The eye, on which we rely so much, is usually not trusted to make divisions between columns etc. It has been given a chance in these tables, from which all lines have been abolished, both here and in provincial and State volumes. Extraneous words, dots, etc., were excised and a uniform color scheme adopted so that the census volumes of 1941 might present a harmonious and uniform appearance on any shelves on which they are gathered together. 

Had the volumes been their usual size there would have been no printing on the face at all. There is no reason why the face of a book should be turned into a title page. The omission would have served both appearance and economy. In order to secure absolute uniformity the position of the lettering on the spine was determined to a fraction of an inch for every line. The spine in the restricted tables is too narrow, and hence the appearance of the legend on the face, but here too it has not been allowed to become merely a title page.

The bindings adopted in 1931 for reasons of economy did not do justice to the importance of these publications, which it is to be remembered are in a way the silent ambassadors of India all over the world. They are not like departmental blue books or committee reports. Actually even a better standard of binding should be given and I had arranged for this but in view of the restricted tabulation and the need for economy agreed to what is described in press language as style VIII. The volumes should be in future and happier times in style VII. Elegance and efficiency can be combined and should be wherever possible and indeed their combination is one of the marks of successful execution. Incidentally it will often be found that they are both compatible with economy. 

M.W.M. YEATS

A -THE INDIAN CENSUS

I - THE RECORD 

The 1931 census coincided with a civil disobedience movement, which occasioned a good deal of localized trouble to certain superintendents particularly however in Bombay. 1940-41 saw also political influences on the census but in opposite direction; since whereas the difficulty in 1941 was to defeat an excess of zeal. 

It can be taken as certain that this single instance operated heavily to secure perhaps the fullest record yet achieved in an Indian census. The whole population was census conscious or at any rate the active part of it. To this extent the public interest was a definite gain and part of the heavy Bombay and Bengal increases is undoubtedly due to under-enumeration in 1931 being overtaken now. 

The interest however was not all beneficial and in some areas the communal preoccupation was no doubt inevitable in view of Indian conditions, but it is important that such preoccupation should not disturb the collection of information. It is necessary however to preserve a sense of   proportion, and fortunately for India the people are far sounder than a perusal of the press or of speeches would imply. 

A census or any other determination must be unaffected by preconceptions or bias if its results are to be acceptable and useful. If for example in an income enquiry there is a suspicion that the furnishers of the basic information have allowed bias to affect the actual returns the result, inevitable and salutary, is that the enquiry is regarded as worthless and its results are used only by biassed publicists and command no general authority or acceptance. Possibly it takes a certain quality of education and temperament to understand such a principle in matters in which personal interest is heavily involved; but it is one of the pre-conditions of a functioning   democracy. Emotion and passion have their place and it is the man who feels deeply who achieves the greatest results. But in political or any other arguments the use of doubtful or suspect figures is like entering into a fight with a cracked lathi; we can deliver no through blow with it. A properly educated mind can make the distinction between the collection of information and its use, but if that is applied as a test then I am afraid that certain elements in India have some way to go before they can be classed as educated. 

From the first I made it clear to all my officers and everyone concerned that our census object was the collection of facts and that while in this effort every citizen was our ally we should never allow a partisan association.

There were two aspects in which communal passion might affect census returns; they were of different importance. Much the more vital was the possibility of an actual influence on the tale of heads. Here we had in support the general reluctance of the decent man, who is no less numerous in India than elsewhere, to utter the barefaced lie that non-existent persons are present in his house. The punitive section of the Census Act entered also as an aid, for this particular falsehood was one admitting of no shade of interpretation and therefore once proved, punishment was inevitable.

The other aspect was the quality of certain individual answers notably as regards language or script. Here we were dealing with a different phenomenon, for the answer to the mother tongue question is broadly speaking entirely within the citizen's power of control. To prove a false answer in a court would be a matter of great difficulty. Moreover sentimental attachment to Urdu or Hindi as the Hindi/Urdu controversy entered, the census returns are worthless; and those passionate Hindus or Muslims who thought that by thus influencing the returns they could secure a valid statistical backing for their communal arguments have met the inevitable end of those who seek to corrupt the form of enquiry. The language and script questions have not been tabulated and I make now a recommendation to the Government of India that they be not tabulated even if the suspended operations are resumed. 

I suggest further that language and script questions be dropped from any future censuses until such time as the population of India is able to respond properly to a factual enquiry on them. The census can collect and deal only with facts not with preconceptions. 

Where Urdu / Hindi was not in question, the language returns are unaffected and can be accepted at once. Even so however I adhere to my recommendation about dropping the language question. The broad dimensions of this distribution of the population are well known and not likely appreciably to change, and in a limited questionnaire and with limited to new and more important aspects on which information is desired. 

To return to the first aspect, the tale of heads, I have already alluded to one important point. Whether from the effects of a long tradition of absolute or bureaucratic government or for other reasons India has tended to look on the census as purely a matter of the Government and its staff. There was a tendency in the more communal quarters to look on the census as purely a matter of the government and its staff. There was a tendency in the in the more communal quarters to look on the census enumerators as the ready tools of faction and to disregard altogether the vastly more critical function of the citizens. This tendency was unsound as well as unjust and received no countenance; I have never found that you develop the best out of men by distrusting them; my experience has been in exactly the other direction and I am to say in this most difficult of censuses the Indian enumerator as a whole responded splendidly to the call. The reduction in enumerators already referred to enabled us to dispense to a greater or less extent according to the region with the less interested, less, competent, or less amenable elements and in  the result India went forward to this critical enumeration with an improved and competent agency and the quality of the return depended on the citizens.

Over far the greater part of the country and in the entire rural areas the citizens responded and only in localized urban areas in the north were definite corruptions observed. It was notable that the great province of the U.P., in many ways a focal unit of India, produced an enumeration record free from any question, alike in the towns and in the countryside. The same of course applies to maintain an evener keel than other parts. 

I was determined not to put my name, or to allow any of my officers to put his name, to a suspect record and before the enumeration was over instructions for scrutiny and purification were in the hands of the officers concerned. The corruptions referred to were dealt with before the tables were prepared. 

I had foreseen the possibilities of acute trouble in certain areas and the changes introduced in the enumeration system while desirable in themselves and representing a notable advance in efficiency and economy, also laid their finger on the weakest spot of the old system in a highly charged atmosphere, namely the basing of everything on a so called one night enumeration which required the free alternation, under circumstances not in any practical sense admitting of check, of an earlier record prepared at leisure. By removing the one night theory (Which in itself had nothing whatever to recommend it in Indian circumstances quite apart from the particular problems caused by communal nervousness) the record with full opportunities for inspection and check. By relating it as far as possible to ordinary residence and dropping the concept that every visitor of a day must be allowed for, we removed again something which would have lent itself obviously to falsification if a corrupt will existed. Finally in the course of 1940. I prescribed certain extensions of the house list which gave a distribution of the persons in each house by sex and age. One object was to facilitate the indent for slips; another was to provide an approximate record in the event of was developments rendering the actual enumeration impossible ; a third  was against the eventuality of corrupted enumeration. This foresight was justified and in the areas where doubt or suspicion arose it was by a study of the house list that we were able to locate the suspected zones and carry out our purification. 

Enumeration was carried out directly on to the slips which were later sorted to produce the tables. This, in itself a major change, meant the removal of the former slips copying stage at which possibilities of error or alteration existed. 

Thus we approached the 1941 enumeration with a much more powerful system than had existed in the past and it was just as well. For I do not believe that any acceptable record could have come in the contentious areas from the one - night operation of the past. 

The issue went to show that even in these regions of turmoil and clam our and communal frenzy the citizens were better than their detractors had imagined. Only in one area was no acceptable record possible, but here too we were able to produce a perfectly sound figure of population. Our house lists had given the general dimensions. The vital statistics for this area of high quality and, a most important fact, are taken by community. Consequently the 1931 figures of community distribution along with the vital statistics and corroborated by the house list gave a sound figure for the population and this has been embodied in the tables. 

There then is the record. Despite, terrific difficulties a sound determination has been secured and the government of India and the country owe a great deal to everyone concerned down to the enumerator in his block. I suggest that never again should the census staff be taken for granted. You can take for granted a man you pay but you cannot treat in that manner a man whom you do not pay and on whom you lay extra, and as on this occasion difficult and contentious duties. 

India has at her disposal a most powerful informational system if she cares to use and develop it. Over wide areas it means that a reliable officer connected with the administration is in touch with the actual man in the village, and by nursing and developing this system capillaries and admirable circulation system from extremities to center could be developed. After this war some of the countries which have spent so much on their censuses may look with longing eyes on India’s advanced and enviable position. But that position must itself be safeguarded.

II - THE 1941 OPERATION

   The small map at the beginning will show at a glance the different degrees of tabulation achieved in various parts of India. Where full tabulation has been done it may be taken that an examination to sample examination indicated by the light blue colouration has in view only the areas in which full tabulation was not carried out. The uncoloured region represents those partial tabulation areas where no Province or State-wide examination of the random sample could be done owing to lack of time or other reasons. 

The main point which emerges at once is that the great population regions of the Indus and Gangas systems in which nearly half the total population of India lies have only a limited presentation in the census figures in the U.P., however, although the 1/50 sample has not been examined for the province as a whole, a statistical study of a part of it has been done and the elaborate economic survey covering 12 months rural life approximately over the calendar year 1941, should go a long way to filling up the gaps in that important province. Apart from the Gangas – Jamuna valley, however a good deal of India will find an effective sample representation from the full operations conducted by States; for the map shows that from the extreme south to the extreme north of India, though not from extreme west to extreme east, there is some appreciable element in which the full course has been carried out, and for this India is indebted to the states. 

Considering India regionally, the areas of full tabulation amount in some cases to considerably more than a sample. For example the beautiful and characteristic are known as Kerala will have a representation considerably more than 50% while Mysore State in itself represents a good half of the very different but equally attractive Kanarese region. The centrally situated Hyderabad state will represent 25% of the Telegu country though not its coastal tracts, while in addition contributing matter to the Kanarese and Maratha pictures. 

The random sample referred to represents another of the major innovations at this census. Over the whole of India every 50^th slip was marked and the original intention had the operations gone their full course was that these slips should be brought together and handled as a separate unit in order to test as fully as possible the validity of a sample in census conditions. It was my intention to apply these tests not only on the political units of India but, where social data were concerned, in which provincial and social frontiers do not coincide, to go on the latter and in fact to use the sample as a means of study of such characteristic social units as Kerala. The truncated operations have of course defeated this but the sample slips have been separately stored and should be available for future study or   use by approved persons or bodies. In some cases the form taken by the contracted operations, e.g., tensile sorting in Sind, the Punjab and Madras, prevented the sample being run continuously through the large aggregations which otherwise would have been possible; but these variations in method are in themselves of value since they will provide material for estimation the advantages of different bases for the random application. A similar variation value will be afforded by the sample taken was 1/20. 

The main purpose of this innovation was to test a method and although practically nothing of this could be done in British India I have hopes that a considerable degree of scrutiny will be applied in those States proceeding to full tabulation. Even in British India some degree of test was applied in the simple form of sorting the sample for communities and comparing the results with those established by the full soft for community as exposed in Imperial table XIII. The agreement in every case was very close for the major divisions and for example in the Punjab the sample revealed a community distribution differing infinitesimally for the mail elements and by considerably less than 1% even for the smaller constituents. So far doubt that community distribution for province could be determined beyond the limits of any necessary accuracy by the sorting of a random sample on this basis. The importance of this for future censuses and their cost is obvious. 

It is not possible to give any reasoned account and criticism of the sample since we have been denied the opportunity of completing the tests we had in mind. I can only hope however that this beginning will be taken further and that before the next census whenever it is, the possibilities of using sample methods extensively will have been seriously considered. 

The 1941 census operations differed widely in their circumstance, methods, and outcome from those of the previous decades and taken all over must represent the most difficult operation of that long and honorable series. One of the last things to be desired in a census is uncertainty; yet that pursued us to the end. It was not till February 1940 that the Government of India decided whether to have a census at all. A still greater difficulty was caused by the delay in deciding how far to go with tabulation. This decision was not reached until after the enumeration was over. Ordinarily preparations for sorting are made months earlier, buildings are booked, staff earmarked, pigeonholes, furniture, etc., arranged long before the enumeration date, the object being to guide the enumeration record straight into the designated sorting office, where it will find a responsible officer and his staff awaiting it. A decision reached only after the enumeration is over meant that none of this preparedness could exist and every Superintendent felt the difficulties this brought. Bombay was perhaps the worst sufferer, for it was impossible to retain building reserved for sorting offices; but Bengal similarly lost an advantageous and suitable building and the tale was repeated over the whole country. The difficulties did not end there; for the staff question in a truncated tabulation was acute, particularly when no reasonable notice could be given. It was not a question of merely tabulation being uncertain but of its degree being unsettled; no one knew whether there would be any tabulation at all or if so how much; and this meant that even tentative preparations could not be made. 

In approaching the question of tabulation, the point   was how to get the most for the least, or in other words, given a certain sum how to use it to the best advantage. The minimum was fairly clear. All Indian census start by a first hand – sorting for sex and community. This indicated at once the minimum effort worth doing at all and the cheapest practicable course. There was no use in doing anything that did not operate on the whole body of slips by this first sort into all India recognizable units and thus elicit two of the chief elements in the census: (1). The distribution by sex and minor unit, and (2). By community. 

I had reason for wishing the whole body of slips to be handled. The main one was to enable the random sample to be extracted, so that, whatever happened ultimately to the main body of slips, its 1/50 sample would be in separate existence and it is disposal of any approved authorities who wished to make use of it. Actually I should have liked to sort for the full age and civil condition table which in Indian conditions of defective or non existent vital statistics and a rapidly growing population is probably the most important of the whole set. And if ever sorting is resumed this should be the one above all others to be done. 

It is often more difficult to do a thing partly than to do it in full, and this applies notably to a census. In order to set free provincial officers I took over the task of seeing their tables through the press. This meant that provincial offices were broken up as soon as the tables were ready in manuscript and no officer or staff remained to deal with queries. Queries, however, invariably arise on census tabled; for since everything must be congruent, even the slightest difference has to be tracked down and either removed or explained. But with the provincial offices no longer in being these conundrums raised disproportionate difficulties, since the local staffs who could have solved them more or less straight off, were no longer there. Some had to remain unsolved and thus for example the details of the minor elements brought under the term “others” in certain provinces must remain undisclosed.

The administrative methods are dealt with in details elsewhere and this report need not dwell on them. It must, however, indicate briefly the general course of the operations followed. The first point for comment is that this census saw more changes in methods than had previously taken place in the whole 70 years since the census began. The chief was the abolition of the old one night theory of enumeration and the next was the abolition of the old schedule and the conducting of enumeration straight on to the slip which was later sorted to produce the various tables. Connected with the last was the complete centralization of printing, the removal of any written language from the enumeration slip and a variety of other connected and consequent changes which produced not only efficiency but substantial economies amounting to over a lakh of rupees despite a war time rise of 30% in paper cost. The first main change enabled us to relate the enumeration far more closely to the existing systems and agencies of the country and brought down the number of enumerators from two million to one, and for British India, from 1½ million to 2/3. The reduction was greatest in Madras and Sind and least in Rajputana and the C.P., and it is significant that efficiency of enumeration was in proportion to the extent of the reduction. The old one night theory was never more than a theory and like most outworn theories it had reached the point of being a danger. It involved putting the whole record collected during preceding weeks, checked and tested, at the mercy of a single night round and whatever the case in previous censuses, that single night round would have produced impossible consequences in a year of tension like 1941. 

The rationale of the census could be summed up therefore as an endeavour to express the whole operation as far as possible in terms of existing divisions, charges and responsibilities and to use the officers operating these as elements in our census generally. In effect an unpaid census has to be based on some such theory and what we did in 1941 was to take this considerably further, and I hope clear the way for a complete expression in 1951.

The form taken by the operations this time followed a double design: 

1.  

   To meet the undoubted stresses and dangers of an enumeration which I knew would arouse contention;

2.  

   To guide the Indian census into more rational channels.

All the changes worked together towards both ends and were designed to that effect. Thus the normal residence basis was impossible under the old one night theory and its implicit expectation of complete simultaneity. This was linked up with developments in the house list, which made that far more of a preliminary census. In turn the whole series of changes produced a position which made the house list yield a population return very close to the actual 1941 figure and more over one of such merits that the enumeration figure could be set against the house list one and any marked deviation justifiably regarded as requiring explanation. In the less urban areas as might have been expected the closeness of the two records was pronounced. In Orissa the floating population floating population was proved to represent something negligible. In the focal, more urbanized and in many ways difficult province the U.P., the difference between house list and final enumeration only in one case exceeded two percent. And in many   fell below even that low figure. In Assam the population records yielded by the house list and the enumeration were almost invariably in close accord, the difference being less than 1½ percent. In the few cases where this figure was exceeded, adequate reasons existed, e.g., the regular accession of outside to sylhet in January and February for fishing and the large floating population in Goal Para. In some of the rustic states the difference was as one or two individuals. 

The house list was taken at a period convenient to the administrator and the general circumstance of the province or state. As its name implies it is based the house. Had enumeration by any ill chance been prevented I could have furnished the Government with a reliable dimensional figure based on this list and I now suggest that the future census taking policy of India base itself on an operation of the house list type and season. 

In pursuance of this idea of fullest articulation with the general  system and needs of the country, we were on the look out for opportunities to use our census momentum to help on desirable allied enquiries. One measure pressed on superintendents was wherever possible to encourage and assist provincial governments or other authorities in economic or other surveys to be carried out along with the census and in association with it. We should offer the planning and the direction of the superintendents while the provincial governments, etc., were invited to assist by contributing staff and in some degree funds. There was no prescription of what should be enquired into; the approach was much more pragmatic. We looked round to see if there was something on which further information, or more information than the census could give was desired; then we tried to work out convenient, practicable and economical methods of doing it. But for the war and the truncated census we should have seen, I think, a considerable development of this and even as it is the achievements are considerable. Quite apart from provincial payment for statistics in which they were interested, e.g., when the governments of Bengal, Bombay and C.P. expended a certain amount on caste tabulation, the latter Government also investigated the returns of handloom activity in the census slips. These, however, were on the census record itself. The outside departures were in a way more significant or certainly had a different significance. 

Of these the chief was the elaborate economic survey carried out in the United Provinces. The plan was worked out by Mr. Sahay and approved by myself and in the main financed by the Provincial Government which put up Rs. 8,000 out of the total amount spent. We contributed Rs.2,500 from census funds and in addition of course the planning of the whole enquiry. One important feature of this enquiry is that it was so framed as to continue after the U.P. census office had been wound up and Mr. Sahay translated to other duties in Delhi. This is a feature to which I attached importance from the first, foreseeing the likelihood of a truncated tabulation or no tabulation at all. Full details will be available from the province, but the essentials can be given here. The scheme covered a year’s activity and enquiry. It was based on a random sample of 300 villages of the plains area of the province. The hill and foothill areas are so different that they were omitted. The scheme gave effect to one standing principle, namely of making the utmost use of existing agencies instead of thinking only in terms of expensive   ad hoc provision. It was carried through, therefore, with men selected from patwaris, rural development organizers, agriculture and cane development employees and schoolmasters. These total about 35,000 in the province, an ample selection base for 300 men to carry out this investigation. Supervision was provided by inspector’s chosen from the co – operative, rural development and agricultural development supervisory staff. These men were chosen first and given specialized intensive training. In turn they trained the selected local investigators. The inspectors themselves were under the supervision of the Assistant Registrars of the Co-operative Department while the Registrar, Director of Agriculture, the rural development officer and of course Mr. Sahay himself were there to act as higher checking authorities during their tours. 

Enquiries were made in relation to each family resident in the random selected villages. A family represented the same definition as produced the census house, thus introducing at once a link with census procedure. There was no attempt at asking figures for the year since the view was that the ordinary villager could  not be expected to give details regarding expenditure another details for a whole year at once. A monthly visit would have been preferable but convenience compelled a quarterly round. 

The information to be produced was net income of rural population, disposal of agricultural produce in rural areas, distribution of land in rural areas for different purposes distribution of cultivated land in different crops, rates of wages in agricultural occupations in rural areas, size and composition of families resident in rural areas, age distribution of children in families in rural areas and size of agricultural holdings. The collection of material was completed by the close of 1941 and its preparation has been put in skilled statistical hands. 

This enquiry should give a comprehensive picture of rural life in the plains area of this important province. Two questions in particular were put owing to developments this time in the main census. Questions 7 and 8 of he census list bear on fertility and represent of the main innovations this year and I think the first time such question have been asked in any census on a country wide scale.  The two tables will provide for the U.P. the age and sex distribution of children in the family, the number of marriages per male, the survival rate of children, the age at marriage and at first child for women, and children’s age at death. Along with this will be the proportion of stillborn children, the sex of the first born and subsequent sex composition, the number of children born per couple and the birth spacing. 

From these we should be able to arrive at an estimate of the specific fertility for this important and representative central block of India. And some approach to genuine fertility rates is a definite desideratum in the conditions of a country like India, dependent mainly on ten yearly determinations and at present in the midst of a heavy increases spell. Information will be collected of all children born to the wife, whether still alive or not; entries will be strictly in order of birth, the age of both parents will be given and the sex of every child. 

A similar though more, restricted enquiry was carried out in Ajmer, and minor studies in Peshawar and one or two other cities. The states also took a prominent part in this effort though I have at the time of writing this brief note no details about their actual performances. 

Tabulation for British India was limited to the first main sort. This produced the population distribution by district, tehsil and town and also the main community record. It yielded also material for these useful documents the village statistics. Printing of these is a provincial liability and in the past they had in may areas never got beyond a single manuscript record. These documents however are of great use in administration and I did my utmost to extend this usefulness and to induce more provincial governments and states to print and distribute them. I am glad to acknowledge a ready response and in many areas the village statistics will appear in print for each village and will approximate in varying degrees to a genuine village directory. Madras of course maintains its old and honorable record as a pioneer  in this field but this time will be joined by other regions, e.g.,  rajputana. 

A list illustrating this point of the production of village statistics in the various provinces and as between 1931 and 1941 is shown at the end of this section. Apart from the great general usefulness of these statistics, they have a special census function, namely, to make easier the preparations for the next census. In particular, with a code number and the pad system in mind, it is important to have a useful and reasonably stable tabulation elements on which to base estimates. Printed village statistics give this and the material whereby to make the desiderated start of block village from the outset. 

The extra material recorded this time in the house lists was grouped in two classes according to whether it exposed the circumstances of the individual villages, e.g., distance from water supply, market school, etc., or whether its interests was more general and descriptive, not requiring individual village detail, e.g., nature of house construction. Information in the first class will appear against the particular village. It had been my intention to deal with the second category by means of a random sample based on British India has defeated the second objective but the states proceeding to full tabulation will I hope fill this gap to some extent. 

The sanctioned tabulation for British   India does not cover caste but even had the full course been                      taken, there would have been no all India caste table. Even in 1931 it was severely limited for financial reasons; the time is past for this enormous and costly table as part of the central undertaking and I share Dr. Hutton’s views expressed ten years ago. With so constricted a financial position and with so many fields awaiting an entry there is no justification for spending lakhs on this detail. 

Here and there however provincial or state governments wished a caste record for administrative reasons. This record did not usually go the length of minor detail but generally contemplated only broader sub divisions. It was made plain from the first that while we were opposed to caste tabulation as a central charge it was quite possible, at the time of community sorting, to take out caste detail provided this were paid for. In the result three provinces, Bengal, Bombay, and C.P. sought and paid for a certain degree for caste sorting. The Bombay requirement was for caste detail by villages and this was provided by Mr. Dracup for the remarkably small sum of Rs. 24,000 which represents, however, a feat of economy unlikely to be repeated. One or two others, and here and there a city, asked for broad literacy figures of certain other information and this too was provided on payment. 

This represents an important departure and on of the most interesting developments of the 1941 census. The central fisc should carry tabulation only of the material required or thought advisable for central purposes; where extra detail is sought for local reasons the census would be ready to assist to the utmost extent in producing it provided payment was made. 

An extension was given to this principle where private interests were concerned. Thus persons interested in the Maithili form of Bihari or in the numbers of Maithila Brahmins were told that they could have these figures extracted provided they deposited in advance the estimated cost. This they did. 

The elements which the center must indefeasibly carry are age, means of livelihood, community, civil condition, literacy, birthplace and mother tongue, where it is decided to produce tables on these matters. Anything else, and any developments or extensions of these should be carried out by the particular interests desiring them. By combining the two in this way conveni9ence and economy can be secured. Elsewhere, and with possibly wearisome iteration, I have stressed the fact that rationalization of the census is overdue. So long as it was regarded as an omnibus in which all seats were free, there was no real selection of passengers. Once a clear decision is taken on, as it were legitimate passengers. Once a clear decision is taken on, as it were, legitimate passengers and extras, the latter will be confined to those who really want to travel with the result of better use of the accommodation available. 

This census has seen the beginning of mechanical tabulation in India. Here again however the departure has taken an individual form. The phoenix system removed every possibility of the exhaustive system removed every possibility of the exhaustive preliminary enquiries necessary before a departure on an all India or even provincial scale could be contemplated and indeed but for preliminary enquiries and discussions by myself with the representatives of the companies in 1938-39, it would not have been possible even to get the Delhi experiment started. For in a mechanical tabulation everything must be thought out and prepared before hand, since the  punchers and the machines can only do and be expected to do straightforward mechanical operations. There is no room for improvisation. The essence of my idea was to see whether at least for urban areas the census tables could not be taken out in the spare time of machines already in use by government departments. The Delhi experiment, conducted against every kind of difficulty, has shown that this quite possible. Ten years hence there will be many more opportunities for putting at least the big cities record into cards. 

The method meant of course the careful adaptation to census needs of the particular form of card etc., in use. But for the war we would have used also the machines of the Military Accountant General. The war however had so swamped this office with work that it had no spare time. The general attitude towards an innovation was of suspicion as it always is in India, and I am all the more indebted to Mr. Sheehy for his ready response to my request. By careful working out we succeeded in getting 3 records into each C. B. R. card, thus reducing the cost of cards by 2/3. A host of problems presented themselves but one or two general conclusions may be of interest. In many census tables there is as in some cricket elevens, a pronounced tail. This applies particularly in birthplace and mother tongue, where different small items may run far into the 10s, even 100s and yet represent only 1 or 2 percent, of the total returns. To save columns and punching a good deal of this can be done by the hand sorting at the time of coding. Coding must be done under the census officer's own direction and instructions. It is a cardinal principle in the use of machines that the person desiring  the return must be absolutely clear what he wants and take all the decisions regarding it. The machine they can say what their machines companies can help in techniques of operations, they can say what their machines can or cannot do, but it is no part of their role to determine objectives. All payments must be on outturn. Actually the C.B.R. machine men are not usually so paid but I insisted on an outturn basis for the census work. 

To use a word that has become rather popular in supply circles, the sorting machines represent the bottleneck and it is rare for a comparatively small unit to be balanced in this respect. Hence of course the desirability of erecting a centralized unit, which would handle all government mechanical tabulation, work, at any rate in one center. This would mean and far more economical use of he spare time of machines. Sorting should be started at the earliest possible moment, as the sorting machines play an important role in verification. And in general for a fundamental rule we go back to what I have said already time and again time: discussion, experiment and re-experiment are here as in other scientific zones the essentials for satisfactory performance and results. 

It was difficult to make an estimate for an under taking of this kind, for which no previous experiences existed as a guide since it was the first operation of its kind in India. The number of cards theoretically required could be got at easily and therefore their cost, although the high wastage by the punchers raised the number beyond expectation. The other elements however were unknown and since all staff payments were to be on outturn, this meant a previous fixing of these rates. All over it was a difficult piece of estimating and the figure arrived at was Rs. 7,000 based on the premises that C.B.R. machines alone could cope with the work. Actually, they did not, and in order to finish of  the Delhi tables along with the others we had to take on Hollerith machines at the end on hire. Had to take on and other circumstances permitted this would not have been necessary. Omitting this however the estimate was only Rs. 228 out. 

Provinces and states where the village statistics are printed.  

  III - THE SYSTEM AND THE FUTURE

            India is apt to take its census for granted, following in this the well-known tendency of mankind to ignore what is near at hand or familiar and concert rate upon the novel or the remote. The peculiar system under which it is administered accentuates this; for the absence of any between census continuity must encourage the tendency towards “ out of sight, out of mind”. Beyond  and outside India the attitude is very different and in some parts the combination of the undertaking and its astonishing cheapness induces the description of it as a kind of administrative miracle. 

            There generally  comes some stage however after which taking things for granted is apt to lead to difficulties. This stage has arisen in the case of the Indian census and certain elements in the present position render it advisable that the citizens as well as the government should devote early and concentrated  reflection to what is involved and to the direction they wish this great undertaking to follow.           

            In other countries, certainly in those with a census history comparable in extent with India’s, the census is a central function carried out like other government activities through paid staffs. In India, which incidentally is from this point of  view rather a continent than a country, the enumeration, the basic part, is carried out by a multitude of citizens in addition to their other duties and for the greatest part unpaid. Thus accidentally for there is no trace of design India has reached the very advanced position of haven't this enormous operation carried out by the people as a civic duty. And that is the first element that is taken for granted ; but it has reached the stage when it must be given some greater reflection if a most powerful as well as a most advanced position is to be maintained. 

            This first item is peculiar to India. The other is of universal application. That is the cardinal fact that it takes two to make a census, the enumerator and the citizen, and that of these two the role of the latter is the more fundamental and vital. The enumerator broadly is a scribe :  in any census it is the citizen's answers which are sought and are tabulated. Yet in India one could hear of read pronouncements showing an impression that the census staff generally, and that therefore the validity of the record depended solely on them. Actually, to repeat a phrase I used in a broadcast, no census anywhere can rise above the level of the citizens. If there were people on Venus and they had censuses, the same would apply. It applies in every attempt to count the people and record particulars about them, and it is not only a folly but a danger and an injustice if this is not realized. Hence the importance in the Indian census of a full understanding by the citizens of their role. The system, if that word can be used here, is in brief that every 10 years some officer is appointed to conduct a census  and  officers to work under him are appointed in each province. The states take corresponding action. These appointments are made at the minimum of time beforehand and within one year questionnaires have to be settled, the whole country divided into enumeration units, a hierarchy of enumeration officers created and trained, millions of schedules or slips printed and astrioutea over the face of the country,  the whole process of  enumeration carried out and checked, tabulation then carried out in offices located in any old place that can be found, on make shift pigeon holes and furniture and with temporary staffs, rushed through the presses – and then, in the third year the whole system is wound up, the officers and the office staffs are dispersed and India makes haste to discard and forget as soon as possible all the experience so painfully brought together. 

            Undoubtedly in  a census there must be aggregations, of staff round about enumeration and tabulation time but this is quite distinct from the question of systematic preparation well in advance. Momentum in the administrative as in the physical world is of the utmost importance and should never lightly be discarded. Continuity is merely another way of expressing momentum. The continuity desiderated is not of officers or staffs but of thought, experiment and preparation. The ad hoc staff of the actual enumeration and tabulation time should come merely as almost automatic expressions of operations evolved long before. One has heard the phrase “the science of administration” and certainly so far as the census is concerned there is much scope for applying the principles of science i.e. actual observation preceded by full and frequently prolonged study, investigation, discussion, and experiment. There is no reason why this should not be done in the all India census and the result would be striking as regards the work that could be  covered and the economies that could be secured. 

            To begin with, the census should be linked up with other scientific activities of the government of India. Its determinations are in many ways the base of departmental and other action. They are in essence an act of observation on a continental scale. Consequently, enquiry should be linked up with other statistical investigations of the government of India and the collecting of information should come under the general direction of those who control such statistical activities. The aegis of the government of India is essential, for an unpaid  enumeration needs  the influence and association of every authority it can command. Undoubtedly mere statistical direction would not be enough in itself, for the Indian census is in essence an enormous and complicated administrative performance, and experience and judgment must have their place in the scheme. But the economics and statistical and other bearing of the collected data are so important and oblivious  that the census as a whole should be formally linked up organically with the main statistical system of the country. How this should be done is primarily an administrative matter. Various association possibilities present themselves; but the main point is the end, not the means. If the end is recognized and pursued the means will to a large extent suggest themselves. The aim is not to keep a particular officer or office in existence, but to keep this integration of the census with the main administrative and informational system of the country a live issue. With this achieved, then what I call the between census operations would be secured and these in present circumstances are at least as important as the actual enumerations themselves. In fact so far as 1951 is concerned, much more so. 

            So much for continuity at the top. There is great scope for continuity also at the bottom and from the first one of my injunctions to the provincial and state superintendents was to seek every means whereby they could secure this continuity in the most convenient way. I do not advocate and indeed discourage any proposals for permanent census departments as such; what I do suggest as meriting constant and close consideration is a study of the administrative essentials for a census, a comparison of these with the continuing system of the province or state, and a study of how most easily the two could be linked up in a regular association, so that the province or state system in its ordinary operation would throw up automatically the administrative necessities for a census. The field for this is wide and varies in opportunity with the nature of the individual administration, but two main objectives stand out: one for the census itself primarily to produce, the other for the general administration of  the province or state. Taken together, the two could provide the basis for any census system these two are 1. village statistics assembled and printed by tehsils and districts and 2. the vital statistics. The first should be produced by the day administration of the country. 

            Until such time as the census is put on a rational basis with ample time allowed for discussion, investigation, etc., the scope for substantial alteration of the questionnaire is not great. Even so however changes of some magnitude were made this time. The whole point is discussed at length elsewhere, but one innovation at least should receive comment in this general account. That is the question on reproduction in regard to every married woman in the while of India two questions were put 1. The number of children born, and 2. Her age at the birth of the first child. Actually the questions were three, for the first was extended also to produce information on the number surviving. Had the operations gone their full course these questions would have been made the starting point for elaborate tests and researches and would moreover have been investigated on the natural region, not on the more or less unnatural province or state. In such matters as fertility, social community matters far more than political association, and for example Kerala for social questions should be treated as one although its component parts are two Indian states, one British district and part of another. The same pronounced individuality does not distinguish Kannada but here too association could be usefully applied and would have been under the originally contemplated scheme. 

            Undoubtedly there was a wide margin of likely variations in the answers given; individual age could not be considered an element susceptible of final determination in Indian conditions and this would apply with even greater force to such matters as are here discussed. On the other hand the mass involved is such as to compensate in an appreciable degree for the limitations of the individual returns. While my main purpose was frankly, in an ancient phrase from the history of my own country which I used in explaining the point to provincial and state  superintendents, to bell the cat: this and other information should appear in every birth or death certificate mutates mutandis. If it were there and the vital statistics were brought up to reasonable completeness, any country has its main population record on tap. I was under no illusion as to the zone of error in the replies; my chief concern was to break the ice and force on Governments and people a realization of the kind of information they ought to have and for which they ought to provide in some more rational an deficient manner than through an All India census determination. 

            The questions were asked for the whole country with practically no difficulty and the ice has been effectively broken. It has been shown that these questions can be asked and will be answered. It is now possible for governments to go ahead and get the information through the proper channels, i.e., the vital statistics, for it must be understood that the inclusion of these questions in the census list  this time does not mean that I regard that as the most suitable place for them. I do not ; and in this as in other respects the census is a primitive and limited method. Whether governments will go ahead on this line I do not know, but at any rate the concept of this as a piece of information which governments require has been effectively put across to the people of India. 

            One of our main objectives at this census was one might say to turn the phoenix into an accepted and familiar feature instead a periodical and disturbing portent. Of the changes introduced into the actual conduct of the 1941 operations all had a positive justification on their own account but all too were designed to serve this great objective. A census commissioner more than most men is required to perform astonishing feats of quick building but perhaps for this very reason he should look more than most men to the future. I have tried to do this by directing the attention of province and state census officers towards the desirability of continuity, inviting suggestions to this end in accordance with the administrative conditions of their province or state, and encouraging them to get these proposals adopted. In this field the best return has come from the states. In some ways of course there was much more scope but it is due to the states to recognize that in general they showed a greater relationship of the desirability of integration than did the British provinces. This applied even in small states and very notably in some of the larger ones. Thus in Rajputana I encouraged Capt. Webb to sketch proposals for preserving some degree of between census continuity and to get these put before the states. As a result nine out of 24 states have declared themselves ready to give effect to the procedure summed up below. 

            The main features are to keep the house list permanent by correcting in annually, to make house numbering permanent and to establish continuity of contact with census problems. The actual detail of application will vary to some extent with the state and its resources and the quality of its administration; but the principle have been accepted. The first point is obvious enough and when in Kishangarh state, one of those which is going to give effect to the scheme, I pointed out to the state census officer on a house wall not only the 1931 number but that of 1921 also, all of them different, and asked why they could not at least have been the same, I think the practical illustration had much effect. The third item is not a matter of permanent census establishment but, as I have said so often, of continuing thought and the integration of the census with the general administrative and statistical system of the state or province. These states instead of dropping the census immediately the tables are out, will keep on their officers, as ex – officio superintendents of census in addition to their other duties during the between census years with the responsibility of following up all census points. Among these would come numbering. They would have the administrative authority to pursue these and any other matters in which experiment, investigation or administrative action had shown itself necessary or desirable as a result of our 1940-41 experiences. 

Scientific advance is the result of imagination applied to knowledge and tested by experiment. This is the case in the census as in other fields. We have to conceive the idea, apply it to our knowledge of the conditions, and then test it in practice. The first two stages have been done and the third will I hope be applied in the states I have mentioned. From the results the whole of India should be able to learn much and I trust that such between census supervising body as exists in India will arrange to secure a regular scrutiny of experience in these sates so as to make the best use of it for the states themselves and for the country as a whole. Every credit is due to these states, some of them quite small, for taking up this development in methods. Their action deserves acknowledgement and gratitude and I strongly recommend the government of India to take a sympathetic interest in these experiments. It would help greatly for example if the Resident in his visits were to enquire how the continuity proposals were faring and to assist them with his advice. 

Another line in which we have tried to secure an improvement is in vital statistics. Here again Rajputana was the scene of one major endeavourer. I sketched out a possible system of birth and death registration and Capt. Webb’s enthusiasm got this  put before the various state authorities. In essence this suggested a full list of desirable questions in any record particularly as regards birth; but the main features was the attempt to relate vital statistics more to the continuing conditions of the countryside and the particular region, to make it easy for a parent or relative to report and to arouse in him the desire to make such a report. The only solution to the vital statistics problem and that which was applied in the U.S.A. in the last decade to improve the many defective regions there is to make the citizen and especially the parent “vital statistics conscious;. I apologies for this unsightly polysyllable but it does express the idea: once the Indian parent really feels that a birth certificate is something his child ought to have, he will give the authorities not peace till he gets it. At present the general attitude towards these certificates is from the reverse direction, namely that they are something which for obscure reasons the authorities demand and which they make arrangements to produce by methods of their own adoption. The parent in this is an entirely passive agent and indeed in many cases  does not enter at all, for the occurrence of the birth is reported by a chowkidar or other village officer who has heard of it. In any proper system the parent should enter as the all important person and the corollary to this in a country like India. Is that his entry should be facilitated to the utmost. 

This means the taking of thought and a definite attention directed toward securing the desired end. It will probably mean also more than merely benevolent interest at the center and this was what both the united states of America and Canada found when they took up the improvement of their vitals statistics; it was only when the center was prepared to contribute and actively assist in practical measures, that these two great federations were able to get the units moving in the desired direction. Local systems and conditions should be studied in order to see how most conveniently to adapt them to the object in view. Wherever there is a Panchayat the possibility of making it the registering authority should be investigated. Honorary registrars could be appointed from retired officers or other suitable persons to whom reports could be made. They could be given counterfoil books on which to make the entries and one foil could be left with the parent as a form of birth or death certificate after any verification considered necessary. There might be a system of post card reports, postage being franked, and the central contribution  could for example enter in this way. But essentially the point is not this or that method as the best but to have the central idea accepted, namely, that the development of country wide good vital statistics is an object which should receive e continuing thought, direction and practical assistance from the center. 

No administration needs or could for that matter make use of the last digits in a country's population and no census  determination however perfect at the moment could ever hope to give these. For within a few minutes in a country the size of India, the last digit has lost all meaning. In fact we are operating in the region of dimensional numbers and our policy and methods should take account of that important but little realized fact. The sooner governments, municipalities and other bodies realize that dimensions are all that is required and all that can be given, the sooner we shall be able to cut loose from this attachments  to digits without significance. The one night round represented perhaps the most glaring instance of sacrifice to theory but by no means the only one. The halving of the number of enumerators over all India, and for British India the considerably greater reduction, represented entirely the  less efficient and amenable elements of the past and went some considerable way towards achieving  that important desideratum in an unpaid census, of making the operations as easy as possible for those who have to carry it out. 

The one night theory has gone, but I would take this change even further. There is no likelihood of Government of India ever paying the census enumerators; not unless something approaching a miracle takes place; and that being  so they are bound to take further this question of suiting the convenience of provinces if they wish the system  to continue. And actually on the merits there is no reason why the census of madras should not be conducted at a different period of the year from that of Northern India, if, as is undoubtedly the case, convenience dictated so. It is only a matter of arrangements worked out in good time to bring this about and to link up a provincial system with the time of the year most convenient for its personnel. Once we are away fro the one night theory we can go on the basis of  ordinary residence in which the floating population which forms so menacing an element and problem on a one night  basis is reduced to easily manageable or, as indent population we want, not the artificialities of a single night. These artificialities have been less in India than in the countries of  the west, where they were enough to defeat any question of a simultaneous system in the United States of America and had created considerable discussion and difficulties in the United Kingdom; but even so they were markedly on the increase. 

The problem of India's census  is one  of dimensions taken along with a fixed low financial roof. In any scientific problem methods are of the highest importance and where the mass is large this importance is enhanced. Actually the ten yearly convulsion represented by a census is essentially a primitive method and with the development of a better and fuller informational system over  the country and wit a fuller application of modern methods it should be possible to reduce considerably the extent or violence of this convulsion. I have dealt at length with this and various other points elsewhere but might repeat here a remark made in a broadcast and in a speech to the Indian statistical Association, namely that the perfect statistic is a by product, something that comes out inevitably, naturally and more or less unobserved as the side result of some recurring phenomenon in the life of the country. The more naturally your information comes out the less it is liable to be affected by predilections or preconceptions. Wherever possible the specific observations should be first hand and the actual quality of the observation should be itself estimated. 

All this needs thought, experiment and discussion and none of these is possible under the present phoenix like conditions of the Indian census. Continuity in administration is of the highest importance and should be observed even in the case of the census ; and indeed one might say particularly in its case because of the longer wave length. This does not mean a permanent census commissioner but some real provision for between census consideration of the result and experiences of the last census and preparation in good time for its successor. The economy argument is the one used in support of the phoenix system but I myself a convinced that if between census preparation was observed and proposals made by the census commissioner and superintendents, instead of being trust into cold storage or not even that, were considered in good time it would be possible to defer the appointment of provincial superintendents for 3 to 6 months, securing thereby an economy which in itself, apart from other consequences, would be of the order of a lakh of rupees. The phoenix system is in fact a financial mistake as well as an intellectual crime. 

Despite the extreme difficulties of this census we were able to introduce more than one change and to carry out experiments in methods, which will be of the greatest value for the future. The principal changes were the non simultaneous enumeration with the results referred to a central data and time (sunrise on 1^st march, 1941) and the cutting out of the entire slip copying stage of the past. 

Another experiment was carried out by the Tonk State at my suggestion and the costing of this has been carefully observed. This experiment used the type of card, which contains holes in different places and is sorted by means of a long needle, which picks out only the cards equipped with a particular hole. This system is already in use in certain offices in various parts of  country but this is the first occasion of its use in a census. The results of this experiment should be studied with the greatest care for it may hold great possibilities. It cannot be said that tonk state was in any way more favored in the quality of its enumerators than other parts of India;  indeed less so. Consequently so far as personnel is concerned, what  could succeed there should be practicable any where. The cost aspect, related to all India dimensions instead of those of a small state of tensile size, will be the matter requiring most careful investigation; but that investigation should certainly be done. 

The one night theory of the past was itself enough to rule out as practicable measures such methods as those and this illustrates how closely the various elements of a census are bound up together. Enumeration methods to a large extent influence possibilities in tabulation. The Indian census is unique in more ways than one. One way is the fact that although technically a purely central undertaking it is inseparable from the fullest use of province and state organization and staffs of every kind, and a use which does not contemplate technicalities of debit. If it did the position would be fundamentally altered. Actually our census is an operation deserving in the highest degree that much abused attribute, All India, for in essence the provinces, states and the whole country put their shoulders to the wheel and carry it through, not without grumbling, not without grumbling, not without friction, but on the whole with an acceptance which is enormously to the credit of the country. I do not think this aspect has ever received the recognition which is its due at least in India. 

Apart from this aspect the system is valuable and powerful, and India, if she likes to take thought and integrate her informational system has in some ways an unequalled instrument at hand. Experience of 1931 showed the obvious dangers of the theoretical consequences of the constitutional separation being pressed excessively and as a result of representations and, I am afraid, a certain amount of persistence on my own part, the government of India approached the provincial governments in the autumn of 1938 with all the cards on the table; whatever the technical attribution of the census it was in essence an all India effort and could be carried on only if the provinces would agree to take their share as in the past. The provinces response was prompt and satisfactory and thus the 1941 census started out at least with one possible source of expense, difficulty, and misunderstanding removed. 

Even as it was, the 1941 census represented a vast amount of pressure and difficulty and although I think that the unique phenomenon of a population itself that the unique phenomenon of a population itself carrying its census can be repeated, it will not be unless there is timely thought and preparation. Something must be done to ease the pressure on those who take the census for us unpaid, and the changes introduced this time must be taken to their logical conclusion. I took as far as was possible with due regard to the novelty of the measures themselves, the not unnatural diffidence of superintendents nearly all of whom were new to census work in any form. And the way is clear I think for the development of a rational system which with minimal or no additional cost-or even possibly a saving-will base the census firmly on the proper agencies operation at the most convenient times. 

The cardinal point is and will always remain the fact that our enumerating agency is unpaid and that payment in any serious form is impossible unless the Government of India are prepared to contemplate a global figure for the census in crores instead of lakhs. It is impossible to stress this too much. People and not only the civilian public talk as if the Indian census is something like that in the U.K. or America where enumerators are paid at definite rates, e.g., so much per head. It is not; and the whole operation must be approached from the point of view that it is essential to keep the demands on our unpaid agencies to the minimum, and by our methods and choice of periods, to make those demands as easily borne as possible. The essentials are - 

  1-   A period enumeration related to the resident population with enumeration travelers, etc., removed and simultaneous record reduced to a minimum.

2-   The period need not be at the same time all over India. What suits Madras does not suit North India.