Welfare State & Insurance Theory
“The fourth class includes the precepts relating to charity, loans, gifts and the like (...) The object of these precepts is clear; their benefit concerns all people by turn; for he who is rich today may one day become poor; and he, who is now poor, he himself or his son may be rich tomorrow”
Moshe Maimonides, “The Guide for the Perplexed”, Ch.XXXV
The quote above, represents one of the most fascinating, powerful (and sadly abused) ideas of the history of the political thought. Of course, charity is a duty for many religions, but before Maimonides, it was justified only in the grounds of superstition and sentimentalism. Maimonides, on the other hand goes beyond this conventional (lack of) arguments, and synthetically describes the theory of social insurance, or even more, the Rawlsian “veil of ignorance” principle.
As you can see, he begins reminding both the rich and the poor that their respective positions are never safe (a classical biblical issue) and then, remind them that the money the rich give to the poor is not more that a payment for the right to get reversed payments if the chances change their face.
Rawls in the same situation would say that supposed you are “under a veil of ignorance” about your position in the social scale, you would agree with a system of social insurance, because the payments you would do supposed you were rich, detract less welfare from you, than the welfare you receive from the payments you get supposed you are poor.
There is, of course a sunken assumption in this insurance argument for charity: moral hazard is supposed to be very small. In fact, if you think that charity will make people lazy and irresponsible, the creation of social insurance is no longer a neutral system of redistribution, not affecting production; but on the other hand, it destroys both the work ethics and the social capital. In other words: if chance is what makes people rich or poor, a big social insurance is justified; if effort is the most significant factor determining income, the social insurance should be small. Probably Maimonides was very optimistic about Jewish work ethics, and no so optimistic about the relation between merit and income.
If social insurance is so good for all participants, why it has to be enforced? Essentially because the fortune is not as volatile as the argument above says: in fact the rich use to remain rich and the poor, poor. The social insurance is good under a veil of ignorance about your position on the social structure. As soon as you know what is your place, if you are rich you don’t want to pay, and if you are poor, you want bigger payments than optimal rawlsian allocation. Let’s make an example:
Let be A and B two inhabitants of a desert island. They know that one of them has a serious illness (A and B have the same probability of being the sick one), and it will need some money to go for a doctor outside the island. They are endowed with a technology of commitment, so if they sign a contract, they honour it.
Supposed both of them are risk averse: then they sign a contract to pay the treatment half and half: that is, they fully insure each other. But what if somebody provides them with a diagnostic system? Of course, the veil of ignorance vanishes, and the healthy one doesn’t insure the sick: there is a total loss of welfare: the welfare they get from the insurance contract (as risk averse agents) disappear; the healthy one doesn’t pay and the ill one takes all the risk. 
This situation lead us to a classical economic theorem: full insurance of exogenous risks is a social optimum. In fact, a perfect social insurance system would be able to divide all the risks an individual faces in her life and divide them in two sets: chosen risks (the individual should pay for) and exogenous risks (that should be socially insured).
Sadly, it is impossible to perfectly implement this principle: for example, in the case of unemployment both luck and effort play their roles. Other cases, as health are clearer: there are good incentives to be healthy, and strong randomness in health: so public health insurance is optimal .But the principle, even difficult to implement is still very powerful, and represents one of the core principles of economics.
 The example is not theoretical: the development of genetic tests will make much more difficult for many people to get insurance.
 In my opinion, the production of health services should be private; but health insurace should be enforced and eventually subsidized.