We may find the head of a business hesitating between two schemes of production , of which one is more efficient than the other , but also more risky .sx It may be worth while to sacrifice some efficiency in order to acquire greater security .sx The chance of total loss will be diminished , and so , perhaps , will the chance of partial loss ; yet since his costs are greater , the chance of any given amount of net gain is also reduced .sx The principles on which such a decision will be based are derived from considerations that we have yet to examine .sx Here it is only necessary to point out that such decisions are made .sx One way of dealing with the uncertainty involved in a given undertaking is to reduce that uncertainty by changing the scale and technique of production .sx But how far this will be done clearly depends on the availability of other methods .sx III .sx Assume , then , for the present , that the form and scale of our firm are given .sx How will the remaining effective uncertainty be borne ?sx Who will suffer the losses and pocket the gains which result from the varying fortunes of the enterprise ?sx There seem to be two possible ways of dealing with the remaining uncertainty .sx ( 1 ) Certain specific risks can be transferred to outside parties ; ( 2 ) the remainder must be borne by those actually co-operating in the firm , bringing their labour or their capital or both .sx In what follows we shall be mainly concerned with the second of these alternatives .sx The first ( including insurance , in a broad sense , and speculation ) is in many ways closely related to the technical adjustment of risks already discussed .sx But it is also in part a distribution of risks among those most willing to bear them , and the operation of that we shall proceed to illustrate in a more general case .sx When this has been done , the analysis of insurance will present few difficulties , and for the sake of brevity will be omitted in this paper .sx A conceivable form of economic structure would be one in which all the parties to a firm shared in the same way in the risks of enterprise .sx Labourers and capitalists would form a co-operative association , each participating in the " ownership " of the business , and each consequently drawing his remuneration in the form of an agreed proportion of the firm's income .sx But in practice such a scheme is highly exceptional .sx Partly for historical reasons , but partly , as we shall see , on account of its usually superior efficiency , enterprise has generally taken another form .sx The co-operating parties are divided into two groups , one only of which receives a remuneration depending on the firm's success ( that is to say , receiving a share of profits) .sx The members of the other receive a remuneration which is not directly contingent on the result of the operations in which they collaborate , but which is fixed before the act of collaboration is performed .sx The first group receives profits ; the second receives wages , interest , or rent .sx The slightest consideration of the history of industrial structure will show that although this division has nearly always .sx existed , it has by no means always , or even usually , coincided with the traditional division of the factors of production .sx In some stages of industrial development it has been nearly true to say that while labour receives wages , capital receives profits ; but it has never been quite true .sx At other times it has seemed appropriate to identify the receiver of profit with the provider of some particular sort of labour , i.e. management or organisation ; but hardly had organisation been promoted to the rank of a separate factor of production when the growth of the salaried manager deprived this scheme of its plausibility .sx To-day we must surely admit that the decision whether any factor of production is to be paid by a share of profit or by a previously fixed reward does not depend on the nature of the service rendered .sx Profit is not in any sense correlative with wages and interest .sx The distinction between wages and interest does depend on the nature of the service ; the distinction between them and profit depends on the way the service is remunerated .sx The division between the two groups is in fact a scheme for the distribution of risks ; but it is not a simple scheme whereby all risks are transferred from one group to the other .sx The group whose payment is fixed in advance does retain some risks , in particular , the risk that when the service has been per-formed , the payment will not be made .sx Legal provisions have reduced this risk to a minimum , but they cannot abolish it altogether , since it is always possible that the enterprise will yield a return which is insufficient for the firm to cover its commitments .sx This is in fact practically the only risk which is left to the capitalist who invests in fixed-interest-bearing securities ; but to the labourer who is being paid in what is logically the same way , it is much less important than the risk that the service that he performs will involve him- in some danger to life and limb , for which Workmen's Compensation will most inadequately compensate .sx In the cases of both labour and capital , then , it is impossible for the party who is engaged at a fixed payment to be relieved of all risks-but the risks are very considerably reduced , and the kind of uncertainty borne is very considerably changed .sx The difference between the kinds of uncertainty borne by the two groups can be easily illustrated by the method of frequency curves .sx Along OX we plot sums of money ; along OY fractions expressing chances .sx Suppose the chance that a particular .sx unit of capital or labour will bring to him who supplies it a gain of 50 , is 1/6 .sx Then we shall plot OM = 50 ( 's ) and PM = 1/6 .sx In this way the whole uncertainty of the result of a particular investment can be expressed in the form of a curve and we can read off from the curve the chance that the investment will give any particular result .sx Where resources are hired for a fixed payment , but there is a possibility that the payment will not be made ( or not made in full ) or that some incidental misfortune will overtake the lender as a result of the investment , then the uncertainty-curve will take the form shown in Figure 1 .sx When resources are exposed to uncertainty , and the reward is to be a share in the net return , whatever it is , the form of the curve will be as in Figure 2 .sx In many cases the curves will be much less dissimilar than those shown in the diagrams , but so long as there is a maximum possible reward , and the chance of that maximum being paid is relatively large , the case falls into the first class .sx ( In either case , of course , there may or may not be the chance of a negative return .sx ) When the first method of hiring resources is used , the question is simply one of demand and supply at various , promised rates of payment .sx Supply will , of course , be considerably affected by the confidence of lenders in the realisation of their expectations .sx Where the second method is used , however , we can no longer treat the problem as one of equilibrium at a rate .sx Yet for any given set of terms ( or uncertainty-curve ) there will be a certain supply of any given kind of resources ( including the resources furnished by the " entrepreneur himself if we are dealing with a private business) .sx Similarly , as we shall see , there will be a certain demand .sx In either case , a change in the terms offered will change the form of the uncertainty-curve , and this change will react on the demand and supply of the resources in question .sx By changing the terms , demand and supply can thus be brought into equilibrium as in ordinary price-theory .sx But whereas in price-theory , the adjustable index is a numerical ratio , here it is a relation or function , and to complete our comprehension of the nature of this equilibrium we must examine how changes in the function react on demand and supply .sx IV .sx What , then , will be the effect of a change in the chances of reward on the supply of resources for a given employment ?sx The simplest possible assumption would make everything depend on the " expectation " of reward , in the sense in which that word is used in the theory of probability-the sum of the products of chances and quantities of return anticipated .sx Of two alternatives the investor would go to that one where the expectation of reward was higher , and thus if the expectation was increased , the supply offered would increase too .sx If this assumption were correct , and if the estimates of investors were either accurate or showed no bias in either direction , then we should arrive at the interesting conclusion that no net share in the National Dividend was to be attributed to any-one for the bearing of risk .sx In the long run , and over large groups of people , successes and failures would balance , so that the return received by the whole body would be the same as if their investments had carried no risk at all .sx But this conclusion , in many ways so convenient , is not really consistent with observed facts .sx For it also follows from this assumption that any two schemes of uncertainty , where the expectation was the same , would be indifferent to the lender .sx And this as a general rule seems highly unlikely .sx It would mean that people would be equally ready to invest in an under-taking that promised a safe 100 , in one which promised equal chances of 200 or nothing , and in one which promised nine chances of 110 to one of 10 .sx It seems fairly safe to say that this is not the case .sx Attempts have been made to arrive at a more satisfactory principle by means of the law of diminishing marginal utility .sx On this basis , the indifference of two uncertainty-schemes would not imply equal expectations .sx That scheme which has the greater proportion of chances of low returns must have a higher expectation , i.e. it must also have a greater proportion of high returns and this must more than counterbalance arithmetically the low ones .sx It is the expectations of subjective satisfaction , not the expectations of money returns , which must be equal if the two schemes are to be indifferent .sx Such a theory is certainly more satisfactory than the crude view which bases everything on money expectation , but it is improbable that it contains the whole truth .sx We may accept diminishing marginal utility as a practically universal law when it is applied to actually realised satisfactions .sx The things which a man can buy with a second 100 are rarely as essential to him as those which he buys with his first .sx But although it would follow from this-if man were a completely rational being - that his desire for 200 would be less than twice as strong as his desire for 100 , can we be certain that it does really follow , in view of the known peculiarities of the human species ?sx And there is one peculiarity which does suggest a powerful tendency in the other direction-the taste for gambling .sx There is no doubt that under certain conditions men are willing to exchange a certainty for a scheme of uncertainties the value of whose expectation is considerably less than the value of the certainty they abandon .sx But this scheme of uncertainties is nearly always of one type .sx When the scheme includes a small chance of a very large gain , " rational conduct " ( based on the law of diminishing marginal utility ) would estimate the value of this chance as very small indeed .sx Practice , however , does just the reverse .sx The philosopher who acted up to his principles would never be willing to give even 1 for a 1/1000 chance of winning 1,000 .sx But men can always be .sx found who will give a good deal more .sx Whether this is due to an unconscious " blind eye " to the chances , or to a conscious preference , may be left to the psychologist to decide .sx