Self-selection can be applied in a variety of situations. We regularly hear about it in the context of relief employment. Only poor people turn up to do unskilled manual labour at minimum wages. Apart from employment, this could be used in a variety of contexts. It is often suggested that by giving coarse grains in the public distribution system, we could weed out some of the relatively rich people, since they would not buy coarse grains. Small benefits by themselves attract people who want it badly. This could be creatively extended into certain other schemes too.

It is not enough to have self-selection. It is equally important to have universal access too if we want to reach out to large sections of the population. Self-selection avoids the problem of inclusion. Universal access is important to ensure that there is no problem of exclusion. The problem of no having universal access is glaring in the context of relief employment. ‘Labour ceilings’ have been imposed in most of the sites of relief employment across the country. These are the represent the maximum number of people who will be employed. These are prepared often at the district level. This has resulted in many deserving people being left out of the employment net. Otherwise reduced amount of employment is given to people who are demanding it. This takes a serious proportion since it is happening in drought affected areas where there has been a complete collapse of livelihood.

The chief reason why identification fails is that we don’t have proper information about he attributes of a person, and an exercise to document it for the nation will be futile. In the case of the most vulnerable population, it is often feasible to identify the deserving by certain social attributes. For example, widows without family support are among the most vulnerable people. It is easy to identify them. Aged people without support, families with disabled people, primary tribes and such other groups can be identified with relative ease. These groups can be identified easily since criteria like age, widowhood, etc. are easy to specify and identify. This would have problems of inclusion, but the its degree is bound to be low. In this method too, it is important to have universal access.