Mr. Speaker, one of the great puzzles surrounding Canada's social programs has been that they have had very little success in reducing the country's social ills. This is so in spite of the fact that expenditures per capita in real terms have increased dramatically since they were initiated during the post war years. What has gone wrong?
This question has occupied my academic research for some time. In the short time available to me today I want to share with you some of my insight on this matter. I do so because from these insights emerge policies appropriate for revisions of Canada's social programs during the present financial crisis and as we enter the 21st century.
It is part of folklore that restaurants that have fire insurance burn more often than those that do not. The insurance industry attributes this increase to what is known as moral hazard phenomenon. In simple language it is due to insurance induced changes in behaviour. The greater number of fires that occur in insured restaurants is due in part to outright arson. However, many additional fires are started because in insured restaurants less care is given to the cleaning of greasy vents and other fire hazards. They are less likely to have sprinkler systems and fire extinguishers.
The Government of Canada provides insurance against the hazards of unemployment, poverty and old age, disability, illness and a wide range of other calamities befalling people. As in the case of fire insurance these kinds of social insurance programs induce changes in the behaviour of the insured that result in higher claims and costs.
Let me illustrate in the context of unemployment insurance. Undoubtedly unemployment and the costs of insurance are raised by some criminal cheating through the filing of multiple claims or receiving benefits while holding a job.
Some claim that such illegal acts are rare while others think they are frequent. By the nature of the crime it is not easy to obtain reliable estimates. At any rate, clamping down on such cheating is not a contentious issue.
Of greatest concern for the present discussion of social insurance reform is the fact that much of the present high unemployment is due to individual Canadians reacting rationally to the changes in the environment in which they operate.
For example, persons with unemployment insurance benefits have higher standards on a new acceptable job than those without. The former will reject jobs with longer commutes, lower pay and higher retraining requirements than the latter.
People who behave in this way do not break the law. They act rationally in response to opportunities created by the government. They can afford to hold out longer for a better job because being unemployed costs less. Such legal and rational behaviour nevertheless increases the rate of unemployment and the cost of the unemployment insurance program.
Similar forces act on the welfare program, as was demonstrated when in the summer of 1993 in Ontario a woman created headlines with the revelation that she quit her job deliberately and went on welfare. She argued that by doing so she enjoyed a higher living standard than when she worked for $40,000 a year.
The media made much out of the question of whether this was true. In my view, this question misses the main point. Consider as a thought experiment that the woman's income was actually reduced by $6,000 by going on welfare. This means that full time work brought her only $500 a month and the loss of time with her family in leisure activities and valuable work in the home.
I know few people who condemn Canadians who make the choice this woman made. She was not violating any law. She was simply taking advantage of choices created by the system. Yet such behaviour swells the number of welfare recipients and raises the cost of welfare programs.
The cost of unemployment insurance is increased even further by the adoption by society of institutions which take explicit advantage of the opportunities offered by the system. Industries with seasonal employment expand. Some governments arrange hiring to accommodate eligibility requirements. They all act rationally and within the law. No one is to blame but the system which allows the creation and exploitation of these institutions.
How large is the cost of social insurance programs due to the insurance induced changes in behaviour and institutions? Social scientists do not have precise answers to these important questions. I have been the coauthor of studies of the effects of unemployment insurance on the unemployment rate and venture to guess that without the increased generosity of the UI system introduced in the 1970s, Canada's unemployment rate today might be about two to three percentage points lower than it is.
More generally, there is little doubt in my mind that insurance induced changes in behaviour explain the puzzle of the post war years, the ever increasing costs of social insurance programs and the constancy of the problems they were designed to cure.
The green book contains a graph of Canada's unemployment rate since the 1950s. It is on a steady upward trend. My analysis suggests that this is due to the effects of insurance induced changes in behaviour, institutions and the overall increased social acceptance of such behaviour.
What does my analysis imply for social policy reform? The answer is found in the practices used by private insurance companies to limit opportunities for risk increasing behaviour. Fire insurance premiums are lower for buildings with sprinkler systems. All forms of private property and health insurance carry deductibles, co-insurance on claims, waiting periods and rates that differentiate between people according to the riskiness of their behaviour. For example, life insurance premiums are higher for smokers than non-smokers; sky-divers face discriminatorily higher premiums.
It is important to note that the private insurance industry knows that moral behaviour cannot be controlled completely. Consequences of it are reflected in insurance premiums. Where control is extremely difficult and the costs vary as in the case of insurance against business losses, premiums have to be so high that no one is interested in buying the insurance. That is why there is no insurance against business losses.
For very noble reasons the architects of Canada's social insurance programs have made only limited use of such private insurance methods for controlling insurance induced behaviour.
To me the most logical solution to the excessive cost of these programs is to introduce more of the true and tried methods of the private insurance industry. Deductibles can be raised. In the case of unemployment insurance and welfare this would take the form of longer waiting periods before benefits are paid.
Co-insurance can be increased by lowering the benefits relative to previous earnings. Pensioners receive benefits only when they are older. The receipt of benefits can be lessened by requiring frequent reporting and evidence of job search, retraining requirements and a host of other measures which are equivalent to increasing co-insurance rates.
Other countries which have had similar experiences to those of Canada have taken such measures. Sweden has cut housing subsidies and brought in vouchers for child care payments. Norway is tightening disability insurance payments. In France social assistance is tied to signing on for a work training scheme. In the Netherlands young people up to the age of 27 have their social assistance set at no more than 50 per cent of the minimum wage. Germany and Italy are raising the age at which pensions can be claimed. In Britain income support is made available only on the basis of need and after means testing of applicants.
I believe that the phenomenon of insurance induced behaviour holds the clue on why in Canada and in other countries of the world the cost of social programs has run out of control. It also provides the answer for stopping these runaway expenditures. We could do better than listen to what other countries are doing to control their problem with insurance induced behaviour.