in view of the fact that the application of mathematics to agriculture introduces an entirely new factor into scientific agriculture in India. Instruction in this branch of agricultural science could be usefully included in the post-graduate courses at Pusa which we have recommended for all recruits to the new superior provincial services, but it can only be given if an officer with the qualifications we indicate above is available.

We do not recommend that any additional subordinate statistical staff should be engaged or that any changes should be made in the present arrangements for dealing with agricultural statistics until a statistical assistant has been appointed to each provincial agricultural department. Otherwise, there is a serious risk that money may not be spent to the best purpose. One of the duties of this officer will, indeed, be to see that the limited funds which will be available for statistical purposes are put to the best use. Ample material on which to base proposals for improvement which involve little or no expenditure exists in the exhaustive reviews of agricultural statistics made by the Board of Agriculture at its meetings in 1919, 1924 and 1925. The possibility of obtaining funds for the strengthening of the superior statistical staff by amalgamating subordinate staff is one that should not be lost sight of.

We have directed our attention chiefly to the statistical requirements of the agricultural departments. We have, however, had occasion to notice, in our enquiries regarding indebtedness and the health of the rural population, the absence of any general statistical organisation at the headquarters of each province. We are aware of the competing claims on such funds as are available which are made by all "nation-building" departments and also of the scepticism which provincial legislatures may feel towards the necessity for, and the value of, statistical work in such matters. We would draw attention, nevertheless, to the increasing importance which is being attached all over the world to statistical research as an aid to the formulation of social policies. We consider that the efforts of Government to promote rural welfare would be greatly assisted by the appointment of a well qualified statistical officer to provincial headquarters to whom would be entrusted the duty of studying all aspects of the economic and social progress of the province. Such an officer would be a valuable correspondent for central workers in the fields of medical research and human and animal nutrition and in him the educational, co-operative, health and other social workers within the province itself should find a most useful guide. He should be of the status of an officer of the new superior provincial services and should receive pay on the scale decided upon for those officers. We do not think that the appointment of such a statistical officer at provincial headquarters should occasion much additional expenditure. He should be a centre round which voluntary workers in the field of economic and social statistics would gather rather than the head of a purely official organisation. In general, we envisage him as acting as director of a bureau of statistical information with a small staff of subordinate officials working under him. For, one of the advantages to be derived from the appointment of such an officer would be that one single

authority would be responsible for co-ordinating and publishing all provincial statistics other than those of crop forecasts and agricultural production. Eventually, indeed, it might be found advantageous to include agricultural statistics in the work of such a bureau and an intimate connection between its work and that of the statistical section of the provincial agricultural departments should be established from the outset. The wider the field a statistical officer, who is a man of insight, is called upon to survey, the more opportunities he has for making those discoveries from the correlation of data derived from widely different sources which are the aim, and frequently the achievement, of all modern statistical work.

The suggestions we have made for the organisation both of agricultural and of other statistics are designed to secure the broadest basis of development with the least possible expenditure. We have carefully considered the recommendations of the majority report of the recent Indian Economic Enquiry Committee which contemplate a total expenditure by provincial governments, collectively, which is estimated at the large sum of Rs. 15 31 lakhs for initial outlay and Rs. 28 40 lakhs for recurring annual expenditure, and we are unable to support the Committee's scheme. We hold strongly that officers of real statistical ability should first be appointed and that subsequent development should come as the result of their recommendations and should be for clearly defined objects which are shown to be both essential and practicable. We are opposed to general censuses of production or large schemes of economic investigation into the condition of the population undertaken with the mistaken notion that the mere accumulation of facts will of itself serve a useful purpose.

CENTRAL STATIS

TICAL ORGANISATION.

539. At present, the statistical work of the Government of India is carried out by the Department of Commercial Intelligence and Statistics. At the head of this department is an officer with the designation of Director General. Agricultural statistics are dealt with in a separate section of the department, under a part time superintendent with a staff of eight clerks, only five of whom are, however, exclusively engaged on the compilation of agricultural statistics.

If, as the result of our recommendations, a Council of Agricultural Research is created, it is clear that it will require statistical information of a character which the existing section concerned with agricultural statistics will be unable to provide. The Institutes of Human and Animal Nutrition will also require special statistical assistance, as will the Central Bureau of Information for irrigation and hydro-electric questions.

Again, although the general tendency towards provincial self-sufficiency must be recognised in the statistical as in other fields and while we have, therefore, laid the main emphasis on provincial development, it cannot be doubted that there are important statistical services which can only be provided by a central agency. Facility of comparison between provincial statistics is essential to their full utility and the

uniformity necessary for comparison is most easily secured, if machinery for co-ordination exists. There are many investigations which have a general as well as a local aspect; for example, the statistical treatment of the relation between weather conditions and the yield of crops.*

We, therefore, think that general considerations are all in favour of reconstituting the Department of Statistics as a separate department. A possible alternative would be to recognise that statistics are required for two quite different purposes, primarily for information and secondarily for correlation with a view to obtaining fresh light on agricultural, health and economic problems. To meet the first of these requirements, arrangements might continue as at present. For the second, an expert statistician might be appointed and placed in control of a small but highly skilled subordinate staff. We believe, however, that any separation of ordinary statistics from what may be termed research statistics would be unsound as well as expensive. Ordinary statistics are the material on which the expert works and the more familiar he is with their origin the better. We, therefore, prefer an organisation separate from that of the Department of Commercial Intelligence which would deal with both classes of statistics. Care should be taken to safeguard the head of the new department from all avoidable routine work by the establishment, within his office, of sections dealing with the various sets of ordinary statistics each under the control of a competent superintendent. As in the case of the provincial organisation, we contemplate that provision should be made for higher statistical enquiry on modest lines and we are unable to agree with the proposal of the majority of the Indian Economic Enquiry Committee that a large central organisation should be created at an estimated initial outlay of Rs. 16 56 lakhs and a recurring annual expenditure of Rs. 29 15 lakhs. All that would be required at first would be the engagement of a statistician of recognised competence who, with his staff, would be appointed and paid by the Government of India. He would be placed in control of the existing statistical staff and would be responsible for suggesting increases of staff as and when such increases were demonstrably justified by the work to be done. As regards the qualifications required of the statistician himself, we would stress the need of obtaining a man of first rate ability and of making, therefore, the pay and conditions of service in the appointment such as would attract an outstanding man. He might be styled the Director of the Central Bureau of Statistical Information. It would, in our view, be useless to engage any but a first class statistician for the duties we have in mind. The officer selected must be of sufficient standing to make his advice acceptable not only to the provincial statistical officers but also to the business world and the informed public. One of his most important duties would be to establish close touch with these very important non-official sections of

46

See, for instance, the evidence of Mr. J. H. Field, M.A., B.Sc., late Director General of Observatories (Vol I, Pt. I, pages 193-208) and the Article on Crop and Weather Data in India and their Statistical Treatment," by Mr. S. M. Jacob, I.C.S., in Vol. XXII, Part IV, of the Agricultural Journal of India.

the community and we trust that provision will be made in the central statistical organisation for boards which would advise on the publication of statistics and their periodical revision, and that representative leading economists, scientists and business men will find a place on these boards. as well as officials of the departments interested. In this way, we trust that a school of statistical interpretation might develop in this country which would have little or no formal connection with Government but would, nevertheless, have access to, and be thoroughly familiar with, official statistical material of all kinds. In fact, there would, we trust, gather round the Central Bureau of Statistical Information the beginnings of a Royal Statistical Society for India.

The whole basis of statistics in India urgently requires broadening. It should rest not on the work of a few government officials, however able, but on the support of the informed public, and, through them, on the recognition by the legislatures and by the general public that modern statistical methods are in a position to make an indispensable contribution to the successful development alike of scientific agriculture and of social administration.

RELATIONS WITH
STATISTICAL

THE

BRANCH OF THE INTER-
INSTITUTE

NATIONAL
AT ROME.

540. It has been suggested to us in evidence that establishment of closer relations with the International Institute at Rome might be productive of particularly valuable results in the statistical field. We deal with this question from a general point of view in Chapter XX, paragraph 578, and there recommend that relations between India and the Institute should be as close as possible and that officers on leave or duty in Europe should take every opportunity of visiting the Institute and acquainting themselves with those branches. of its activities which may have a special bearing on their own duties. In this recommendation, we include statistical officers and other officers interested in statistical methods. We agree with the conclusion of the Board of Agriculture in 1925 that statistical information of value to India over and above that made available in its publications is to be found in the archives of the Institute. But we do not consider it. necessary for the purpose of utilising this material, any more than we have held it to be necessary on general grounds, that there should be a whole-time representative of India on the Permanent Committee of the Institute. There are so many important directions in which improvements can be effected in Indian statistics, and, given money and men who have been properly trained, the ways in which these improvements can be made are so well known, that the money which would be required to meet the cost of a permanent representative at Rome can be better spent in India. We desire, however, to make it clear that we are in favour of the attendance from time to timeof representatives from India for special questions.

We are also informed that the Institute is organising a world census of agricultural production to be made in 1931 and that it has been suggested that India should participate in it. We agree with this suggestion provided its acceptance involves no large expenditure. We cannot

accept the view that, even if the cost of collecting the information were considerable, participation would necessarily still be worth while. This, however, is a matter in which the Government of India will look for authoritative advice from the Director of the Central Bureau of Statistical Information whose appointment we recommend in paragraph 539.

CONCLUDING OBSER

VATIONS.

541. In advising concentration on a moderate programme of official improvement, as compared with comprehensive and costly enlargement of the statistical services such as that contemplated in the majority report of the Indian Economic Enquiry Committee, we do not wish to underrate the assistance that unofficial agencies can give. As interest in statistical methods, promoted by the activities of bodies such as the Board of Economic Enquiry in the Punjab, increases, there is every reason to hope that substantial landholders, commercial firms and groups of cultivators associated co-operatively for the furtherance of their interests in a particular crop, such as cotton or jute, will be able to render valuable assistance. We agree with the Indian Cotton Committee of 1919 that sufficient use has not hitherto been made of such non-official agencies. But nothing is to be gained by ignoring the fact that, in the present stage of development in this country, no extensive use can be made of a system of paid reporters of agricultural information such as obtains in the rural statistical organisation of Great Britain and the United States of America. It is for this reason that we have forborne to compare the Indian system of collecting agricultural statistics with the systems obtaining in other countries.

Until a reliable voluntary reporting system can be built up, we also consider that the establishment of a government agency separate from the revenue agency, for the collection of statistics, would impose an unjustifiable burden on the public purse. Even if far greater accuracy were thereby obtained, we have grave doubts whether its attainment would be worth the cost, but, in fact, we see no likelihood that substantially greater accuracy would be achieved in this way. Indeed, it is quite possible that the immediate result would be to make returns less, rather than more, accurate. For the agency employed would need to be literate and the number of literates required would make it necessary to employ many who were entirely unversed in agricultural matters. We think that solid progress is far more likely to be achieved by the adoption of the policy advocated by many of the witnesses who appeared before us, and which we ourselves recommend in paragraph 538 above, namely, to strengthen the staff which checks and interprets the data collected by the primary reporters.

SUMMARY OF CON

542. The conclusions and recommendations in CLUSIONS AND RECOM- this chapter may be summarised as follows:

MENDATIONS.

(1) The date of publication of the Agricultural Statistics of India should be expedited (paragraph 525).

(2) Separate figures for each district should again be given in the Agricultural Statistics of India (paragraph 525).