The road to privatisation

1 NFT 2/1997 The road to privatisation by Lars Nilsson, Kammarkollegiet These are my points and observations that are discussed and, hopefully, substantiated in this paper: • Private and government sectors are gradually converging in practices and management. Risk awareness is likely to be an area that generally needs to be improved both in the government sector and in the private sector. A holistic management approach to risk is something to be advocated both in the government sector and in the private sector. Administrative, financial and economic control of risk is likely to need more attention than pure technical aspects when treading the road of privatisation. Risk financing programs must be quick and easy to imple- ment in existing and former government agencies. The captive concept is as valid in the government sector as in the private sector as a sophisticated tool for risk management. The spin-offs of a captive may not be so attractive in the government sector as in the private sector. A concept for introduction must perhaps be even better prepared in the professionally investigative and scrutinizing environment of the government sector. The security and reinsurance aspects could be quite different in a government risk financing framework compared with a conventional captive. Agencies should in my experience be expected to be quite keen on receiving and adopting risk management practices. 1 Scope and focus The heading of this session is the road to privatisation. Privatisation means the trans- formation or transition of an organisation and its activities from the state or budget sector to the private sector. The final step is when a state owned company is put up for sale to private owners and the transaction finalised. Such a process is indeed very complicated The article is based on a lecture held at the 5th Luxem- bourg Rendez-Vous, June 4 and 5, 1996. to implement and comprises a vast number of tasks. The end result implies a significant change in many respects. I will deal with just one specific aspect of those. That is the aspect of risk management. My present position, starting in December 1993, is being in charge of the development of risk management in the Swedish govern- Lars Nilsson 2 meaning • identification evaluation control financing I will give my personal views on differences and changes pertaining to the above areas and how related issues could be addressed. What I can and will do is to share with you my experience up to the present moment. I will attempt to make an analysis of common- alities and differences and to make some sort of conclusion that could hopefully contribute to the inspiration, understanding and devel- opment of both privatisation projects and successful implementation of risk manage- ment in the government sector. Hopefully this paper might at least meet some expectations from the rendez-vous in reference to the subject. 2 Development in management and in risk 2.1 The species of risk Over the years there have been numerous ways of analysing and demonstrating the na- ture and scope of risks a company is exposed to. Traditional risk management – I suppose it could be called that by now – made a point of focusing a certain class of risks. This class of risks has been designated e.g. pure risk, static risk and fundamental risk. Risk management has its origin in the insu- rance world. It is however a development driven by the growing insight, on the side of the insurance buyer, of the inadequacy of insurance alone as a measure of protection for a company. The world was made a dichotomy with pure risks being the one part and speculative, dy- namic or commercial risks being the other part. In my opinion this dichotomy was never very well defined, although it presented a ment sector. My previous experience is insu- rance and risk management in the private sector for more than twenty years. Obviously my experience with the very complex gov- ernment sector is limited in comparison. In my new role I have found it necessary to review my positions and thinking in some respects. We are all familiar with the risk exposures in a commercial operation and how we pro- fessionally handle the situation, what the tools are and so on. That is the end of the road. But what does the beginning look like? From where do we start? What are we going to see down the road? Can we in any way prepare for such a project? My own experience is drawn more or less from two points, the starting point and the end point. I have not been involved in an actual transformation process of the kind within an organisation. I will give my own comments on the scope and role of risk management in the goverment sector as such versus the private sector. I will touch upon the general development of risk management as I see it. I will of course account for the background, the development and the solutions now being implemented in Sweden. As we will see the kinship with a captive program is apparent. There are some interesting variations due to the government sector environment and application, though. What is indeed risk management in the government sector? How should the scope be defined? How can the range from national defence to national museums be handled? We are just in the beginning. I will relate my discussion to three main areas. These are risk exposure risk perception and attitude risk management In doing so I will also relate to the conven- tional steps in risk management as a basis, 3 workable concept. A remark I would like to make here, is that according to the definitions, a dynamic risk was explained to hold two possible outcomes, gain or loss, whilst a static risk could only bring loss. This theory obviously stems from the insurance world where a company fre- quently is, or was, called a risk. I would say that a risk by universal understanding is some- thing negative and that alone. By the same token something that brings the two possible outcomes, gain or loss, is called a business enterprise. The dynamic risk is in my opinion the outcome that the enterprise will not be commercially successful. Statisticians, however, may define risk as the possible variation from the mean. This would include a positive deviation also. I do not agree with that. In risk there is an inherent negative meaning. However, I agree totally with the view that uncertainty, unpredictabil- ity, as described by the variance, is the risk. But there is a material difference between genuine statistical curve analysis and busi- ness. In business, positive variation is not a risk. Anyway, the philosophical discussion about risk will go on. 2.2 The winds of change The world has developed a lot since risk management was first launched. The world has become more complex. The term risk management is today used more and more frequently by e.g. investment portfolio man- agers of financial institutions. The financial risks were however not included in traditional risk management. At least many of its practi- tioners were not dealing with them. Financial management has grown more important and requires its own specialists. For a number of years financial income has been even more significant than that of the core business in some industries. We have seen the legal environment change. There is a wide-spread and growing concern and awareness regarding conservation of the natural environment. Product liability and environmental impairment liability have strongly influenced the risk situation for the business community and the society. The industrial processes have changed. The rapid progress in the area of information tech- nology has brought new opportunities and new risks. It has also brought a large number of new professional careers, in our field, among others, the IT-security specialists and officers. That is however not the only new position that has created competition on the scene for the risk manager. There are the already exist- ing positions in safety and security. Then there are new areas relative to process and transportation hazards, various EC safety di- rectives and last but not least quality manage- ment. The role of the risk manager has always been a topic. In my opinion that is now more relevant than ever before. It is however, in my opinion, not so much the title and the role of a position called the risk manager that is of interest here, it is rather the scope and focus of risk management. 2.3 Aggravated risk in industry and commerce The ultimate risk for a company is to be eliminated from the market. For the risk man- ager in industry and commerce the conse- quential losses, the dynamic effects of static risks have always provided an important set of arguments for loss prevention. Now the market-place is still more com- plex and quick in reaction. Internationalisa- tion, deregulation and fierce competition make companies even more vulnerable. Product life-cycles and lead-times are shorter. Sub- contractors are selected on a quality and reli- ability basis. Concentration may even lead to just one supplier, where companies yesterday for back-up reasons by policy had at least two. But goods can be ordered and delivered from some other place on earth today just as quickly 4 closely related and should not be treated in different chambers. The tactic and operative measures should be handled further out in the organisation according to the policies and guidelines issued by management. A well known illustration of the goal of risk management is achieving the minimum cost of risk. As may be concluded from the preced- ing discussion, the total cost of risk is in practice impossible to calculate. The cost of risk concept is per se limited in its practical application. That will be further discussed later on. I would however like to venture to suggest a somewhat different angle to the goal. It should rather be to help to optimize the organisation´s value output. The value we are talking about here, is the buyer’s per- ceived value of the output. The minimum cost of risk does not necessarily coincide with this optimum. This ambition would definitely be more in line with top management strategy and mar- ket positioning. Risk management needs in my opinion a more dynamic image and less of what too often is perceived as boring and obstructive protection measures. as from the local region yesterday. Long- standing relations are tried. Customer loyalty tends to diminish. 2.4 Future strategic management function In other words, the dynamic effects of acci- dents are more important than ever to man- age, and that argument for risk control and loss prevention is more valid than ever. It may however not be the traditionally tailored risk management or risk manager that will be needed in the future. To some extent I think that risk managers have withdrawn, more or less voluntarily, to the insurance area. Risk managers, I sometimes fear, may also not have developed professionally into the areas of strategic management and marketing or business administration. An area, now rapid- ly developing, of which risk management is an integral part, is quality management. Risk management should, in my view, for the future be designed as a top management strategic function, closely related to financial and market perspectives and comprising both “static“ and “dynamic“ risk aspects. They are MAXIMUM VALUE OUTPUT protec- tion curve output value curveloss curve cost of protection/ loss/ output value volume of protection 5 In this context the risk manager has to get familiar with the business process and identify the crucial elements. These are then to be protected in a proper way. I am sure that support for such ideas will be welcomed from the business area managers in charge. It may even be that the most financially heavy in- vestment in a production facility, that nor- mally would receive attention from risk man- agers and insurers, is not the key competitive advantage in a company. It is perhaps just a necessary piece of equipment for manufac- turing the hardware that is part of the total delivery to the buyer. It could just as well be that the servicing organisation or another quality is the success factor. This is of course especially true in a service company, where the hardware contents of the product is nil. 2.5 A holistic view Again, quality, occupational safety and health and all aspects of risk should be integrated and managed together. I will attempt to illus- trate below how I think that the spectrum of “non-desired production“ and wastage in a wide sense is composed. What we tend to look at as accidents and risks is one end of the quality aspect. Accidents are in most cases man-made in one meaning or other. Such accidents are the extreme of inferior quality. Constantly we have to pay for inefficiency or wastage as operating expenditure. It is all budgeted for. As long as we meet our budgets everything seems all-right. We must however wake up and remember to approach loss pre- vention from that side. New developments as BPR, business process re-engineering, certi- fication in accordance with ISO quality stand- ards and so forth, offer an opportunity to address these issues. That opportunity should not be neglected by the risk management community. Low and high risk in this context refers to high and low predictability, i.e. low risk means high predictability and vice versa. The horizontal arrows approximately indi- cate how an organisation often is structured and focused. Obviously there are both gaps and overlaps in the areas where the arrows meet. From an organisational and efficiency point of view I am not certain whether to prefer gaps or overlaps. In all experience both situations lead to frictions and waste of ener- gy within an organisation. Therefore risk ineffi- ciency inter- rup- tions low risk (small events) high risk (severe events) mgtwast- age OPERATIONAL EXPENSE/ BUDGET RISK FINANCING QUALITY ”FORTUITOUS” RISKS de- fects/ errors ? SAFETY medium risk SECURITY HOLISTIC RISK MANAGEMENT? PRODUCTION RISK MANAGEMENT 6 basis. That basis is in my opinion growing. One of the challenges in government sector risk management is to create the awareness that the difference is smaller than is generally perceived. In the government sector there has by and large been no positions as risk manager, pos- sibly because government agencies did not buy insurance. That is however not to say that risk control has been non-existent. In effect, in some agencies the situation has been very much to the contrary from a technical view- point. But the financial aspect has been very much neglected. I will revert to this. 3 Risk chart 3.1 Introductory remarks Under this section I will make my personal attempt to position the phases in a transition from the budget sector to the private sector in a risk chart, in order to illustrate their relative situations regarding risk exposure and risk management. It may not at all present any new knowledge in the area. A number of aspects and dimensions is jammed into the chart. Therefore they have to be explained in some detail even if most of it should be rather self-evident. Firstly I will account for the risk chart. Then I will mark the positions. In the process I will make a short over-view of some aspects and developing factors that I find essential in order for me to arrive at that positioning. 3.2 X-axis The chart (next page) illustrates the various risk areas or types of risk that any company or organisation is exposed to. It does not specific- ally state whether a risk is static or dynamic, but it is obvious that from the left to the right at the bottom we move from static to dynamic. Resources and activities are components in the accomplishment of an organisation´s mis- sion. All risks, accidents and events, direct or management as well as other aspects of man- agement must be clearly defined in terms of responsibility. Risk management in practice should con- tribute to develop a more totally efficient organisation. In practice that has happened in some cases where we have been involved in risk analysis projects. The arrows also illustrate my earlier obser- vations about how many players there now are in the field of risk management in a wide sense. A conclusion, that should not be too far-fetched, is that a steering committee or something similar on top management level must take better control of risk management as a part of general management. The rewards are in risk awareness, strate- gic protection, quality, cost control effi- ciency and subsequent efficacy and com- petitive advantage. Another question indicated in this picture is how top management manages their own quality and their own risks. 2.6 The government sector If the risk situation in fact has been exacer- bated in the private sector, what is then the situation in the government sector? No doubt there is a similar development. There are some obvious fundamental differ- ences, though. The government sector is not under competition. There is usually no profit requirement. On the other hand there is an increasing demand for cost-effectiveness, and resources are generally shrinking. At the same time public demand for services is increasing. Media tend to sharpen the way the govern- ment sector is watched. One of the risk areas, that no doubt is becoming more significant, is liability. Legislation within the EC and in member countries will continue to work in this direction. The issues and the assets at risk are there- fore not the same, but there is a common 7 indirect, related to resources and activities are designated activity exposures. This is also the area where traditional insurance exists. For- tuitous risks may be another designation. A different remark is however that not many risks are strictly fortuitous. In most cases the human factor is the cause. In many cases it is also true that the ultimate responsibility could be brought back to management. In this era of quality management, it could even be possi- ble to identify accidents and losses as sheer lack of quality and waste of resources. Inte- gration between RM and TQM is already happening. I am strongly advocating that approach as far as various aspects of safety, security and process reliability are concerned. In Sweden quite a number of agencies have introduced or implemented TQM. The fact that Sweden, has become a member of the EC, has no doubt been a driving force for that development. Agencies operating in the areas of e.g. drugs and food are now under compe- tition in the sense that a company may obtain an approval or a certificate from any such an agency within the community. That means that price and quality now are issues for the future position of an agency. When an agency embarks on a quality improvement project, they can be expected to do so with ambition and with perfection as a goal. That lies within the culture of an agency. The same has hap- pened with market orientation in agencies that have been more or less forced to expand their volume of trading. Legal risk is intended to designate all as- pects of legal exposure, following an activity exposure event or not. Financial risk designates all risks related to pure financial activities and other pure finan- cial losses. Intangible assets are becoming more and more essential in today´s business and media financial intan- gibles hidden ECONOMIC RISKS strategic commer- cial direct con loss -extra exp -lop traditional insurance area -currency -political -investment r i s k operational expense range low severity high freq. high severity low freq. ”legislative, regulative, policy-based framework area” quali- tative risk asses- ment sce- na- rios visible art of loss legal liability activity exposures < f o r t u i t o u s > < q u a l i t y > indirect 8 world. Enormous sums are invested in brands, images, logotypes and the like. Public confi- dence is an extremely volatile asset, both in the public and private sectors. Commercial risks are all aspects of an organisation´s failure to execute its mission on the market. 3.3 X-Y-area Commercially insurable risk has been illus- trated as an area. Insurance is an important dynamic factor in business and in private life. Many large projects would not have been realised but for insurance. For most people insurance is necessary when buying a home. Without insurance the other player, the bank, would not extend its funds for a loan. It is however obvious how many areas and as- pects of risk that are outside the insurance area. This representation is merely qualitative and the areas do not represent their relative importance. However, it reflects how the total risk spectrum has developed with emerging significance of certain risks and how relative- ly small the insurable area is. Insurance prod- ucts have not developed much beyond the traditional products. This also reflects the need for organisations to come up with new solutions of their own. Structured risk smooth- ing programs and captives are obviously al- ternatives that should be attractive. It may also serve as an inspiration for the insurance industry to develop their range of products and services. Offensive competitors in the banqassurance sector seem to be the banks rather than the insurers. Another area is marked for legislation, reg- ulations and policies. The characteristic of this area is that activities and measures here are generally not evaluated and decided against cost-benefit risk management criteria. Re- quirements have to be met irrespective of any such considerations within the individual or- ganisation. Cost-efficiency must however be accomplished in the implementation process. Failure to meet requirements could entail legal liability, fines, problems in the area of industrial relations or the loss of confidence in the eyes of the public or of the market. In government agencies it is frequently laid down by regulation that a security officer must be appointed. That is of course always the case if the agency is involved in national security or defence matters. 3.4 Y-axis Along the y-axis has been marked the com- bined aspect of degree of transparency, time effect and strategic importance of an event. The “tip of the iceberg“- metaphor has been frequently used for accidents. The normal accounting and reporting in an organisation does not generally capture the effects of an accident. This is the more obvious in an agency where assets have not been recorded but written off directly in a cash reporting system. The argument for a separate loss reporting system including all ramifications is as valid as ever. It is unfortunately not very practical to run a permanent system of that kind. It is generally not even possible to capture all data theoretically to be included in a “cost of risk“ concept. I will revert to that. I have discussed the matter with the Auditor-General concern- ing government agencies, and they will not support such a system. I agree. I find it better to make an in-depth follow-up on any major event as you go along. Outside the diagram is indicated where visible costs are located in the realm of high frequency-low severity. In this area normal budgeting procedures and cost-benefit calcu- lations are applicable. This is an area of oper- ational expense and consequently no risk in financial terms. As severity increases and frequency dimin- ishes we move up into the range of hidden and strategic losses. Here we have a non-budget- 9 able situation and consequently a risk in fi- nancial terms. Statistics have no credibility and we turn to more qualitative methods, in the end using risk scenarios. 3.5 An illustration Two conventional losses are illustrated in the chart to explain further how it can be used. Again, this is not new. It is just about how to demonstrate in terms that hopefully points out the value of risk management as a strate- gic function. An accident resulting in the loss of a leading scientist or a fire in a unique prototype may eventually entail strategic loss, i.e. the same type of loss as commercial risks. 4 The privatisation process 4.1 Phases The road to privatisation could be considered to consist of a number of different phases. I have identified the following. For each one I will give at least one illustrative example from Sweden, which hopefully could be rel- evant also for other countries. These phases are in turn: 1.Pure agency (exercise of authority). Example: Internal Revenue Service. 2.Agency involved in trading activities. Example: National Bureau of Statistics. 3.Trading agency in a market under deregulation. Example: Swedish Rail. 4.State owned enterprise (SOE, a wholly owned separate legal entity) in a market under deregulation or in a free market. Example: Swedish Electricity Generating Board, Telia (former Swedish Telecom), Swedish Institute for Testing Materials. 5.Privatised company. Example: Celsius, Bure (ven cap), Nord bank. Although these phases seem to be very dis- tinct, the transition from one phase to another could very well be gradual and also represent a mixed situation. ECONOMIC RISKS fire unique eqpmt loss of earnings loss of market • ? accident lead scientist ? ? ? activity exposures visible hidden strategic human resources property direct indirect liability conseqential loss 10 tors that must not be overlooked. There are in some respects very distinct differences. For a budget sector entity it is generally not relevant to speak of a business concept but rather a mission or an instruction giving the tasks or duties. Agencies operate on an assignment by the government. The majority of the agencies will still be budget funded by appropria- tions. A private company is directly dependent on its customers for its funding. An agency is not. An agency cannot go bankrupt. A private company is a profit-oriented or- ganisation. An agency is not. A private company is directly dependent on its image or market perception. An agency is fundamentally not. Mismanagement may lead to the replacement of the director gen- eral. This has been known for a fact quite a number times quite recently. The duties of the agency could of course in an extreme case be transferred to another agency. But basically, the duties would have to be per- formed by someone, and the job for the individual employee would typically not be endangered. That difference by and large creates a signif- icant variation in culture and attitude. Al- though efforts have been made in some agen- cies to change this situation, in my opinion it can never be totally eliminated, due to the nature of the activities. These differences also explain a variation in risk perception, risk exposure and risk management between public and private or- ganisations. Agencies involved in commercial activities During the office of the former non-socialist government, agencies were encouraged to embark on various trading activities, proba- bly for ideological reasons. One reason may 4.2 Discussion Pure agency reform A management reform concerning adminis- tration and financial management in an agency could be considered to be analogous to the first step in a development towards privatisa- tion. However, for a general agency the de- velopment will not be expected to go any further than that. The following discussion will form a basis for more detailed analyses of evolving differences in risk and risk manage- ment moving from public to private. Such a discussion should emphasize a market and process oriented perspective, but also include a cultural perspective. A management reform of the kind would include the following components and steps. The extent and application of these factors will vary with the size and task of the agency. Agencies are made totally responsible for their own operations. Means and resources belonging to the state and necessary for their duties are allocated to the individual agencies. The agencies operate more on an outcome basis than on an input basis. Strategic planning, budgets, profit and loss accounts and balance sheets to be required and produced. Asset records will be pro- duced. Assets will be evaluated and depre- ciated according to generally accepted rules instead of being directly written off. Cash reporting is changed into accrual ac- counting. Agencies are to take up loans for develop- ment projects and to pay back with interest. Buyer-seller processes will be introduced. Customer focus and service management will be introduced. Various management - isms and development trends from the pri- vate sector will be looked upon with great interest. Agencies are to be expected to or wishing to operate as close as possible to the private sector. There are however some differentiating fac- 11 have been to reduce the budget volume, the need for appropriations or to allow cut-downs. That policy has resulted in a number of trad- ing activities among agencies. For instance, some museums have indicated that it is necessary for them to arrange various exhibitions in order to raise funds for other high-interest projects. Some agencies are dependent on fees for most of its income without having the status of a trading agency. In many cases the commercial activities are not kept apart from the regulatory activities in terms of reporting and responsibility. Projects are accepted without written contracts. There is often no limitation of liability. The practice of non-insurance has led to a ”happy go lucky”-attitude. There is a danger in the atti- tude that agencies can go along with almost anything. The state does not want to insure and so be it. The state will cover it all. That could even go as far as civil contract works. In other cases separate areas of responsibil- ity have been formed but the financial back- ing still resting with the ”parent agency” since there is only one legal entity, which ultimately is the state. My impression is, that the enthusiasm over the new opportunities however has been greater than the professionalism in the respect of doing business and coping with risks. There is no doubt, though, about the professional- ism in the subject matter in which agencies are involved. The present government is probably less encouraging in this respect and will in any case want a clear line of demarcation between the duties and responsibilities of the agency and its trading activities. The new internal risk financing option for agencies, as described later on, will clearly contribute to a commercial and professional development. An agency will have to consider various aspects of their commercial deci- sions, among others the risk exposure and the cost of financial protection. The risk survey projects we have been involved in have been very well received among agency in-house counsellors, who for years on end have tried to establish manu- script contracts for various trading activities. When the risk survey has pointed out the risk inherent in performing projects without ap- propriate contracts, they have felt a positive support which hopefully will lead to substan- tial improvements in their practices. 4.3 Organisation culture Are there any significant differences between the private sector and the public sector? It may seem more than obvious that the obser- vations I will offer here are my own personal ones and that my experience is very limited. I have already indicated some factors that are relative to the commercial and profit- oriented angle in the private sector. Another set of factors are relative to the regulations-oriented situation in agencies. Agencies are generally operating in accord- ance with the constitutional and legal frame- work. Each agency has its own letter of in- struction by resolution of the government. That means, for one, that once a task is included in the regulations, then it is promptly address- ed. Before that it is more of a pending situa- tion. Sometimes this means that there is a lack of initiative to get started. That is however not to say that development and improved effi- ciency are not encouraged or demanded. In a basic regulation governing all agencies it is even expressly stated that it is the duty of every agency to enhance efficiency, review procedures and initiate improvements. In some technical respects I have found that agencies, on the contrary, are better equipped than their private counterparts. That could be said for utilization of IT, keeping up to date with new releases of both software and hard- ware. In the area of security you will find a lot of the latest gadgetry in agencies. 12 The pride and traditions of being a civil servant are very much alive. The hallmark of the Swedish civil servant is the untouchable, impartial and just treatment of the citizen and the matters at hand in every respect. That really deserves to be admired and to be con- tinued. The draw-back is that the process might be rather time-consuming in the ambi- tion to do everything correctly. There is gen- erally an ambition to perform in a thorough and professional manner. Once a regulation is effective or a decision has been taken the matter is carefully and loyally pursued. That also goes for risk management projects. Most agencies perform types of activities that are likely to be found within the profes- sional services industry. The pulse in the commercial stream of business is however very seldom felt. Although comparisons be- tween customers and a service partner on the one hand and individuals approaching an agency on the other might be totally relevant in some cases, I dare say, it is mostly not. In some circumstances these comparisons may even be considered to be misplaced. Employ- ees in agencies might even find them ques- tionable or even ridiculous. Generally it is the management that tries to implement certain concepts from the private sector without prop- er adaptation. Nonetheless the job situation in agencies might be extremely stressful. Cut-backs in staff and changes in the regulations create balances and complaints. The stress could also arise from the burden of important deci- sions under the exercise of authority concern- ing various issues for enterprises or the tax- payer. 4.4 Risk financing Terminology In my view, risk financing connotes the trans- fer or sharing of risk with an outside party. The financial planning or budgeting of losses predictable within reasonable limits, I refer to as loss funding. For a state both these concepts are relevant depending on the out-look. For the state re- taining aggregate losses it could strictly mean a cash flow situation, pay-as-you-go. It could also mean loss funding, if there is an element of budgeting. The term self-insurance should moreover be reserved for a structured system, not just a non-insurance practice on a pay-as- you-go basis. For the individual agency it is a situation of risk financing if the state by agreement ac- cepts to indemnify the agency for certain classes of losses. From no insurance to all insurance? Risk financing is an entirely new concept in an organisation that has been formally pro- hibited to purchase insurance. Privatisation in such a case could probably mean going from no insurance to full insurance. No doubt there are some possible pit-falls in such a process. For one thing, while being an agency under a management reform im- plying full responsibility for losses, insurance may have been perceived as a panacea. The risk is that the privatised organisation will rather insure too much, once it is allowed. When first introducing risk management and insurance matters the novelty of such arrangements is obvious. It could be said for some private companies, at least it was true earlier, that purchasers of insurance had little experience and based their programs on their experience as private individuals. I am sure that the overall professionalism has improved significantly in the private sector. In a newly privatised public organisation that learning period has just begun. Insurance companies and brokers could also be seen to represent the real business world in a way that could be tempting in itself. The privatised organisation might be keen on insurance as the missing tool for the full- 13 fledged commercial operator. Some agencies were also very creative and bought policies that basically were not allowed. I have come across a number of these. At least in Sweden, the insurance industry looked at the government sector with great expectations, when it was made known that the non-insurance policy might be lifted. The well-known marketing activities were well under way. Insurance is an industry with a tradition of spending, forgive me for saying so. Civil servants mostly are very sensitive about such activities, since they would want to keep the integrity intact. I find it essential that an internal risk financ- ing program identifies the special circum- stances and needs involved. To copy the commercially available solutions is a quick and practical way. Such a concept should however be carefully evaluated. Especially the administration and collection of informa- tion is an area where procedures should be cut to a minimum. Wordings and provisions should be kept simple. Some of the shortest and broadest wordings I have ever come across originate from some Australian states. In very few words is stated that the agreement covers damage to or loss of property plus liability. That’s it! Introduction of risk management It is of course very important that the true risk management focus is not lost in the transition. If risk management was not among those good management practices in use before the privatisation, the more important it is to intro- duce it in the transition process, but to intro- duce it in a proper way. It is however my experience that agencies are very quick to appreciate and apply risk management principles and not overly de- mand the internal risk financing solution of- fered. A well balanced loss retention level should be fixed in any organisation. A holistic view on all risks and the awareness of the limited scope of commercial insurance should be established. A captive company is by far the most so- phisticated tool available for both administra- tive and financial control of risk. A captive could also be used for other risk financing or loss funding purposes than those commonly related to commercially insurable risks. However the privatised organisation could of course be just as well off with a less sophisticated tool meeting the same criteria, e.g. a captive account, dummy captive, asso- ciation captive or rent-a-captive. The forma- tion of a captive involves capital investment and it is not self-evident that it is the best way of utilizing capital resources. 4.5 On the principle of non-insurance Most states around the world adhere to a principle of non-insurance. States are ”long term operations” and are generally financial- ly strong in comparison with private enter- prises. States have the prerogative to collect taxes. The spread of risks is normally excel- lent. Consequently there is no need to seek external financial support e.g. in the commer- cial insurance markets for such risks and to spend the tax-payer´s money on insurance premiums. In actuarial terms there is no financial risk, because the state can afford to absorb any variations in the outcome and operates long- term at an expected value level. States can afford to apply the pay-as-you-go practice. There are of course exceptions to the rule. One of them is New Zealand. The explanation for this is however not based on the discovery of any new actuarial principles but rather very practical solutions in a period of extensive reform work. I have also learned that some Australian states are very advanced in the implementation of risk management. They 14 are very explicit in their non-insurance poli- cies. Some of them have nonetheless actually decided to buy natural catastrophe cover for their own property on a high excess and stop- loss level. 4.6 Cost of risk In the context of risk management the cost of risk is a concept that is often discussed. It is a theoretical concept that is easy to compre- hend. It is put forward as a tool for optimizing risk management. In practice it is in most cases not practicable because expense items are not at all reported in a way consistent with the implied cost of risk reporting requirements. That fact can not be expected to change. The second problem is that expenditure and loss can not be isolated along those lines. Prevention expenditure is often integrated in other functions. The situ- ation is also depicted in the risk chart in the visible, hidden and strategic cost levels. Another problem is that cost is an accrued expense and should be expected to appear in the P&L account. Most of the risk exposures are not costs but more of balance sheet items, much like liabilities or guarantees. A risk that is transferred to an external party for a premi- um however appears as a cost item. In order to estimate a ”cost”, the risk expo- sures would have to be phased out over a period of time, the length of which is not known. The ”cost” would vary largely with the anticipated number of years, but probably still be significant. For a commercial opera- tion it would not be feasible to set aside funds to meet that ”cost”. Hence the existence of insurance. But for a state it could be enough to estimate the exposures in order to prepare and prevent. But it would still not be neces- sary to set aside funds for the worst case. In practice the cost of risk concept is appli- cable only in the area of insured risk and only in a limited way because all the relevant costs low risk high risk RISKTRANSFER/ RISK FINANCING = operational expense (budget) COST OF RISK IN ”P&L ACCOUNT” (hidden & other costs excl.) losses + premium + (cost) deductible + (est. cost) exposure (no cost) RETAINED RISK (act. cost) = cost of risk= ??? (no cost, no credibility, balance sheet item) protection + (cost) high severity low severity protection + (cost) protection + (cost) Swedish National Judicial Board for Public Lands and Funds 15 cannot be included. It could be perceived as an indicator in this area. Comparisons be- tween different companies are however near- ly impossible because of the variations in programs and of the insurance market and the offers from its individual players. The cost of risk concept is consequently limited to an area where risk exposures can be transformed into expenses for a price. In practice, in the private sector this is only possible for insurable risks. The next issue is then, what should be included in a cost of risk reporting system? To go beyond what is cap- tured within an insurance scheme is generally very difficult to keep alive. It is also hard to get. In the government sector the problems are the same. There is however a possibility to broaden the risk financing program in com- parison with insurance. But an extensive re- porting scheme is not practicable. Moreover there is an obligation for an agency that proposes new administrative procedures to make an assessment of the consequences in terms of cost and benefit. My judgment is that comprehensive cost of risk reporting is diffi- cult to advocate because of the extra cost involved compared to what could be achieved. 4.7 Risk positioning Based on the above discussion, I have marked in the risk chart, where different risks and positions relative to the evolutionary chain of privatisation will appear. These positions are only my own judgments. The private company (P) acting in an free market is obviously exposed to the full range of risks in all dimensions. The pure agency (A) has a limited exposure due to the parameters of its activities. There is of course no commercial risk at all. The trading agency (TA) is located in be- tween. In terms of intangibles there is no doubt that financial intan- gibles hidden RISK POSITIONING strategic commer- cial visible art of loss legalactivity-based f o r t u i t o u s PPPP A A A A A TATATA TATA TA Swedish National Judicial Board for Public Lands and Funds A=pure agency TA=trading agency P=private enterprise free market 16 the confidence in an agency, be it from the public, media or from the government level is of utmost importance for its future. Above all media attention may lead to loss of confi- dence from the other two. The risk may not be ultimate for the agency but could very well be for its executives. In a way this aspect is comparable with commercial risk for a pri- vate company. If the company does not con- duct its business very well qualitywise or in an appropriate manner the customers will turn elsewhere. The financial risk is very limited. Within the general category of financial risk I have also included political risks. For an agency there is the very special political risk that it will be discontinued due to a change in policy or administration. That aspect is not illustrated here. The direct legal risk as defined here is in practice extremely small for an agency. General liability stemming from the owner- ship of property, buildings and real estate is not very significant. In terms of risks relative to human resourc- es, property, equipment and other means of carrying out the daily business, agencies are generally less exposed than the average enter- prise. Most agencies are pure office occupan- cies in rented premises. There are some exceptions. E.g. the IRS is extremely dependent on its computers. Other agencies with specific and different risks are e.g. the Coast Guard and the Police. The government involvement extends to some unexpected areas like camps, vehicles and equipment in the Antarctic regions, space research rockets launched from bases in the North of Sweden or an ice-breaker bound for the North Pole in a forthcoming summer expedition. Trading agencies tend to fall between the pure agencies and private companies. Some trading agencies are of course impressive operations with the same risk exposures and profiles as private industrial enterprises. In this category we will find e.g. the Civil Avi- ation Authority and the Swedish Rail. Some severe liability risks are to be found here. 5 Risk management in The Swedish government sector 5.1 Starting point Towards the late 80´s and early 90´s, a man- agement reform was started in the Swedish government sector. The reform brought with it new practices in terms of administration and economic control. Departments and agen- cies were to be operated in more close resem- blance with the private sector. They were to be operated as economic units and in some cases also like profit centres. The last model was of course primarily to be applied on trading agencies. Accrual accounting was introduced. Profit and loss accounts and bal- ance sheets were to be prepared. Especially the task of establishing a bal- ance sheet was not very easy, since the asset values had not been recorded. Investments and purchases had been written off directly against the budget. Consequently a valuation project had to be performed. The project had to be limited to only a fraction of existing assets, namely those with a purchase value in excess of approx. 900 GBP and with a re- maining life of at least three years. One can imagine that a traditional insurance quotation would take some time and effort to go through considering the volume and scatter of pre- mises and property. The planning process was changed. Budget appropriations were transformed to a blanket basis within limits rather than the previous detailed model. Budget surplus or deficit was allowed to be carried forward. A consequence was however that the strong balance sheet of the state now became frag- mented. Agencies had the responsibility for 17 the funds ands and resources allocated to them, although these technically still belonged to the state. Now the heads of departments and agencies started to worry about risks. What was to happen if an agency lost its property in an accident? The state had always upheld a policy of non-insurance. If a loss occurred that could not be financed by the agency itself, the normal procedure had been to approach the government for a supplementary allocation. The government had, and still has, by provi- sion from the parliament, a special account at its disposal for unexpected expense. The number of cases had been low, though. Either the actual accident rate was low or did the budgets allow losses to be absorbed rather comfortably. Anyway, no extremely large losses had occurred. The major losses that had occurred over the years were usually related to the railway or the defence forces. However, by and large there was no report- ing of losses within the agencies themselves. The cost was consequently unknown, both on an aggregate level and on an individual level. 5.2 Course of development Initial study Due to the effects of the management reform and the demands from some agency execu- tives the government initiated a study in order to clarify the situation. The study was submit- ted in 1992. A number of measures were proposed but not immediately implemented. However the government put forward a pro- posal for new guide-lines in 1993. Adopted strategy In the new guide-lines carried by the parlia- ment in early 1993, the fundamental new statement of principle was that all losses were to be borne by the agencies themselves. It was then foreseen, or at least hoped for, that the non-insurance practice could possibly be lifted by July 1, 1994. There had been demand and expectation in this direction from some large trading agencies. It was also pointed out that, although losses had been few, there was some room for im- proved risk awareness and loss prevention and that risk exposures and losses were large- ly unreported and unknown. These observa- tions were based on findings of the initial study. However, it was also stated that insu- rance should be avoided at length and that risk analysis and loss prevention should be given priority. At the same time The Swedish National Judicial Board for Public Lands, hereinafter the SN, was assigned its task of developing a new risk management framework. The as- signment covered four different areas: general recommendations for risk manage- ment support for individual agencies development of a risk financing system purchasing of common insurance arrange- ments if needed The steps and time-table of the project are displayed in condensed form in the following table. Activity Date Inception July -93 Guidelines carried by parliament Introductory phase Aug -93 Questionnaire ” Questionnaire Jan -94 Preliminary study April -94 Assignment on further investigations May -94 Government decision on risk analysis in agencies June -94 Recommendations by the SN June -94 Implementation seminars Sep-Dec-94 Feasibility study report Nov -94 Period for risk analysis Jan -95 Final report Febr, 1995 18 Questionnaire ” Result and feed-back report on risk analysis April -95 Decision by parliament April -95 Implementation project May -95 “Market research“ Sep -95 Basic EDP-project Oct -95 Final design phase Regulations in preparation Regulations effective Jan -96 Introduction seminars Jan-Feb -96 Agreement no. 1 Feb -95 Claim no.1 Feb -95 Further development in close cooperation with agencies Scope of risk management In a private company the scope of risk man- agement should really not be restricted. But how should risk management be defined in the government sector? A number of agencies do nothing but man- age risks on a collective and societal basis. Among these we will find the Road Transit Authority, the Nuclear Inspectorate or the Food and Drugs Administration Agency. For example, the way our roads and intersections are designed affects us every day, not to mention the intersections between road and railroad. Already at an early stage, suggested by the SN and included in the general guide-lines, risk management was defined to encompass mainly the internal perspective of an agency, i. e. its own resources and means. That is comparable to the perspective a private com- pany would have. The connecting risk be- tween the internal and external perspective is inherent in the liability an agency would have for its actions, as shown above. Introductory phase The government decree (see time table p.17- 18) on risk analysis meant that agencies had to prepare a study in about 8 months time. In order for the agencies to get focused on the project, the general guidelines directed the work towards serious risk. Serious risk was defined as those scenarios resulting in the agency being unable to perform its functions even after cut-downs and reorientation of its activities. The emphasis was put on management is- sues and organisational aspects. Support for the process was produced as a set of questions, a risk review questionnaire or ”risk diagnosis”, based on knowledge of vulnerabilities and weaknesses expected to be present in various organisations. A scenar- io method was also developed, mainly based on past experience from the private sector. In cases when we have worked together a73 300 authorities and agencies a73 260 000 employees a73 army, police, justice, customs, coast guard, aviation, railroads, roads and bridges, tax, social benefits, education, national heritage............... SCOPE AND SPECTRUM ”From JAS to Picasso” ”What has happened can happen!” RISK MANAGEMENT IN GOVERNMENT SECTOR SCOPE OF AGENCY RISK MANAGEMENT INTERNAL PERSPECTIVE AGENCY SOCIETY EXTERNAL PERSPECTIVE CASUALTY AS CROSS- BORDER RISK AGENCY RISK MANAGE- MENT 19 with agencies in order to support a risk sur- vey, we have found it very practical to start with the risk diagnosis questionnaire. The questionnaire lists a large number of ques- tions sub-divided into sections that pertain to areas that are known to lodge risks. Such areas are, not surprisingly, e.g. information security, fire safety, liability and risk manage- ment policies. These models had to be very general in order to fit everybody. The usual effect is that such ideal omnipotent models do not fit any- body. They have both been used with reason- able success, though. The results of the project were described as in the picture below. The goal is to create awareness throughout the organisation, pre- vent strategic losses, to achieve financial pro- tection for non-budgetable losses, develop standards and guide-lines for security and quality and finally to result in cost savings on an operational level. Risk profile Due to the non-insurance situation there are generally no risk profiles or risk descriptions. Consequently, that has to be remedied in the initial phase. It should require some prepara- tions though, since it cannot be expected that an agency immediately understands why and how. It is important that this survey expands beyond insurable risks. There are on the other hand some agencies that have safety as part of their mission. An example of this is the Swedish rail. The Swed- ish rail has its own risk manager since a number of years and operates also a captive. The engineers and operations people have always had as their goal to prevent accidents. Accidents that have occurred have been me- ticulously investigated and lessons learned have been transformed into new safety meas- ures. Cost has not been the issue here. As a result there is an abundance of infor- mation concerning technical matters. It is not until quite recently that the corresponding financial data have been gathered and ana- lysed. Accidents and losses have historically simply been perceived as operational expense and there has been no reporting. That is now rapidly changing. At least for the time being it does not seem likely that the Swedish rail will be transformed into a SOE. In practice though it has to operate like one in all re- spects. Loss reporting One of the underlying factors for the decision concerning risk management was the lacking quality in the reporting of losses. In fact there were no general procedures at all laid down for loss reporting. Losses had historically been perceived as part of the expenditure. The Auditor-General had not paid any interest to the matter. This is of course a different situation com- pared to the private sector, where at least due to insurance the insurable losses are recorded. Another issue is whether there is a specific reporting of losses in the accounting system. Generally there is not, neither in the govern- ment nor in the private sector. That is why the concept of the hidden losses has come up as a significant issue. In order to estimate the losses, we included an inquiry in the questionnaires. We asked for losses during the last three-year period. It was confirmed that the agencies in general did not report losses in a systematic way. One excep- tion is work related bodily injury. In such a case reporting is required by law. By and for SERIOUS RISKS (non-predictable high severity) awareness prevention financing ELIMINATION OF SUB-STANDARD PROTECTION AND QUALITY IMPROVED COST-EFFICIENCY AND QUALITY strategic level tactical level operational level 20 large the reporting was a matter of what could be remembered. The proposal for a risk management system The proposal finally submitted by the SN could easily be depicted as above. It implies taking into operation a system that will produce the data for risk manage- ment. The lack of reporting was one of the original problems. The risk financing sub- system will help to produce the necessary data for improved risk management. As can be understood the cost of risk reporting will be included only as far as provided by the financing scheme, at least for a start. In the picture is also indicated the close integration that preferably should be accom- plished between risk management, quality (Q) and occupational safety and health (OSHA). Management is responsible for all these processes. New regulations Since January 1, 1996 a number of regula- tions are in effect replacing former ones. The regulations can simply be visualized as follows. RISK MANAGEMENT RISK ANALYSIS (requirement) LOSS PREVENTION (requirement) RISK FINANCING (option) COST OF RISK REPORTING (requirement) THE FRAMEWORK The essence is in condensed form: Risk management for government agen- cies requires risk analysis, loss prevention and cost assessment to be performed and offers the option of a risk financing agree- ment with the SN. In the regulation is also established that the state retains its risks and that agencies cannot purchase commercial insurance. The SN is entrusted with the authority to allow exemp- tions from the non-insurance policy. ANALYSISCONTROL MANAGEMENT (Q,OSHA) FINANCING reporting services finance identification evaluation active programs monitoring feed-back INTEGRATED RISK MANAGEMENT SYSTEM 21 5.3 Risk financing scheme Risk financing in this context refers to the situation of the individual agency. For the state it is a loss funding or loss budgeting scheme. All agencies have the option to enter a risk financing agreement with the SN. The option is not open to other legal entities than the state, thus excluding organisations like SOEs or various foundations. The scheme and the pricing is characterised by the following elements. risk analysis required sum at risk selected by the agency, equal to maximum indemnity loss sensitive charges minimum deductible type of activities risk management standards technical protection level Entry requirements A number of agencies had shown an obvious interest in the risk financing option. That was well established in their responses to the questionnaires accounted for in the project table. In order not to lose the risk management strategy, it seemed important to build risk awareness and risk analysis into the system. We would try to avoid the ”don’t worry, it’s insured” syndrome. However, launching the system without unnecessary delay was a priority. Therefore it did not seem practical at this point to intro- duce general requirements or standards for risk analyses or loss prevention other than those given in the earlier recommendations in June 1994. Another problem would of course have been to evaluate reports and to visit all premises. The loss sensitive charges are expected, at least in part, to make unnecessary any exten- sive inspection activities. Again, it would have delayed the implementation. Moreover, the majority of agencies are low in risk pro- file. For those that are not, it is at least a limited effort to make a more detailed assess- ment. One could however demand at least maxi- mum losses to be estimated for each risk to be transferred to or shared with the SN. The agencies had been through their first exercise in risk analysis during late 1994 and early 1995. Maximum loss estimates were then introduced as a significant parameter in a possible internal insurance system. Therefore the ”EMLs” of the agencies were already there, at least as a first guess. In order to emphasize the management is- sues and to make the agencies participate in an active way for their cover, the selection of the risk sum rests entirely with themselves and the selected amount serves as the limit of indemnity. This way of placing the responsibility has actually not met any objections at all. Scope of cover The cover is divided into three sections, prop- erty damage, consequential loss and liability. Property damage The wording is intentionally very broad. It affords full all risks cover for all kinds of property anywhere in Sweden, stationary or in transit. Besides, only a few agencies have the responsibility for real estate and build- ings. The others rent their premises. Extended cover endorsements will be giv- en for other situations at additional charges. The cover includes property owned by the state and put to the disposal of an agency. It also includes property not owned by the state but under an agreement of indemnity, e.g. leased property or property taken into custody. A consequence of the wording is that there is full cover available also for money and financial instruments and for fidelity crimes. 22 Such a cover will require an extra endorse- ment and an audit for high limits. The only exclusion is wear and tear and other damages that are not sudden and acci- dental. This section covers the reinstatement of any damaged function by replacement or re- pair. The rationale behind this is that any agency performing its duties must be able to recover as quickly as possible after an acci- dent. It is not the property itself that is impor- tant but rather the function. The activities would have been restored anyway and the state would eventually have paid for it one way or another. In case of a small loss, it is only to be expected that the damaged property is re- paired. In case of a total loss the agency is free to use the risk sum as they find appropriate, e.g. by substituting a main-frame for a LAN with a client server. The risk amount is recommended to be calculated to cover the new replacement val- ue of the area expected to constitute the EML. Loss adjustment in the property area is expected to be rather light. The agency may in most cases after notification be entrusted with the restoration of the damaged functions. The claim could in most cases simply be paid against presented documents. Consequential loss Consequential loss is covered both for extra expense and for loss of profits under a word- ing which in practice in accordance with a normal market wording. The agency must select the risk sum and the time period. Casualty The wording covers liability. Contractual liability is covered under a special endorse- ment. The only exclusion under this section is liability arising out of exercise of authority. Right from the start it seemed obvious that liability arising out of actions by the police, customs etc. had to be excluded. Eventually all aspects of exercise of authority was ex- cluded. Minimum deductible The minimum deductible has been set in order to screen out single stolen or lost PCs. For a large agency that is a small amount, for a small agency it can be painful enough, especially if repeated. Rating structure Rating curves have been calculated on the basis of just two parameters, the total number of employees and the selected risk amount. The charge is thus calculated on “an EML basis“. This fact takes into consideration the spread of risk within an agency. It also makes unnecessary the total valuation exercise. In principle there are different curves for the different sections and extended endorse- ments. A special set of curves has been calculated for the charge discounts related to deducti- bles. The discounts have intentionally been set in order to create incentives for both higher deductibles and loss prevention meas- ures. That means that they are generally larger than seems to be the market practice. Loss sensitive charges The loss ratio is calculated against 70 % of the basic aggregate level of the annual charges for moving three year cycles. The portion will be reduced by increased deductibles. There is a symmetric variation applied of maximum 15%. After 8 years of no claims the charge bottoms out at 30 % of the initial level, given that all other factors remain the same. The 30% portion is kept aside for administration, claims handling, development projects and security allocations. There is no cap on increases of the charge in case of a continuous unfavourable loss situa- tion. 23 concept so far is a provisional arrangement to be evaluated after some few years it is a simple solution. For the SN it seemed a per- fectly feasible arrangement. Other factors of some significance are that as the arrangement now is operated it does not require a license from the finance inspector- ate and it does not fall under the legal frame- work pertaining to commercial insurance. The similarities between the proposed con- cept, now in operation, and a normal captive are obvious. The SN substitutes the direct insurer in its position. The SN is however not a normal insurer and has e.g. no self- interest in whether a certain agency has chosen to join the system or not. The charges paid are forwarded to the cap- tive account, which is the actual insurer. The role of the SN is that of the captive manager in a somewhat unorthodox combination with an underwriter. The SN on the other hand takes a definite interest in being perceived as a professional partner in risk management by other agencies and will try to extend the support and services demanded. Open and complete reporting and customer satisfaction are the only ways of achieving a top mark. A steering committee is under formation Large agencies will be underwritten indi- vidually. Other factors The remaining factors included in the ele- ments of the financing scheme will still have to be developed. They will be generated with- in the system. Agencies have been grouped together for obvious reasons, e.g. universi- ties. But as of now there are no data to use and we do not know if there are any diverse factors within the groups. One of the spin-offs of the financing scheme is to help to produce the knowledge and the actuarial figures. As a matter of fact, in one instance in Australia a similar program was launched with a nil de- ductible in order to lure out loss data. The captive account arrangement Structure and operation The combined purpose of loss funding for the state as a whole and risk financing for the individual agencies has been built into a cap- tive account arrangement. The main structure is shown below. The government included in its instruc- tions for the project that the risk financing system was to be arranged within the legal entity of the state and not in the form of a company as a separate legal entity. Since the agency level aggregate retention level per occurrence deductible per loss retention annual aggregate retention working layer FUNDING MECHANISM special appropriation credit lines with The National Debt Office 24 AGENCY A AGENCY B AGENCY C AGREEMENTS, CHARGES, SERVICES, INDEMNITIES CAPTIVE ACCOUNT NDO CREDIT LINE NDO GVT UNEXPEC- TED EXPENSE ACCOUNT RISK FINANCING FRAMEWORK IN-HOUSE RISK SMOOTHING ON A COST-PLUS BASIS NDO= National Debt Office Swedish National Judicial Board for Public Lands and Funds normal loss levels. If need arises, the Govern- ment can resolve to make use of a special authorization by the Parliament for unexpected purposes. In even more extreme cases a spe- cial approval by the Parliament would be needed. The arrangement is in parts similar to what is now being offered in the market as financial reinsurance or finite risk programs. Alternative arrangements What will eventually become the final ar- rangement is pending the evaluation. A dif- ferent approach has been proposed for a re- form concerning the organisation of workers´ compensation benefits in the public sector. The solution is in parallel with the develop- ment in the private sector, for which a former consortium now has been discontinued. The situation is quite different between these two areas. In the case of workers´ com- pensation there is a commitment which falls under collective agreements between employ- ers and employees. with its members recruited from a broad spec- trum of agencies. The committee is to be compared with the board of a company with the corresponding tasks. Credit line Instead of the normal captive retro-cession there is a credit line to the National Debt Office. The credit will serve as a smoothing vehicle if the liquidity of the account is ex- ceeded. Agencies will then collectively have to pay back in instalments together with an interest. Agencies will initially be grouped together in classes, gradually reviewed as data develops. If it becomes apparent that a level of charges is incorrect it will have to be adjusted. The credit line has to be approved by the Parliament. Considering the national debt situation in general there is a desire not to carry along unnecessary large liabilities or guarantees. There is no need for a credit line to absorb the most severe risks. The credit line is more to be considered as a cash facility for 25 also be lost and they would have problems in competition with other players if starting to write general business. Queensland, another Australian state, has introduced a premium-less system with com- pulsory retention levels per loss and in the aggregate depending on the size of the agency. The retention levels function in practice as stop-loss covers. The main reason for this concept was said to be to avoid costly admin- istration. A comment is that in the process something could be lost. The information inherent in the premium level and a loss- sensitive rating scheme would be missing and that fact might reduce the incentive for loss prevention. Conclusions So far it could be said that this whole concept has been well received. It has been introduced in a fairly short time. Sometimes it could take even longer to implement a captive program in an industrial group. It is perhaps not really fair to make that comparison. In this case the central decision-making in the form of regu- lations and the instruction-oriented culture have been highly facilitating factors. However, I dare say that the type of simple approach and the very small need for administration is sel- dom seen in the conventional market, which still tends to work traditionally. The initial study had investigated the non- insurance practices of other states, predomi- nantly in Europe but also in New Zealand. The policy of non-insurance was reported to be absolutely prevailing. The one exception was New Zealand, where agencies were in- structed to seek insurance in the commercial market on their own account. No explanation was reported in the study. Therefore one of the first steps was to further investigate the New Zealand situation more in detail. It then was established that the NZ government had found it necessary to do so in the course of an extremely rapid and radical management re- form when there had been little time for fine- tuning of various issues. In New South Wales in Australia there is an example of a state-owned captive that after some years was privatised. The captive was intended to be a vehicle for introducing risk management. After the initial phase it was decided that it could continue to operate on its own. In my experience it is however a risk that the true risk management focus will be lost and more of an insurance-orientation will follow, if a sound risk management has not been established and supported by generally accepted practices. There might also be some problems for a former state-captive to act in the open marketplace. Their own focus might