An interactive demonstration of adverse selection

This graphic shows the purchasing decisions of a certain number (default is 500) insureds under conditions in which, by adjusting three lines, the insurer can group potential insureds into four pricing categories. The x-axis of the graph shows the risk posed by an insured. The y-axis of the graph shows the aversion to risk of an insured.  Each insured has some risk level (x) and some risk aversion level (y). Insureds who end up purchasing insurance are colored green and insureds who end up not purchasing insurance re colored red. The top right of the graphic contains an inset graph containing a Histogram of the welfare of all the potential (and actual) insureds. The x-axis shows the welfare (higher is better) and the y-axis shows the number of such insureds. The number atop the inset is important: it is a "spectral measure" of the well-being of society given an inequality measure chose by the user. Higher is better.

Overhead refers to the load on insurance policies due to expenses. Making the overhead slider take on negative values permits you to examine the effect of insurance subsidies on adverse selection.

Mean expected loss deals with how high mean losses are.

Heterogeneity of expected loss refers to the "spread" of the distribution of losses. You should see what happens to adverse selection as the heterogeneity of the insurance pool decreases and increases.

Mean risk aversion shows how afraid the average insured is of risk. See what happens to adverse selection as one increases or decreases risk aversion.

Heterogeneity of risk aversion refers to the mean aversion to risk of the insurance pool. 

Inequality weighting affects only the spectral measure of well-being shown in the inset graphic. Increasing the weighting will result in greater punishment of unequal distributions of welfare. Decreasing it will result in inequality becoming progressively less important.