number of remedial options available and the less effective they are. Action must intensify as soon as possible.

2.13 Testing is a vital part of the remedial process, but it is the most time-intensive. If organisations do not have the time or resources for testing, they must plan for contingencies.

2.14 The Year 2000 problem is everybody's problem and no organisation can stand aside from it. While the detail of the problem is technical, ultimately its solution lies in a strategic and business continuity⁸ response at the board of governance and chief executive level.

2.15 This problem is unique. It never has happened before and, knowing what we know now, it is unlikely ever to happen again. This means that there is no readily available store of experience or tool-box for managers to consult in order to address their Year 2000 problems. Even if they understand the issue, and many do not, they do not know where to turn for help.
The international perspective 2.16 Clearly, interlinkages and supply chain problems are not confined within New Zealand. This is a global problem. US President Bill Clinton says it is critical that the United States Government do its part in addressing this challenge⁹. British Prime Minister Tony Blair warned earlier this year that the Year 2000 problem is one of the most serious facing British business and the global economy today. Its impact, he said, could not be overestimated.¹⁰

2.17 International commentators have recorded particular concern with the lack of response to the Year 2000 problem in some Asian countries and the possible consequent economic impacts on the global economy.¹¹

2.18 For example, Dr Ed Yardeni, chief economist for Deutsche Bank Research in the US, believes that there is a 70 per cent chance of a severe global recession which could be as severe as the 1973-74 downturn. He estimates that the production of goods and services in the US could fall by 5 per cent and that on a worldwide basis it could fall by US$2 trillion.¹²

2.19 A chart listing some of the responses to the Year 2000 problem by other governments around the world is appended to this report (Appendix II). While the responses may differ in detail, their overall message is clear; the management of this issue is critical to the continued wellbeing of any nation's social and economic infrastructure, demanding government attention and involvement at the highest levels.

2.20 The international perspective takes on additional significance for New Zealand. We are the first country in the world to experience the ticking of the clock from 11:59 pm on 31 December 1999, to 12:00 am on 1 January 2000.Whether we welcome it or not, the world's focus will be on New Zealand. In order to reassure our major trading partners, we need to manage the Year 2000 problem successfully. Successful management calls for recognition of 3 principal facts. They are: Year 2000 represents a significant problem requiring a great deal of work on the part of both the public and private sectors; the problem, if it is left unmanaged, has the potential to pose major risks to our social and economic infrastructure and the health and safety of our citizens; and time to manage the problem is limited and pressing. As we move towards Year 2000 and beyond, solutions will decrease in their effectiveness.

The investment for readiness and the costs of failure 2.21 We have not undertaken any analysis of the scale of investment needed for readiness against the costs of failure. International estimates are speculative to say the least, ranging from a worldwide investment of $US600 billion¹³ to $US1.6 trillion¹⁴.

2.22 The only possible conclusion is that the investment is significant. One estimate puts the cost to New Zealand government departments at $40 million¹⁵; the consultancy group Coopers and Lybrand advised the Government Administration Committee¹⁶ that New Zealand businesses will spend an estimated $450 million on Year 2000 compliance¹⁷.

2.23 One estimate puts the cost of failure for New Zealand at $2.5 billion for one month's failure to $10 billion for four months' failure.¹⁸ We are unable to judge the accuracy of this estimate, nor do we wish to offer up an estimate of our own.

2.24 The point is this. The potential impact of failure on our social and economic infrastructure, the health and safety of our citizens and our international reputation is huge. Our aim must be to prepare ourselves as well as possible at the lowest possible cost. This is the starting point for this Prime Ministerial Task Force.

FOOTNOTES

A wide range of devices contains computer technology and the programming necessary for them to operate. return
To a computer, all information is ones and zeroes. Useful information, composed of letters, numbers and dates, must be rendered in some pattern of ones and zeroes for computers to be able to use it. While there are only a few obvious ways of translating numbers and letters into computer codes, a wide variety of ways of encoding dates has been used. Some of these allow room for only the last two digits of the year.return
Many machines use computer processor chips for control. Examples are video recorders, microwave ovens and some cars. Not all these devices perform calculations on dates or even know the date, and some of these which do are correctly programmed. Others will fail due to the Year 2000 problem. return
Programming whose incorrect operation might threaten human life, e.g. aircraft control systems. return
Andrew Kyte: The Year 2000 Crisis: A framework for Embedded Systems, The Gartner Group 1997. return
6 Plans to minimise the adverse effects resulting from possible events. return
Computers both within and between organisations are often connected together. The largest example of this is the Internet which comprises tens of millions of separate, interlinked computers worldwide. return
The practice of planning in advance to deal with an unexpected, adverse event such as fire to ensure that an organisation can continue in business. Also called a disaster recovery plan. return
President Clinton: Executive Order: Year 2000 Conversion, 4 February 1998.return
Prime Minister Tony Blair: Introduction to Millenium Bug Campaign, Action 2000,1998. return
Stuart Kennedy: Gartner Sees Y2K Misery in Asia, Computer Daily News, 25 March 1998 return
Dr Ed Yardeni: Year 2000 Recession? www.yardeni.com/y2kbook.html, 2 July 1998. return
The Gartner Group as cited by Dainty Duffy in Software Magazine: Year 2000 News and Analysis, Field Report, 15 April 1998. return
Capers Jones, Chairman, Software Productivity Research Inc: The Global Economic Impact of the Year 2000 Software Problem, 1997 return
Government Administration Committee: The Year 2000 Inquiry into the Year 2000 date coding problem, p14, April 1998. return
Governmment Administration Committee: op cit, p13. return
There are a number of definitions of Year 2000 compliance including the British Standard DISC PD 2000-1. They amount to being able to deal correctly with all dates in the next few years. return
Submission to Government Administration Committee by consultant, Michael Ross. return

MINISTER'S HOME PAGE || EXECUTIVE HOME PAGE

The nature of the problem	2.1	The essence of the Year 2000 problem is that the computers on which we have come to rely so heavily may not work as intended.
	2.2	This is not about one computer or piece of equipment containing computer technology¹. It is about any product or system that works using computer technology, and the resulting impact of any widespread failure on essential goods and services. It could involve any machine or system from the family PC to a city's sewerage system.
	2.3	The Year 2000 problem - also described as the "millennium bug" - has arisen because many organisations throughout the world have information technology (IT) systems which were programmed using a two-digit representation² for the year.
	2.4	This problem does not affect all computers, all systems and all machines or only the one date - 1 January, 2000. The Year 2000 problem involves mainly those machines and systems regulated by embedded computers³ whose programming does not correctly deal with dates in the next century. Errors may occur where two digit year dates have been used (so that 2001 cannot be distinguished from 1901), by the failure of the chip to recognise 2000 as a leap year or specific dates have been used to convey indefinite information, such as 9/9/99 to denote the end of a computer file. These errors - unless identified and corrected - could have significant consequences for the organisations concerned and the wider communities and countries in which they operate Affected systems and machines will behave unpredictably or simply fail.
	2.5	Some embedded systems do not incorporate a date function. These will not present problems nor will those systems already correctly programmed with a four-digit year date. Some systems have been programmed incorrectly and these will present problems.
	2.6	The biggest difficulty of all is identifying in advance those systems that will present problems and those systems that will not.
	2.7	The difficulties created by those systems that do not correctly deal with the change in century will range from the mildly inconvenient (not being able to access a building or use a lift), to critical situations threatening life or personal safety (failure of hospital life support or water systems) or the national economy (failure of the financial system).
	2.8	It is not a problem confined to each organisation nor is it capable of being remedied entirely on an individual basis. Maintaining the supply of critical goods and services could present major problems for many organisations. An organisation's own systems may be fault-free, but those of its suppliers may not be. To paraphrase John Donne, no organisation is an island.
	2.9	The incidence of failure may not be high. Much safety-critical⁴ software has been developed according to stringent engineering design and implementation criteria. The Gartner Group estimates that there will be 50 billion embedded systems in use worldwide and that around 1 to 3 percent of these will have Year 2000-related failures leading to shutdowns, erroneous results or chaotic behaviour. Of this, a fraction are in mission- critical systems, leaving in the order of 25 million systems which must be repaired worldwide.⁵
	2.10	Uncertainty over the extent of possible failure is a key element of the Year 2000 problem, getting in the way of a "perfect" solution. We believe it is important to recognise that there will be problems. As a country, we need to do the best we can, in the time available, to manage a difficult situation. That means identifying the critical areas and being assured that they will not present problems or, if they do, that there are robust contingency plans⁶ to manage the situation effectively.
	2.11	Testing to see that everything will work according to plan is not easy. Testing is both time and resource intensive. Sometimes it is difficult to find the embedded chips and, once found, the mechanism for changing the date on an embedded device may not be obvious. Even devices with the same model number may differ simply because the chips used in their construction vary. Finally, in interlinked systems⁷ problems may only occur when all elements of the system are moved forward simultaneously. Effective testing of these systems may be particularly difficult when not all segments of a system are under the control of a single organisation, such as New Zealand's electricity supply network.
	2.12	It is important to recognise that the longer the delay in addressing Year 2000 problems, the fewer the number of remedial options available and the less effective they are. Action must intensify as soon as possible.
	2.13	Testing is a vital part of the remedial process, but it is the most time-intensive. If organisations do not have the time or resources for testing, they must plan for contingencies.
	2.14	The Year 2000 problem is everybody's problem and no organisation can stand aside from it. While the detail of the problem is technical, ultimately its solution lies in a strategic and business continuity⁸ response at the board of governance and chief executive level.
	2.15	This problem is unique. It never has happened before and, knowing what we know now, it is unlikely ever to happen again. This means that there is no readily available store of experience or tool-box for managers to consult in order to address their Year 2000 problems. Even if they understand the issue, and many do not, they do not know where to turn for help.
The international perspective	2.16	Clearly, interlinkages and supply chain problems are not confined within New Zealand. This is a global problem. US President Bill Clinton says it is critical that the United States Government do its part in addressing this challenge⁹. British Prime Minister Tony Blair warned earlier this year that the Year 2000 problem is one of the most serious facing British business and the global economy today. Its impact, he said, could not be overestimated.¹⁰
	2.17	International commentators have recorded particular concern with the lack of response to the Year 2000 problem in some Asian countries and the possible consequent economic impacts on the global economy.¹¹
	2.18	For example, Dr Ed Yardeni, chief economist for Deutsche Bank Research in the US, believes that there is a 70 per cent chance of a severe global recession which could be as severe as the 1973-74 downturn. He estimates that the production of goods and services in the US could fall by 5 per cent and that on a worldwide basis it could fall by US$2 trillion.¹²
	2.19	A chart listing some of the responses to the Year 2000 problem by other governments around the world is appended to this report (Appendix II). While the responses may differ in detail, their overall message is clear; the management of this issue is critical to the continued wellbeing of any nation's social and economic infrastructure, demanding government attention and involvement at the highest levels.
	2.20	The international perspective takes on additional significance for New Zealand. We are the first country in the world to experience the ticking of the clock from 11:59 pm on 31 December 1999, to 12:00 am on 1 January 2000.Whether we welcome it or not, the world's focus will be on New Zealand. In order to reassure our major trading partners, we need to manage the Year 2000 problem successfully. Successful management calls for recognition of 3 principal facts. They are: Year 2000 represents a significant problem requiring a great deal of work on the part of both the public and private sectors; the problem, if it is left unmanaged, has the potential to pose major risks to our social and economic infrastructure and the health and safety of our citizens; and time to manage the problem is limited and pressing. As we move towards Year 2000 and beyond, solutions will decrease in their effectiveness.
The investment for readiness and the costs of failure	2.21	We have not undertaken any analysis of the scale of investment needed for readiness against the costs of failure. International estimates are speculative to say the least, ranging from a worldwide investment of $US600 billion¹³ to $US1.6 trillion¹⁴.
	2.22	The only possible conclusion is that the investment is significant. One estimate puts the cost to New Zealand government departments at $40 million¹⁵; the consultancy group Coopers and Lybrand advised the Government Administration Committee¹⁶ that New Zealand businesses will spend an estimated $450 million on Year 2000 compliance¹⁷.
	2.23	One estimate puts the cost of failure for New Zealand at $2.5 billion for one month's failure to $10 billion for four months' failure.¹⁸ We are unable to judge the accuracy of this estimate, nor do we wish to offer up an estimate of our own.
	2.24	The point is this. The potential impact of failure on our social and economic infrastructure, the health and safety of our citizens and our international reputation is huge. Our aim must be to prepare ourselves as well as possible at the lowest possible cost. This is the starting point for this Prime Ministerial Task Force.