upon measures intelligently. In Oregon and California, where the joint operation of the compulsory referendum on constitutional amendments and of the optional referendum on statutes has thrown a comparatively heavy burden on the voters, and where the official bulletins are said to be carefully read by many of them, there is far more evidence of discrimination between measures than in such a state as Missouri, where the burden of the referendum has also been comparatively heavy and where there has hitherto been no official bulletin of information.

It is not easy to determine just how effective the official campaign bulletins are as a means of educating the electorate. In Oregon the use of the pamphlet for the publication of arguments upon measures referred to the voters is optional with the supporters and opponents of the measures. In only a small proportion of cases are both affirmative and negative arguments published. Half of the referred measures have been published in the pamphlet without any arguments. Nor is there any relation between the publication of arguments and the results of the voting on the measures. Of the measures accompanied by affirmative arguments alone more were accepted than rejected, and of the measures accompanied by negative arguments alone, more were rejected than accepted. But the total number of measures referred to the electorate by means of the optional referendum is comparatively small, and the relation between the publication of arguments and the decision by the voters uncertain. There has been no perceptible tendency on the part of the opponents and supporters of measures referred to the voters by means of the optional referendum towards a more general use of the privilege of publishing arguments in the official bulletin. Apparently those who are most concerned have not yet been convinced by experience with the Oregon pamphlet that official publicity pays, nor have they been convinced that it does not pay. In California arguments on each side of every measure referred to the electorate must be published in the pamphlet, together with the text of the measures, and provision is made for the preparation of the arguments by members of the legislature or other responsible persons. A comparison of the California and Oregon pamphlets shows that the arguments and information laid before the voters in the former state are more complete and more adequate than

in the latter, and the California pamphlet is presumably a more effective aid to the voter.

Though the degree of effectiveness of the official bulletin as a medium of publicity cannot be accurately determined, and doubtless varies in different states according to the character and condition of the voters, it cannot fail to be of considerable value in the development of the referendum as an educational discipline. The clearest evidence of this is the marked increase in the proportion of voters attending the polls who have voted on constitutional amendments proposed by the legislature that has occurred since the adoption of the optional referendum and the publication of an official bulletin. In California during the ten years 1899 to 1908 the average vote on the fifty-one measures submitted by the legislature, indicated as a percentage of the total attendance at the polls, was forty-three per cent. In 1914, when the system of direct action upon measures by the electorate was subjected to the severest test ever imposed in any state, the result was as follows:

Despite the heavy burden laid upon the voters by the combined operation of the initiative and referendum, they voted more generally upon measures submitted by the legislature than ever before. In Colorado during the ten years 1899 to 1908 the average recorded vote upon the seventeen measures submitted by the legislature was thirty-seven per cent of the total recorded vote. The initiative and referendum were adopted in 1910, a year before their adoption in California, but there is no official bulletin or other provision for publicity except in the newspapers. The result of the voting on measures in 1914 was as follows:

Doubtless the greater increase of interest in measures submitted by the legislature in California than in Colorado since the adoption of the initiative and referendum cannot be explained wholly by the adoption of better means of publicity. The evidence, however, such as it is, tends to confirm the presumption that an official bulletin like that of California will materially help to arouse the interest and inform the intelligence of the voters.

The tendency among the states which have adopted the optional referendum is towards the adoption and further development of the official bulletin. In 1914 official bulletins were published in eight states, and the voting on measures was more general in those states than in the states where there was no official bulletin. Of the various bulletins, that published in California was the best, both in form and in substance.

OPTIONAL V. COMPULSORY REFERENDUM

Whether the optional referendum has yet so proved its worth as to justify the complete abandonment of the compulsory referendum may be questioned. With respect to various matters of detail the practice of the several states differs, and further experience may be necessary before a standard form of the optional referendum is developed. Yet it is already clear that the optional referendum, under the proper conditions and rightly used, satisfies the two tests of good government. As Mill phrases it, "government is to be judged by its action upon men and by its action upon things; by what it makes of the citizens and what it does with them." Judged by either test, the optional referendum upon legislative enactments has proved in the main a better governmental device than the compulsory referendum on constitutional amendments proposed by the state legislatures. It

is a more efficient instrument both for the correction of misrepresentative action on the part of the legislatures and for the education of the voters themselves.

The substitution of the optional for the compulsory referendum upon constitutional amendments proposed by the state legislatures would have the effect of diminishing the existing distinction between constitutional and ordinary statute law. Now the process of constitutional amendment upon legislative initiative, as established in almost all states, requires the approval of all proposed amendments by the people. If that approval were dispensed with, except in the case of such amendments as should be brought before the electorate under the optional referendum, the status of amendments not referred to and formally approved by the electorate might seem less secure than that of those ordinary laws expressly approved by the electorate. To be sure, the legislatures are usually forbidden to propose constitutional amendments by bare majorities, but legislatures are also forbidden to adopt some kinds of ordinary legislation by bare majorities. In fact, the original distinction between constitutional and statutory law has already been so far impaired by the insertion of ordinary legislation in the state constitutions that the preservation of the distinction in its present form is of doubtful utility. It is for that very reason that it is proposed to substitute the optional for the compulsory referendum on constitutional amendments. The legislatures have shown themselves incapable of maintaining the distinction in any logical form, and for the revival of the traditional distinction, which is important, the public may confidently look to its more direct representative, the electorate. Whether the compulsory referendum should be retained for certain classes of amendments, as it now exists for certain classes of ordinary legislation, is a question that will be considered presently. It is enough to point out here that the vital distinction between constitutional and statutory law does not reside in the character of the procedure for its enactment, but in the importance attached to the substance of the law itself.

There can be no doubt that the referendum is now permanently established among the political institutions of the states. There is no question of abandoning it. The only questions concerning which there are still serious differences of opinion relate to the form

in which, and the conditions under which, it shall be used. In the majority of states the referendum still exists only in the mandatory form, and is applicable only to constitutional amendments, and, in some of these states, to certain classes of ordinary legislation. In a large minority of the states it exists also in the optional form, and is applicable to most of the ordinary legislation. A comparative study of the operation of the referendum in each of its forms throws much light on the problems which still remain unsettled.

The evidence indicates that the referendum, like any other political institution, is an imperfect instrument, which works better under some conditions than under others. The first limitation upon its most effective use is one of number. The greatest number of measures hitherto brought before the voters at one time by the operation of the referendum is forty-two. These were all constitutional amendments, proposed by a constitutional convention in Ohio in 1912, and submitted to the voters at a special election. The greatest number of measures of legislative origin hitherto brought before the voters at one time by the operation of the referendum is thirty-one. Of these twenty-seven were submitted by direction of the legislature and four by means of the optional referendum at the regular election in California in 1914. In both states the action of the electorate was generally conceded to have been deliberate and on most of the measures intelligent. Yet it was clear that the number of measures was greater than the voters could easily dispose of. There was a general opinion that the burden laid upon the electorate was excessive. The average number of measures brought before the voters by means of the optional referendum alone is not more than two or three, and the evidence indicates that this number of measures can be easily handled by the voters. When the number becomes excessive, the voters have a way of voting "no" on all or most of the measures without much regard to their several merits. This remedy has been most conspicuously applied in Missouri. In that state the presence of much statutory matter in the constitution occasions the submission of numerous amendments, and inadequate provision is made for the information of the electorate. In 1914 eight constitutional amendments were submitted by the legislature and four legislative measures