Still, scientific method could be pushed further than it has been in the comparison of occupations. The statements of the various interested parties would be a valuable guide in the estimate of occupations. Furthermore, only the major relationships between occupations would have to be taken into consideration. For example, if the question at issue was whether the wages of miners were too low as compared with wages in other industries--that is to say, whether a demand on the part of the miners for an improvement in their relative economic position was justified--only the most important of mining occupations would have to be taken into account in reaching a decision. There would be small risk of error in applying a decision, based upon a study of the work performed and of the income received in the most important mining occupations, to the less important mining occupations also. And indeed such would prove probably the only practicable policy. Furthermore, revision of the existing differentials would be undertaken only when the case for revision seemed definite and clear. As for example, it was clear in England before the war, that railroad labor was underpaid; or, as was clear to the whole of the recent President's commission on the wages of coal miners, that the wages of the miners were too low, relative to wages in other industries--though the commission differed on the amount of wage increase to be awarded. But perhaps the most significant answer to those objections which rest on practical grounds is the fact that any wage level that might be set for any occupation under the proposed principles would be but the minimum standard wage for that occupation. And no element in the whole policy of wage settlement should stand in the way of the payment of a higher wage than that fixed by the central authority for any type of work. Thus no fear would have to be entertained that any industry would be faced with a shortage of labor due to the difficulty of getting precise knowledge on which to base wage differentials. Here, indeed, we approach very close to that other objection which may be put forward on theoretical grounds. Which objection is that all attempts at revision of existing wage differentials would involve a risk of producing, on the one hand, a shortage of certain kinds of labor, and, on the other hand, an oversupply of other kinds. It is reasoned that in spite of every effort of careful calculation of wage differentials, some danger of over or undersupply of certain kinds of labor will always be present.