by groups. And it cannot be doubted that the statutes and their provisions for insurance have been drafted in the reasonable expectation that private insurance carriers also would vary their premium charges in accordance with the same general principle.

And in actual practice one of the most interesting and gratifying developments under the compensation system in America is seen in the careful and unremitting endeavors of all insurers-state funds, insurance companies, and mutual associations alike-to work out a sound rule for making premium rates depend in part upon the employer's efforts or success in safeguarding his work-places. Both the principles of discrimination adopted and the practical measures for carrying them into effect vary. There are open questions as to the relative weights to be assigned to the accident rates actually developed in experience (experience rating) and to conformity to prescribed standards of safety (schedule rating). There are also unanswered questions as to the extent to which individual conditions and experience may be considered without departure from the sound essential principle of all insurance; that is, the principle of pooling the individual in the collective group of the insured. It is not necessary here to outline the excellent work which has been done by numbers of very capable insurance experts in their attempts to solve such problems. But it must not be forgotten that much of the best thought of the insurance companies and of the compensation commissions is devoted to the working out of plans which may combine sound insurance procedure with the utmost practicable allowances and penalties for observance and neglect of approved safety measures.

And there are still other ways, at least as important, in which compensation insurance has conduced to the safety of industrial employees. Once the insurer has contracted for his premiums, it is clear that he stands to gain the more-to save the more of the premiums for his own profit-the more he can reduce the accidents of the insured below the numbers upon which the premiums were computed. And, at the same time, he may encourage his insured to hope that a reduced accident rate will mean a reduced premium for the next policy period. Thus we find that, as a matter of the most obvious business interest, all sellers of compensation insurance seek constantly to improve their risks, by inspections, by counsels of safety, and by the preparation and urging of all manner of safety

policies and appliances. Very fortunate, indeed, it is that corporations with resources as great as those of the large insurance companies find it unmistakably good business to work for the safeguarding of industry. Nothing has done more to clear private sellers of compensation insurance from the charge of heartlessly exploiting human misfortune for their own profit than their intelligent and persistent endeavors to make work-places ever safer. Nothing can be more likely to preserve for them their present rights in a large and expanding field of enterprise. At first individually, company by company, but more recently chiefly through their specially constituted associations and bureaus, and sometimes in friendly conference with representatives of the state compensation commissions, private insurance carriers have done a most admirable work in devising and recommending plans for the safe construction, equipment, and management of factories, mines, and other work-places. Their provisions for the human element, that is, against the so-called moral hazards, are scarcely less elaborate.

Whenever, as often under the statutes, those who employ large numbers are allowed to dispense with formal insurance and to "carry their own risks," workmen's compensation tends in the simplest and most direct way to promote safety. In all such cases employers know that each injury means so much direct cost to themselves and that, therefore, it is for their own immediate business advantage to reduce accidents to the lowest possible figures. Essentially so it is also with all genuinely mutual insurance associations.

The conclusion, therefore, is fully warranted that compensation insurance, with its merit ratings, does contribute largely to the contemporary movements for industrial safety. And both insurance and the merit ratings were contemplated in the workmen's compensation legislation generally, even in the states where they were not expressly required by the statutes. Here, then, is the happiest consequence of the compensation system. It certainly is well if many millions of money have been added to the sums formerly received as indemnity and solace for the losses and sufferings entailed' upon workmen's families by industrial accidents. But it is much better in every way if, as is not unlikely, some few thousands of lives are saved each year and some scores or hundreds of thousands of injuries prevented.

It must be confessed that, except as they did bring in compensation insurance and its merit ratings, the compensation laws have not done much for industrial safety. Not often have they prescribed rules or standards of safety or even conferred powers of inspection. This, however, has been because such matters have been left to distinct statutes and distinct offices. In California there was a temporary change of policy. For the revised compensation act of 1913, replacing substantially the act of 1911, was made up in considerable part of sections devoted specifically and exclusively to the direct promotion of safety in industry; and the act itself was legally designated the "Workmen's Compensation and Safety Act." But in 1917 the safety sections were entirely removed to a separate statute. Other states, as a rule, have followed the plan of separate statutes from the first. At present there are only about a dozen states which make any provision for accident prevention in their compensation acts. Doubtless there are real advantages in some measure of coördination or unification in the two lines of activity. But the present degree of separation is not as great as might appear. For in many of the states there are industrial commissions with general authority over all administration of laws affecting labor, workmen's compensation laws, factory-inspection laws, safety laws, and the rest. And it can make but trifling difference whether public regulations which a given office is to administer are found in different parts of one statute or in different statutes.

In the critically important matter of fixing compensation-insurance rates, whether for the state funds or for private insurers, the framers of the statutes usually have thought it sufficient to take account of the actually developed accident experience of industries or groups of industries. Rarely have they thought it well to sanction schedule ratings; that is, ratings based upon conformity to approved standards of safety. The terms of typical statutes, as in Ohio, unmistakably contemplate no other principle of merit rating than such as may be derived from the accident experience of the insured. There are statutes, as in Kentucky and Virginia, which in general terms authorize "merit rating," while others authorize or direct "schedule rating," with or without a fair implication that the words

1 As a matter of fact, legislatures and their draftsmen knew very little about the different principles and methods of merit rating.

are to be taken in their technical meaning. Only in a few of the states, as in Colorado from the first, in Michigan, New Jersey, Pennsylvania, and now since 1917 in Washington, is there express and definite sanction for the adjustment of premium rates with reference to physical and moral conditions in plants, as distinguished from accident rates revealed in experience. Accordingly, the state commissions, practically limiting themselves to experience ratings in their administration of the state funds, have been much less efficient than private insurers in their use of a most powerful force for the promotion of industrial safety. The provisions of the Pennsylvania statute and of recent amendments in New Jersey and Washington may indicate that this important fact is coming to be recognized in public places. If so, a change for the better may be in prospect.

But, even without large use of this efficacious means, some of the state commissions have produced gratifying results. They have conducted extensive educational campaigns; they have made inspections, examined and recommended safety devices, established safety exhibitions and museums, and widely scattered helpful printed materials. The California Industrial Accident Commission long has issued a valuable illustrated California Safety News, which cannot have failed to do much good. The same commission conducted a special campaign for safety in mining industries, with the apparent result that fatal accidents in California mining fell from 20 in 1916 to 10 in 1917. In 1914 the Massachusetts Industrial Accident Board carried on for three months a campaign for safety in establishments with some 55,000 employees. And a comparison of accident rates for half-year periods before and after the campaign shows a reduction of 20.8 per cent in the total number of accidents. It is by discriminating attention to experiments and reports like the two just mentioned-not by observing the broad, general movements of accident rates-that one may come to his best conclusions as to the effects of the compensation laws in making industry safer. . . .

So far, then, it would appear that the compensation system has had no appreciable direct effect upon business. Nor, in the nature of the case, can there have been any large, direct effects. The single new influence to work directly is the weight of the compensation costs as an addition to the expenses of doing business. And, on the average, these costs cannot have figured for much.

Clearly their magnitude is shown by the premiums for compensation insurance. And these, varying from a minor fraction of 1 per cent of the pay roll in most textile mills to several per cent in structural iron work, mining, and other specially dangerous operations, may perhaps be generalized at about 1 per cent, or a little more.1 And such a cost, a cost of such amount and of such character, can have no great effect upon the total expenses of producing, upon necessary prices of products, or upon demands for goods.

The addition of 1 per cent to labor costs becomes an addition of about one fifth of 1 per cent to total costs when account is taken of the fact that labor, as in manufacturing, figures at about one fifth of all costs. In other industries, as in mining, labor counts for much more. But the costs of compensation, of compensation insurance, are not a net addition to the employers' business expenses. Something employers paid in former days on account of industrial accidents. The largest item was the amount paid for employers' liability insurance. But there were other items also-the small sums given occasionally in direct and gratuitous settlements of claims, the costs. of litigation, and the rest. So that the immediate net increase in the costs of production or business must be much less than the present costs of compensation insurance. In fact, there have been official claims for several of the states that the new system costs employers no more than the old. Something still further must be allowed in reduction of the fair charge against compensation. Better medical care for injured employees and larger cash benefits for them make the interruptions of their work less prolonged and less disturbing to the orderly course of the business.

When all due allowances have been made for these reactions of the compensation system, as for others which might be added,

1In Montana, where employments cannot be especially free from dangers, 1 per cent is the official estimate of the costs for the two years 1917-1918. In West Virginia, where also mining is an important industry, the general average of premiums for 1916-1917 was 1.41 per cent. In Wisconsin during 1913-1916 it was 1.33 per cent. In the exclusive state funds the premiums or assessments are lower. But the question of true costs is so much involved with other subordinate questions that it cannot be answered in a convincing manner without elaborate expositions and discussions. The variety of Wisconsin industries and the different agencies of insurance operating in that state perhaps make Wisconsin figures as instructive as any.

2 Montana Industrial Accident Board, 1918-1919, pp. 28-29.