This write up will seek to give a critique of an article “Insurance Versus Self-Insurance: A risk Management Perspective”. This is a journal of risk and insurance. The article was written by Patrick L. Brockett, Samuel H. Cox, Jr. and Robert C. Witt and published in June, 1986. The article seeks to expound on what self insurance really mean, its technicality and the condition under which a firm can opt to self insure a number of its exposure units instead of insuring them. They consider both self insurance and insurance in their study. They clearly demonstrated how firms can obtain maximally tight bounds on the expected utility of self insurance irrespective of the amount of information about the loss severity distribution they have (Patrick et al, 1986).
In justifying their study, the authors have noted the difficulty that most risk managers experience in making decisions on whether their firms should opt for self insurance or not. They sought to give various options which the managers can use to carry out a detailed scientific analysis. The authors have also given clear explanations of the various concepts that they use. Some of such concepts are risk management, self insurance, insurance (Patrick et al, 1986).
The authors clearly explained the approaches they utilized in their study such as an expected utility approach. This approach enabled them to derive the worst and best possible case distributions and to obtain the upper and the lower bounds on the expected gain or loss of utility resulting from self insurance. The technique also helped them obtain the number of exposure units the company must have for self insurance to be financially feasible. They also gave the way in which the possibility of the occurrence of bankruptcy could be accessed. They clearly described how the expected utility model would be used to translate utility bounds into dollar bounds on the amount the firm would appropriately pay for any insured loss while having incomplete information about the loss distribution (Patrick et al, 1986).
Giving their statistical framework of analysis, they stated that though they would use a generalized non-specific exposure unit, they chose on an automobile which is normally exposed to liability peril during one year. This framework was the most appropriate according to them. They also used clear and appropriate variables such as the total number of accidents and their severity, the random amount of disconnected loss, the random number of accidents, the premium, and the finite upper bound.
The authors clearly stated their assumptions. They assumed that each automobile had a given severity distribution while the losses from different accidents remain independent. They also assumed that the severity distribution for a given exposure unit is independent of that of all other exposure units.
Even though this study has expressed a high level of professionalism on the part of the authors, the article still had some weaknesses. One of the ways in which it could be improved is by providing a clear delineation of the abstract. Second, even though the authors have used sources which look credible and relevant, some of the sources were as old as thirteen years during the time of study and may have not been updated to contain the latest advancements on the topic of study. The authors have also failed to put in-text citations leading the readers to guess what information would have been borrowed from which source (Patrick et al, 1986).
The authors have managed to present their study with precision and clarity. From the onset of their study they made their objective very clear while also explaining the problem statements to justify the need for their study. They also adequately designed and used various appropriate methods in reaching conclusions. Their procedures are presented in details making them easy to be duplicated by the reader. The study can therefore form a basis for further research today on the topic today.